The income of nations;: Theory, measurement, and analysis: past and present; a study in applied economics and statistics

M:M.

RUTGERS THE STATE UNIVERSITY o f N E W JERSEY

SOUTH JERSEY LIBRARY

COLLEGE OF SOUTH JERSEY CAM D EN 2, N. I

T H E INCOME OF NATIONS Theory, Measurement, and Analysis: Past and Present

THE

INCOME OF NATIONS THEORY, MEASUREMENT, AND ANALYSIS: PAST AND PRESENT

A Study in Applied Economics and Statistics

PAUL S T U D E N S K I Professor Emeritus of Economics, N ew York University

Washington Square • NEW YORK U N IVER SITY PRESS 1958

© 1958

BY NEW YORK UNIVERSITY PRESS

LIBRARY OF CONGRESS CATALOGUE CARD NUMBER:

57-6376

MANUFACTURED IN TH E UNITED STATES OF AMERICA

30498

To Nastinka, my wife, whose spark and wisdom have guided me through life

Preface This is a work of many years. It was started in the middle 1930’s as a long-range project and was carried on intermittently after that, parallel with other research and teaching engagements. Its preparation was interrupted several times by untoward de­ velopments. Each time the work was re­ sumed, the manuscript had to be recast to take care of the newer events in the field and of the newer materials that had become available. The final form of the work is different from the one designed originally. It empha­ sizes the historical background of the exist­ ing national income concepts, estimates, and critical appraisals of them more than did the earlier version. It is also much more selective in its factual and statistical presen­ tations. The passage of time, even more than the conscious efforts of the author, has helped to enrich the work. It is impossible to name all the indi­ viduals and organizations that assisted in the preparation of this book. Foremost ac­ knowledgment is due to Professor Julius Wyler of the New School for Social Re­ search, who was co-author over a period of several years but was forced by poor health to terminate the collaboration. His contri­ bution to this volume was great. Acknowl­ edgment is due next to Professor Simon Kuznets, who lent encouragement to the project from the beginning and never failed to give support when called upon to do so. Professor Jacob Marschak, as chairman of a special committee of the Conference on Research in Income and Wealth, gave valu­ able advice during one stage of the work.

Great debt is owed to the authorities of New York University who sponsored the project from the start; and to the Rocke­ feller Foundation, which on two occasions during the past twelve years made financial grants for the study. Acknowledgment is also due to the Federal government, whose long defunct Works Progress Administra­ tion provided needed clerical assistance dur­ ing the 1930’s; to the National Bureau of Economic Research, which rendered like as­ sistance intermittently during the 1940’s; and to the staffs of the New York Public Library, New York University Library, and the New York State Library for the many courtesies extended in the use of their books and document collections. Among the several research assistants who aided in the project over these past twenty years none have been more helpful than the two who were associated with the final, most crucial stages of the work: Miss Dorothy Dorfman, who helped to bring up to date many of the country chapters, fac­ tual statements, and references; and Mrs. Rosalie Becker, who helped to edit the final manuscript; both performed their tasks with extraordinary thoroughness and rare intelli­ gence and skill. Special acknowledgments are made in some of the individual country chapters to the foreign experts who reviewed them when they were still in preliminary draft form. P aul Studenski

New York City September, 1958

Contents In tro d u c tio n

IMPORTANCE OF NATIONAL INCOME ESTIMATES 1. Nature of Recent Develop­ ments 2. Organization of This Vol­ ume: Its Historical Part 3. The Theoretical and Meth­ odological Part 4. Individual Analyses of Thir­ teen Selected Countries’ Estimates 5. Developments in Sixty-Six Other Countries 6. Statistical Appendix

3 3 5 5 6 6 7

Part One. HISTORY DEVELOPMENT OF CON­ CEPTS 1. The Sequence of the Main Ideas 2. The Mercantilist Concept 3. Origination of National In­ come Concept by Dissent­ ers from Mercantilism: Petty, King, Davenant, Boisguillebert, Vauban 4. The Physiocratic Concept of Production and of Na­ tional Income: Quesnay and Others

11 11 13

13

15 IX

5. Italian Economists Going Their Own Way: Galiani, Verri, and Palmieri 6. Adam Smith’s Material Pro­ duction Concept 7. The Return to the Compre­ hensive Production Con­ cept: Lauderdale, Storch, Marshall, and Others 8. Marxian National Income Theory a. Revival of the Smithian Material Production Concept Under a Marx­ ian Imprint b. Marx’s Theory of Value c. Marx’s Recognition of the Correspondence Be­ tween Income and Product d. Marx’s Distinctions Be­ tween Gross and Net Product 9. The Contrasting Treatments of the Two Concepts in the Present Century 10. The Keynesian Dynamic Ap­ proach

25

DEVELOPMENT OF FIRST ESTIMATES IN ENGLAND IN THE SEVENTEENTH CENTURY

26

1. Petty—the Originator of Na­ tional Income Estimates

26

16 18

20 21

21 23

23 23 24

CONTENTS

X

a. Petty’s Part in the Inven­ tion of Statistics or “Political Arithmetick” b. Circumstances That Led Petty to His Discovery c. Petty’s Estimates d. Petty’s Statistical Con­ clusions e. Petty’s Generalizations Often Running Ahead of His Statistics 2. Gregory King’s Novel Esti­ mates: A Neglected Mile­ stone a. Sources of King’s Interest in Economic Statistics and in National In­ come Estimates b. Elaborateness of King’s Estimates c. Bases of King’s Estimates: All Three Phases Cov­ ered d. Estimates of the Coun­ try’s War Potential e. King’s International Com­ parisons f. Regrettable Neglect of King’s Great Work 3. Davenant—Popularizer of the Estimates and Pleader for Better Statistics a. Exposition of the Uses of the Estimates b. Exposition of Sources, Methods, and Skills Required for the Prep­ aration of Estimates c. Pleas for Better Statistics and for Freer Access to Them d. Davenant’s International Comparisons 4. Summary ELABORATIONS IN ENG­ LAND IN THE EIGHT­ EENTH CENTURY 1. The Period of Neglect of “Political Arithmetick,” 1700-1770 a. The British Merchant’s Estimate for 1713-21 b. Sir Matthew Decker’s Es­ timates for 1740 c. The Anonymous French­ man’s Estimates of 1764

2. The Revival of Interest in Political Arithmetic In­ cident to the Industrial Revolution a. Arthur Young’s Estimates of 1770—A Signal of a New Era in Eco­ nomic Statistics b. Mr. Pulteney’s Amateur­ ish Estimate 3. The Influence of the Napo­ leonic Wars and of the In­ come Tax Proposal a. Pitt’s Estimates 4. Dr. Beeke’s Estimates of 1799-1800 a. Beeke’s Estimates of In­ come from Agriculture and Certain Other Sources b. Beeke’s Estimate of La­ bor Income c. Beeke’s Estimate of In­ come from Internal Trade d. Beeke’s Observations Re­ garding the Distribu­ tion of National In­ come e. The Merits of Beeke’s Estimates 5. Dr. Bell’s Estimates of 1799-1802 a. The Source of Bell’s In­ terest in National In­ come b. Bell’s Critique of Pitt’s Income Tax and His Substitute Proposal c. Bell’s National Income Estimate 6. Gentz’s Recalculations Based on the Smithian Concept 7. Comparison of the Results of the Estimates of the Two Centuries

26 27 28 28 29 30

30 32 33 33 33 37 37 37

38 38 39 40

40

40 40 41 41

4

EARLY ESTIMATES IN FRANCE: BOISGUILLEBERT AND VAUBAN 1. Boisguillebert—Good Eco­ nomic Theorist but Poor Statistician a. A Mild Man Turning into a Severe Critic of His Government b. Official Disapproval of Boisguillebert’s Views

41

41 43 43 43 43

44 45 46

46 47 47 47 48 48 50 51

52

52 52 53

CONTENTS

c. Boisguillebert’s National Income Estimates 2. Marshal Vauban— Military Genius Turning Into a Fiscal Reformer and Na­ tional Income Estimator a. Vauban’s Life b. Vauban’s Proposed Tax Reform c. Estimates of the Revenue from the Tithe and of Agricultural Income d. Vauban’s Proposal for Annual Enumeration (Population, Income, and Wealth) e. Vauban’s Estimates of Revenue from Pro­ posed Taxes on Specific Money Incomes f. Estimates of Revenues of the Third and Fourth Funds g. Vauban’s Estimate of To­ tal Revenue as a Base for Estimating National Income h. Appraisal of Vauban’s Estimates 3. A Lapse of Interest in Na­ tional Income Analysis (1707-58) 5

CONTRIBUTIONS OF FRENCH PHYSIOCRATS: QUESNAY, LAVOISIER, AND OTHERS 1. Quesnay’s Hypothetical Esti­ mates and Invention of the Money-Product Flow Analysis a. A Physician Becomes an Economist b. The Theory of the “Ta­ bleau Economique” c. The “Tableau” 2. Theoretical Analyses by Forbonnais, Condillac, and Turgot 3. Estimates Published During the Immediate Pre-Revo­ lutionary Era: Du Pont, Tolosan, Young, and Oth­ ers 4. Lavoisier’s Comprehensive and Systematic Estimates a. Lavoisier’s Life—a Great Chemist Becomes a

XI

54

54 55 55 56

56

57

5.

58 6. 59 59 60

61

61 62 62 64 65

66 68

7. 8.

Practical Social Scien­ tist b. Lavoisier’s Perception of the Importance of Good National Income Estimates c. Lavoisier’s Concept of National Income d. Three Major Types of Data Used by Lavoisier for His Estimates e. Lavoisier’s Estimate of the National Income and of Its Distribution f. Lavoisier’s Estimate of the “Net” or “Taxable” Income g. Appraisal of Lavoisier’s Estimates Lagrange’s Estimates of France’s Annual Con­ sumption—A Side Ven­ ture by a Great Mathema­ tician Other Estimates of the Revo­ lutionary Period Estimates by D’lvernois Summary of the Estimates, 1690-1798

THE THREE LOST RUSSIAN ESTIMATES OF THE END OF THE EIGHTEENTH CENTURY 1. The Tremendous Transfor­ mation of the Country During the Century a. Territorial Expansion b. Growth of Population and Economic Activity c. Advancement of Educa­ tion d. The Rule of Bureaucracy e. Influx of Scholars and Technicians from Abroad 2. The Growth of Statistical Data and Inquiries During the First Half of the Cen­ tury a. The First Censuses of Population b. The First Vital and Eco­ nomic Statistics c. The Economic and Statis­ tical Inquiries of the First Half of the Cen­

tury

69

70 71 72 73 74 75

75 76 76 77

78

78 78 79 79 80 80

80 80 81

82

xii

CONTENTS

3. The Great Improvements in Statistics Under Catherine the Great 83 4. The Convergence in Russia of Three Schools of Statis­ tics Developed Abroad 84 a. The French Mathematical School 84 b. The German Descriptive School 85 c. The English Political Arithmetic School 85 5. The “Voluntary Economic Society” 87 6. The “Society of Friends of the Literary Sciences” 89 7. The Birth of the Three Na­ tional Income Estimates 90 8. Hermann’s Estimate 90 a. His Life and Writings 90 b. His Statistical Handbook 91 c. Hermann’s National In­ come Estimate 92 93 9. Radishchev’s Estimates a. His Life and Writings Up to His Exile to Siberia 93 94 b. His Writing in Exile c. His Work After His Re­ turn from Exile 95 d. Radishchev’s National In­ come Estimate 96 10. The Estimate Presented in the “New World Geogra­ phy” 97 a. The Identity of the Au­ thors and Nature of the Work 97 b. Reasons for Including a National Income Esti­ 98 mate c. The National Income Es­ timate of the “New World Geography” 99 11. Appraisal of the Three Esti­ 100 mates NEW APPROACHES IN ENG­ LAND IN THE NINE­ TEENTH CENTURY 1. Factors Responsible for the New Approaches I. Estimates by the Net Output Method, 1811-46 1. Colquhoun’s Estimates of 1812-14 a. Colquhoun’s Approach b. Sources of Colquhoun’s Estimates

101 101 102 102 102 103

c. Colquhoun’s Applica­ tion of the Smithian Concepts d. Examples of Colqu­ houn’s Estimating Methods e. Colquhoun Echoes the Pleas Made by Davenant and Lavoisier for Governmental Estimates f. Critical Appraisal of Colquhoun’s Esti­ mates 2. Projections and Interpre­ tations of Colquhoun’s Estimates by John Gray and Simon Gray 3. Lowe’s Estimates in the 1820’s: the First to be Expressed in Constant Prices 4. Pebrer’s Projections in the Early 1830’s 5. Influence of the Contro­ versy Between Agricul­ tural and Manufactur­ ing Interests in the 1840’s on Inquiries into National Income a. McCulloch’s and Por­ ter’s Lack of Faith in National Income Estimates b. Spackman’s National Income Estimates: a Defense of Agricul­ ture II. Estimates by the Income Dis­ tributed Method, 18461900 1. Smee’s Estimates of 1846 —The First to Be Based in Part on In­ come Tax Returns 2. Baxter’s Estimates of 1868 a. More Solid Estimate Than Any of Its Predecessors b. Baxter’s Concept of National Income c. Baxter’s Calculations 3. Estimates by Giffen and Others III. Summary of the British Esti­ mates of the Nineteenth Century

104

104

106 106

106

107 109

110

110

110

111

111 114 114 115 115 117

117

CONTENTS

MULTIPLICATION OF ESTI­ MATES IN FRANCE IN THE NINETEENTH CEN­ TURY 1. Estimates of the Napoleonic Era a. Ganilh’s Estimates b. Montalivet’s Colored Re­ port of 1813 c. Chaptal’s Estimates d. Say’s Contributions to National Income Anal­ ysis 2. The Period of Lack of Inter­ est, 1815-50 3. The Period of Revived In­ terest, 1850-99 a. Moreau de Jonnes b. Break with Tradition: Leroy-Beaulieu, de Foville, Coste, Parliamen­ tary Commission FIRST DEVELOPMENTS BE­ YOND ENGLAND AND FRANCE 1. Development in the United States a. Tucker’s Estimates for the United States for 1840-50 b. Spahr’s Estimates for the United States for 1890 2. The First Estimates in Aus­ tria 3. Why Was Germany a Late­ comer to National Income Estimating? 4. Australia Producing Official Estimates: A Major Event in World Statistical Devel­ opment 5. Developments in Other Countries 6. Absence of National In­ come Estimates in Russia Throughout Most of the Nineteenth Century 7. The First International Col­ lations: Levi and Mulhall 8. Summary 10

UNIVERSALIZATION IN THE TWENTIETH CENTURY; RETROSPECT OF PAST DEVELOPMENT 1. The Period 1900-1917

119 120 120 122 123 123 124 125 125

126

129 129 129 132 134 134

135 137

137 140 141

142 142

x iii

a. England, France, United States: Bowley, Stamp, Colson, and King b. German Estimates Com­ ing to the Fore: May and Helfferich c. Influence of Austro-Hungarian Estimates: von Fellner d. Australia, New Zealand, and Norway: Coghlan and Others e. Five New Country Esti­ mates: Japan, Switzer­ land, the Netherlands, Italy, and Bulgaria f. National Income Esti­ mates Coming into Prominence in Russia 2. The Extraordinary Flourish­ ing of Estimates During the Period 1918-39 a. Private Institutional Re­ search Supplementing Individual Effort b. A New Group of Schol­ ars: Colin Clark, Kuznets, Gini, and Others c. Estimates Become Official in Nine Countries d. National Income Re­ search Elevated to a Central Place in Eco­ nomics 3. Universalization, Enrich­ ment, and Standardization of Estimates During and After World War II a. National Income Esti­ mates Become the Core of Governmental Fiscal and Economic Planning b. Equal Importance Attrib­ uted to Estimates of National Expenditure c. Shift in Emphasis from Net to Gross Product Figures d. Development of “Sector Accounts” : Complete Social Accounting Ap­ proach e. Standardization of Con­ cepts by International Agreement and Exam­ ple

142 144 144 145

145 145 149 149 150 150

151

151

151 153 153

154

154

CONTENTS

x iv

f. Development of the Input-Output Approach g. Universalization of Prep­ aration of Estimates During the Postwar Period 4. Forces That Influenced the Development of the Past Three Hundred Years a. Interest and Initiative of Individual Scholars b. Advance of Economic Theory c. Wars, Revolutions, and Severe Economic Crises d. Agitation for Fiscal Re­ form e. Rapid Technological Changes in the Econ­ omy f. Class and Group Interest Conflict g. International Commercial and Political Rivalries h. Improvement in Statisti­ cal Data and Methods i. Introduction of the In­ come Tax j. Development of Govern­ mental Fiscal and Eco­ nomic Planning k. Influence of International Organization

155

155

4. 158

5. 158 158 158 158 158 159 159 159 159 159 160

Part Two. THEORY AND METHODOLOGY 11

BASIC CONCEPTS 1. The Meaning of National Income a. Distinction Between Eco­ nomic and Noneco­ nomic Production; Be­ tween Economic Wel­ fare and Welfare Gen­ erally b. The Social Nature of the Concept c. National Income as a Double Flow: That of Product and That of Money Income 2. The Structure of National Income a. Combination of Two Types of Magnitudes

3.

163 163

163 165

166 168 168

6. 7.

b. Importance of a Synthe­ sized Three-Phase Treatment: Produc­ tion; Distribution; Dis­ position National Income as an Ag­ gregate of Net Outputs National Income as an Ag­ gregate of Factor Shares National Income as the Ag­ gregate Value of Final Products at Factor Cost and Also at Market Prices Net Income from Abroad Summary

12 THE

AREA TION

OF

168

169 170

171 173 174

PRODUC­

1. The Comprehensive Produc­ tion Concept a. Inclusion of Value of Foods and Other Goods Produced on the Farm for the Farmer’s Own Consumption b. Unpaid Services of House­ wives and of Other Members of the House­ hold c. Inclusion of Services of Owner-Occupied Dwell­ ings d. Treatment of Services of Other Private Consumer Durables e. Treatment of Services of Government Properties f. Treatment of Services of Durable Properties of Churches and Other Benevolent Organiza­ tions g. Other Issues 2. Adam Smith’s Restricted Material Production Con­ cept 3. The Marxian Restricted Ma­ terial Production Concept a. Emphasis on the Sole Productiveness of Hired Labor b. Clearer Distinction Be­ tween the Various Ele­ ments of Gross and Net National Income c. Specific Definitions of the Area of Production

175 175

176

177 178 179 180

181 181 181 183 183

184 184

XV

CONTENTS

d. Application of the Marx­ ian Concept in Soviet Russia’s Estimates e. Conceptual Differences Among the Marxists f. Critique of the Marxian Concept 4. The Restricted Market Pro­ duction Concept 5. Conclusion 13

FINAL VS. INTERMEDIATE PRODUCTS IN THE MAR­ KET ECONOMY 1. The Empirical Rule 2. Unresolved Theoretical Issues a. Fisher’s and Gini’s Re­ strictions b. Shoup’s, Lowe’s, and Kuznets’ Distinctions 3. The Special Problem of Fi­ nancial Intermediaries a. Banks b. Life Insurance Companies 4. Summary

14

TREATMENT OF GOVERN­ MENT 1. Treatment of Government Services a. Forms of Presentation in Different Phases of the National Income b. Method of Evaluation c. Classification of Govern­ ment Expenditures d. Treatment of Government Services Dependent on the Philosophy of Gov­ ernment Functions e. The View That All Gov­ ernment Services Are Intermediate Products: Matolcsy and Varga f. The View That All Gov­ ernment Services Are Final Products: Ger­ hard Colm g. The Theory That Govern­ ment Services Are Partly Final and Partly Intermediate Products: Kuznets’ Position 2. Treatment of Taxes a. Classification of Taxes

b. Treatment of Taxes Un­ der the Comprehensive Production Concept 201 c. Treatment of Taxes Un­ der the Material and Market Production Concepts 203 3. Treatment of Subsidies to Business 203 4. Treatment of Contributions to and Payments from Social Security, Pension, 204 and Welfare Funds

185 186 187 187 188

188 188 190

15

191 191 192 193 194 194 194 194 195 195 16

196

197

198 201 201

205

1. Various Aggregates and Their Interrelations 205 a. Nature of the Various Aggregates 205 b. Interrelationships of Var­ ious Aggregates 207 2. Social Accounts 209 a. Types of Accounts 209 b. The Area of the Greatest Usefulness of the So­ cial Accounting Ap­ 212 proach c. The Need for Balance Be­ tween the National In­ come and the Social 213 Accounts Approaches 3. The Nation’s Economic Budget 214 4. The Input-Output Approach: Leontieff 215

190

195

VARIOUS AGGREGATES, ANALYSES, AND SOCIAL ACCOUNTS

INTERTEMPORAL AND IN­ TERNATIONAL COMPARI­ SONS 1. Comparisons Over Time a. Deflation of Total Cur­ rent Value by an Over­ all Cost of Living Index b. Deflation of Current Val­ ues of Different Items by Different Indexes c. Evaluation of Current Physical Outputs in Base Year Prices 2. Forecasts, Projections, and Plans a. Forecasts: Long Range and Short Range; and Planned Goals

217 217 218 218 219 220

220

CONTENTS

XVI

b. Extent of Estimating In­ volved c. Influence of the State of Economic Data d. Degree of Accuracy of Income-distributed Method e. Degree of Accuracy of the Net Output Method f. Degree of Accuracy of Income-disposed or Fi­ nal Products Method g. Summary 3. Extrapolation and Interpola­ tion a. Census Figures Projected on the Basis of Less Complete Annual Data b. Value Aggregates Pro­ jected on the Basis of Changes in Quantities and Values of Com­ ponent or Related Items c. Net Value of Output Pro­ jected on the Basis of Variations in Gross Value and of Either a Constant or a Varia­ ble Net-gross Ratio d. Substitution of Suppos­ edly Co-ordinate Series for Unknown Data e. Projections Based on General Appraisal of Business Conditions or Arbitrary Assumptions 4. Reliability of the Final Out­ come of the Estimates a. Qualitative Appraisals b. Measures of Degree of Accuracy 5. Classification of Estimates of Various Countries Ac­ cording to Their Relia­ bility

b. Hypothetical Projections 221 c. Two Recent Projections in the United States 221 3. Problems of International Comparisons 224 a. The Misleading Nature of Conversions into a Sin­ gle Currency on the Basis of Foreign Ex­ change Rates 224 b. Adjusted Comparisons by Gilbert-Kravis and by Colin Clark 225 c. Other Approaches 226 4. Effects of Intertemporal and International Differences in the Distribution of Na­ tional Income 227 5. Actual International Com­ parisons 227 a. The Relative per Capita Incomes of the Coun­ tries 228 b. Differences in the Indus­ trial Origins of Income 233 c. Differences in the Dis­ tribution of Income Among the Factors of Production 237 d. Expenditures of the In­ come for Different Ob­ jects 240 e. Differences in the Rate of Growth of Income and Their Underlying Causes 244 f. Summary 250 17

ESTIMATING METHODS 1. Main Characteristics of the Three Methods a. Functional Interdepend­ ence of Methods b. Choice of Methods for Different Divisions of Economic Activity c. Dependence of Method on the Structure of the Economy d. Statistical Short Cuts e. Classification of Estimates by Method 2. The Processes for Achieving Accuracy Under Each Method a. Coverage and Evaluation

251 251 251

251

252 253 253

254 254

18

THE NET OUTPUT METHOD 1. Characteristics of the Method 2. Classification of Industrial Divisions 3. Estimating the Gross Value of the Output a. Agriculture, Including Fishing and Forestry b. Mining c. Construction

255

255 256 257 257 258 258 259

259

260 260

260 261 261 261

263 264 264 264 266 267 269 269

xvii

CONTENTS

d. Manufacturing e. Transportation and Com­ munication f. Trade g. Services of Dwellings h. Services of Financial In­ termediaries i. Services of Professions and Other Divisions Es­ timated by the Incomedistributed Method 4. Deductible Costs a. General Approach b. Costs of Materials, Serv­ ices, and Taxes c. Depreciation 5. Summation and Final Ad­ justment of the Net Value of Output INCOME-DISTRIBUTED METHOD I. Income-paid-out Variant 1. Classification by Type of Income 2. Labor Income 3. Mixed Income 4. Capital Income a. Dividends b. Undistributed Profits c. Changes in Inventories d. Interest e. Rent II. Income-received Variant 1. General Characteristics and Application 2. Adjustments of Income Tax Data a. Additions to Assessed Incomes b. Deductions from As­ sessed Incomes c. Other Adjustments to Assessed Incomes THE EXPENDITURE OR FI­ NAL PRODUCTS METHOD 1. General Characteristics 2. Classification of Final Ex­ penditures or Products a. Private Consumer Goods and Services b. Domestic Capital Forma­ tion: Private or Com­ bined Private and Gov­ ernment

269 270 270 271 272

272 273 273 273 274 276 276 277 277 278 280 280 281 281 281 282 284 285 285

c. Current Government Ex­ penditures on Goods and Services 291 d. Net Capital Investment Abroad 292 3. Retail Sales Value and Family Budget Survey Methods 292 a. Retail Sales Value Method 292 b. Family Budget Survey Method 293 4. Special Problems of Estimat­ ing Domestic Capital For­ mation 294 5. Commodity Flow Method of Estimating Consumer Ex­ penditures and Investment 294 a. Manufactured Products (Table 20-1) 294 b. Agricultural, Mining, and Other Non-manufactured Products (Table 20 - 2 )

c. Services d. Adjustments for Interna­ tional Accounts and Government Expendi­ tures e. Partial and Simplified Commodity Flow Esti­ mates

295 296

296

297

Part Three. ESTIMATES FOR SELECTED COUNTRIES 301

288

Explanatory Note to Part Three 1. Reasons for Selecting Thir­ teen Countries’ Estimates for Detailed Individual Review 2. Distinguishing Character­ istics of These Countries’ Estimates 3. Organization of the Coun­ try Chapters

288

CANADA

303

288

References 1. History I. Net Output and Incomepaid-out Estimates 1. General Methodolgy 2. Net Output Portion of the Estimates

303 303

285 285 287 287

289

290

301

301 303

304 304 305

x v iii

CONTENTS

3. Incom e-paid-out P arts of the Estim ates a. L abor Incom e b. C apital Incom e c. Incom e of the Selfem ployed 4. G eographical D istribu­ tion of Personal In ­ come 5. Gross N ational P roduct and G ross N ational E xpenditure II. Estim ates of N ational E x­ penditure by the Final P roducts M ethod 1. G ross N ational Expendi­ ture at C urrent Prices a. C onsum er Expendi­ tures b. G overnm ent Expendi­ tures c. G ross D om estic In ­ vestm ent d. N et Investm ent A broad 2. G ross N ational Expendi­ ture at C onstant Prices a. T he N atu re o f the C onstant D ollar Se­ ries and Im plicit Price Indexes b. D eflation o f Personal E xpenditure on C onsum er G oods and Services c. D eflation o f G overn­ m ent Expenditure on G oods and Serv­ ices d. D eflation o f G ross D om estic Invest­ m ent e. D eflation of Exports and Im ports 22

306 306 307

4. F in al P ro d u ct o r Expendi­ tu re Estim ates 23

310

311

312

313 313 313 314 314 315 315

315

316

316

316 317

IR E L A N D

318

References 1. H istory 2. Concepts and G eneral M eth­ ods 3. N et O utput and Income-distributed Estim ates a. A griculture, Including Fishing b. M anufacturing c. T rade and T ransport d. O ther Services e. Rest of the W orld

318 318 318 318 318 321 321 322 323

24

323

TH E NETHERLANDS

325

R eferences 1. H istory 2. Concepts and Sources 3. Incom e-distributed Estim ates a. T axable Incom es b. T ax-exem pt Incom es; E x ­ em ptions fo r C hildren c. T ax E vasion d. C orrection fo r T im e Lag in D ividend Paym ents e. T ran sfer Paym ents 4. Subsidiary N et O utput Esti­ m ates a. A griculture and Fishing b. M ining, Public Utilities, an d C onstruction c. M anufacturing d. T rad e e. Banking and Insurance f. Services o f Dwellings g. Shipping h. O ther T ransportation and C om m unications i. G overnm ent Services j. O ther Services k. N et Incom e from A broad 5. Breakdow n of N ational In ­ com e by Types of Incom e Paym ents; G ross C apital F orm ation 6. D evelopm ents Since 1940

325 325 326 326 327

331 331

SW E D E N

334

327 328 328 328 328 328 329 329 330 330 330 330 330 331 331 331

R eferences 334 1. H istory 334 I. P rew ar E stim ates 335 1. C oncepts in the P rew ar Estim ates 335 2. G eneral Estim ating M eth­ ods 335 a. P eculiar C haracter­ istics 335 b. G ross V alue of O ut­ put and D eductions fo r M aterials and Services 337 c. D eduction o f D ep re­ ciation 337 d. N et V alue o f O utput 338 3. Estim ates by Industrial Divisions 339

x ix

CONTENTS

a. A griculture, Forestry, Fishing, and H unting b. M anufacturing, M in­ ing, and H andicraft c. C onstruction d. T rade, Finance, and Professions e. T ransportation and C om m unications f. Services of Consum er D urable G oods g. G overnm ental Services h. D om estic Services 4. C onsum ption and C apital F orm ation 11. P ostw ar Estim ates 1. C oncepts and Basic M ethods 2. Estim ates fro m the E x­ penditure Side 3. Estim ates from the In ­ com e Side

346

SO V IE T RU SSIA

349

References 1. H istory a. E arly Beginnings F o l­ lowed by a C entury of Inactivity b. P rokopovitch’s Estim ates o f 1906 and 1917 c. Official Soviet Estim ates D uring the P eriod of T heir F ree Publication d. Shift to a Policy of Se­ crecy e. C ritical A ttitudes T o­ w ard Soviet Estim ates A broad 2. C oncept o f N ational Incom e 3. Statistical Sources 4. D ivision by E conom ic A c­ tivity 5. Estim ates at C urrent Prices a. Incom e from A griculture b. Incom e from Large-scale Industry c. Incom e from Small-scale H andicraft and D o­ m estic Industry d. Incom e fro m C onstruc­ tion e. Incom e from Freight T ransportation f. Incom e from R etail T rade 6. D istribution and Redistribu-

339 340 341 341 342 342 343 343

7.

343 344 345 345

349 350

350 351

8.

9.

10. 11. 12.

tion of N ational Incom e a. Incom e D istributed by State-owned Enterprises b. Incom e D istributed by K o lkh o z and O ther Enterprises c. R edistribution of N a­ tional Incom e d. Reconciliation of D istri­ bution an d Redistribu­ tion of Incom e C om putation o f N ational In ­ com e at C onstant Prices a. T he “D ynam ics o f Gross O utput” M ethod b. T he “Index o f the Price of the P ro d u ct” M ethod c. D irect C om putation of N et V alue in C onstant Prices d. D istortions U n d er All T hree M ethods O verstatem ents D ue to O ther th an M ethodological F ac­ tors Colin C lark’s R ecom puta­ tions o f the Soviet N a­ tional Incom e Estim ates by W yler Corrections by Bergson and Associates Prospects fo r the Official P reparation of a MoneyF low Balance Sheet and of Social A ccounts Sum m ary

362 363

365 366

367 367 367 367

368 368

369

371 372 372

373 373

351

13.

351

GERM ANY

374

References 1. H istory I. P rew ar Estim ates by Incom edistributed M ethod 1. Concept of N ational In­ com e in the P rew ar Estim ates 2. Estim ates fo r 1891-1913 3. Estim ates for 1925-31 a. Statistical Sources b. L abor Incom e c. A gricultural E n tre­ preneurial Incom e: F arm Incom e d. N onagricultural E n tre­ preneurial Incom e e. C apital Incom e f. Taxes and Fees 4. Estim ates fo r 1932-38

374 374

352 352 353 354 354 355 357

358 359 361 362

375

375 376 376 376 377

378 378 378 379 379

CONTENTS

XX

5. Estimates by Size of In­ dividual Incomes II. Postwar Net Output Esti­ mates 1. Estimates for the Former Reich for 1936 by the Net Output Method 2. Computation of West Germany’s Share of the Reich’s National Income in 1936 3. Extrapolation of the Es­ timate for 1936, by the Net Output Method, to 1948 4. Subsidiary Estimates of Distribution and Ex­ penditure of National Income 5. Subsequent Improvement of Estimating Methods 6. Estimates Prepared in 1956-57 27

e. Value Added in the Saar and in the Public and Social Security Sectors 2. Estimates Prepared by the Income-distributed Method 3. Estimates Prepared by the Final Expenditure or Products Method 4. Further Improvements of Estimates in 1954 and 1955

381 382 383

384

385

28

385 388 389

FRANCE

391

References 1. Recent History 2. Concepts I. Estimates Prepared During 1947-51 1. Estimates by the Net Output Method 2. Estimates by the Incomedistributed Method 3. Estimates by the Final Expenditure or Prod­ ucts Method a. Final Product of the Enterprise Sector b. Gross National Prod­ uct and Expenditure 4. Degree of Reliability of the 1947-51 Estimates II. Estimates Prepared in 195255 1. Estimates Prepared by the Net Output Method a. Value Added in Pri­ vate Industrial and Agricultural Pro­ duction b. Value Added in Other Branches of Pro­ duction c. Value Added in Trade d. Corrections for Tax Evasion

391 391 392 392 393 395 395 396 398 398 400 400

400 401 401 401

29

402

402 402 406

YUGOSLAVIA

408

References 1. History I. Prewar Estimates (Djuricic et al. and Official) 1. Concepts and Methods a. Agriculture, Mining, and Construction b. Manufacturing c. Transportation and Communications d. Trade and Finance II. Postwar Estimates a. Agriculture, Fishing, and Forestry b. Mining, Manufactur­ ing, and Electrical Industry c. Trades and Handi­ crafts d. Transportation and Communications e. Construction f. Trade

408 408 409 409 409 409 410 410 411 411 412 412 413 413 413

TURKEY

414

References 1. History I. Estimates Prepared During the Period 1935-50 1. Concepts 2. Estimating Methods a. Agriculture b. Forestry c. Manufacturing, Min­ ing, and Other In­ dustry d. Trade, Finance, and Transportation e. Services of Dwellings, Government, Pro­ fessions, and Others

414 414 415 415 415 415 416 417 418

419

xxi

CONTENTS

II. Estimates Prepared in 1951— 54 1. General Approach 2. Estimates by Industrial Divisions a. Agriculture and For­ estry b. Manufacturing and Other Industry c. Construction d. Trade, Finance, and Transportation e. Services (Professions, Dwellings, Govern­ ment) and Rest of the World III. Methods of Deflation of the Income Series

1. Consumer Expenditures 2. Gross Domestic Invest­ ment 3. Net Capital Formation 4. Savings Available for Net Capital Forma­ tion 5. Projections Under the Second Five-Year Plan

419 419 419 419 421 421 421

31

422 423

INDIA

424

References 1. History 2. Concepts I. Net Output Estimates 1. General Methods and Sources a. Methods in Rao’s Two Estimates b. Methods of Official NIU Estimates 2. Agriculture, Forestry, Fishing, and Mining a. Rao’s Estimates b. Official NIU Estimates 3. Manufacturing and Hand­ icrafts a. Rao’s Estimates b. Official NIU Estimates 4. Trade, Finance, and Transportation a. Rao’s Estimates b. Official NIU Estimates 5. Services of Dwellings a. Rao’s Estimates b. Official NIU Estimates 6. Government, Professions, and Domestic Service a. Rao’s Estimates b. Official NIU Estimates 7. Net Income from Abroad a. Rao’s Estimates b. Official NIU Estimates 8. Reliability of the Esti­ mates II. Final Expenditure or Product Estimates (Mukherjee and Ghosh)

424 424 425 425

32

425 426 428 428 430 430 430 431

432 432 433 433 433 433 433 434

434 436 438 439 440

Reference 1. Labor Income 2. Profits 3. Interest and Rent 4. Income from Agriculture 5. Net Income from Abroad

440 440 441 441 441 443

BRITISH WEST INDIES AND MALAYA

443

References 1. Importance of Benham’s Es­ timates I. Estimates for British West In­ dies II. Estimates for Malaya 1. Estimates of the Net Out­ put 2. Estimates of Final Expendi­ tures 3. Comparison of Estimates by the Net Output and Final Expenditure Meth­ ods 4. Estimates of Private Ex­ penditure and Saving

425

431 431 432 432 432 432

BRAZIL

434

443 443 444 445 445 448

448 449

Part Four. DEVELOPMENTS IN SIXTYSIX OTHER COUNTRIES

33

Explanatory Note to Part Four

453

DEVELOPMENTS IN THE UNITED STATES, UNITED KINGDOM, AUSTRALIA, NEW ZEALAND, AND SOUTH AFRICA (with Bib­ liographies)

455

1. 2. 3. 4. 5.

United States United Kingdom Australia New Zealand Union of South Africa

455 457 458 460 460

XXii

CONTENTS

DEVELOPMENTS IN EIGHT­ EEN EUROPEAN COUN­ TRIES I. General Review 1. Developments in the Free Enterprise Econ­ omies 2. Developments in the Six Communist Countries II. Reviews of Developments in Individual Countries (with Bibliographies) 1. Italy a. Prewar Estimates b. Postwar Estimates 2. Switzerland 3. Belgium 4. Luxembourg 5. Denmark 6. Norway 7. Finland 8. Iceland 9. Austria 10. Hungary 11. Czechoslovakia 12. Poland 13. Bulgaria 14. Rumania 15. Albania 16. Spain 17. Portugal 18. Greece

461 462 462

36

462 464 464 464 465 466 468 469 469 471 471 472 472 473 474 475 477 478 478 479 479 480

DEVELOPMENTS IN LATIN AMERICA AND THE CAR­ IBBEAN (with Bibliographies)

481

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

482 482 483 483 ' 484 484 484 484 485 485 485 486 486 486

Cuba Puerto Rico The Dominican Republic Mexico Guatemala Nicaragua Honduras El Salvador Costa Rica Panama British Guiana Colombia Venezuela Ecuador

15. 16. 17. 18. 19. 20.

37

Bolivia Paraguay Uruguay Argentina Peru Chile

487 487 487 487 488 488

DEVELOPMENTS IN THE NEAR, MIDDLE, AND FAR EAST (with Bibliographies)

489

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

Egypt Israel Lebanon Syria Pakistan Ceylon Burma Thailand (Siam) Indonesia Hong Kong China Japan a. Estimates Before World War II b. Postwar Estimates 13. Korea 14. Taiwan (Formosa) 15. Philippines

490 491 492 492 492 493 494 494 494 494 495 497

DEVELOPMENTS IN AFRICA

500

1. 2. 3. 4. 5. 6. 7.

Nigeria Belgian Congo Kenya Northern Rhodesia Southern Rhodesia Nyasaland Ghana and Tanganyika

497 497 499 499 499

501 501 501 502 503 504 504

STATISTICAL APPENDIX: National Income Aggregates of Eighty-Seven Countries at Mid-Century 1950-55 Explanations Summary Table

507 508

NOTES

513

INDEX

545

T H E INCOME OF NATIONS Theory, Measurement, and Analysis: Past and Present

Introduction IM PO R TA N C E O F N A TIO N A L INCOM E ESTIM ATES first of all, by a considerable clarification and refinement of the basic national income concepts. The result has been achieved through the work of Bowley, Stamp, Clark, Keynes, Hicks, and Stone in England; Kuz­ nets, Gilbert, Colm, Copeland, and Gold­ smith in the United States; von Fellner, Matolcsy, and Varga in Hungary; Jostock in Germany; Dahlgren, Lundberg, and Ohlson in Sweden; Frisch in Norway; Chernomordik and Kolganoff in Soviet Russia; Perroux and Gruson in France; and Gini and Barberi in Italy. Secondly, the preparation of national in­ come estimates, which prior to the 1930’s was confined to some twenty of the most industrially advanced countries, has become practically universal by 1957. Everywhere it is now a regular function of government, with annual or even more frequent publica­ tion of results. Thirdly, methods of estimating national income have been perfected to an extraordi­ nary degree with the aid of greatly enriched statistical data. As a result, the estimates have gained greatly in substance and reli­ ability. They have been extended in an everincreasing number of countries to cover all the three phases of national income or prod­ uct—its origin in the various industrial divi­ sions, its distribution to the various factors

The last thirty years or so, and particularly the decade following World War II, have witnessed an extraordinary growth of na­ tional income estimating and analysis in all parts of the world. As the next few decades are likely to bring even further develop­ ments in the field, a study of national in­ come estimates would seem to be both timely and important. An examination of the origins of national income estimating, its evolution from rather crude beginnings to a fairly advanced analytical tool and planning aid, its present structure and uses, and the possibilities for its further improvement must be rewarding to anyone concerned with the future of economic analysis and measurement. Such an examination is attempted in this book. The history, theory, and practice of national income estimates are surveyed on a world-wide scale. The subject matter is treated both functionally by topics and re­ gionally by examination of work done in various countries. A uniform presentation has been the objective throughout.

1. Nature of Recent Developments The development of national income esti­ mates since World War I and even more since World War II has been characterized, Notes begin on p. 515. 3

4

THE INCOM E OF NATIONS

of production, and its final expenditure or disposition for various types of consumption and investment. Current estimates embrace a wide range of aggregates differing from one another in the degree of their “gross­ ness” or “netness” or in their functional or other characteristics, but otherwise inter­ related in most ways. Suffice it to mention among them national income at factor cost, national income at market prices, gross na­ tional product, personal income, and dispos­ able income. Margins of error in the esti­ mates have been reduced considerably as a result of the greater availability of direct statistical data on the various components of national income and the consequently lesser dependence on substitute data, guesses, and conjectures. In the fourth place, the usefulness of na­ tional income estimates as tools for eco­ nomic analysis has been greatly enhanced through the development of subsidiary or associated double-entry social accounts that show the incomes, expenditures, and sur­ pluses or deficits of the various sectors of the national economy—namely of its busi­ ness sector, personal or household sector, government sector, and international or “rest of the world” sector. The usefulness of the estimates has also been reinforced by the development of the input-output analysis of national production. National income esti­ mates together with the subsidiary social accounts and input-output breakdowns of the flow of production have come to be used everywhere as the most important in­ struments for analyzing the operations of the national economy and for formulating both short- and long-range fiscal and eco­ nomic policies and programs. In the fifth place, the development of an­ nual series of national income estimates has made it possible to measure the current achievements of the national economy in comparative terms, i.e., as rates of increase in total production and in its various com­ ponents from one year to another. At the same time, the accumulation of historical data on prices prevailing at different periods has made possible meaningful comparisons of the national income of a country over fairly prolonged periods of time. In the sixth place, with the spread of national income estimates throughout the world and with the establishment of the United Nations, the Organisation for Euro­ pean Economic Co-operation, and other in­ ternational political and economic organiza­

tions, international comparisons of national incomes have assumed fresh importance. New techniques for making such compari­ sons have been devised, particularly with a view to overcoming the distorting effects of foreign exchange rates and the disparities in the price and product structures of different countries. In the seventh place, with the annual publication of estimates by government and with extensive discussion of them in news­ papers and journals, the concept of national income has gained wide currency in many countries. Its main substance has come to be reasonably well understood by the lay­ man even though its specific details may still escape him. The professional economist, no matter what his specialization, cannot prac­ tice his art competently nowadays without a knowledge of current changes and some notion of potential changes in the magni­ tude and structure of the income of his country and of their implications for the eco­ nomic welfare of the nation. Nor can public officials participating in the formulation of public policies discharge their responsibili­ ties properly without some knowledge of these facts. An understanding of them is also required of the businessman, farmer, and responsible labor leader, inasmuch as their activities are dependent in a thousand ways upon the operations of the national economy; and these, in turn, are most clearly revealed in the current estimates of national income. In the eighth place, as a reflection of these changes in both professional and lay attitudes, national income theory and anal­ ysis have come to occupy a central position in modern economics. Schumpeter aptly called national income analysis the “aggre­ gative analysis,” while Frisch gave it the name “Macroanalysis.” 1 They were both emphasizing that this type of economic analysis deals with large aggregates—total income, total consumer expenditure, total investment, total wages, and the like—rather than, as did the traditional, purely theo­ retical and nonstatistical economics, with merely representative single units—the be­ havior of the representative business firm, the worker, the consumer, the “economic man,” etc. Economics texts have been rewritten in recent years to give greater prominence to national income theory and analysis; and some of the newer texts have even used it as a starting point for the treatment of the whole subject of eco­

IM PO RTANCE OF NATIONAL INCO M E ESTIM ATES

nomics. Graduate courses in national in­ come have been introduced in the curricula of the economics departments of many uni­ versities. The literature on national income has grown by leaps and bounds. Biblio­ graphical references to current publications, published by the International Association on Research in Income and Wealth, have become numerous enough to fill an annual volume.

2. Organization of This Volume: Its Historical Part The present volume attempts to survey and appraise this global development. It attempts to do so both in its historical con­ text and in its current theoretical and prac­ tical phases. The volume consists of four parts and a statistical appendix: Part I, presenting the history of national income concepts and estimates from their beginning to the present; Part II, surveying the cur­ rent theoretical and methodological prob­ lems in the field and the solutions so far provided for them; Part III, reviewing in detail the estimates of thirteen selected countries illustrative of different concepts, estimating methods, and different types of economies; Part IV, presenting the recent developments and current status of national income estimation and analysis in 66 other countries; and the Statistical Appendix giv­ ing the most important national income sta­ tistics for some 87 countries for the period 1950-55. The tables prepared by the various national income estimators, recently as well as in the past, are reproduced in the form in which they appeared in the original docu­ ments. No attempt is made to correct any unaccountable discrepancies that may ap­ pear in their figures, or otherwise to alter them. The historical part, consisting of ten chapters, reviews the development of na­ tional income concepts and estimates over the past three hundred years, from the time William Petty first originated the idea and prepared the first estimate. It shows how during these three centuries, through the efforts of many pioneering economists and statisticians, the various conceptual and methodological estimating problems have been solved, until the basic concepts have become much clearer and more relevant and the estimates have become more mean­ ingful, reliable, and useful.

5

An attempt is made throughout this whole historical account to examine each new idea and each new major estimate as it relates to the political, social, economic, and cultural conditions, and to the state of development of economics and statistics of its country at that time. Not only is each successive major estimate described in fair detail as to its underlying concept, meth­ odological approach, sources of data, forms of presentation, and results, but attention is paid also to the personality of its author, his motivations, competence, and methods of work. An attempt is also made to indi­ cate the various political, social, economic, and other factors that have influenced the development of national income estimates, sometimes positively, sometimes negatively, and to appraise their relative importance. The primary purpose of this historical account is to help the reader toward a better understanding of the conceptual and prac­ tical problems that still beset the prepara­ tion and interpretation of national income estimates, and, perhaps, even to provide in­ formation of value to present and future estimators and analysts.

3. The Theoretical and Methodo­ logical Part The theoretical and methodological part of the book contains ten chapters. The first six are primarily theoretical and the second four primarily methodological. However, it is not always easy to distinguish between theoretical and methodological problems. For example, are the problems of inter­ temporal and international comparisons of national income theoretical or methodo­ logical? Insofar as they are concerned with the identification of incomparable items in such comparisons, they are theoretical; in­ sofar as they deal with the ways of over­ coming these incomparabilities, the prob­ lems are methodological. The two sets of problems must be dealt with jointly, and are in the final analysis both theoretical and methodological. Similar considerations apply as well to many other problems dis­ cussed in these ten chapters. Chapter 11 deals with basic concepts. It distinguishes between economic and non­ economic production, and between eco­ nomic and noneconomic welfare, inasmuch as the one is reflected in or measured by

6

THE INCOME OF NATIONS

national income while the other one is not. The discussion seeks to show that the con­ cept of income embodied in national in­ come is different from that incorporated in the concept of individual income: that it includes certain items, such as imputed in­ comes, not included in the concept of indidividual income, and, on the other hand, excludes others ordinarily included in the latter; in a word, that it is a social concept —a measure of the social product—and not a mere aggregate of individual incomes. The text shows national income as a flow of products and of money incomes, and it shows that this flow is circular and has three phases—that of production, that of distribution, and that of expenditure or disposition. Chapter 12 defines in considerable detail the area of production covered under each of the two leading modern concepts of na­ tional income—the comprehensive produc­ tion concept of modern liberal economics and the restricted material production con­ cept of Marxian economics. It shows what items of economic activity and income are included and what excluded in each case, and examines critically the reasons lying behind their inclusion or exclusion. Chapters 13 and 14 deal with the distinc­ tion made between final goods available for consumption and investment and interme­ diate goods (raw materials, etc.) used up in the production of the former—a distinction that must be made if double counting of the same goods is to be avoided. One of these two chapters treats this problem as it manifests itself in the market economy; the other, as it arises in the case of government. The latter chapter analyzes also the treat­ ment of taxes and government transfer pay­ ments (subsidies, social security payments, interest on government debt, etc.). Chapter 15 analyzes the various income aggregates, showing their interrelations, uses, and relative importance, and reviews the modern social accounts and input-output analyses and their uses. Chapter 16 deals with the problems of intertemporal comparisons, i.e., largely with conversions of national income series ex­ pressed in current prices into figures calcu­ lated at constant prices; and with the prob­ lems of international comparisons arising out of the disparities in foreign exchange rates, price and product structures, require­

ments of the people, and many other com­ plicating factors. Chapter 17 analyzes general methodo­ logical problems. It identifies the three basic methods of estimating national income—the net output method, the income-distributed method, and the final expenditure of prod­ ucts method—and shows under which con­ ditions the use of each is preferable. It also treats certain problems met in handling sta­ tistical data that are common to all three, such as the problem of interpolation or extrapolation of bench-mark data for cer­ tain years to others for which no direct data are available. Each of the last three methodological chapters— 18, 19, and 20—is concerned with one of these three methods. The prob­ lems of classification of the elements of the national income peculiar to each estimating method are examined, as are the statistical sources employed, and the ways of handling these sources or bridging the gaps in avail­ able data.

4. Individual Analyses of Thirteen Selected Countries’ Estimates Part III is largely a case book. In twelve chapters (21 to 32) it presents detailed analyses of estimates for thirteen countries. These estimates are all of comparatively recent date. Most of them are presented as prepared during the postwar period; a few as they were constructed immediately be­ fore World War II. Each estimate is pre­ sented as far as possible with all its attend­ ing background.

5. Developments in 66 Other Countries Part IV presents a panoramic survey of the recent developments in national income estimation and analysis in sixty-four coun­ tries other than the thirteen whose national income estimates were surveyed in detail in Part III. It examines the developments in each one of them but touches only upon the high spots in each case. Chapter 33 deals with the five English-speaking coun­ tries, other than Canada and Ireland cov­ ered in Part III; Chapter 34— with eighteen European countries not dealt with in the country chapters; Chapter 35—with twenty

IMPORTANCE OF NATIONAL INCOME ESTIMATES

Latin American and Caribbean countries, other than Jamaica and Brazil, treated in Part III; Chapter 36 covers fifteen countries in the Near, Middle, and Far East, other than India and Malaya; and Chapter 37 concerns itself with eight countries on the African continent for which national in­ come estimates have been recently devel­ oped, other than the Union of South Africa treated with the English-speaking group in

7

Chapter 33 and Egypt treated with the Near Eastern countries in Chapter 36.

6. Statistical Appendix The statistical appendix gives the prin­ cipal national income aggregates for some 87 countries in their respective currencies for the period 1950-55. The various parts of the text throughout are interrelated by means of copious cross references.

Part One. History

1. DEVELOPMENT OF CONCEPTS The concept of national income was first formulated in the seventeenth century by Sir William Petty in England and by Pierre le Pesant Sieur de Boisguillebert in France. It was a natural extension to the nation as a whole of the well-established custom in those countries of comparing the wealth and power of individuals of substance or rival­ ing houses of nobility in terms of their yearly incomes. From the very beginning, this broader concept took on the specific meaning of the monetary value of the na­ tion’s annual production and consumption. This meaning was further extended, in time, to include annual additions to the nation’s capital. The development of national income esti­ mates, since Petty prepared the first one in England nearly three hundred years ago, has depended as much on the formulation of meaningful concepts of national income as on the accumulation of reliable statistics and the development of effective statistical tech­ niques. The development of meaningful con­ cepts, in turn, has depended on the formu­ lation of rational theories of economic pro­ duction, inasmuch as national income is merely a monetary expression of national product. Several theories of economic pro­ duction, and hence of national income, have been advanced during these past three hun­ dred years, but of them only three have sur­ vived: the comprehensive production con­ cept, which includes among the objects of production both material and immaterial goods (i.e., commodities and services); the restricted material production concept, which includes only materia! goods; and the re­

stricted market production concept, which includes both material and immaterial goods, but only to the extent that these are produced for the market.

1. The Sequence of the Main Ideas It is sometimes stated that the material production concept antedates the compre­ hensive one, but this is not true. The seventeenth-century founders of modern political economy, England’s William Petty and France’s Sieur de Boisguillebert, conceived of production and national income in com­ prehensive terms as embracing both goods and services. The later restricted material production concept of Adam Smith suc­ ceeded in temporarily crowding the compre­ hensive production concept off the scene, leaving some historians with the impression that it had come first. The third theoretical survivor, the restricted market production concept, is an altogether recent development and is of comparatively minor importance. The course of development of these con­ cepts of production and national income, like the development of thought on other important economic questions, was not al­ ways smooth. There were many rough spots, blind detours, and frustrating periods of little or no progress. The three-centurieslong development of thought can be roughly divided into six periods or stages, each one characterized by a different notion of the nature of production and national income. The first of these was the mercantilistic doc­ trine of the sixteenth century, which viewed national wealth as a stock of gold and pro­

Notes begin on p. 515.

11

12

THE INCO M E OF NATIONS

vided no clearly discernible notion of na­ tional income. In the late seventeenth cen­ tury, Petty and Boisguillebert formulated their more rational theory that viewed wealth as a stock of both consumable goods and the means of their production, and that also originated the concept of national in­ come as a flow of goods and services. Theirs was a truly comprehensive production con­ cept that defined production as the creation of all useful things whether material or im­ material. However, their theory reached only a limited audience and never influenced the official architects of economic policy of their time. The third period was marked by the mideighteenth-century development in France of the physiocratic doctrine that narrowed the definition of production considerably. Ac­ cording to this theory, agriculture was the mother of all production and, in fact, the only truly productive activity, for it alone was capable of producing a “revenu net” to producers. National income, according to this doctrine, consisted wholly of the “net product” of agriculture. The fourth stage of development saw the emergence of the the­ ory of material production first formulated by Adam Smith. This doctrine dealt the final blow to mercantilism at the same time that it exposed the glaring fallacies of the physiocrats. It swiftly gained popularity throughout the Western world, and for the next hundred years or so served as the basis for most national income estimates, com­ pletely overshadowing other notions of production and national income. In the last quarter of the nineteenth cen­ tury, the fifth and sixth theoretical develop­ ments occurred. They paralleled rather than succeeded each other and were representa­ tive of two completely different trends of thought. One involved a repudiation of the Smithian viewpoint and a return, instead, to the earlier comprehensive production con­ cept of the seventeenth century. The other, authored by Karl Marx, involved reformu­ lating and significantly extending Smith’s original restricted material production con­ cept. The first of these theoretical posi­ tions was quickly accepted by all economic schools except the Marxian, and in the twentieth century was adopted as the con­ ceptual base for national income estimates in all countries but those dominated by Marxist philosophy. The second conceptual apparatus, reworked in Marxian philosophy

and language, was absorbed into the official dogma of all Marxist countries and was adopted as the base for their national in­ come estimates. In the mid-1930’s another doctrine put in a brief appearance. Formulated by the Polish economists Kalecki and Landau, and the Hungarians Matolcsy and Varga, this re­ stricted market production concept did not emerge from the main stream of develop­ ment and was not related to either of the major trends. It did not attract any follow­ ing to speak of in Hungary or Poland and seems to have made no more lasting impres­ sion on economic thought in other coun­ tries.1 Finally, after World War I, and particu­ larly since 1940, a new, dynamic concept of national income came into being as a result of the influence of J. M. Keynes and his followers. Under this concept, national income, in addition to being a measure of past or current national production, was also seen as a forecast of or program for future national production. Thus, at the end of several centuries of effort we are left with just two main and one subsidiary concept of national income, and we are using them for new dynamic purposes. The merits and demerits of the two main concepts were actively debated throughout the nineteenth century in West­ ern Europe, but in the present century they are, to all practical purposes, no longer a subject of discussion. All economists with the exception of the Marxists have agreed on the soundness of the comprehensive pro­ duction concept. Disagreement among them is limited to details of its application, i.e., to the treatment of certain items of income or product. Among the Marxists there is a similar unanimity of judgment as regards the soundness of the restricted material pro­ duction concept, but this unanimity, at least in Soviet Russia and some of her satellites, is more required than voluntary. The rea­ sons for the widespread acceptance of the comprehensive production concept and the limited appeal of the material production concept cannot be understood by a simple examination of the logic supporting each viewpoint. They can be understood only in reference to the historical forces that have helped to shape the two concepts. With this broad panoramic survey to pro­ vide background we can now explore each of the conceptual developments in detail.

1 . D E V E L O P M E N T OF CO NCEPTS

2. The Mercantilist Concept The mercantilist school of economics, which reached its zenith in England and France in the sixteenth and seventeenth centuries, had a well-defined though errone­ ous concept of national wealth, and no no­ tion whatever of national income. As ex­ pressed by John Hales (d. 1571), Thomas Mun (1571-1641), and Josiah Child (16301699) in England, and by Jean Bodin (1530-1596), Barthelemy de Laffemas (1545-1611), and Antoine de Montchretien (1576-1621) in France, mercantilist doc­ trine held that overseas trade and shipping were the two most productive occupations because they were the most capable of in­ creasing the country’s stock of gold. Gov­ ernment was an important agent for in­ creasing the nation’s wealth through the controls it was able to exercise over exports, imports, and gold movements as well as over its domestic production and trade. In­ directly, by emphasizing the inherent unity of a national economy and the importance of centrally directed economic effort, the mercantilists laid the foundations upon which their successors could build a con­ cept of national income.

3. Origination of National Income Concept by Dissenters from Mercantilism Petty and Gregory King in England, and Boisguillebert and Seigneur Sebastien le Prestre de Vauban in France, the first social scientists to advance the notion of national income, broke away from the mercantilist school of thought and constructed fiscal and economic theories and programs for their countries far in advance of their times. Petty’s tracts and Charles Davenant’s sum­ mary of King’s national income estimates appeared in England a few years before the publication of Boisguillebert’s work and some ten to sixteen years before the ap­ pearance of Vauban’s in France. It is quite possible, therefore, that the two Frenchmen had read the works of Petty and of Charles Davenant and were influenced by them. Boisguillebert’s second book, Factum de la France, contains definite evidence of his having read one or both of the English works.2 However, it is also possible that Boisguillebert and Vauban developed their particular ideas of national income entirely

13

independently of Petty and Gregory King. William Petty (1623-1687) was the true originator of the concept of national in­ come. At least no mention of this concept by any earlier political economists has been discovered so far. In his Verbum Sapienti, thought to have been written in 1665 but not published until 1691, four years after his death, and in his Political Arithmetick, believed to have been written in 1676, and similarly published posthumously,3 Petty defined the “income of the people” as the sum of the “annual Value of the Labour of the People” and of the “Annual Proceed of the Stock or Wealth of the Nation,” antici­ pating the modem distinction between labor income and capital income. He identified capital income as the sum of rent and other income from ownership of real and per­ sonal property, interest, and profits. He in­ cluded the services of civil and military government officials in national income as well as the services of professions and other occupations. He also defined national in­ come as the sum of the “Annual Expense of the People” and of the surplus remain­ ing after the expense, laying down the broad propositions that “where a people thrive, there the income is greater than the ex­ pense,” 4 and that “what we call the Wealth, Stock, or Provision of the Nation, being the effect of the former or past labour. . . . ” 5 In estimating this national income—a thing never done before—he sought to measure the extent of the nation’s economic power and welfare, or as he put it, “to show . . . that the King’s subjects are not in so bad a condition as discontented Men would make them” and also to show “the great effect of Unity, industry, and obedience, in order to the Common Safety, and each Man’s particular Happiness.” 6 The one possible fault in his construction of national income theory—which was corrected by his imme­ diate successor, Gregory King—was an in­ adequate treatment of the sources of in­ crease of the nation’s wealth. Gregory King (1648-1712), who was more a statistician than an economist, was the first truly scientific estimator of national income. In his Natural and Political Ob­ servations and Conclusions upon the State and Condition of England (dated 1696, but circulated only in manuscript form for the next century except for a summary of it made public by Charles Davenant in 1698), King used the terms “annual income of the nation,” “annual expense of the nation,”

14

THE IN CO M E OF NATIONS

and the “yearly increase in the nation’s wealth.” 7 He first prepared separate esti­ mates of per capita income, expenditure, and savings for each social and economic class in England, totaling these to obtain a single national income aggregate. In this way he obtained an estimate of the distribu­ tion of national income as well as estimates for each of the three items surveyed. His computations were, in effect, a national bal­ ance sheet foreshadowing the social accounts of today. (See Chapter 2, Section 2, for a fuller discussion of King’s work.) King’s concept of national income was broader than Petty’s and his statistical estimates and analyses of national income were much more elaborate and precise. In many re­ spects they were on a par with those devel­ oped in the twentieth century. The comprehensive production concept of national income thus advanced by Petty and King became the basis of a whole series of national income estimates made by other explorers of the subject in eighteenth-cen­ tury England (see Chapter 3). Boisguillebert (1646-1714) and Vauban (1633-1707) did for France what Petty and King did for England: namely, intro­ duced the concept of a measurable national income, provided a broad and rational defi­ nition of economic production, and initiated the first estimates of their country’s national income. The two of them must be given equal credit for this pioneering work, with the distinction, however, that Boisguille­ bert’s contributions were mainly theoretical, whereas Vauban’s were principally statis­ tical. In his Detail de France, written some time in the early 1690’s and published in 1697, and in his Factum de France, published in 1707, Boisguillebert laid down his distinc­ tion between national income (“revenu na­ tional”) and income of the state (“revenu du roi”), identifying the latter as a deriva­ tive of the first, and defining production, national income, and national wealth in the broadest possible terms. Thus, he postu­ lated the “basic principle” that “consump­ tion and income are one and the same thing”; that “consumption is the foundation of all wealth,” inasmuch as “all the riches of the world, whether belonging to the sov­ ereign or to his subjects, are useless if they are not consumed”; and that “the growth of the national income is proportional to the progress of consumption.” 8 Wealth, he wrote, is merely “the power to procure for

oneself the means of a comfortable life as much in luxuries as in necessities.” 9 It con­ sists of a “complete enjoyment not only of the necessities of life, but also of all super­ fluities, including all things catering to the pleasures of the senses which the continued corruption of the human heart invites and refines from date to date.” 10 In a primitive economy in which land is sterile there are but three or four occupations, but in an advanced country, he wrote, “there is a multitude of them, beginning with those of the baker and the tailor who produce neces­ sities and ending with that of the actor who represents the last word in luxury and su­ perfluity,” for the task of the actor is “to flatter the ears and to satisfy the spirit often by a fanciful tale which everybody knows has no relation to reality; and only a people who feel perfectly secure in their necessities can afford to buy pleasant lies. Just as in times of opulence one reaches for the serv­ ices of actors as the highest mark of luxury, so when misery strikes a people their serv­ ices are the first to be dispensed with in an effort at a retrenchment of expenditures.” 11 In proclaiming the services of actors to be a mark of a high level of national income, Boisguillebert was much closer to twentiethcentury thought than was Adam Smith, who, guided by his Scottish philosophy of thrift and austerity, declared a hundred years later that actors were unproductive laborers. Boisguillebert further asserted as a basic principle that “all occupations in a given country, whatever may be their nature . . . work for each other and maintain them­ selves reciprocally” and that “once having become a part of the social body in the course of historical evolution, they can no longer be disjoined or separated without causing a fundamental alteration in the whole social body.” 12 Thus, with a clarity unequaled in any economic writing of the next hundred years or more, Boisguillebert expressed the concept of the internal unity of the economic system and the interde­ pendence of its various parts. Naturally, he classified agriculture as the primary branch of production but he did not claim that it was the only productive occupation. In this respect he was wiser than the physiocrats were half a century or so later. But he made one possible mistake: unlike some of his British contemporaries he did not sufficiently emphasize the importance of saving and capital formation in achieving greater na­

1 . D EV E L O PM E N T OF CO NCEPTS

tional prosperity. This failure was probably due to Boisguillebert’s overriding desire to improve the economic status of the lower classes forthwith rather than to increase the wealth and power of his country in the long run. The most urgent economic problems of the time arose out of the intolerable tax burden that was ruining production and plunging the common people into ever greater misery; and it was this condition that he sought to ameliorate. While identifying national income with a flow of goods and services, Boisguillebert also defined it as a flow of money incomes. He believed that national income consisted of two approximately equal parts: (a) in­ come from property (“revenu des fonds”), such as land, houses, mills, toll houses, revenue-producing public offices, and mon­ eyed capital, in other words incomes from rent, interest, profits, and fees; and (b ) income from “industry,” meaning income from labor. Among the latter he included the income of peasants, artisans, factory workmen, petty tradesmen, and professional people.13 The similarity between Boisguille­ bert’s and Petty’s ideas is striking, and, as already stated, may possibly have been due to Boisguillebert’s acquaintance with Petty’s work. Vauban in his book, Dime Roy ale, pub­ lished in 1707, spoke of national income as the “income of the Kingdom” (revenu du royaume), similarly treating it as the source from which the tax revenue of the King was derived; and he defined it just as broadly as Boisguillebert did, as being com­ posed of incomes from agriculture, com­ merce, industry, and all other occupations.14 Moreover, he was the first in France to at­ tempt to give national income a definite statistical expression (see Chapter 4, Sec­ tion 2).

4. The Physiocratic Concept of Production and of National Income The physiocrats of eighteenth-century France, as represented by Frangois Quesnay (1694-1774), Mercier de la Riviere (17201793), Comte de Mirabeau (1715-1789), and Dupont de Nemours (1739-1817) ad­ vanced a concept of the nature of produc­ tion, national income, and wealth that was much narrower than Boisguillebert’s. First of all, it was a materialistic concept, defin­

15

ing national income and wealth as an aggre­ gate of consumable commodities only. Sec­ ondly, it treated agriculture as the only truly productive occupation by virtue of the fact that it was the only one returning more to producers than their investment of capital and labor, actually yielding them a net in­ come (“produit net”) that represented the free contribution of nature. This notion of net income, later to play such an important role in economic theory as “net rent,” was first conceived by the physiocrats. As to the nonagricultural occupations, the physiocrats took the position that although these were not necessarily useless, they were sterile in the sense that they did no more than return their costs. The physiocratic doctrine did not develop in intellectual isolation from its economic, political, and social environment. It was, rather, a logical expression of its time and place. First of all, France was then the rich­ est agricultural nation on the European con­ tinent and its most prosperous inhabitants were landowners who regarded themselves as the most productive members of society. Secondly, France’s industry and overseas commerce had been ruined by a series of unsuccessful wars, while her agriculture had gone unscathed. As a result, the French economists, in their debates with the Eng­ lish, whose country was growing ever more prosperous industrially and commercially, tended to emphasize their nation’s superior­ ity in agricultural pursuits. Most of the physiocrats were themselves landed pro­ prietors, albeit the “left wingers” of that class, and their outlook was unmistakably that of the agriculturist. One of the most important contributions that the physiocrats made to the concept of national income was Quesnay’s construction of the “Tableau Economique.” This “Tab­ leau” was a “model” demonstrating the flow of national income in the economic system. The flow was shown to begin with the farm­ ers and their workers (the productive class), then to reach the landowners (the control­ ling and distributive element in the econ­ omy), and finally to end with the people engaged in the remaining nonagricultural activities (the sterile class). In thus tracing the flow of national income, the physiocrats emphasized the great role that capital played in its formation. At that time large capital investments in improvements of agricultural land, livestock, and agricultural implements were markedly raising the productivity of

16

THE INCOM E OF NATIONS

French agriculture. By emphasizing the role of capital in the formation of national in­ come, the physiocrats made a new and last­ ing contribution toward a sounder and more realistic concept of national production and national income. The physiocratic concept of national in­ come provided the basis for the famous estimate made by the scientist Antoine Laurent Lavoisier (1743-1794) at the order of the French National Assembly in 1791, which decapitated him for this and other services three years later (see Chapter 5, Section 4, below).16

5. Italian Economists Going Their Own Way: Galiani, Verri, and Palmieri In Italy, during the eighteenth century, quite independent of the main currents of economic thought that swept through Eng­ land, France, and the rest of Europe during that period, a very vigorous development of economic theory took place. When one reads the writings of the Italian economists of that century, whose breadth of thought and brilliance of exposition was not sur­ passed by any of their English or French contemporaries, one is even further con­ vinced that Adam Smith led economics astray. Unlike Smith, these economists traveled the main road of the comprehen­ sive production concept as it had been laid out by the philosophers of antiquity and scholasticists of the Middle Ages, and had been most clearly mapped out by Petty and Boisguillebert. They did not wander down the blind alley of Smith’s material produc­ tion concept nor were they detoured by the appeals of physiocratic doctrine. They rejected the physiocrats as vigorously as Smith, and even earlier than he did, without making his mistake of reapplying the dis­ tinction between productive and unproduc­ tive labor in a new and equally misleading manner. They declared all labor resulting in the production of useful and desired things to be productive, without insisting that these things take on material form; and they identified national income as the value of the annual production of all such com­ prehensively conceived useful and desired things. Inasmuch as the physiocrats held manufacturing in particular contempt, the Italian economists concentrated on defend­ ing it, but were ready to apply the same

defense to other types of nonagricultural activity as well. a. Galiani Ferdinando Galiani (1728-1787), the highly gifted and learned Neapolitan abbe and one-time (1765-1769) Secretary of the Neapolitan Embassy in Paris, rose to the defense of manufacturing as early as 1769. He did this in his Dialogues sur le Commerce des Bles (Dialogues on the Trade in Grain), some twelve years after Quesnay had published his article on the subject in the French Encyclopaedia, in which he fully outlined his economic doctrine. Written and published in French, and acclaimed by Vol­ taire as combining the virtues of the best works of Plato and Moliere and as being as readable as the best of novels, Galiani’s Dialogues achieved great popularity among the members of French society because of its elegant style and the clarity and charm of its exposition. In taking this position in opposition to the physiocratic doctrine, Galiani merely extended the views on the nature of pro­ duction he had expressed in a treatise on money, Della Moneta, published in Italy eighteen years before (1751) when he was only twenty-three years old. In this treatise he defined wealth as the “possession of any­ thing desired more by others than by its possessor.” In a manner reminiscent of Bois­ guillebert’s writings of more than a century before, with which he must have been ac­ quainted, Galiani declared that “among such useful and desired things man is most useful to other men” and that “then come the foods used for consumption, the clothes, the habi­ tation, and lastly, all those comforts that satisfy the secondary pleasures of men.” 16 In his Dialogues, Galiani took issue not only with Quesnay’s advocacy of the repeal of the prohibition of the exports of grain, but with his general economic theory as well. Galiani sought to demonstrate that manufacturing, far from being a “sterile occupation,” was actually more productive than agriculture.17 Galiani’s versatile interests and lively spirit eventually carried him away from economics into the fields of philosophy, po­ litical theory, literature, and practical pub­ lic administration. b. Verri Pietro Verri (1728-1797) covered a much wider field of economics than Galiani

1 . D EV E L O PM E N T OF CO NCEPTS

and concerned himself more closely with the nature of production and the sources of national economic prosperity. The son of a wealthy and eminent member of the admin­ istration of Milan, which was then a part of the Austrian Empire, Verri was well edu­ cated; and developed an early interest in scholarly pursuits. After brief military serv­ ice, he entered the civil administration of Milan where he apparently won the respect of Prince Kaunitz (1711-1794), the prac­ tical head of the Austrian government, one of the greatest European diplomats of the time, and a patron of science and the arts. In the course of his concern with public affairs, Verri became deeply involved in the exploration of basic questions of commer­ cial policy of benefit to his state of Milan and in inquiries into the broader problems of economics generally, and proceeded to develop his views in print. Together with a group of other gifted young men, Milanese patriots like himself, he founded a literary and scientific society and launched a peri­ odical, 11 Caffe, modeled after the English Spectator. In that publication appeared many articles of lasting importance on a variety of scientific and literary subjects. In 1771, Verri published his Medittazioni Sull’ Economia Politica 18 in which he ad­ vanced a definition of production and na­ tional income that was broader and, in some respects, better than that Adam Smith was to provide in his Wealth of Nations five years later. Verri, like Galiani, did not limit production to the creation of material things. He wrote: “In every nation the in­ habitants consume annually not only food, but also clothes, and furniture, and every­ thing that can be of service to men. The sum of the values of these things constitutes the nation’s annual consumption. Similarly, every nation produces annually by its agri­ culture, the use of animals, and work of artisans, such things as grain, clothes, fur­ niture, etc. The sum of the values of these things constitutes the nation’s annual re­ production. . . . When a nation produces more than it consumes, its wealth increases . . . if it consumes more than it produces, its wealth decreases.” 19 Anticipating Adam Smith, Verri held that the manufacturer or artisan receives in the price of his product more than a mere re­ imbursement for his outlays for raw ma­ terials, labor, and his own consumption; and that this surplus constitutes an addition to the nation’s wealth. Refuting the physio­

17

cratic doctrine of the sterility of manufac­ turing production, he maintained that the nation was wealthier after such production than it was before; and cited as proof the fact that the “artisan, if he is able and in­ telligent, changes his level of living; and if he cannot achieve this improvement for him­ self does so for his children.” He contrasted the artisan’s situation with that of the peas­ ant, who generally “is compensated for his work only by the amount he consumes” and “seldom ends his days any richer than he began” and who, in fact, “may descend from twenty generations of equally poor peasants.” Verri further asserted that all wealth has its origin in labor, and that pro­ duction is not the creation of matter but merely the modification of existing matter to suit human needs. He wrote: All the phenomena of the universe, whether produced by the laws of physics or by the hand of man, do not result in creation but only in a change of matter. To combine and to separate are the only elements involved in the concept of production.

Refuting the physiocratic designation of manufacturing as a sterile occupation, Verri wrote: Just as production of wealth and value takes place when the seed, the earth, the air and the water turn into grain, so it takes place when by the hand of man the inside of an insect becomes velvet or pieces of metal are organized into something else. Entire cities and states live off the production of this so-called sterile class, whose production includes the value of the raw material, proportionate consumption of the labor used plus the profit of whosoever under­ takes the manufacture.20

c. Palmieri A few years later, in 1787, Giu­ seppe Palmieri (1721-1794), distinguished scholar, military figure, and fiscal admin­ istrator in the Neapolitan government, pub­ lished a treatise on public welfare, Pubblica Felicita, which, barring the works of medi­ eval scholasticists, was a notable precursor of the welfare economics of the twentieth century. In this work, and in one on na­ tional wealth, Della Ricchezza Nazionale, which followed it in 1792, Palmieri ad­ vanced his thoughts on the nature of pro­ duction and the sources of national eco­ nomic prosperity, which, like Verri’s, were unique for the time. Palmieri had read Adam Smith, whose book was available at

18

THE INCOM E OF NATIONS

the time, but his contact with Smith’s ideas did not divert him from his own original and broad approach.21 All the prominent Italian economists of this period, including Gian R. Carli (17201795) and Cesare Beccaria (1738-1794), advocated freedom of enterprise and, within certain limits, a governmental policy of laissez faire. Verri was particularly out­ spoken in favor of this policy.22 The interest that all these Italian econo­ mists expressed in the nature of production, national income, and wealth, however, never went further than a definition of basic prin­ ciples. None of them was interested in sta­ tistics and none attempted to measure the size and structure of his state’s production and national income. Owing to the barrier of language, their influence, unfortunately, did not spread far beyond the limits of Italy.

6. Adam Smith’s Material Produc­ tion Concept Adam Smith (1723-1790) was greatly influenced by physiocratic thought. He stud­ ied it closely during his prolonged sojourns in Paris and his numerous visits with its leading exponents. His great book published in 1776, Wealth of Nations, shows the ear­ marks of this influence. He acknowledged his indebtedness to the physiocrats when, in reviewing their doctrine and pointing out some of its errors, he concluded nonethe­ less that “this [physiocratic] system, how­ ever, with all its imperfections, is, perhaps, the nearest approximation to the truth that has yet been published upon the subject of political economy . . .” (Book IV, Chap­ ter IX, p. 642).23 Adam Smith adopted the physiocrats’ dis­ tinction between “productive” and “unpro­ ductive” labor, merely modifying it to in­ clude in the “productive” category all labor engaged in the production of material goods. He criticized the physiocrats for classifying manufacturing, trade, and transportation as sterile occupations, insisting that these other branches of “material goods production” were just as capable of returning a net in­ come to the producers as was agriculture, this net income taking the form of profits instead of net rent. Adam Smith derived from the physiocrats a large part of his analysis of capital formation (including the breakdown of capital into operating and fixed, and the concept of the reproduction

of capital). He also borrowed from them a part of his theory of wages. He was also influenced by the physiocrats in his development of a distinction between net and gross national income. The physio­ crats had fumbled the distinction by con­ fusing net agricultural rent with net national income. Smith introduced a broader and more realistic distinction that became firmly established in later economic theory. He defined the gross income of a nation as the aggregate value of all products created dur­ ing the period of a year (gross of the costs of raw materials and other duplications), and the net income as the value remaining after deductions of these duplicated costs. Thus Adam Smith wrote: The gross revenue of all the inhabitants of a great country, comprehends the whole produce of their land and labour; the neat revenue, what remains free to them after deducting the ex­ pense of maintaining; first, their fixed; and secondly, their circulating capital; or what, without encroaching upon their capital, they can place in their stock reserved for immediate consumption or spend upon their subsistence, conveniences and amusements. Their real wealth too is in proportion, not to their gross, but to their neat revenue.24

In this particular passage Adam Smith did not mention the additions to capital as a part of net national income. But in other passages he did refer to them, and, in fact, emphasized that such additions constituted the main source for increasing national pro­ duction and property.25 In a famous and often quoted passage, Adam Smith laid down his distinction be­ tween what is and what is not productive labor, as follows: There is one sort of labour which adds to the value of the subject upon which it is bestowed: there is another which has no such effect. The former, as it produces a value, may be called productive; the latter, unproductive labour. Thus the labour of a manufacturer [i.e., manu­ facturing worker] adds, generally, to the value of the materials which he works upon, that of his own maintenance, and of his master’s profit. The labour of a menial servant, on the con­ trary, adds to the value of nothing . . . the labour of the manufacturer fixes and realizes itself in some particular subject or vendible commodity, which lasts for sometime at least after that labour is past. . . . The labour of the menial servant, on the contrary, does not fix or realize itself in any particular subject or vendible commodity. His services generally per­ ish in the very instant of their performance,

1 . D EV E L O PM E N T OF CO NCEPTS

and seldom leave any trace or value behind them, for which an equal quantity of service could afterwards be procured.26

In accordance with this broad thesis, Adam Smith classified as productive labor­ ers those engaged in agriculture, manufac­ turing, commerce, and the transportation of goods, while rating as “unproductive” labor­ ers the whole civil and military personnel of government, the professions, the domes­ tics, and others engaged in the performance of personal services and the services of dwellings. He considered the national prod­ uct to be constituted solely of commodities, and the national income (or “neat revenue”) to be composed of wages, rent, and profit (including interest) derived from the pro­ duction of these articles (Book II, Chapter II). The wages and salaries, profits, and other forms of income earned in the render­ ing of services were not a part of the “neat revenue” of the nation. They were a deriva­ tive revenue drawn from the revenue cre­ ated by the productive labor, and repre­ sented merely an expenditure on the part of productive labor. Services of dwellings, just as any stock of goods used for imme­ diate consumption, are also not a part of the national income, inasmuch as these services afford no revenue or profit.27 Adam Smith viewed the economic process not as the circulation of consumable and investible goods and income, but wholly as the circulation of entrepreneurial fixed and operating capital. The entrepreneur’s ad­ vances of wages and other operating ex­ penses are returned to him, and profits are created for him in the process of produc­ tion, thereby insuring the continuation of the process. Only “productive labor” has this capacity of reproducing its own value and creating profit. Labor engaged in the rendering of services does not reproduce its costs or produce profits. Although such labor “has its value and deserves its re­ wards,” it represents the consumption and not the production of wealth. The “neat revenue” is spent by laborers and masters “upon their subsistence, conveniences and amusements” and, implicitly, on taxes.28 In this way part of the original “neat revenue” is redistributed as the incomes of the un­ productive laborers. Finally, within the framework of his definition of “productive labor,” he maintained that a nation is richer to the extent that it spends more of its in­ come on durable commodities and less on

19

perishable things, and also to the extent that it saves more of its income for the accumu­ lation of capital (Book II, end of Chapter III). While materially contributing to the clari­ fication of the concept of national income in some respects, Adam Smith was also re­ sponsible for introducing new sources of confusion that were to plague economists for many years to come. Thus, he errone­ ously concerned himself with wealth pri­ marily in terms of a “stock of goods,” seriously neglecting its aspects as a “flow” of utilities, i.e., of national income. In this respect his analysis was a backward step— backward from the physiocrats, and back­ ward from Petty and his other predecessors more than a hundred years earlier. Adam Smith’s differentiation between productive and unproductive labor on the basis of the materiality or immateriality of its product, i.e., his restriction of the concept of pro­ duction to material objects, viewed in a broad historical perspective, was also a major error. It was, perhaps, more an error in terminology than in fundamental thought, but it certainly was most unfortunate in its consequences. Adam Smith’s followers ac­ cepted his distinction literally, but when Karl Marx breathed new life into it over a century later, he gave it at the same time a wholly different meaning that was de­ structive of the ethical foundations of capi­ talism. Smith’s introduction of the restricted material production concept, in the light of the influence it exerted over the next cen­ tury, can be said to have been a serious interruption in the logical development of economic analysis begun by William Petty and to have retarded the growth of national income theory and practice. Speaking of Adam Smith’s contribution to economic theory generally, Schumpeter wrote that his “chief task was to combine and develop the speculations of his French and English contemporaries and predeces­ sors” but that he had not done this job as well as had, say, Turgot in France and Beccari in Italy; and he added: “The blame is at his door for much that is unsatisfactory in the economic theory of the subsequent hundred years, and for many controversies that would have been unnecessary had he summed up in a different manner.” 29 This criticism may be just as aptly applied to Adam Smith’s contribution to the theory of national income. Adam Smith’s book achieved instant pop­

20

THE INCOM E OF NATIONS

ularity not only in England and in the United States, but also in other lands in which it appeared in translation. His con­ cept of production was adopted by most political economists of the so-called “liberal” school of the succeeding half a century or more, rapidly displacing both the mercantilistic and physiocratic doctrines. It par­ ticularly suited the economic thinking of a period characterized by the rapid develop­ ment of manufactures and by the emphasis on the possession of a large supply of ma­ terial goods as the basis of national pros­ perity. Smith’s doctrine of production, wealth, and income was adopted by Ricardo, Malthus, James Mill, John Stuart Mill and, with some qualification, by R. D. Baxter in England; by Sismondi, Daire, and Baudrillart in France; by Hufeland, Jakob, Kraus, Rau, and Neuman in Germany;30 and most of the national income estimates prepared during the succeeding three quarters of a century were based on it—those of Chal­ mers, Colquhoun, Lowe, Pebrer, Spackman, and in part, Baxter in England; Chaptal, Moreau de Jonnes, and Block in France; and Tucker in the United States.

7. The Return to the Comprehen­ sive Production Concept Some criticism of Smith’s material con­ cept of production appeared in England, France, and Germany soon after the pub­ lication of his book. It was made by econo­ mists who, on the whole, enthusiastically accepted most of the other parts of Smith’s “liberal” economics; and their critical com­ ments grew more numerous and telling in time. The first to criticize the doctrine were James Maitland, Lord Lauderdale (17591839) in England,31 and Germain Garnier (1754-1821), translator of Smith’s work, in France. They were joined by Heinrich Storch (1766-1835), eminent member of the St. Petersburg Imperial Academy of Sci­ ences, whose works (1806-1827) translated into both French and German had a wide reading public throughout the Western world;32 and by J. B. Say (1767-1832) 33 in the later editions of his work, Charles Ganilh (1758-1836) 34 and Destutt de Tracy (1754-1836),36 in France. Next, in the second and third quarters of the nine­ teenth century, the doctrine was subjected

to further devastating criticisms at the hands of J. R. McCulloch (17 89-18 64) 38 and N. W. Senior (1790-1864) 37 in England; by Antoine Auguste Walras (1801-1866),38 Pellegrino Rossi (1787-1848),39 and Charles Dunoyer (1786-1863) 40 in France; and by F. B. W. Hermann (1795-1868) 41 and Wil­ helm Rosher (1817-1894) 42 in Germany. All these writers conceived of production as serving consumption, and of national product as comprising both commodities and services. The distinction Smith drew between durable and nondurable utilities, though valid for other purposes, appeared to them to be wholly irrelevant to the issue of what constitutes production and what does not. Smith’s followers were unable to provide any effective answer to these criticisms and, for the most part, remained silent; or else conceded in the end (as did R. Dudley Baxter in England) 43 that Adam Smith was mistaken in this part of his theory. Alfred Marshall (1842-1924), in his Eco­ nomics of Industry, published in 1879, and in his later treatises, championed the com­ prehensive production concept, and by the force of his authority influenced all modern economists (Marxists excepted) to adopt it. At the same time, Marshall also confirmed the importance of distinguishing between “gross” and “net” national income and of avoiding double counting. Thus, he wrote: Everything that is produced in the course of a year, every service rendered, every fresh util­ ity brought about is a part of the national in­ come. Thus it includes the benefit derived from the advice of a physician, the pleasure got from hearing a professional singer, and the enjoy­ ment of all other services which one person may be hired to perform for another. It in­ cludes the services rendered not only by the omnibus driver, but also by the coachman who drives a private carriage. It includes the services of the domestic servant who makes or mends or cleans a carpet or a dress, as well as the results of the work of the upholsterer, the mil­ liner, and the dyer. We must however be careful not to count the same thing twice. If we have counted a carpet at its full value, we have already counted the values of the yarn and the labour that were used in making it; and these must not be counted again. Suppose however a landowner with an an­ nual income of £10,000 hires a private secre­ tary at a salary of £500, who hires a servant at wages of £50. It may seem that if the in­

1 . D E V E L O P M E N T OF CO NCEPTS

comes of all these three persons are counted in as part of the net income of the country, some of it will be counted twice over, and some three times. But this is not the case. The land­ lord transfers to his secretary, in return for his assistance, part of the purchasing power de­ rived from the produce of land; and the secre­ tary again transfers part of this to his servant in return for his assistance. The farm produce, the value of which goes as rent to the landlord, the assistance which the landlord derives from the work of the secretary, and that which the secretary derives from the work of the servant are independent parts of the real net income of the country; and therefore the £.10,000 and the £500 and the £50 which are their money measures, must all be counted in when we are estimating the income of the country. But if the landlord makes an allowance of £500 to his son, that must not be counted as an inde­ pendent income; because no services are ren­ dered for it. . . .44 In his later, more comprehensive work he elaborated upon the same subject as follows; The labour and capital of the country, acting on its natural resources, produce annually a certain net aggregate of commodities, material and immaterial, including services of all kinds. The limiting word “net” is needed to provide for the using up of raw and half-finished com­ modities, and for the wearing out and depre­ ciation of plant which is involved in produc­ tion: all such waste must of course be deducted from the gross produce before the true or net income can be found. And net income due on account of foreign investments must be added in. This is the true net annual income, or rev­ enue; or, the national dividend: we may, of course, estimate it for a year or for any other period.46 By this time, the Smithian concept of material production appeared to be in full retreat. Its fallacies had become apparent to most economists who otherwise accepted Smith’s theories. Among them were: in Eng­ land (in addition to Marshall) 46 Edwin Cannan (1861-1935);47 in Germany, Adolph Wagner (1835-1917);48 in France, Alfred de Foville (1842-1913), Pierre Paul Leroy-Beaulieu (1843-1916) and C. Colson (1853-1939). A number of national income estimates based upon the comprehensive production concept appeared toward the end of the century. Among them were those of Wil­ liam Smart (1899) in England, and Louis Wolowski (1871), Alfred de Foville (1891), and C. Colson (1899) in France, and Pokrovsky (1897) in Russia. M. G. Mulhall (1836-1900) in his Dictionary of

21

Statistics (London, 1883) and in his Indus­ tries and Wealth of Nations (London, 1896) presented national income estimates for 22 countries based on the comprehensive pro­ duction concept (see Chapter 9, Section 7). The only important figure in modern eco­ nomics who still adhered to Smith’s concept of production by the second quarter of the twentieth century and who built his national income estimates upon it was the Hun­ garian economist and statistician, Friedrich von Fellner. But he had only a few fol­ lowers—mostly in Czechoslovakia, Yugo­ slavia, and Hungary. Today, so far as avail­ able information would indicate, no esti­ mates are being made on the basis of the Smithian concept.

8. Marxian National Income Theory a. Revival of the Smithian Material Pro­ duction Concept Under a Marxian Im­ print Just as Smith’s material production con­ cept was about to be cast out, it suddenly received powerful support from a wholly unexpected quarter: Karl Marx (1818— 1883), the founder of modern socialism, in his posthumously published books, Das Kapital, Vol. II (1885), and Theories of Surplus Value (1904), offered a new ver­ sion of Smith’s concept. Its kinship to the original was not fully apparent at first, for it was cloaked in different language and Marx’s followers vigorously denied its re­ lationship to Adam Smith’s idea, claiming that it was entirely original. The kinship is undisputable, however, and was recognized by Marx himself in his more elaborate treatment of the subject in Theories of Surplus Value.*9 Karl Marx went to great lengths in this book to prove that Adam Smith was entirely correct in his particular distinction between productive and unpro­ ductive labor and that all his critics were in error. He merely pointed out certain in­ consistencies and irrelevancies in Smith’s concept, claiming credit only for eliminating them from its otherwise sound substance. No economist in the nineteenth century did as much as Marx, in his then unpub­ lished manuscript, to defend the Smithian doctrine against its critics, or, as he called them, “detractors.” Under Marx’s aegis the Smithian doctrine of material production

22

THE INCOM E OF NATIONS

gained a new lease on life. More than that, it became an ideological cornerstone of the socialist movement—a use that Adam Smith could scarcely have anticipated. Marx was the first to point out that Adam Smith’s distinction between productive and unproductive labor contained two unrelated ideas—one that viewed labor as a supplier of profits and capital to the entrepreneur; and the other that considered labor in terms of its production of utilities for consump­ tion. He maintained that only the first view of labor was meaningful for economic anal­ ysis of the operations of the capitalistic sys­ tem, and that in following the second no­ tion, Adam Smith was in error. He wrote: Adam Smith’s adversaries have neglected his first explanation (based on the profitability or non-profitability of labor) in order to seize upon his second (based on the nature of the use value of the product), whose contradictions and non-sequiturs they have emphasized. In order to polemicize against them the more comfortably, they have insisted upon the ma­ terial form of labor and above all upon the fact that labor must fix itself in a more or less durable product.50

Thus, Marx denied that the form of the product—material or immaterial—has any­ thing to do with the distinction between productive and unproductive labor. The fact that a commodity has a more lasting use value than a service, he insisted, does not make the labor involved in its manufacture productive. He wrote: The concrete character of the labor, and therefore of its product, do not, as such, play any part in this division of labor into produc­ tive and unproductive. Chefs and waiters, for instance, are productive laborers in the sense that their labor is converted into capital for their employer. . . . That labor is productive which produces capital. . . . The use value of a commodity in which the labor of a produc­ tive laborer is realized may be of the most trifling nature. The material result has no rela­ tion at all with this quality [of materialization of productive labor] which is simply the ex­ pression of a social condition of production. The result has its origin not in the content nor in the produce of the labor but in its deter­ mined social form.51

In Theories of Surplus Value, Marx gave full credit to Adam Smith for having been the first to introduce the proposition that under the capitalistic system only that labor is “productive” which produces profits and capital for the entrepreneur, and he criti­ cized Smith only for not adhering to this

definition more consistently and for not drawing from it all pertinent conclusions. Recognizing in that manuscript, written in the 1860’s, that services as well as goods can be produced under a capitalistic system in a way to produce profits and capital for the entrepreneur, and could, therefore, be “productive labor,” Marx chose, nonethe­ less, in all his subsequent writings to asso­ ciate productive labor with the creation of material goods alone. Adam Smith may have committed this error unwittingly, not realizing that some services were already being conducted capitalistically, but Marx did so knowingly. He understated the ex­ tent to which services were conducted capi­ talistically in his day (passenger services of railroads and steamship companies, and communication services of telegraph agen­ cies). He wrote: . . . in the case of all the performances of artisans, actors, professors, doctors, parsons, etc. . . . capitalist production is very limited and occurs only in certain spheres. In institutes of learning for example, the professors can be mere wage workers hired by the entrepreneur. Such factories of learning are numerous in England. Vis-a-vis the entrepreneur, they are productive laborers, although they are not so vis-a-vis the students [Marx did not entertain much respect for the learned institutions and professions]. The director exchanges his capital for their labor power and enriches himself by this operation. The same is true of theatrical and concert producers . . . an actor is a pro­ ductive laborer to his producer. But all these phenomena of capitalist production are insig­ nificant compared with the whole. We can therefore disregard them altogether.52

Marx chose to disregard them, because he could develop his thesis of the materiali­ zation of surplus value into capital more easily in the case of commodities than in the case of services. In defending Adam Smith’s material concept of production against the criticisms of the economists who expounded the comprehensive concept of production, Marx wrote: “The commodity is the most elementary form of bourgeois wealth. To say [as Adam Smith did] that that labor is productive which produces commodities is therefore responsive to the even more elemental point of view that labor is productive only if it produces capital.” 53 It is significant that the followers of Marx never bothered with the fine distinctions he drew in Theories of Surplus Value (between the materialization of the surplus value in the product and the materialization of its

1 . D EV E L O PM E N T OF CO NCEPTS

use value), which permitted the inclusion of services in “production.” To them “ma­ terial production” meant the production of material goods at all times, and it alone deserved to be called “production.” At any rate, the material production con­ cept as Marx revised it became firmly im­ bedded in socialist theory, and provided, more than half a century later, the basis of all national income estimates prepared in countries dominated by the Marxian doc­ trine.

b. Marx’s Theory of Value Closely associated with the reformulated doctrine of material production was Marx’s theory of exploitation of labor in which he definitely departed from Adam Smith. The values of all commodities, he insisted, were merely “congealed labor.” Profits, including interest, were merely the “surplus value” created by labor and appropriated to him­ self by the entrepreneur; they were the wages withheld by the entrepreneurs from the workers.54 The only productive labor was that of the worker. The entrepreneurs were the parasitic class. National income, super­ ficially consisting of wages, profits, interest, and rent, was in reality income created by the laboring class alone. Where the physio­ crats elevated the cultivator to the sublime position of the creator of all wealth and sup­ porter of all other occupations in the nation, Marx elevated the industrial laboring class to this lofty position, thrusting a banner of revolution into its hand at the same time. Unless one understands Marx’s revolution­ ary political program, one cannot fully un­ derstand the reasons for his adopting the materialist concept of production.

c. Marx’s Recognition of the Correspond­ ence Between Income and Product Marx improved on Adam Smith in recog­ nizing the identity of the income and prod­ uct aggregates. He defined the “newly pro­ duced value” (value added) in income terms, as the sum of wages, profits, and rent, while defining it in product terms as the sum of consumer goods and net investment goods. He identified gross value as the sum of wages, profits, rent, and capital replace­ ments and also as the value of the new product plus the value of capital replace­ ment. In line with this, he recognized that national income can be measured by either the product or the income method, i.e., by

23

taking the gross value of the total product less material expenses and capital replace­ ment or by aggregating the incomes derived from production. Finally, Marx also noted that replace­ ment of capital is never a part of income, but always flows back into the economic system in the form of capital. The following quotations from his Das Kapital show his line of reasoning: The value of the annual product in com­ modities, just like the value of the commodities produced by some particular investment of capital, and like the value of any individual commodity, resolves itself into two parts: Part A, which replaces the value of the advanced constant capital, and Part B, which presents itself in the form of wages, profit and rent. . . . Part A . . . never assumes the form of rev­ enue . . . [it] always flows back in the form of capital, and of constant capital at that . . . In order to avoid useless difficulties, it is necessary to distinguish the gross output and the net output from the gross income and the net income. The gross output, or the gross product, is the total reproduced product. With the exception of the employed but not con­ sumed portion of the fixed capital, the value of the gross output, or of the gross product, is equal to the value of the capital advanced and consumed in production, that is, the constant and variable capital plus the surplus-value, which resolves itself into profit and rent. Or, if we consider the product of the total social capital instead of that of some individual capi­ tal, the gross output is equal to the material elements forming the constant plus variable capital, plus the material elements of the sur­ plus product, in which profit and rent are ma­ terialized.55

Finally, he restricted the term “net in­ come” to the net return of the entrepreneur or, in other words, to surplus value. He criticized Adam Smith for applying the term “net income” to the sum of wages, profits, and interest as being inconsistent with the philosophy of the capitalist system that looks to the production of a “net in­ come” for the capitalist only. He denied that society in the capitalistic system was get­ ting any “net income” or that the produc­ tion of it for society was the goal of capital­ ist production.

d. Marx’s Distinctions Between Gross and Net Product Marx drew a clearer and more elaborate distinction between gross national income and net national income than did Adam Smith. His terminology was sometimes con­

24

THE IN CO M E OF NATIONS

fusing, but his theory was consistent. H e distinguished between “the value of the p roduct” and “the new value” or the “newly produced value” (the equivalent of w hat is nowadays generally spoken of as “value added” ). T he “value of the product” ac­ cording to him was the sum o f “the new value” and the replacem ent of used-up capi­ tal (today generally term ed “depreciation” ). Thus, the “value of the product,” in his term inology, was equivalent to w hat we call today “gross national p ro d u ct” (G N P ). A t the same tim e, M arx distinguished between w hat he called “gross product,” w hich included all the duplicated values of raw materials, supplies, and other “inter­ m ediate products” (although he did n o t use the latter te rm ), replacem ent of used-up capital, and the “new value” produced du r­ ing the year, w hich he identified as consist­ ing of wages, profit (including in te rest), and rent, which he called, som ew hat m islead­ ingly, “gross incom e of the society,” and w hich we call “net national incom e.” The gross income is that portion of value and that portion of gross product measured by it, which remains after deducting that portion of value and that portion of the total product measured by it, which replaces the constant capital advanced and consumed in production. The gross income, then, is equal to the wages (or to that portion of the product which is to become once more the income of the laborer) plus the profit plus the rent. On the other hand, the net income is the surplus-value, and thus the surplus product, which remains after the deduction of the wages, and which, in fact, represents the surplus value realized by capital and to be divided with the landlords, and the surplus product measured by it. . . . There is only this difference between the product of the individual capitalist and that of society: From the point of view of the individual capitalist the net income differs from the gross income, for this last includes the wages, whereas the first excludes them. Viewing the income of the whole society, the national income consists of wages plus profit plus rent, that is, of the gross income. But even this is an abstraction to the extent that the entire society, on the basis of capitalist production places itself upon the capitalist standpoint and considers only the income divided into profit and net rent as the net income.56 In the latter portion of this statem ent, M arx m isinterpreted the point of view “of the entire society” in a capitalistic system. N either at the tim e he w rote n o r subse­ quently did th a t society place itself “upon the capitalist standpoint” on the point at

issue and consider “n et profit and n et rent as . . . n et incom e.” C apitalist society as expressed in the writings of its leading econom ists always drew a distinction be­ tw een the private incom e of the capitalists and th e social incom e. T he society viewed its incom e then, as it does today, as con­ sisting of wages, interest, and rent as well as profits. This m ay n o t represent the true “social dividend” in the sense of a net gain in econom ic w elfare, b u t bare entrepreneu­ rial profits are even less representative of such a gain an d are n o t identified w ith the social dividend by any responsible m em bers o f capitalist society.

9. The Contrasting Treatments of the Two Concepts in the Present Century Today, th e division o f allegiance between the two m ajor production concepts is draw n along political lines. A ll nations and econo­ mists w ho are n o t follow ers o f M arx’s doc­ trine subscribe to the com prehensive p ro ­ duction concept and build their national incom e estim ates on its broader, m ore ra ­ tional foundation. W ith the organization of the U nited N ations after W orld W ar II, and its entrance into the field of national incom e estim ation, the com prehensive production concept achieved the status o f an interna­ tional standard. N ational incom e analysis, organized around the com prehensive production con­ cept, is being constantly refined, re-exam ­ ined, an d broadened to achieve greater theoretical consistency and m ore reliable techniques and to reflect m ore accurately the productive activities of our age. N ew breakdow ns of national incom e are being developed to show n o t only its origin in the various branches o f production and its re­ distribution am ong individuals and groups, but also its expenditure on various types of final consum ption and investm ent products. A series of new incom e aggregates, varying in the degree of their “grossness” o r “netness,” are being introduced to perm it a m ore diversified analysis of th e operations of the national econom y. Finally, a series of “sector” accounts and “input-output” analyses of production has been developed to show the m onetary and p roduct flows in the national econom y and in each o f its parts. A t the sam e tim e greater use is being m ade of national incom e estim ates th an

1 . D EV E L O PM E N T OF CONCEPTS

ever before. Both governm ents and private groups have com e to rely m ore heavily upon them in the form ulation of their production program s and econom ic and financial plans. All these developm ents have occurred within the fram ew ork o f the com prehen­ sive concept during the past half-century. By contrast, the progress m ade during the same period u nder the m aterial production concept in elaborating national incom e theory or in perfecting its use in analysis has n o t been startling. T he concept is essen­ tially no different from w hat it was eighty years ago w hen K arl M arx adopted and altered Sm ith’s original form ulations. I t still defines national incom e as it did then, as the value of annually created com m odities, less duplication and depreciation. T here has been no expansion of the concept, and no new breakdow ns o f national incom e have been introduced. T he breakdow ns are still lim ited to the “net value” of the output of the several branches of m aterial production and to the distinction betw een current con­ sum ption and capital replacem ent and in­ vestm ent, and betw een the socialized and the nonsocialized sectors of the economy. The uses of national incom e estim ates have probably been extended som ew hat in the C om m unist countries in connection with their launching o f long-range plans of eco­ nom ic developm ent, but this is only a sur­ mise, fo r there is not m uch published evi­ dence on the subject. T he quality of the statistics used in th e preparation of the esti­ mates m ust have been im proved consider­ ably, b u t even on this subject, inform ation is sparse.67 T he steady developm ent o f national in­ come analysis in both theory and practice under the one concept, and its practical stagnation under the other, is not inherent in the nature of the concepts. T he differ­ ence is directly related to the different de­ grees o f freedom of thought in the coun­ tries involved. H ad the developm ent of the two concepts proceeded in the reverse m anner, the m aterial production concept taking root in the non-C om m unist countries and the com prehensive production concept in the C om m unist countries, the m aterial production concept w ould probably have evolved into som ething m ore realistic and richer in content, while the com prehensive production concept, in the hands o f the Com munist countries, w ould probably have resulted in as rigid an analytical tool as the other is now.

25

10. The Keynesian Dynamic Approach A ny account o f the development of n a­ tional incom e concepts w ould be incom ­ plete w ithout m entioning the contributions J. M . Keynes and his followers have made. A n appraisal of these contributions is diffi­ cult because they are so recent. A nother fifty years or m ore m ay have to pass before their full im pact on econom ic thought and on national incom e concepts and analysis becomes apparent. In the meanwhile any com m ent on the subject m ust necessarily be guarded. Keynes (1 8 8 3 -1 9 4 6 ) gave a new orien­ tation to m odern econom ics and in so doing also affected m odern thinking in the field of national incom e analysis. H is principal con­ tribution lay in the distinction between variable or strategic factors and related or dependent subsidiary phenom ena; and in a dem onstration th at by altering the size or direction of the first, governm ent and pri­ vate groups m ay alter the size and structure o f the entire national income. Therein lay the substance of his discovery. A m ong the strategic or independent factors, Keynes first included only fo u r: volume of con­ sum ption, volum e o f investm ent (which together w ith the other form s the national inco m e), quantity of m oney (bank de­ po sits), and the interest rate. L ater when, u nder the influence of th e G reat D epres­ sion and W orld W ar II, governm ent finance had expanded trem endously in the economy, Keynes added two m ore strategic factors— governm ent spending and tax collections. H e show ed th a t the volum e of em ploym ent at any given tim e— a m atter of deepest con­ cern to the nation— depended on changes taking place or being m ade in the size and direction o f one or m ore of the strategic factors m entioned. A lthough Keynes him self was not inter­ ested in statistics and was n ot particularly skillful in using them , his followers were. T hey used his theoretical approach in de­ signing various models of national income, each based upon a different assumed volume o f this o r th a t strategic factor and each, accordingly, showing a different size and structure of national income. By selecting the model representing the situation that in their judgm ent was m ost likely to develop or was m ost desirable, K eynes’s followers

26

THE INCO M E OF NATIONS

used these m odels either as forecasts of the future national incom e o r as guides to fu ­ ture governm ental and private group eco­ nom ic and financial policies and program s of action. T his new approach, to w hich Schum peter, Frisch, and others gave the nam e o f “m acro-econom ics” or “econom ic dynam ism ,” has given a new m eaning to the concept of national incom e.58 Historically, the Keynesian approach is linked to the dynam ism of the physiocratic doctrine and of Q uesnay’s Tableau Eco-

nom ique, in w hich the size and structure of the gross national incom e of a pre­ em inently agricultural country was shown to be capable of an infinite num ber o f varia­ tions, depending on the size and m anner of allocation of the “p ro d u it n et”— the one strategic facto r in this type of economy. So far as the fu tu re is concerned, the K eynesian approach paves the way fo r still fu rth er innovations in national income con­ cepts and analysis, the exact n atu re of which it is impossible to predict at this time.

2. DEVELOPMENT OF FIRST ESTIMATES IN ENGLAND IN THE SEVENTEENTH CENTURY Estim ates of national incom e originated in England. T hey first appeared in the seven­ teenth century during one of the m ost criti­ cal periods in E ngland’s history. English­ m en w ere locked in a deadly civil w ar over the w ay in w hich they should be governed; and the shape of the nation’s political life was never m ore uncertain. M oreover, E ng­ land’s position as the w orld’s greatest sea pow er was being threatened by repeated at­ tacks from H olland and F rance. N o w onder her scholars w ere concerned w ith determ in­ ing their country’s resources and its capa­ bilities to survive both its internal and ex­ ternal ordeals. N o t only did the first estimates of E ng­ land’s national incom e appear at an early date but these first estimates, prepared by Sir W illiam Petty and G regory King, had a breadth and sweep to them th a t was not duplicated by any estim ate developed in England for the next century o r m ore.

1. Petty— the Originator of National Income Estimates W illiam Petty (1 6 2 3 -1 6 8 7 ) was the first m an to estimate the incom e of a nation. It was in 1665, in an essay called V erbum Sapienti, that Petty prepared an estim ate of the then current national income of England, following it eleven years later, in Notes begin on p. 517.

an essay called Political A rith m etick, w ith an o th er such estim ate. In b oth essays, he supplem ented the estim ates w ith other sta­ tistical calculations of E ngland’s resources and com pared them w ith those of F rance and H olland. Both these essays w ere circu­ lated privately fo r years and were first pub­ lished only in 1690 and 1691, three and fo u r years after P etty ’s death.

a. Petty’s Part in the Invention of Statis­ tics or “Political Arithmetick” A few years before w riting the V erbum Sapienti, Petty had helped John G rau n t (1 6 2 0 -1 6 7 4 ) to w rite his Natural and Po­ litical O bservations M ade upon the Bills o f M ortality (L ondon, 16 6 2 )— the first sys­ tem atic treatise on vital statistics, w hich was later recognized as th e beginning of the discipline of statistics. By applying the sam e m ethods of quantitative observation and analysis to econom ic phenom ena th at G rau n t had used fo r population changes, P etty was able to enlarge the scope of this new tool. H e called it aptly, “Political A rith ­ m etick,” em phasizing w ith this nam e two things: first, th a t it dealt w ith m atters re­ lating to society; and secondly, th a t it dealt w ith them quantitatively, or, as he p u t it, “in term s of N um ber, W eight or M easure,” rath er th an in “C om parative or superlative W ords.” H e insisted th a t only quantitative

2 . D E V E L O P M E N T OF FIRST ESTIM ATES IN ENGLAND

expression was precise, as it alone dealt w ith observable and verifiable facts. T he nam e th at Petty gave this new science was used fo r the next hundred years until it was displaced near the end of the eighteenth century by the term “Statistics,” derived from the Latin phrase ratio status.1 H e stated th a t “the M ethod I take to do this, [to reason by num ber, weight, and m easure] is n ot yet very usual. . . .” 2 B ut he was convinced th at in tim e it w ould be generally accepted inasm uch as it was in harm ony with N atural Law. In 1662 he played a leading role in founding th e Royal Society, w hich was patterned after th e F ren ch A cad­ em y of Sciences except fo r th e fact th a t it was m ore concerned w ith the N atu ral than w ith the Social Sciences. P etty insisted tim e and again at the sessions o f the Society th at all indefinite generalizations should be avoided in discussion and th a t “no w ord m ight be used but w hat m arks either nu m ­ ber, weight, or m easure.” 3

b. Circumstances That Led Petty to His Discovery P etty’s discovery th a t the incom e o f a nation as well as other phenom ena of its econom ic life could be m easured w as n o t altogether accidental. It was, in a sense, as m uch a product of the tim es as o f a single m an. T h e air of E ngland and W estern E u ­ rope was perm eated w ith the scientific spirit. T he R eform ation, the Revolution w ith its tem porary installation of a R epublic, and other violent changes in E ngland’s govern­ m ent h ad liberated the hum an spirit and enthroned experim entalism as th e best ap­ proach to the solution of m ost h um an p ro b ­ lems. Coincidentally, great im provem ents had been m ade in recording the acts o f pri­ vate individuals and public officials, m aking possible m ore precise observations o f the workings of society. D uring the m iddle of that century, following the plague, P arlia­ m ent passed a law requiring every parish to m ake accurate registration of all births and burials, and following the R estoration, num erous changes w ere m ade in the coun­ try ’s system of taxation which, fo r the first time, provided data on the num ber of houses, the acreage of land, the num ber of livestock, and the consum ption of certain widely used articles of food. D uring the same general period, the first o f a series of navigation laws was passed, establishing tighter governm ental controls over E ng­

27

land’s foreign trade, and incidentally re­ sulting in m ore accurate recording of its volum e an d nature. T he political turm oil of the period presented an opportunity to men o f speculative m ind to use these newly available social and econom ic data to verify the so-called “n atu ral laws” generally be­ lieved to govern th e operations of m ankind, and to form ulate plans fo r the better m an­ agem ent of society’s affairs. W illiam Petty was at the very center of these tu rb u len t developments. A n adven­ turous and bold spirit endow ed with a per­ ceptive m ind, he was peculiarly well suited to interpret these developm ents in scientific term s. H e had acquired a lust fo r science during his youthful years o f vagabondage through W estern E urope and his studies of science and m edicine at the Universities of Paris and Leyden. H e was appointed pro­ fessor o f anatom y at the University of Ox­ ford upon his retu rn to England, b ut he resigned in 1650 to take the job o f physi­ cian to C rom w ell’s arm y and fam ily in Ire­ land. Eventually he was elected to P arlia­ m ent, acting as one of the leaders of C rom ­ w ell’s p arty in the H ouse. Later, he served fo r m any years as chief lan d surveyor and tax adm inistrator of Ireland, occasionally returning to E ngland on business and never failing to becom e em broiled in violent po­ litical disputes. Petty’s interest in m easuring E ngland’s national incom e and in appraising her re­ sources grew o u t o f his experience as tax adm inistrator, his broad concern with the nation’s public affairs, and, strange as it m ay seem, his m edical career. H e sought to apply the techniques of physical science to the field of social science, likening the social body to the hum an body, and the tool of “Political A rithm etick” (statistics) to the scalpel of the surgeon. H e m ade his atti­ tude particularly clear in a w ork called The Political A n a to m y o f Ireland, apparently w ritten in the early 1670’s but only pub­ lished posthum ously with some of his other w orks in 1691.4 In estim ating the national incom e and resources of England, Petty had several objectives in m ind. F irst of all, he w anted to prove m athem atically th a t the State could raise a m uch larger revenue from taxes to finance its peace and w artim e needs, and th a t it could do so by m ore equitable and less burdensom e form s of taxation. This he attem pted to do in the V erbum Sapienti (1665) and in his Treatise

28

THE INCOM E OF NATIONS

o f Taxes and Contributions (1 6 6 2 ). Sec­ ondly, Petty w anted to disprove once and fo r all the notion that England h ad been ruined by the Revolution and foreign wars and was no m atch, either m ilitarily o r com ­ mercially, for H olland and F rance.5 T his he did in his Political A rith m etick (1 6 7 6 ). In the Preface to th a t w ork he w rote: I have . . . thought fit to examin the fol­ lowing Perswasions, which I find too currant in the World, and too much to have affected the Minds of some, to the prejudice of all, viz. That the Rents of Lands are generally fall’n; that therefore, and for many other Reasons, the whole Kingdom grows everyday poorer and poorer; that formerly it abounded with Gold, but now there is a great scarcity both of Gold and Silver; that there is no Trade nor Employ­ ment for the People, and yet that the Land is under-peopled; that Taxes have been many and great; that Ireland and the Plantations in America and other Additions to the Crown, are a Burthen to England; that Scotland is of no Advantage; that Trade in general doth la­ mentably decay; that the Hollanders are at our heels, in the race of Naval Power; the French grow too fast upon both, and appear so rich and potent, that it is but their Clemency that they do not devour their Neighbors; and finally, that the Church and State of England, are in the same danger with the Trade of England; with many other dismal Suggestions, which I had rather stifle than repeat.6

c. Petty’s Estimates P etty’s m ethod of estim ating national in­ com e was simple. H e approached the m eas­ urem ent from the point o f view of con­ sum ption, adopting the basic assumption th a t the annual incom e of a country is equivalent to its “annual expense.” H e was aw are of the fact that in years of peace and prosperity, nations saved p a rt of their in­ com e and added it to the stock of wealth, but he apparently considered this p ropor­ tion to be relatively small. A t any rate he did not attem pt to estim ate it. Starting w ith this m ain assum ption, Petty proceeded to estim ate the expense of the people for “Food, H ousing, Cloaths, and other necessaries,” fixing it at an average of AVid. per diem, o r £ 6 13.?. Ad. p er annum for every m an, w om an, and child. E stim at­ ing the total population of England and W ales at 6 millions, he arrived at a total annual expense fo r the country o f £ 4 0 millions. Next, turning to the incom e side, he estim ated that £.1 5 millions of the am ount spent m ust have com e from prop­ erty. H e arrived at this figure by estim ating

the annual incom e from each category of property as listed in the available records. T hus, he calculated the total acreage of agricultural land and its probable annual rental; the total n um ber of houses in the country, in L ondon, and in the smaller u rb an centers and villages, and their p ro b ­ able annual rentals; incom e from the own­ ership of cattle, from shipping, fro m com ­ m erce, etc. H is breakdow n of the £ 1 5 mil­ lions incom e from p ro p erty was as follows: £ 8 millions from land, £ 2 . 5 from houses, and £ 4 . 5 from other property.7 T h e re­ m aining £ 2 5 millions of the national in ­ com e he attributed to agricultural, indus­ trial, com m ercial, and oth er labor. H e as­ sum ed th a t one h alf of the population earned an average incom e of Id . p er head per day by its labor, exclusive of 52 Sun­ days, 26 holidays, and days of recreation o r illness. T his added up to £ 8 6s. 8d. per year p er capita— yielding the £ 2 5 millions total fo r incom e from labor. H e further divided the laboring class into six num eri­ cally equal categories, the first earning 2d. per diem, the second Ad., the third 6d., the fo u rth 8d., the fifth 10d., and the sixth 12d.8 In th e Political A rith m etick (1 6 7 6 ), Petty am ended this estim ate, raising the an­ nual expense p er head to £ 7 per annum an d “the w hole expence” o r incom e o f the nation (still estim ated at 6 million souls) at £ 4 2 m illions.9 In Political A rith m etick, Petty also of­ fered figures on the population and resources o f F ran ce and H olland, b ut owing to lack of d ata he did n ot estim ate their national incomes. H e p ut F ran ce’s population at 13.5 millions against the com bined popula­ tion of E ngland, Scotland, Ireland, and the P lantations of A m erica of 9.5 to 10 m il­ lions, and, on th e basis of his evaluations o f the productivity of various occupational groups, decided th a t despite her popula­ tion advantage, F ran ce could n o t be any w ealthier th an E ngland and h er territories.10

d. Petty’s Statistical Conclusions O n the basis of his estim ates of E n gland’s incom e and expense and other statistical analyses o f h er resources com pared with the resources of F rance and H olland, Petty advanced a n um ber of generalizations of im portance. In V erbum Sapienti h e dem on­ strated th a t if all property ow ners and p ro ­ ductively em ployed people in England were taxed at 10 p er cent of th e ir incom e, the State w ould obtain an annual revenue of

2 . D E V E L O P M E N T OF FIRST ESTIM ATES IN ENGLAND

£ 4 millions. This sum w ould be am ple fo r peacetim e needs and be considerably in excess of the am ount then being raised. If the tax were raised to 17.5 per cent o f in­ come, the State w ould obtain a revenue of £ .7 millions, enough to take care o f the requirem ents of an extraordinary w ar as well. Petty also m aintained th a t universal taxation proportionate to incom e w ould be m ore equitable and less burdensom e on the econom y than the existing “irregular taxa­ tion” th a t bore heavily on trade w hile leav­ ing other segments of the econom y almost untouched. H e showed th a t proportionate taxation could be realized by using con­ sum ption taxes in addition to taxes on rents, polls, and wages, inasm uch as con­ sum ption was related to incom e, and he ac­ tually calculated the share o f total revenue th a t each type o f tax w ould produce under this system.11 In his Political A rith m etick Petty ad­ vanced “ten Principal Conclusions,” which were, he held, if n o t actual m athem atical truths, a t least “Suppositions” capable of verification in th e fu rth er advance o f “th a t knowledge I am at.” 12 N early all these conclusions w ere refutations of the pessi­ mistic evaluations o f E ngland’s econom ic situation m ade by contem porary writers. These conclusions, to each o f w hich Petty devoted a chapter, w ere presented by him in the following striking form : The Principal Conclusions of This Treatise are, Chap. I That a small Country, and few People, may by their Situation, Trade, and Policy, be equivalent in Wealth and Strength, to a far greater People, and Territory. And particularly, How conveniences for Shipping, and Water Carriage, do most Eminently, and Fundamentally, conduct thereunto. Chap. II That some kind of Taxes, and Publick Levies, may rather increase than di­ minish the Common-Wealth. Chap. I ll That France cannot, by reason of Natural and Perpetual Impediments, be more powerful at Sea, than the English, or Hollanders. Chap. IV That the People, and Territories of the King of England, are Naturally near as considerable, for Wealth, and Strength, as those of France. Chap. V That the Impediments of England’s Greatness, are but contingent and removeable. Chap. VI That the Power and Wealth of England, hath increased above this forty years.

29

Chap. V II That one tenth part, of the whole Expence, of the King of England’s Sub­ jects; is sufficient to maintain one hundred thousand Foot, thirty thousand Horse, and forty thousand Men at Sea, and to defray all other Charges, of the Government: both Ordi­ nary and Extraordinary, if the same were reg­ ularly Taxed, and Raised. Chap. VIII That there are spare Hands enough among the King of England’s Subjects, to earn two Millions per annum, more than they now do, and there are Employments, ready, proper, and sufficient, for that purpose. Chap. IX That there is Mony sufficient to drive the Trade of the Nation. Chap. X That the King of England’s Sub­ jects, have Stock, competent, and convenient to drive the Trade of the whole Commercial World. s o u r c e : William Petty, Economic Writings, edited by C. H. Hull, Cambridge, England, 1899, vol. 1, pp. 247^18.

A t the tim e, it was said th a t the K ing of F rance h ad been offended by these conclu­ sions, w hich w ould indicate th at Petty’s tracts w ere probably read in France. H is essays w ere found to be indiscreet by E ng­ lish officialdom as well, w hich explains why m ost of his essays rem ained unpublished during his lifetime. P etty ’s son, L ord Shelborne, alludes to this censorship in unm istak­ able term s. W hen finally publishing P etty ’s Political A rith m etick, w ith the K ing’s per­ mission, in 1690, L ord Shelborne w rote in his dedication of this w ork “to the K ing’s M ost E xcellent M ajesty” th a t “had n ot the D octrins o f this Essay offended France, they had long since seen the light, and had fo u n d Follow ers, as well as im provem ents before this time, to the advantage perhaps o f M an­ kind.” 13 H e also w rote: “1 presum e to offer what m y Father long since writ, to shew the w eight and im portance o f the English Crow n. It was by him stiled Political A rith ­ metick, in as m uch as things o f G overn­ m ent, and o f no less concern and extent, than the G lory o f the Prince, and the happi­ ness and greatness o f the People, are by the O rdinary Rules of A rithm etick, brought into a sort o f o f D em onstration. H e was allowed by all, to be the Inventor o f this M eth o d o f Instruction; where the perplexed and intri­ cate ways o f the World, are explain’d by a very m ean piece o f Science. . . .” 14

e. Petty’s Generalizations Often Running Ahead of His Statistics Petty was a bold and speculative analyst, whose generalizations often ran ahead of

30

THE INCO M E OF NATIONS

his statistics. H e was aw are of this fact him ­ self and frequently cautioned the reader n o t to accept some of his observations as proved facts, but to treat them as hypotheses deserving fu rth er testing and verification. T he uses to w hich P etty p u t his national in­ com e estimates and other calculations were even m ore im portant than the estimates and calculations themselves, for they encouraged future generations of statisticians to im ­ prove their data and techniques so that their findings would carry greater weight. D r. H ull rightly says about P etty’s economic writings that they . . . exhibit both the strength and the weak­ ness of his characteristic method. When his terms of number, weight and measure result from an actual enumeration they are gener­ ally of value, for he has a considerable capacity of segregating the really significant factors of an economic problem. But the difficulties in the way of enumeration were great, and in his eagerness for results he often resorted to cal­ culations which were nothing more than guesses.” 15

2. Gregory King’s Novel Estimates: A Neglected Milestone In 1696, six years after the publication of Petty’s estimates, a novel estim ate o f E ng­ land’s national incom e and expense was privately circulated am ong a small group of political economists. This m anuscript, entitled Natural and Politicall Observations and Conclusions upon the State and C ondi­ tion o f England, was prepared by G regory King, a hitherto unknow n official in a heraldry office who was also a m apm aker, engraver, and surveyor. C harles D avenant, who happened to see the m anuscript King had m ade available to a few interested per­ sons, imm ediately recognized its value. A p­ parently with G regory K ing’s consent, he m ade copious excerpts from it and incor­ porated them, w ith due acknow ledgem ent to their author, in a tract on the revenue and trade of England that he was then prepar­ ing fo r publication (See Section 3 below ). F o r a century thereafter these excerpts were the only parts of K ing’s w ork available to the general public. As a result, K ing’s esti­ m ates, although very different from Petty’s, never displaced them or exercised as m uch influence upon other students of the sub­ ject. Only after the trac t was published in full by G eorge Chalm ers in 1802, as an appendix to his E stim ate o f the Com para­

tive Strength o f G reat Britain, was the im ­ portance of K ing’s w ork realized by schol­ ars, and its influence on the developm ent of statistics felt.

a. Sources of King’s Interest in Economic Statistics and in National Income Estimates G regory King was in m any ways the op­ posite of Petty. H e was m odest and unas­ suming, m eticulously careful w ith his fig­ ures, and extrem ely cautious in his conclu­ sions. H e allowed the figures to tell their ow n story, adding only a m inim um of generalizations an d expressing these in the tersest possible term s. H e was the prototype o f the m odern statistician. A lthough King was less w ordy th an Petty, his estimates were m uch m ore detailed and, probably, m ore accurate. T his was due as m uch to th e im provem ents m ade in statistical rec­ ords during the intervening qu arter of a century as to the m ore systematic nature of K ing’s m ind. A lthough K ing m akes no m ention of P etty ’s w orks in either of his two tracts (the o ther is titled, O f the N a va l Trade o f E ng­ land a0 1688 and the N ational Profit then arising th ereb y), their influence upon him is unm istakable. H e treats the sam e sub­ jects, often uses the same term s, and arrives at some of the same conclusions. Professor B arnett, editor of the A m erican edition o f K ing’s w orks (1 9 3 6 ), says, “We have no inform ation as to the date at w hich K ing becam e interested in Political A rith­ m etic.” 16 B ut there is no doubt th a t this interest grew o ut o f K ing’s early w ork in h eraldry as clerk to Sir W illiam Dugdale, the fam ous antiquarian and heraldry expert, for this w ork brought him into contact with public records. H is interest was fu rth er stim ulated by later activities as surveyor, cartographer, and finally, as registrar of the College of A rm s and Official of the L ancaster H erald. “It is possible,” says Bar­ nett, “th a t the enactm ent in 1695 of the law im posing a tax on m arriages, births, and burials was the beginning of his interest in the subject,” inasm uch as he acknow l­ edges in his N atural and Political Observa­ tions th a t the assessment of this tax p ro ­ vided a new and im portant source of infor­ m ation “. . . upon the K nowledge of the tru e N u m b er of People in England and such oth er circum stances relating thereunto . . 17 B arnett fu rth er says th at “there is

2.

31

D E V E L O P M E N T OF FIRST ESTIM ATES IN ENGLAND

Table 2-1. A Scheme of the Income & Expence of the several Families of England Calculated for the Year 1688

N um ber of Families 160 26 800 600 3,000 12,000 5,000 5,000 2,000 8,000 10,000 2,000 8,000 40,000 140,000 150,000 16,000 40,000 60,000 5,000 4,000

R anks D egrees Titles and Qualifications

H eads P F am ­ ily

Tem porall Lords Spirituall Lords Baronets K nights Esquires Gentlem en Persons in offices Persons in offices M .rch.ts & T raders b y Sea M .rch.ts & T raders b y Sea Persons in th e Law Clergy M en Clergy M en Freeholders Freeholders Farm ers Persons in Sciences & Lib. a rts Shopkeep.rs & Tradesm en A rtizans & handycrafts N aval officers M ilitary officers

40 20 16 13 10 8 8 6 8 6 7 6 5 7 5 5 5 4H 4 4 4

6,400 520 12,800 7,800 30,000 96,000 40,000 30,000 16,000 48,000 70,000 12,000 40,000 280,000 700,000 750,000 80,000 180,000 240,000 20,000 16,000

SX

2 ,6 7 5 ,5 2 0

N um ber of Persons

511,586 50,000 364,000 400,000 35,000

Common Seamen Labour.8 People & outserv.ts Cottagers & Paupers Common Souldiers

849,000 V agrants

3 3H 3H 2

150,000 1,275,000 1,300,000 70,000

3H

2,7 9 5 ,0 0 0 30,000

849,000

2,8 2 5 ,0 0 0

Y early Income P Fam ily £ 2800 1300 880 650 450 280 240 120 400 200 140 60 45 84 50 44 60 45 40 80 60 £ 67 £ 20 15 6 10 14 10 10 £ s 10 10

Y early InT otall of Ex­ come th e E states pence P or Income H ead p H ead £ 448,000 33,800 704,000 390,000 1,200,000 2,880,000 1,200,000 600,000 800,000 1,600,000 1,400,000 120,000 360,000 3 ,360,000 7,000,000 6,600,000 960,000 1,800,000 2 ,400,000 400,000 240,000

£ 70 65 55 50 45 35 30 20 50 33 20 10 9 12 10 8 15 12 10 10 20 15

£ 60 55 51 46 42 32 27 18 40 28 17 9 8 11 9 8 11 9 9 18 14

34 ,495,800 £ 1,000,000 5 ,460,000 2 ,000,000 490,000

£ s 12 18 £ 7 4 10 2 7

£ 12 £ 7 4 2 7

8 ,950,000 60,000

3 5 2

3 9 3

9 ,010,000

3 3

3 76

So th e G eneral A ccount is £ 511,586 849,000

Increasing th e w ealth of th e Kingd. Decreas.g th e w ealth of th e Kingdom

1,360,586

N e at Totalis

so u rc e :

sx

2,6 7 5 ,5 2 0 2,8 25,000

4Ho 5 ,5 0 0 ,5 2 0

67 10 10 32

If to be well apprized of the true State and Condition of a Nation, Especially in the Two maine Articles of it’s People, and Wealth, be

s

12 18

£

10 10 10 10 10

s 10 12 5 10

s d

12

Totall Increase p annum

£ s d 10 10 4 4 3 2 10 3 2 10 5 3 1 1 1 10 5 1 10 10 10 2 1

64,000£ 5,200 51,200 31,200 90,000 240,000 120,000 60,000 160,000 240,000 210,000 12,000 40,000 280,000 350,000 187,000 40,000 90,000 120,000 40,000 16,000

18 s d 10 2 5 10

2 ,4 4 7 ,100£ Decrease 7 5 ,000£ 127,500 325,000 35,000

£

4

5 62,000£ 60,000

4 6

622,000£

1

£

s d 18

2 ,4 4 7 ,100£

9 ,010,000

3 3

3 76

46

622,000

4 3 ,505,800

7 18

7 11 3

69

1,825,100£

Two Tracts b y G regory King, edited by G. E. B arn ett, Baltimore, 1936,

ground fo r believing th a t K ing’s w ork in Political A rithm etic m ay have led to his em ploym ent in public office,” fo r he spent the last ten years o f his life as Secretary to the Com ptrollers of th e A rm y A ccom pts and as one of the “Com m issioners to state King W illiam’s debts.” In the Preface to his w ork, K ing set down his reasons fo r thinking th a t esti­ mates of the incom e and population of a nation were o f im portance and value, but that such estimates could only approxim ate the truth. H e w rote:

£

34,495,800

10

In­ crease p H ead

p.

31.

a Piece of Politicall Knowledge, of all others, and at all times, the most usefull, and Neces­ sary; Than surely at a Time when a long and very Expensive Warr against a Potent Monarch (who alone has stood the shock of an allyance, & Confederacy of the Greatest part of Christendome) Seems to be at it’s Crisis, Such a Knowl­ edge of our own Nation must be of the Highest concern: But since the attaining thereof (how necessary & disireable soever) is next to im­ possible, We must Content our selves with such near approaches to it as the Grounds We have to go upon will enable us to make. However if haveing better Foundations than heretofore for Calculations of this kind, We have been Enabled to come very near the Truth; Then doubtlesse the following observa­ tions & conclusions will be acceptable to those

32

THE INCOM E OF NATIONS

who have not entirely given up themselves to an Implicit Beleif of Popular Falcehoods. But the Vanity of People in overvaluing their own Strength is so natural to all nations as well as ours; That as it has influenc’d all former Cal­ culations of this Kind both at Home and Abroad, So if these, even these Papers may be allowed not to have erred on that hand, I am of opinion they will not be found to have Erred on the other.18

b. Elaborateness of King’s Estimates King began his estimates with a new “Calculation” of the population of England and Wales, showing its distribution by sex, age, occupation, geographic district (in­ cluding apportionment as between London, other cities, market towns, villages, and hamlets) and annual rate of growth. He arrived at a figure of 5.5 millions, a some­ what lower figure for the total population of England than Petty’s.19 Next, he related the population of England and its rate of growth to the populations of France, Hol­ land, and the world as a whole and the ap­ proximate rates of their growth. Finally, he calculated “the Annual Income and Expence of the Nation as it stood in 1688” (the year of the second English Revolution and one immediately preceding the outbreak of war

with France), and the annual excess of in­ come over expense added to the stock of wealth, thus filling an important gap in Petty’s calculations that had omitted the annual savings of the people. King built his estimate by first computing separately for each occupational or social group the number of families, the number of persons in each family, and the average income, expense, and saving, if any, per family. He then compiled an aggregate of these items for each group before adding the group totals together. Thus he obtained not only the total annual national income, expenditure, and saving, but their distribu­ tion by social and occupational groups. The resulting calculation, reproduced in Table 2-1, was remarkable in the breadth of its conception and richness of its coverage. It went far beyond any of Petty’s calculations. King supplemented the results of this calculation with a breakdown of national income by type of income, as Petty had done, following, however, a somewhat dif­ ferent classification. He placed income from the rent of land and dwellings and from other “heriditaments” in one group, while combining income from profits and from labor in another group. He estimated the

Table 2-2. The Annual Income, & Expence of the Nation as it stood An.° 1688 T hat the Yearly Income of the Nation a.° 1688 was T hat the Yearly Expence of the Nation was T hat then the Yearly Increase of wealth was T hat the Yearly Rent of the Lands were about of the Burgage or Houseing about of all other Heriditaments about

43,500,000 Sterlg. 41,700,000 1,800,000 10,000,000 2,000,000 1,000,000 In all 13,000,000

T hat the yearly Produce of Trade arts & Labour was ab.4 30,500,000 In all 43,500,000 T hat the Number of Inhabited Houses being about The Number of Families about And the Number of People about The People answer to 4J4 P House, & 4 p Family. T hat the Yearly Estates or Income of the several Families answer

1,300,000 1,360,000 5,500,000

In common To about And about T hat the yearly expence of the Nation is about And the yearly Increase about

£ s. d. 32 p 7 18 p 7 11 4 p 68p

T hat the whole Value of the Kingdom in Gen.11 is about

650,000,000 Stlg.

Family head head head

£

s o u r c e : T w o T ra cts

by Gregory King, edited by G. E. Barnett, Baltimore, 1936, p . 30.

2.

33

D E V E L O P M E N T OF FIRST ESTIM ATES IN ENGLAND

number of dwellings and assumed the num­ ber of families to be but slightly in excess of the number of dwellings, and each dwell­ ing to contain 4.25 persons and each family 4 persons. He recapitulated the results of his calculations as shown in Table 2-2.

c. Bases of King’s Estimates: All Three Phases Covered King drew heavily on fiscal data in con­ structing his estimates. He found the books of collections of the graduated poll tax par­ ticularly useful because this tax was general in application and fairly well enforced. For each local collection district these books showed the number of individuals in each occupational or social group, their ages, sex, size, and the composition of their families, and gave an indication of the probable range of their income. He also found useful information in the collection records of other taxes such as the temporary levies (called “aids”) imposed on rent from land and houses, profits from trade, and salaries of civil servants; the assessments on mar­ riages, births, and burials; the “hearth tax” that was levied on all homes and gave ac­ count of the number of families; the in­ ternal excises on beer, wine, glass, salt, etc.; as well as the customs duties. The fact that England’s tax system was probably more diversified than that of any other country at the time, and was in all likelihood better administered than that of any other coun­ try, made the tax records exceptionally useful for calculating the country’s national income. For his estimates of national expense and saving, King used nonfiscal information in addition to fiscal data. He employed data on the quantity and value of the principal agricultural crops (wheat, rye, barley, oats, peas, beans, and other vegetables), animal products (milk, cheese, butter, wool, meat, hides, etc.), accretions of live stock (cattle, sheep, hogs, etc.), consumption of beer, etc. The following two tables that he drew up, one on crops and one on the consump­ tion of meat, illustrate the detailed and comprehensive nature of his computations. It will be noticed that in estimating the value of crops, King deducted the cost of seed. In other words, he sought to calculate the net value of output, or as he called it, “the neat Produce.” In modern terms King’s estimate can be said to have covered all three phases of na-

T h e P roduce of th e A rrable L and I th u s E stim a te of B ushells W h eat R ye B arley O a ts Peas B eans V etches & c.a

12 8 25 16 7 4 1

M illions M illions Mill." M ill.8 Mill." Mill." Mill."

p B u sh .11 V alue at at at at at at at

3.* 6.d 2,100,000£ 2.’ 6.d 1,000,000 2.* 2,500,000 1.' 6 * 1,200,000 2." 6.d 875,000 2.» 6.d 500,000 2.* 100,000

73 M illions a t 2.« 3 d

8,275,000

T h is is o nly th e n e a t P roduce Exclusive of the Seed C orn w .oi in som e so rts of g rain being n ear a 4 .th of th e P roduce in o th ers a 5.th m ay in gen­ eral be reckoned a b o u t 17 mill." of Bushells m ore, w hich m akes th e whole produce to be 90 Mill." of B ushells w .ci a t 2.* 3.d p B ush in com m on is F u ll 10 m illions Sterlg. s o u r c e : T w o Tracts b y G regory K ing, edited by G . E . B a rn e tt, B altim ore, 1936, p. 36.

tional income: its production, distribution, and consumption. The first two phases were covered completely, and the third, consump­ tion, somewhat less fully.

d. Estimates of the Country’s War Potential King did not see his job as finished with the calculation of England’s national income and expenditure for one year. He went on to construct an annual series of estimates— the first and only such series to be con­ structed anywhere for the next hundred and fifty years. He not only included past years, but projected the series into the future, an­ ticipating the modern practice of forecast­ ing by some two hundred and fifty years. In his projections, he considered the effects that the war then being waged with France would have on England’s income and re­ sources. In Table 2-4 below, “The state of the Nation An.0 1695,” it can be seen that he estimated that the six-year-long-war had reduced the annual income of the na­ tion from £43.5 to £42.5 millions and was resulting in an annual excess of expenditure over income, i.e., reduction of the national capital, by about £ 3 millions a year.

e. King’s International Comparisons His concern with assessing England’s ability to win the war with France prompted King to calculate the national incomes of France and Holland. Aside from Petty’s fragmentary speculation about them, they

34

THE INCOM E OF NATIONS

Table 2-3. An Estimate of the yearly consumption of Flesh in the Nation

Beeves & Calves Sheep & Lambs Swine & Piggs Deer & Fawns Goats & Kidds Hares & Leveretts Rabbits & Conies

Number of the yearly consumption

W .tof each carcass

800,000 3.200.000 1.300.000

200 lb. w.t 32 46 70 36

20,000 10,000 12,000 2 , 000,000

2K

Price of a pound weight 1. d3.[ l

Price of each beast

Value of the yearly consumption

Weight of the yearly consumption

1 18

1,520,000 960.000 750.000 35,000 4,000 900 42,100

208,000,000 w.t 102,400,000 59,800,000 1.400.000 360,000 30,000 1.700.000

3,302,000 600,000

373,690,000 24,000,000 400,000

6 11 6

1 15 76 16

5

7,342,000 Tame Fowle Wild Fowle

a t 6.d p pound a t 12.d p pound

20,000

In all 3,922,000£

398,090,000 w.t

Which for 5% Mill." of People is, £ s. d. In value!

14 * p ™n ' . /l2 p diem fere

lb oz. In weight I ^ ^ p ann. I 3% p diem.

But for 2,700,000 Persons being the value of those who eat flesh constantly, the fores.d Proportion of 398.090.000 Pound weight of Flesh yearly spent as Food comes to 6% ounces p head p diem & 147 pound weight p head p annum; Besides Dutch Beef, Westphalia Bacon &c.° The remaining 2,800,000 Persons not eating of Flesh being these 200.000 40,000 260.000 1.280.000

Infants under 13 months old. Sick Persons. Part of 700,000 Persons, who feed on fish a t Least 2 days in 7. Part of 1,760,000 Persons contain'd in 440,000 Families who by reason of their Poverty do not contribute to church or Poor, & consequently eat not flesh above 2 days in 7. 1.020.000 Part of 1,200,000 Persons contain’d in 400,000 Families who receive Alms & consequently eat not flesh above once a Week. 2,800,000 s o u r c e : T wo

Tracts by Gregory King, edited by G. E. Barnett, Baltimore, 1936, pp. 38, 39.

were apparently the first calculations ever made of either French or Dutch national income. Unlike Petty, King credited France for the year 1688 with nearly double Eng­ land’s income and Holland for the same year with between one third and one half of England’s. He estimated the annual de­ crease of income and consumption of capi­ tal in France, as a consequence of the war, at about the same proportion, i.e., at about one twelfth part of the income, or £ 6 mil­ lions a year. On the other hand, he estimated an exactly reverse effect of the war on Hol­ land: namely an annual increase of income of about half a million pounds a year, “so that Holland in general is richer than at the beginning of the Warr by 6 million [pounds a year],” or by one third of her prewar in­

come. He constructed the table, “That the Expence of the 3 Nations . . . ,” to dem­ onstrate the effects of the war on the three economies, showing a marked reduction in the levels of consumption for the French, as contrasted with but slight reductions for the English and Dutch (see Table 2-5). He wrote: “. . . That from the year 1688 To 1695 England has decreas’d in People 50,000 France 500,000 And Holland is In­ creas’d 40,000. That England is Decreas’d in its Income a Million, France 10 Millions, But Holland is Increased Half a Million.” 20 The implicit conclusion was, of course, that the people of France were suffering much more severely from the war than the people of England, and that France, there­ fore, was bound to sue for peace sooner.

2 . D E V E L O P M E N T OF FIRST ESTIM ATES IN ENGLAND

35

Table 2-4. The state of the Nation An." 1695 T hat the Present Income of the Nation is a Million lesse than it was A.° 1688 and is 42% Mill." Sterlg. now but about T hat the Yearly Expence is about 40% Mills. & the Taxes 5 Millions In all 45% Mill." Sterlg. 3 Mill." Sterlg. T hat the Kingdome does now yearly decrease T hat if the Warr continue to An.° 1698 Inclusive The Yearly Income will in Prob­ 38% Mill." Sterlg. ability Be But The Expence 3l * Mm.-}In a1142H MaL< SterlgTaxes 4 Mill." Sterlg. The Yearly Decrease According to the Following Scheme

An." 1688 1689 1690 1691

Annuall Income of the Nation 43,500 43,600 43,700 43,800

Annuall Expence of the Nation

Ordinary Revenue of the Crown

Extraord.ry Taxes actually raised

Annuall Expence In all

Increase or Decrease of the Nation

41,700“ 41,500“ 41,500.” 41,400.”

2,000,000 1,800,000 1,800,000 1,700,000

3,000,000 4,000,000 4,000,000

41,700,000 44,500,000 45,500,000 45,400,000

Incr. 1,800,000 Deer. 900,000 Deer. 1,800,000 Deer. 1,600,000

1692 1693 1694 1695

43,800 43,600 43,100 42,500

41,200.” 41,000.” 40,800.” 40,500.”

1,700,000 1,600,000 1,600,000 1,500,000

4,000,000 4,000,000 5,000,000 5,000,000

45,200,000 45,000,000 45,800,000 45,500,000

Deer. Deer. Deer. Deer.

1696 1697 1698

41,600 40,200 38,500

40,100.” 39,300.” 38,500.”

1,500,000 1,400,000 1,400,000

4,500,000 4,500,000 4,000,000

44,600,000 43,800,000 42,500,000

Deer. 3,000,000 Deer. 3,600,000 Deer. 4,000,000

1,400,000 1,400,000 2,700,000 3,000,000

Hence We may Inferr That in 7 years From 1688 To 1695 Inclusive the Taxes have amounted to effectually But th at the Kingdome is scarce actually decreased So that by Industry & Frugality there has been saved full That by the year 1698 Inclusive The Taxes will in 10 Years have amounted to in all probability effectually

29 Mill.8 Sterlg. 13 Mill.8 16 Mill.8 42 Mill.8

And the Kingdome will be actually Decreas’d 23% Millions T hat after the Year 1695 The Taxes actually rais’d will fall short every year more & more to th at Degree, T hat the Warr cannot well be sustain’d beyond the year 1698 upon the Foot it now stands, unlesse 1. 2. 3. 4. 5. 6.

The Yearly Income of the Nation can be Increas’d. Or the Yearly Expence Diminish’d. Or a Forreign or Home Credit be obtain’d or Establish’d. Or the Confederacy be Inlarg’d. Or the State of the Warr alter’d. Or a General Excise, in effect Introduc’d.

Now whereas by the Foregoing Scheme, The Wealth of the Kingdome seems to be actually Decreas’d almost 13 Mill.8 Sterlg. between 1688 & 1695 Inclusive, And will Probably decrease by 1698 Inclusive above 10 Mill." & a half More, In all about 23 Mill." and a half in 10 Years; The said Decrease seems to be thus chargeable. ( C ontinued on p . 36)

Table 2-4. The states of the Nation A n.0 1695 ( C o n tin u e d )

Remain.6 Stock an.° 1695

The Stock of Decrease by the Kingdom the year 1688 1695

Remaining Stock anno 1698 3,000,000 1,500,000

8,500,000 3,000,000 500,000 4,000,000 1,500,000 10,500,000

4,000,000 400,000 1,600,000 500,000 2,500,000

4,500,000 3,000,000 100,000 2,400,000 1,000,000 8,000,000

28,000,000

9,000,000

19,000,000

6,000,000

13,000,000

33,000,000 25,000,000

3,000,000 1,000,000

30,000,000 24,000,000

3,500,000 1,000,000

26,500,000 23,000,000

86,000,000

13,000,000

73,000,000

10,500,000

62,500,000

Coyn’-l Silver Coyn’d Gold Uncovnd Silver & gold Wrought Plate Rings &c.“ Jewells Furniture Apparell &c.“ Stock for Trade Consumption &c.a The Live Stock in Cattle &c.°

Decrease by the year 1698 1,500,000 1,500,000 100,000 1,200,000 200,000 1,500,000

1,200,000 800,000 6,500,000

Hence it follows T hat if the stock of the Nation which was 86 Millions Sterlg. A.° 1688 viz.1 about double to the yearly Income & expence, shall be decreas’d to 62 Mill.8 & a Half by A.0 1698. The Warr cannot well be sustain’d longer than that year for these Reasons. 1. For that the Mony of the Kingdome will then be but 4% Millions viz.* but one Tenth of the annual Expence, Lesse then which cannot circulate the Whole. 2. T hat the Wrought Plate will be little above a Million Consequently nothing to be spar'd further than that article. 3. T hat 7 Mill.8 in Jewells, Household stuff, Furniture, Apparell &c.° is the least quantity We can imag­ ine that Article reduceable unto, The bedding of the Kingdome amounting to one Half of that Summ. 4. T hat if the stock of the Kingdome, in Shipping, Forts & castles and in naval and military stores and appointments, & for Forreign Trade, and home consumption, and all the Branches of that Article be reduc’d from 33 to 26 Millions; If it should be further Lessen’d, the Nation cannot be secure, Trade cannot be carried on, Nor a Sufficient Stock of Provisions left to supply us in times of Difficulty. 5. T hat if the Live Stock of the Nation which will then be Diminish’d a 12.th part, should be further Diminish’d, I t may occasion an excessive Rice of the Price of wool, Leather, Flesh, Butter, and cheese, not much short of a Famine; Unlesse the Number of People Decrease Proportionably; the effect whereof will be equally pernitious. so u r c e ;

Two Tracts b y Gregory King, edited b y G. E. Barnett, Baltimore, 1936, pp. 46-49.

Table 2-5. T hat the Expence of the 3 Nations may be thus proportion’d for the years 1688 & 1695 D y et 1688

Incident Charges

A pparell

Generali Income

Generali Expence 4 1 ,700,000£ 80,500,000 15,750,000

Increase

4 3 ,5 0 0 ,000£ 84,000,000 17,750,000

1 ,8 0 0 ,000£ 3,500,000 2,000,000

Engl.'1 France Holl.d

2 1 ,3 0 0 ,000£ 41,000,000 6 ,4 00,000

1 0 ,400,000£ 18,500,000 3 ,0 0 0 ,0 0 0

1 0 ,000,000£ 21,000,000 6 ,3 50,000

Engl.d France Holl.d

68,700,000 21,000,000 38,000,000 6,20 0 ,0 0 0

32,400,000 10,200,000 16,000,000 2,8 00,000

37,350,000 14,300,000 26,000,000 8,400,000

138,450,000 45,500,000 80,000,000 17,400,000

147,250,000 42,500,000 74,000,000 18,250,000

8,800,000 3,000,000 6,000,000 850,000

D ecrease D ecrease Increase

65,200,000

31,000,000

49,700,000

147,900,000

141,750,000

6,150,000

D ecrease

Lastly as to th e General A ccount of Eng.d France, & HolI.d for th e years 1688 & 1695 I have added this fu rth er Scheme.

1688

1695

so u rc e :

Engl.d France Holl.d

Eng.ld France Holl.d

N um ber of People

Yearly Income p H ead £ s d

P u b lick R evenue & Taxes £

Taxes p Head £ s d

A nnunall C onsum ption besides Taxes £

5,500,000 14,000,000 2,200,000

7 18 6 3 8 14

2,0 00,000 10,500,000 4 ,7 5 0 ,0 0 0

7 3 15 2 3 2

39,700,000 70,500,000 11,000,000

21,700,000

17,250,000 6 ,5 00,000 17,500,000 6 ,9 00,000

£ 1 1 3

15 10 s d 4 5 1 7

121,200,000

5 ,4 50,000 13,500,000 2 ,2 40,000

6 15 9 £ s d 7 16 5 18 8 2 9

21,440,000

6 12

30,900,000

1 8 10

117,000,000

39,000,000 62,500,000 10,500,000

Consum pt.n p H ead £ s d

Annall Increase In all £

7 4 5 9 5

1,800,000 5,000,000 2,00 0 ,0 0 0

5 £ 7 4 4

8,80 0 ,0 0 0

11 s 3 18 13

6 8 7 18 4

£

2 9

D . 3,000,000 D . 6,00 0 ,0 0 0 In. 850,000

5 9 4

D . 6,15 0 ,0 0 0

T w o T racts b y G regory King, edited b y G. E. B arnett, B altim ore, 1936, p . 55.

36

9 d

Increase p H ead £ s d

8 s 11 8 7

1 d Deer. 10 Deer. 7 Incr.

5 9 D e er.

2 . D E V E L O P M E N T OF FIRST ESTIM ATES IN ENGLAND

37

f. Regrettable Neglect of King’s Great Work

a. Exposition of the Uses of the Estimates

It is regrettable that King’s brilliant work went unpublished for more than a century. Had it been published earlier it might have influenced some of England’s economic and political writers of the succeeding century to make similarly comprehensive analyses of the country’s national income and expense and to emulate King in his concern for ac­ curacy.

In his Discourses on the publick revenues and on the trade of England, published in two volumes in 1698, Davenant devoted a whole chapter to an exposition “of the use of Political Arithmetic,” which he defined succinctly as . . the art of reasoning by figures, upon things relating to govern­ ment.” 22 He wrote:

3. Davenant— Popularizer of the Estimates and Pleader for Better Statistics Charles Davenant (1656-1714), an in­ fluential writer and political figure, contrib­ uted a great deal to the advancement of scholarly interest in national income esti­ mates. He was quick to realize the value of King’s estimates, and gave them promi­ nent publicity in his own work. Without his aid, King’s estimates might have remained in complete obscurity for the next century. Davenant was eloquent in pointing out the value of national income estimates, and he pleaded with public officials to keep im­ proving their records so that better and more meaningful estimates of the country’s resources could be prepared in the years to come. Born in comfortable circumstances, Davenant was educated first at Oxford and then at Cambridge, where he received his Doctorate of Laws at an early age. Soon thereafter, he entered upon a public career. Three times elected to Parliament—in 1685, 1698, and 1700—he also held the offices of commissioner of excise (1683-89) and in­ spector general of imports and exports (1705-14). A gifted writer, he published many essays on public revenues, public debt, and trade that were republished later in the eighteenth century as his complete works. His books were widely read in Eng­ land both at the time they first appeared and in later years.21 An ardent admirer of Petty and King, Davenant held a curious mixture of liberal and Tory economic ideas. He was not a pro­ found thinker, but he was a persuasive writer and was read as widely as any other English economic writer of the time.

A great statesman, by consulting all sort of men, and by contemplating the universal pos­ ture of the nation, its power, strength, trade, wealth and revenues, in any counsel he is to offer, by summing up the difficulties on either side, and by computing upon the whole, shall be able to form a sound judgment, and to give a right advice: and this is what we mean by Political Arithmetic.23

In plainer and more convincing words than those used by either Petty or King, Davenant argued the value of estimates of national income and wealth. He particularly stressed their utility in: (a) formulating effective tax policy; (b) preparing plans for the mobilization of manpower and economic resources in time of war; (c) gauging the enemy’s resources during war; and (d) formulating a foreign trade policy. Davenant developed these points as follows: The wealth of all nations arises from the labour and industry of the people; a right knowledge therefore of their numbers, is neces­ sary to those who will judge of a country’s power and strength. . . . For without doubt, it must very much help any ruler to understand fully that strength which he is to guide and direct; since he may thereby know how many are fit for war, what hands support the commonwealth by their labour and industry, and what sort of men are idle and useless in it. And this is so far from being a matter of meer speculation as some think, that very many conclusions may be drawn from thence, useful and reducable to practice. For if the numbers of the people can be truly stated, if they can be divided into proper ranks and classes, if it can be distinguished who are rich, who are easy, who can barely subsist, and who, by reason of poverty, can contribute little in any tax or aid to the public; if all this can be thoroughly understood in the laying any excise or new imposition, a very near computation may be made what such a duty should produce to the government, sup­ posing it to be duly collected, and under a good management.24

38

THE INCOM E OF NATIONS

b. Exposition of Sources, Methods, and Skills Required for the Preparation of Estimates Davenant did more than demonstrate the usefulness of national income estimates; he also showed, to the benefit of future esti­ mators, the sources, methods, and skills re­ quired for their preparation. In doing so, Davenant filled in the gap left by both Petty and King, for neither of them had docu­ mented his figures adequately or revealed his methods in satisfactory detail. More than that, Davenant displayed extraordinary per­ ception in identifying the peculiar features of the British fiscal system that not only enabled England to pioneer in national in­ come estimating, but made it possible for her to continue as a leader in this field for the next two centuries. He wrote; “There are few places which afford better helps for computation than England does at present.” Davenant demonstrated the truth of this by reviewing the kind of information revealed by each type of tax record, as exemplified by the following passage: The excise is a measure by which we may judge, not only of what the people consume, but, in some sort, it lets us into a knowledge how their numbers increase or diminish. The customs are the very pulse of a nation, from which its health, or decays, may be observed. The hearth-money has given us a view, certain enough, of the number of families, which is the very ground-work in such speculations; and these 3 revenues must be the better guide to computers, because the accounts of them are fairly kept and stated, and because the respec­ tive branches have been under so exact a man­ agement, that perhaps their utmost produce is known and understood. The 1st poll-tax, 1 Gul. & Mar. was paid with great alacrity and affection to the govern­ ment, and very few avoided payment; so that from those books we may not only see the numbers, but divide the people into such proper classes and ranks, as may in a manner shew the wealth and substance of the whole kingdom; . . . The aid of 1 & 2s. per pound, 1 Gul. & Mar., which was more carefully collected than any that has been since granted, has given us a great insight into the rent of land and houses in England.25

Displaying a keen understanding of the problems involved in preparing estimates of national income and other statistical com­ putations, Davenant defined the skills needed for such work. He wrote: He who will pretend to compute, must draw his conclusions from many premises; he must

not argue from single instances, but from a thorough view of many particulars; and that body of political arithmetic, which is to frame schemes reduceable to practice, must be com­ posed of a great variety of members. He who will arrive at this art, must look into all the public revenues; he must under­ stand something of their management; he must not be a stranger to the product and manufac­ tures of every country and place; he must know what goods we export, and what foreign com­ modities are imported to us; and only from this general view, he must frame any scheme that may be useful to the public. A contemplation of one object, shall give him light into things per­ haps quite of a different nature: for as in com­ mon arithmetic, one operation proves another; so in this art, variety of speculations are help­ ful and confirming to each other.28

c. Pleas for Better Statistics and for Freer Access to Them Davenant did not claim to possess these rare qualities personally, reserving a more modest function to himself. But there is in this art a sphere for lower capacities to move in, who presume no further than to find out and prepare materials, to be made use of by abler hands, and who think it praise enough to them, if they can make tools for skillful artists to work with withal; and of this rank is the writer of these papers.27

While claiming that England’s statistics were better than any other country’s, he still felt they were inadequate for preparing quality estimates. “In this art [of computa­ tion of the] wealth, stock, product, consump­ tion, shipping, exportations, and importa­ tions [of the country, etc.],” he wrote, “the most difficult point is to find good materials, and to have footing, probably sure, to fix our reasonings upon; for where our premises can be certain, our conclusions shall be al­ most undeniable. . . 28 Concerning recent improvements made in the government’s statistics Davenant noted: . . . in order to this, the house of lords did, in 1695, lay a most excellent foundation, which we hope some able head and good genius will so improve, as by this aid to find out that balance of trade which has been so often talked of. Their lordships have directed the commis­ sioners of the customs to draw out an account of all the exportations from London, and the outports, to every distinct country; and also of all the importations to London, and the out­ ports, from every distinct country, for the years 1694, 1695, 1696, and so downwards. Every commodity is under a separate head, and the

2 . D E V E L O P M E N T OF FIRST ESTIM ATES IN ENGLAND

drawbacks upon re-exportation are taken notice of, and the value is set down, where the duty on goods is ad valorem. And the whole is put into a very good method. . . ,29

He wanted this Parliamentary order ex­ tended so that the “books may be made up in the same method” for every year as far back as 1672 so as to cover periods of peace as well as war, periods in which foreign trade was at its height as well as those in which it was at its lowest ebb. Davenant also wanted to have made available to the gen­ eral public the facts about the nation’s trade and other phases of its life. He complained that when he was preparing his own work on public revenues and trade, he was denied access to important sources of information: “The matters to be treated of, are intricate and difficult, requiring the sight of many accounts, which the Author cannot come at; (to whom the commissioners of excise have refused any inspection into their books) so that, swimming against the tide, he is more willing, than able, to serve the public.” 30

Annual income: Before the War Since the War Revenues Debts

war, the French government must have run up a debt of 100 millions sterling, which would require an annual tax levy of 13.5 millions per year to take care of interest and redemption. Drawing again upon the figures worked out by Petty and King, Davenant wrote, “That there are reasons to think the annual income of Holland, from land, houses, trade and manufactures, is about £18,250,000. That during this war, the Dutch have paid above l/3 d of the an­ nual income of their country every year to the public, or per ann. 6,900,000,” 33 and that the debt of her government to her own people must stand at around £ 2 5 millions. He concluded that these heavy debts would require such heavy taxes that neither France nor Holland would want to become en­ tangled in another war for many years to come. “To help the reader’s memory,” Davenant drew up the following “scheme” (table) of the annual income, government revenues, and public debts of England, France, and Holland: 34

England

France (sterling)

Holland

44.000.000 43.000.000 3,355,472 17,552,544

84.000.000 81.000.000 13,500,000 100,130,000

17.500.000 18.250.000 4,750,000 25,000,000

d. Davenant’s International Comparisons In addition to publishing King’s estimates of the comparative national incomes, popu­ lation, and resources of England, France, and Holland, Davenant also offered his own estimates and commentaries which, in many respects, were more complete.31 He wrote, “There are reasons to think that the general rental or annual income of France, from land, trade and manufactures, amounted to, in time of peace, about £ 84,000,000 sterling. This general income we deduce from that consumption, which such a number of people (as France contained before the war) must of necessity make, to sustain life.” After reviewing the various sources of French public revenue, he con­ cluded, “And it makes up the sum . . . of 16,100,000 per ann., which is near the 5th penny of 84,000,000 annual income, which . . . by the war, loss of trade, and expulsion of the protestants, was perhaps reduced in 1697 to 77,000,000 annual income.” 82 He further concluded that, as a result of the

39

Davenant reached the general conclusion that the people of England were much less burdened with taxes than the people of France; that peacetime taxes in England were “but little above l/20th part of their annual income” and in time of war “not quite l/8th part,” or, “little more in pro­ portion to its annual income during the war, than France paid in times of peace.” He re­ peated Petty’s observation that in time of peace a people may pay one fifth of their income in taxes and in time of war pay one third and still be prosperous, for what counts is not how heavy taxes are but how wisely or unwisely they are spent. Like Petty, Davenant pointed to the fact that Holland spent a large proportion of her tax revenue to support the Navy, i.e., the country’s over­ seas trade, thus materially contributing to the citizens’ well-being. Because of this, he said, no Dutch citizen begrudges the govern­ ment this portion of the levy, . . but willingly lays it aside for the state, as the penny best employed, because it gives a

40

THE INCOM E OF NATIONS

value to all the rest [of his income].” 35 Finally, Davenant looked forward to the time when, having paid off their public debt, . . the people of England may come to pay, as heretofore, to the government, but about l/20th part of their annual income; which would put the land interest, trade, and manufactures in a good posture, and a flourishing condition.” 36

4. Summary The national income estimates and anal­ yses of Petty, King, and even of Davenant, who was their intellectual debtor, were broad and penetrating, and were to remain unequaled in scope and quality for the next two hundred years. Colin Clark, in the

Preface to the first edition of his Conditions of Economic Progress (1940), aptly says that “economics was started on the right lines by Gregory King and Sir William Petty at the time of that astonishing flowering of the English scientific spirit in the later seven­ teenth century.” After observing that Pro­ fessor Hogben had “recently honoured Sir William Petty in the field of population” by calling his own work “Political Arithmetic,” Clark says, “In the field of economics, too, I think we need to return to the tradition of these two brilliant pioneers from whom even now we have much to learn.” 37 There is no doubt that the full significance of Petty’s and King’s work in terms of the subsequent development of both economics and statis­ tics has not yet been fully appreciated.38

3. ELABORATIONS IN ENGLAND IN THE EIGHTEENTH CENTURY After its brilliant start in the seventeenth century, further development of Political Arithmetic and work on national income estimates in England came to a stop in the beginning of the eighteenth century. The general sense of domestic well-being and the relaxation of international tension contrib­ uted, no doubt, to the decline of interest in national income estimating. But, toward the end of the century, under the unsettling pressures of a great many new factors—the Industrial Revolution, the outbreak of the American Revolution and the loss of the American colonies, the challenge of revo­ lutionary France to the English social order, and, finally, the outbreak of the Napoleonic Wars—interest in national income estimates and Political Arithmetic was revived.

1. The Period of Neglect of “Po­ litical Arithmetick,” 1700-1770 During the first seventy years of the eighteenth century, only three estimates of national income seem to have appeared in England. All three were mere projections Notes begin on p. 518.

of Petty’s and King’s figures, roughly ad­ justed to supposed changes in population, agricultural productivity, and the volume of foreign and domestic trade. None of them exhibited any originality of approach or contributed new data, and none would have been accepted by the founders of “Political Arithmetick” as a worthy specimen of their art.

a. The British Merchant’s Estimate for 1713-21 In 1721, in a publication called The British Merchant; or, Commerce Preserved, an article appeared containing an estimate of the country’s national income during the period 1713-21. The author, described by the publisher as “an Inspector General of Customs,” attempted to relate the country’s foreign trade to its income and wealth. He estimated England’s population at 7 mil­ lions, as against Petty’s and King’s estimates of 5.5 to 6 millions. On the other hand, he took Petty’s figure of £ 7 per annum for per capita income as if no changes had taken place in productivity or the value of

3.

ELABORATIONS IN ENGLAND

money during the intervening forty to fifty years. He thus arrived at a total national income of £ 4 9 millions, which he did not attempt to break down into components or otherwise verify. Sir Robert Giffen, eminent British statistician of the late nineteenth century, in commenting upon this estimate, felt that its population figure was too high, but that the national income total may have been approximately right.1

b. Sir Matthew Decker’s Estimates for 1740 The estimate made by Sir Matthew Decker (1679-1749) in an essay published in 1744 on Causes of the Decline of the Foreign Trade, etc. was slightly better.2 A Hollander by birth and resident of England since 1702, Decker was an important mer­ chant, director of the powerful East India Company, and, for a time, member of Par­ liament. An ardent free trader, he published two essays advocating the repeal of customs duties, proposing that they be replaced by a general tax on houses and buildings (ex­ empting the houses of the poor) and by excises on luxuries. To support his proposal, Decker prepared an estimate of the then current national income (1744). He set the population at 8 millions— a figure that Gif­ fen felt was one million too high— and the per capita income at £ 8 per annum, which may or may not have been right, arriving at a national income total of £ 6 4 millions. Decker gave no other particulars for his estimate, except to state his figure for the annual rental from land and buildings, which he set at £ 2 0 millions as against the Petty-King figure of £ 1 2 .5 -1 4 millions. Giffen believed that this estimate of rental income was about right, and that Decker’s total of 64 million pounds compared “not unfairly with the preceding estimates, there having been no material change of prices in the interval to be allowed for.” 8

c. The Anonymous Frenchman’s Esti­ mates of 1764 Giffen also mentions a third estimate made in the middle of the eighteenth cen­ tury that appeared in a French book pub­ lished in 1762, and translated into English in 1764. In its English edition, it was en­ titled A General View of England from the

year 1600 to 1762 to A.M.L.C.D. by M.V.C.M. According to the translator’s preface, the author “is (upon pretty sure

41

grounds) supposed to be a French gentle­ man, who, several years ago, resided for some time in England, and who, within these last ten years, was at the head of the finances of France.” 4 From Giffen’s de­ scription of this work, however, it appears that the author estimated only some ele­ ments of the national income of England, such as rental income, which he set at £ 1 5 .5 millions for land and £ 4 .5 millions for houses, or at the same total of £ 2 0 millions as did Matthew Decker.

2. The Revival of Interest in Po­ litical Arithmetic Incident to the Industrial Revolution The Industrial Revolution, which created a whole set of new social and economic problems, also brought about a revival of public interest in “Political Arithmetick” and in national income estimates. Through the efforts of Arthur Young (1741-1820) and John Sinclair (1754-1835), consider­ able improvement was made in British sta­ tistics. Although Sinclair never extended his statistical researches to the estimation of national income, he helped all future work in that field by the publication in 1791 of a twenty-one-volume Statistical Account of Scotland, and by inducing Pitt and other officials to place the collation of basic na­ tional statistics on a firmer foundation.

a. Arthur Young’s Estimates of 1770— A Signal of a New Era in Economic Statistics In 1770, Arthur Young prepared an esti­ mate of England’s national income which was widely different from all previous eight­ eenth century estimates. Even though it, too, was somewhat casual by modem stand­ ards, it was far more systematic than any of the earlier efforts, and was based on a substantial body of data. Young, scion of a distinguished and wealthy family of landowners, showed little inclination towards academic studies in his youth, preferring lighter pastimes. But in his twenties, having lost his fortune in unsuc­ cessful farm management, he became intent on learning all about scientific agriculture and proceeded to visit the better managed farms in England and on the Continent. From that time on Young was active as both writer and scientist, two activities that

42

THE INCO M E OF NATIONS

occupied him for the rest of his lengthy life. He set out to convey to the English farmers what he was learning, and, being a keen observer and a fluent writer, he succeeded extraordinarily well in this undertaking. His works became immediately popular not only in England, but also on the Continent where some of them appeared in translation. In 1793 these activities brought him an ap­ pointment as Secretary to the newly created national Board of Agriculture— an office he held to the end of his life.5 Young’s interests extended far beyond agriculture to include public finance, gen­ eral economics, politics, and culture. At one time he became interested in reforming the country’s tax system through the introduc­ tion of a general income tax and the aboli­ tion of the church tithe. To prove that under his system more revenue would be produced than under the existing one, he needed an estimate of the national income and pre­ pared one himself. He presented this estimate in his Six M onths Tour Through the North o f Eng­ land.6 It was, in many ways, a novel one, built along different lines from the estimates of Petty, King, and Davenant. First of all, it was built up from production data Young collected in the course of his travels and on his observations of changes in produc­ tion, rather than, as were the others, from fiscal data. Secondly, his classification of branches of production was more elaborate than theirs and incorporated the newer in­ dustrial branches of production. At the same time that Young’s estimates signaled a new era in statistical collation, they suffered from certain defects. They were undocumented, for one thing, and contained far too many “suppositions.” Young, himself, did not give too much credence to the accuracy of his figures, as evidenced from the following extraordinary admission with which he ended his statistical exhibit: Now the most inattentive eye must be able, at the slightest glance, to specify abundance of various kinds of income omitted in this table; but I by no means aim at an accuracy in a matter that requires it not: All I would en­ deavour to show, is that the income of the whole people is a very great sum, compared to all public wants! and that it, in all probability, amounts to considerably more than an hundred millions.7

Young’s estimates (expressed in pounds) are reproduced here in full:

Table 3-1. Arthur Youngs Estimate of 1770 Agriculture The income we have from products specified amount to Suppose woods, timber, inland fish­ eries, parks, mines of all sorts, yield a product of

60,000,000 6,000,000 66 , 000,000

Manufactures The average of five accounts now before me makes the value of the labour added to our wool to amount to The labour bestowed on leather, exclusive of the consumption in the article wear and tear in hus­ bandry, consisting of shoes, breeches, coaches, chairs, har­ ness, &c. &c. &c. Suppose The manufactures of lead, tin, iron, copper &c. is one of the first, if not the greatest in the kingdom. Suppose the labour is Flax and hemp, glass, paper, and porcelaine Silk and cotton must be consider­ ably more than Besides these articles there are all the earnings of the whole body of artisans th a t are scattered (except in the hardware way), such as carpenters, masons, cab­ inetmakers, upholsterers, gla­ ziers, &c. &c. with an infinite number of shopkeepers: The whole aggregate of labour, ex­ clusive of the preceding manu­ factures, must be prodigiously great: However, th at we may not exaggerate let us suppose it, including all trades, not be­ fore specified, a t

7.000.000

4.000.000

.

6 000.000 2 , 000,000

1,500,000

6,500,000 27,000,000

Commerce The amount of the income arising from commerce can only be con­ jectured: But when we consider th a t it includes that, not only of the merchants, b ut also of all the numerous bodies em­ ployed by them, such as sailors, shipbuilders, boatmen, writers, porters, servants, with a vast number of & cetera’s, it must certainly be very considerable, suppose The public revenue, exclusive of the interest paid to foreigners The interest of the savings in agri­ culture, manufactures, and com­

10,000,000

9,000,000

3 . ELABORATIONS IN ENGLAND

merce, exclusive of the public funds, which are included in the last article; and the sums bor­ rowed by farmers, such as mortgages, bond-debts, &c. &c. &c. suppose Law, physic, the fine arts, litera­ ture, &c. &c. cannot create an income of less than

5,000,000 5,000,000

Recapitulation The soil Manufactures Commerce Publick revenue Sums a t interest Law, physic, &c. Total income of England

66 , 000,000

27.000.000

.

10 000.000

9.000.000 5.000.000 5.000.000 122,000,000

s o u r c e : Arthur Young, A Six Months Tour Through the North of England (anonymously pub­ lished), London, 1770, vol. IV, pp. 543-47.

b. Mr. Pulteney’s Amateurish Estimate Nine years later, in 1779, at the climax of England’s war with the American colo­ nies, a certain Mr. Pulteney published an essay entitled, Considerations on the Present State o f Public A ffairs and the M eans o f Raising the Necessary Supplies, which con­

tained, among other things, an estimate of national income. This estimate, however, was not a scientific one.8 Pulteney estimated the population of the country at 7 millions, which was much too low for the time, and took the similarly obsolete figure of £.7 10.s\, derived from Petty’s and King’s cal­ culations, for per capita income. He thus arrived at a ridiculously low national in­ come estimate of £ 5 2 .5 millions. Appar­ ently, Pulteney was either unacquainted with Arthur Young’s contemporary estimate of £ 1 2 2 millions or chose to pay no attention to it.

3. The Influence of the Napoleonic Wars and of the Income Tax Proposal At the end of the eighteenth century, England’s involvement in the struggle for world power placed a great strain on her economy, and forcefully pointed up, once again, the need for having good national income estimates at the disposal of her statesmen. George Rose (1744-1818), who was a member of parliament from 1784 till

43

his death, secretary to the treasury in 17821801, and assistant and close intimate of Pitt’s, in a preface to his first work, pub­ lished anonymously in 1793, expressed the anxiety shared by most of England’s citi­ zens over the financial and economic posi­ tion of their country at the end of the War with the Colonies.9 As a result of this gen­ eral concern a new series of national in­ come estimates began to appear in the last years of the century. The new estimates were founded upon richer statistics than those that preceded them, and were keyed to a single objective —that of providing a basis for estimating the yield of a general income tax. Moreover, they were based upon the wholly new classi­ fication of incomes that had arisen out of the application of the income tax. The ma­ jority of subsequent British estimates de­ rived the bulk of their data from income tax statistics and followed the classifications of income used in these statistics.

a. Pitt’s Estimates In 1798, William Pitt, while proposing the imposition of a general income tax, prepared an estimate of the total income of the people of England, Wales, and Scotland, which he proposed to bring under his tax. He classified these incomes in twelve groups, showing for each class the total income, the sum to be exempted, and the amount of revenue to be obtained. Since he proposed to exempt all labor income from the tax, Pitt did not attempt to estimate it. Hence, his national income total of £ 1 2 5 millions was far short of the actual one. Pitt’s table is reproduced as Table 3-2.

4. Dr. Beeke’s Estimates of 1 7 9 9 1800 The Reverend Henry Beeke, a parson possessing unusual facility in handling sta­ tistics, who believed strongly in the sound­ ness of Pitt’s proposed tax, undertook to verify and correct Pitt’s figures, publishing an exhaustive study on the subject in 1799, following it a year later with a revised edi­ tion of the same work.10 Little is known of Dr. Beeke except that he was born in Devonshire in 1751, was educated at Cambridge, where he received his Master’s and Doctor of Divinity degrees, and served as professor of modern history at Bristol from 1801 and as Dean from

44

THE INCOM E OF NATIONS

Table 3-2. Computation of the Income of Great Britain, as Stated by Mr. P itt in the House of Commons; distinguishing the Part likely to be Taxable under the Act Annual Income (£) Landlords’ rents, 40,000,000 of cultivated acres estimated a t 12*. 6d. per acre Tenants’ profits a t three fourths Tithes Mines, [inland] navigation, & timber Houses Professions Proportion fr. Scotland Income from possessions beyond sea Interest on funds, after deducting foreign property, and sums issued to commissioners as sinking fund and interest to capital re­ deemed Profit on foreign trade, suppose 15 £ per cent. on 80,000,000 £ of capital insured Ditto home trade, a t 15 £ per cent. Other trade

Deduction (£)

Taxable Income (£)

25,000,000 19,000,000 5,000,000 3,000,000 6,000,000 2,000,000 5,000,000 5,000,000

J-f; 5,000,000 20,000,000 % 13,000,000 6,000,000 1,000,000 4,000,000 3,000,000 % 1,000,000 5,000,000 2,000,000 5,000,000 5,000,000

15,000,000

% 3,000,000

12,000,000 18,000,000 10,000,000 £125,000,000

12,000,000

12,000,000 18,000,000 10,000,000 23,000,000

102,000,000

s o u r c e : Rev. Henry Beeke, Observations on the Produce of the Income Tax and on Its Proportion to the Whole Income of Great Britain, 2d ed., London, 1800, p. 5.

1814. No account exists, or has come to our attention, explaining how Beeke hap­ pened to acquire the understanding of con­ temporary economics and skill in handling figures that characterized his entire work. Beeke corrected certain errors in Pitt’s table and added an estimate of labor in­ come to Pitt’s total, setting it at £ 1 1 0 mil­ lions. After these corrections he arrived at a national income total of £ 2 1 7 millions, on the basis of which he recalculated the probable yield of Pitt’s proposed income tax. He indicated that its low yield was due not to any insufficiency of the annual income of the people, but to the lowness of Pitt’s proposed tax rates. He estimated the rev­ enue at about £ 6 .2 5 millions instead of at the £ 7 .5 millions estimated by Pitt. He showed that at this rate it would amount to but 3 per cent of the national income; and claimed that the national income could bear a higher rate of levy.11 Beeke’s revenue estimate proved more accurate than Pitt’s, as shown by the actual collection of the tax in the succeeding year.

a. Beeke’s Estimates of Income from Agriculture and Certain Other Sources Like Pitt, Beeke estimated the rental in­ come from land (received by some 200,000 landowners) by multiplying an average rental per acre by the number of cultivated

acres. However, he adopted lower figures for both the cultivated acreage and the rental per acre, reaching a total of £ 2 0 millions, instead of Pitt’s estimate of £ 2 5 millions. Beeke also lowered Pitt’s estimate of £ 1 9 millions for the income of tenant farmers to £ 1 5 millions. Whereas Pitt cal­ culated these incomes at three fourths of the landlords’ rents, Beeke set them at two thirds on the ground that the income tax law prescribed a new method of computa­ tion of farm profits. Another contemporary estimate and a modern criticism of these early estimates suggest that whereas Pitt’s original figure for agricultural income was too low, Beeke compounded the error by reducing Pitt’s figure still further. (See Sec­ tion 5c below.) He also cut in half Pitt’s estimate of £ 5 millions for the tithe in­ come of churches and universities, a figure that Pitt had taken from Arthur Young’s works. On the other hand, Beeke believed that Pitt’s estimate of £ 3 millions for the in­ come from mines, inland navigation, and timber was too low, and raised it to £ 4 .5 millions.12 Beeke also raised Pitt’s estimates of the rental income from residential and commercial buildings from £ 6 to £ 1 0 mil­ lions, but agreed with his estimate of £ 2 millions for the income from professions. He considered Pitt’s setting of the income

3.

ELABORATIONS IN ENGLAND

of Scotland at one eighth that of England and Wales to be too low, and raised it to one fifth of their income. He reduced Pitt’s estimate of £ 5 millions for the income from overseas possessions by £ 1 million to take account of offsetting payments to foreign investors that Pitt had failed to con­ sider. He took issue with Pitt for including the interest payments on the domestically held public debt, holding that these were offset by payments of interest to foreign holders of Britain’s public debt. Beeke reduced the estimate of profits from foreign trade that Pitt had computed at 15 per cent of the total British foreign capital investment of £ 8 0 millions, i.e., at £ 1 2 millions. Beeke calculated these profits more directly from current figures on the value of exports and imports, reaching a figure of only £ 8 millions for them. Pitt had calculated the income from in­ ternal trade by applying a 15 per cent rate of return to the estimated £ 1 2 0 millions of capital investment in domestic trade. To the resulting figure of £ 1 8 millions he added a further £ 1 0 millions for the tax­ able income of other trade “including the lowest retail,” thus reaching a total of £ 2 8 millions for the two items combined.13 Beeke was not satisfied with this method and calculated the income from trade as the trader’s profit on the total national ex­ penditure in the home market. To do this, however, he first had to complete his com­ putation of income from all sources other than trade.

b. Beeke’s Estimate of Labor Income The most difficult item in Beeke’s com­ putation was labor income. He first esti­ mated the population of England and Wales at 10.5 to 11 millions, after considering that it was generally agreed to have been 6.5 millions in 1700 and, on the basis of militia lists, was probably better than 8 millions by 1757. He assumed that the population had probably increased another million by the end of the American war, and another two millions since then, increasing at the rate of about 100,000 a year.14 On the basis of sample studies of the population in London and certain other counties, Beeke decided that the total population could be broken down into 2.2 million families of 5 persons each. He further subdivided this number into two groups. The first group of 700,000 fami­ lies consisted of those earning more than

45

£ 6 0 a year. These were middle and upper income families from the propertied and trading classes. Beeke classified the remain­ ing 1.5 million families as low income fami­ lies, consisting of manual laborers, cottagers, and small shopkeepers. In this larger group there was usually more than one wage earner per family, including women and children. Beeke decided, for estimating pur­ poses, to set this at an average of 1.5 wage earners per family, with each wage earner bringing in £ 3 0 a year or a total of £ 4 5 a year per family. Using these figures, Beeke computed the total labor income at £ 6 7 .5 millions.10 To this he added Pitt’s allowance of £ 2 millions for the “profits of profes­ sions” and “a considerable sum for the value of labour of domestics retained solely or principally on account of their master’s oc­ cupation, as clerks, shopmen, servants in husbandry, etc.” He further explained that “the master is allowed to make a deduction of their maintenance and wages, in esti­ mating his own income; in which, therefore, the cost of their labour is not included” and that, therefore, their wages and maintenance “must be added to that [the wages] of the 1.500.000 families before mentioned.” Beeke also took account of the fact that . . . many of the occupiers of lands who in­ habit houses not reckoned among the 1,500,000 just mentioned, are working, not merely super­ intending farmers; and I do not imagine that the value of their manual labour ought to be confounded with their farming profit, or was intended to be included in it. For it would be great injustice to accumulate in one income the personal labour and farming profit of a man renting only 30 pounds a year; and yet allow another renting ten times as much, to deduct for his bailiff.10

Finally, he said, one must also add “the salaries, fees, and occasional gains by per­ sonal service of all above the lowest class, whether employed by government or by in­ dividuals, [which] are incomes of industry, of which few persons in active life are with­ out some portions.” Admitting that it was more difficult to estimate the labor income of the 700,000 families in the upper and middle income group than the labor income of the 1,500,000 low income families, Beeke fixed it at an average of £ 4 5 a year for each of the 700.000 families, or at £ 3 1 .5 millions. Adding this amount to the £ 6 7 .5 mil­ lions, he arrived at a total labor income for England and Wales of £ 9 9 millions, saying

46

THE INCO M E O F NATIONS

at the same time that he was “more inclined to believe that this is too small a sum, than that it errs’ in the opposite extreme.” 17 Next he estimated the population of Scot­ land at one seventh of that of England and Wales, noting that wages in Scotland were lower than in England. For this reason he estimated the labor income of Scotland at one ninth of that of England and Wales ( £ 9 9 millions), or at £ 1 1 millions. After all these additions, Beeke’s total for labor income of Great Britain was £ 1 1 0 millions.

c. Beeke’s Estimate of Income from In­ ternal Trade Next, Beeke estimated the income from capital, exclusive of that from trade, at £ 9 0 millions. By adding this to the labor income he obtained a total of £ 2 0 0 millions as representing total national income exclusive of home trade. He allowed £ 3 3 millions for savings and direct taxes, deducting the latter sum from the £ 2 0 0 millions total. This left £ 1 6 7 millions available for ex­ penditure in the market for consumer goods, an aggregate similar to “disposable income” in modern national income accounting. On this turnover he allowed a 15 per cent profit to traders, equaling £ 2 5 millions. To this he further added “the profit of traders on each other’s expenditure,” i.e., 15 per cent on five sixths of £ 2 5 millions, or another £ 3 millions, so that “the whole profit on all internal trade, including the lowest re­ tail, will be 28 million pounds.” 18 This £ 2 8 millions figure was the same as the one Pitt had offered. Beeke, however, in reviewing the basis on which the calculation was made,

decided that 15 per cent was too high a trader’s profit and decided that 10 per cent would be a more realistic figure, taking into account offsetting losses and other expenses. In addition, he was of the opinion that the whole capital turned over in eight months in­ stead of in twelve. In the light of these two considerations, Beeke reached the conclu­ sion that profit to traders from both retail and internal trade could amount to no more than £ 1 8 millions, and that the whole na­ tional income “. . . will be about 218 mil­ lion pounds sterling.” 19 Beeke’s estimate of the total income of the people and the yield of the proposed tax is presented as Table 3-3.

d. Beeke’s Observations Regarding the Distribution of National Income On the basis of his estimates, Beeke ad­ vanced certain novel generalizations regard­ ing the distribution of the national income that represented an attempt to formulate a general law of the distribution of income. His first observation, which was empirical in nature, was that “the wealth of Great Britain is far more equitably divided than has commonly been supposed.” His second generalization, more in the nature of a hypothesis, was that “the natural and moral causes of unequal division of property, and income, have, when undisturbed by par­ ticular laws, or customs, a tendency to make the shares greater in proportion, as the pro­ prietors, when divided into classes, become less numerous, in a regular order; and . . . that this regular order to which the division of property, so far as influenced by natural

Table 3-3. Beeke’s Estimate of the National Income and the Yield of the Income Tax

Land Rents Farming Profits Tithes Mines, etc. Houses Scotland Foreign Possessions Funds Foreign Trade Shipping Home Trade Labour

Clear Income £ 20,000,000 15,000,000 2,500,000 4,500,000 10,000,000 8,500,000 4,000,000 15,000,000 8,000,000 2,000,000 18,000,000 110,000,000

Gross Assessment £1,500,000 220,000 225,000 300,000 500,000 450,000 400,000 1,200,000 750,000 180,000 800,000 500,000

£217,500,000

£7,025,000

About (per cent) 7.5 1.4 9.0 6.6 5.0 5.2 10.0 8.0 9.3 9.0 4.4 0.4

Exempted or Abated £ 500,000 1,280,000 25,000 150,000 500,000 400,000 000,000 300,000 50,000 20,000 1,000,000 10,500,000

About (per cent) 2.5 8.6 1.0 3.4 5.0 4.8 0.0 2.0 0.7 1.0 5.6 9.6

£14,725,000

s o u r c e : Observations on the Produce of the Income Tax and on Its Proportion to the Whole Income of Great Britain, 2d ed., London, 1800, p. 136.

3.

ELABORATIONS IN ENGLAND

and moral causes only, is continually ap­ proximating, must be an arithmetical pro­ gression.” 20 By arithmetical progression, Beeke meant a distribution in which the society was divided into ten groups, with the richest containing ten times fewer peo­ ple than the poorest class but receiving aver­ age incomes perhaps ten times as large as the average per capita incomes of the people in the poorest group. He added, “I do not state that such a progression ever actually exists, because, in fact, its existence is al­ ways more or less impeded by political acci­ dents and institutions; but I only state, that there exist certain natural and moral causes which are always actively tending to pro­ duce it. These natural and moral causes are —unequal ability and diligence in acquiring property—unequal prudence and success in preserving it—unequal numbers inheriting it.” 21

e. The Merits of Beeke’s Estimates Beeke’s work deserves much more credit than it has received in the past.22 His esti­ mate of the income from agriculture may have been too low, resulting, therefore, in an underestimated national income total, but his estimating procedures taken as a whole were methodical—more so, in fact, than those of Arthur Young. They repre­ sented a return to the type of systematic and detailed estimates made a hundred years before by Gregory King, and contrasted sharply with the type of “guess work” in­ dulged in by most of the British national income estimators of the eighteenth cen­ tury.

5. Dr. Bell’s Estimates of 1 7 9 9 1802 a. The Source of Bell’s Interest in Na­ tional Income Another contemporary of Pitt’s to pre­ pare a national income estimate was Dr. Benjamin Bell (1749-1806), a famous Edinburgh surgeon, author of a number of notable medical works, practical as well as scientific agriculturalist, keen analyst of the current economic scene, and a respected friend of Adam Smith. Bell was prompted to prepare a national income estimate for the country, or more correctly to prepare a correction of Pitt’s estimate, by a desire to demonstrate statistically that Pitt’s pro­ posed income tax would be wholly inade­ quate to produce the needed increases in

47

the government’s revenues as well as effect the desired changes in the distribution of the national tax burden among the differ­ ent classes of British society. He wanted to show that a much more generalized income tax, progressive in its rate schedule, was needed to accomplish these ends. Born on his father’s large farm, Bell de­ veloped an early interest in both the prac­ tice and study of agriculture. Soon after completing his medical studies and settling down to practice in Edinburgh, Bell ac­ quired a large farm near the city, establish­ ing his family there during all but the coldest months, and spending a few days each week in its management and espe­ cially in conducting experiments in scien­ tific cultivation. His large surgical practice forced him to travel extensively throughout the country, and he made it his business to make inquiries wherever he went about the state of local agriculture. In accordance with his lifetime habit of committing all note­ worthy observations to paper, he accumu­ lated a great deal of material on the sub­ ject. “Having early seen with regret,” he says in the preface to one of his later works (Essays on Agriculture, Edinburgh, 1802), “that the Agriculture of the kingdom did not keep pace with its population, he judged it to be his duty to suggest his opinion to those who might have it in their power to remedy so great an evil.” And so, in 1783 he began the publication, anonymously, of a series of essays on agriculture. Bell attributed the depressed state of agri­ culture of the second half of the eighteenth century—characterized by low prices, low profits, abandonment of many farms, and, generally, the failure to put all usable land under cultivation—to the corn laws, which tended to keep farmers cultivating corn rather than more economically produced crops and to use ancient and obsolete meth­ ods; and to the huge size of the national debt and the overtaxation of the land and farm products. He advocated greater diver­ sification in cultivation, particularly the cul­ tivation of rice, potatoes, and garden vege­ tables, which he believed to be more suited to the conditions and needs of England of the time; and he urged the government to repeal the com laws and to provide sub­ stantial bounties to farmers diversifying their crops and increasing the yield of their land. He also urged the government to re­ peal all the existing internal excises and customs duties, and replace them with a universal income tax, levied at graduated

48

THE INCO M E OF NATIONS

rates, which would relieve the farmers of some of the tax burden improperly imposed on them and shift it to other, more lightly taxed classes, and would provide enough revenue to enable the government to finance his proposed program of bounties to farm­ ers and to begin to pay off the national debt at a substantial rate. In the preface to his Essays on Agriculture (see note 23), he wrote: . . . till measures are pursued for the quick reduction of the National Debt, there is reason to fear, that the Agriculture of the country will not meet with that encouragement which it re­ quires; and, if it be not by a well-regulated Tax on Income, no plan has ever been mentioned, that, in his opinion, can furnish funds for the payment of any considerable part of that debt.

In the same preface, Bell claims that Adam Smith fully agreed with his main ideas and encouraged him to put them in book form. He was hampered by his pro­ fessional duties in completing the work, but when William Pitt made public his income tax proposal, Bell, finding it thoroughly inadequate, proceeded forthwith with the publication of his still not fully completed manuscript. Thus, in 1799, Bell’s Essays on the Income Tax appeared anonymously.23

b. Bell’s Critique of Pitt’s Income Tax and His Substitute Proposal Bell found five main faults with Pitt’s in­ come tax: (1) that the tax was not suffi­ ciently universal, inasmuch as it exempted all incomes below £.60 a year, i.e., practi­ cally all labor incomes; (2) that it was imposed on specific categories of income rather than directly on total incomes of in­ dividuals; (3) that it was imposed at pro­ portional rather than progressive rates; (4) that its rates were much too low and, hence, would not produce enough revenue to ac­ complish the needed objectives; and (5) that it was assessed through county govern­ ments on the basis of the declarations of individuals of their various types of in­ comes, rather than through smaller jurisdic­ tions and by direct assessment. He proposed instead that the exemptions be lowered to £ 1 5 a year so that laborers would contribute to the public treasury just as all other citizens did; that the tax be imposed on the total income of a person and be levied at progressive rates starting with 1 or 2 per cent and rising gradually to 10 per cent in the medium brackets (in­ comes of £ 2 0 0 to £1,500) and to 33 per

cent in the higher brackets (£45,000 in­ comes and up); that the rates be so gradu­ ated that the whole tax produce 10 per cent of the national income; and that the tax be assessed by elected citizens in the parishes, with the taxpayer being given the privilege to protest an unfair assessment. He also proposed that in fixing the rates considera­ tion be given to the differences in the de­ gree of “Permanency of the sources from which the incomes proceed . . . [so that] income of the same extent arising from the daily exertion of individuals in their various professions, as being more uncertain . . . [would be taxed at lower rates than] in­ comes arising from permanent funds, such as land, houses, mines, mortgages,” etc.24 thus anticipating by a hundred years the plan of tax rebates on earned income intro­ duced in the British income tax by Lloyd George. Following in the footsteps of Vauban (see Chapter 4, Section 2), Bell proposed sub­ stituting the income tax for all existing con­ sumption levies, and further recommended the modification of its rates from time to time, in accordance with changes in the conditions of the country and in the re­ quirements of the government, thus empha­ sizing the principle of elasticity in the con­ struction and application of a tax system. Bell’s ideas on taxation were considerably in advance of those of Adam Smith and his other contemporaries. In his advocacy of progressive taxation, he was a pioneer, preceded only by Montesquieu (see his Esprit des Lois, Livre XIII, Chapter 7) and by Rousseau (De L ’Economie Politique, Oeuvres, Paris, 1826, vol. IV, p. 441) in France, but by no one in England. More­ over, whereas these two philosophers ex­ pressed the principle of progressive taxation only abstractly, Bell sought to give it prac­ tical expression in an actual project of an income tax. In this respect he seems to have been the first anywhere.

c. Bell’s National Income Estimate To demonstrate the inequitableness of the existing tax system, the inadequacies of Pitt’s income tax proposal, and the supe­ riority of his own broader tax proposal, Bell prepared his own estimate of the British national income and of its distribution among the different classes of population. His concept of national income, like Pitt’s, was the traditional one of the aggregate of all incomes from any source whatsoever. Despite his respect for Adam Smith, he was

3 . ELABORATIONS IN ENGLAND

not, apparently, much influenced by the latter’s material production concept. At least, he does not mention it anywhere in his work or introduce it in any of his computations. Bell arrived at a much higher national income total than Pitt did, not only because he included the bulk of labor incomes, but also because he set the incomes from agri­ culture and certain other sources much higher (see Tables 3-1 and 3-3). His total was £.236 millions as against Pitt’s £ 1 2 5 millions. At the proposed average rate of 10 per cent, his proposed income tax would have yielded £ 2 3 .6 millions as against the £ 7 .5 millions Pitt’s tax was designed to raise (or the £ 6 million odd that it actually produced). Bell maintained that the coun­ try could sustain the burden of such a levy, insisting that no tax of a lesser magnitude could finance the war without a further in­ crease in the national debt, not to mention reducing the debt. He also presented a de­ tailed computation showing that under the existing tax system, agriculture was bearing a share of the tax burden wholly out of pro­ portion to its share in the national income; and that this inequity would be corrected under his proposed tax to the advantage of the whole economy. In the preparation of his estimates, Bell followed Pitt’s general arrangement, adding two categories of income to the latter’s twelve—income from money on bonds and other personal securities, and income from

49

labor above £ 1 5 a year. His other impor­ tant departure from Pitt’s estimate was in the higher figures he set for agricultural rent ( £ 4 0 millions) and for farmers’ profits (£ 3 5 millions). He estimated the income from labor very roughly, taking the total population of Great Britain, which he said “is not accurately known” but which he assumed to be in the neighborhood of ten millions, and assuming that about two mil­ lions of them were engaged in labor at in­ comes of £ 1 5 to £ 6 0 , with the average amounting to £ 3 5 , thus yielding a figure of £ 7 0 millions for this group. In his second book, Essays on Agriculture, he revised his original estimates only slightly, as can be seen from Table 3-4 below. He raised his estimate of the labor incomes from £ 7 0 millions to £ 7 5 millions by raising the number of recipients but assum­ ing a slightly lower average income. This and a few other minor rearrangements raised his estimate of the total national in­ come from £ 2 3 6 millions to £243 mil­ lions. His first proposal of a steeply graduated income tax must have been subjected to considerable criticism. He, therefore, mod­ erated it in his new work by raising the rates on the lower incomes to yl5, or close to 7 per cent of income, and lowering those in the highest brackets from 33 to 20 per cent. Under this schedule he expected the tax to produce £ 2 6 millions.

Table 3-4. Dr. Bell’s Estimate of National Income of Great Britain in 1799 (Incomes below £15 excepted) (Million £) Original Estimate (1799) Rents of land Profits of farming Amount of tythes Income from money in public funds From money in bonds and bills, including capi­ tals in banks From rents of houses Income of professional men From canals, mines and fisheries Incomes spent in Britain arising from prop­ erties in other countries From the retail trade over the nation From the capital employed in foreign trade From the capital employed in domestic trade Income of farmers, manufacturers (industrial workers) and others from £15 to £60 a year Total national income over £15 s o u r c e s : Three Essays on Taxation of Income, London, 1799, Benjamin Bell, Edinburgh, 1802, p . 143.

Revised Estimate (1802)

40.0 35.0 5.5 15.5

40.0 35.0 4.0 17.0

6.0 6.5 2.5 5.0

10.0 4.0 3.0 5.0

5.0 5.0 12.0 28.0

5.0 5.0 12.0 28.0

70.0

75.0

236.0 p.

243.0

74; and Essays on Agriculture

by

50

THE INCO M E OF NATIONS

Bell’s first book appeared in March 1799, apparently a few months before Beeke’s; but his second book appeared two years after the publication of Beeke’s second edi­ tion, yet Bell makes no reference to Beeke’s work in his new volume. Whether Bell had seen it, therefore, is uncertain. But if he had, he apparently must have believed his own calculations to be superior. A com­ parison of Tables 3-3 and 3-4 shows that Bell’s estimates of all three types of income from agriculture—rent, farmers’ profits, and the tithe—were substantially higher than Beeke’s, and that his national income total was also considerably higher. Phyllis Deane believes that Bell’s estimate was more cor­ rect than Beeke’s.25 This may well be true for Bell’s knowledge of agriculture was su­ perior to Beeke’s. On the other hand, his general statistical technique was certainly inferior and his interest in statistics as such was limited. This is evidenced by the fact that even though in the preparation of his second book he had opportunity to examine the assessments under the income tax which by that time had become available and to demonstrate the accuracy of his original estimates with the aid of this information, he merely contented himself with the re­ mark that “I have not as yet perceived, either from further attention to the subject, or from the result of the assessment hith­ erto, that in any material degree [my ob­ servations] have been wrong.” (Essays on Agriculture, p. 152.) Beeke, having a much greater appreciation of statistics, would probably have done otherwise under the same circumstances. Bell’s lack of statistical interest and proficiency does not detract much, however, from the great contribution that he made to the art of estimation and analysis of the national income—a contri­

bution which, until the appearance of Miss Deane’s article in 1955, had gone totally unnoticed in the annals of British national income estimation. Neither Giffen in his comprehensive historical review of British estimates, nor Bowley in his comprehensive historical bibliography of them, ever men­ tioned Bell’s w ork;26 nor did any of the other nineteenth- and twentieth-century Brit­ ish estimators do so.

6. Gentz’s Recalculations Based on the Smithian Concept In 1800, Friedrich Gentz (1764-1832), eminent German publicist and admirer of Adam Smith, published a critical review of the contemporary estimates of Britain’s na­ tional income.27 In addition to commenting on the ones prepared by Beeke, Pitt, and others, Gentz presented an estimate of his own. He agreed with most of Beeke’s fig­ ures (of which he had only the first set, pub­ lished in 1799), but rearranged them in keeping with Smith’s concept of productive and unproductive labor. Taking the view that the incomes of unproductive workers represent an expenditure of the national in­ come rather than an element of the income itself, he eliminated them from the com­ putation. Accepting Beeke’s calculation of £ 9 0 millions for the income from labor for England and Wales (£ 1 0 1 millions for the whole of Britain), he estimated that about one third of it represented labor income in agriculture, one third labor income in other productive occupations, and one third in­ come of unproductive labor. Accordingly, he reduced Beeke’s figure for the income from labor by some £ 3 4 millions, and submitted the following calculation for the current income of Great Britain: £ Sterling

1. Product of land Income from land in the proper sense of the term in­ cluding rent of proprietor’s income of farmers and agriculture labor income Income from mines, forests, and fishing Products of land in Scotland 2. Product of manufacturing industry and commerce Internal trade External trade Salaries of workers in productive occupations 3. Income from the possessions in the two Indies

100,000,000 5,000,000 15,000,000 28,000,000 12,000,000 30,000,000 5,000,000 195,000,000

s o u r c e : Friedrich Gentz, Essai sur I’Etat Actuel de VAdministration des Finances et de la Richesse Nationale de la Grande Bretagne, London and Berlin, 1800, p. 73.

3.

51

ELABORATIONS IN ENGLAND

Gentz’s speculations on the relationship between total government expenditure and tax revenue in 1799 and national income were rather interesting. He took Rose’s esti­ mate for total expenditure (£ 5 9 .6 millions) and for total tax revenue (£ 3 6 .1 millions), setting expenditures at 30 per cent of the national income and tax revenues at 18 per cent.28 He further estimated that if the war should end soon, tax revenues could be re­ duced to £24.7 millions, or to approxi­ mately 12.5 per cent of the national income, within a few years after peace was restored. In noting that Gregory King estimated England’s tax burden for the period of the war with France at only 10 per cent of na­ tional income (and at 7 per cent before the w ar), Gentz observed that it had been much more difficult one hundred years earlier to carry the much smaller tax burden than it was for the country to take care of the higher tax burden currently. Gentz wrote: “Today when the most exhaustive war in which England had ever been engaged places a much heavier tax burden on the nation than was ever placed upon it, she [England] is able to bear it without any serious impairment of any branch of its industry and encroachment upon the na­ tional prosperity.” 29

7. Comparison of the Results of the Estimates of the Two Centuries We can now bring together, in a single table, the results of the thirteen estimates discussed above and known to have been prepared in England during the two cen­ turies under review. These figures show an increase in the per capita national income from £ 7 in 1676 and £ 8 in 1700, to about £ 9 in 1744 (after correction of Decker’s overestimate of population) and to £ 2 0 in 1799. The 1799 figure reflects, in part, the great rise in prices that took place toward the end of the eighteenth century. Giffen says, however, “I should doubt whether the additions for the rise of prices, comparing 1700 with 1800, ought to be more than 50 per cent upon the figures of the former period.” If this is true, then the per capita income of 1800 expressed in the prices of 1700 must have been around £ 1 3 , or more than 50 per cent higher than a century be­ fore. The greater part of this increase took place in the second half of the eighteenth century, during which the Industrial Revo­ lution produced an unprecedentedly rapid industrial and commercial growth.

Table 3-5. Estimates of National Income of England and Wales in the Seventeenth and Eighteenth Centuries National Income Estimator and Year of Publication 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

William Petty, 1667 William Petty, 1676 Gregory King, 1695 Davenant, 1695 “British Merchant,” 1721 Matthew Decker, 1744 Anonymous French Writer, 1762 Arthur Young, 1770 Mr. Pulteney, 1779 William Pitt, 1798 Henry Beeke, 1799 (incl. Scotland) Benjamin Bell, 1799-1802 (incl. Scotland) Friedrich Gentz, 1800 (incl. Scotland)

a Exclusive of labor income. b Exclusive of incomes below £15.

Year for Which Prepared 1667 1676 1688 1688 1713-21 1744 1762 1770 1779 1798 1799 1799-1800 1799

Estimated Population (millions) 6 6 5.5 7.0 8.0 7.0

11.0 11.0 10+

Total (million £) 40 42 43.5 44.0 49.9 64.0 125.0 52.5 1 25.0“ 217.0 243.0 b 195.0

Per Capita (£) 6 13s. 4d. 7 7 18^. 8 7 10s. 20 approx. 22 20 approx.

4. EARLY ESTIMATES IN FRANCE: BOISGUILLEBERT AND VAUBAN to delve into the subject of national income, was one of the world’s poorest statisticians even for the time when Political Arithmetic was in its swaddling clothes. Boisguillebert made sound and devastating criticisms of the French fiscal system in conjunction with his proposals for reforming it, but he was completely ineffectual when it came to dem­ onstrating his ideas statistically.

National income estimating began much later in France than in England and de­ veloped there at a much slower pace. This was so for several reasons. Most important, the basic information necessary for the preparation of national income estimates was decidedly inferior. Public records were generally more poorly kept and were less accessible for scrutiny by private individ­ uals than they were in England. In par­ ticular, France’s tax system, incorporating vast exemptions for the rich and highly arbitrary assessments, yielded much fewer data on incomes than England’s more uni­ versal and proportionate tax system. In ad­ dition, the French government was much less inclined to tolerate private inquiries into the country’s resources than the more demo­ cratic government of England. It is signifi­ cant that the first explorations of France’s national income and wealth were made and published in England, and when the first two estimates were attempted in France, the government suppressed them. Obvi­ ously, the atmosphere was anything but conducive to free scholarly inquiry. The development of national income esti­ mating in France began in the 1690’s with the appearance of some broad guesses made by Boisguillebert in one of his major eco­ nomic treatises. These were followed in the first decade of the eighteenth century by more precise but only partial estimates pre­ pared by Marshal Vauban. For the remain­ ing first half of the century, no work of any importance on French national income, either theoretical or statistical, was done.

a. A Mild Man Turning into a Severe Critic of His Government Born in Rouen, the capital of Normandy, Boisguillebert spent his entire life in this small provincial city. He received a good classical education and had a flair for writ­ ing. In 1674, at the age of 28, he published a translation from the Greek of a twovolume work of Xenophilus, following it the next year with the publication of a similar translation from Herodotus. Next, he published a four-volume historical novel on Mary Stuart, Queen of Scotland. While engaged in literary work he entered public life, for by 1690 he had risen to the office of lieutenant governor of his city. By that time he had become interested in social and economic problems. The inefficiency and inequities of the existing tax system at­ tracted his particular attention and he be­ came intent on persuading the government of the necessity of revising the system com­ pletely. He began to write a book. It was his contention that since 1661, when Louis XIV decided to be his own prime minister, the government of France had been treating its citizens as if they were a conquered people. The King’s ministers, he said, were “bent on raising revenue espe­ cially for extraordinary expenses, not by increasing the incomes and wealth of the people, but by decreasing them everywhere and destroying them almost completely in certain cases.” They were “following but one certain rule: that of destroying twenty

1. Boisguillebert— Good Economic Theorist but Poor Statistician Pierre le Pesant de Boisguillebert (16461714), the first economic theorist in France Notes begin on p. 519. 52

4.

EARLY ESTIM ATES IN FRANCE

units of a private income in order to secure one unit for the King.” Boisguillebert urged a return to the earlier precepts of Henry IV, that “the wealth of the subjects consti­ tutes the whole basis of the State’s pros­ perity,” that “the Sovereign can receive as revenue only a proportion of the income of his subjects,” 1 and that his true interest lies in maximizing the incomes and wealth of his people by wise fiscal and other measures. Boisguillebert advocated repealing all the existing taxes except two. He wanted to keep the royal “taille,” revising it so that it would be rid of its existing exemptions and could be levied on all landowners, farmers, and workers proportionately to the amounts of their property and income.2 He proposed that it be collected by locally elected offi­ cials on the basis of annual declarations. He also wanted to retain the poll tax, imposing it on all adult individuals at a low uniform rate.3 He maintained that under his system, the State would receive more revenue, while agriculture and industry would once again grow prosperous. At the same time Boisguillebert also criti­ cized the government’s mercantilistic poli­ cies, advocating instead a laisser faire atti­ tude towards private enterprise. Although he did not mention the Law of Supply and Demand as such, he implied its basic rule. He spoke of “balance,” “harmony,” “pro­ portion,” and “equilibrium” being achieved when economic competition among men is allowed free reign. He emphasized that “every exchange must be profitable to both parties” as they “dispense with their super­ fluities in return for things they lack” and that the exchange must be made at mutually satisfactory prices. He contended that “such mutually profitable exchanges and the im­ provement of the productive efforts gener­ ally can take place only under conditions of competition and freedom of producers.” 4 Boisguillebert was the first economist to criticize vigorously the policy inaugurated by Colbert in 1650, then still pursued by the government, of prohibiting exports of grain. He contended that it interfered with the growth of French agriculture and should be reversed in favor of the former policy of complete freedom in the exporting and importing of grain.0 Sixty years later in 1757, Quesnay and other physiocrats took the same position (see Chapter 5, Section la ). Yet it was not until 1861, more than two hundred years after its enactment, that

53

this highly debatable economic policy was finally abandoned.

b. Official Disapproval of Boisguillebert’s Views Having worked out his main ideas, Bois­ guillebert obtained an audience in Paris with Pontchartrain, the Minister of Finance. Due Louis de Saint-Simon (1675-1755), the famous historian of the reign of Louis XIV and contemporary of Boisguillebert, described the interview and its consequences as follows: Anticipating the Minister’s quick temper, Boisguillebert opened his talk by begging the Minister to be patient, saying that he undoubt­ edly will first take him for a fool, but next will find some merit in his observations and will in the end agree completely with his proposals. The Minister after listening to Boisguillebert for a while and being taken aback by so much free advice being offered to him and being inclined to be caustic, suddenly exploded in laughter saying to Boisguillebert that he cer­ tainly agreed with the first point of his intro­ duction and thereupon brusquely terminated the interview. Boisguillebert returned to Rouen, not at all disconcerted over his rebuke, and proceeded even more persistently than ever with the completion of his book. He took it next to the new Finance Minister, Chamillart, who found considerable merit in it and spent several sessions with Boisguillebert trying to put his project into adoptable form. Encour­ aged by this reception Boisguillebert took his manuscript to Marshal Vauban without know­ ing that the latter was working on his own book containing some of the same ideas. Vauban, being more interested in achieving practical re­ sults of benefit to the people than in asserting his own authorship, lent help to Boisguillebert towards the perfection of his book. . . . In the meanwhile Chamillart informed Boisguillebert that the outbreak of the war ruled out all pos­ sibility for the immediate effectuation of the latter’s project.6

In 1697, Boisguillebert anonymously pub­ lished his book abroad under the title Le Detail de la France. During the previous year he had anonymously published an abridgement of it, also outside the country, under the title La France ruinee sous le regne Louis X IV (France Ruined under Louis XIV). The book passed unnoticed in France, and he followed it in 1707 with the publication of another work called Factum de la France, and with several shorter pamphlets, including one called Le SupplSment au Detail de la France in which he urged the government, in vigorous terms,

54

THE INCO M E OF NATIONS

not to delay the imperatively needed tax reforms till after the war.7 In the meanwhile, Marshal Vauban com­ pleted his own book, La Dime Royale (The Royal Tithe), and published it in 1707. The appearance of Vauban’s book created a sen­ sation at court, and he was promptly ban­ ished from it. The Marshal was pronounced to have lost his wits, and his book was ordered suppressed by government edict. All existing copies were to be destroyed and its circulation was prohibited on the threat of strict penalties. The stir over Vauban’s “in­ discretion” incidentally brought Boisguille­ bert’s works to official notice where they suffered the same fate—suppression. Bois­ guillebert himself was dismissed from his official post at Rouen and was ordered into exile to Southern France. This punishment, however, did not prevent Boisguillebert’s continuing to publish his own works and republishing Vauban’s clandestinely abroad.8 c. Boisguillebert’s National Income Estimates Boisguillebert’s main contention was that ever since Louis XIV assumed control of the affairs of state in 1661 when France was at the apex of her economic prosperity, she had been declining steadily. In his first book he estimated the national income of France for 1661 at 1,400 million livres and the loss of her annual income by 1690 at 500 mil­ lions, thus implying that national income during that year stood at 900 million livres. This was a gross underestimate, as the ec­ clesiastical tithe alone amounted to consid­ erably more than 100 million livres a year. He estimated the public revenue for 1661 at 85 million livres and for 1680 at between 112 and 115 million livres, thus contend­ ing that the tax burden, relatively to na­ tional income, had increased sharply. He gave no details for his estimate, saying merely “if there is any error in this compu­ tation, it could not be very large.” 9 Bois­ guillebert’s estimate of a 500-million-livre decline in national income over the three decades in question seems to have been based solely on the claim that “the prices of all properties have declined over that period by about one half.” 10 Even if his estimate of the decline in property values was cor­ rect, it would not necessarily mean that all prices and incomes had declined corre­ spondingly. Ten years later, in his second book, Bois­

guillebert corrected his low estimate, assert­ ing that the national income in 1661 had been 3,000 million livres and that it had declined 1,500 million over the intervening 40 years. He offered no better statistical documentation in support of this estimate than he had for the one before. The follow­ ing passage from his second work is typical. The loss of one half of the wealth of France as much in real property as in income from labor [since 1660] . . . has as many witnesses as there are people in this kingdom, not to speak of the evidences furnished by registers of deeds and contracts. . . . It is maintained here that the national in­ come has decreased since 1660 by more than 1,500 million livres: let the reader not be aston­ ished by this figure of hundreds of millions of livres! Just as it is possible to calculate the income produced by a house, farm, or a vil­ lage, both as it increases and decreases, so it is possible for any person competent in such mat­ ters, to compute the income of an entire king­ dom. That income has been computed in Eng­ land, which cannot be but one-fourth of that of France, in whatever way it may be measured . . . and it is pretended to amount there to 700 million livres. As for France, people who will specialize in this type of calculation will find that a country which furnishes a revenue to the State of 150 million livres in some years and to the church one amounting to more than 300 million livres a year, must have a national income running into many hundreds of million livres.11

It is clear that in this passage Boisguille­ bert referred to the estimates of English in­ come appearing in the books of Charles Davenant that must have been available in France at that time. It is also evident that this 1,500-million-livre figure for the de­ cline in national revenue (a figure which Voltaire ridiculed in 1751 in his Century of Louis XI V) 12 was exaggerated. It is unques­ tionably true that a substantial decline in national income had taken place, but it could hardly have been of such proportions. It was due to a sharp drop in the price level.

2. Marshal Vauhan— Military Genius Turning into a Fiscal Reformer and National Income Estimator Marshal Vauban (1633-1707), who was the second Frenchman to attempt the meas­ urement of his country’s national income, was neither a good economic theorist nor

4.

EARLY ESTIM ATES IN FRANCE

a good statistician. But, being an engineer by training, he brought to the task the same respect for precision that characterized his military work. He had few data with which to work, and in the end succeeded in measuring only certain parts of the na­ tional income rather than the total. Even though his estimate was only a partial one, he must be regarded as the founder of the statistical approach to the measurement of social and economic phenomena in France. a. Vauban’s Life Born in Burgundy of a noble but impov­ erished family, Vauban was a self-educated man. He entered military service at the age of seventeen, where he displayed a penchant for engineering and an ability to master the hardest problems in geometry and physics. He became interested in the building of fortifications, exhibiting a unique genius in that field, and playing a major role in the construction of most of the fortifications of the time, including the great ones of Dun­ kirk, Lille, Metz, and Strasbourg. Vauban also developed new tactics for attacking and defending fortifications and he consistently took command of the operations on the battlefield. Wounded in battle many times, appointed director of all of France’s fortifi­ cations, decorated by Louis XIV with every conceivable order of distinction, and rec­ ognized throughout the civilized world as the greatest military engineer of the time, Vauban was named a Marshal of France in 1703. A great patriot, and one of the most gen­ erous and unselfish of men, Vauban had broad interests extending far beyond engi­ neering and military science. He was a firm believer in enlightened absolutism and was always searching for ways to increase the glory of France and the well-being of her people within the framework of a benevo­ lent monarchy. During his lifetime he drew up many memoranda on a variety of sub­ jects that he submitted directly and privately to his beloved King. After he had retired, he collected the more important of these memoranda in a manuscript of some ten or twelve volumes, under the title Oisivetes (Idlenesses). Some of these volumes got lost; others were published only in the nine­ teenth century.13 During his retirement, Vauban became deeply disturbed over the social, economic, and fiscal problems of France, grieving par­ ticularly over the marked impoverishment

55

of the common people. Like Boisguillebert, whose work he read in manuscript form be­ fore its publication, he ascribed their misery primarily to the inequities of France’s tax system. He agreed with most of Boisguille­ bert’s diagnoses of its weaknesses and also with most of his prescribed remedies, but he sought to give these remedies greater precision and to demonstrate their practica­ bility more forcefully. In the Preface to the Dime Royale, ad­ dressed to the King, Vauban described the contemporary social and economic condi­ tions of France in the following dismal terms, which could have hardly pleased the King. I have found that in recent years close to one tenth of the population has been reduced to a state of beggary; . . . of the remaining nine tenths, five are unable to give alms to the latter, because they themselves have been re­ duced to pretty nearly the same unfortunate condition; of the remaining four parts three are badly off, being encumbered by debts and law suits; and that the tenth part—embracing all officers of the army, magistracy, clergy, high nobility, the whole military and civil govern­ ment personnel, the upper merchant class, the middle class “rentiers”—does not contain even 100.000 families; and I am sure that I do not exaggerate if I say that there are not even 10.000 families large or small, that can be said to be in absolutely comfortable situations.14

b. Vauban’s Proposed Tax Reform Vauban’s remedial program for the re­ moval of tax inequities and frauds was founded on the “chief principle . . . that all incomes ought to contribute to the sup­ port of the State proportionally” 15 instead of having the small incomes carrying al­ most the entire tax burden as was the prac­ tice at that time. In his simple scheme Vauban would have abolished most of the existing taxes, substituting for them a “Royal Tithe” 16 or gross income tax, universal in its scope, levied at a uniform rate of one twentieth to one tenth of income, depending on general economic conditions. The tithe was to consist of two parts, each of which would constitute a separate “fund.” The first part would be a levy on all agricul­ tural produce, and would be collected in kind. These collections would form the “First Fund” of the State. The second part would be a levy on all money incomes. This would be collected in money and would constitute the “Second Fund.” It would apply to incomes from the rent of houses,

56

THE INCOM E OF NATIONS

profits of businesses and grain mills, the operation of public properties, m oney wages, pensions, and fees of governm ent offices. In addition to this basic incom e tax, he pro ­ posed to retain the tax on salt, separating its proceeds into a “T hird F u n d .” H e also suggested keeping, w ith some modifications, the existing system of registration fees and other adm inistrative charges as well as the profits of the public dom ain, postal charges, custom s duties, and taxes on luxury goods such as “tea, coffee, chocolate, brandy, gilt coaches . . . huge and ridiculous P eri­ wigs,” etc. T o these he proposed to add a tax on wine, cider, and beer consum ed in public houses. These revenues were to con­ stitute a proposed “F o u rth F u n d .” V auban’s proposals w ere m ore practical than those of Boisguillebert, particularly his levy in kind on agricultural produce, which w ould have been easier to collect than Bois­ guillebert’s m oney levy. V auban’s ta x re­ form program m ay have been economically practical but it was certainly n o t practical politically. Effectuating his scheme would have deprived the “farm ers-general” and their num erous assistants of their jobs. T heir influence and th at of their friends at court was enough to bring V auban into disgrace. Crushed by his estrangem ent from the King, V auban died a few m onths later.

c. Estimates of the Revenue from the Tithe and of Agricultural Income T o prove the practicability of his system, V auban estim ated each category of taxable incom e, and the am ount of revenue that w ould be raised from each at various rates. V auban proposed that the largest am ount of tax revenue should com e from the tithe on agriculture. This was to be sim ilar to the existing ecclesiastical tithe, but levied at a different rate. It was to be collected in kind directly from the cultivators before paym ent of the ecclesiastical tithe o r of any ren t to the landowner. Inasm uch as the rate of the tithe was to be definite and fair, V auban felt th a t the farm ers would pay it m uch m ore willingly. H e envisaged th at m any individuals of hon­ orable reputation, including m em bers of the clergy and the local gentry, w ould apply for the office of “tax farm er” and perform the required duties for a m oderate rem unera­ tion. H e also suggested th a t the state pro ­ vide a barn in every parish w here farm ers could deposit their tithes in kind. V auban com puted the probable yield of

his royal tithe in k ind on the basis of an estim ated value o f the annual produce of lan d fo r the entire country. F ro m the exist­ ing records of the departm ental “intendants,” he calculated th e total area of the nation and the p roportion of th at area that was arable. H e found th e total area to con­ tain approxim ately 30,000 square leagues, o f w hich 24,000, o r fo u r fifths, w ere culti­ vated, o r at least kept as pasturages or as fallow land. H e then determ ined the average yield of crops p er square league from the records fo r certain sections of N orm andy, taking into account the poorest as well as the m ost fertile lands. H e projected th e re­ sulting figure of the value of the crops for N orm andy into a total fo r the country ac­ cording to the proportion th a t N orm andy’s arable area bore to th a t of the nation. His calculation of th e revenue yield of a oneten th tithe in N orm an d y was 12 million livres. F o r the entire nation, figured on a p roportionate area basis, he expected the yield at the same one-tenth rate to am ount to 120 m illion livres.17 T his would indicate th a t he estim ated the annual value of the crops fo r all of F ran ce to be 1,200 million livres. V auban verified his calculation of the yield of th e tithe in k ind by another m ethod, likewise applied experim entally to N orm an d y and then blow n up to national proportions. F ro m local records covering certain N o rm an parishes, he found th a t the ecclesiastical tithe, levied at one eleventh, was yielding 5,600 livres in revenue per square league. H e m ultiplied this sum by th e 24,000 square leagues o f cultivable land in th e country as a whole, arriving at a total probable revenue from the ecclesiastical tithe of 134.4 m illion livres.18 A pparently, in o rd er to be conservative, he next reduced this estim ated yield to 120 million livres, th u s obtaining a total corresponding to the one he had reached by the first m ethod. A century o r so later one national incom e estim ator (G a n ilh ) criticized V auban’s cal­ culations as being inflated on the ground th a t they were based on the yields of one o f the m ost fertile provinces o f France.

d. Vauban’s Proposal for Annual Enumeration (Population, Income, and Wealth) T he second category o f incom es to be “tithed” u n d er V au b an ’s plan consisted of all nonagricultural incom es. V auban insisted th a t these incom es be subject to taxation,

4.

EARLY ESTIM ATES IN FRANCE

as otherw ise “m ore th an one half o f the total income of the inhabitants of the K ing­ dom w ould be paying nothing tow ards the support of the State.” This, he said, “would be an obvious injustice to the others; be­ cause all citizens being equally protected by the State, each of them has an obligation to contribute to the needs of the State in pro ­ portion to his incom e— w hich is the basis of this System.” H e em phasized th e p rin ­ ciple of taxation according to ability to pay, saying: . . the m ore a person rises above others in station by birth o r own dignity, and the m ore property he owns, just that m uch m ore is he in need of the protection of the State, and is vitally concerned to m ake sure th a t the State is m aintained in honor and authority— a thing w hich can be accom plished only w ith the aid of an ap­ propriate expense. T o accom plish this [uni­ versality of contribution] all th a t is neces­ sary is to ascertain the incom e of each person so th at his due tax m ight be appro­ priately calculated.” 19 V auban proposed th a t each person be re­ quired by law to declare the am ount o f his income, subject to severe penalties fo r fail­ ing to tu rn in a tru e report. These returns would then be exam ined by com petent stateappointed individuals receiving adequate re­ m uneration fo r their w ork.20 A t the same tim e, V auban was well aw are o f the fact th at the enforcem ent of any such plan hinged upon the existence o f an accu­ rate enum eration of all the inhabitants of the nation w ith an appropriate distinction between those receiving incom e and those w ho w ere dependent upon them fo r their sustenance. H e classified all inhabitants into three categories— “m en o f arm s,” including the nobility; “m en of gowns,” em bracing m agistrates, other adm inistrative officers, and the clergy; and the com m on people, including m erchants, artisans, farm ers, and com m on laborers. Such an enum eration, he said, w ould n o t be difficult. A ctually, it w ould have been available had each m em ber of the clergy perform ed his p roper duties in keeping an accurate and curren t record o f all the souls in the parish assigned to his charge. In the absence of such accurate church records, V auban proposed the insti­ tution by the State of an orderly enum era­ tion system. H e recom m ended th a t in each parish an individual be designated by the State to keep an annual register of all in­ habitants, recording in each case the inhab­ itan t’s occupation o r social status and the num ber in his household—wife, children,

57

servants, etc. H e proposed to supplem ent this register w ith another one setting forth fo r each parish the num ber of occupied hom es; the num ber of uninhabited, ruined, and newly built houses; the num ber of horses, oxen, cows, pigs, sheep, fowl, etc.; the area u n d er cultivation, in pasture, or left fallow; the n um ber of vines; the acreage of w oodlands, and the volum e of cut tim ­ ber; the num ber of mills of various kinds, taverns, etc. These local registers would then be tu rn ed over to the officials of the larger jurisdictions and eventually to the central authorities of the nation to be con­ solidated into provincial and national tabu­ lations. V auban em phasized th at such a curren t annual reco rd of the population and its w ealth w ould be o f im m easurable value to the King and to his m inisters not only fo r tax-collecting purposes, b u t also fo r form ulating policies fo r the advance­ m ent of the w elfare of the nation generally. T he individuals preparing the original parish records w ere to be given a distinguished title, and w ere to receive a m oderate com ­ pensation from the m em bers of the parish th a t could be payable in kind if n ot in money. Leaving no detail to chance, V auban even drew up suggested schedules fo r the parishes an d provinces to use in recording and consolidating the statistical data.

e. Vauban’s Estimates of Revenue from Proposed Taxes on Specific Money Incomes In the absence of such detailed inform a­ tion on the num ber, occupations, and w ealth o f the people necessary fo r the estim ation o f th e revenue from th e m onied incomes o r “Second F u n d ,” V auban prepared, as best he could, his own estimates of occu­ pational groups and their average incomes from reports of the “intendents” of the provinces fo r the preceding years and from other sources. H e calculated the total popu­ lation at 19.1 m illion.21 F o r some groups he prep ared no estimates of incom e on the ground th a t they w ould be taxed u n d er the agricultural tithe. Thus, he did not estimate directly the rents received by landlords, the interest received by mortgagees, o r the in­ com e of the church from other sources th an its tithe. In some cases, such as w ater transportation, he estim ated the revenue di­ rectly, w ithout first calculating the income. V auban crudely estim ated the am ount of rent produced by city and town dwellings.

58

THE INCO M E OF NATIONS

H e included m oney rents paid by tenants and rents in kind received by ow ners occu­ pying their ow n dwellings. H e figured that there were some 800 cities and towns in the country, with an average of 400 houses each, giving a total of 320,000 city and tow n dwellings. H e estim ated the average annual rental value of each house, after deducting one fourth of it fo r the cost of repairs, at 100 livres, giving him a figure fo r rental income from such dwellings of 32 million livres and a tax revenue, calculated at a rate of one twentieth, of 1.6 m illion livres.22 V auban estim ated the incom e fro m grain m ills by assuming th a t there were two mills fo r every square league, o r 60,000 in all; and th at the incom e of each mill for the support of the ow ner and his assistants, after allowing one fourth for repairs, was ap­ proxim ately 248 livres. This indicated a total income from the country’s grain mills of 14,850 thousand livres. V auban m entioned other types of mills— sawmills, foundries, oil mills, hem p mills, textile mills, paper mills, pow der mills, etc., but m ade no at­ tem pt to estim ate their incomes, merely saying that they w ould produce a substan­ tial revenue. H e com puted the revenue from the tithe on owners of ships and barges at the rate of 5 sous (25 centim es) per ton, apparently assuming th a t a tax at this rate would bear a p roper ratio to the incomes of the owners. H is estimate of revenue from the tithe on fisheries was given in a lum p sum w ith­ out any explanatory details. V auban had no difficulty in getting the figure of the interest received by the holders o f governm ent bonds. H e set it at 20 m illion livres, and he estim ated the incomes of state officials and pensioners at 40 m illion livres. M uch m ore difficult was the calculation of the total wages o f dom estic servants rang­ ing from those of the lowest station to the stewards of great estates. V auban not only believed that servants should pay a propor­ tionate tax on their wages, but th a t those masters who were not paying their dom es­ tics should be required by law to do so. A t any rate, he estim ated the num ber o f do­ mestic servants of both sexes at 1,500,000 and their total wages, figured on the basis of a m inim um average rate of 20 livres a year, at 30 m illion livres, w hich w ould pro­ duce a revenue of 1.5 to 3 millions. N ext, V auban calculated th e incom es fro m fees earned by various types o f public officials, judges, prosecutors, notaries, regis­

trars, etc. These em olum ents exceeded 1,500 livres a year in m any cases, and the num ber of such officials ran into m any thousands. H e estim ated the total sum at 10 million livres, producing a tax revenue of from 500,000 to 1 m illion livres.23 In co m e fro m com m erce was estim ated by V auban at 40 m illion livres a year. This figure apparently was a deliberate u nder­ statem ent on his part. V auban was exercised over the unhealthy condition of com m erce an d industry in F ran ce and sought to rem ­ edy it by taxing these groups as lightly as possible. Thus, h e distinguished between “useful” and “n ot useful” com m erce and proposed to im pose m inim um taxes on the form er, in violation of his own principle of universally proportional taxation. V auban fo u n d it m ost difficult to esti­ m ate the incom e of w orkm en— both artisans an d com m on laborers—who, together w ith their families, com prised nearly h alf of the total population, o r about 8,250,000 per­ sons. Tw o thirds o f them , however, were aged people, w om en and children who were n o t earning m uch, if any, income. H e cal­ culated the productively engaged m em bers of the labor group at 2,750,000. However, he deducted some 750,000 from this num ber fo r people w orking in the fields and vine­ yards w ho, he said, w ere indirectly taxed through th e tithe on th e produce of the soil. Consequently, h e was left w ith the figure of 2 m illion representing the artisans and com m on laborers w orking in th e cities, towns, and villages. U sing inform ation on the earnings of representative occupations, he arrived at an average wage fo r the entire laboring group o f som e 90 livres a year, m aking a total wage incom e o f 180 million livres. V auban believed, however, th a t the people included in this group w ere living at the m inim um subsistence level and could n o t b ear th e regular rate of taxation. H e proposed, therefore, to tax them at only one thirtieth of th eir incom e, yielding a rev­ enue of 6 m illion livres. Thus, all th e revenue sources, w hich to ­ gether constituted th e second fund, were estim ated by V auban to produce u n d er a tax of one tw entieth of incom e (w ith the excep­ tions noted) the sum of 15.4 m illion livres.24

/. Estimates of Revenues of the Third and Fourth Funds A lthough V auban recognized the objec­ tionable features of the salt tax, h e consid­ ered them to be outw eighed by its advan­

4.

EARLY ESTIM ATES IN FRANCE

tages as a universal levy, and proposed to retain it. H e w rote: The Salt is a Manna, which GOD has given as a Blessing to Mankind; and therefore it would seem that no Imposition ought to be laid upon it. But there being a necessity for Levy­ ing Money upon the People, for supplying the pressing Occasions of the State, there has been no better Expedient thought of for doing it, with a just Proportion, than by laying an Im­ position on the Salt; because every Family must make a consumption of it, less or more, accord­ ing to their Number and Circumstance; the rich having many servants and eating well, using much more of it than the poor. This is why there are few states where they do not place an Imposition on Salt, but they do so more moderately than in France, where it is very badly regulated.25 If im posed at a m oderate rate, V auban felt th at the salt tax was n o t as objectionable as it seemed at first glance. But, w hile rec­ om m ending the retention of the tax, V auban proposed a num ber of adm inistrative m eas­ ures th a t would correct its glaring defects, and suggested th a t its rate be cut in half, if n ot m ore. T he first, second, and th ird funds were to com prise the variable o r flexible revenue of the State. T he rates of their levies were to vary according to the requirem ents of w ar or peace o r other changes in th e con­ ditions of the country. T he fourth fund, on the other hand— com prised of revenues from the public dom ain and public enterprises, customs, excises, and registration dues— was to constitute the fixed revenue of the State. T he rates of its charges were not designed to vary with political and econom ic condi­ tions.

g. Vauban’s Estimate of Total Revenue as a Base for Estimating National Income V auban sum m ated the revenues of the four funds “at a Royal tithe o f one tw en­ tieth” in the following tabulation: 26 Livres First Fund, or tithe of the fruits of the earth 60,000,000 Second Fund, or tithe on incomes from homes, invested capital, in­ dustry, commerce, and labor generally 15,422,500 Third Fund, or Tax on Salt at 18 livres per minot 23,400,000 Fourth Fund, or fixed revenue 18,000,000 Total

116,822,500

59

This tabulation seems to have suggested to some later w riters th at V auban estimated the total incom e of the people of F rance from all sources at twenty times this sum, o r at 2,336,450,000 livres. However, he gave no specific figure fo r aggregate income to m atch his figure fo r total expected revenue. Such a sum m ation of incom es either was outside his im m ediate interest, w hich was to dem onstrate the productivity of his pro­ posed tax system, o r else, as is very likely, he found th a t the task required statistical adjustm ents and reconciliations that were beyond his powers. Some thirty years after the publication of V auban ’s estim ate, however, the economist D utot, in his book Reflexions Politiques sur le C om m erce et les Finances 27 attributed a national incom e estim ate of 2,336 million livres to V auban. D u to t did n ot give any exact reference to the portion of V auban’s book in w hich he found this figure; n o r did he give the com ponents of this alleged cal­ culation.28 Eugene D aire, who republished V auban’s w ork in 1843, challenged D u to t’s statem ent, m aintaining th at he could not find any such estim ate in the 1707 edition of V auban’s w ork.29 D utot m ust have de­ rived this figure indirectly from the various items in V auban’s com putation. A t any rate, we likewise have not been able to find this, o r any other, sum m ation of incomes in any edition of V auban’s w ork th at we have been able to examine. O ur own addition of V au­ b an ’s figures does n ot produce the total quoted by D utot. N o r does the list of the types of incom e given by V auban appear to be com plete enough to w arrant calling their total a national incom e total.

h. Appraisal of Vauban’s Estimates T here is a basic inconsistency between V auban’s com putation o f incom e from land at 1,200 m illion 30 livres, the total nonagri­ cultural incom e of 400 m illion livres, and his statem ent, m ade m ore th an once in his text, th at nonagricultural incomes are about equal to agricultural ones. V auban’s esti­ m ates w ere obviously crude. T he statistical data available at the tim e were n o t adequate raw m aterial fo r producing m ore refined estimates, nor was V auban equipped by his experience and station in life to carry out the type of detailed calculations required fo r the preparation of a first-rate estimate. Some elements in his calculations may have been grossly overestim ated, others seri­ ously understated, but V auban had no guide-

60

THE INCOM E OF NATIONS

posts by which to check his calculations. For example, his estimates of income from agricultural production were conceptually closer to “gross income” than to “net in­ come.” He made no attempt to deduct from the gross value of production (total produce multiplied by price), the value of seeds set aside for next year’s production, of feed furnished to working animals, costs of maintenance, repairs and replacement of permanent farm structures, and the costs of replacement of stocks of animals de­ pleted by age, illness, and mortality. Vau­ ban, like his contemporaries, did not seem to be aware of the need for distinguishing between the “gross” and the “net” value of agricultural production. No one in France —not even Boisguillebert—had pointed out the distinction at that time. Another fifty years or more had to elapse before it would be made by Quesnay in the second half of the eighteenth century. Although Vauban’s estimates were some­ what fragmentary in nature and not too well co-ordinated, they were an important mile­ stone in France’s development of national income estimating. They were important be­ cause (a) they represented the first concrete attempt in France to estimate, if not the entire national income, at least its main elements; and (b) because they forcefully demonstrated the utility of national income estimates and the importance of government participation in the organization and main­ tenance of better statistical records so that scholars could perform their estimating tasks more completely in the future.

3. A Lapse of Interest in National Income Analysis ( 17 0 7 -5 8 ) During the fifty years or so following the death of Boisguillebert and Vauban, no im­ portant analysis of the national income of France appeared. Public interest and schol­ arly explorations in political economy dur­ ing this period shifted from fiscal and com­ mercial problems to monetary and banking problems. This was a natural reorientation in the light of the history of those years. Louis XIV, upon his death in 1715, had left his successor a ruined agriculture, industry, and commerce, an empty treasury, and a huge public debt. The debt amounted to 2.4 bil­ lion livres,31 which was just about equal to

the national income suggested by Vauban’s revenue calculations. More than half of the State’s revenue was siphoned off to pay the interest charges on the debt, making the orderly financing of the State’s other re­ quirements impossible and producing new and ever larger annual deficits. The virtually bankrupt State proceeded with a series of repudiations of its debts, devaluations of its existing paper currencies, and the issue of new paper money. To the utter distress of the common people, currency values were sinking while the price level rose sharply. John Law (1671-1729), the financial wiz­ ard who promised to save the state through colossal banking operations, failed disas­ trously in his undertakings, dragging down with him thousands of innocent investors. But Law gave birth to a whole school of political economy that continued to search for a solution to the ills of society through monetary measures.32 His ablest disciples were Melon (1675-1738) and Dutot (exact dates unknown). Dutot made some observa­ tions on the national income of his time (his book appeared in 1738; the third edi­ tion in 1754), but these observations were made by him only incidentally to his ex­ position of the relationship that the quan­ tity of money and the rapidity of its cir­ culation must bear to the trade of the na­ tion, its income, and turnover of wealth. Dutot wrote: 33 Livres Marshal Vauban in his Dime Royale shows us th at the national income of the country amounted to 2,336,450,000 Goevin de Rademont, royal tax receiver, in his Traiti de la Dime Royale says th at the total annual income of the country from all properties, houses, and enterprises amounts to 2,494,939,194 Taking the middle between these two figures we arrive a t a figure of 2,415,694,597 This income, distributed among the 20 mil­ lion people contained in France, yields a per capita annual income of only 120 livres, or 6 and sous a day, which shows that the income is not large and, in fact, is insufficient.

Thus, it is clear that in 1754 inquiries in France into her national income had not advanced beyond those made by Vauban and Boisguillebert, fifty years before.

5. CONTRIBUTIONS OF FRENCH PHYSIOCRATS: QUESNAY, LAVOISIER, AND OTHERS By the late 1750’s the excitement over irre­ sponsible government credit and money policies and the equally reckless private banking and investment operations of John Law, including his Mississippi Bubble, was ended. Scholars once again turned their at­ tention to considerations of the fiscal prob­ lems of the state, the inequities of the tax system, and the size and distribution of the national income. Many inquiries were made along these lines in the last decades of the monarchy and especially during the Revo­ lution. The most important studies were those emanating from the physiocratic school of political economy, particularly those made by Quesnay, Du Pont, Lavoisier, Le Trosne, and Lagrange.

for that matter, any other great agricultural country. The Philosophie Rurale contained even more of such hypothetical estimates. To what extent some of them were designed to represent the actual national income of France at that time is not clear. Nor is it clear which of the figures in the Philosophie Rurale were Quesnay’s and which might have been Mirabeau’s elaborations of his master’s calculations. Quesnay attempted to devise the tools that would make it possible to estimate the size, flow, and distribution of national in­ come for any country at any time. Whether he succeeded or not is relatively unimpor­ tant. What is important is that he attempted to do it. Future economists and statisticians developed better methods for estimating na­ tional income, but they merely proceeded, more successfully it is true, upon Quesnay’s original proposition that all economic so­ cieties have a common physiology whose functioning lends itself to standardized analyses, both qualitative and quantitative in nature. Quesnay’s “Tableau Economique,” as he called his tabular analysis, was accepted by his followers as one of the most important inventions of the age, equal in importance to Newton’s discovery of the law of gravi­ tation. Without attempting to dispute their enthusiastic evaluation, it is plain that the “Tableau” was revolutionary in its implica­ tions. Its chief characteristic was that it treated national income in dynamic terms as a flow of funds and products from one sector of the economy to another, showing the various transactions occurring in the course of this flow in a series of interrelated accounts. Quesnay’s “Tableau” anticipated by just about two hundred years our modern statistical analysis of national income and expenditure by “sector accounts.” His anal­

1. Quesnay’s Hypothetical Estimates and Invention of the M oney-Product Flow Analysis Francois Quesnay (1694-1774) is gen­ erally acknowledged to have invented the money-product flow analysis of national income—the approach that is at the very core of modern economic analysis. His con­ temporaries (Voltaire, Lavoisier, and oth­ ers), as well as some early nineteenth-century writers (Ganilh, etc.), credited him with having made a definite estimate of the national income of France, or rather of her gross agricultural income, in the Philosophie Rurale. This work, originally published anonymously in Amsterdam, and generally listed under Victor Mirabeau’s name, is supposed to contain a substantial amount of Quesnay’s writing. Quesnay developed a whole series of hypothetical national in­ comes postulated on a variety of situations that could arise in the life of France or, Notes begin on p. 520.

61

62

THE IN CO M E OF NATIONS

ysis of national income in terms of hypo­ thetical cases was also a forerunner of the modern practice of constructing models of potential national income under varying conditions. In more than one way, Quesnay, like Petty, King, and Boisguillebert, was a pathfinder. The substance of his approach is not easily explained. We must, therefore, beg the indulgence of the reader in laboring with us through the following somewhat extended exposition of it.

a. A Physician Becomes an Economist Quesnay was born in the country, and was said by his contemporary biographers (Mirabeau and others) to have retained a rustic physical appearance and outlook throughout his life. Having developed an interest in medicine in his youth, he com­ pleted his formal education in surgery and established himself in practice in the prov­ inces. Being a scholar and a man of action, he was not content with the mere practice of medicine, and when he was still in his thirties, published his first book of scientific research. In this book he took issue with certain notions about the effects of bleeding held by some of the most powerful figures in the surgical profession. Prompted to move to Paris by the grow­ ing demands for his services, Quesnay be­ came Secretary of the then newly founded Academy of Surgeons—a position that en­ abled him to publish a variety of scientific articles on a wide range of medical subjects. These publications earned him election to membership in a number of scientific acad­ emies, including the Royal Society of Eng­ land. He became, first, one of the court physicians; next, the personal physician to Madame de Pompadour; and, finally, chief physician to the King. A man of wide in­ terests, he extended his studies to the fields of mathematics, political science, and eco­ nomics, and became a central figure among the philosophers and literary men of his time. Quesnay’s interest in political economy seems to have been prompted by his con­ viction that a good political, social, and eco­ nomic order was of utmost importance to the happiness of mankind. He applied to this field of knowledge the same methods of reasoning that he had so successfully used in his medical researches. As the fiscal situa­ tion was at the root of the nation’s difficul­ ties, he turned his attention to it in particu­ lar, analyzing it, however, in terms of the

total economic structure of society. His so­ lution was wholly original. He advocated replacing all existing levies with a single tax, or “impot unique,” to be levied on the rents paid by the cultivators to the landowners. Quesnay, like Boisguillebert, sixty years before him (see Chapter 4, Section la), took issue with the government’s policy of prohibiting the export of grains as a means of keeping grain prices and hence industrial wages low, and of stimulating the produc­ tion of luxury goods such as silk fabrics, lace, Gobelins, etc. He argued, instead, for free export of grain and free trade in gen­ eral, and for the maximum possible expan­ sion of agricultural production and of manu­ factured articles for general consumption, such as linens, other common textiles, leather goods, etc., as a better basis for national prosperity.1 Quesnay was the founder of “Physioc­ racy,” and the acknowledged leader of this school. He died, broken in body but vigorous in mind, a few weeks after having the satis­ faction of seeing his favored disciple, Tur­ got, elevated to the position of Minister of Finance.

b. The Theory of the “Tableau Economique” Quesnay’s most important economic work on the subject of national income was his “Tableau Economique,” written especially for the King’s consumption. Completed in 1758 and published two years later, its exact title was A nalysis o f the A rithm etical For­ m ula o f the E conom ic Table o f the D istri­ bution o f the A n n u a l E xpenditure o f an A gricultural N ation.

Quesnay opened his work with the state­ ment that “a nation consists of three classes of citizens: the productive class, the proprie­ tors, and the sterile class.” 2 He defined the first as being comprised of the cultivators of the soil; the second as including not only the landowners and their managers and other aids, but also the Church as the re­ cipient of the tithe, and the Sovereign, in­ cluding the entire civil and military estab­ lishment protecting life and property and promoting the welfare of the citizens. He defined the third or sterile class as including artisans, merchants, manufacturers and their workmen, professionals, artists, and all peo­ ple employed in other occupations. He made it clear that he called the last class “sterile” not because its members did no useful work, but because they merely reproduced the

5.

CONTRIBUTIONS OF FRENCH PHYSIOCRATS

value of their labor and of the materials they used without creating any true surplus. In analyzing agricultural production, Quesnay distinguished between three types of outlays or “avances” : a. for original improvements of land (“avances fonciers”) entailing clearing land of all obsta­ cles to cultivation, drying up marshes, planting of useful trees, and other works which, once done, require little if any subsequent redoing; b. for investments in structures, livestock, and working animals (“avances primitifs”) which require annual repair, maintenance, and re­ placement; and c. for annual operating capital, including both the cultivator’s advance of his own and his family’s labor, i.e., his and their annual con­ sumption outlays and his advance of the cost of primary materials such as seeds, feed, etc. (“avances annuels”), similarly requiring annual reproduction.

The last two kinds of advances, accord­ ing to Quesnay, constituted the cultivator’s annual reinvestment in reproduction (“re­ prise”). He estimated the annual repairs and replacement of worn-out agricultural capital necessary to maintain the original total stock at about one tenth of its capital amount, and insisted that if such repairs and replacements were not provided for, agriculture would eventually disintegrate. He also insisted that annual reproduction of operating capital must include the main­ tenance and replenishment of reserves of crops and feed necessary to protect the cul­ tivator against natural catastrophes such as drought, flood, hail, etc. In addition to providing this reproduction of advances, Quesnay claimed that the an­ nual cultivation must also provide (a) a “net product” or “net income” for the “pro­ prietors” (landowners, Church, and State), and (b) a product for the support of the sterile class in return for which the culti­ vators can obtain the manufactured prod­ ucts and services required by them in their work. The “net income” left to the land­ owners after the payment of taxes could be spent for their own consumption and for new investments in agriculture, including improvements of structures, additions to livestock, etc. When it is so invested, wrote Quesnay, the ability of the land to yield more net income is enhanced and every­ body is benefited—the landowner, the State, the Church, the cultivators, and the mem­ bers of the sterile class. In line with this notion, Quesnay and his followers advo­ cated that the State spend a portion of its

63

taxes for the improvement of agriculture as the surest means for enriching the nation and the State’s own treasury. From these general premises, Quesnay developed a set of figures representing the flow of advances, receipts, and “net in­ come” of a “great state operating at the highest state of its agricultural develop­ ment.” He defined this nation as having a population of 30 millions, an area of 130 million “arpents” (acres) of varying quality, a stock of wealth of 12 billions “necessary to sustain that territory in good condition,” and an annual production of 5 billions.3 These conditions represented an achievable potential for France, which at that time had a population of 22 to 23 millions, an area of 105 million “arpents,” and a gross national income, in Quesnay’s terms, of 3 billion livres. Quesnay insisted that his fig­ ures were hypothetical only in the sense that they related to a hypothetical case, but that otherwise they faithfully reproduced rela­ tionships actually existing in “nature.” The 5 billion product created by the pro­ ductive class in such a nation, wrote Ques­ nay, is produced with an annual outlay of 2 billions by the productive class, of which 1 billion represents its outlays for its own consumption and 1 billion its outlays for primary materials, including replacement of its reserves of crops. Of the remaining 3 billions, 2 are the income furnished to pro­ prietors and 1 is the agricultural product sold to the sterile class in exchange for 1 billion’s worth of manufactured goods and such services as transportation, trade, re­ pairs, professions, etc. The “proprietors” can spend one half of their 2 billion income for agricultural products and half for the prod­ ucts of the sterile class. However, instead of spending their 2 billions in this way, the proprietors can invest part of it in improve­ ments for agriculture, thereby guaranteeing its continuance. Quesnay described the circular flow of national income in terms of this particular “hypothetical case” as follows: The total of 5 billions, divided between the productive class and the proprietary class, being expended annually in a manner insuring the perpetual annual reproduction, resolves itself into 1 billion of expenditures by the proprietors for purchases from the productive class and one billion of their expenditures for purchases from the sterile class. The productive class which sells its products for three billions to the other two classes, provides two billion of this as income to the proprietors and spends

64

THE IN CO M E OF NATIONS

one billion for purchases from the sterile class; thus the sterile class receives two billions which it spends on purchases from the productive class for articles of consumption for its mem­ bers and for primary materials needed in its own production; and the productive class itself spends two billions of its product [for its own consumption and production advances], which completes the total expenditure or consumption of 5 billions of an annually reproduced product. Such is the regular order of the distribution of the expenditure of 5 billions which the pro­ ductive class causes to be recreated annually through the expenditure of 5 billions of annual reproduction.4

Quesnay then presented the following “arithm etical distribution of the expendi­ ture” in a “T ableau,” prefacing it w ith this explanation:

income and go down obliquely to the one [producer] and to the other [sterile] class. At the end of each line is the sum which the pro­ prietors expend in purchases from each of these classes. Reciprocal trade between these two classes [sterile and productive] is also indicated by dotted lines which descend obliquely from the one [purchaser] to the other [seller], . . . at the end of each line is the sum which each class receives from the other, in the course of their reciprocal trade, for their expenses. Finally, the calculation ends up on each side with the total sum of receipts of each of the two classes. It can be seen in the case given that if the distribution of expenditures follows the described pattern, the receipts of the pro­ ductive class, including its advances, are equal to the total annual reproduction, and that cul­ tivation, wealth and population remain un­ changed, neither increasing nor decreasing.5

At the top [left] . . . is the total of the ad­ vances of the productive class which were ex­ pended during the preceding year in order to create the harvest of the current year. Below is a line which separates this sum from the column of sums received by this class. On the extreme right are the sums received by the sterile class. At the center top is the sum of the income which is distributed—to the right and to the left—to the two classes which spend it. The distribution of the expenditure is indicated by dotted lines which start from the sum of the

If the proprietors were to spend more money for purchases from the productive class than for those from the sterile class in order to im­ prove the fertility of the soil and to increase their income, these increased purchases from the productive class will have to be regarded as an addition to the advances of that class. The expenditure of the income, under con­ ditions of prosperity, is supposed to be distrib­ uted equally between the productive class and the sterile class, while the productive class de-

c. The “Tableau”

Q uesnay th en m ade these generalizations:

Table 5-1. Formula of the “Tableau Economique" Total Reproduction: 5 billions Annual Advances by the Producer Class 2 billions. Sums used to pay the income and to repair and replace capital

Income for the Proprietors (Landowners, Church, and State) 2 billions m • * " r"~~’ “ ~

_

“ *7.~«»1 billion

1 billion1 billion-- " 1 billion- ' " *

Expenditure of annual advances 2 billions

Advances of the Sterile Class _ . -1 billion

• - * ' ' V^

* * * *1 billion Total 2 billions of which one half is retained by this class for advances of the following vear

Total 5 billions Oeuvres Economiques et Philosophiques de F. Quesnay, edited by Auguste Oncken, Paris, 1888, p. 316 (in French). For a more elaborated version of the Tableau, supposed to be the original one, see the facsimile in French, F. Quesnay, Tableau, etc., Macmillan & Co., London, 1894. so u rc e :

5.

CONTRIBUTIONS OF FRENCH PHYSIOCRATS

votes only one third of its expenditure to the sterile class, because expenditures of the culti­ vator are less disposable than those of the pro­ prietor; but the more agriculture languishes, the more is it necessary to devote a portion of dispensable expenditures to the re-establishment of agriculture.6 O f the 2 billion incom e, two-sevenths w ent to the State in taxes, four-sevenths to the landowners, and one-seventh to the C hurch. Q uesnay w rote: “D ifferent states of prosperity and adversity o f an agricultural nation offer a m ultitude of other cases th an the one cited, each characterized by a differ­ ent set of figures founded upon a type of calculation m ost appropriate to it.” 7 In an­ other w ork published in 1766, called Probleme Econom ique, Q uesnay furnished other sets of figures for his “T ableau E conom ique,” reflecting the condition of a nation m aking less effective use o f its resources, i.e., pro ­ viding m ore incom e to proprietors and less to the sterile class, larger investm ents by proprietors in agricultural im provem ents and, hence, a higher ratio of production to the annual advances of producers than 2% to 1 o f the case above, thus resulting in a rising national incom e; and also figures showing a reverse condition resulting in a declining national incom e.8 Q uesnay’s followers contended th a t this “T ableau” m ade it possible fo r anyone trained in its application to calculate the state of a nation from a few pertinent fig­ ures. “F o r exam ple,” w rote the C ount d’A lbon in his M oge du Quesnay in 1775, . . . given the total crops, the amount of the net product and the distribution of the people’s expenditures, it is easy to determine the number of people in each of the three classes and what is their relative well-being. If, on the other hand, one is given the size of the population, the distribution of the people’s expenditures, and the amount of the net product, one will know perfectly well what are the total crops, the amount of income received by the culti­ vators for their own consumption and capital expenses, and the distribution of the population among the various classes.9 A ccording to Lavoisier, Q uesnay in the Philosophie R urale set th e to tal annual “gross” agricultural incom e o f the country at approxim ately 2.75 billion livres, the total population at about 25 millions, and the per capita gross agricultural incom e at 110 livres. A t least this is the conclusion that Lavoisier drew from the Philosophie some thirty years later.10 It should be noted th a t if one assigns the

65

activities of the landowners, the State, and the other nonagricultural occupations the sam e characteristic of productivity that Q uesnay accorded to the cultivators, Ques­ n ay ’s figures w ould yield a considerably larger total of national production than the 2.75 billion livres.

2. Theoretical Analyses by Forbonnais, Condillac, and Turgot Tw o political economists who w ere not physiocrats, Forbonnais and Condillac, pub­ lished analyses of the operations o f the national econom y th a t similarly stressed the flow of incom e and expenditures, the flow of production and consum ption, and the reproduction of w ealth. T urgot, a third econom ist, w ho occupied a position midway betw een the physiocrats and nonphysiocrats, published a unique an d pioneering work, also stressing this approach. However, none of these three im portant w riters furnished any figures on national income. Frangois Forbonnais (1 7 2 2 -1 8 0 0 ) was the son of a w ealthy m erchant. H e began his studies of political econom y at an early age, and being a m an of independent means, was able to pursue them w ithout hindrance. H e becam e well acquainted w ith the E ng­ lish econom ists— M un, Petty, D avenant, and others— and was deeply influenced by them . E ventually he entered governm ent service, becom ing Inspector G eneral of the M int and instituting im portant reform s in that office. In 1754 he published the E lem ents du Com m erce, in 1758 the R echerches et Considerations sur les Finances, and in 1767 the Principes Econom iques, all of w hich w ere original w orks o f considerable m erit. In the Principes E conom iques he took sharp issue w ith Q uesnay’s concept of p ro ­ duction and national income. Forbonnais considered national incom e to consist of all currently produced useful things, as re­ flected in the n et incomes o f all occupied persons. This n et incom e is the residue of gross incom e left to each individual after paym ent fo r the aid secured from other individuals; an d national incom e is, there­ fore, th e sum of these net incomes. It m ust be divided into portions supplying mere sub­ sistence and those supplying surpluses. H e viewed incom e from agriculture as prim ary o r natural incom e because it provides neces­ sities; incom e from industry as secondary or artificial incom e because it supplies super­

66

THE IN CO M E OF NATIONS

fluities and other additional utilities; and felt th a t the two together m ade up th e na­ tional income. F orbonnais thought th a t it was wrong to consider the cultivators alone as producers, as the physiocrats did. Sub­ sistence, he felt, has the first claim on na­ tional income, even before the claims o f the State. Taxes should be levied only o n the surplus portions of incom es, and persons having only subsistence incom es should not be taxed at all. Forbonnais’ concepts of national incom e were derived partly from English thought and partly from F rench ideas. T hey had their roots in the theories of Petty, King, and Boisguillebert, and represented a view­ point that in the long ru n proved m ore lasting than that of the physiocrats. Forbonnais believed in rule by the p ro p ­ ertied class, but he also stressed the respon­ sibilities that the propertied class m ust bear fo r the welfare of the general com m unity. H e believed in proportional taxation, and when elevated to the nobility, refused to take advantage of the tax exem ption privi­ lege. By a public declaration in 1764 he subm itted all his property to the imposition of the Taille. H e believed in protectionism , ascribing the grow th of English foreign com m erce to her adherence to that policy. H is analysis of national incom e was purely theoretical, and he m ade no attem pt to esti­ m ate the size of the actual national incom e, its com position, and its distribution.11 Etienne B onnet de Condillac (1715—1780) was prim arily a philosopher. H e was well read in the w orks of the English and other E uropean philosophers (H obbes, Locke, Descartes, Spinoza, Leibnitz, etc.) and was the author of several books th a t w ere influ­ ential in the developm ent o f W estern E u ro ­ pean thought. In the field of political econ­ om y he w rote a com prehensive and original treatise, L e Com m erce et L e G ouvernem ent (1 7 7 6 ), which, by curious coincidence, con­ tained some of the same ideas A dam Smith developed in his W ealth o f N ations pub­ lished in England during the same year. Condillac held that all occupations, in­ cluding services, w ere equally productive, differing in this respect from both the phys­ iocrats and A dam Smith. H e insisted that all w ealth has its roots in labor; th a t the wealthiest nation is the one that has the widest variety of occupations; that it is just as senseless fo r one nation to prefer agri­ culture to m anufactures as it is for another to prefer m anufactures to agriculture— that

in th e m atter of econom ic occupations “a nation should prefer none, b u t should oc­ cupy itself w ith everything.” 12 H e traced th e circular flows o f production and con­ sum ption and o f incom e and expenditure, but, like Forbonnais, did n o t attem pt to esti­ m ate the real national income. A . R . J. Turgot (1727—1781) published a m om entous w ork in 1769-70 on the fo r­ m ation and distribution o f national w ealth.13 H e was sym pathetic to the basic ideas of the physiocrats, but did n o t subscribe to them in full. Schum peter recently described his book as one whose “theoretical skeleton is, even irrespective of its priority, distinctly superior to the theoretical skeleton of the W ealth o f N ations” and whose “capital theory . . . anticipates m ost of the nine­ teenth-century w ork.” 14 T urgot was well acquainted w ith the English economists and translated th e w orks of M un and T ucker into F rench. A lthough he follow ed Ques­ n ay ’s doctrine o f the “revenu n et” as a p roduct of agriculture, he attributed m uch greater im portance to industry th an did Q uesnay o r any other physiocrat. In 1774 he was appointed M inister of F inance and, during his tw enty m onths of tenure in th a t office, effected a series of extraordinarily im portant financial reform s. H e was driven out o f th a t office n ot so m uch by dissatisfaction on the p art of the C ourt as by “p opular resistance of the rural proletariat and o f craft guilds” to his de­ crees of internal free trad e in grains and his m easure abolishing the monopolistic guilds.15

3. Estimates Published During the Immediate Pre-Revolutionary Era: Du Pont, Tolosan, Young, and Others In the 1770’s and 1780’s, the earlier the­ oretical analyses of national incom e were follow ed by an equally rich ou tp u t of sta­ tistical estim ates of the size and distribu­ tion of national incom e. M ost of these esti­ m ates w ere prepared by political econo­ mists in conjunction w ith specific proposals fo r am eliorating the increasingly desperate financial condition of France, b ut the au ­ thors w ere often unclear as to w hat they m eant by the term “national incom e” or how they reached their p articu lar statistical results.

5.

CONTRIBUTIO NS OF FRENCH PHYSIOCRATS

Voltaire (1 6 9 4 -1 7 7 8 ), in his w itty pam ­ phlet, L ’H o m m e aux Quarante E cus (T he M an of F orty C row ns), published in 1768, quoted a national incom e figure of 2.4 bil­ lion livres for F rance. T his figure was sup­ posed to have been com puted on the basis of the average value of the agricultural produce per cultivated acre of land. A ppar­ ently, V oltaire took this figure from Phi­ losophie Rurale. D ividing this total by an assum ed population of tw enty millions, he got a per capita incom e for the country of 120 livres a year, or 40 crow ns (the “crow n” o r “ecu” equaling 3 liv res).1® T he physiocrat, L e Trosne (1 7 2 8 -1 7 8 0 ), in his D e I’lnteret Social (1 7 7 6 ), offered an estim ate of 3.1 to 3.4 billion livres for the incom e of F rance fo r th a t tim e, but according to D ’lvernois (see Section 7 be­ low ) this figure did n o t include the incom e of the clergy from the tithe, the salaries of civil and m ilitary personnel of governm ent, o r the interest on the public debt.17 Pierre Sam uel D u P ont de N em ours (1 7 3 9 -1 8 1 7 ), one of the m ost brilliant and versatile followers of Quesnay, in a m em o­ randum prepared in 1785 fo r the C om ­ m ittee on A griculture (w hich is generally considered to have been the precursor of the D epartm ent of A griculture established during the R evolution), estim ated the total gross agricultural production of F rance at 2.5 billion livres and the net rental o r tax­ able incom e at 1.5 billion livres. A few years later, in 1789, in a speech delivered before the N ational Assembly on the state of the finances of the N ation, he estim ated the gross national incom e of the country at 4 billion livres, but its net agricultural in­ come, again, at 1.5 billion. A t th a t tim e the Assembly was considering a proposal to impose a 25 per cent direct tax on n et rent from land. In arguing against this rate of levy, D u Pont quoted the following figures: The annual value of the total product of the Kingdom embracing the annual reproductions of grain, wines, oils, fruits, vegetables, flax, hemp, hay, animals, wood, fishing, and prod­ ucts of mines and quarries, has been calculated many times in different ways, all of them con­ firming each other. The lowest estimates place this value at 3,200 million livres. The highest and probably the more correct ones bring it up to four billions, and this is little when com­ pared with the value which such a wonderful land could have produced if well cultivated; it is also little when compared with the product of one of the neighboring nations which has not as yet attained a perfect government. Of

67

this produce of 4 billions, 2.5 billions are ab­ sorbed by the expenses of cultivation and ex­ ploitation, including the subsistence of the cul­ tivators and of other agents of productive labor as well as of salaried persons, workers, mer­ chants and artisans who live at the expense of the producers—cultivators, fishermen, and miners. This part of the product of the land and of the rivers is not susceptible of taxation. It is the indispensable compensation of produc­ tive labor: any levy upon it would inevitably alter the source of the wealth of the society and cause a greater loss of revenue to the state than gain. . . . The expenses being at least 2.5 billion in a total product of 4 billions, the net or taxable income cannot exceed 1.5 billion. This 1.5 billion is divided between the public treasury, the ecclesiastical tithe, and the land­ owners and proprietors of other productive works, both maritime and subterranean. The division is made in such a way . . . that the state receives approximately half of the net income. . . . Therefore, there remains between 700 and 800 millions net income in the hands of the proprietors . . . whose number is im­ mense and includes many who are very poor and are unable to bear any additional imposi­ tions such as the proposed levy of one fourth of their income. I believe that there are not even 300 million of income belonging to propri­ etors who are sufficiently wealthy to be able to pay a tax of one fourth. The most that could be raised would be 75 million and even then only probably over a two-year period, as dem­ onstrated by the Minister of Finance. In other words, the levy could not produce more than 40 millions in 1790.18 In 1789, Tolosan, an Inspector of Com ­ m erce, anonym ously published a book on the F ren ch econom y in w hich he included estimates o f the incom e from agriculture; incom e from industry, including labor in­ come, operating expenses, and profits; and the gains from colonial trade, including the slave traffic. H e estim ated the total national product at 3 billion livres.19 Such com petent nineteenth-century statisticians as M oreau de Jonnes considered T olosan’s statistical survey and estimates to be m ore authorita­ tive th an any others o f the pre-Revolutionary and Revolutionary period.20 D ’lvernois in his book of 1799 quoted M inister E tienne Claviere as having simi­ larly estim ated the national incom e at 3 billion livres in his book, Public Faith (1 7 8 8 ).21 A rth u r Y oung (1 7 4 1 -1 8 2 0 ), in a book about his travels in F ran ce,22 furnished an estim ate o f both the gross and the net agri­ cultural incom e of F rance in 1789. H e m ade no estim ate of the country’s total n a­

68

THE INCO M E OF NATIONS

tional income, although he did give scat­ tered figures fo r m any categories o f wealth, property, and industrial activity. Y oung m ade two calculations of the gross income from agriculture. In the first, he divided F rance into seven districts crossing provincial boundaries, each with a dom i­ n ant soil type— consisting of rich loam , heath, m ountain, chalk, gravel, stone, and sand. In order to ascertain the proportional area of each of the several soil divisions into w hich he had divided the kingdom , he proceeded as follows. “I procured a copy of [a] m ap to be m ade on a sheet o f paper of equal and sim ilar thickness . . . and then cut out, w ith a fine pair of scissars [sic], the several divisions, w hich w ere first weighted separately, and afterw ards th e w hole to ­ gether. A ll F rance [131.7 m illion acres] weighted 413 weights, equal to one fourth of a grain.” 23 T he acreage o f each category of soil was then calculated on its m apw eight basis. On the basis of private discussions with F rench agriculturalists and his personal ob­ servations, Y oung determ ined an average productivity per acre fo r each soil category. H is first estim ate o f incom e from agricul­ ture was as follows: 24

Acres (millions) Rich Loam 28.4 25.5 Heath 28.7 Mountain 16.6 Chalk 3.8 Gravel 20.4 Stone 8.3 Sand 131.7

Average Product per Acre £ s. d. 213 9 y 1 8 9% 1 99 1 80 1 0 6J£ 1 15 0 1 11 4K

Total Product, Value (million £) 76.3 36.8 42.7 23.2 3.9 35.7 13.0

1 15 l J i

231.6

2

Y oung prepared his second calculation of gross agricultural incom e by redividing the total area into six categories of dom inant agricultural product, deducing the total gross income from each as follows: 25

Acres Arable land (grain production) Vines Woods Meadows and rich pasturage Lucerne, etc. Pastures and wastes

70.0 5.0 19.9 4.0 5.0 27.2 131.0

Y oung declared th at it was easier to cal­ culate the gross th an the n et product ( re n t) .26 However, on the basis of personal observations ( “m y notes on th a t subject” ), he did estim ate an average re n t fo r each category. H is estim ate of ren t o r n et product w as: 27 Acres (millions) Arable and Lucerne Woods Vines Meadows Waste

Rent per Acre £ s. d.

Total (million £)

75.0 19.9 5.0 4.0 27.2

0 0 3 2 0

15 12 16 3 1

7 0 6 9 9

57.4 11.9 19.1 8.8 2.4

131.0

0

15

10

99.9

A ll estim ates of the period (1 7 6 0 -9 0 ) w ere devoid o f detailed breakdow ns and gave little inform ation regarding the con­ cept o f national incom e em ployed in their preparation. M ost of them , however, were influenced by the physiocrats, and lim ited national incom e to agricultural income.

4. Lavoisier’s Comprehensive and Systematic Estimates L avoisier’s estim ate of national income, m ade fo r th e N ational Assembly, was alto­ gether different from either V auban’s in­ com plete estim ate o r the crude an d u n ­ docum ented estim ates o f th e period 1 7 6 0 90. T o this task, begun in 1784 and com ­ pleted in 1791, Lavoisier tried to bring the same m eticulous scholarship an d m ethodical procedure th a t characterized his great w ork in chem istry, but, being less fam iliar w ith econom ic problem s, he inadvertently m ade some very grave errors. Lavoisier’s estim ate was based on a m ore solid factual foundation th an th a t o f any previous estim ator. H e assem bled m ost of his d ata through correspondence w ith the Average rrouucL (livres) 40 175 16 100 100 10 40

Total Agricultural Product (million livres)

(million £)

2,800.0 875.0 317.6 400.0 500.0 271.5

122.9 38.2 13.9 17.5 21.9 11.9

5,164.1

226.2

5 . CONTRIBUTIONS OF FRENCH PHYSIOCRATS

farm er-generals in the various provinces, w ith m any of w hom he had becom e p er­ sonally acquainted w hen he also was a farm er-general. H e m ight have been fam iliar w ith th e estimates prepared in E ngland by Petty, D avenant, G regory King, and A rth u r Young, although there is no direct evidence to th a t effect. Y oung had visited Lavoisier in Paris in 1787 and reported on this visit in his E uropean travel book. T he project would have interested Y oung, w ho p ub­ lished his own estim ates of the English and, in p art, F rench national incom es, and Lavoisier might have discussed it w ith him , but this is pure conjecture. In preparing his estim ate, Lavoisier cross­ checked his figures w ith the aid of different data. H e recognized the danger o f double counting, and contributed significantly to the identification of this problem . H e ad­ vanced the idea th a t double counting could best be avoided by approaching the evalua­ tion of production from the point o f view o f consum ption. In effect, he h ad touched on the essence of the im portant distinction between final goods and interm ediate o r p roducer goods, w hich is critical in m odern economics and national incom e account­ ing. Lavoisier’s estim ate was prophetic in another respect as well, in th a t it forecast the eventual preparation of national in­ com e estimates by governm ent itself. A t the same tim e th a t it extended the frontiers o f econom ic statistical analysis, Lavoisier’s estim ate suffered from certain defects. It was built on the physiocrat’s re­ stricted concept o f national incom e as the gross value o f agricultural production and of the n et or taxable incom e as the ren t of land. H is total, therefore, was a great u nder­ estimate. B ut although Lavoisier did not include nonagricultural item s in the national income, he did not fail to estim ate them , as separate items, just the same. H e was m uch m ore conscious o f the im portance o f in­ dustrial production than w ere m ost other physiocrats. H is estim ate was low because he fu rth er restricted agricultural produce to grains and m eat. H e failed to include the value of dairy products, vegetables, fish, and fowl. H is estim ate o f the “produit n et” o r taxable income was apparently also low. M oreover, through no fault of his own, but because of the deficiencies of the available statistics, Lavoisier’s estim ate involved m any debatable assumptions and speculations, the weakness of w hich he him self readily ad­ m itted.

69

a. Lavoisier’s L ife— A Great Chemist Becomes a Practical Social Scientist A n to in e L aurent Lavoisier (1 743-1794) was the son of a prosperous Parisian m er­ chant.28 H e developed an early interest in science, an d was perm itted by his fam ily and econom ic circum stances to devote himself entirely to his studies. H is notable discoveries in physical sciences, and especially in chem ­ istry, earned him election to m em bership in the F rench A cadem y of Sciences at the age of twenty-five, and, later, recognition every­ w here as a fo u n d er of m odern chemistry. In ord er to have an assured income with which to support his extensive an d expen­ sive research laboratory, Lavoisier, in 1768, bought the office of a provincial “farm ergeneral” or tax collector. T he “Farm -G eneral” operated very m uch like a stock com ­ pany. E ach farm er-general was a p a rt ow ner of the enterprise in addition to managing his own branch. Lavoisier’s entrance into this business was th e m ost fateful step in his life. T he office of farm er-general— there were sixty of them — was thoroughly hated by the com m on people, and twenty-five years later, during the Revolution, he had to pay fo r this m istaken investm ent w ith his life. Lavoisier was said to have adm inistered his tax collection office w ith singular fair­ ness, probity, and generosity, b ut this did not outw eigh the strong popular prejudice against his having held it at all. W hile a farm er-general, Lavoisier learned m uch about the inequalities of the existing tax system and becam e an ardent advocate of tax reform . H e also developed an interest in th e n atio n ’s econom ic problem s and learned to appreciate the im portance of de­ veloping better econom ic and fiscal statistics in the country fo r form ulating m ore en ­ lightened public policies. In 1777, Lavoisier bought a farm and began to experim ent w ith new m ethods of agriculture. This personal experience ena­ bled him to understand agricultural prob­ lems better. H e becam e one of the leaders of the newly founded Royal Society of A gri­ culture (originally established in 1761 as the Society o f A griculture of the “gener­ ality o f P aris” ), and w rote several notable papers fo r it on ways to correct the back­ wardness of F ren ch agriculture. In 1785 he was appointed Secretary of th e im portant G overnm ent C om m ittee on Agriculture, presided over by V ergennes and Calonne.

70

THE IN CO M E O F NATIONS

T he Com m ittee prepared m any plans, but none o f these was adopted even in the face of the growing econom ic disorganization and approaching political catastrophe. T he governm ent was incapable of any construc­ tive action. In 1784, Lavoisier began to assemble data on population, national in ­ com e, production, and consum ption to be used in form ulating plans fo r im proving the existing fiscal system and revising the governm ent’s econom ic policies. A fter the outbreak o f the Revolution, th e N ational Assembly to w hich he had become an alternate delegate, appointed Lavoisier to an official Commission charged w ith preparing reorganization plans fo r the Treasury. A t the same tim e th at he pre­ pared several reports fo r the Commission, he was prevailed upon by his associates in the A cadem y to com plete his calculations o f the national incom e and to subm it them to the N ational Assembly fo r the use o f its T axation Com m ittee. H is work, subm itted in 1791, was praised by the Com m ittee, and the N ational A ssem bly ordered it to be printed, but none of its conclusions was adopted. As soon as the Jacobins gained control over the N ational Assembly, Lavoisier’s fate was sealed. H e was arrested in 1793 and, together w ith twenty-seven other farm ersgeneral, was charged w ith crimes against the people. H is despairing friends vainly pointed out th at he had adm inistered the duties of that office w ith exem plary fair­ ness and th a t he had resigned from it years before. Similarly fruitless were the interven­ tions on his behalf by the A cadem y o f Sci­ ences, w hich cited his great scientific con­ tributions to hum anity. T he fact th a t he had occupied the odious office of farm er-general was sufficient to condem n him to death together w ith the rest of the farm ers-general. References to his scientific contribution were said to have been dismissed curtly by the chairm an of the tribunal with a statem ent that “the Republic does not need scientists and chemists; the course o f justice cannot be interrupted.” Lavoisier, who was at w ork in prison collecting his m ain w orks fo r fu ­ ture publication, asked for a few days’ de­ ferm ent o f his execution to perm it him to com plete this task, but even this request was denied. On M ay 8, 1794, Lavoisier ascended the guillotine. O n learning of his execution, his friend, the w orld-fam ous m athem atician L a­ grange, was said to have stated to the illus­

trious astronom er D elam bre: “It took them b u t a m om ent to cut off this head, b u t it will take m ore th an a century fo r the world to produce another one like it.” 29

b. Lavoisier’s Perception of the Impor­ tance of Good National Income Estimates L avoisier called his estim ate L a Richesse Territoriale de France (T h e T erritorial W ealth of F ra n ce ), b ut to him “Richesse” m eant the flow of w ealth, i.e., th e national incom e o f the country, n o t the stock of w ealth. T his concept of w ealth as a flow cam e o ut of the m ainstream o f F rench rath e r th an English political economy, w hich p referred to trea t w ealth as a stock. In the Preface to his estim ate, Lavoisier explained how he started w ork on it in 1784, endeavoring to go beyond the estimates p re­ p ared by D u P o n t fo r the Vergennes-Calonne C om m ittee, an d to provide a m ore solid factual basis fo r calculating the m onetary value of the n atio n ’s production and con­ sum ption.30 H e w rote th a t he h ad “resum ed and interrupted this w ork at least tw enty tim es during this period,” being diverted from it by m any other tasks, and th a t it was due only to the urgent request of the N a­ tional A ssem bly th at he was m aking his w ork available fo r publication in its existing incom plete an d im perfect state. Lavoisier observed fu rth e r th a t “the kind of com bination and calculation th at I have sought to give examples o f here, is the basis of all political economy. This science like m any others began w ith m etaphysical dis­ cussions: its theory has advanced; b u t its practice is still in its infancy, and the states­ m an lacks at all times the facts on w hich to base his speculations.” 31 H e addressed the following eloquent plea to th e N ational A s­ sem bly to establish an agency w ithin the governm ent th a t w ould regularly collect the d ata necessary fo r the preparation an d pub­ lication of reliable annual estimates of n a­ tional incom e: Let the representatives of the French nation, these men of genius whose works will merit the admiration of the nations . . . realize how much more sure could have been their moves if the philosophers who have preceded them had prepared in advance the materials neces­ sary for the erection of the structures proposed by them and had founded their works on facts instead of on mere reasonings! It now is left to these representatives of France to establish an agency in the government which would un­

5.

CONTRIBUTIO NS O F FRENCH PHYSIOCRATS

dertake to prepare a balance sheet of the results of agricultural production, of commerce, and of growth of population; and would depict in a single summary table the current situation of the country, its wealth in people, in production, in industry, and in the accumulation of capital. Such a grand agency which does not now exist in any country can be established only in France. The National Assembly has but to wish it, and it will come into being. The pres­ ent organization of the French state is pecul­ iarly well suited to the organization of such researches. The General Administration can, through the medium of the directories of the departments and districts, easily reach to the farthest branches of the political tree, namely the municipalities: using the patriotic commu­ nication of this kind, it can obtain any neces­ sary information and undertake any type of useful survey.32 H e accom panied his plea w ith an outline of the type of inform ation such an annual national account should contain and a dem ­ onstration o f th e uses to w hich it could be put, saying: There is a way of giving a greater degree of clarity to this kind of work: it consists in the preparation for a given base year of a general balance sheet of the total production of the country. Each type of product will have in it a separate section. The agriculture of the coun­ try will be treated just as if it were a domain belonging to a single individual having charge of all the receipts from production and de­ ciding how these receipts should be employed. Thus, taking for example the section on wheat, he would be in charge of the receipts of the entire crops (estimated at 14 billion pounds) and will next allocate their disposition under various items of expense such as: Delivered to the cultivators of the country for use as seeds Delivered to the cultivators for their annual subsistence Delivered to the harvesters for defrayment of the expenses of the harvest Delivered to the collectors of the tax Delivered to the landowners as rent for the right of farming Similar sections would be provided for all other types of production. This general ac­ count in kind will be supplemented by an ac­ count in money in which all the others will be joined. The account of woolens, hemp, flax, and all primary materials of industry would be par­ ticularly interesting as it would reflect the point of contact between agriculture and com­ merce. It would show that the values of the products of commerce and of industry are equal to the values of their consumption; so that, for example, to sell cloth abroad is to

71

sell wool and wheat, with one difference only, namely, that the nation fabricating the cloth, gains in population as it has more individuals who produce cloth and consume wheat. A work of this nature will contain in a few pages the whole science of political economy; or, rather, it would do away with the further need for this science; because the results will become so clear, so palpable, the different ques­ tions that could be raised would be so easily solved, that there would no longer be any differences of opinion. This account will not be brought to a state of perfection all at once: it would contain errors at first, but time will furnish the means of correcting them. Once such an account for agriculture has been established for a base year, it would be a simple matter to construct a similar one for each succeeding year. Then one would be able to see the influences of the abundance of crops on the national wealth, how much taxes can the country support in a good year and what tax reliefs should be provided in a bad one; how much could a country export in a given year without risk, etc. These general accounts which would be ex­ tended to include the whole population and the balance of all commerce would furnish a veri­ table thermometer of the national prosperity; and every legislator would be able at a glance to see in these summary tables the good as well as the bad that has resulted from the actions of the previous legislatures.33

c. Lavoisier’s Concept of National Income H aving indicated the difficulties surround­ ing the preparation of his work, Lavoisier explained his concept o f national income. Agreeing in the m ain w ith the physiocrats, he distinguished between three types of n a­ tional incom e aggregates: a. an aggregate of all products in kind, both primary and final, both of agriculture (includ­ ing the grass growing in the pastures and the increases in livestock) and of all other occupa­ tions supported by agriculture. This aggregate would contain much double counting and would not lend itself to summation in a single con­ sistent total; b. an aggregate representing “the real na­ tional income” (“revenu reel du Royaume”) expressed in money v alu e34 and free of all duplications (Lavoisier called it “gross national income”); and c. the “net national product” or “net na­ tional income,” which is also the “taxable na­ tional income” and which consists of the agri­ cultural rent or surplus payable to the land­ owner, the church, and the public authority. (Inasmuch as Lavoisier’s report was intended to help determine the amount of revenue from

72

THE INCO M E O F NATIO NS

the direct tax which the National Assembly could raise from the landowners and the church, the calculation of this “net” or “taxable” in­ come was one of its main objectives.) W ith the aid of a few examples, Lavoisier sought to explain the type of double count­ ing he was trying to eliminate. “T he value of the hay and of the oats consum ed by animals em ployed in cultivation is contained in the value of the crops raised w ith their labor. It would, therefore, be a m istake to include the value of the hay and oats as

( b ) his figures show ed a slight num erical preponderance o f m ales over fem ales; (c) people over 60 appeared to constitute but 7 p er cent o f th e total population; and (d ) th e noblem en num bered only 83,000 o r only 3 p e r cent o f the population, and of them , only 18,323 w ere identified as being capable of bearing arm s. H is table showing the distribution o f the population by occu­ pation an d social status deserves reproduc­ tion, as it sheds light on the econom ic struc­ tu re of F ren ch society of the tim e:

T ab le 5-2. T ab le of th e In h a b ita n ts of F rance, w ith identification of th e ir social s ta tu s a n d o ccu p atio n , including in each categ o ry m en, w om en, a n d children 36 P opulation of th e cities a n d boroughs L aborers, farm ers, v alets, o th e r dom estics, shepherds (m en, w om en a n d children included) D aily laborers occupied in b a rn s d uring th e w in ter a n d in th e fields du rin g th e sum m er, earthw orkers, m asons a n d all o th ers living a t th e expense of agricu ltu re, including th e ir fam ilies W ine-grow ers a n d th eir fam ilies Salaried w orkers em ployed b y w ine-grow ers a n d ow ners of v in ey a rd s M erch an ts, tavern-ow ners, surveyors of cities a n d boroughs, m archals, C artw rights, harness a n d saddle m akers, living a t th e expense of agricu ltu re, including m en, w om en a n d children Sm all p ro p rieto rs living m o stly off p ro p e rty incom e Sailors, d aily laborers of all kinds em ployed in m an u fa ctu re s located outside th e cities, quarryw orkers, m iners, carriage a n d w agon m akers, noblem en, ecclesiastics, a n d th e ir dom estics, living o u t­ side th e cities F re n c h A rm y T o ta l

such in the national incom e. These items are not real income. T hey are only annually reproducible item s o f expense.” 35

d. Three Major Types of Data Used by Lavoisier for His Estimates A s a first step in his estimates, Lavoisier prepared three basic sets o f data: one on population, one on the num ber o f useful animals, and one on land usage. H e referred to the population estimates m ade by M oheau and M ichandiere that gave the F rench population by cantons, but indicated that they were generally inade­ quate. H e sought to achieve a m ore elabo­ rate estimate showing the population not only by provinces and departm ents, but also by age groups, sex, occupation, m arital status, and urban and ru ral residence. W ithout reproducing all his population tabulations, it can be noted th a t (a ) he w ound up with a population of 25,000,000, w hich was m uch higher than any previous estimate, and yet was probably too low;

8 ,0 0 0 ,0 0 0 6 ,0 0 0 ,0 0 0

4 ,0 0 0 ,0 0 0 1 ,7 5 0 ,0 0 0 800,000

1 ,8 0 0 ,0 0 0 450,000

1 ,9 5 0 ,0 0 0 250,000 2 5 ,0 0 0 ,0 0 0

Lavoisier estim ated the total n um ber of useful anim als, classifying them as: horses used in plow ing; those used in o th er agricul­ tu ral labor w here plow ing was done by oxen; those used in Paris and o th er cities fo r transportation; arm y horses; oxen; cattle used to fatten into beef; m ilking cows; sheep; pigs; etc. H is th ird m ajor calculation dealt w ith the total cultivated area o f the country; the n um ber of plow s in use (classified according to the type of anim al pulling the p lo w ); the average n u m b er o f arpents plow ed with each type of plow; and the average yield of each type of cultivated arpent. H e esti­ m ated th e total area of F rance at 105 m il­ lion arpents, w ith 64.8 million, o r tw o thirds being cultivable and only 28.2 m illion ac­ tually u n d er cultivation. H e fu rth e r subdi­ vided the cultivated area on the basis o f his d ata on the num ber an d kinds of plows into th a t w orked by oxen, th a t w orked by horses, th a t w orked during one p art o f th e season by oxen and during th e oth er by horses, the

5.

CONTRIBUTIONS OF FRENCH PHYSIOCRATS

portion of the land left fallow (18.6 m illion arpents) and the portion left in pasture (18 million arp e n ts).37

e. Lavoisier’s Estimate of the National Income and of Its Distribution W ith these basic data in hand, Lavoisier used three m ethods of estim ating th e n a­ tional incom e (or, m ore correctly, the “gross agricultural incom e” ) : (a ) the total pro­ duction of grain, m eat, and w ine; ( b ) the total per capita consum ption and expendi­ tu re o f the whole population and of its dif­ ferent classes, and o f specific kinds o f con­ sum ption and expenditure taken separately; and (c) total daily earnings and expendi­ tures of certain types of gainfully em ployed persons. As far as possible, he used the re ­ sults obtained by one m ethod as a check against those obtained by the others. H e estim ated the total annual production of grain at 14 billion pounds (7 raised by the use of plows pulled by horses and 7 by those pulled by o x en ). H e broke this p ro ­ duction down according to its uses as fol­ lows: Billion Pounds Seeds C u ltiv a to rs’ own consum ption T em p o rary help a n d o th er o p e ratin g expenses T ith e a n d o th e r taxes R ent T o ta l

2 .3 0 .9 3 .5 2 .9 4 .4 1 4 .0

Thus, the cultivator’s ow n consum ption am ounted to less than 7 p er cent of the total crop (92.6 million livres). Lavoisier eval­ uated the portion available fo r hum an con­ sum ption (11.7 billion pounds of w heat, rye, and barley) at 2 sous (one tenth of a livre) p er pound, o r at a total of 1,167 million livres.38 F o r the consum ption of m eat in cities, Lavoisier used slaughterhouse d ata (arriving at a total m eat consum ption in cities of 6.5 millions of carcasses of different kinds, aggregating 689 m illion pounds, not count­ ing fo w l), while for villages he used infor­ m ation on the total livestock of various kinds and on the estim ated num ber o f each being slaughtered each year. In this w ay he ar­ rived at a total m eat consum ption fo r F rance o f 1.2 billion pounds, o r one tenth of the w eight o f the consum ed bread.39 Lavoisier had a m ore difficult tim e in es­ tim ating the consum ption of wine, and ad­

73

m itted th at his figures m ight be off, one way o r the other, by one quarter, one third, or even one half. T he best estimate, he said, was 4.5 m illion pints of w ine p er day, not counting cider, o r a total of 1.6 billion pints a year. Starting w ith these figures, w hich covered the bulk of the people’s consum ption (the rest being fo r clothing and housing— com ­ paratively small items in this type of agricul­ tural econom y), he proceeded to calculate the per capita average consum ption expendi­ ture by different classes of th e population, and fo r all classes taken together. H e started w ith the general proposition th at m en con­ sum e m ore th an w om en and th a t in a fam ­ ily com posed of husband, wife, and three children below ten years of age, the fath er consum es almost as m uch as the rest of the fam ily; and th at fam ily consum ption varies w ith its circum stances and econom ic status. A portion of the individuals in the country districts do n o t eat any m eat; inhabitants of Paris and the larger cities eat from 6 to 7 ounces of m eat a day; those living in small cities eat 4 to 5 ounces; and those in country districts consum e on the outside 2 ounces— m ost of their nourishm ent consisting of bread, vegetables, fruits, butter, cheese, and milk. M oreover, the consum ption of bread varies from year to year, as the crops are abundant o r poor, and as the price o f bread or flour rises o r falls, the poorer people eat­ ing m ore bread w hen it is cheap and less w hen its price rises. A ll these facts, he said, m ake any precise com putation o f the aver­ age consum ption o f the people difficult. Lavoisier concluded th at am ong the rural fam ilies at the low est subsistence level ( “les plus m iserables” ), the per capita consum p­ tion (m en, w om en and children included) am ounted to 60 livres a year, providing one pound of bread per person a day, with bread consum ption representing approxim ately 70 p er cent o f the total m oney value of the fam ilies’ consum ption.40 F o r th e average F ren ch ru ral fam ily he arrived at a per capita consum ption of 117 to 120 livres annually. F irst, he assumed th at th e country­ w ide average consum ption of husbands just about equaled a soldier’s pay and m ainte­ nance, w hich then am ounted to 251 livres a year. H e anticipated a possible objection to the effect th at the average should be fixed m idw ay between the pay of a soldier and that of an officer, b ut dismissed it on the ground th a t an officer’s pay has to provide not only fo r his own m aintenance but also fo r th at of his m anservant and his horse.

74

THE INCOM E O F NATIONS

Actually, he said, the officer’s own consum p­ tion, except fo r clothing, is n o t m uch higher than that of the soldier. N ext, he assumed th a t the w ife’s consum ption in the average fam ily am ounted to two thirds of her hus­ ban d ’s, or to 169 livres and 6.8 sous a year (1 livre having 20 sous); and that the con­ sum ption of three children below ten years of age in such a fam ily also totaled two thirds of the father’s, or sim ilarly 169 livres 6.8 sous. Thus, Lavoisier arrived at a total annual consum ption fo r the average fam ily of five m em bers of 587 livres and 13.4 sous, or at the per capita annual consum ption of 117 livres 2.8 sous. “T o support such a con­ sum ption,” w rote Lavoisier, “it is necessary for the head of the fam ily and his wife to earn together 1 livre 18.3 sous (o r 38.3 sous) per day, excluding Sundays and holi­ days, which is certainly the m axim um that a m anual labourer (and his w ife) could ex­ pect to earn.” 41 H e conceded th a t this aver­ age per capita consum ption fo r the country m ight be as m uch as 120 livres a year; but he did not believe th a t it could be higher. This corresponded to V oltaire’s earlier cal­ culation of forty crowns (120 livres) a year per individual. Lavoisier believed th a t Vol-

products o f the “sterile class.” H e ended w ith a figure o f 1.2 billion fo r this “n et” incom e.43 A dm itting th a t he was unable to estim ate the value of certain products sold in cities by the agents of landow ners (eggs, butter, cheese, fruits, vegetables, and o ils), he conceded th a t the “n et” or “taxable” in ­ com e could be a little above 1.2 billion livres, b ut he did n o t think th a t it could be m uch above this sum. H e concluded, therefore, th a t the N ational Assembly’s p ro ­ posed direct tax o f one sixth of the n et in­ com e from land could n o t produce m ore th an 180 m illion livres or, at the utm ost, 200 m illion; th a t the proposed tax on personal property could n o t produce m ore than 40 m illion livres; and th a t the two measures taken together could n ot yield m ore than 230 billion in 1791.44 These tw o taxes were to replace th e h ated and now abolished “taille,” poll tax, an d salt tax. Lavoisier’s conclusion as to the possible yields of the two direct taxes was disappointing to some m em bers of the Assembly who h ad counted on obtaining a m uch larger revenue from these sources. Lavoisier sum m arized the results o f his calculations in th e following table: 45

Table 5-3. Final Results Evaluated in Money General product of the territory This product cannot be expressed in terms of money, a t least in its totality and is, therefore, entered here merely as a Memorandum Portion of the national product (“produit territorial”), con­ vertible into money and free of all double counting: it is the Livres total product consumed by the inhabitants 2 ,7 5 0 ,0 0 0 ,0 0 0 Net or taxable product, at a price of wheat a t 2 sous a pound or as 24 livres a setier 1 ,2 0 0 ,0 0 0 ,0 0 0 Portion of the net product absorbed by direct and indirect taxes 6 0 0 ,0 0 0 ,0 0 0 Portion of the net product remaining to the landowners 6 0 0 ,0 0 0 ,0 0 0 taire’s figure (supposedly taken from Ques­ nay) was too high fo r the period to w hich it was supposed to apply, but th a t his own figure was just about right for the year 1789. A t these per capita figures, the total con­ sumption fo r the entire population of 25 million could only be between 2.5 and 3 bil­ lion livres. H e was satisfied th a t this was a correct estimate. This, he said, “is th e real income of the nation free o f all duplica­ tions.” 42

f. Lavoisier’s Estimate of the “Net” or “Taxable” Income In order to obtain the “n et” o r “taxable” income o f the country, Lavoisier h ad to de­ duct the consum ption and expenses of the cultivators, including their expenses fo r the

A ccording to L avoisier’s estim ate, m ore th an h alf of the national product of 2.75 billion livres was absorbed by the consum p­ tion of the people and the expenses of cul­ tivation, and the surplus of 1.2 billion was split equally between th e public treasury and the proprietors (ch u rch and land­ ow ners). Lavoisier follow ed this general sum m ary w ith a series of supporting tables showing, in considerable detail, th e consum ption ex­ penditures of th e inhabitants of Paris by type of goods (on the basis of the records of the “octroi,” or custom s at the gates of the city ); total national production of the chief crops and th eir various em ployments; production of silk; anim al production; fam ­ ily budgets of peasants, city laborers, etc.

5.

CONTRIBUTIO NS OF FRENCH PHYSIOCRATS

g. Appraisal of Lavoisier’s Estimates Lavoisier’s estim ate represented a great advance over those p repared in F rance be­ fore. It was based upon m ore elaborate statistical data th an h ad ever been assem­ bled, and involved m ore com petent statis­ tical techniques th an h ad ever been applied before. B ut it contained serious defects, some o f w hich had roots in the faulty physi­ ocratic doctrine, and others in the poor qual­ ity o f the inform ation available at the time.

5. Lagrange’s Estimates of France’s Annual Consumption— A Side Venture by a Great Mathemati­ cian Lavoisier’s estim ate was fu rth e r amplified two years after his execution by his close friend, Joseph L ou is Lagrange (1 7 3 6 1 813). A great m athem atician and close friend o f m ost of the philosophers of the tim e (d ’A lem bert, C ondorcet, and o th ers), Lagrange had a long and distinguished ca­ reer. D uring a tenure as professor at the artillery school in T u rin he w on tw o prizes from the F rench A cadem y of Sciences. In ­ vited by Frederick th e G re at to com e to Berlin he spent there tw enty years as D i­ recto r o f th e Prussian A cadem y o f Sciences. H is great work, L a M ecanique A nalytique, was w ritten there. A fter the death o f F re d ­ erick the G reat he returned to F rance at M irabeau’s invitation, and in 1794 was ap­ pointed a professor at the newly established Ecole Superieure N orm ale. In 1796, Lagrange paid his respects to Lavoisier’s w ork and m em ory by publish­ ing anonym ously his A rith m etique Politique. In this w ork he attem pted to m ake L a­ voisier’s estimates m ore m eaningful as in­ terpretations of the standards of living of the population by filling in some o f the gaps in his figures. H e did n o t estim ate the w hole national incom e of F rance, lim iting him self to per capita food consum ption, w hich he deem ed to be the best index of a country’s relative prosperity o r poverty.46 In this essay he divided consum ption into three m ain categories: food; clothing; and housing, including heat and light. T he first he subclassified into vegetal and anim al foods. U nlike Lavoisier, he included am ong vegetal products all garden vegetables and fruits in addition to grain, and am ong ani­ mal products he included dairy products as

75

well as m eat. But, like Lavoisier, Lagrange om itted fowl, eggs, and fish. H e converted the vegetal foods into w heat on the basis of their nourishm ent value and converted all anim al products on the basis of nourish­ m ent values into their equivalents in beef. A ll beverages were converted into their equivalents in wine. W hereas Lavoisier esti­ m ated total vegetal consum ption a t 11.7 billion pounds of breadstuffs and total m eat consum ption at 1.2 billion pounds, L a­ grange, w ith his m ore inclusive calculation, estim ated them at 12.8 billion pounds for breadstuffs and at 3.7 billion pounds for anim al products, i.e., at a 10 p er cent higher figure fo r th e first category and at a 200 per cent higher figure fo r the second.47 N ext, L agrange attem pted to determ ine the total vegetal and m eat consum ption nec­ essary fo r th e sustenance of the people, using to th a t end three sources: (a) food rations distributed to the m em bers of the arm ed forces; ( b ) records of the “octroi” (custom s) offices regarding the quantities and prices of foods brought through the city gates from the farm districts; and (c) evalu­ ation of the annual produce of the land, after adjustm ent fo r im ports and exports. Lagrange found the daily ration of each m em ber of the arm ed forces to consist of 28 ounces o f b read an d h alf a pound of m eat. H e considered each pound o f bread to be equal to a pound of w heat, inasm uch as the w eight of w heat lost in processing was fully offset by the w eight of the w ater added to th e dough and retained in baking. Thus, he said, the soldier requires 1.75 pounds of w heat (b read ) a day. B ut the soldier is one of the physical elite, and his consum ption represents the m axim um in­ dividual consum ption. T aking into account the fact th a t w om en and children consume less than the soldier, L agrange arrived at an average per capita consum ption for the na­ tion o f 1.4 pounds of w heat per day or 511.3 pounds p er year; and of 0.4 of a pound of m eat per day or 146 pounds per year, still on the basis of the m axim um con­ sum ption represented by the soldier. N ext, he considered the available data on the food brought into Paris each day (which then had a population of 600,000). By con­ verting them into w heat and meat, he ar­ rived at a per capita annual consum ption for the population of Paris, o f 435 pounds of w heat and 208 pounds of meat. Finally, he com puted the total annual production of foods in F rance, on the basis

76

THE INCOM E OF NATIONS

o f the data on total cultivated lands and their yields and of the total livestock and its annual slaughter. On this basis he arrived at a per capita consum ption fo r th e nation of 585 pounds of w heat and 80 pounds of m eat, representing a low er level of con­ sum ption than the soldier’s. L agrange then com bined these figures into the weight ratios of bread to m eat con­ sum ption per individual as follows: fo r the soldier 7 weights of bread to 2 of m eat; for the inhabitants of Paris 21 of bread to 10 of m eat; and fo r the w hole of F rance 15 weights of bread to 2 weights of m eat. H e held that the ratio of the weights of these two categories of consum ption constituted a true m easure of the poverty or w ealth of a nation, because well-being is dependent prim arily on food. H e concluded th a t F rance was producing enough w heat but not enough m eat for her people’s nourishm ent, and ad­ vocated m easures that w ould increase the nation’s anim al production. H is estim ate was ingenious, but was m uch too abstract and circum scribed fo r practical use. One hundred and fifty years later, in 1940, Colin C lark devised an “international u nit” for m easuring the national incomes of various countries th a t consisted of the am ount of goods and services purchasable in the U nited States o f A m erica fo r $1, or exchangeable fo r them . This was a m ore refined form of L agrange’s approach of m easuring national incom e in term s of per capita quantities of bread and m eat con­ sum ption and of the ratios of these two goods.48

6. Other Estim ates of the R evolu­ tionary Period A s the R evolutionary governm ent con­ tinued to be severely pressed fo r m oney to finance its m ilitary and civil expenditures (despite its repudiations of the debts and pension lists of the M o narchy), its interest in th e national incom e and in the taxable portion of it rem ained keen. A rnould (1 7 5 0 -1 8 1 2 ), econom ist and di­ rector of the board of com m erce under the Revolutionary governm ent, published a twovolume w ork called Balance de Com m erce in 1790 in w hich he estim ated the rental income from land at 1 billion livres. In 1798 he prepared a special rep o rt on the subject fo r the N ational Assembly contain­ ing a review of a variety of estimates, in­ cluding those prepared by D u P ont and

other physiocrats. In this rep o rt he fixed the rental or taxable incom e fo r 1789 at 1.5 billion livres as com pared w ith L a­ voisier’s estim ate of 1.2 billion livres, and his ow n 1 billion estim ate m ade in 1790.49 H e also referred to an estim ate m ade by d ’A gier in a rep o rt to the C onstituent A s­ sembly w hich fixed th a t incom e at 900 m il­ lion livres. H is conclusion was th a t the “tax­ able” incom e on the eve of the Revolution could n o t have been less than 1.1 billion livres b ut th a t as a result of increased costs of production and a drop in exports, it m ust have dropped since then by 450 m illion and could n ot possibly exceed 650 millions cur­ rently. H e concluded th a t if the landow ners w ere to pay the State the 580 m illion livres tax th a t the A ssem bly planned to impose upon them , they w ould have only 70 million left. H e concerned him self w ith exploring the m eans by w hich agricultural production and rental incom e could be increased so th at larger am ounts w ould be left fo r investm ent in agricultural im provem ents at th e same tim e th a t a larger revenue could be obtained fo r the State.

7. E stim ates by D ’lvernois In 1799, Frangois D ’lvernois (1 7 5 7 1 842), a F ren ch noblem an who escaped to E ngland, published a w ork in both F rench an d English on the losses inflicted upon the F ren ch people by the Revolution and the N apoleonic W ars. In this w ork he reviewed the existing estim ates of F ran ce’s prerevo­ lutionary incom e and offered one o f his ow n.00 A lthough he was obviously biased in some o f his observations regarding current conditions in F rance, his figures of the pre­ revolutionary national incom e deserve at­ tention, fo r he was an econom ist o f con­ siderable com petence. H e published other w orks on econom ics in E ngland and was knighted by the K ing o f E ngland before his retu rn to F ran ce in 1814. D ’lvernois was neither a physiocrat nor a follow er o f A dam Sm ith, and his concept o f national incom e did n ot suffer, therefore, from the narrow ness of either school of economics. H e included in national incom e the incom es of governm ent employees, both civil and m ilitary, of professions, profits of m anufacturers, and incom e from interest on the public debt. But he did not include wages of artisans and journeym en. T his omission was probably due m ore to difficulties of estim ating this incom e th an to theoretical considerations. H e stated his reasons as fol-

5.

77

CONTRIBUTIONS OF FREN C H PHYSIOCRATS

lows: “A lthough th e wages of artisans and journeym en are, to a certain point, n o t only taxable, but actually taxed, they cannot be included in the general account o f th e in­ come of a nation, w hich we should thus multiply w ithout end. This difficulty, how ­ ever, deserves to be m entioned, to show that we ought not to pursue calculations o f this nature w ith any expectation of attaining to the exact truth, but m erely as an illus­ tration of the subject.” 51 D ’lvernois estim ated the national incom e of F rance for 1789 at 3 billion livres. U n­ fortunately, he did n o t docum ent his figures beyond furnishing the following elem ents of his total: 52 Millions of Livres Net income from land Income of farmers Rent on houses in towns Income of West India planters Interest on public debt; annuities, dividends from securities Income of clergy received in tithe Salaries of civil and military em­ ployees and pensions Profits of professions, trades, etc. Profits of manufacturing and trad­ ing establishments and wages of their clerks and workmen

1,100 300 200 100 300 70 100 200 700-750 3,000-3,050

8. Summary of the Estimates, 16 9 0 -1 7 9 8 F ro m th e account in this and the pre­ ceding chapter it can be seen th at by the end o f the eighteenth century, after a hun­ dred years o f theoretical discussion about the natu re and sources of national income, F rance was in the possession of an extraor­ dinary n um ber of national income estimates. H ow ever, th eir statistical bases were ex­ ceedingly w eak. In the theory of national incom e, F ran ce was, perhaps, m ore ad­ vanced th an E ngland; b u t in the statistical substance o f h er estim ates she was m any years behind. T he physiocratic doctrine, w hich tem porarily dom inated F rench eco­ nom ic thought, produced three concepts of national incom e th a t com peted w ith one another fo r first place, an d it was n o t al­ ways clear w hich of them was being used in a particular estim ate. W hat progress was m ade tow ards clarifying the concept of n a­ tional incom e u nder the physiocrats was, in some respects, offset by this unw arranted confusion. W e can now bring together in a single table the principal estimates prepared in F rance during the periods reviewed in this and the preceding chapter, classifying them according to their underlying concepts.

Table 5-4. French National Income Estimates, 1690-1798 Year of Estimate

1690-1700 1768 1776 1789

1798

Estimator

Boisguillebert Vauban Voltaire Le Trosne Du Pont Lavoisier Tolosan Clavi&re D ’lvernois Arnould Young Arnould

National Income Gross Presumably of All Agricultural Gainfully Employed n Income billion livres

Net Rental Income

(millions)

1.5 2.4 2.4 3 .1 -3 .4 4.0 2.7 -3 .0

Population

18 20 1.5 1 . 2 - 1. 5

3 5.2 0.7

25

6. THE THREE LOST RUSSIAN ESTIMATES OF THE END OF THE EIGHTEENTH CENTURY in Russia tow ards the end o f the eighteenth century w ere m any, b u t am ong them these w ere outstanding: first, the trem endous p o ­ litical, social, econom ic, and cultural ad­ vances the country m ade during the eight­ eenth century an d particularly during the long an d brilliant reign o f C atherine the G re at (1 7 6 2 -9 6 ), w hich filled the hearts o f all literate Russians w ith pride b u t which, in some respects, particularly those relating to th e status o f the peasantry, w ere incom ­ plete and, hence, disappointing to some of th e m ost liberal-m inded persons am ong them ; second, the great progress m ade dur­ ing th a t entire century, b ut particularly dur­ ing C atherine’s reign, in econom ic an d sta­ tistical scholarship and records th a t m ade b ro ad inquiries into the state of social and econom ic affairs of th e country an d into the size of its national incom e at this tim e tech­ nically possible; and third, the extraordi­ narily liberal m ood of th e governm ent dur­ ing the greater p a rt o f C atherine’s reign, n o t entertained by any governm ent o f Russia before o r since, w hich perm itted social and econom ic inquiries to be conducted w ith considerable freedom .

Connected by a m ultitude of threads with the developm ent of national incom e esti­ m ates in Britain and F rance, b u t firmly rooted in Russian soil, w ere three national incom e estimates m ade in Russia tow ards the end o f the eighteenth century. These estimates attracted little attention at the tim e, and as they were n o t follow ed by any others in the succeeding century, soon passed into oblivion. O ne of them , lay B. F . J. H erm ann, appeared in a section of a book published in G erm an in St. Petersburg w hich had only a lim ited circulation. T he second, by A. N. Radishchev, was incorpo­ rated in a private m anuscript unknow n at the tim e and published only 113 years later; and the third, prepared by a group of anony­ m ous scholars, appeared in a book of politi­ cal and economic geography w hich was sup­ pressed by the governm ent im m ediately after its publication, and w hich rem ained u n ­ know n to the general public fo r the next 150 years. T he existence of these three estimates was thus little know n until Professor M. V. Ptucha, m em ber of the Soviet A cadem y of Sciences, brought them to light in the first volume of his H istory o f Statistics o f the U .S.S.R., first published in 1946.1 T he appearance of these estim ates was fated to rem ain an isolated event in the history of Russian economics, statistics, and national income estim ates fo r m ore than a hundred years. T his was so because these first estimates appeared during a unique period of extraordinary freedom in social inquiry u nder a m onarchy th a t had, fo r this short while, been swept off its feet by the spirit of liberalism th a t reigned over the w hole of W estern E urope at th a t tim e.

a. Territorial Expansion P eter the G re at’s opening up o f Russia to intercourse w ith W estern E u ro p e was probably the greatest event in w orld history of the end o f th e seventeenth and beginning o f the eighteenth centuries. Its im portance was obviously especially great fo r Russia herself. W ithin less th an a century Russia changed from an inaccessible barbaric coun­ try o f little consequence in w orld politics an d com m erce into a great E uropean po­ litical, m ilitary, and com m ercial pow er to be taken into account by other W estern pow ers in all m ajor m atters of political and com m ercial policy. R ussia was growing in all directions during these years, partly th ro u g h independent conquests o f lands form erly belonging to Sweden an d Turkey, and partly through th e participation w ith

1. The Trem endous Transforma­ tion of the Country During the Century T he factors th a t were responsible fo r the appearance o f national incom e estimates Notes begin on p. 521. 78

79

6 . THREE LOST RUSSIAN ESTIM ATES

Prussia and A ustria in the conquest and partition of Poland. H er territory now reached northw ard along the G ulf o f F in ­ land, w estward tow ards and along the coast of the Baltic Sea and the boundaries of Prussia and A ustria, an d south and east to the Black Sea (a stone’s throw from the M ed iterranean), th e C aspian Sea, and the foothills of the Caucasian M ountains.

b. Growth of Population and Economic Activity A fter having been stationary fo r years, and even declining occasionally as a result o f pestilence o r w ar, R ussia’s population during the eighteenth century alm ost dou­ bled, rising from some fifteen to some thirty m illions.2 A p a rt o f this increase was due to the annexation o f foreign lands and peo­ ples, b u t m ost of it w as th e result o f im ­ provem ents m ade in the standard o f living an d health of th e people and the resulting drop in m ortality rates. P . I. Liashchenko gives the following figures o f Russian popu­ lation an d its com position during the cen­ tury:

and other goods. Some of this new produc­ tion was developed by state-owned enter­ prise, b ut m ost was organized by the land­ ow ners on their estates largely with their own serfs as laborers o r by the new middleclass entrepreneurs in the cities either with purchased slave labor or w ith free labor. T he system of internal transportation and com m unication was vastly improved, and b oth internal and external trade grew apace. P aper m oney was introduced by the gov­ ernm ent and the newly established banks lent m oney to landowners, m anufacturers, and m erchants. St. Petersburg was built with m arble and granite as one of the m ost beau­ tiful capitals and aggregations of sum ptuous palaces in the w orld. T he system of prim i­ tive or domestic econom y was everywhere being displaced by m onetary exchanges, and the national econom y as a w hole was rapidly becom ing capitalistic. T he new w ealth was finding its way into the coffers o f the state, the large landowners, the higher govern­ m ent officials, and the larger m erchants and m anufacturers. T he small landow ners and m erchants were n ot gaining m uch, while the

T ab le 6-1. G ro w th of R u ssia’s P opulation 1722-1815 (m illions)

Y ear F irs t census, 1722 T h ird census, 1762 F ifth census, 1796 S ev en th census, 1815

In B asic A rea 14

In A nnexed A rea —

a

29 30.5

a 7 14. 5

T o ta l 14 19 36 45

U rb an (m illions)

%

0.5

3.6

—

1. 3 1. 7

—

3.5 3.8

R ural (m illions) 13. 5 1 4 . 5» 34 . 7 43.3

% 96. 4 —

96.4 95 . 2

Serfs (m illions) —

7.6 b 20. 0 20. 8

% —

52.4 b 55. 5 46.2

“ N o t d ifferentiated betw een p rim a ry a n d annexed areas. b G re a t R ussia a n d S iberia alone; th e p e r c e n t is of ru ra l po p u latio n only. SO U R C E:

P. I. L iashchenko, Isto ria Narodnago K hosiaistva S S S R , M oscow, 1956, vol. 1, p. 396.

T he settlem ent and cultivation o f the vast “black earth” belt of the U krainian steppes (th at had been w rested from T urkey) alone added greatly to the bread and cereal sup­ ply of the country and to th e volum e of its exportable com m odities exchangeable fo r foreign goods. The developm ent of m ining and smelting w orks in the U ral m ountains and other parts o f Siberia also added greatly to the country’s w ealth (especially as iron becam e one of its chief exportable com ­ m odities) as did the grand developm ent of shipbuilding, quarrying, m anufacturing of brick and cem ent, the increase in residen­ tial and other building, and the grow th of m anufacturing of w oolen, linen, leather,

position of the peasants, u nder the existing systems of serfdom , continued to be about as hopeless as ever.

c. Advancement of Education E ducation and the sciences were also greatly encouraged by the governm ent, but here again their benefits accrued mostly to the privileged classes. P eter the G reat es­ tablished in St. Petersburg, Moscow, and other cities the first secondary schools, called “gym nasia,” and the first m ilitary and naval schools or academies. H e also founded the A cadem y of Sciences in St. Petersburg in 1725, as b oth a research center and the first university in the country. A ttached to

80

THE INCO M E OF NATIONS

it were two gymnasiums designed, in part, to prepare students fo r adm ission to study in the Academ y. A fter P eter’s reign, but following the lead set by him, the U niversity of Moscow was established in 1755, the M ining Institute was opened in St. Peters­ burg in 1773, and m ore gymnasiums w ere started in St. Petersburg, Moscow, and other cities in various years. N early all of these higher and secondary institutions were es­ tablished prim arily fo r the benefit of the children of the privileged classes; but one or two semimilitary and quasi-secondary schools w ere also open in p art to children of com m on people. T he education of the low er classes was provided largely through the parochial schools under an act of 1721. Finally, under C atherine the G reat, the first com m on schools w ere established in the principal cities providing from two to five years of instruction to the children o f the low er classes.

d. The Rule of Bureaucracy Politically, the rule of the high nobility, w hich was tem porarily revived after the death of Peter the G reat, was definitely curbed under C atherine the G reat and sup­ planted by the rule of a centralized bureauc­ racy recruited from various classes of the population, including the small nobility, the middle classes, and the com m oners. This bureaucracy, w hich perm eated the adm inis­ trative structure of the governm ent from the highest councils around the throne down to the lowliest office in the rem otest portion o f the country, provided the throne with m uch firmer support than the high nobility was either willing to give o r capable of giving.3

e. Influx of Scholars and Technicians from Abroad All these progressive developm ents in adm inistration, m ilitary arts, industrial pro ­ duction, trade, and education, w ere substan­ tially furthered by the influx into the coun­ try, particularly in the days of C atherine the G reat, of thousands of im m igrants from Sweden, H olland, D enm ark, G erm any, A us­ tria, France, Switzerland, Italy, and other lands. Some of them cam e to Russia p ri­ m arily in search of personal fortunes or pow er; others in search of adventure; still others in search of opportunity to do con­ structive scientific work.

2. The G rowth of Statistical D ata

and Inquiries During the First Half of the Century In the sixteenth and seventeenth centuries, Russia h ad only some scattered statistical d ata com piled sporadically by the official scribes of the M oscow governm ent w ho vis­ ited the various cities and provinces of the country, recording a variety of adm inistra­ tive, geographic, social, economic, an d fiscal facts. Especially im portant am ong these were the censuses or assessments of land, houses, and people m ade in 1646 and 1678 fo r the levy of the national land and house taxes.4 O ther im portant records w ere the local m aps, sim ilarly developed by the scribes, and the “G reat M ap o f Russia” evolved from them and repeatedly revised.5 In the seventeenth century an illum inating w ork appeared in Sweden, describing a wide variety o f facts about Russia. I t was p re­ p ared by a form er Russian scribe, I. K. K otuchihin (1 6 3 0 -1 6 6 7 ), a m em ber of sev­ eral missions sent abroad by M oscow to negotiate treaties, who, on one of these missions, chose to escape and m ade his way to Sweden.6 Some two hu n d red years later, in 1860, a volum inous and exceedingly well com ­ posed m anuscript on various aspects of the R ussian state was discovered, p repared by a C roatian theologian, G. K rizanic (1 6 1 7 1 6 6 3 ), w ho h ad com e to Russia apparently to propagate Catholicism, b u t was prom ptly arrested and sent to Siberia, w here he spent eighteen years in exile and w rote his survey. In the first p a rt of this w ork he dealt m ainly w ith the w ealth and econom ic structure of the country, while in the second he treated its governm ent and population. In both parts he incorporated m any critical rem arks and penetrating generalizations.7

a. The First Censuses of Population T he systematic collection of statistics be­ gan during the reign of Peter the G reat (1 6 8 2 -1 7 2 5 ) and particularly in the latter portion of it. T he prim e reason for th e im ­ provem ent in collection of statistics was fis­ cal in nature. P eter needed m ore revenue th an he was receiving in order to shore up a treasury exhausted by his m any w ars and other costly undertakings. T he old house tax was producing a dim inishing revenue, as th e n um ber of houses had shrunk as a result o f the destruction o r abandonm ent of

6 . TH REE LOST RUSSIAN ESTIM ATES

m any o f them during this period o f general distress. M any people avoided paying the house tax by giving up their houses and moving, w ith other fam ilies, into one larger, or enlarged hom e. T hus, the house census of 1710 showed th a t the num ber of houses in the country had decreased since the house census of 1678 by some 20 p er cent and th at m any houses had 25 to 30 inhabitants. U nder these circum stances, P eter ordered, in 1718, a changeover from the house tax to a poll tax, w ith the am ount p er head varying according to the social status of the individual (a m ale a d u lt), i.e., according as he was a peasant belonging to the state, a peasant belonging to a private landow ner, a petty urban tradesm an, a m em ber o f a craftsm an’s guild, o r a m erchant. In th e case of peasants the head tax was payable by the landlords, who, in turn, collected it from the peasants; but the landlords w ere held responsible fo r the paym ent o f th e full am ount to the state. T o impose this levy, P eter ordered th a t a census be taken of the entire population, m ale and female, adult and child, ablebodied and invalid, alike. A ll landlords w ere required, under penalty of death, to m ake accurate return to the local civil authorities of all their serfs, domestics, and other peo­ ple, w ith indication of their nam es, sex, w hether adult o r child, and other such in­ form ation. In addition to serving a fiscal purpose, the census was also to provide a basis fo r setting the quotas of recruits for the arm y from each district o r village. But the returns w ere slow in com ing and w ere patently incom plete. In 1722, therefore, P eter ordered th a t th e census should be com pleted w ith the aid of the troops and th at th e results should be tabulated in each district in a prescribed m anner by the com ­ m anding officer and the tabulation be then sent to the Senate fo r final collation. In this m anner the first census was finally com ­ pleted in 1724. A t the sam e tim e, P eter is­ sued an order th a t sim ilar censuses be taken in the future every tw enty years.8 T he nu m ­ ber of the m ale adults returned fo r a given district was to be used as a basis fo r the assessment of the head tax upon th a t dis­ trict fo r the next tw enty years. If in the m eantim e the num ber of people subject to the tax in the district decreased, the assess­ m ent was to be distributed over the sm aller num ber, with the resultant increase in the burden fo r each living person; if the nu m ­ ber increased, the opposite result w ould

81

follow, the burden per head becoming lighter. Subsequent inspections of the re­ turns of the first census revealed consider­ able under-reporting by landowners. The offenders, however, w ere only fined for these breaches of the law. A lthough the census of 1718-24 was im perfect, it constituted nonetheless the m ost advanced statistical collation of popu­ lation ever carried out in the world up to th a t tim e, as no other country had as yet taken a census. In 1744, years after P eter’s death, but u nder the law prom ulgated by him, the sec­ ond population census was taken, likewise by the m ilitary. A prelim inary report by the Senate, m ade in 1742, had indicated th a t in m any sections of the country the population had decreased through death o r the running away o f the serfs, th a t the total num ber of individuals subject to the levy had decreased by 2.8 million; and th at the uncollected taxes now aggregated the colossal sum of 6 m il­ lion rubles. T he census was com pleted in 1747, m ore accurately and quickly th an the first, and a new assessment was based upon it.9

b. The First Vital and Economic Statistics D uring the reign of P eter the G reat the first steps were taken tow ards the collection of vital and econom ic statistics on a nation­ wide basis. In 1722, acting through the newly established Synod (H igh Council of the G reek O rthodox C h u rc h ), Peter ordered all parish priests to keep a book of records o f all m arriages, births, and deaths occur­ ring in th eir parish and to transm it sum ­ m aries of them to th eir bishops every four m onths, w ho in tu rn were required to trans­ m it them to the Synod. U n d er this order, vital statistics began to accum ulate in the Synod offices. B ut they rem ained there for years, unclassified and unsum m ated until, finally, in the latter quarter of the century scholars began to exam ine them in an effort to obtain m ore m eaningful figures on the rates of birth and m ortality.10 A t about the same tim e, while establish­ ing a system of local adm inistration of cities under central supervision, patterned in p art on the E uropean models, Peter also ordered the city officials (m ayors, police chiefs, etc.) to m aintain a cu rren t register of all the city inhabitants, noting their sex, age, occupa­ tion, num ber o f persons in the family, num ­ ber o f domestics and other employees. This

82

THE INCOM E OF NATIONS

register was to be used for the levy of the city poll taxes and also for nonfiscal pu r­ poses. T he local officials were required to tabulate these data and to transm it the tabu­ lations at stated intervals to the C entral B oard for City Affairs. These tabulated records eventually provided an invaluable source of inform ation on the grow th of city populations and changes in occupations. Inasm uch as the cities were centers of trade and m anufactures, these records also shed considerable light on the grow th of the nation’s industry and com m erce.11 In fostering the developm ent of m ining and metallurgy, partly through state ow ner­ ship and partly through grants of m onopoly rights to large landow ners, P eter the G reat also required the m anagers o f all m ining and smelting w orks to m aintain certain records of their operations, thus laying the foundation fo r the collection of national economic statistics fo r this im portant in­ dustry.12

c. The Economic and Statistical Inquiries of the First Half of the Century T he statistics w hich thus began to be as­ sembled during P eter’s regim e and w hich continued to be assembled under the regimes of his successors until the ascendance to the throne of C atherine the G reat in 1762, were far from being com plete. T hey w ere n o t well collated by the central agencies; n o r were they published. It rem ained fo r individual scholars, generally adm inistrative officials of some kind, to dig up these records in the archives and to derive some m eaningful summ ations from them . W ith the aid of these archives an exten­ sive literature descriptive of the various as­ pects of the country began to be prepared by scholars of distinction. All of them in one way o r another were associated w ith some adm inistrative office or with the A cadem y of Sciences in St. Petersburg, w hich actively fostered such w orks and either published them in its proceedings o r kept them in m anuscript form in its files fo r the con­ venience of other scholars. M ost of these w orks were in the nature o f political and econom ic geographies rath er th an statistical com pilations or studies as these are com ­ m only conceived today. Because of the in­ com pleteness of the figures o r the inaccessi­ bility of records containing them , the schol­ ars used verbal descriptions m ore generally than figures.

P articularly outstanding examples o f this kind of w ork w ere those w ritten by I. K. K irillov (1 6 8 9 -1 7 3 7 ), chief secretary of the Senate and an indefatigable w orker who published a new m ap o f Russia at his own expense and spent ten years preparing a volum inous m anuscript descriptive o f al­ m ost every aspect o f the life of the country. This, how ever, rem ained in the archives of the Senate in unpublished form till 1831, w hen it was finally m ade available to the general public as an im portant historical docum ent.13 A nother great scholar of this period was the dynam ic V. N . T atishchev (1 6 8 6 -1 7 5 0 ), gifted geographer, historiographer, political econom ist, statistician, and adm inistrator, a liberal statesm an, an d one o f the best edu­ cated m en, w ho h ad spent several years in Sweden on missions fo r Peter and had be­ com e thoroughly acquainted with the tech­ nological and social advances m ade there. H e w rote m any geographic reports and an extensive history of Russia fo r the A cadem y o f Sciences, w hich were either published in the volum es of the A cadem y o r filed w ith it in m anuscript form . In his last years he was u n d er investigation by the governm ent fo r his advanced views and was confined to his country estate.14 Econom ics, as a science proper, was wholly undeveloped in Russia during the first half o f the century. Some m ercantilist w ritings in foreign languages did seep through to Russia and becam e know n to h er best educated m en, who generally knew F rench or G erm an, and one w ork of do­ mestic origin did appear during th a t tim e. It was w ritten by I. T. Pososhkov (1 6 6 5 1725) and was called P overty and Wealth. T he au th o r was a self-educated peasant who h ad becom e a prosperous m erchant b ut who h ad never lost his sense of identity w ith the com m on m an. H is book, w ritten in simple b ut tren ch an t language, was a pow erful plea fo r the abolition of serfdom and other form s o f oppression and fo r an end to the abuse of pow er by the ruling class. Like Bois­ guillebert and V auban in F rance before him (w ith whose w orks he probably was n o t ac­ q u ain ted ), Pososhkov em phasized th a t a rich state treasury could exist only in con­ junction w ith a prosperous population. But he favored a m ercantilist policy fostering the developm ent of m anufacturing by p ro ­ tective tariffs, regulation by governm ent of the quality o f dom estic products, and the

6 . THREE LOST RUSSIAN ESTIM ATES

im position of fines on substandard products. H e published his book at his own expense in 1724 and presented it to P eter in the hope th at it w ould receive sym pathetic hearing from him . Instead, P ete r broke into a rage, ordered the book to be suppressed fo rth ­ w ith, and threw the author into prison, w here he died shortly afterw ard.15

3. The G reat Im provem ents in Sta­ tistics Under Catherine the G reat D uring the long reign of C atherine the G reat, often described as the “G olden Age of Russia,” the governm ent placed the col­ lection of statistics on a m uch broader and firm er foundation th an th a t established by Peter the G reat; and scholars w ere encour­ aged to m ake better use of them in devel­ oping policy recom m endations aim ed at ad­ vancing th e well-being of the people and the prosperity of th e state. Intense interest in population theories and econom ics devel­ oped in governm ental circles and am ong the educated classes generally, resulting in the appearance of num erous well-organized studies. Im m ediately after h er ascendance to the throne in 1762, C atherine was confronted w ith th e problem o f organizing the third population census to be taken u n d er the law. She becam e convinced th a t taking it w ith the aid of troops was thoroughly bad. First o f all, it was exceedingly costly. E ach of the previous censuses u nder this system had cost a m illion rubles; and the T reasury could scarcely bear such an expense. Sec­ ondly, the use of the troops had caused con­ siderable perturbation in the country every tim e. Peasants w ere frightened by the ap­ pearance of the soldiers, believing (n o t w ith­ out reason) th a t the purpose o f listing them was to recruit them into the arm y. Conse­ quently, the peasants hid themselves in the woods and swamps or ran aw ay across the frontier into P oland just to escape being listed. C atherine conceived o f a better solu­ tion: she dispensed w ith th e use o f troops and ordered all proprietors to rep o rt to the district civil authorities a list o f all their people and serfs w ith the requisite inform a­ tion about their sex, age, etc. These lists w ere then to be transm itted to the provin­ cial governors fo r collation and fo r further transm ittal to the Senate. Proprietors who

83

had failed in the past to m ake com plete re ­ ports w ere to be forgiven provided they fully com plied w ith the new order. Surpris­ ingly enough, this new m ethod worked com paratively well, producing m ore com ­ plete an d b etter organized returns which lent themselves better to sum m ation.16 In 1775, C atherine reorganized and tight­ ened up the w hole system of provincial and m unicipal adm inistration. T he num ber of provinces was increased from tw enty to fifty. E ach province or "gubernia” was di­ vided into a netw ork of small districts ( uyesd) , an d these into still smaller ones (v o lo st), and at the head of each unit was placed a civil an d police official of appro­ priate rank. T he cities were given lim ited powers of self-governm ent th a t were to be exercised by them through an elected m u­ nicipal council and m ayor. T he right to vote was given only to the property-owning por­ tion of the city population. This portion was small (only 6 per cent of th e total, in St. Petersburg) and was subdivided according to the social an d econom ic status of the indi­ viduals into six categories, w ith the upper groups consisting of the wealthiest citizens being given dom inant representation. The operations of these organs of self-governm ent w ere m ade subject to close supervision by the provincial governor and his aids. Police adm inistration was kept outside the jurisdiction of the city councils and mayors, being responsible only to the provincial gov­ ernor. This system of centralized bureauc­ racy an d som ew hat shadowy m unicipal selfgovernm ent provided a fairly effective ap­ paratus n ot only fo r the transm ission of orders from above, but also for the collec­ tion of adm inistrative and economic statis­ tics and other inform ation from below.17 A t the same tim e, C atherine instituted other econom ic reform s. T erm inating one by one the private monopolies granted by P eter the G reat, C atherine issued a m ani­ festo in 1775 establishing the freedom of private m anufacturing and trade and declar­ ing specifically th at “any person has the right to institute all kinds o f looms and m achines and to m anufacture with them all kinds of goods.” On the other hand, in fo r­ eign trad e she applied protectionism even m ore vigorously th an it had been applied before. In governm ent finances, she avoided the im position of new levies, resorting in­ stead to large internal and external loans and to the issuance o f pap er m oney— in other

84

THE INCOM E OF NATIONS

words, to pure inflation. Starting in 1772 with the issue of 1 m illion rubles of redeem ­ able paper money, by 1786 C atherine had put into circulation 100 m illion rubles of irredeem able paper currency and by 1796 157 millions, w ith the result th at the paper currency depreciated, in term s o f coin money, 30 per cent, and that the prices of all goods rose.18 But C atherine’s reign was far from being consistently liberal even in its early days. C ontrary to her originally declared inten­ tions to lighten the burdens of the serfs and even to em ancipate them altogether, she m ade their burdens m ore onerous. M oved by a desire to insure fo r the throne the com plete loyalty of the nobility, she gave the landowners even m ore absolute pow er over their serfs than they h ad before, af­ fording the peasants no protection w hatso­ ever against inhum ane treatm ent by their m asters; and she extended serfdom through the whole of U kraine, where serfdom did not exist before. These m easures brought on a bloody rebellion am ong the peasantry in the southeastern provinces, w hich was led by Y em elian Pougatchev (1 7 3 0 -1 7 7 5 ), a D on Cossack, who pretended to be Tzar P eter III, C atherine’s husband, m iraculously escaped from his would-be assassins, and w ho prom ised to liberate all peasants who would join him. A lthough the rebellion was suppressed eventually and Pougatchev was executed, the m em ory of it haunted C ath­ erine to the end of her life, causing her to be in perpetual terror o f another rebellion and to apply severe repressive m easures on the slightest provocation, blotting out com ­ pletely her early liberal record. D uring C atherine’s regim e, a new group of Russian scholars was com ing to the fore who studied statistically and otherw ise the social, economic, and adm inistrative devel­ opments of this new era. T hey were young persons w ho had been trained in foreign universities or who had otherwise become fam iliar w ith the latest in foreign statistical thought and literature. A t the same tim e, a large flock of able foreign statisticians and political scientists was com ing into Russia, either for lim ited appointm ents at the St. Petersburg A cadem y of Sciences o r other institutions of higher learning, o r to stay with them for life. They were attracted to Russia by the vast opportunities to carry on constructive scientific w ork opened up to them under the new liberal regim e.

4. The Convergence in Russia of Three Schools of Statistics D e­ veloped A broad T he three schools of statistics th a t had developed in the w orld at the tim e, con­ verged or, in a way, collided w ith each other in Russia during this period. One of them was th e school o f “political arith­ m etic” developed in E ngland by John G rau n t an d W illiam Petty (as detailed in C hapter 2 ab o v e ); the second was the F ren ch school o f m athem atical statistics based on th e theory of probability; and the th ird was the G erm an Conring-A chenwall school of descriptive treatm ent of the facts o f governm ent. T he convergence in Russia a t this tim e o f these three schools rep re­ sents one of th e m ost interesting incidents in the history of statistics. W e are con­ cerned w ith this incident here, however, only to th e extent th a t it has a bearing on the history of national income estimates.

a. The French Mathematical School T he F ren ch school need n o t be dwelt u p o n here, fo r it h ad the least im pact on the developm ent of Russian statistics in the eighteenth century an d played no hand w hatsoever in the initiation there o f national incom e estim ates. It was represented in Russia by D aniel Bernoulli (1 7 0 0 -1 7 8 2 ), m em ber o f the fam ous Swiss fam ily of m athem aticians, w ho spent eight years in Russia (1 7 2 5 -3 3 ) as a m em ber and p ro ­ fessor of th e St. P etersburg A cadem y of Sciences, b u t continued his collaboration w ith the A cadem y throughout th e rest of his life, contributing to its publications some o f his m ost fam ous m athem atical w orks on the subjects of longevity and on the theory o f probability. Bernoulli had practically no following in Russia at the tim e, fo r his w ork was m uch too theoretical and abstract. A s Professor P tu ch a rightly puts it, “Russia of the second h alf of the eighteenth century needed works of a w holly different n atu re from w hich any educated person could obtain interesting and useful inform ation . . . and w hich could convince adm inistrative circles and public opinion of th e im portance and usefulness o f the developm ent and collation o f basic statistical records.” 19 A m ong oth er exponents of the F rench m athem atical schools o f statistics whose nam es w ere know n in Russia, b ut whose

6 . TH REE LOST RUSSIAN ESTIM ATES

w ork did not m ake too m uch of an im pres­ sion on the Russian statisticians at the tim e was A ntoine D eparcieux (1 7 0 3 -1 7 6 8 ) who sim ilarly wrote on longevity and even de­ veloped one o f the first m ortality tables in the world.

b. The German Descriptive School T he G erm an school concerned itself p ri­ m arily with collecting governm ental and adm inistrative facts and was closely allied to the field of public law. It was a purely descriptive approach, using verbal exposi­ tion m uch m ore extensively than figures, in­ asm uch as precise figures w ere difficult to obtain and others could be only “estim ated” and, hence, were subject to error. It con­ tented itself w ith statem ents o f facts, and avoided any explanation o r interpretation of them , leaving this task to statesm en or to philosophers, contending th a t explana­ tions and interpretations involved opinions and were, therefore, unscientific. T his school of statistics was originated in G erm any in the m iddle o f the preceding century by H erm ann C onring (1 6 0 6 -1 6 8 1 ), b u t was developed in fullest detail and m ost effec­ tive form only in 1748-49 by G ottfried A chenw all (1 7 1 9 -1 7 7 2 ) o f the U niversity of G ottingen. It was Achenwall w ho coined the w ord “Statistik” and gave it currency, although it cam e to be used, in later years, very differ­ ently th an he h ad intended, to m ean a study of exclusively num erical data. A chenw all defined statistics as “the study o f the po­ litical arrangem ents of the m odern states,” thus conceiving it as a practical study of the facts of governm ent and public adm in­ istration m ore than the science of statistics as it exists today. A chenw all’s books w ere widely read by Russian scholars, even though they w ere w ritten in G erm an; and Russian sum m aries of them soon appeared in Russian books. A m ong the m any Russian follow ers of the G erm an school w ere M . M . Shcherbatov (1 7 3 3 -1 7 9 0 ), w ho in 1755-77 w rote a m anuscript entitled Statistics R elating to Russia, thus being one of the first Russians to adopt A chenw all’s nam e fo r this new science. Shcherbatov covered an exceed­ ingly wide range of political, social, and econom ic subjects, m ostly in verbal term s. H e was a gifted m an, but he expressed an extrem ely reactionary point of view. Shcher­ batov was the spokesm an of the highest

85

aristocracy and criticized the m onarchy for being too dem ocratic. E ven m ore im portant was the twentyone volum e H istorical Description o f the Russian Trade o f M. D. Tchulkov (1 7 4 0 1793), w hich was published in 1788 with the financial aid of I. I. Golikov (1735— 1801), a w ealthy m erchant and himself a historian and statistician of note. In his statistical approach, T chulkov stood m id­ w ay between the G erm an and the English schools. In his econom ic views he showed the influence of A dam Sm ith’s works. H e favored the developm ent of m anufacturing and com m erce on the grounds th a t they w ere far m ore productive th an agriculture and generated m ore national w ealth.20 A m ong th e G erm an scholars w ho had com e to Russia to teach at the A cadem y of Sciences and other institutions of higher learning and w ho w ere the exponents of the Conring-A chenw all statistical school w ere A. F . Biisching (1 7 2 4 -1 7 9 3 ), author o f a fam ous multi-volum e com parative sta­ tistical study o f the governm ents of various countries, published originally in G erm an b u t soon translated into several other lan­ guages, including Russian; I. G. Reichel (deceased in 1 7 7 8 ), professor at Moscow U niversity w ho w rote the first Russian text­ book in statistics closely following the A chenw all line; and A . L. Schlotzer (1 7 3 5 1809) w ho spent fo u r years in Russia (1 7 6 1 -6 5 ) b u t published a num ber of sta­ tistical w orks about Russia after his return to G erm any w here h e took over A chenw all’s chair at G ottingen; and A. B. H upei (1 7 3 6 -1 8 1 9 ), author of the two-volume w ork, w ritten in G erm an, The Structure o f the Russian State (1 7 9 1 -1 7 9 3 ) about whom m ore will be said later.21 T he w orks o f these Russian and foreign scholars, w ritten largely in the vein of the G erm an statistical school, were of great help in the developm ent o f adm inistrative sta­ tistics in Russia, b u t they contributed com ­ paratively little to any understanding of the facts w hich they presented or of the forces th a t controlled these facts. In this latter re­ spect they were alm ost wholly sterile.

c. The English Political Arithmetic School T he English school of political arith­ m etic was com pletely different. It em pha­ sized num bers in the treatm ent of social, economic, and political phenom ena, sought

86

THE INCO M E OF NATIONS

to discover their cause and effect relation­ ships, and, while being thoroughly em piri­ cal, engaged in bold “estim ates,” specula­ tions, and generalizations, and, in its m ost im portant applications, was closely allied with political economy. It was well know n to the Russian political scientists o f the second half of the eighteenth century both through the w orks in English of John G raunt, W illiam Petty, and Charles D ave­ nant and through the writings of the la tte r’s followers in H olland, Sweden, and F rance and the few followers in G erm any, m ost of w hich were available in Russia. Particularly influential am ong these fo r­ eign works w ere those of the Swedish acad­ em ician P. V argentine in vital statistics; the book of J. P. Siissmilch (1 7 0 7 -1 7 6 7 ), the G ottliche O rdnung (1 7 4 1 ), w hich was influenced by G raunt and Petty and w hich even went beyond them in the application of the same approach to dem ographic specu­ lations and to economics; and the fam ous w ork of J. F . Bielfeld (1 7 1 7 -1 7 7 0 ), The Political Instructions, w hich appeared in F rench in 1760 and was soon translated into G erm an, Russian, and Italian, w hich gave a fairly com plete and wholly favor­ able exposition of the school of political arithm etic and of its contribution to the theory of population and to econom ics.22 One of the chief exponents of this school of statistics am ong Russian scholars of the tim e was M . V. Lom onosov (1 7 1 1 -1 7 6 5 ), the giant o f Russian social and physical sciences and the reform er o f th e Russian language. T he son of a peasant living in the depths of the A rchangel province, a runaw ay from hom e, educated first in a Russian m onastery and next sent to G er­ m any under a scholarship from th e St. Petersburg A cadem y of Sciences, he studied under Christian W olff (1 6 7 9 -1 7 5 4 ), the great liberal philosopher and rationalist. Lom onosov eventually becam e the leader o f the St. Petersburg A cadem y and a corre­ spondent of the greatest E uropean scien­ tists o f the tim e. H is contributions along the lines of political arithm etic in both dem ographic and econom ic studies relating to Russia becam e considerable, particularly during the period 1757-65.23 In th e last years of his life, Lom onosov launched a broad “political and econom ic study of the country” to be conducted at the A cadem y w ith the aid of his disciples, th a t fa r sur­ passed in scope any ever attem pted in the country before. It was so broad and delved

so deeply into the existing social and eco­ nom ic system th a t it frightened the govern­ m ent, w hich ordered the A cadem y to take th e project o u t of Lom onosov’s hands and even to drop it altogether.24 A n o th er outstanding exponent of the same political arithm etic approach am ong R ussian social science scholars of the tim e was A. N . R adishchev about w hom m ore is said later (see Section 9 below ). A large group o f political arithm eticians also col­ lected in an d around th e “V oluntary E co­ nom ic Society” founded in St. Petersburg in 1765 and the “Society of Friends of the L iterary Sciences” w hich Radishchev helped to establish there in the 1780’s. A m ong the foreign-born scholars then resident in Russia w ho used the “political arithm etic” appro ach to social an d eco­ nom ic studies was W . L . K ra ft (1 7 4 3 1814), professor at the St. Petersburg A cad­ emy. A lthough of G erm an parentage and a graduate of the U niversity of Tubingen, he was n o t a follow er o f the Achenwall school. A n active m em ber of th e V oluntary E co­ nom ic Society, he w rote several notable papers fo r the A cadem y dealing w ith the population problem , em phasizing the im ­ portance o f th e econom ic factor. H is fu n ­ dam ental thesis was th a t the w ealth of a nation consisted prim arily of its people and o f th eir well-being; th a t an increase in population was an im portant factor in eco­ nom ic grow th; th a t the interests of th e state an d o f the people are identical; and th a t an enlightened governm ent m ust devote all its efforts to th e advancem ent of the well-being o f its people. K ra ft set up an index with fifteen item s in it, all dealt w ith in n um er­ ical term s, th a t p u rp o rted to m easure the grow th o f th e population an d th e advance­ m ent o f its w elfare. H e called this index “a political therm om eter th a t can indicate to th e rulers o f a co untry the degree o f wel­ fare o f its people even in the rem otest sec­ tions o f the country, and the changes cur­ rently occurring in th eir well-being as well as in th eir social needs.” This index, he thought, w ould also indicate to th e authori­ ties th e m easures by w hich they could “ac­ tively an d energetically influence th e con­ servation of and grow th in the n u m b er of th eir subjects . . . and in the well-being of the m ultitude o f its citizens.” 25 T hus, in a way, K ra ft anticipated th e m o d em statis­ tical indices used by governm ents as a basis fo r the form ulation o f th eir social and eco­ nom ic policies; an d extended the political

6 . THREE LOST RUSSIAN ESTIM ATES

arithm etic approach fa r beyond the bound­ aries established by any social o r econom ic statistician anyw here at th a t tim e. A n o th er foreign-born scholar who con­ tributed to econom ic surveys of the politicoarithm etic type was B. F . J. H erm ann, author o f the first national incom e estimates in Russia about w hom m ore will be said in Section 8 below. In this period another scholar o f G er­ m an parentage w ho, a t least in his earlier writings, showed inclinations tow ards p o ­ litical arithm etic w as H en ry (H einrich) Storch (1 7 6 6 -1 8 3 5 ). H e published, in 1794, a book in G erm an entitled A Picture of St. Petersburg which, although statistical in n a­ ture, was w ritten in fascinating style and attracted especially w ide attention abroad. Storch was born in Riga and was educated at the universities of Jena and H eidelberg, specializing in philosophy, jurisprudence, and political science. H e cam e to St. Peters­ burg in 1787 and becam e successively pro­ fessor of history and literature a t T he M ili­ tary Academ y, official o f the m inistry of foreign affairs, professor o f statistics at the A cadem y of Sciences, and, beginning in 1813, vice-president of th e A cadem y. In his Picture of St. Petersburg, he wrote, “A n educated E uropean citizen will n o t be content w ith th e m ere know ledge o f the size of the population o f a country. H e will w ant to know the tendency o f th a t popula­ tion to increase o r decrease and also the fac­ tors w hich facilitate o r ham per its growth. This knowledge, w hich is attained through the application o f political arithm etic, is founded on the explorations o f the rates o f m arriages, births, and m ortality and o f th e health o f a people.” 26 A m ong the m any interesting items in this notable w ork was one presenting the budget of a well-to-do gentlem an who com es to St. Petersburg to set up a tem porary residence and has an income of 194 rubles a m onth. T he budget consisted of ten categories o f outlay and was used by Storch fo r a num ber o f interest­ ing conclusions regarding the m odes of life of the rich, the m iddle-incom e group, and the poor, and the effects o f these on econom ic welfare. Storch followed this w ork w ith the p ub­ lication in 1801-3, likewise in G erm an, of a nine-volume Historico-Statistical Pic­ ture of the Russian Empire 27 in w hich he likewise used politico-arithm etical analysis. Eventually, Storch gave up statistics and devoted himself w holly to political econ­

87

omy, publishing a six-volume treatise on political econom y and various essays in the field, in th e G erm an, French, and Russian languages, w hich earned him an interna­ tional reputation.

J. The “Voluntary Econom ic

Society” N o b etter evidence o f the prevalence in Russia at th a t tim e of an intense interest in econom ic subjects can be found than the creation, w ith C atherine the G re at’s encouragem ent, o f the “V oluntary E co­ nom ic Society” (V olno Ekonom icheskoie O bshchestvo) in St. Petersburg in 1765. T he proclaim ed purpose of this society— an essentially private body— was “to p ropa­ gate am ong the people useful and necessary know ledge relating to agriculture and house construction.” A. I. H odnev, secretary and official historian o f the Society in the 1860’s, credits C atherine the G reat w ith the idea of its creation, citing th e fact th at its founder, I. I. T aubert, was C atherine’s librarian, and th a t all th e fourteen original m em bers of it w ere m em bers of her court.28 O n the other hand, P rofessor P tu ch a treats C atherine’s establishm ent of the Society as her w ay of bypassing th e A cadem y o f Sciences, w hich could n o t be easily controlled, and also as a m eans of providing a substitute fo r L o­ m onosov’s m ost disturbing study.29 T he organization’s m em bership grew rap ­ idly in the succeeding years and soon in­ cluded some of the m ost progressive m en in society, occupying high positions in the governm ent and keenly interested in so­ cial and econom ic questions. T he “political arithm etic” approach found im m ediate sup­ p o rt am ong them . T he funds o f the Society cam e from m em bership dues and gifts. T he Society soon acquired its own building in w hich to hold its meetings, house its library, and carry on its vast correspondence and other w ork. Its funds were insufficient to m aintain a staff o f researchers the w ay the A cadem y of Sciences did, and so the So­ ciety adopted the expedient of offering m onetary prizes fo r the best studies on set subjects and prom ising to publish the prize m anuscripts in its volumes. T he first list of subjects announced by the Society fo r com petition contained 65 eco­ nom ic topics. One of these, carrying a prize of 100 gold pieces offered by Prince Z. G. Chernyshev, called fo r the construction of

88

THE INCOM E OF NATIONS

a “budget fo r an individual residing in St. Petersburg o r M oscow, having an incom e of 3 to 12,000 rubles a year, and no hom e or estate in the country, w ith all pertinent details as to the num ber of servants, and m anagem ent of the household.” 30 T he w in­ ning paper was published in parts 21 and 22 of the “T rudy” (W orks) of the Society and is an invaluable record fo r historians of the prices of com m odities and the m ode of living o f the upper-class Russian city dweller. It is too bad th a t the list did not include sim ilar assignments to construct budgets fo r the m iddle and low er classes. It m ay be noted th at the estimates of R us­ sia’s national incom e fo r 1790, described below, indicated an average incom e fo r a fam ily of five of only 50 rubles a year. “D uring the 1760’s,” says Pokrovsky, “the Russian landow ners h ad n o t yet decided w hat was better fo r them — to m anage their farm s themselves o r to let the peasants m anage them and to becom e themselves m erely the recipients of rent. T he existence of this uncertainty was reflected in th e fo r­ m ulation of certain projects fo r the em an­ cipation of serfs, w hich found considerable favor among some m em bers of the highest nobility presiding over the affairs of the V oluntary E conom ic Society.” T herefore, am ong the subjects included in the list was also one stated as follows: “W hat is pref­ erable for the public, to have the peasant own the land he tills, or to have him own only movable property, and how far should his rights of ow nership to the other type o f property extend?” T he prize— 100 gold pieces and a gold m edal—was w on by a certain Bearde de L ’A bbe, “doctor of th e­ ology and civil law from the U niversity of A achen,” who developed the thesis th a t “it is preferable that the peasant should own the land,” but that to do so “he m ust first be liberated.” Considerable controversy de­ veloped am ong the directors of th e Society as to w hether a w ork propounding such a disturbing doctrine should be published. T he advocates of its publication finally won, largely on the ground th a t the em ancipa­ tion of the serfs proposed by the author was to take place only gradually and with full com pensation to the ow ners.31 A nother subject on the Society’s list of them es called fo r a com plete description of the operations of a large estate includ­ ing a set of detailed instructions from the ow ner to the m anager covering every m atter

entrusted to his care, such as the m anage­ m ent o f fields, collection of dues in m oney and labor from the peasants, the operation o f the mills and distilleries, the care of the m anor house, and other such m atters. This question brought fo rth two rem arkable p a­ pers, both of w hich were published by the Society in 1770. O ne was by P. N . Rytchkov, an agricultural official in the U ral re­ gion, and the other, by A. T. Bolotov, the outstanding agronom ist and w riter on agri­ cultural subjects o f the tim e. Both of these w ere read w ith interest by contem porary agriculturists. Like the other essays m en­ tioned, they are of great historical im por­ tance.32 A nother device w hich the Society adopted in its early years to encourage econom ic research in the country was to send a ques­ tionnaire to the provinces eliciting inform a­ tion from local officials and private citizens on local econom ic conditions and problem s an d offering publication and prizes fo r the best reports. It developed this m ethod on an even greater scale in later years. T h e questionnaire sent o ut in 1791 is w orth quoting. It consisted of six parts, each containing a n um ber of questions. P art one related to the external character­ istics o f th e territory; p art two, to its n at­ u ral resources and their uses; p art three, to the ethnographic com position of the popu­ lation, its custom s, etc.; p a rt four, to the activities and life of the city population, its m erchants, m anufacturers, artisans, labor­ ers, etc.; p a rt five, to th e m ethods o f agri­ culture, num bers of livestock, agricultural im plem ents used, and the extent o f eco­ nom ic well-being of the proprietors and peasants; and p a rt six, to the natu re and am ount of goods shipped out of the prov­ ince o r brought in. These questionnaires n ot only brought fo rth extrem ely valuable reports fo r publication by the Society, b ut also brought to it new corresponding m em ­ bers.33 T he Society grew in strength from year to year. It continued to operate throughout the nineteenth century and into the first two decades of th e next, closing its doors only 153 years later in 1918, w ith th e ad­ vent of the Soviet regim e. Curiously enough, in 1907, it published the first national in­ com e estim ate fo r Russia to have been prepared since the publication, 110 years earlier, o f the three estim ates described fu rth er below.

6 . TH REE LOST RUSSIAN ESTIM ATES

6. The “Society of Friends of the Literary Sciences” In 1784, a very different type of organi­ zation from th e V oluntary Econom ic So­ ciety was founded by a group of intellec­ tuals w ho did not belong to high society and to the u pper adm inistrative circles. T hey held com paratively radical views and drew inspiration from the rising tide of p o ­ litical revolt in F rance. A m ong them w ere A. N . Radishchev, an adm irer of W estern philosophers and political econom ists; M . I. Antonovski, an able publicist; N . D . Zhulkowski, an econom ic geographer; and other students of politics and economics. They called their organization the “Society of Friends of the L iterary Sciences,” but their interest lay in other directions than th e fine arts. T hey were interested prim arily in the exploration of the econom ic resources, so­ cial institutions, and history of their country, and in the dissem ination o f inform ation along these lines am ong the people at large, all w ith the hope o f encouraging a yearn­ ing fo r a better political, social, and eco­ nom ic order than the one th a t was then in effect. A ccordingly, the leaders of th e Society conceived the plan of preparing fo r publi­ cation an encyclopaedic volum e th a t would show the physical, social, econom ic, po­ litical, and m oral conditions prevailing in the various lands of the w orld, including th eir own, and thus set up better m odels fo r Russia to strive for. T hey called this projected w ork A N e w D escription o f the W orld ( “Novieisheie Zem pleopisanie” ) ; and divided the task of preparing it am ong themselves. A substantial p art of it was to be devoted to a statistical description of Russia.34 In the m eanw hile the Society proceeded w ith the publication of a m onthly journal called The D iscoursing Citizen, w hich touched upon a wide range o f political and econom ic subjects. This publication, however, lasted only one year (1 7 8 9 ). It was charged by the governm ent w ith prop­ agating “dangerous and harm ful ideas” and was ordered to suspend publication. T he large work, “Zem pleopisanie,” was not ready until 1795. It is described m ore fully in section 10 below. Suffice it to note here th a t the approach of this great w ork was strictly that of “political arithm etic.” It was imaginative, analytical, bold, and dynam ic;

89

in a w ord, wholly different from the fam ous w ork of Biisching w hich, in accordance with the dictates o f the G erm an school, was strictly descriptive and “pedestrian.” Professor N . L. Rubinstein, one of the m odern Soviet historians, thus summarized, from a M arxian point o f view and in M arxian term inology, the tren d of Russian econom ic thought of the second half of the eighteenth century: The great forward strides of the Russian economy during the eighteenth century mani­ fested themselves most directly and clearly in the advancement of the social sciences and par­ ticularly of the economic science. The first stage in this theoretical advance during the middle of the century was the transition from mercantilism to physiocracy. It was in this latter doctrine that the program of free trade and of a partial emancipation of industrial enter­ prise from government control, embodied in the economic policies of the reign of Catherine the Great, found its first theoretical justifica­ tion. In its second phase, however, physiocracy was embraced by the reactionary Russian land and serf-owning nobility as a defense and idealization of the “village economy” and of its rule by a feudal master. But towards the end of the century, spurred on by the development of capitalism came still one more theoretical development—the acceptance of the capitalistic theory of classical economics of Adam Smith. These capitalistic bourgeois elements found ex­ pression in the multi-volume Historical De­ scription o f Russian Trade by M. Tchulkov, in the speech of the undeservedly forgotten Russian economist N. A. Tretiakov, and in the economic works of A. N. Radishchev, who based his fundamental views on the concepts of economic liberty and national wealth.35 Professor Liashchenko, em inent Soviet econom ic historian, m akes the point that the popularity of th e F ren ch physiocratic doctrine am ong the conservative high no­ bility of Russia in the last quarter o f the eighteenth century was based on a com plete m isunderstanding on their p art of its true essence. H e w rites: The Russian serf-owning aristocrats who had been abroad and had read Quesnay, Turgot, and other physiocrats could not understand the progressive, anti-feudal, and bourgeois sub­ stance of their theories. They were deceived and captivated by the glorification of agricul­ ture by the physiocrats and by the feudal ap­ pearance of their system. They declared them­ selves to be exponents of physiocracy, inter­ preting the latter as a defense of the bulwarks of feudalism. . . . Even Catherine the Great in her “Nakaz” expressed herself as favoring physiocracy. . . . But neither Catherine, nor

90

THE INCOM E OF NATIONS

the ruling class of serf-owners raised the ques­ tion as to whether the productive powers of agriculture could really be developed under the system of serfdom. . . . When the French physiocrat Mercier de la Riviere, after visiting Russia, wrote that the real reason for Russia’s disorder was “despotism, serfdom, and igno­ rance,” Catherine the Great replied that he was “talking nonsense” and forbade his re­ entry into Russia.36

The liberal elements among the high nobility, the spokesmen for the merchant classes, and the new intelligentsia of Russia, who found their prophet in Adam Smith, on the other hand, were all in favor of the liberation of the serfs, the development of manufactures and free enterprise, the estab­ lishment of a constitutional government; and some of them, were even in favor of free trade. 7. The Birth of the Three N ational

Income Estim ates The leading scholars of this period of tre­ mendous intellectual fermentation who took part in the economic and statistical analyses and discussions were fully acquainted with the concept of national income as it was developed in England in the writings of Petty, Davenant, Arthur Young, and Adam Smith, as well as in those of the French political economists and physiocrats. The writings of some of them (Radishchev and the members of the Society of Friends of the Literary Sciences, in particular), show that they were acquainted not only with that concept but even with the actual esti­ mates of national income that had been prepared in England and France on the basis of these concepts. It is not at all sur­ prising, therefore, that some of them should have attempted to prepare a national in­ come estimate for Russia similar to those which had been prepared in those two coun­ tries; and that three such estimates should have been made in Russia in the 1790’s.37

8. Hermann’s E stim ate a. His Life and Writings B. F. J. Hermann (1755-1815) was bom and educated in Austria. He was an able mineralogist and the author of a number of works in that field. But he was also a student of population phenomena and of social and economic affairs generally, a stat­

istician, a gifted writer, and a competent administrator. Educated at the University of Graz, Hermann was employed in the mining industry of Austria immediately after graduation and soon made an extensive voyage through Hungary, Italy, and Ger­ many with the double purpose of becoming acquainted with the technology of their mining industries and also with their social and economic conditions generally. As a result of this trip he wrote a book of travel, filled with broad observations on many sub­ jects; and another book, published in Leip­ zig and St. Petersburg in 1782, containing extensive statistical and other information on nearly every aspect of the life of Aus­ tria.38 A mention of this latter book is im­ portant, because it served as a model for a like statistical work about Russia that he prepared eight years later. Following a lectureship position at the University of Vienna and the publication of his first books, Hermann came to Russia in 1782 at the express invitation of the St. Petersburg Academy of Sciences. Almost immediately after his arrival he was sent to the Ural Mountains to direct the construc­ tion of mines and smelting works. He found time from his engineering and administra­ tive duties to collect extensive demographic, economic, and administrative information not only about this section of the country but about many others which he visited during this period. He made numerous con­ tacts in the provinces with administrative officials, who made their records available to him. In 1784, Hermann returned to St. Petersburg, and soon thereafter was ap­ pointed to a professorship of mineralogy at the Academy of Sciences. During the next few years, with the data collected in St. Petersburg and in the prov­ inces, Hermann prepared several economic and statistical papers, mostly in the German language, for publication and other uses by the Academy. In 1787 he published in Ger­ many a book of statistical information about Russia and her neighboring lands which had wide European circulation; and he followed it in 1790 with the publication, likewise in German, of a comprehensive handbook of statistical and other information about the Russian Empire as a whole.39 It was a book of some 500 pages and represented the most practical statistical compendium ever pre­ pared about the country. -It was especially useful to the foreigners visiting the country or doing business with it. Hermann made it

6 . TH REE LOST RUSSIAN ESTIM ATES

clear that much of his data came from orig­ inal documents customarily considered con­ fidential. To avoid identification of the offi­ cials who made these documents available to him and possible embarrassment to them, he presented some of these data in round figures. In 1794, Hermann made another trip to Siberia, returning to St. Petersburg Academy in 1796 to resume his professorship there and to become, successively, a member of the Board of Mines, inspector of the Mining Institute, and finally, in 1801, chief of the Ekaterinburg Mining Administration. In 1811 he made his last trip to Siberia and in 1815 he died in St. Petersburg, leaving behind a large list of statistical and engi­ neering publications.

b. His Statistical Handbook Hermann’s statistical handbook consisted of seven parts: (1) territory, administration, population, etc.; (2) physical characteristics; (3) natural resources; (4) agriculture; (5) mining; (6) manufacturing; (7) commerce. In the first part Hermann dealt primarily with the growth of Russia’s population and its composition, correcting what he deemed to have been errors in the calculations made by other authors, such as Voltaire and Plestcheiev. Using the results of the previous censuses, he arrived at the following total population figures: 14 millions in 1722, 16 millions in 1742, and 20 millions in 1762, including 1.7 millions for the population of the Ukraine, Finland, and Estonia. Using the statistics of the 1782 census, he arrived at a total for that year of 28 million, of which 26.4 were subject to the poll tax and 1.7 were not; and he projected these figures to 1788 (the year to which most of his in­ formation related), arriving at what he deemed to be a conservative figure of 30 millions for that year. It was not a bad estimate, as can be seen from the results of the 1796 census (see section lb, above). He estimated that during Catherine the Great’s regime the population had increased so far by 10 millions, of which 7 were by natural reproduction and 3 by annexation— also a conservative figure. He verified his population estimates with the aid of the statistics of births and deaths in certain provinces. Hermann advanced certain broad but unfortunately erroneous generalizations re­ garding the rate of growth of Russia’s popu­ lation. Thus, he contended that during the

91

last decades Russia’s births were about twice as numerous as her deaths, resulting in an unprecedented increase in her population at the rate of approximately 2 per cent, or at a doubling of her population every thirty to forty years. If this rate should be main­ tained in the future, he postulated, Russia’s population by the end of the eighteenth century would reach 40 millions, by 1820— 60 millions, and by 1882—230 millions. Both his basic conclusion regarding the current rate of population growth and his projections on that basis involved gross exaggerations. Actually, the annual rate of population increase for the country as a whole, according to Kraft’s calculations, averaged during the period only 0.78 per cent.40 But while Hermann’s projections of popu­ lation growth were fantastic, his emphasis on the great importance which economic factors played in population growth was correct. Hermann defined the economic wel­ fare of a people very concretely in terms of their levels of consumption as deter­ mined by their resources, place of habita­ tion, and period of historical development. He recommended that every provincial ad­ ministration be required to send an annual report to the Senate in tabular form showing “the relative annual production of various goods in the area under its jurisdiction, the quantity of bread on hand, the numbers of livestock, etc., and the relation [of the out­ put] to the people’s needs.” Such tables when obtained from all the provinces and put together by the Senate, he said, “would show to the ministers the current situation of the Empire as accurately and clearly as a mirror.” 41 In this statement, Hermann advanced the same idea Charles Davenant had expressed in England nearly a hundred years before and that Lavoisier would ex­ press some three years later in France. This conception of measuring economic welfare, whether original with him or not, is indica­ tive of the high quality of Hermann’s eco­ nomic thinking or practical judgment. In a supplemental article published in 1791 in Zimmermann’s Annalen der Geo­ graphic und Statistik (vol. 9, pp. 249-65), Hermann presented a picture of the income and expenditures of the Russian govern­ ment which was of considerable interest, inasmuch as these matters were generally kept secret. He set the tax revenue at 40 million rubles, with 15.6 millions of it com­ ing from the poll tax, 1.5 millions from the

92

THE INCOM E OF NATIONS

business licenses, 10 millions from the tax on alcohol, 6 from custom’s duties, 2 from the salt duty, 3.5 from mining and coinage profits, and 1.4 from miscellaneous sources. The tax revenue thus amounted to 1.3 rubles a year per head. The expenditure was set in his table at 35 millions, thus resulting in an annual surplus to the treasury in peace time of 5 million rubles. Of this expendi­ ture, 14 millions, or 40 per cent, went for national defense; 6 millions were for the maintenance of the civil offices and military commands in the provinces; 4 millions were for the support of the imperial court; 2 mil­ lions were for the central civil administra­ tion in St. Petersburg and for the embassies abroad; 2 for the administration of the pub­ lic domain; 1.5 for the costs of operation of state-owned mining and other enterprises; 1.5 millions for transportation and commu­ nications; 1 million for the administration of the church; 1 million for education; 1 million for police, hospitals, and welfare institutions; and 1 million for the interest on the government debt. A. W. Hupei in his book published a year later (see next section) gave a different schedule of tax revenue, aggregating 46 million rubles (Ptucha, op. cit., pp. 269-70). c. Hermann’s National Income Estimate Hermann built his national income esti­ mate as Petty had, largely from the con­ sumption point of view. Assuming the popu­ lation to be 30 millions, as stated above, he broke it down into 6 million families, with five persons in each. He estimated next that these 6 million families between them con­ sumed 576 million pud (a pud containing 40 pounds) of bread a year, or about 96 pud, or 3,840 pounds, per family per year —a calculation which meant an average per capita consumption of a little over 2 pounds of bread a day. Under “bread” the Russian writers of the time included, in addition to rye bread, which was the principal food of the population, a wide variety of cereals consumed in various forms (for comparable estimates of average bread consumption in France during the same period— 1.3 to 1.6 pounds per person—see Chapter 5, sections 4e and 5 above; the lower bread consump­ tion in France, however, was offset by a higher per capita consumption of garden vegetables, fruits, and meats). Valuing the bread at an average price of 25 copecks per pud—% o 0f one copeck per pound—Her­ mann obtained a total value for the national

bread consumption of 144 million rubles a year. To this bread consumption, Hermann next added the consumption of other foods, such as alcohol, salt, animal products, vege­ tables, etc., as well as other products, such as furs, cordage and other hemp products, linen, tobacco, iron, copper, and other met­ als—using appropriate average prices in each case. He used only material products, ignoring the services of domestics, profes­ sional persons, the clergy, and government employees, thus departing from Petty’s, King’s, and Davenant’s estimates. His con­ cept of national income thus closely fol­ lowed Adam Smith’s. There can be no doubt that Hermann, just as any other cultured man of the time interested in economics, was well acquainted with Smith’s great book (German translations of which had been available since 1776 and French, Italian, Spanish, and Danish since 1779, in every continental country, including Russia) and that he was deeply influenced by this work. Hermann presented his estimate in the fol­ lowing tabulation: 42 Million Rubles Bread [incl. cereals] Alcohol, 5 million buckets a t 3 rubles each Salt, 12 million pud a t 35 copecks Gold, silver, copper, iron, and other metals Furs Hemp, flax, tobacco, linen, vegetable oils, etc. Fuel wood, construction materials, lum­ ber, tar, grease, etc. Livestock, leather, wool, milk, and gar­ den vegetables Fish products Total

144.0 15.0 4.2 8.75 5.0 30.0 20.0 58.1 15.0 300.0

The average per capita income, according to Hermann’s estimate, thus amounted to 10 rubles a year. His estimate, he said, was minimal. For example, his calculation of the value of bread consumption was based on a price of 25 copecks per pud, which, he said, was a price at which bread could no longer be bought in many sections of the country. Nor did his calculations, he said, include the bread that was being exported. Finally, he said, one should add to the fore­ going sum of 300 million rubles the total money in circulation, which he estimated at 230 million rubles (76 in gold and silver,

6 . THREE LOST RUSSIAN ESTIM ATES

54 in copper, and 100 in paper notes), and which, he said, was increasing annually at the rate of 5 million rubles. But he did not explain why such an addition was necessary. In this fashion, “the total annual wealth of the country,” he concluded, amounted to 530 million rubles and “not any less.” 43 Actually, this aggregate was completely meaningless, inasmuch as the money in cir­ culation included in it was not a part of the national wealth. For a further discussion of this error, see Section 10c below. Hermann’s contemporary, A. W. Hupei (1736-1819), who in 1791 published the first volume of a work covering some of the same ground but handled very differently,44 severely criticized Hermann’s attempt at estimating the national income. Hupei, born in Germany and educated at the universities of Weimar and Jena, was an important mem­ ber of the Lutheran clergy of Liefland, as well as a publicist and prodigious writer, well-known both in Russia and abroad. He was keenly interested in the practical prob­ lems of government and was an ardent fol­ lower of Achenwall. In the aforementioned review of Hermann’s work, Hupei wrote that “Hermann’s conclusions regarding the annual wealth of the country appear to have been written in haste in a hit or miss fash­ ion, without sufficient prior consideration of the theoretical problems involved.” The figures, he says, were not founded on facts. The quantity of bread consumption was ex­ aggerated. The fact that babies and some segments of the population—such as the Arctic “Samoyeds”— did not consume bread, he thought, was not taken into ac­ count. On the other hand, the fact was not considered that grain was used to make beer and to feed the animals and that this was also an element of the national income. The evaluation of the products in prices in­ volves difficulties, he said, because of price fluctuations, and these were not accounted for by Hermann. Hupei questioned Her­ mann’s estimates of alcohol consumption and the accuracy of the prices at which he evaluated it. He also charged Hermann with omitting the values of a variety of animal products, as well as those of forestry, and quarry products. Anyone who ventures on an evaluation of national wealth, he said, must first develop a clear concept of what national wealth is. Some of Hupei’s criti­ cisms were just, but others revealed a state of confusion on the subject in his own mind. Hupei denied the possibility of calculat­

93

ing the national wealth and national income of a country, saying: “In some countries statisticians have attempted to calculate the national income, or at least, the incomes from all products derived per inhabitant. Such exercises are good only to kill time; they seldom satisfy an intelligent student.” 46 Hiipel’s negative attitude toward national income estimates was in line with the spirit of the German statistical school, but it was basically wrong.

9. Radishchev’s Estim ates a. His Life and Writings Up to His Exile to Siberia A. N. Radishchev (1749-1802) came from a noble family of well-to-do land­ owners. Having demonstrated his keen in­ telligence and love for study, he was sent abroad at the age of seventeen to study at the University of Leipzig. He spent five years there immersing himself in jurispru­ dence, philosophy, history, natural sciences, and languages and became thoroughly im­ bued with liberal ideas. In 1777, some time after his return to his home in Russia, he entered government service as an employee in the Board of Trade, serving under Count A. P. Vorontzov, one of the most cultured and liberal statesmen of the time. By 1790, under the latter’s tutelage, he had risen to the by no means unimportant position of director of the Custom House of St. Peters­ burg. During those thirteen years of service in an economic agency, Radishchev learned much about foreign trade and also about statistics.46 As a statistician, Radishchev leaned more toward the English analytical and specu­ lative school than towards the German de­ scriptive school. Although he appreciated the importance of precise facts and figures, he was interested in these only insofar as they helped to explain the social and eco­ nomic phenomena of his time. As M. V. Ptucha puts it: “Radishchev used figures to solve fundamental social problems, but when these were unavailable, with the courage possessed only by great minds, he resorted to political arithmetic, using approxima­ tions.” 47 In his economic thinking he was influenced by Adam Smith, as he favored free enterprise and free labor (see Section 6 above). During the 1780’s Radishchev took part in the organization of the “Society of

94

THE INCOM E OF NATIONS

Friends of the Literary Sciences” described in Section 6 above, in the launching of its monthly journal, The D iscoursing Citizen, and in the planning of its other future proj­ ects. He wrote a brilliant leading article for the journal’s first issue, entitled “What Is a Citizen,” in which he gave expression to the liberal and democratic ideas that were coming out of France, America, and other Western countries at that time. He also started two economic and statistical surveys in the spirit of the political arithmetic. One dealt with taxes in the St. Petersburg prov­ ince. In it he attempted to summate all the national and local taxes levied in that prov­ ince, within the city and in the country proper, to calculate their burdens for the various classes of the population on a per capita basis, and to appraise their social and economic effects. The second study at­ tempted to present the topographic, demo­ graphic, administrative, and other features of the province. Neither study was com­ pleted by him as he had also begun in the meanwhile to write another book which he deemed to be more important, The Voyage fro m St. Petersbourg to M oscow . Following its publication he was arrested and was sent into exile in Siberia where he no longer could obtain the necessary materials for the completion of his two earlier statistical works. The Voyage was both a work of social protest and a proclamation of faith in the rights of man, inspired by the ideas of Rous­ seau, Voltaire, the American Revolution, and the revolutionary movement in France. In this book, speaking through the voice of an imaginary voyager holding conversations with various people and observing various incidents in the course of his travel, Radi­ shchev, in simple and moving language, sought to depict the poverty and oppression of the peasants and to arouse the reader’s sympathy for them. He pleaded with the monarch to liberate the serfs and to correct existing injustices in their treatment, warn­ ing the monarch that if this were not done, the monarchy might be swept away by just such an uprising of an aroused and wrath­ ful populace as was taking place at that very moment in France. Taking advantage of a decree issued by Catherine the Great a few years before per­ mitting private individuals to own presses and to print books, Radishchev installed a press in his own home and printed his book in an issue of 100 copies. He had originally

submitted a milder version of his manu­ script to the censor and had secured per­ mission to publish it, but subsequently while the work was on the presses he added sev­ eral passages that were much stronger in tone. Soon after the book appeared in the shops, some copies fell into the hands of members of the highest government circles and cre­ ated a considerable stir among them. Cather­ ine the Great herself had a look at it and became incensed: here was a clear call to rebellion. By that time, frightened by the developments in France, she had turned reactionary and decided on strong measures against all “dangerous thoughts.” An order was issued to seize all unsold copies of the book (only 25 had apparently been sold), and to arrest the author. Radishchev was charged with inciting people to rebellion and engaging in high treason, and he was condemned to death. This sentence, how­ ever, seemed much too severe to many in­ fluential people, and was, therefore, com­ muted to a ten-year exile to Siberia and the deprivation of the culprit of all his civil rights. Of the confiscated books a few were pre­ served in the archives for only trusted eyes to see; and the others were destroyed. For the next 117 years its publication and cir­ culation in Russia remained prohibited. Two editions appeared abroad—one in London in 1858 and one in Leipzig in 1876. Radi­ shchev’s work was known by name as a revolutionary piece of writing to all cultured people in Russia of the nineteenth century, although its exact contents were known to only very few.

b. His Writing in Exile On his way to Ilimsk, his assigned place of exile in Siberia, Radishchev was allowed to stop for a while at Tobolsk. There, through the intervention of the governor, who was his former friend, he was allowed considerable freedom of movement. He took advantage of this opportunity to as­ semble data for two works—a statistical study of the province of Tobolsk, drawn along the lines of his uncompleted manu­ script on the province of St. Petersburg; and a treatise on Russia’s trade with China. The first work was completed by him on the spot; the second—in 1794, in Ilimsk. Radishchev had been interested in the subject of trade with China since the days of his service in the board of trade and in the custom house in St. Petersburg, but here

6 . TH REE LOST RUSSIAN ESTIM ATES

in Tobolsk, through which most of the trade from China flowed to European Russia, he was able to obtain first-hand information. In embarking upon this study, Radishchev had occasion further to mull over Adam Smith’s ideas on the sources of national wealth and income. On May 2, 1791 he wrote a letter from Tobolsk to his friend and patron in St. Petersburg, A. R. Vorontzov, asking him to send him Condorcet’s Commentary on Adam Smith’s Wealth of Nations which had then appeared in the Bibliotheque de VEconomie Politique in France. In his study of the trade with China (which he put into the form of a commu­ nication to Count Vorontzov) he attempted to show that this trade was of comparatively little importance to the economy of Russia. In making this point, he made his calcula­ tion of Russia’s national income. In October 1791, while still stationed at Tobolsk, Radishchev received from Voron­ tzov a copy of the just-published book by Hermann, presenting a statistical picture of the Russian Empire. In a letter to Vorontzov dated March 24, 1792 and dispatched from Ilimsk, Radishchev compared Hermann’s work to the famous geographer T. G. Georgi’s description of St. Petersburg. Not­ ing that Hermann’s work was written more in the spirit of the British analytical school that was dearest to his own heart, whereas Georgi’s was written in the vein of the Ger­ man descriptive school with which he had much less in common, Radishchev nonethe­ less expressed himself in favor of Georgi’s work. “Although both books are mere com­ pilations,” he wrote, “Georgi’s fulfills its purpose better. While abstaining from judg­ ments, he is at least consistent in his atti­ tude. He talks about everything that he sees or hears; he is all ears and eyes; he does not engage in half-statements. On the other hand, Hermann engages in a task that is beyond him; he draws conclusions but does so unsuccessfully; he would like to reason, but he does not dare. One feels that he de­ ceives the reader and uses flattery. His con­ clusions are often false, particularly when they deal with current affairs. . . . For ex­ ample, in speaking of Her Majesty’s decree allowing individuals to own presses and print books, he says that the Empress’s desire in doing so was to foster a love for the sci­ ences among the people; and he stops there as if he really had said something extraor­ dinary. His silence over some related facts is an insult to the intelligence.” 48

95

Radishchev also criticized Hermann for his fantastic forecast that by 1882 Russia would have more population than China, saying that the author forgot that “political institutions like things of nature, age through usage” and that “if a human being in his youth lives wastefully, he invites premature old age and death” and that “the same holds true of political bodies.” No better example of the depth of Radishchev’s po­ litical, social, and economic thought is needed than what is contained in these few passages of his letter to Vorontzov.

c. His Work After His Return from Exile In 1796, after Catherine’s death and Paul’s succession, Vorontzov was able to obtain a further commutation of sentence for his friend, permitting him to return from Siberia and to settle on his estate under local police surveillance. During the next four years, 1797-1800, while thus settled in the country, Radishchev wrote another remarkable work, based on what informa­ tion he could obtain on the spot. He called it “A Description of My Estate, Domain, Village, or Whatever You May Wish to Call It.” Actually, it was a description not of his particular estate, but of what he con­ sidered to be a representative estate, using pertinent materials from his own village. This representative estate or village was described by Radishchev as containing 360 peasants, 1400 “desiatin” of land of which 700 were under cultivation under the threefield system, and 150 homes. Every aspect of the economic life of that community was surveyed and a balance sheet of its income and outlay was constructed, showing how much the people produced and consumed annually, how much actual money they earned and had to pay out in taxes and other dues, and how much they saved and added to their capital, or “dissaved” by the end of the year. Radishchev accounted for all capital outlays of this village, in the form of additions to livestock, fertilizer, seed, etc. He computed the total capital of the village at 8,200 rubles of which nearly half were represented by livestock; and he evaluated the labor of the peasants in summer cul­ tivation at 3,500 rubles (20 rubles per man, 10 per woman, and 5 per teen-ager). He computed the money earnings of 50 persons leaving the village for winter work in the town or city annually; and he wound up with a net income for the village of 1,500 rubles, or 4 rubles per peasant for the year,

96

THE INCOM E OF NATIONS

to be added to his capital or to be spent by him and his family for holiday celebrations. In this “case study,” Radishchev sought to disclose in concrete terms the main evils of the existing system of serfdom and of the feudal economy generally, identifying the factors that made the peasants poor and ill, while making the masters relatively well-off and healthy. He showed how the State weak­ ened the village economy by its compulsory recruitment of peasant youth; how in order to earn money with which to pay taxes and other money dues the peasants had to en­ gage in subsidiary occupations; and also how they had to marry off their young boys to older girls so that the latter could help in field work and home care; and how this, in turn, resulted in the births of feeble in­ fants and in a high rate of mortality among the peasant population. In this particular statistical work Radishchev applied the methods of political arithmetic most effec­ tively.49 With the accession, in 1801, of Alexander I, who showed some liberal tendencies at the beginning of his reign, Vorontzov was able to obtain a complete pardon for Radi­ shchev and permission for him to re-enter government service. Radishchev was ap­ pointed to the Special Commission, created through the efforts of the gifted and liberal minded statesman M. M. Speransky, to de­ vise a new civil, criminal, and administra­ tive code for Russia. Catherine the Great, during the first years of her reign, had dreamt of doing the same through the issuance of her famous “Nakaz,” (large portions of which she had lifted bodily from Montes­ quieu’s works), and through calling to­ gether a mammoth commission of elected delegates; but she failed completely in this effort. The head of Alexander’s new com­ mission was the exceedingly able and pro­ gressive Swiss and English educated admirer of English statistics and economics, young Prince V. P. Kotchubey (1768-1834). Radishchev expected much good to come out of this commission and threw himself headlong into its work. He prepared a com­ prehensive memorandum outlining a broadgauged plan of collection of statistical and other data from the provincial offices. The data were to cover almost every aspect of the life of the nation having a bearing on the civil, criminal, and administrative law.50 The breadth of his project, however, fright­ ened some of his immediate superiors, who curtly reminded him that if he persisted in

the entertainment of such wild projects he would soon find himself returned to Siberia. Utterly dejected, frustrated in his efforts, and broken in both spirit and body, Radi­ shchev returned home and ended his life by taking poison. Thus ended tragically the life of one of the most gifted social scien­ tists and statisticians of Russia and one of the most noble and courageous spirits of the age. Of all Radishchev’s works, only one was published during his life and that one was immediately suppressed. His other works never got beyond the manuscript stage. His name remained but a legend, and his works remained hidden from the public for more than a hundred years after his death. Finally, after the Revolution of 1905, when the government was forced momentarily to relax its oppressive controls over the thoughts and actions of the people, all of Radishchev’s principal works were pub­ lished. Still later, under the Soviet regime, they were republished many times, and Radishchev himself became honored as the country’s first revolutionary and friend of the working classes.

d. Radishchev’s National Income Estimate Radishchev must have had at least two good reasons for including an estimate of Russia’s national income in his study, The Trade with China: (a) he wished to show quantitatively the relative unimportance of the Chinese trade to Russia, and there could be no better way to do it than to express that trade as a percentage of the value of the country’s total annual production; and (b) he had read Hermann’s book in which such an estimate of the value of national production was presented and he had found this estimate to require some correction. To these two reasons may be added two others: (c) Radishchev was acquainted with the national income concepts and estimates de­ veloped in both England and France, and he must naturally have been inclined to favor the introduction of this same concept and these same types of calculations into Russian economic and statistical analyses; and (d) in his earlier fiscal and economic surveys of the province of St. Petersburg he had dealt to some extent with the annual income and wealth of the people and it was, therefore, natural that he should now try to apply this same type of analysis to the national scene.

6 . TH REE LOST RUSSIAN ESTIM ATES

Radishchev did not use the term “national income” (narodny dokhod). He used in­ stead the term “anually produced wealth of the country” which he apparently derived from Adam Smith. His estimating methods were somewhat different from Hermann’s. His population estimate was lower—only 26 millions as against Hermann’s 30 millions— and, in fact, was a gross underestimate (as may be seen from the table in Section lb above); but his estimate of the annual per capita bread consumption was about 60 per cent higher than Hermann’s—namely, 3.3 pounds as against the other’s 2.1. Thus, Radishchev estimated the consumption of rye bread at 15 puds (600 pounds) per person per year and the consumption of all other breadstuffs and cereals at a like amount. In this way, he obtained a total bread consumption for the nation of 390 million pud. He valued it at substantially higher prices per pud than Hermann did, and, moreover, at higher prices for cereals than for rye. His average price was 38 copecks per pud, or almost 1 copeck per pound. Thus, he obtained a total value for the bread consumption for the country of 150 million rubles as against Hermann’s figure of 144 million rubles—a result not very much different in its total but different in its composition and in its relation to the total estimated population. To this 150 million rubles, Radishchev added an equal sum for the total value of all other products, thus arriving by a some­ what different process and for a much smaller estimated population at the same total national income of 300 million rubles as Hermann arrived at. His per capita in­ come per year— 11.6 rubles—was higher than Hermann’s 10 ruble figure.51 Radishchev’s estimate was not as detailed as Hermann’s. This was understandable, inasmuch as he did not have any statistical documents at Ilimsk on which to draw for material. But in three respects his estimate was methodologically sounder than Her­ mann’s: (a) he did not repeat Hermann’s error of adding the amount of money in circulation to the value of annual material production; (b) he used the amount of money in circulation, however, as a verifica­ tion of the accuracy of his national income estimate; and (c) he used the figure of gov­ ernment revenue for the same purpose. Thus, taking 200 million rubles as the approximate amount of money in circula­ tion, and assuming that each ruble on the

97

average turns over three times a year, i.e., “represents three different things,” he ar­ rived at a total turnover of trade of 600 million rubles a year. In proportion to this sum of trade turnover, the sum of 300 mil­ lion rubles for the value of the annual pro­ duction seemed to him just about correct. Next, taking the total annual revenue of the government at 40 million rubles, he declared that this revenue represented probably one fifth of the total money incomes of the peo­ ple, which could thus be placed at 200 mil­ lion rubles, or at about the sum of money in circulation. This figure, too, seemed to check out with the estimate of a 300 million ruble annual production. Radishchev’s crude methods of verifying his national income estimates are clearly reminiscent of the methods used by Petty and Davenant a hun­ dred or more years before; and were prob­ ably taken by Radishchev from their works. Radishchev sent his manuscript to Voron­ tzov, but whether the latter showed it to others, is not known.

10. The E stim ate Presented in the “N ew W orld G eography” a. The Identity of the Authors and the Nature of the Work The third estimate of the national income of Russia appeared in 1795 in a work pub­ lished anonymously in St. Petersburg under the impressive title: “A N ew Description o f the W orld and o f A ll o f Its F our Parts In ­ cluding a Statistical Description o f the R u s­ sian E m pire in a M anner N ever Previously A ttem pted : Composed from most authentic

sources and newest and best works by a group of Russian scientists, made available through the efforts of I. Glazunov, book dealer attached to the Imperial Academy of Sciences, St. Petersburg, 1795” (reading in Russian as the Noveisheie Zem pleopisanie, etc.). It will be noticed that the title em­ phasized that the work was written by R u s­ sian scientists. This was done to distinguish it from other earlier studies prepared mostly by foreigners, especially the Germans, and bearing a distinctly foreign tenor and point of view. This book was thus designed to express the Russian point of view. The Preface to this work identified its authors as members of the “Society of Friends of the Literary Sciences” and the publishers of the discontinued journal The

98

THE INCOM E OF NATIONS

Discoursing Citizen.52 Later historians spe­ cifically mentioned as among the probable authors M. I. Antonovski, the editor of the defunct journal, and N. D. Zhulkowski, sec­ retary of the Admiralty Board; and also ex­ pressed the belief that Radishchev had col­ laborated with them in the late 1780’s in planning the work and may even have con­ tributed some chapters to it before his ar­ rest. However, there is no direct evidence to that effect. One thing is certain in light of the tremendous scope of the undertaking: the work was the result of several years of assiduous endeavor on the part of a group of scholars. Having been taught by the example of Radishchev’s book to fear similar suppres­ sion, the authors took special pains to give their work an aura of devotion to the throne. They adorned the title page with an emblem showing Catherine the Great sitting in the center of a resplendent sun captioned: “Her Light Will Shine Through the Ages.” More­ over, they dedicated the work to Catherine’s grandson, the Grand Duke Alexander, whom she preferred as a successor to her mentally unbalanced son Paul. Wisely, too, the authors preferred to remain anonymous in case a legal prosecution would be insti­ tuted against their work. The book contained about 1,000 pages and consisted of five parts. The first, which was the smallest, was an elementary text in geography for young people; part two, the largest, was a statistical description of European Russia—her physical features, resources, history, population, government, economic activities, arts, sciences, and other such matters; part three, which was also large, gave similar information, but in ab­ breviated form, for each of the principal Western European countries; part four dealt with the Asiatic, African, and North and South American countries; while part five dealt with the ancient world. The work was of grandiose proportions and gave full ex­ pression to the authors’ broad conceptions of geography as “a science that shows the natural and civil arrangement of the land surface.” In the chapter on Russia’s population, the authors noted that its rate of growth had improved considerably during the preceding several decades and that if this improve­ ment continued, “Russia will have a hun­ dred years hence, in 1892, two hundred and thirty million people.” In their treatment

of the political, social, and economic sys­ tems of the country, the authors gave vent to their liberal ideas, condemning despotism, the misuse of power by officials, and the system of serfdom. In the chapter on France, they gave a full exposition of the ideas of the French Revolution, including those of the “natural and inalienable rights of man,” liberty, and equality; the thought that “in a society no power can exist that is not derived from the people and is not founded on their welfare”; and the principle of the separation of legislative, executive, and judicial powers. When the work appeared, members of the inner government circles, including Catherine the Great herself, were astounded to find such daring passages in it. An order was issued to seize all printed copies of the book. In Moscow alone 359 copies were confiscated by the police, only 36 having been sold by the book dealers during the few days it was on sale. The censor who authorized its publication was questioned, but he pleaded oversight, suggesting that the “criminal” parts might have been in­ serted after he had read it. By a subsequent order of Catherine’s all the seized copies of the work were delivered to the Academy of Sciences with a request that it designate a committee to review the work and to strike out from it “all inaccuracies and forbidden statements . . . inimical to the preserva­ tion of the monarch and of the existing social order . . . and especially the entire section on France.” The committee was ap­ pointed by the Academy, but after consid­ ering the matter it refused to act as censor and to emasculate the work. Thus, the New World Geography remained unchanged, but only very few copies of it remained in pri­ vate circulation. These were eagerly read during the succeeding half-century or more by progressive Russian scientists and by all revolutionary-minded people who could lay their hands on it. It was popular among the “Decembrists” of the first quarter of the nineteenth century and still later among the “Petrashevtsy” of the 1840’s.

b. Reasons for Including a National Income Estimate The reasons which led the authors of this New World Geography to include an estimate of the national income of Russia can be readily surmised. First of all, the writers were “political arithmeticians,” ac­

6.

THREE LOST RUSSIAN ESTIM ATES

quainted with the national income estimates prepared in England by Petty and others. Secondly, some of them might also have been acquainted with the estimates prepared in France shortly before by Du Pont, La­ voisier, and others (see Chapter 5, Sections 3-5, above), for they must have made every attempt to get all important pieces of printed matter coming out of that country during this momentous period in its life. Thirdly, they must have wanted to refute some of the false notions that were being circulated abroad regarding the extreme poverty of the Russian people, and thought that national income estimates were a good means to that end. Fourthly, they were acquainted with the national income estimates prepared by Hermann and published in his book, and would naturally be anxious to give as much information on the subject as he had, if not more.

c. The National Income Estimate of the “New World Geography” Chapter 19 in Part II of the Geography contained in a general discussion of Russia’s commerce, the estimate in question.53 Ex­ cept for the assumption of a larger popula­ tion and, accordingly, of a correspondingly larger production, the estimate in the Geog­ raphy followed Hermann’s estimate very closely. The authors sought to measure not only “the value of the country’s annual produc­ tion of new goods,” or national income in the sense of the national product, but also a larger aggregate composed of the sum of this national income and of the amount of money in circulation, which they called somewhat euphemistically “the national wealth circulating annually and receiving a substantial accretion annually.” In thus treating money as a part of national wealth, these authors merely repeated Hermann’s error, which, in fact, was the original error of the mercantilists and was still current at the time. Adam Smith, just as Petty before him, called attention to this confusion, when he wrote that “the great wheel of circulation is altogether different from the goods which are circulated by means of it. The revenue of the society consists altogether in those goods, and not in the wheel which circulates them” ( Wealth of Nations, Book II, chap­ ter 2). But Hermann and the authors of the Geography might not have read the Wealth of Nations with due attention; and

99

it is also possible that they might have con­ sciously blown up the figure through this particular addition in order to produce a more impressive total. The authors evaluated “the total annual production of new goods,” or national in­ come proper, at 579 million rubles, or at nearly twice the sum reached by Hermann and Radishchev. This total was composed of three main aggregates: (a) the value of bread production (including cereals), (b) the value of all other finished products of the country, and (c) the value of exports of raw materials. The largest single item in the computa­ tion was “bread,” conceived broadly to in­ clude all kinds of cereals (barley, oats, buckwheat, etc.). The authors calculated its consumption at 1,152 million pud a year, on the assumption that the population aggregated 60 millions—which was a gross overstatement (see Section lb, above)—and that the per capita bread consumption was 19.2 pud (768 pounds) per year, or 2.1 pounds a day. Although they did not ex­ plain where they got this figure of per capita bread consumption, it is clear that they took it from Hermann, who was the first to advance it (see Section 8c, above). It will be noticed that their figure (1,152 million pud) for total national bread con­ sumption was exactly double the 576 mil­ lion given by Hermann and that this dou­ bling was due to their assumption of twice as high a population. They evaluated this consumption, however, at the higher price of 30 copecks per pud as against Hermann’s 25 copecks. Thus they reached a total of 345.6 million rubles for the value of the national annual bread consumption, which was exactly 120 per cent above Hermann’s figure. Adding to this the value of exported grain and of grain used in distillation, they wound up with a figure of 350 million rubles as representing the value of the country’s annual grain and cereal production (after deducting the value of the seeds used for planting). Next, the authors calculated the values of all other finished goods—whether pro­ duced by agriculture, fishing, mining, or industry—as well as the values of exported raw materials. For the first, they developed a total of 194 million rubles and for the second one of 35 million, thus making a total of 229 million for the two together. Unfortunately, in the only summary of the

100

THE INCOM E OF NATIONS

Geography, available in the United States, —the one published in Ptucha’s book,54— no details of these latter calculations are given. In the absence of these details it is impossible to compare these parts of their computations with those of Hermann’s. It is remembered that Hermann gave a figure of 156 million rubles for the value of all other products than bread. In his estimate, therefore, bread constituted 48 per cent of the total national income (144 million in a 300 million total). In the estimates in the Geography, on the other hand, bread repre­ sented 60 per cent of the national income (350 million in a 579 million total). The difference between these two proportions is considerable. Adding all these three items together— the value of grain production, the value of the output of other finished goods, and the value of exports of raw materials—the au­ thors of the Geography wound up with a total value of all annually newly produced goods, or national income, of 579 million rubles. In adding to the national income total the amount of money in circulation, the authors used a different figure from Her­ mann’s. Whereas Hermann calculated the money in circulation at 230 million rubles, the authors of the Geography calculated it at 360 million rubles (allowing considerably more for both coin and paper circulation). By adding this figure to the 579 million rubles of material goods, they obtained a total of 939 million rubles, which they said, represented “the total national wealth cir­ culating annually.” Fortunately, the authors of this estimate were apparently just as much aware as Hermann was of the fact that the addition of the amount of money in circulation to the national income proper could serve only decorative purposes. For they estimated the per capita income of the population without considering the money in circulation. But they made an arithmetical error in the lat­ ter computation (unless Ptucha quotes their estimate incorrectly), for they arrived at a per capita annual income of 9 rubles and 7 copecks, whereas a division of 579 million income by a 60 million population should have yielded a per capita income of 9 rubles and 65 copecks. At any rate, at either figure the result was not much different from Her­ mann’s per capita of 10 rubles, but consid­ erably below Radishchev’s estimate of 11 rubles and 60 copecks.

11. Appraisal of the Three Estimates The three estimates of national income prepared in Russia towards the end of the eighteenth century and which were the first ever to have been prepared there, contained patent imperfections. Yet, by and large, they were no more deficient than those prepared in England or in France at that time. The notions as to what properly constituted the national income entertained in Western Eu­ rope ranged from the broad concepts of Petty, King, Davenant, and Arthur Young (as the aggregate of all incomes in money or in kind earned by people from every type of economically remunerable pursuit) to the narrow notion of the produit net from agri­ culture propounded by the physiocrats, and the likewise narrow concept of material production advanced by Adam Smith. It should not be surprising, therefore, that the Russian statisticians of the end of the cen­ tury, who must have been aware of some of these conceptual differences, should have found themselves somewhat at sea in their own attempts to give a theoretical and sta­ tistical expression to this important measure of the economic activity and welfare of a people. Nor is it at all surprising that these Rus­ sian scholars, having made their computa­ tions, were unable to draw any significant conclusions from them. There were no na­ tional income estimates of earlier vintage with which theirs could be compared so that any changes that had occurred in the country’s national income could be ascer­ tained and the annual rate of change for the future estimated. Nor were there sufficient data to permit the authors to prepare cor­ responding figures for earlier periods. The figures obtained for Russia could not be readily compared with those calculated for, say, England or France, for they were ex­ pressed in different currencies, and a trans­ lation of them into rubles would not have produced meaningful results. Only one such comparison made in a single currency had ever been made anywhere previously. This was Gregory King’s comparison of the Dutch, French, and English national in­ comes in English pounds, reproduced by Davenant a hundred years earlier and this calculation was not too well known. Had the Russian scholars been allowed to pursue their economic and statistical studies

7 . N E W APPROACHES IN ENGLAND

further, they would undoubtedly have come forward in another ten or twenty years with another set of national income estimates that would have been more deftly executed and that, taken in conjunction with the earlier ones, would have provided a basis for the measurement of the advances made by the economy during the interval, and would also have served many other useful purposes. Instead, all further statistical studies of such a broad and fundamental sort were discouraged by the unenlightened policies of the succeeding monarchs. The school of “political arithmetic” was placed on a list of politically harmful sciences. At the St.

101

Petersburg Academy, at the universities, and in the administrative offices, the non­ committal German school of statistics was enthroned in its place. Practically all traces of these early estimates soon vanished: Her­ mann’s book in the German language sank into obscurity; Radishchev’s estimate never saw the light of day; and the estimates of the New World Geography were soon forgotten as the least exciting parts of a suppressed and forbidden book. No further steps to estimate the national income of the country were taken in Russia during the following one hundred years. (See Chapter 9, Section 6.)

7. NEW APPROACHES IN ENGLAND IN THE NINETEENTH CENTURY During the nineteenth century and particu­ larly its second half, estimates and analyses of national income in Great Britain became more numerous, more reliable, and more meaningful. Estimates were placed on a more consistent conceptual basis; clear dis­ tinctions between gross and net income were drawn; double counting was more carefully avoided; better statistics were employed; techniques of combining data or filling miss­ ing data were refined; the methods for in­ terpreting the estimates became more scien­ tific; and the forms of presentation of the estimates were made more understandable.

1. Factors Responsible for the N ew A pproaches A great many factors were responsible for the progress just described, but most important among them were the momentous changes taking place in the whole political, social, economic, and cultural fabric of British society, and in the role Britain played in political and economic affairs of the world during that time, perhaps, happiest of all the centuries of England’s history. Notes begin on p. 523.

Among the internal changes, particularly outstanding were those produced in the eco­ nomic well-being of the nation by the war effort and wartime inflation, the postwar deflation, and the burdens of the huge war debt; the decline of agriculture and rise in manufacturing and shipping which contin­ ued throughout the century; the pauperiza­ tion of the peasantry and the small farmer incident to the further spread of enclosures; the progressive abandonment by both small and large landowners of cultivation of their lands, first following the repeal of the Corn Laws in the 1840’s and later, on a much larger scale, under the pressure of the mass importation of cheap grain from the Ameri­ can prairies beginning in the 1870’s; the reduction or repeal of most customs duties as part of a move to lower the cost of living and, hence, the cost of labor to the manu­ facturer so that manufactured goods could be exported at lower prices and manufac­ turing would flourish; the establishment of the income tax and the ensuing controversies over rates and exemptions; the growing ur­ banization with its attendant social problems of crime, ill housing, and disease; the growth of large industrial and financial fortunes and

102

TH E INCOM E OF NATIONS

abuses of economic power; the advent of industrial crises and periodic unemployment; the emergence of the labor struggle and the rise of trade unionism; the eruption of the Chartist movement, socialist agitation, and reform movements, resulting in a series of great liberalizing enactments by Parliament such as the Poor Law Amendment, the Mu­ nicipal Reform Act, the Electoral Reform Act, the Public Health Act, the Housing Act, the Free Education Act, the Factory Acts, the Employers Liability Act, the Trade Union Act, and other such ameliatory measures. Equally important stimuli to the making of inquiries in her national income were provided by the momentous changes which took place during that century in England’s external affairs; namely, her acquisition of new colonies and markets, the expansion of her shipping trade, the colossal growth of her foreign investments, the emergence of London as the world’s money market, and England’s confrontation, toward the end of the century, by new and powerful commer­ cial rivals, the United States and Germany. All these and other developments invited the type of synthesized quantitative analysis that only statistical inquiries into the changes in the size, structure, and distribution of the country’s national income could provide. At the same time, the improvements made during that century in the collection and publication of government statistics, in­ cluding the inauguration of decennial cen­ suses, and the advances made in economic theory affecting the concepts of production, consumption, distribution, investment, and saving, greatly facilitated the task of schol­ ars who undertook to prepare national in­ come estimates. The accompanying spread of interest in economic and statistical sur­ veys of every type lent them encouragement in their ventures. Most of the English estimators of that century, especially in the first half, sought to use their national income estimates to prove some important point, whether it be the country’s great war potential and its economic supremacy in the world, or its need for a particular type of economic or fiscal reform. Only a few—and then only during the second half of the century—pre­ pared their estimates without any objective in mind other than their intrinsic informa­ tional value, allowing the estimates to speak for themselves. The national income estimates of England of the nineteenth century fall into two

groups, each associated with a different pe­ riod: first, those of Colquhoun, John and Simon Gray, Lowe, Pebrer, and Spackman of the period 1811-46, all of which were based on production statistics and on the net output method, and most of which fol­ lowed the Smithian material production concept; and second, the estimates of Smee, Baxter, and Giffen of the period 1846—1900, which were based on the comprehensive production concept, and were prepared with the aid of income tax data and occupational statistics by the income-distributed method.

I. Estimates by the Net Output Method, 1811-46 The estimates of the period 1811-46 made a brilliant start with the spectacular work of Patrick Colquhoun, and, very largely, followed the lines laid out by him.

1. C olquhoun’s Estim ates of 1 8 1 2 -1 4 Colquhoun’s estimate, published in a handsome volume of superlative dimen­ sions, befitting a statistical atlas, appeared in 1814, the year England celebrated her vic­ tory over the French “tyrant and usurper.” It bore the impressive title: A Treatise on the W ealth, Power and Resources o f the British E m pire in E very Quarter o f the W orld Including the E ast Indies; with the further sub-title: The R ise and Progress o f the F unding System Explained; with Obser­ vations on the N ational Resources fo r the beneficial E m p lo ym en t o f a redundant P opu­ lation, and fo r rewarding the M ilitary and N a va l Officers, Soldiers, and Seam en, fo r their Services to their Country during the late War. Illustrated by Copious Statistical Tables, constructed on a new plan and ex­ hibiting a collected view o f the different subjects discussed in this work. The first

edition sold out very quickly and was fol­ lowed by a second edition in 1815 pub­ lished “with additions and corrections.” 1 While covering every major aspect of the internal and external economic and fiscal life of the country, Colquhoun concen­ trated on the evaluation and analysis of the “annually produced new property,” its sources, composition, and distribution.

a. Colquhoun’s Approach Patrick Colquhoun (1745-1820) was a mixture of economist and social reformer

7.

N E W APPROACHES IN ENGLAND

and had a flair for publicity. Like Adam Smith, he was a native of Scotland. Start­ ing his career as an official of the City of Glasgow, he sponsored there “many projects aimed at the development of industry.” 2 In 1789 he moved to London, where he became a magistrate and quickly immersed himself in the social problems of that great city. In 1797, he published A Treatise on the Police o f the M etropolis exposing the inadequacies of that organization in dealing with the city’s growing crime. In 1806 he published two more books, one entitled A N ew and A ppropriate System o f E duca­ tion fo r the Labouring People, and the other A Treatise on Indigence, in which he out­

lined measures for alleviating the condi­ tions of the poor. Each one of these works attracted considerable attention. But none won for him the fame that did the Treatise on the Wealth, P ow er and Resources o f the British E m pire. The latter combined hand-

somely-laid-out statistical tables of almost pictorial appearance with beautiful English prose, in which the author extolled the greatness of England, the liberality of her economic policies, and the unlimited poten­ tialities of her development. In the Preface to the second edition of his work, Colquhoun emphasized its “nov­ elty of a design.” It was an attempt, he said, such as had never been made before, “to develop the general system of national economy” for a widely extended empire in a “concise and connected manner.” In a manner reminiscent of Petty’s and King’s prefatory statements to their own works more than a century before, Colquhoun stressed its timeliness. He expressed the hope that his work, dealing as it did with a wide range of subjects, being “no less curious than useful” and being “systemati­ cally arranged and brought within the nar­ rowest compass, supported and elucidated by every existing accessible official fact, might afford some assistance to the national credit at this important crisis” and that it might also tend “to tranquilize the minds of those who may have been disposed to enter­ tain gloomy apprehensions respecting the resources of the Empire under the immense pressures occasioned by a war of unexam­ pled expence and duration.” He alluded to the fact that the war had not yet been completely terminated, as Napoleon at that very moment had just escaped from Elba, making it necessary for “the Nations of Europe to array themselves in arms to op­ pose once more the hydra which had for

103

so many years disturbed the tranquillity of the world.” Colquhoun expressed a hope that “a Work embracing the whole range of political economy may prove useful and interesting, not only to the statesman, but also to the land-holder, the agriculturist, the m erchant, the trader, the m anufacturer and the stock-holder

b. Sources of Colquhoun’s Estimates Colquhoun’s estimates were based largely on the then just published data of the Second Decennial Census of Great Britain and Ireland, held in 1811, as well as on the earlier published data of the First Census conducted in 1800. The appearance of these two censuses was a great event in the life of the British nation. Although a group of scholars, such as Arthur Young and John Sinclair (whom Edmund Burke contemptu­ ously called “economists, sophisters and cal­ culators”),3 had for years urged Parliament to provide for the holding of a census, it was only in 1800 that Parliament finally or­ dered the first census. It was undoubtedly spurred on to such action by the example of the United States, which had made it constitutionally obligatory to hold a census every ten years, and by the example of Napoleon who had ordered the holding of a census for France. The government put the administration of its first two censuses of 1800 and 1811 into the competent hands of John Rickman, one of the ablest statis­ t i c a l of the time. Colquhoun acknowledged his indebted­ ness to these census figures 4 and to Rick­ man personally, but he also used a variety of other sources. For Ireland he was en­ tirely dependent on other sources inasmuch as the part of the second census covering that country had not been completed by the time of his writing. He relied to a large degree on the publications of the Board of Trade, Board of Agriculture, and the Treasury. Unfortunately, he offered very meager documentation for his figures, in most cases leaving the reader in ignorance as to the sources of his information. It is clear from an examination of Col­ quhoun’s computations that he often devel­ oped broad figures from very sketchy data. McCulloch, writing a few years later, criti­ cized Colquhoun’s work severely. Robert Giffen, reviewing it in the 1880’s, attributed much more merit to it, while noting at the same time that “many of his [Colquhoun’s] details are no doubt somewhat fanciful” and

104

THE INCOM E OF NATIONS

that he unfortunately was “not recognising sufficiently the special objects of such inves­ tigations and the necessary unsuitability of the figures for general purposes owing to the inherent difficulty of the data.” 5

c. Colquhoun’s Application of the Smithian Concepts Colquhoun was a follower of Adam Smith. He defined national income in terms of the material production concept as the “new property annually created in the Brit­ ish Empire by the labour of the people em­ ployed in agriculture, manufactures, trade, commerce, navigation, fisheries, and other branches of productive industry.” 6 He did not include in national income the incomes of government employees, professions, bank­ ing, and other service occupations, or serv­ ices of dwellings, or interest on government debt. His concept of national income was, therefore, much narrower than the concepts employed by the English estimators from Petty through Beeke, and his totals are not altogether comparable with theirs. Colquhoun’s estimates were not devel­ oped primarily from fiscal data as the pre­ vious estimates had been, but from popula­ tion, production, and trade statistics. He constructed a separate table for each branch of material production—an innovation in national income estimates. In each case he attempted to distinguish between the gross value of production and its net value by deducting the cost of raw materials. In this respect, he followed the same procedure as Arthur Young and Lavoisier. Like them he made no deduction for the consumption of fixed capital.

d. Examples of Colquhoun’s Estimating Methods The merits as well as the weaknesses of Colquhoun’s estimates can best be shown in his estimates of the net value of agricultural production. The part of his pertinent table that deals with the most important crops is reproduced. It will be noticed that Colqu­ houn did not cite any sources for his fig­ ures of either physical quantities or average prices. He multiplied the one figure by the other in each case, and from the resulting gross value deducted the cost of seeds and fertilizers. He did not deduct the cost of feed furnished working animals, or the cost of depletion of the stock of animals (whether through illness, age, or mortality), or of the wear and tear of other agricultural

capital. He estimated the value of animal products by adding the value of hired labor employed to care for the animals to the value of the feed. In estimating the net value of manufac­ tured goods, Colquhoun used a variety of short cuts. For example, he calculated the net value of cotton goods by taking their reported average export value for the years 1810, 1811, and 1812, which was £ 2 9 mil­ lions, and deducting £ 6 millions for the cost of raw materials, leaving £ 2 3 mil­ lions.7 Apparently, he made no deduction for repairs and maintenance of equipment or for depreciation of capital. Colquhoun placed the total value of ma­ terial production for 1812-13 at £430.5 million, distributed as follows: 8 Million £ Agriculture 216.8 Mines & Minerals 9.0 Manufactures 114.2 Inland Trade 31.5 Foreign Commerce and Shipping 46.4 Coasting Trade 2.0 Fisheries (exclusive of Newfoundland) 2.1 Chartered & Private Banking Estab­ lishments 3.5 Foreign Income Remittances 5.0 Total

430.5

It will be noted that Colquhoun credited agriculture with producing about one half of the value of the total national material product. Having computed the “new property cre­ ated annually in Great Britain and Ireland,” Colquhoun proceeded to distribute this total as a consumable product among the various classes of the estimated 17 million popula­ tion of these two countries. He frankly ad­ mitted that in developing this distribution he had to rely considerably “upon observa­ tion and conjecture.” For a sample group of 1,765 families in the laboring classes, he computed a total income per family of only £ 4 5 to £ 4 8 a year, or only slightly more than £ 1 0 per person. For 500,000 families in the farmer and lesser freeholder groups, he showed a per family income of £ 1 0 0 to £ 1 2 0 a year, or approximately £ 2 0 per person. He credited the nobility with an average family income of £18,000. The total laboring group, numbering almost half the popula­ tion of the country (8 millions out of 17), was estimated to be receiving total annual incomes of only £ 8 1 .5 millions, or only

Table 7-1. Reproduction of a Part of Table 3 of Colquhoun’s Treatise on the Wealth, Power, of the British Empire, e/c., 2d edition, London, 1815, p. 89

Resources

AN A T T E M P T TO E S T IM A T E T H E N E W P R O P E R T Y C R E A T E D A N N U A L LY IN G R E A T B R IT A IN A N D IR E L A N D Arising from th e Use of C apital combined w ith H um an Labour an d M achinery, in raising various sorts of G rain and Animal Food, and in converting various R aw M aterials in to M anufactures and Fabrics, as derived from Agriculture, M ines and Minerals, Manufactures, Inland Trade, Foreign Commerce and Shipping, Coasting Trade, Fisheries, and Foreign Income. (1812.)

A G R IC U L T U R E Quarters 9 .1 7 0 .0 0 0 a t 6 .3 3 5 .0 0 0 — 16,950,000 — 685,000 — 1.8 60.000 —

G R A IN , viz:—W heat (A) Barley Oats Rye Beans and Peas

s. 70 37 29 43 38

d. 6 * per Q uarter. do. do. 10 do. 10 averaged ditto.

H AY, GRASS, STRA W and F IT C H E S . F IE L D T U R N IP S . PO T A T O ES for Consum ption, including P otatoes for C attle and E xportation. G A R D E N ST U FFS— estim ating 15,000 acres to be in G ardens, including W allfruit, a t 401, per Acre. N U R S E R IE S — ab o u t 5,000 Acres a t 50/. per Acre. O R CH A R D S— ab o u t 100,000, averaging an Acre each, 100,000 Acres a t 20/. per Acre.

£

Exclusive of the Seed and D ung and other M a­ nure, b u t including Labour in th e Field and in H ar­ vest, and th e L abour of Ploughm en, Ploughboys, G ardeners, C arters, W ag­ goners, R eapers of Corn, M owers of Grass and Corn, N urserym en, &c.

32,324,250 11,719,750 24,577,500 1,501,291 3,611,500 £73,734,291

(B)

(8 9 , 200,000 14, 200,000 115 923,626

,

600,000 250,000

,

2 000,000

C ontinued

£195,907,917

EX P L A N A T O R Y N O TE S

(A) Species of Grain

E stim ated Average Each of th e Population Person A ver­ of G reat B ritain an d Ireland aged Q uarters

W heat Barley O ats Rye Beans and Peas

9 ,0 0 0 ,0 0 0 1,500,000 4 ,5 0 0 ,0 0 0 500,000 500,000

T otals

16,000,000

1 IX 1H IX l

Consumed by M an Q uarters

Consumed by Animals Q uarters

Used in Beer and Spirits Q uarters

9,0 0 0 ,0 0 0 1,875,000 6 ,7 50,000 625,000 500,000

210,000 10,200,000 59,000 1,360,000

4 ,250,000

18,750,000

11,829,000

4,250,000

Used in various M anu­ factures Q uarters 170,000 1,000 171,000

T otal of Q uarters 9.170.000 6.335.000 16,950,000 685,000 1.860.000 35,000,000

PR O G R E SSIV E C O N SU M PTIO N C onsum ption One Y ear Six M onths Three M onths Six weeks One m onth Two weeks One W eek One D ay

W h eat Q uarters 9 .1 7 0 .0 0 0 4 .5 8 5 .0 0 0 2,2 92,500 1 ,1 46,250 764,166 382,083 191,041 27,291

O ther G rain Q uarters 25.830.000 12.915.000 6.4 57.500 3,2 2 8 ,7 50 2 .1 5 2 .5 00 1,076,250 538,125 76,875

T otal Q uarters 35,000,000 17,500,000 8.750.000 4.375.000 2,916,666 1,458,333 729.166 104.166

(B) A ccording to a n intelligent practical F arm er in Norfolk, Horses in farm ing cost for their Feeding in Grass 22 weeks (from th e 12th M ay to th e 13th Oct.) a t 4s. 6d. per week, 41. 19s. H ay 2 cwt. per week, 13/. 10s. O ats 2 bushels per week, 11/. 5s. T otal 291. 14s. Cows and F a t Bullocks, grass 27 weeks a t 3s. per week 4/. Is. H a y in th e w inter season 25 weeks 1H cwt. per week a t 3s. 6d. per cwt. 61. 11s. 3d. T urnips half an acre, 31. T otal 13/. 12s. 3d. Sheep 6d. a-head per week for the year 1/. 6s. M oderate as th is com putation m ay be, we have as a n average preferred a still less in order to place th e m at­ te r beyond all dispute, th u s:— Horses in G reat B ritain and Ireland, estim ated a t 1,800,000, a t 45s. each for grass, 4,050,000/.: for H ay a t 61. 10s. each 11,700,000/. exclusive of Straw , which a t 5s. for each horse would a m o unt to 450,000/.— T otal 16,200,000/. H orned C attle in G reat B ritain an d Ireland, estim ated on 10,000,000 a t 40s. each for grass, 20,000,000/.; for H ay a t 60s. each, 30,000,000/.; for field T urnips a t 20s. each, 10,000,000/.; to which m ay be added Straw a t 4s. each, 2,000,000/.— T otal, 62,000,000/. Sheep and Lam bs in G reat B ritain and Ireland, estim ated a t 42,000,000, a t 10s. each for grass 21,000,000/.; and for Turnips a t 2s. each, 4,200,000/.— T otal, 25,200,000/. N am ely:— F o r H ay 41,700,000/.:—Grass 45,050,000/.—Straw 2,450,000/.— T otal (including Fitches) 89,200,000/. And T urnips 14,200,0001. P otatoes are the chief article of food in Ireland; of course th e consum ption is very considerable. This article, tak en a t a very m oderate average for th e consum ption of each fam ily per week th roughout G reat B ritain and Ireland, am ounts to 11,923,626/. The P otatoes used for c attle and exportation every year m ay be fairly estim ated a t 4,000,000/.— T otal 15,923,626/. * The several kinds of G rain are ta k e n a t 50 per cent, less th a n the average prices in the Public M arkets of the 12th Septem ­ ber, 1812.

105

106

THE INCOM E OF NATIONS

one fifth of the total national income. “From about 1820 on, critics of the capitalistic regime and later the socialists,” says Max Beer, “found in Colquhoun’s treatise the facts and figures they required to demon­ strate that the system of private property is based on the plundering of the labouring poor.” 9

e. Colquhoun Echoes the Pleas Made by Davenant and Lavoisier for Govern­ mental Estimates Like Davenant, and especially Lavoisier, Colquhoun addressed a plea to his govern­ ment to undertake the official preparation of national income estimates. His plea was so similar to Lavoisier’s in its wording as to suggest strongly that it might have been copied from it (see Chapter 5, Section 4 b ): In the p u rsu it o f a species o f statistical know ledge, interesting to ev ery civilized nation a n d p a rticu la rly above all o th ers to th e B ritish E m pire, it is a m a tte r o f re g re t th a t th e official m aterials a re so scanty. . . . A p e rio d m ay p erh ap s arriv e a fte r th e a u th o r o f this w ork shall be m o u ld ered in th e dust, w h en th e sam e ground m ay be gone o ver by som e ab le r pen u n d e r th o se su p e rio r advantages w hich the progress o f society w ith resp ect to official facts m ay afford, so as to exh ib it to fu tu re gen era­ tions view s w hich shall be official o f th e value o f the new p ro p e rty c rea te d every y e a r fro m the lab o u r o f th e people, in a ll th e different occupations in w hich th ey m ay be em ployed. Such estim ates, w ere it possible to p ublish th em annually, w ould p rove a m ost c o rre c t b a ro m e te r by w hich the resources o f th e State co u ld be m easured, and its p o w er a n d opulence ascer­ tained. It w ould e n ab le th o se assigned to exe­ cu te the pow ers o f g overnm ent to discover a t once all vacillations w ith resp ect to th e actual situation o f the E m p ire, so as to ap p ly rem edies w here a re tro g ra d e is discovered by such en­ c ouragem ents as should a p p e a r b est calculated to secure the co n tin u an ce o f a course o f p ro s­ perity. . . . O f all b ran ch es con n ected w ith this m ost interesting subject [of political econ­ om y] the a scertain m en t o f th e n a tio n a l incom e, o r th e p ro p e rty c rea te d y e arly fro m th e lab o u r o f the people, is the m ost im p o rta n t; since on the produce o f this la b o u r every th in g depends w hich constitutes th e existence, th e c om fort, th e pow er, and th e security o f th e sovereign and the people.10

f. Critical Appraisal of Colquhoun’s Estimates Colquhoun gave no separate figures for Great Britain and Ireland, nor did he indi­ cate what proportion of his various totals related to the one and what proportion to the other country. This made it difficult to

compare his estimates with those made for previous years by Beeke and others for Great Britain alone. Giffen, nonetheless, at­ tempted a comparison of Colquhoun’s and Beeke’s figures. He referred to the fact that “the Income Tax income of Great Britain about 1812 was 130 millions sterling, and adding about a fourth for Ireland, which seems to have been the calculation of the relative resources of Ireland to those of Great Britain at that time, we get a total of 165 million.” 11 He concluded next that “this figure, according to modern experi­ ence of the relation of Income Tax income to total income would justify an estimate of such total income at about 350 millions; and assuming, what seems to be the case, according to Mr. Beeke’s estimate and other estimates at the time, that the Income Tax income bore a less proportion to the total than it now does, a figure of about 400 millions at least as the national income in 1812, at the high prices then ruling, would not appear to be far off the mark.” Giffen further concluded that Colquhoun’s 400 mil­ lion total for Great Britain and Ireland for 1812 compared well with Beeke’s estimate of £ 2 1 8 millions for Great Britain alone for 1800, saying: “Allowing for the increase of population to 1812 and for a rise of prices, etc., in the interval, this total in the latter year would not be far short of 320, and an addition of one fourth for Ireland would bring the sum up to 400 millions.” Although he found fault with Colquhoun’s methods, Giffen believed that his estimate was only about 8 per cent too high.

2. Projections and Interpretations of Colquhoun’s Estimates by John Gray and Simon Gray Colquhoun’s estimates exerted consider­ able influence on popular British opinion about the country’s wealth and its mode of distribution, and on the development of esti­ mates of Britain’s national income in the years to follow. With the repeal of the income tax in 1816, one of the best sources of data for estimating national income was lost, and for the next twenty-five years or so national income estimators had to depend on less informative sources. Few among them were ready to repeat the extraordinary feat per­ formed by Colquhoun in searching through production and trade data and engaging in multiple calculations of a most hazardous

7 . N E W APPROACHES IN ENGLAND

nature. Most of them were content to pro­ ject Colquhoun’s figures to later years by means of indexes of one kind or another. In 1823, John Gray (1799-1850?), Brit­ ish radical reformer, in his Lecture on H u ­ m an H appiness (1825, reprinted in 1831) undertook to prove, on the basis of Colqu­ houn’s figures, that labor, which was held by Colquhoun himself to be the source of all wealth, was being robbed of four fifths of its produce and that one half the national income was being distributed to “Persons who are not only unproductive, but useless members of society.” 12 The unequal ex­ changes between labor and capital, wrote Gray, were bound to dislocate the entire system of national production. To remedy this he advocated establishing a semicommunistic society in which the marketing system would be nationalized through the organi­ zation of national warehouses in which pro­ ducers would deposit their goods and from which they would receive in return paper money issued by a national bank at prices fixed by chambers of commerce. Another contemporary of Colquhoun, Simon Gray, a critic of Adam Smith’s dis­ tinction between productive and unproduc­ tive labor as well as of the Malthusian theory of over-population, in his book The H appiness o f States presented an analysis of Great Britain’s national consumption, based largely on Colquhoun’s national income estimates. Eliminating some of Colquhoun’s double counting and also allowing for some capital investment, Gray arrived at a figure of £ 2 8 0 millions as representing the total consumption of the people. Gray held all classes to be productive. In a letter ad­ dressed to J. B. Say in France and added to his book in 1817, Gray also made some comparisons between England and France as regards the rate of their economic re­ covery from the effects of the war and capital accumulation, contending that in both respects France was far behind Eng­ land.13

3. Lowe’s Estimates in the 1820’s: the First to Be Expressed in Con­ stant Prices In 1822-23 Joseph Lowe, up to that time a little known writer, published a book entitled T he Present State o f E ngland in Regard to A griculture, Trade, and F i­ nance; with a Comparison o f Prospects o f England and France, which attracted con­

siderable attention and which was recog­

107

nized by later economic theorists (Jevons and others) as being, in some respects, a landmark.14 One chapter in this book was devoted to the presentation of a new esti­ mate of the national income of the United Kingdom built upon Colquhoun’s figures. Lowe projected those figures to the year 1822-23 with the aid of data of the 1821 Census, the Report of the Parliamentary Agricultural Committee of 1821, and a va­ riety of other official and unofficial sources. After allowing for changes in prices (which he estimated to have dropped 60 per cent for agricultural products and 40 per cent for manufactured goods during the period 1812-22), and after correcting Colquhoun’s supposed error in including £ 8 0 millions of the cost of feed furnished to horses, cattle, etc., Lowe arrived at a national income total for 1822 of £ 3 5 0 millions at current prices for Great Britain and Ireland combined. He concluded, in other words, that the drop in prices since the end of the war more than offset the increase in production. Lowe used the national income figures to calculate the burden of existing taxation. He related them to the taxable income re­ maining after deducting from national in­ come the subsistence incomes that could not bear any taxation. He did this for the national incomes of specific years in the preceding three decades, arriving at the in­ teresting figures of Table 7-2, calculated first at current prices and next at prices of 1792: 15 T a b le 7-2. L ow e’s C alcu latio n s of N a tio n a l T a x a ­ ble Incom e a n d T a x B urden, 1792-1823

Y ears

P ublic B urdens

O u r N a tio n a l R evenue o r T ax a b le Incom e

(m illion £ ) A t C u rre n t Prices G re a t B rita in O nly 125 220 300

1792 1806 1814

22 60 80

1823

G re a t B rita in & Irela n d 64 260

1799 1806 1814

G re a t B rita in O nly 22 125 46 170 50 188

1823

G re a t B rita in & Ireland 50 200

P ro p o rtio n of B urden to R evenue (perC 6nt)

18 27 27

25

A t Prices of 1792

18 27 27

25

108

THE INCOM E OF NATIONS

To Lowe, therefore, belongs the credit of having been the first estimator to express a chronological series of national income fig­ ures in constant prices. Phyllis Deane interprets Lowe’s calcula­ tions as indicating that he “postulated a rise in real incomes at the beginning of the Na­ poleonic wars— a rise which averaged about 10.5 per cent per decade if we measure it up to 1806— a stagnation during the second part of the war and an appreciable decline after the end of it.” 16 Actually, Lowe noted that real incomes had increased consider­ ably since the restoration of peace, ascrib­ ing this specifically to three factors: “The additional produce on the part of the hands restored to labour by the peace; The in­ crease of our population since 1812; and The progress of improvement in agricul­ ture and manufactures, by which the same number of hands are enabled to produce a considerably larger quantity.” 17 Lowe was especially interested in com­ bating Malthus’ theory, or its crude inter­ pretation, that the resources of any country are limited and that an increase in popula­ tion inevitably brings in its wake a reduc­ tion in the standard of living. On the basis of his national income estimates, he con­ tended that England’s volume of production had increased since 1792 at least as rapidly as population. These and other observa­ tions, bearing upon some of the most funda­ mental economic issues of the time, earned for Lowe a sure place, even if a modest one, in the history of economic thought. In pursuing this line of analysis, Lowe projected the population of the country to the year 1850 at its then existing rate of growth of 1.5 per cent a year. He also pro­ jected France’s population to the same year at its then existing rate of increase of 1 per cent a year. He concluded that “even as to population, we shall probably overtake our ancient rival ere another generation pass away.” 18 Finally, Lowe devoted considerable space to the effects of both the previous inflation and the then current deflation on produc­ tion, consumption, and the government’s fiscal operations. He was particularly con­ cerned over the increase in the burden of the public debt on the shrinking money value of the national income. He proposed that the government enact a tabular stand­ ard that would control the value of money

and the payments of both capital incomes (including interest payments on government debt) and labor incomes. This tabular standard was to be based on the “power of purchase” of money in terms of basic com­ modities. He proposed that the application of this standard to holders of both govern­ ment and private obligations be made vol­ untary. He argued that in the long run creditors would be as much benefited by such a stabilization of the value of money as would the debtors. Lowe was not the first person to make such a suggestion. Schumpeter points out that Sir George S. Evelyn had made essentially the same pro­ posal before the British Statistical Society in 1798, twenty-five years before Lowe. But Lowe expressed the idea far more clearly than did Evelyn and he implemented it with better technical tools.19 In the face of increasing burdens of taxa­ tion under conditions of sharp deflation, Lowe advocated the resumption of govern­ ment borrowing. He maintained that “one of the chief features in the great transition from war to peace, was an increase of dis­ posable capital, and considering the magni­ tude of this increase, we may well ques­ tion, whether government ought not, sev­ eral years ago, to have made a demand on the monied interest for a loan, rather than on the public for taxes.” He believed that such a policy would be of benefit to the monied interests, inasmuch as “to take a few millions annually out of the money market would, doubtless, operate in some measure to retard the fall of interest.” 20 Thus Lowe may well be credited with hav­ ing been one of the earliest advocates of “compensatory,” inflationary deficit financ­ ing of government expenditures during pe­ riods of deflation, anticipating Keynes by more than a hundred years. Following in the footsteps of William Petty, Gregory King (see Chapter 2, Sec­ tions lc and 2b and c ), and Simon Gray (Section 2 above), Lowe prepared “A Table of Annual Consumption” of the Nation, contending that it was just as important to know how the national product is used as to know how it is produced. His table, as he readily admitted, was largely a rearrange­ ment of Colquhoun’s figures, except that it related to a different year. Here is how it appeared in his book: 21

7.

N E W APPROACHES IN ENGLAND

Table 7-3. Lowe’s Table of National Expenditure or Consumption of Great Britain and Ireland for 1823 Expended on the produce of the soil for the food of man, or for purposes of manufacture £120,000,000 On the produce of the mines 10,000,000 On manufactures for home consumption 70,000,000 On houses built or repaired; on furniture; and on improvement of land on whatever is termed in low real property 30,000,000 On all goods imported, whether for consumption, such as tea, sugar, coffee; or for manufacture, as wool, hemp, iron 70,000,000 On all commodities or products not comprised in the preceding 50,000,000 Total consumption

£350,000,000

Lowe also attempted to estimate the an­ nual saving or investments in the sense of “addition to our national capital.” He wrote that if such addition was to be considered in the limited sense of investments in money securities, the amount would be very small. “But if we take in a more liberal sense the difference between the revenue and expendi­ ture of the nation, if we consider as saving or as increase of our stock, all that is laid out on the improvement of land, the build­ ing or repair of houses, the increase of fur­ niture, and if to these we add interest on money saved, we shall find on the whole, an addition to our national capital of 50 or 60,000,000.” 22 Finally, Lowe projected Colquhoun’s esti­ mate of total national wealth to 1823, but he did not consider the evaluation of wealth as important to economic analysis as the evaluation of national income. Lowe’s estimates of national income, both in their character and in the uses to which they were put, were of considerable merit. While they lacked the spectacular features of Colquhoun’s calculations, they were used for a superior type of theoretical analysis.

4. Pebrer’s Projections in the Early 1830’s In 1833, ten years after the appearance of Lowe’s book, Pablo de Pebrer published a lengthy work in which he projected Colqu­ houn’s estimate of national income by more or less comparable methods. Pebrer was a

109

native of Spain and, although he emigrated to England, he apparently retained his in­ terest in the affairs of his homeland, dedi­ cating his book to Spain’s newly crowned Queen Christina. He called his book Taxa­ tion, R evenue, tics, and D ebt Practical Plan Liquidation o f

Expenditure, Power, Statis­ o f the British Em pire and a fo r A p p lyin g Them to the the N ational D eb t ,23 Pebrer

projected Colquhoun’s estimates by gener­ ally raising them one third. He based this percentage increase on data from the 1831 census and other government and private sources, using these, however, to modify the percentage for individual branches of pro­ duction wherever he felt it necessary. Thus, he raised Colquhoun’s figure for mining in­ come 200 per cent, but for manufacturing income only 50 per cent; assumed no in­ crease in income from transportation and distribution; and postulated a 50 per cent re­ duction in income from government service. In this fashion he arrived at a national in­ come figure for Great Britain of £ 5 1 5 mil­ lions for 1830.24 He also projected Colqu­ houn’s estimate of national wealth, furnish­ ing elaborate tables to support his findings of a one-third increase in national income since 1812. Phyllis Deane says: A c o n sid era tio n o f th e b ro a d stru c tu ra l changes im plicit in P e b re r’s estim ates suggests th a t th e re w as a m a rk e d fa ll in th e p ro p o rtio n o f th e n a tio n a l incom e abso rb ed by the m ining, d istrib u tio n and tra n s p o rt g roup o f industries, a slight rise in th e sh are o f m an u factu res. T hese a re plausible results, if it is assum ed th a t the w artim e inflation benefited those engaged in d istrib u tio n a n d tra n s p o rt m o re th a n th e p ro ­ ducers o f com m odities. P e b re r’s estim ates fo r agricu ltu re seem to reflect th e view th a t the re a l o u tp u t o f this in d u stry h a d increased sub­ stan tially o v e r th e p e rio d : his estim ates fo r g rain crops, fo r exam ple, involve an estim ated increase in th e u nw eighted volum e o f o u tp u t am o u n tin g to n e arly 46 p e r c en t.25

Pebrer’s estimates and discussions largely concentrated on one objective—to prove that the national debt and the interest charges on it were imposing a crushing burden on the common people through the existing regressive taxes and were prevent­ ing the expansion of Britain’s production. Pebrer maintained that the only way to re­ duce this burden and to free the economy was to liquidate the larger portion of this debt quickly by the imposition of a huge capital levy. He proposed a levy of 9.25 per cent of the value of all real and personal

110

THE INCOM E OF NATIONS

property to be paid off over a period of two years. Such a levy would produce £ 5 0 0 millions and would reduce the debt by more than one half in one stroke.26 He main­ tained that the imposition of such a capital levy would benefit not only the large body of consumers, but also, in the end, the propertied classes who would have to pay the levy.

5. Influence of the Controversy Be­ tween the Agricultural and Manufacturing Interests in the 1840’s on Inquiries into National Income The sharp debate between the representa­ tives of the agricultural and manufacturing interests over the government’s foreign trade policy that had been going on for some decades (see Chapter 3, Sections lb and 5a, above), and which erupted with new vio­ lence during the 1830’s and 1840’s, led to a round of new inquiries into the national in­ come of the country and the relative outputs of the different branches of the economy.

a. McCulloch’s and Porter’s Lack of Faith in National Income Estimates John Ramsay McCulloch and George R. Porter, the two most influential English economists and statisticians of the 1830’s and 1840’s, had no faith in national income estimates as such, but placed complete re­ liance on specific statistics of the growth of output of given branches of the economy, expressed both in physical and monetary terms. McCulloch (1789-1864), author of the Principles o f Political Economy (London, 1825), at one time professor at the newly founded University of London, and a lead­ ing free-trader and protagonist of industrial development, published, in 1832, a compre­ hensive statistical compendium, intended to displace Colquhoun’s work.27 In this book, he showed, among many other things, that manufacturing was rapidly gaining ascend­ ance over agriculture, and contended that its development should be further encour­ aged by the government through the repeal of the Corn Laws and the further reduction in the prices of foodstuffs. Although Mc­ Culloch did not calculate the national in­ come—and, in fact, declared (in his cus­ tomary doctrinaire manner) that “it is quite

impossible to form anything like a correct estimate of the total aggregate income of any extensive country”— he did give figures on the growth of agricultural, manufactur­ ing, and other production over the years. He was particularly severe in his criticisms of Colquhoun’s work, contemptuously sug­ gesting that many of his figures had been taken out of the “Arabian Nights Enter­ tainments.” Porter (1792-1855), eminent statistician and economic publicist of his day, and, like McCulloch, an advocate of the vigorous development of manufacturing and the abandonment of the protection of agricul­ ture, followed McCulloch’s publication, a few years later, with one of his own, which was even more comprehensive in scope. In this work, The Progress o f the Nation,28 he likewise presented extensive statistics on the growth of national production, and particu­ larly of its manufacturing sector, and argued that the cheapening of foodstuffs and the growth of manufactures would eventually benefit the agriculturalists themselves by providing them with a wider local market. Chief of the newly established Statistical Department in the Board of Trade, one of the founders of the Royal Statistical Society and of its Journal, author of many articles dealing with current economic issues, mar­ ried to David Ricardo’s sister, Porter exerted a powerful influence on the development of statistics and on government economic pol­ icy of the time. But in his entire compen­ dium of 800 pages he did not mention even once the term “national income,” and his only reference to Colquhoun was related to the latter’s statement before a Parliamentary committee in 1816 on the prevalence of drunkenness in London in 1744.

b. Spackman’s National Income Estimates: a Defense of Agriculture A vigorous challenge to McCulloch’s and Porter’s economic statistics and free trade arguments was presented in 1846 by W. F. Spackman, an author with a knowledge of statistics, a strong believer in agriculture as a prime source of national prosperity, and an advocate of the continuance of the Corn Laws and of agricultural protective duties generally. In a book published during that year,29 Spackman sought to demonstrate that contrary to the claims of McCulloch and Porter, agriculture was still the prime occupation of the nation and was entitled

7 . N E W APPROACHES IN ENGLAND

to continued protection by the government against imports of cheap grain from abroad. To prove this, Spackman not only reviewed McCulloch’s and Porter’s figures of the relative numbers of persons employed in the various occupations and the values of their outputs, but also constructed an estimate of the national income and of the contribu­ tions made to it by the various branches of production. Spackman followed the Smithian concept of the national income as the net value of material production. He used occupational breakdowns of the population census of 1841 and other information, criticizing, however, the makers of that census for following Porter’s recommendation to lump together, in one occupational group, the persons gainfully employed in manufactur­ ing, commerce, and trade. He contended that this consolidation was designed “to conceal the weakness of the manufacturing interest and at the same time to oust the agricultural interest from that precedence which, in all former times, had been ac­ corded to it in the national accounts.” He built his national income estimate by examining closely all available data on the gross output and cost of raw materials in the several branches of material produc­ tion. But in his eagerness to prove the con­ tinued importance of agriculture, he over­ estimated the net value of its product. Thus, he did not deduct the cost of seed and feed from the gross value of agricultural output. Nor did he compute separately the values of the services of trade and transportation, but included them instead in the value of the final products. By means of these de­ vices, he was able to show that agriculture was responsible for more than half of the national income— a result that was far from accurate. His estimate was as follows: 30 T able 7-4. S p a c k m a n ’s E stim a te of N a tio n a l Incom e of th e U n ited K ingdom for 1841 A griculture £250,000,000 M an u factu res, d e d u ctin g th e value of th e raw m a te ria l 127,000,000 M ining In te re s t 37,000,000 C olonial In te re s t 18,000,000 Foreign C om m erce (including th e shipping in te rest) 10 p e r c e n t on th e a m o u n t of o u r E x p o rts a n d Im p o rts 15,000,000 Fisheries 3,000,000

£450,000,000

111

Spackman also attempted to prove, by means of exceedingly extensive computa­ tions, that agriculture, while contributing twice as much to the national income as manufacturing, paid more than three times as much in direct and indirect taxes, namely £ 1 3 .9 millions as against the latter’s £ 4 .4 millions (pp. 2 6 -2 8 ).

II. Estimates by the Income Dis­ tributed Method, 1846-1900 In 1842, with the reintroduction by Rob­ ert Peel of the income tax, after a lapse of twenty-six years, a new epoch opened up for estimating national income. The income tax was first introduced as a temporary measure to fill the gap in public revenue created by the repeal of the Corn Laws and the repeal or reduction of many other cus­ toms duties, but it was regularly re-enacted thereafter, and finally, in 1874, was made permanent. The tax was recognized not only as a highly productive source of revenue but also as a much more equitable mode of rais­ ing it than were the consumer taxes. Labor income was exempted from its application, just as it had been under Pitt’s tax. Taxable incomes were classified under five schedules: Schedule A embracing incomes of land­ owners and from houses, railways, and mines; Schedule B, those of land occupiers, on half the rent; Schedule C, those from public funds (bonds, etc.); Schedule D, those from trades and professions, and for­ eign property; and Schedule E, those from public offices. The tax thus covered, thor­ oughly and universally, the middle and up­ per classes. The new records of income tax collections provided far more reliable data for use in estimating national income than did the production and trade statistics on which Colquhoun and his successors relied. The first person to prepare a national income estimate with the aid of these new data was William Ray Smee; the second, Robert Dudley Baxter; the third was Sir Robert Giffen.

1. Smee’s Estimates of 1846— The First to Be Based in Part on In­ come Tax Returns Little is known about William Ray Smee. The British Dictionary of National Biogra­ phy does not list him; nor does Palgrave’s

112

THE INCOM E OF NATIONS

Dictionary of Political Economy mention him. The libraries reveal the publication by him of three brochures of which only the first, treating the income tax and containing an estimate of the national income for 1846, seems to have attracted some attention, and then only many years later.31 The title pages of his pamphlets list him as a “Fellow of the Society of Antiquarians of London”— a designation which may or may not be in­ dicative of his occupation. The subject of his third pamphlet advocating an increase in the small salaries of government employ­ ees suggests that he might have been a civil servant; on the other hand, his concern might have been entirely disinterested. A serious injury he sustained in a railroad accident in 1852 may have been responsible for the shortness of his literary career. An examination of his three publications shows him to have been well read in history and economics, and possessed of considerable ability in handling statistics. Smee’s estimates of national income were made for Great Britain only. In preparing them he used the data of the 1841 popula­ tion census, with its breakdown of the popu­ lation by occupational groups, as well as the returns under the new income tax that had just become available. He prepared the estimates primarily to demonstrate the in­ adequacies of Peel’s income tax and the superior merits of his own proposal of a broader income tax. Peel’s income tax applied only to in­ comes in excess of £ 1 5 0 , i.e., to a far more restricted group than Pitt’s, which applied to all incomes over £ 5 0 . It was also levied at a much lower rate, namely at only 2.93 per cent ( £ 2 18s. Ad. per cent). Peel had expected it to yield £ 3 .8 millions, but it actually produced £ 5 .4 millions during the first year of its operation. It enabled Peel not only to reduce or repeal a great many duties, but also, for the first time in years, to balance the budget and even to realize a sizable budgetary surplus for debt reduc­ tions. But to Smee these results appeared insufficient. He wished to see a complete abandonment of all customs duties and in­ ternal levies on consumption, including the tax on beer and the one on windows. These duties and taxes produced together an in­ come of £ 7 .8 millions a year. Smee, with the aid of his national income estimate, sought to demonstrate that a completely universal income tax at the same rate of 2.93 per cent would produce a revenue of

£ 1 4 .2 millions which would be more than enough to do away with all of these other levies. The income tax returns cited by Smee showed that the tax was paid by 500,000 individuals (318.5 thousand men, 169.5 thousand women, and 20 thousand minors of either sex), and that their total assessed income was £ 1 8 5 millions. Smee estimated the total population of the United Kingdom for 1846 at 20 millions (7 per cent higher than reported by the 1841 census), and the total number of gainfully employed or income-receiving persons at 8.2 millions. This meant that the income tax was paid by only 6 per cent of the total number of income recipients in the country, and that in order to obtain a figure of the national income it was necessary to ascertain the incomes of the remaining 94 per cent. To do this, Smee divided the total num­ ber of income recipients into three groups: the upper income group, which he identi­ fied for convenience’s sake as embracing all individuals with incomes above £ 1 5 0 , i.e., all payers of the existing income tax; the middle income group, which he defined as comprising those receiving between £ 5 0 and £ 1 5 0 a year; and the low income group, having incomes of less than £ 5 0 . The idea of dividing society into upper, middle, and “lower” classes was not new, but its expression in terms of the size of individual incomes was new, and so was the application of this classification of in­ comes to the measurement of the national income. In fact, this classification of in­ comes represented a very ingenious ap­ proach to the problem of estimating na­ tional income, greatly facilitating the work involved. It was followed by R. Dudley Baxter, in the preparation of his estimate of national income, twenty-two years later. Why Baxter did not give credit to Smee for introducing this approach is not clear. That Baxter took it from Smee is evident, for he was acquainted with Smee’s estimate, quoting the latter’s figure in his work (see Section 2 below). Taking the occupational distribution of the 1841 Census, which was given for men and women separately, Smee allocated the number of men and women in each occu­ pation to the appropriate income class, in­ dicating in each case, as far as conditions permitted, the average income in each occu­ pation for men and women separately. For example, taking the 300,000 male farmers

7.

113

N E W APPROACHES IN ENGLAND

T a b le 7-5. Sm ee’s E stim a te s of N a tio n a l Incom e of G re a t B rita in in 1846, w ith B reakdow ns b y Sex, Age, a n d Incom e Class (R e arra n g ed from Sm ee’s T ab u latio n s)

Incom e G roup

N u m b e r of A verage Incom e R ecipients ( £ per y ear) (th o u san d £ )

I. U pper Incom e M ale F em ale M inors

467 200 200

T o ta l I I . M iddle Incom e M ale Fem ale M inors

100 66 20

T o ta l I I I . L ow er Incom e M ale Fem ale M inors

28 15 11

T o ta l G ra n d T o ta l U p p er Incom e M iddle Incom e L ow er Incom e T o ta l N a tio n a l Incom e so u rc e:

T o ta l Incom e (m illion £)

318.5 169.5 12.0

148.7 3 3 .9 2 .4

50 0 .0

185.0

1 ,2 2 6 .9 394.3 2 0 0 .0

122.7 25 .7 4 .0

1 ,8 2 1 .3

152.4

3 ,3 8 7 .7 1 ,1 6 3 .2 1 ,2 6 3 .7

9 5 .7 16.7 13.5

5 ,8 1 4 .6

125.9

500 .0 1 ,8 2 1 .3 5 ,8 1 4 .6

185.0 177.5 125.9

8 ,1 3 5 .9

4 8 8 .4

W . R . Sm ee, The Incom e T ax, L ondon, 1846, pp. 4-10.

and grazers recorded in the census, he placed 20,000 of them, according to the income tax returns, in the upper income group, and the remaining 280,000 in the middle income group; he classified the 946,812 male agricultural workers in the lower income group, identifying their average in­ come, at the same time, at 8s. per week; the 589,078 male nonagricultural laborers he classified also as belonging to the lower income group, but credited them with a somewhat higher average income of 11s. per week; the 2,182,167 male proprietors and employees in commerce, trade, and manufactures he split three ways: 106,500 in the upper income group, 800,000 in the middle income group, and 1,275,667 in the lower income group, identifying the latter’s incomes at an average of 15s. per week. Of the 63,409 male persons in the professions, he placed 30,000 in the upper income group and 33,409 in the middle income group. Next, Smee reassembled his distributions of men, women, and minors in the various

occupations and income groups in one grand table. For each income group, he gave the numbers, the average incomes, and the total incomes for men, women, and minors. Adding all these incomes together, he ob­ tained the national income total. This total was £ 4 8 8 .4 millions. In Table 7-5 above, the final stage of Smee’s calculations is re­ produced in convenient and easily under­ stood form. Space does not permit a repro­ duction here of all the intermediate stages of his calculations. Next, Smee proceeded to verify his figure of £ 4 8 8 .4 millions by calculating the yield of Pitt’s income tax for 1811 by the same procedure. Inasmuch as the 1811 popula­ tion was six tenths the size of the 1846 population, he reduced his calculation of the incomes over £ 5 0 in 1846 (£ 3 5 2 .5 millions) to the latter proportion, thus ob­ taining a figure of £ 2 1 7 .5 millions. From this figure he deducted £ 1 6 .5 millions to account for certain exemptions granted un­ der Pitt’s tax. The resulting figure of £ 2 0 1

114

THE INCOM E OF NATIONS

millions was £.31 millions above the actual assessed incomes ( £ 1 7 0 millions) in 1811. Smee attributed this difference to the in­ crease in the per capita incomes since 1811, as well as to the better enforcement of the collections of the tax in 1846; and he con­ cluded that his estimate checked out well with the historical experience. He also com­ pared his estimates with Lord Liverpool’s estimate of 1822 of £ 2 5 0 to £ 2 8 0 millions (“Debate on Agricultural Distress,” House of Lords, Hansard, 1822); and similarly concluded that the comparison, allowing for differences in population and other factors between the two periods, confirmed his esti­ mate. Smee’s methods of verification may have been somewhat crude, but in their resort to independent series they represented a step in the right direction. In her recent study, Phyllis Deane de­ clared that Smee’s estimates “on the whole . . . look fairly plausible with the excep­ tion of those for domestic servants, which are clearly much too low.” 82 Smee believed that all classes of society, including the poorest, should contribute to the public treasury in proportion to their means. He allowed for only one exception from this rule, saying that female workers in the low income group should be exempted inasmuch as they were “either the wives or daughters of the labouring man, never liv­ ing by themselves, and for the most part with their families.” Accordingly, he urged that the tax be made universal (with the last mentioned exception); and that the same rate of 2.9 per cent, which he believed to be reasonable, be applied to all income re­ cipients. Although he believed that in the case of the individuals in the lower income group receiving weekly or daily wages “the tax could be collected without much diffi­ culty from their employers,” he proposed nonetheless to collect it directly from the individuals, doing so not via a percentage tax, but via a head tax, ranging from 13s. to 30s. a year and indirectly bearing about the same per cent ratio to their incomes. Smee took issue with McCulloch’s posi­ tion that all national income estimates were imaginary and rested on no solid foundation and that “it is quite impossible to conduct such estimates in a reliable manner.” Re­ ferring to his own undertaking, namely, cal­ culating the national income as a basis for estimating the probable yield of a given tax, Smee asserted that “it is not necessary . . . that a perfect estimate should be made.”

Even an approximate one, derived from a careful analysis of occupational statistics, he asserted, would provide a better basis for the purposes in question than any other possible guide. Smee’s national income estimate was the first in England— and that meant the first anywhere— to have been developed in some part, at least, from actual income tax re­ turns. Pitt’s, Beeke’s, and Bell’s estimates were developed from mere anticipations of the yields of a prospective income tax, while Colquhoun’s estimate was developed from production data. Within the framework of the use of both income tax data and occu­ pational statistics, Smee’s estimate was de­ veloped by a wholly original approach, was completely systematic, and was in many ways more solid than any made by anyone before. It represented a substantial contri­ bution to the advancement of estimating methodology. Even though his work has been seldom mentioned (no reference to it is made by either Giffen or Bowley), it has not been lost, for traces of its influence can be found in later estimates by other statisticians.

2. Baxter’s Estimates of 1868 a. More Solid Estimate than Any of Its Predecessors In 1868, Robert Dudley Baxter (18271875), economist and statistician of con­ siderable repute, presented a paper before the Statistical Society on the “National In­ come of the United Kingdom,” publishing it in book form during that same year. He built his estimate upon income tax statis­ tics (as related to the incomes of the upper and middle class), census data, a variety of surveys on wage rates in various laboring occupations, data on dwelling rentals re­ vealed by the records of collections of local property taxes, and other information. His estimate was thus based on data relating to distributed incomes rather than to the value of production. In its methodological ap­ proach it was a descendant of Smee’s and Beeke’s estimates and, looking further back, of Petty’s and Gregory King’s. In the light of what was to follow, Baxter’s work was a true ancestor of the twentieth-century es­ timates made by A. L. Bowley, Josiah Stamp, Colin Clark, and the Central Statistical Office of the United Kingdom. Baxter’s estimates were more solid than

7.

N E W APPROACHES IN ENGLAND

both Beeke’s and Smee’s. First of all, Baxter had at his disposal data on population and occupations collected in several successive industrial censuses. Secondly, he had the use of official data on the collection of the income tax over a period of twenty-five years, as well as a wealth of information provided by numerous surveys. Baxter’s work was characterized by a careful choice of figures, close reasoning, and the utmost conciseness, lucidity, and elegance of ex­ position. In the introduction to his work, he pictured the task of estimating national income by means of the following analogy: T h ere is in th e A tla n tic a n islan d — th e P e a k o f Teneriffe— w hich rises fro m th e sea in a p y ram id al fo rm to th e h eig h t o f 12,000 feet. . . . A n in h ab itan t is scarcely aw are o f its real p ro p o rtio n s: fo r if h e lives a t its foot, he sees chiefly the lo w er em inences w hich rise im ­ m ed iately above h im ; a n d if he clim bs th e h eights h e is a p t to lose sight o f th e b ro a d base w hich sp read s below . . . . I h a v e o ften th o u g h t th a t such an islan d is a good em blem o f a w ealthy state, w ith its long low base o f lab o u rin g p o p u latio n , w ith its u plands o f the m iddle classes, a n d w ith th e to w ering peak s and sum m its o f th o se w ith prin cely incom es. T he difficulty is, to a scertain th e re la tiv e dim ensions o f these m o u n ta in zones. . . . W e m u st sail o u t in to th e offing, till w e can see th e islan d as one gran d w hole, and realize its tru e p ro ­ po rtio n s.33

Elaborating further upon the usefulness of national income estimates, Baxter wrote: T h ere can scarcely b e a n in q u iry m o re in­ teresting to th o se w ho ta k e a prid e in th e ir c o u n try th an th e investigation o f th e statistics o f o u r N a tio n a l In co m e. W h a t a re th e m eans a n d aggregate w ages o f o u r lab o u rin g p o p u la ­ tion; w h at the n u m b ers a n d aggregate profits o f the m iddle classes; w h a t th e revenues o f o u r g re at p ro p rieto rs a n d cap italists; a n d w h a t the p e cu n iary strength o f th e n a tio n to b e a r th e burdens a n n u ally falling u p o n us? W h a t cap ital in lan d and goods a n d m oney is sto red up fo r o u r subsistence, a n d fo r carry in g o u t o u r e n te r­ prises? W h at is th e relativ e m ag n itu d e o f o u r N a tio n a l D e b t? W h a t pro g ress h a s b e en m ade since the beginning o f the c en tu ry in th e in ­ crease o f o u r incom e and th e a cc u m u la tio n o f savings? A n d w h at a re th e risks to w hich o u r w ealth is exposed, a n d th e p re ca u tio n s th a t o u g h t to be tak e n fo r o u r ow n p ro tec tio n and fo r the safety o f p o sterity ? 34

Then, in a few masterly strokes, Baxter outlined the estimator’s problems: T h e m aterials fo r such an in q u iry a re a b u n ­ dantly am ple; b u t th e ir e n o rm o u s m ass re n d ers it difficult to p re sen t th em clearly a n d in sm all

115

com pass. T h e long catalogue o f occupations of th e people, a n d th e infinite v ariety o f wages even in th e sam e occupations, can only be ap ­ p re ciate d by those w ho have endeavoured to red u ce th e m to o rd e r. M in u te accuracy is u n ­ a tta in a b le , and w e a re obliged to w ork b y gen­ e ra l averages. T h e g re at object is to re n d er those averages tru stw o rth y a n d sim ple, and th a t th ey should n o t be undigested m asses o f figures, o r m ere lists o f u n co n n ected totals, b u t co ­ h e re n t a n d lucid. N o r o u g h t im p o rta n t facts to re st u p o n m ere assertion; th e authorities fo r the facts, a n d the reasons fo r th e calculations, o u g h t in every case to be given, so th a t the re a d e r m ay re fe r a n d verify fo r him self.35

b. Baxter’s Concept of National Income Unlike Colquhoun, Baxter did not agree with Adam Smith’s classification of labor into productive and nonproductive. He con­ sidered this classification to be “in reality of a superficial character, and inaccurate, because every class contributes something to production.” He believed that “it would be more accurate to classify the different occupations into the two heads of Produc­ tive and Auxiliary, and to drop the term Non-productive,” 36 but he compromised by introducing a threefold classification: pro­ ductive labor that includes all Adam Smith’s components, except retail trade; the aux­ iliary class, that includes retail trade, trans­ portation of people, and services of hous­ ing; and the unproductive, that includes gov­ ernment services, professions, domestic service, and the holding of government debt. He included the incomes from all three categories in national income, calling this aggregate “gross income” as distinguished from “net income,” which he defined as the sum of the incomes of the productive class and of a part of the incomes of the auxiliary class. This threefold classification of labor and singular distinction between “gross” and “net” income were not much of an improve­ ment on Adam Smith.

c. Baxter’s Calculations As a first step in his calculation of the national income, Baxter took the population figure of the 1861 census and broke it down into the number of income recipients and the number not having income, subdividing the former, in turn, into the number in the upper, middle, and laboring classes, includ­ ing dependents in each case. Next, he pro­ jected these figures on the basis of the past annual rates of growth to the year 1867. His population figures for 1867 were as

116

THE INCOM E OF NATIONS

follows: 21 millions for England and Wales; 3.2 millions for Scotland; 5.6 millions for Ireland; totaling 29.7 millions for the United Kingdom as a whole.37 He then calculated the incomes of the upper and middle classes on the basis of the Income Tax, keeping in mind that in­ comes below £.100 were not taxed (and that some people in the middle class had in­ comes below that sum) and that an exemp­ tion of £ 6 0 was allowed on incomes up to £ 2 0 0 . He had to make adjustments not only for these exemptions but also for the fact that a recipient of income could be taxed under more than one schedule and that “the total of the numbers of ultimate taxpayers in each Schedule by no means represents the net number of individuals charged.” 38 The wages of manual labor, which were not subject to the income tax, were esti­ mated by Baxter on the basis of census fig­ ures of the number of wage earners in vari­ ous occupations as well as on data derived from various published sources on average earnings in such occupations. These in­ cluded Professor Leone Levi’s comprehen­ sive study of Wages and Earnings o f the Working Classes (London, 1867) and his own direct investigations. In estimating the average earnings, he took account in each case of differences in the rates of earnings of men, women, and children as well as in the number of days or hours ordinarily worked by them during the year. Baxter ended with the following income figures for 1 8 6 7 :39

million, or 10 per cent, receiving incomes of over £ 1 0 0 . Baxter also broke down the total income according to its sources, showing that about one third was derived from capital and two thirds from labor and profits: 41 M illion £ F ro m C a p ita l U p p er a n d m iddle classes F ro m P rofits a n d E arn in g s U pper a n d m iddle classes M a n u a l lab o r class

280 .0 209 .5 324 .5 53 4 .0

T o ta l

81 4 .0

His final distribution, by the three classes of labor first mentioned, was as follows: 42 M illion £ I

II III

P ro d u c tiv e Classes A gricu ltu ral M a n u fac tu rin g , m ining, e tc.

165.8 3 1 3 .9

4 7 9 .6 196.0 138.4

A uxiliary Classes N o n p ro d u c tiv e Classes

8 1 4 .0

T o ta l

The income of the productive classes he considered to be wholly “original earnings,” that of the auxiliary classes to be partly original and partly derivative, or “second­ hand,” and that of the unproductive classes to be wholly derivative. Accordingly, while

A verage Incom e per

E n g lan d & W ales S co tlan d Irela n d

N um ber of Incom e R ecip ien ts (m illions)

A m ounts of Incom e (m illion £ )

Incom e R e cip ien t (£ )

H ead of P o p u latio n (£ )

9 .8 1 .4 2 .5

662 74 78

68 53 31

32 2 3.5 14

13.7

The division of income among the several classes and sources of income that emerged from his calculations for the whole United Kingdom appears as Table 7-6.40 It will be noticed that Baxter’s calculation showed 12.4 million income recipients, or 90 per cent of the total number, receiving incomes of less than £ 1 0 0 a year and 1.3

designating the £ 8 1 4 millions aggregate as the “gross income of the nation,” he esti­ mated the “net income,” comprising that of the productive class and part of the aux­ iliary class, at between £ 5 5 0 and £ 6 0 0 millions. Except for his misleading breakdown of national income into three classes of labor

117

7 . N E W APPROACHES IN ENGLAND

Table 7-6. Baxter’s Calculations of the National Income of the United Kingdom for 1867 N u m b e r of Incom e R ecipients (thousands) U p p er a n d M iddle Incom es I L arg e incom es (1) £5,000 a n d u p * (2) £1,000 to 5,000 II M iddle incom es £300 to £1,000 I I I Sm all incom es (1) £100 to £300 Below incom e tax : (2) U n d er £100

A m ount of Incom e (m illion £)

8 .5 4 8 .8

126.2 8 3 .3

178.3

8 7 .7

1 ,0 2 6 .4

110.9

1 .4 9 7 .0

8 1 .3

2 .7 5 9 .0

4 89 .5

1 .3 4 5 .0

6 6 .4

5 .0 8 7 .0

160.6

4 .5 2 9 .0

9 7 .6

T o ta l

1 0 ,9 6 1 .0

3 24 .6

G ra n d T o ta l

1 3 ,7 2 0 .0

814.1

T o ta l M a n u a l L ab o r Class IV H ig h e r skilled labor a n d m an u factu res £50 to £73 V L ow er skilled lab o r a n d m anufactures £35 to £52 V I A g riculture a n d unskilled labor £ 1 0 10s. to £36

* T h e £5,000 incom es included £30 m illions of incom e of 400 com panies.

and the erroneous distinction between gross and net income, Baxter’s estimate was su­ perb.

3. Estimates by Giffen and Others Sir Robert Giffen (1837-1910), the lead­ ing British statistician of the last quarter of the century, continued the systematic work begun by Smee and Baxter of estimat­ ing national income on the basis of income tax statistics. From 1868 to 1876 Giffen was assistant editor of the influential London Economist (of which the celebrated Walter Bagehot was the editor). From there he moved to the government, becoming chief of the Statistical Department in the Board of Trade and eventually the first head of the newly created Labour Department. In 1886, Giffen directed the first national cen­ sus of wages and was responsible for many improvements in English economic statistics. He was much more interested in the growth of wealth (or capital) than in the growth of national income, and his book, The Growth o f Capital, is a classic.43 So, too, are his papers delivered before the Statistical Society on the “Progress of the Working Classes in the Last Half Century” (London,

1883 and 1889). He was eventually knighted for his scholarly and administrative work. In 1878, in a paper delivered before the Statistical Society, Giffen estimated the na­ tional income of the United Kingdom for the year 1874-75 at £ 1 ,2 0 0 millions; and in The Growth o f Capital, in 1889, he esti­ mated it for 1884-85 at £ 1 ,3 5 0 millions. Giffen did not pursue Baxter’s distinction between “gross” and “net” national income on the basis of the types of labor receiving the income, and his figures included all in­ comes. At the end of the century, in 1895, A. L. Bowley began publication of his compre­ hensive studies of “Changes in the Average Wages in the United Kingdom since 1880.” In the twentieth century, he made similarly exhaustive studies of changes in the national income.

III. Summary of the British Estimates of the Nineteenth Century We can now put together the main esti­ mates of national income prepared in Eng-

118

THE INCO M E OF NATIONS

Table 7-7. Estimates of National Income of the United Kingdom 1800-89

E stim a to r a n d Y ear of P u blication 1. H en ry Beeke, 1799-1800 2. B enjam in Bell, 1799-1802 3. P a tric k C olquhoun, 1812-15

Y e ar C overed b y the E stim a te 1799 1800 1812-14

4. Jo h n G ray, 1825 5. Joseph Lowe, 1822

1818 1822

6. P ablo Pebrer, 1833

1831

7. W . F . Spackm an, 1847b

1841

8. 9. 10. 11. 12.

W . R. Sm ee, 1846b L eone Levi R. D udley B axter, 1868 R o b e rt Giffen, 1878 R o b e rt Giffen, 1889

1846 1858 1867 1874-75 1884-85

N a tio n a l Incom e -— ---------------------Per T o ta l C a p ita (m illion £ ) (£ )

C o n cep t a n d Basis of E stim a te

P o p u latio n (th o u san d s) a

d istrib u te d incom es d istrib u te d incom es v alue of m ate ria l production v alu e of consum ption v alue of m ate ria l production value of m ate ria l production v alue of m ate ria l production d istrib u te d incom es

14,943 14,943 17,110

272“ 300“ 430 d

18.2 2 0 .0 2 5 .0

2 0,0 0 0 2 1,2 0 0

350 350 d

2 0 .5 1 7.0

2 4,2 0 0

559

2 3.1

18,527 b

450 b

2 4 .3 »

2 0 ,1 6 5 b

488 b 600“ 814 1,200 1,350

2 4 .4 b

d istrib u te d incom es d istrib u te d incom es d istrib u te d incom es

29,700 32,4 0 0 35,7 0 0

2 7 .4 3 7 .0 3 8 .0

a P o p u latio n figures a re e stim a te d from th e n e are st decennial census, a n d include rough e stim a te s for Irelan d . P opulation figures for 1800 a re ta k e n from th e A bstract o f A nsw ers and R eturns to the P opulation A ct 1800, G re a t B rita in G eneral R eg ister Office, L ondon, 1801, p. iv ; for 1811 th e y a re ta k e n from th e Abstract of A nsw ers and R eturns to the P opulation A ct 1811, G re a t B rita in G eneral R egister Office, L o ndon, 1812, p. xi; for 1821, 1831, 1871, 1881, 1891, th ey a re ta k e n from th e A n n u a l A bstract of Statistics, N o. 89, C en tral S ta tistica l Office, L ondon, 1952, p. 7, a n d from o th e r sources. b G re a t B rita in only; to include Irela n d a n o th e r £42 to £45 m illions w ould h a v e to be a d d ed to the na tio n a l incom e. Q uoted in B axter, op. cit., p. 66. 0 B eeke's e stim a te w as £218 m illions for G re a t B rita in alone; w e h a v e a d d ed one fo u rth of th is sum for Ireland. B ell’s e stim a te w as for incom es over £15. d T o avoid double counting Lowe d ed u cted £80 m illions fo r th e feed of horses a n d c a ttle , w hich C olquhoun did n o t deduct. T h e ir estim ates, therefore, a re n o t com parable. 6 Q uoted in B axter, op. cit., p. 65.

land during the nineteenth century. We in­ clude Dr. Beeke’s and Dr. Bell’s estimates of 1799-1802 in this series, inasmuch as they were published at the century’s turning point. (See Table 7-7.) In examining this series, it must be re­ membered that the estimates included in it are not fully comparable. Five of them (Colquhoun’s, Gray’s, Lowe’s, Pebrer’s, and Spackman’s), were based on the restricted material production concept, whereas the others were based on a comprehensive pro­ duction concept and were, therefore, much more inclusive. Furthermore, some of them were calculated at more or less inflated prices, whereas the others were computed at preinflation prices or at deflated or sta­ bilized prices. The figures for Ireland in­ cluded in the United Kingdom’s national income and population totals are often very

rough, and in two cases— Spackman’s and Smee’s estimates— they are not included. The per capita income figures are even less comparable than the income totals. During the last quarter of the century the concept of national income was greatly clari­ fied, first by Alfred Marshall (1842-1924) of Cambridge in his E conom ics o f Industry (London, 1879), which was later expanded into his Principles o f E conom ics; later, in 1893, by that incomparable teacher at Ox­ ford and eventually at the London School of Economics, Edwin Cannan (1861-1935), in his H istory o f the Theories o f Production and D istribution (London, 1893); and finally, by William Smart of the University of Glasgow in his D istribution o f Incom e (London, 1899). The effect of their work was to rid the concept of the national in­ come both of the misconceptions introduced

8 . M U L TIPLIC A TIO N OF ESTIM ATES IN FRANCE

by A dam Sm ith, and o f the equally m is­ leading corrections o f them m ade by Bax­ ter. S m art’s book, m ore than any other, con­ cerned itself w ith the question o f w hether the existing distribution of incom e was equitable or not. Sm art had been influenced by th e publication tw o years earlier o f the W ebbs’ book, Industrial D em ocracy, and appropriately dedicated his own to John St.

119

L oe Strachey (1 8 6 0 -1 9 2 7 ), the great pub­ licist, editor of the influential Spectator, and advocate o f m oderate social reform . T he nineteenth-century development of national incom e estim ating in England was rich in both theoretical advancem ent and practical application, and as the century ended, E ngland was still the w orld’s leader in the field.

8. MULTIPLICATION OF ESTIMATES IN FRANCE IN THE NINETEENTH CENTURY T he developm ent of national incom e esti­ m ates in F rance during the nineteenth cen­ tu ry proceeded along lines very different from those th a t characterized their develop­ m ent in E ngland during the sam e tim e. M any m ore estim ates w ere p repared in F rance, but these w ere less well founded either theoretically o r statistically, and con­ sequently were of m uch less practical value th an their English counterparts. D uring the first h alf of th e century, F rench estimates w ere based wholly on production data. T hey w ere designed p ri­ m arily to m easure the rate o f grow th of agricultural and industrial production, but, being influenced by restricted and divergent concepts of national incom e, actually m eas­ u red only a part of the entire national p ro d ­ uct and, m oreover, not always the sam e part. Some of them m easured the gross value of industrial and agricultural production; oth­ ers, their net value; still others, only the net product o f agriculture alone. T hey scarcely deserved, therefore, to be called national incom e estimates. D uring the second half o f the century, and particularly the last quarter o f it, esti­ m ates becam e m uch m ore consistent and meaningful. F irst o f all, they began to deal w ith the total national incom e of th e popu­ lation derived from all occupations and sources instead of w ith a m ere p a rt of it. By th a t time the last advocates o f physioc­ racy had passed away, and the new crop of economists and statisticians— Block, de Notes begin on p. 524.

Foville, Leroy-Beaulieu, and others— fa­ vored the com prehensive production con­ cept of n ational incom e. Secondly, under the pressures of growing agitation fo r fiscal reform and growing social conflict between the laboring an d proprietary classes over wage rates, w orking conditions, an d the equitableness of th e existing social order, estim ators becam e m ore concerned w ith the distribution o f national incom e th an w ith the m easurem ent of national production, and began to base their studies m ore on fiscal, occupational, and business data than on production statistics. W ith the indus­ trialization of the F rench economy, the in­ com es o f the population could be m ore clearly differentiated as labor incomes, capi­ tal incom es, and mixed incomes; and con­ siderable inform ation on these incomes be­ cam e available in financial and business re­ ports. T his tended to facilitate the prepara­ tion of national incom e estimates based on distributed incomes. In their general orientation tow ard the incom e-distributed approach, the F rench estimates of the second half of the century w ere rath e r sim ilar to the ones prepared in E ngland, b u t because F ran ce had no in­ com e tax, her incom e-distributed estimates never achieved the same degree of perfec­ tion as did the English estim ates of the same period. T he developm ent of national income es­ tim ates in F rance during the nineteenth cen­ tury was deeply influenced by the rapidly

120

THE INCOM E OF NATIONS

changing political scene. W ith each new regim e the availability o f official statistics fluctuated and the pertinence of inquiries into the national incom e was re-evaluated. T he development of national incom e esti­ m ating can be divided accordingly into three periods: (1 ) the N apoleonic era (1 8 0 0 1 5); (2 ) the period of the R estoration and Louis Philippe’s reign (1 8 1 5 -5 0 ); (3 ) the period 1850 to 1900 associated w ith the regim e of N apoleon III and w ith the first quarter of a century or so of the Second Republic.

1. Estimates of the Napoleonic Era D uring the N apoleonic era, both scholarly and public interest continued in national in­ com e estimates, w hich had originally grown out of the fiscal and political crises of the “ancien regim e” and the Revolution, even though the reasons fo r and direction o f this interest were now different. N ational incom e estimates were no longer called fo r in order to determ ine the potential revenue to be gained under existing or proposed taxes. T hey were now prepared w ith the single purpose of m easuring the econom ic strength of France so as to com pare it w ith the eco­ nom ic strength of her political and com ­ m ercial rival— England— and to reveal the rate of the country’s econom ic recuperation from the devastation created by the Revo­ lution. A substantial im provem ent in official statistics during this era contributed to the development of estimates. D uring its early days, the R evolutionary governm ent, giving expression to V auban’s urgings of nearly a century before, h ad m ade com prehensive plans fo r collecting statistical inform ation to be used in form ulating fiscal and economic plans. It also had created in 1790 a M inistry o f the Interior w ith a spe­ cial Bureau of Statistics, placing at its head the ardent and able statistician, N eufchateau, and later, the fam ous m athem atician and astronom er, L a Place. In 1791 the govern­ m ent had ordered the holding of a general census, but it was never actually held. T he generally disorganized state of the revolu­ tionary adm inistration, lack of funds, and political considerations apparently prevented conducting the projected census. N apoleon, who cam e into pow er in 1799, continued w ith certain lim itations to foster the developm ent of statistics. In 1799 he prom ulgated a decree ordering a general census for 1800, but this was n o t carried

o ut any better th an the N ational A ssem bly’s proposed census of nine years before. As Professor F au re p u t it in his historical sketch of F ren ch statistics: [Although Napoleon] said “Statistics are the budget of things and without a budget there is no safety” . . . we know that he liked sta­ tistics in the way that Louis XIV liked them, that he demanded statistics of his prefects as the great king had demanded them of his intendants, on the condition they should not fur­ nish fuel for indiscreet curiosity and the mis­ placed criticisms of those whom he called in the worst sense of the word, “ideologues.” 1 On th e other hand, N apoleon gave con­ siderable authority to his M inister of In ­ terio r to collect an d publish com prehensive statistics on population, industry, and agri­ culture. In 1800 he appointed Jean Chaptal, the fam ous chem ist and professor o f the U niversity of M ontpellier, as M inister of the Interior. C haptal took his appointm ent seri­ ously and proceeded to m ake a com prehen­ sive collection o f p refectural statistics. In 1804 he was succeeded by M ontalivet, who was m ore o f a politician th an a scientist. A n enum eration of population, though not based on a house-to-house canvassing, was carried o u t u n d er N apoleon fo r the years 1801 and 1811, an d several volum es of prefectural statistics of industrial and agri­ cultural production w ere published by C hap­ tal by 1804 and by M ontalivet by 1813. T hese p refectural reports varied greatly in quality. M oreover, n ot all the prefectures tu rn ed in th eir reports in tim e fo r publica­ tion, an d in order to arrive at national totals the M inistry h ad to fill the gaps in the data by m eans of rough estimates, m aking the accuracy o f th e final totals questionable. B ut even though these statistics w ere im ­ perfect, they w ere superior to those avail­ able during the R evolution and especially to those prep ared u n d er th e old m onarchy. T hey enabled a new crop of estim ators to prepare better estim ates o f national incom e th an w ere possible before. O utstanding am ong these new scholars w ere Charles G anilh an d Jean C haptal himself.

a. Ganilh’s Estimates C harles G anilh (1 7 5 8 -1 8 3 6 ), a lawyer by profession, was a fo rm er tribune in the R evolutionary governm ent and, u n d er N a ­ poleon, an advocate to th e P arliam ent of Paris (a superior political c o u rt), who turned to fiscal, econom ic, and statistical w riting.2 In his System es de VEconomie

121

8 . M U LTIPLICATIO N OF ESTIM ATES IN FRANCE

Politique G anilh devoted a w hole chapter to the theory of national incom e, taking strong issue w ith A dam Sm ith’s m aterial production concept of national incom e, and supporting instead the com prehensive p ro ­ duction concept advanced by E ngland’s Petty, King, and Beeke, and F rance’s own Boisguillebert and V auban.3 In his last and m ost im portant w ork, L a Theorie de VEconomie Politique, G anilh de­ voted the entire first volum e to an estim ate of the national incom e o f F rance in 1789. In th at estimate, using new er statistical data and a m ore com prehensive concept o f p ro ­ duction, he tried to correct th e earlier esti­ m ates m ade by Lavoisier, Lagrange, and others for th e same year.4 In addition, he com pared his own new estim ates fo r th a t year w ith D r. Beeke’s estim ate fo r E ngland fo r 1798. Follow ing Lavoisier’s approach in essence, but m aking num erous corrections, G anilh prepared separate tables fo r each m ajor branch of production— agriculture, m ining, m anufacturing, and com m erce. F o r each branch he gave the value of its fixed capital; the value of the dom estic raw m aterials em ­ ployed in it; the cost o f labor; the gross value of the product; and the net value (w hich he erroneously defined in accordance w ith the physiocratic and Sm ithian doctrines as the n et rent o r net p ro fit); th e num bers of w orkers em ployed; their average earn­ ings; and the num ber o f persons dependent on those earnings fo r their support. G anilh firmly believed in the effective alliance of econom ics and statistics fo r the exploration o f social problem s, saying: “These two sciences are indispensable to each other. T hey lend support to each other and draw from their convergence a con­ sistency, a pow er, and a brilliance w hich they could never attain in isolation.” 5 In his estim ate of the gross value o f agri­ cultural production, G anilh included the values of the following items om itted by Lavoisier: vegetables, fruit, dairy products, fowl, eggs, honey, raw m aterials supplied to m anufacturers (hides, flax, silk, tobacco ), and horses and m ules supplied to nonagri­ culturalists and to governm ent. H e arrived at a gross value o f agricultural production o f 3.9 billion livres, o r 1.3 billion m ore than Lavoisier and 300 m illion m ore th an A rth u r Y oung.6 H e evaluated the earnings of the cultivators (num bering 3.5 m illion families o r 18 million people) at 2 billion livres (560 livres per fam ily), w hich was about

50 p er cent higher th an Lavoisier’s figure.7 T he rem aining sum, according to his calcu­ lation, was m ade up of 233 million livres fo r the value o f seeds and 400 million livres fo r other advances of cultivators, thus leav­ ing a “n et” o r “disposable” incom e of 1.3 billion livres. T he latter sum in tu rn was divided into 600 m illion livres for taxes; 75 fo r church dues; 130 fo r m aintenance of agricultural property; and only 490 million to cover paym ents to landowners, profes­ sions, m anufacturers, m erchants, and other kinds of services— a category representing a population of 7 millions. In some parts of his estim ate, G anilh fol­ low ed th e physiocratic approach. In others, he departed from it. H is final sum m ations showed considerable breadth, but in making them he seemed to follow the Sm ithian m a­ terial production concept even though he had rejected it in the theoretical portions of his w ork. D espite this lack of consistency, his w ork represented an im portant advance in national incom e estimating. One of the m ost interesting features of his estimates was th e attem pt to m easure the distribution of incom e in F ran ce— an attem pt in w hich he was influenced by D r. Beeke’s work. The m ain elements of his sum m ary are repro­ duced in Table 8-1. Table 8-1. Main Results of Ganilh’s 1812-15 Estimates for France, 1789 Livres (millions) Total population: laboring classes (18 million in agriculture) middle classes superior classes

22.0

2.4 2.0

26.4

(billions) Total capital employed in French production: in agriculture in industry and manufacturing in domestic commerce in foreign commerce

44.0 2.5 .5 .25

47.2

(millions) This capital helped to sustain the production and consumption of the following population: in agriculture in manufacturing in commerce ( Continued on p. 122)

18.0 3.6 3.0

24.6

122

THE INCOM E OF NATIONS

(billions) This capital and labor of this popu­ lation created a product of the following gross value: in agriculture in industry and manufactures in domestic commerce in foreign commerce The costs of this production were as follows: in agriculture in industry and manufactures in commerce

3.9 .8

.7 .3

5.8

2.4 .6 .5

3.5

(no. of livres) This sum of laboring income dis­ tributed among the individuals belonging to that class and their dependents gives the following income to each individual in it: in agriculture in manufactures in wholesale trade in domestic trade in foreign trade

132 160 314 205 191 (billions)

The net product created by the capital and labor of France in 1789 was distributed as follows: in agriculture in industry and manufactures in domestic wholesale trade in domestic retail trade in foreign trade The portion of the population re­ ceiving this income number 2 million, so th at each individual in this group received

1.3 .3 .3 .05 .2

2.1

1,034 livres

s o u r c e : Ganilh, La Theorie de L ’Economie Politique, Paris, 1815, vol. I, pp. 206-209; vol. II, pp. 41, 44.

G anilh concluded that the 5.8-billion-livre gross income of F rance in 1789 was divided as follows: the laboring population, repre­ senting 2%e of the total population, received only three fifths of this sum, w hereas the propertied classes, containing only %e of the population, received two thirds of it; that the laborers were receiving a per capita in­ com e of 120 livres, w hereas the propertied classes received 900 livres. H e com pared these results with those revealed by D r. Beeke for England in 1798, concluding that the levels of income in F rance w ere m uch

low er and the inequalities in the distribution of incom e m ore pronounced than in E ng­ lan d .8 G anilh concluded his estim ate w ith the following two general observations relating to F ran ce’s national incom e in 1789 and to her cu rren t resources fo r w aging w ar: It is seen that the territory of France, one of the most fertile in the world, operated with a fixed capital of some 50 billion livres, cul­ tivated by a purely agricultural population of 18 million individuals and with the aid of an operating capital of 4 billions, was yielding an annual gross income of only 4 billion livres and a net income of only 1.3 billion; and that this income, considered in relation to both this agricultural population and to all other classes of population, provided to all but a meager subsistence, insufficient for their prosperity and incapable of raising the people of France to the power and grandeur consistent with the first rank which it occupied among the nations of Europe. France’s wealth, prosperity, power and splendor had their principal foundation elsewhere than in her territorial wealth. . . ,9 For the past ten years, despite her internal dissensions, France was able to resist and win over a coalition of all European nations di­ rected against her . . . [Her miraculous pow­ ers] did not come from any increase in her territory or in her population, or in the size of her armies . . . in every one of these re­ spects she was outmatched by her adversaries . . . [Her powers] came from the mass of capital accumulated over the years through her maritime commerce. It alone bore the costs of the revolution, her victories, and her con­ quests.10 It was G anilh’s contention th a t F ran ce’s fu tu re prosperity lay in a revival o f her foreign com m erce. G anilh’s national incom e estim ates w ere conceptually w eak and probably contained substantial statistical errors, but th ey w ere conceived m ore broadly and w ere m ethod­ ologically m ore advanced th an any F rench estim ates up to th a t tim e.

b. Montalivet’s Colored Report of 1813 W ith N apoleon returning hom e from his crushing defeat in Russia, M ontalivet, his M inister o f Interior, in a desire to bolster N apoleon’s spirit, subm itted to P arliam ent a hastily p u t together statistical rep o rt on the condition o f F ran ce’s agriculture, indus­ try, and population. H is rep o rt was based o n such prefectural retu rn s as he h ad been able to assemble by th a t tim e. M ontalivet estim ated the total value o f production (us­ ing the new currency) at 7 billion francs,

8 . M U LTIPLICATIO N OF ESTIM ATES IN FRANCE

of w hich 1.3 represented the value of indus­ trial production.11 H is rep o rt painted an u n ­ duly rosy picture of the country’s econom ic condition. A ctually, m any industries had suspended production because the continen­ tal blockade h ad deprived them o f their needed raw m aterials and m arkets. A t the same tim e a country-w ide crop failure had resulted in fam ine. But M ontalivet w rote just the same: “T he population has contin­ ued to increase; industry has m ade new progress; never have the lands been better cultivated; the m anufactures m ore flourish­ ing; at no period in ou r history has w ealth been m ore distributed am ong the different classes o f society!” 12 c. C h a p ta l’s E stim a te s In 1819, Com te Jean C haptal (1 7 5 6 1832), ex-M inister of the Interior and m em ­ b er of th e Royal A cadem y of Science, who by th a t tim e had becom e an elder statesm an, published a com prehensive tw o-volum e w ork on F rench agricultural and industrial production and c o m m e rc e 13 in w hich he com pared the country’s production during the years 1800-12 w ith production before the Revolution. A lthough he dealt mostly w ith physical quantities of products, taking them up item by item w ith m eticulous de­ tail, h e also sum m ated them in m oney values in physiocratic term s as the “gross incom e o f the country” and as its produit net o r taxable income. H is w ork was based m ainly on the prefectural statistics devel­ oped during his and M ontalivet’s m inisterial regimes. H e also used a variety of other contem porary docum ents. F o r th e earlier periods he used the estim ates prepared by Lavoisier, Lagrange, Y oung, A rnould, Tolosan, G anilh, etc. H e attem pted to de­ term ine not only the rate of grow th of the various branches of production but also the factors th a t were responsible fo r the differ­ ences in their growth. T he details of C haptal’s w ork w ere m ore valuable than his physiocratically conceived totals. F o r agricultural production, C haptal estim ated the physical quantities o f each type of production and then m ultiplied each quantity by its average price, thus obtaining a figure of 4.67 billion francs fo r total gross value of agricultural output. F rom this sum h e deducted the value of seed, wages of hired help, repairs of structures, m ainte­ nance o f equipm ent, etc., annual loss of horses and other animals, and the nourish­ m ent o f both m en and animals. These de­

123

ductions am ounted to 3.33 billion francs, leaving a disposable or taxable net income from agriculture of 1.34 billions of francs.14 W ith sim ilarly detailed com putations of quantities and prices, Chaptal obtained a total gross value of industrial production of 1.8 billion francs, o r 500 m illion m ore than M ontalivet. F ro m this sum Chaptal de­ ducted 416 million francs for domestic raw m aterials and 186 m illion fo r im ported raw m aterials; 844 m illion for wages; 192 mil­ lion fo r expenses for repair, m aintenance of equipm ent, heat and lighting, and in­ terest on debt; leaving 182 million francs of entrepreneurial profits.15 Twenty-five years later, M oreau de Jonnes, the em inent F ren ch statistician, took issue w ith some of C haptal’s estimates. H e believed the estim ate of 1.8 billion for the gross value of industrial production to have been 40 p er cent too high. H e ascribed the erro r partly to the fact th at C haptal ac­ cepted the prefectural reports m uch too u n ­ critically, and partly to C haptal’s “southern tem peram ent,” w hich inclined him tow ard exaggeration.16 d. S a y ’s C o n trib u tio n s to N a tio n a l In c o m e A n a ly s is Jean Baptiste Say (1 7 6 7 -1 8 3 2 ), a pros­ perous m erchant w ho becam e a fam ous econom ist, published his im portant Traite d ’E conom ie Politique in 1807. This book was extrem ely influential in th e develop­ m ent of econom ic thought n o t only in F rance b ut also in the m any other coun­ tries in w hich its translations were pub­ lished. A lthough he was uninterested in esti­ m ating national incom e, Say contributed m aterially to clarification of its concept. Originally inclined tow ard accepting the Sm ithian m aterial production concept, he eventually discarded it and becam e an ex­ ponent of the com prehensive production concept. H e also dem onstrated the fallacies o f the physiocratic concept of national in­ com e, writing: The sum of all the incomes of individuals comprising the nation constitutes that nation’s income. It is equivalent to the gross value of all the products of the nation . . . because this value is distributed in its totality to either one or another group of producers . . . , ex­ cept that the product arising from the nation’s foreign commerce must be entered at a net figure. . . . The grave mistake of the eight­ eenth century economists was that they con­ sidered the national income as being constituted

124

THE INCOM E OF NATIONS

of the net product of the earth, falsely assum­ ing that the nation’s consumption consisted of this net product; as if the nation did not con­ sume the whole value that it was creating. . . . I must repeat over and over again the funda­ mental truth that the gross income of the so­ ciety is the same thing as its net income.17 Say was a theoretician, not a statistician. H e had no interest in statistics and seldom used statistical data in support of his theo­ ries. H e believed th a t econom ics and sta­ tistics were wholly unrelated disciplines and he criticized those w riters w ho interm ixed them . H e was extrem ely skeptical about the usefulness of national incom e estimates, m aintaining that they consisted o f values of different natures. “If we evaluate the in­ com e of the whole of F rance in francs,” he w rote, “and if the francs in P aris do not purchase as m uch goods as th e francs in the provinces, we are adding up units w hich are not hom ogeneous. T herefore, if I esti­ m ate the national incom e of F rance at 8 billion francs, I have in my sum francs of all kinds of value.” 18 Beyond offering this figure of 8 billion francs w ithout m uch if any item ization, Say m ade no contribution to national incom e estimating. Several economists quote Say’s estimates of national incom e for 1820 at 7.8 billion francs and of the national w ealth fo r the same year at 120 billion francs, but they give no references to the w orks in w hich these figures appeared; and w e have not been able to find them .19 In view of Say’s lack of interest in statistics, it is not likely that his estim ate w ould have been very elaborate.

2. The Period of Lack of Interest, 1815-50 W ith the fall of the N apoleonic regim e, the further developm ent of statistics and national income estimates was interrupted. F o r some tw enty years (1 8 1 5 -3 5 ) there was scarcely any activity in the field. The period of the Bourbon restoration was a period of reaction. Louis X V III (1 8 1 5 -2 4 ) and especially Charles X (1 8 2 4 30) did not believe in statistics. They re­ garded statistics as an evil invention of the Revolution and a likely source of subver­ sion. In 1821, instead of the originally de­ creed census, a m ere rough estim ate of the current population was prepared in the M in­ istry of the Interior. A t the same tim e, the

B ureau of Statistics was abolished and its statistical records w ere destroyed. T he reign of Louis Philippe (1 8 3 0 -4 8 ) was, at least in its first years, under the able leadership o f Prim e M inister Thiers (jo u r­ nalist an d historian as well as statesm an), m oderately liberal in outlook. It gave F ran ce eighteen years of unprecedented econom ic advance and prosperity. Thiers believed in statistics, and early in the 1830’s re-established the B ureau of Statistics, ap­ pointing as its head M oreau de Jonnes, a brilliant statistician, able w riter, an d ad­ m inistrator, w ho did outstanding w ork in th at office during the next two decades or m ore. T he first tru ly effective enum eration of the population, held u n d er M oreau de Jonnes’ adm inistration in 1836, used, for the first tim e, a form designating inhabit­ ants by fam ily and by household. A t the same tim e, th e preparation of an agricul­ tural and industrial census was ordered. T he agricultural census was com pleted and published by 1839, b u t the industrial census was delayed until 1845. T h e publication of the results was even fu rth e r delayed by the Revolution o f 1848 and by the political turm oil associated w ith the installation of the regim e of N apoleon III. It was not until 1856 th a t the industrial census was published. This great statistical event was Jonnes’ suprem e accom plishm ent, as the census provided the first tru ly authentic and com plete data on F rench industry.20 D uring this period (1 8 3 6 -5 6 ), several econom ic and statistical tracts appeared con­ taining partial, if n o t com plete, national incom e estim ates o f one kind o r another, based on w hatever fragm entary statistics were available. C harles D upin (1 7 8 4 -1 8 7 3 ), m athem atician and onetim e m em ber of the C ham ber o f D eputies, published several books in the late 1820’s and in the 1840’s, containing a series of national incom e esti­ m ates from 1780 to 1840 w ithout, how ­ ever, furnishing any explanation as to how he arrived at these figures. H is estimates fo r the various years w ere: 21 Billion Francs 1780 1790 1800 1810 1820 1827 1830 1840

4.0 4.6 5.4 6.3 7.9 8.4 8.8 10.0

8 . M UL TIPLIC A TIO N OF ESTIM ATES IN FRANCE

J. M . D utens (1 7 6 5 -1 8 4 8 ), who had spent several years in E ngland, was an engi­ neer, a lecturer on social problem s, and one of the last ardent physiocrats in France. In 1842 he published a book containing a historical series of national incom e esti­ mates, b ut like D upin he failed to give any explanation of their bases. It is probable th at his estimates related only to the gross value of agricultural production. D utens’ figures were as follows: 22

Year 1790 1815 1835

National Income (billion francs.) 8.4

Per Capita National Income (francs) 110

190.5 10.6

201

In 1846, J. H . Schnitzler (1 8 0 2 -1 8 7 4 ), traveler, journalist, tu to r to royal families, historian, and resourceful com piler of sta­ tistics, published a four-volum e com pen­ dium of F rench statistics, containing partial estimates of F rench national incom e and w ealth.23

3. The Period of R evived Interest, 1 8 5 0-99 D uring the regim e o f N apoleon III (1 8 4 8 -7 0 ) and especially after F ra n ce’s m ilitary debacle of 1870, during the second Republic (form ally established in 1875), the organization o f F re n ch statistics was fu rth er im proved, u nder the influence of the country’s com m ercial and industrial ad­ vances of the tim e. T his progress, in turn, w orked to benefit the preparation o f n a­ tional income estimates. P opulation censuses continued to be taken every five years. In 1876, the fam ily form was supplem ented by an individual report, and in 1881, the census was ordered to be held in all parts o f the country on the same day. T he census o f 1881 included a m ore elaborate enum era­ tion of the population by occupation than any of the form er censuses, and provided a treasurehouse of inform ation for estim ators o f income distribution. T he m ost im portant national incom e esti­ m ates of this period w ere those of de Jonnes, V acher, Cochut, de Foville, Leroy-Beaulieu, Coste, and D oum er.

a. Moreau de Jonnes In 1856 M oreau de Jonnes published an adm irable little book in w hich he p u t to ­ gether the results of the agricultural, indus­

125

trial, and population censuses prepared by him during the preceding twenty years. H e did n o t preten d to give com plete n a­ tional incom e totals. H e gave only the total incom es from agriculture, industry, and handicrafts, and then only those o f the laboring population. H e apparently consid­ ered the existing statistics on capital in­ com es m uch too unreliable for estimating purposes. In this volum e Jonnes com pared his 1850 estim ates w ith those prepared by Tolosan fo r 1789, w hom he considered to have been the ablest statistician of the Revolutionary period. So far as his own figures were con­ cerned, Jonnes explained th a t 75 p er cent o f them w ere authentic, 12 per cent were taken from the industrial census o f Paris prepared by the Paris C ham ber o f Com ­ m erce, and 12 per cent cam e from the southern prefectural reports. Some of his estim ates are shown in T able 8-2. Jonnes concluded his survey with the fol­ lowing com m entary: These figures prove the error of those who believe that an immense rise in national wealth has taken place over this period. The tremen­ dous social and technical advances of the 19th century have produced a rise in labor incomes of only one third. Each family obtains today for its labor 100 francs as against the 67 which constituted its remuneration sixty years ago. But the division of large landed estates has enabled the cultivators to acquire numerous parcels of land and to derive from them through hard and intelligent labor, a supple­ mentary remuneration to that paid them by the more opulent classes of society. We have shown that the 24 million inhabit­ ants of the rural communities consist of 5.3 million families of 4% persons each. An ex­ amination of tax records shows that one half of them possess rural property bringing in, on the average, an annual income of 105 francs and paying a real estate tax of 21 francs. This property is nothing but a cottage with a vege­ table garden, some potato, hemp, clover or buckwheat fields. . . . At the utmost, these little properties add but 200 to 300 million francs a year to the incomes of these agricul­ turalists. . . . Instead of earning merely 562 francs a year from their ordinary labor, these families now earn a combined income from such labor and from their properties of 667 francs, thus having but about 2 francs a day for their needs throughout the year. This is the whole measure of the comfort these rural dwellers have been able to achieve. To obtain it they must exploit to the utmost every small­ est bit of their land. The division of land, accomplished by the

126

THE INCOM E OF NATIONS

Table 8-2. Main Results of Moreau de JonnSs’ Estimates of the Value of National Production of France in 1850, Compared with Tolosan’s Estimate of 1789 1789 (Tolosan) (millions) (per cent) Population, by classes Total laboring other classes

17.0 9.0

(65) (35)

30.3 5.7

(84) (16)

26.0

100

36.0

100

24.0 2.5 3.8 5.7

(67) (7) (10) (16)

36.0

100

agricultural manufacturing handicrafts other classes

Francs (billions) Gross value of production agricultural manufacturing handicraft Labor income agricultural manufacturing handicraft

Wages of laboring class per family (4.5 persons) per day per year so urce:

1850 (Jonnes) (millions) (per cent)

Francs (billions)

2.0 0.9 0.9

(50) (25) (25)

8.0 4.0 4.5

(50) (25) (25)

3.8

100

16.5

100

1.0 0.5 0.5

(50) (25) (25)

3.0 0.8 0.9

(65) (19) (16)

2.0

100

4.7

100

Francs

Francs

1.48 542.0

1.93 706.00

Moreau de Jonnes, Stalistique de l}Industrie de la France, Paris, 1856,

Revolution, has incontestably produced pro­ found changes by doing away with inherited estates as large as provinces, where a single family had 100 times as much land as it could cultivate. In recompense, the new possessors of the land, who labored so long under the feudal and clerical yoke, have delivered France from the terrible famines which decimated her popu­ lation every third year—and who, above all, have furnished over the past sixty years eight millions of armed men for the defense of their country’s independence and the maintenance of her integrity.24

b. Break with Tradition: Leroy-Beaulieu, de Foville, Coste, Parliamentary Commission In 1874, M aurice Block (1 8 1 6 -1 9 0 1 ), one of the outstanding F rench political sci­ entists of the tim e, review ed the mistakes m ade by m ost of the national incom e esti­

pp.

338-44.

m ators o f the previous hundred years in a book called D ictionnaire de la Politique. All of them , he wrote, erred in taking a p a rt of the national incom e fo r the whole, namely, only net rent, or only entrepreneurial profits. H e em phasized, as J. B. Say did before him , th a t national incom e consists o f the net value of the national output— net value in the sense of th e com bined values of labor and of th e services o f capital. H e proposed th a t the best w ay to m easure national in­ com e was to take the sum of the values of all final products (goods and services). In this w ay alone, he felt, could double count­ ing and the om ission of pertinent items be avoided.25 B ut despite his obvious interest in and fam iliarity w ith the field, Block of­ fered no over-all national incom e estimate o f his own. F o r a few m ore years, the traditional physiocratic an d m aterial production ap-

127

8 . M U L TIPLICATIO N OF ESTIM ATES IN FRANCE

proaches continued to dom inate national in­ com e estimating. In 1878, D r. L. V acher prepared an estim ate of F ran ce’s national incom e and national w ealth in w hich he follow ed Lavoisier’s approach and used the physiocratic distinction betw een gross and net income. B ut in th e early 1880’s a defi­ nite break with this tradition occurred, w ith Leroy-Beaulieu and de Foville leading the way, and national incom e began to be com ­ puted as the sum of incom es earned by in­ dividuals in all occupations. In 1881, P aul Leroy-Beaulieu (1 8 4 3 1 916), em inent professor of political econ­ omy, first at the E cole Libre and later at the College de F rance, “econom ic jo u r­ nalist w ithout a p eer” as characterized by Professor Seligman, and founder o f the im ­ p o rtan t journal E conom iste Frangais, p ub­ lished an estim ate o f the national incom e of F ran ce that was constructed as an aggre­ gate of individual incom es.26 H e started w ith an estim ate of the total incom e o f the P arisian population, basing it on d ata on local assessments of dwelling rents and per­ sonal properties. H e then blew up the re­ sulting figure to a national incom e total, set­ ting it at 25 billion francs. H e also used his figures to dem onstrate the falsity of the com m on notion th a t in m odern society “the rich grow always richer and the poor be­ com e always poorer.” H e asserted th a t quite the contrary is true, “there being a definite tendency in m odern society tow ards lesser inequality in the distribution of incom e and w ealth.” 27 In 1883, A ndre Cochut, contributor on econom ic questions to the R evue des D eux M ondes, published an estim ate of F rench national incom e th a t was higher th an any previous one. H e set it at 37 billion francs fo r 1881, com puted at the current inflated prices. H is com putation was based on the occupational census statistics o f 1876 and 1881.28 A lfred de Foville (1 8 4 2 -1 9 1 3 ) was an econom ist and statistician of international repute. As chief o f the B ureau of Statistics in the M inistry of Finance, he reorganized the M inistry’s statistics. In an illum inating book published in 1887, this lucid w riter, distinguished fo r his com m on sense, esti­ m ated the national incom e of France, as a total of all individual incom es, at betw een 20 and 25 billion francs.29 D e Foville doubted the possibility of arriving at a m ore definite estim ate based on the kind o f data at hand. H e arrived at these u pper and low er lim its of th e national incom e fo r the following

reasons: (a ) the total national and local governm ental revenue in 1887 was in ex­ cess of 3.5 billion francs; ( b ) the ratio of taxes to incom e in representative families, according to Leroy-Beaulieu’s investigation, ranged from 10.8 p er cent in a Parisian w orkingm an’s fam ily to between 13 and 17 p e r cent fo r a m illionaire (depending on the com position of the m illionaire’s w ea lth ); and (c ) assum ing an average tax burden of 15 per cent, the national incom e w ould be 23.3 billion fran cs.30 D e Foville also m entioned an estimate of incom e from capital m ade in 1885 by the P arliam entary Com mission on Fiscal Re­ form , placing it at 8 billion francs.31 A dding 15 billion francs to this sum for labor in­ com es an d m ixed incomes, de Foville ar­ rived at a confirm ation of his own figure. In addition to his over-all estimate, de Foville developed a distribution of the n a­ tional incom e total by size of individual incom es, based on data on the num ber of funerals of different classes. H e chose these data on the perfectly plausible assumption th a t each incom e group h ad its own p a r­ ticular class of funeral. T h e distribution of incom e th a t he developed on this basis did n o t differ m aterially from Leroy-Beaulieu’s distribution.32 In 1890, A lfred Coste, m em ber of the Statistical Society of Paris and later, in 1895, Technical A ssistant to the Parliam en­ tary C om m ittee on a Proposed Incom e Tax, published a detailed distribution of national incom e fo r 1886 based on occupational sta­ tistics and “coefficients of productivity of agriculture and industry.” T aking de Foville’s estim ate of a 20-25-billion national income as his starting point, h e arrived at the fol­ lowing distribution, exclusive of the incomes of governm ent em ployees: 33 Income Recipients Francs (millions) (billions) 1. Workers 2. Small proprietors and others, with resources and incomes not exceed­ ing those of workers, self-employed 3. Capitalists in the proper sense of this term includ­ ing larger proprietors, industrialists, and professions Total

10.3

8.0

3.7

4.0

3.7

10.5

17.7

22.5

128

THE INCO M E OF NATIONS

Leroy-Beaulieu’s m ethod of estim ating the distribution o f incom e on the basis of local property tax data fo r the Parisian population was adopted by the M inistry of Finance and by Professor Clem ent Colson (1853-1939) as a basis fo r an estim ate of the incom e distribution fo r the w hole of F rance. In 1896 P aul D oum er, M inister o f F i­ nance, in his report on the project o f an income tax, presented a distribution o f in­ come as follows: 34 N u m b er of Incom e R ecipients P e r (thousands) C en t Below 2,500 francs F ro m 2,501 to 10,000 francs F ro m 10,000 to 100,000 francs Above 100,000 francs T o ta l

T o ta l Incom es (m illion francs)

Per C en t

9,187

86.0

12,432

56

1,303

12.2

5,244

24

184

1.7

3,778

17

546

3

3 10,677

0.03

22,000

A s can be seen from this exposition and from Table 8-3, the national incom e esti­ m ates prepared during the nineteenth cen­ tury in F rance differed widely in their con­ ceptual bases as well as in their arithm etical results. By the end of th a t century, F rench national incom e estimates possessed greater reliability than they had at any tim e before, especially since they now utilized d ata on distributed incomes that the now industrial­ ized econom y of F rance provided in sub­ stantial quantity. T able 8-3 recapitulates the various national incom e estimates of the nineteenth century reviewed above.

T ab le 8-3. E stim a te s P rep ared in F ra n c e D u rin g th e N in e te e n th C e n tu ry * N am e of E s tim a to r

Y ear of P u b li­ c atio n *

C h a p ta l M o n ta liv e t Say D u p in D u te n s S chnitzler M oreau de JonneSs V acher L eroy-B eaulieu C ochut D e Foville C oste D oum er

1819 b 1813 1820 1827 1842 1846 1856 1878 1881 1883 1887 1890 1896

N a tio n a l Incom e T o ta l (billion francs) 6.1 7.0 7.8 8.4 10.6 7.7 16.5 26.0 25.0 37.1 25.0 22.5 22.0

(p roducts) (p roducts) (products)

(products) (p roducts) (incom es) (incom es) (incom es) (incom es) (incom es)

* P o std a tin g b y one to nine years th e y e ar covered b y th e e stim ate. b E s tim a te for th e p eriod 1800-12. * L avergne a n d H e n ry {ibid., p. 76) presen ted th e follow ing series of n a tio n a l incom e e stim a te s fo r F ra n c e t h a t c o n ta in som e estim a te s t h a t we w ere u n ab le to find a n d h ave n o t included in o u r list, b u t on th e o th e r h a n d does n o t include som e c o n ta in ed in o u r list. T h e ir list follows:

F o u rn ie r de F laix Poussielgue H ip p o ly te P assy T eisserenc de B o rt R ouvier E lisee R eclus V acher T alan d ie r Vignes L eroy-B eaulieu L eT reso r de la R o q u e B onnet C hailley F o u rn ie r de F laix B allue D e Foville P e y tral C oste Com m ission de l ’im p o t su r le revenu

Y e ar

M illions

1789 1817 1849 1871 1874 1877 1878 1878 1880 1883 1883 1883 1884 1885 1886 1886 1888 1890

4,250 3,756 6,000 10 to 15,000 16,000 25,000 9,452 25,000 11,169 25 to 30,000 37,000 35,000 30,000 25,000 8,000 20 to 25,000 16,191 22,500

1895

22,500

9. FIRST DEVELOPMENTS BEYOND ENGLAND AND FRANCE T ow ard the m iddle of the nineteenth cen­ tury, the extensive w ork done over the years in national incom e estim ating in England and F rance prom pted statisticians in other countries to attem pt the preparation of like estimates for their own countries. A dvances in transportation and com m unication had brought the nations of the W estern w orld into closer contact w ith one another, m aking it possible for them to exchange their in­ tellectual and scientific heritages. T his shar­ ing of ideas included the concept o f national incom e and the m easurem ent o f its aggre­ gate. T he broadening o f com m ercial and political com petition, bringing the U nited States, G erm any, and Russia to the fore as the challengers o f the older pow ers, also tended to foster interest in the subject in o ther countries. T he nineteenth century saw the preparation of national incom e estimates in the U nited States, in A ustria-H ungary, in some of the G erm an states; tow ard the end of th at period in some o f the Swiss cantons, in G reece, in India, in the new colonies of A ustralia, and in Russia. T he end o f the century also saw the publication in E ngland of the first com prehensive international com ­ parison of national incomes.

T ucker h ad m ade a thorough study of the six decennial censuses th a t had been held betw een 1790 and 1840. A m ong the m any things he tried to discover in these censuses was the size and structure of the country’s national incom e and the changes th a t had taken place. T he 1840 census was broader in scope and included m ore pertinent eco­ nom ic data th an any previous A m erican census. Its scope had been broadened by Congress on the specific recom m endation of P resident V an Buren. D eeply concerned over the severity of the business crisis of 1837, and anxious to find the means for preventing sim ilar econom ic catastrophes, the President, in his annual message of 1838, asked w hether the scope of the census “m ight n o t be usefully extended by causing it to em brace authentic statistical returns of th e great interests specifically entrusted to or necessarily affected by the legislation o f Congress.” Congress responded favorably to the P resident’s request and provided th at the m arshals should “retu rn in statistical tables . . . all such inform ation in relation to mines, agriculture, com m erce, m anufac­ tures, an d schools, as will exhibit a full view of th e pursuits, industry, education, and re­ sources o f the country.” T h e returns w ere published in three volum es covering popula­ tion, industry and com m erce, and pension­ ers. This action by the P resident and Congress was probably the first instance of the exer­ cise o f federal pow ers to m ake possible broad econom ic inquiries into the affairs of the people of the U nited States. A s D r. Cum m ings, Statistician of the U.S. Bureau of th e Census, w rote in his historical sketch o f the Statistics of the U nited States, “this suggests th at the scope of the census may constitutionally em brace the general welfare of the com m unity and virtually removes all lim its to the range of statistical inquiries by the federal governm ent.” 1 “T he range of the inquiries at this cen­ sus,” says D r. Cum mings, “seems to have

1. D evelopm ent in the U nited States a. Tucker’s Estimates for the United States for 1840-50 In the U nited States, the first estim ate of n ational income was prepared in 1843 by G eorge T ucker (1 7 7 5 -1 8 6 1 ), professor of m oral philosophy at th e U niversity o f V ir­ ginia, form er m em ber o f the U nited States Congress, and an outstanding econom ist of the tim e, w ho w rote, am ong other works, a notable book on th e L aw s o f Wages, P ro f­ its, and Rent. A ctually, T ucker’s estim ate included only the net value of m aterial pro­ duction, which he apparently equated w ith the national incom e. Notes begin on p. 525. 129

130

THE IN CO M E OF NATIONS

greatly exceeded the range of adm inistrative capacity and control.” M any com plaints fol­ lowed concerning inaccuracies contained in the results. Congress ordered an inquiry that led to the creation in 1849 of a Bureau of Statistics w ithin the D epartm ent o f Interior and to the institution of better m ethods for collecting data in succeeding censuses. It was on the basis of the 1840 census that T ucker wrote, in 1843, his Progress o f the U nited States in Population and Wealth in F ifty Years, supplem enting it in 1855 w ith additional chapters extending his fig­ ures to cover the data of the 1850 census. A m ong the several chapters o f this book, one is devoted to the m easurem ent of the annual incom e or product, and one to the estim ation of the national wealth. U nder annual product, T ucker dealt only w ith the m aterial product. W hether he did this for theoretical reasons o r because of the lim ita­ tions of the census data is not clear. In the chapter dealing with th e “annual incom e” o r “product of the country,” T ucker pointed out some of the difficulties of preparing an estim ate of th at annual in­ come even with the aid of th e rich data m ade available by the new census. One of these difficulties, he w rote, was “those di­ versities and fluctuations of price, from w hich no country is exem pt” and w hich “are particularly great in this country,” especially since “articles of raw produce, w hich vary in price, from year to year, far m ore than m anufactures, constitute here the unusually large proportion of m ore than two thirds of the whole annual product.” M oreover, in so large a country, “differing so widely in soil, climate, density of population and easy access to m arket, the price o f the same com m odity varies considerably am ong the different states in th e same year,” and in the larger states even am ong different locali­ ties. “T he price of the m ore bulky com ­ m odities at one place of production [may be] m ore than twice as high as the price they bear at another.” T o m ake a fair aver­ age price, “it is necessary to take into ac­ count the quantities produced in the several parts, as well as the difference of price.” 2 A m ong other difficulties, T ucker m en­ tioned the lack of inform ation on the con­ tribution to the annual revenue m ade by livestock and other fixed capital, and the incom pleteness of other data on the costs of raw m aterials th at have to be deducted

from gross value in order to avoid double counting. T u ck er explained his m ethods of evalua­ tion in detail. In cases w here the census gave only quantities of articles, he took “the m ark et price at the place produced, or w here the p roducer transports it by his own lab o r,” ascertaining this price from local inform a­ tion. A lthough, technically speaking, he should have used the prices of 1840, the year o f the census, he preferred to take “a m edium between the prices of 1840 and those of 1843,” inasm uch as the prices of 1840 w ere greatly inflated by the overissue of state banknotes. In estim ating the annual p roduct of live­ stock, T u ck er assum ed it to be one fourth of its gross value. F ro m the gross value of dom estic m anufactures, w hich were included in the products o f agriculture, he deducted one h alf fo r raw m aterials. In estim ating the annual p ro d u ct o f com m erce, consisting of wages and profits, he assum ed them to be equal to 25 p er cent of the invested capital. “W ithout doubt,” he w rote, “this greatly exceeds the rate of profits in the wholesale an d foreign trade, b ut it is also far short o f the retail trade, in w hich fo r th e m ost part, the capital is tu rn ed over several times in the y ear.” H e checked his calculations against census inform ation showing that m ore th an 100,000 families w ere engaged in com m erce and declared th a t “a less profit th an th e one supposed w ould n o t be ade­ quate to th e support o f th a t num ber, in a style of living w hich far exceeds the aver­ age rate of th a t of the w hole com m unity.” F ro m the gross value o f m anufactured products, T ucker deducted one th ird fo r the value o f raw m aterials, leaving two thirds fo r wages and profits, noting th a t “these are th e average proportions in the official state­ m ents o f the m anufacturers of N ew Y o rk .” H e m ade an exception in th e case of mills, deducting three fourths instead o f one third fo r raw m aterials. H e m ade a separate esti­ m ate fo r printing and bookbinding, allowing 25 p er cent fo r invested capital and $200 fo r each employee. In estim ating th e annual products of mining, fisheries, an d forestry, T u ck er took the w hole value at the place of production, or place o f sale by the p ro ­ ducer, “the value being m ade up o f the profits of land, of labor an d o f capital.” In all his estimates, T u ck er gave the prices at w hich he estim ated the principal products of each state, “so th a t everyone

9.

131

FIRST D EV E L O PM E N T S BEYOND ENGLAND AND FRANCE

m ay correct the estim ate w herever he deems it erroneous.” Finally, he noted th a t the census om itted several products “whose ag­ gregate value w ould m ake no insignificant addition to the total am ount,” such as: (1 ) blades of Indian corn used as fodder fo r horses and cattle; (2 ) peas and beans; (3 ) flax seed; (4 ) broom corn; (5 ) sum ach; (6 ) honey; and (7 ) feathers. T u ck er built his estim ate state by state, totaling them to give the follow ing national incom e or product: 3 V alue of A n n u al P ro d u c ts (m illions) A griculture M a n u fac tu rin g Com m erce M ining F o re st Fisheries T o ta l

$

V alue per Person

6 5 4 .4 2 3 9 .8 7 9.7 4 2 .4 1 6 .8 1 2 .0

$ 3 8 .1 6 13.99 5 .7 0 2 .4 7 .98 .70

$ 1 ,0 4 5 .1

$ 6 2 .0 0

T he p er capita product varied regionally as follows: N ew E n g la n d M iddle S ta te s S o u th ern S ta te s S outhw estern S ta te s N o rth w e stern S ta te s

$84 76 52 61 41

N atio n al A verage

$62

T ucker also show ed the differences in the industrial structure of the different regions in the following tabulation:

edition o f his w ork. F o r agriculture, he found an increase of 70 per cent in the net value o f the product, o r about double the increase in population, b u t he ascribed it m ore to a rise in prices th an to a rise in quantities. F o r m ining, m anufacturing, the m echanical arts, and com m erce he found a 100 p er cent increase. This increase was m ore th e result of an increase in output (largely due to the increase in the num ber em ployed) th an of an increase in prices. F o r fisheries and forests he found a smaller rate of increase. F o r the n et value of the com bined p ro d u ct of the five branches of production— agriculture, mining, m anufac­ turing, com m erce, fisheries and forests— T ucker estim ated the increase at 74 p er cent. In other words, he estim ated the total n et value o f the annual product of 1850 at $1,986 m illion as against his 1840 estimate of $1,063 m illion, and the p er capita net value of the annual p roduct of 1850 at $87 as against the corresponding 1840 figure of $62.4 T uck er’s estimates, considered within the lim itations set by his exclusive use of m a­ terial production, w ere exceedingly well car­ ried out. T hey w ere far superior to C olqu­ h oun’s, Low e’s, o r P eb rer’s estimates in England— a fact th a t m ay be ascribed as m uch to the superiority o f A m erican statis­ tics at the tim e as to the superior com pe­ tence of T ucker as a statistician and econo­ mist. In the next several decades, A m erican statistics m ade striking progress. As G eneral F rancis A. W alker (1 8 4 0 -1 8 9 7 ), fo r years director of the census and the m ost em inent

P e r C e n t P ro p o rtio n s of A nnual P ro d u c t in V arious S ta te s

A griculture M an u fac tu rin g C om m erce M ining F o re st Fisheries

N ew E n g lan d

M iddle

S outhern

S outhw est

N o rth w e st

T o ta l

40 44 7 2 2 5

55 26 10 7 2

80 8 7 2 2

80 8 10 1 1 —

67 18 10 3 2

62 22 9 4 2 1

100

100

100

100

.3

.5

Twelve years later, using data from the 1850 census, w hich in some respects was less com plete than the 1840 census, T ucker projected his estim ate of national product fo r 1840 to the year 1850. In 1855 he p ub­ lished th e results in an appendix to a new

100

.2

100

A m erican statistician and econom ist of the tim e, p u t it in a speech before the In ter­ national Statistical Institute in 1893, “a strong passion fo r statistics early developed itself in the life of o u r people. . . . N o governm ent in the w orld has ever lavished

132

THE IN CO M E OF NATIONS

m oney and labor m ore generously upon sta­ tistical inquiry, nor has any people ever responded m ore cheerfully and patiently in this respect.” 5 But despite this continuous im provem ent in A m erican statistics, no other A m erican statistician or econom ist during this period attem pted to carry forw ard T ucker’s n a­ tional income or to prepare independent ones. W hen, finally, in the 1890’s, an at­ tem pt was m ade, it em phasized the distribu­ tion of national incom e rather than its origin. T he reasons fo r this forty-year lapse of interest are not easy to determ ine. T he slow developm ent of economics in the U nited States m ay have had som ething to do with it. M any of the statistics in the nineteenth century w ere designed to supply inform ation o f im m ediate practical value, and were collected w ithout any clear notion of their possible future use. It was fortunate, however, th a t the data were assembled, even if it was done w ithout any realization of their value. Y ears later, tow ard the tu rn of the century, they were used to great advan­ tage in a variety of im portant studies, in ­ cluding a historical series of national income estimates.

b. Spahr’s Estimates for the United States for 1890 In 1896, D r. Charles B. Spahr (1 8 6 0 1904), econom ic publicist and associate editor of the liberal m agazine O utlook, p ub ­ lished a book on the distribution of w ealth in the U nited States,6 in w hich he presented estimates o f the size and distribution of the w ealth and incom e of this country in 1890. Spahr had studied abroad and was well ac­ quainted w ith E uropean econom ic thought and literature, including the English, F rench, and G erm an national incom e esti­ m ates. H e was interested prim arily in de­ term ining w hether the inequality in the dis­ tribution of w ealth and income in the U nited States was increasing or diminishing, and if so, to w hat extent and in w hat m anner. In addition, he critically reviewed the studies of the distribution of w ealth and incom e m ade in E ngland by Baxter and Giffen, and in F rance by Leroy-Beaulieu and several other economists and statisticians. Spahr built his estimates largely on the censuses of 1880 and 1890, draw ing m ost heavily on the 1880 census, w hich contained a considerable am ount of data on the value of real and personal property in addition

to the usual data on the num ber of persons in each occupation. H e also used govern­ m ent reports on agricultural production and wages paid agricultural w orkers; the reports o f b oth the N ational and State labor bureaus on industrial production, wages, and profits; m iscellaneous official o r private studies; and his own personal observations. Spahr defined incom e in the broadest possible term s to include all incom e flowing to th e inhabitants of the country from any source. H e opened his analysis w ith the statem ent th a t “In the U nited States, de­ spite the absence of incom e-tax returns, we find perhaps the m ost com plete and satis­ factory statistics in the w orld respecting the aggregate of the national incom e.” 7 Spahr took the gross value o f agricultural production as estim ated by the D epartm ent o f A griculture for 1889, deducting from it three fourths of the reported value of corn, oats, and hay as being fed to anim als; add­ ing to it the value o f anim al products (as reported by the Senate R eport 986 for 1892), th e increase in the value of farm stock (as reported by the D epartm ent of A g ricu ltu re), and the ren tal value o f hom es occupied by farm ers. H e calculated the net value o f m anufacturing and m echanical trad e production largely from the reports o f the M assachusetts B ureau of L ab o r Sta­ tistics fo r 1890, w hich w ere based on com ­ prehensive returns of m anufacturing estab­ lishm ents on their total wages, profits, and interest. H e m ade adjustm ents fo r other states in accordance w ith the available data on existing differentials in wages between N o rth and South, E ast and W est. F o r value added in mining, Spahr used the reports of the state labor bureaus of the m ining states. F o r railroads he used the reports of the Interstate C om m erce Com m ission. F o r re­ tail stores he assum ed th e annual wages and profits per person engaged to be about equal to those in the factory industries. A s regards the earnings of professions, Spahr was guided by the official inform ation on teach­ ers and m inisters; and fo r wages of domes­ tics and unclassified laborers he used “com ­ m on observation.” T he am ount of data he had available on wages was com paratively rich. H e estim ated incom e fro m capital on th e basis o f census d ata on th e values of various types of property; estim ated rates of retu rn ; gross rents of houses in cities, th eir operating costs, and n et rents.8 S pahr’s final estim ates 9 were as show n in Table 9-1.

9.

133

FIRST D EV E L O PM E N T S BEYOND ENGLAND AND FRANCE

T ab le 9-1. S p a h r’s E stim a te of th e N a tio n a l Incom e of th e U n ite d S ta te s for 1890 N u m b e r E ngaged (thousands) A griculture M ines M a n u fa c tu re rs a n d m echanical trad e s R ailroads O thers in tra d e o r tra n s p o rta tio n T eachers M inisters P hysicians a n d law yers O th e r professions S e rv a n ts a n d laborers All o th ers U rban real e s ta te (including stores a n d factories) T o ta l

T o ta l Incom e (m illion £) 2 ,6 0 0 * 210 b 2 ,7 9 0 °

8,4 9 7 350 5 ,0 9 1 (w age-earners 4,650) 462 2,8 6 3 342 88 195 320 3,3 5 7 1,172

630 d 1 ,5 7 0 e 86f 80* 234 » 260» 670 i 470 * 1,200 ‘

22,735

10,800

a C ost of seed, fertilizers, etc., a n d expenses for stock a n d im p lem en ts offset b y fuel a n d b e tte rm e n ts ; $140 m illion included for house re n t. b W ages co m puted a t $370 (official e stim a te for coal m ines); profits tak e n a t 60 p e r c en t of aggregate w ages as in 1880. 0 W ages $360; profits a t tw o th ird s of ag g reg ate wages a s in M assach u setts. d N e t profits official, w ages, etc., sam e per c e n t a s in 1880. ° N e t p ro d u c t th e sam e p e r head a s in m anufactures. 1 W ages a t $250 (official e stim a te for public school in 1889, $230). 8 Salaries, $900 (official estim a te s for M e th o d ist C hurch). h E arnings, one th ird m ore th a n m inisters, or $1,200. ‘ E arnings a t $800. ‘ W ages a t $200. k E arn in g s a t $400. 1 6.66 per c en t on e stim a te d value.

Spahr then rearranged all these data to show the distribution of labor incom e and capital incom e by size o f fam ily income, obtaining the following results: 10

receive nearly one-fourth of the national in­ come, while fifty percent receive barely onefifth . . . the general distribution of incomes in the United States is wider and better than F a m ily Incom es

N u m b er of families A verage incom e from labor T o ta l incom e, m illions A ggregate incom e from labor, m illions A ggregate incom e from capital, m illions

U n d er $1,200

$1,200-5,000

$5,000 a n d O ver

T o ta l

1 1 ,0 0 0 ,0 0 0 $380 $4,8 0 0

1 ,3 0 0 ,0 0 0 $1,2 0 0 $2,8 9 0

200,000 $3,5 0 0 $3,1 1 0

12,5 0 0 ,0 0 0

4 ,2 0 0

1,560

700

6 ,4 6 0

600

1,330

2 ,410

4 ,3 4 0

Spahr concluded his analysis w ith the following observations: We find that more than five-sixths of the income of the wealthiest class is received by the 125,000 richest families, while less than one-half of the income of the working-classes is received by the poorest 6,500,000 families. In other words, one percent of our families

$10,800

in most of the countries of western Europe. Despite this fact, however, one-eighth of the families in America receive more than half of the aggregate income, and the richest one per cent receives a larger income than the poorest fifty percent. In fact, this small class of wealthy property owners receives from property alone as large an income as half of our people re­ ceive from property and labor.

134

TH E IN CO M E OF NATIONS

Spahr disputed Leroy-Beaulieu’s thesis th a t the tendency in m odern society is to ­ w ard less econom ic inequality. H e found, at least in all the countries he examined, including England, the U nited States, and F rance herself, that the contrary was true, the tendency being tow ards greater concen­ tration of incom e and w ealth.11 H e m ain­ tained th a t the distribution tends to be m ore o r less unequal, depending on the society’s institutional arrangem ents; and th a t these can be controlled. In the last two chapters of his book Spahr reviewed the national and local tax systems in the U nited States. Finding that the federal tax system o f the tim e bore down m ost heavily upon the poor, and th a t taken to ­ gether with other financial legislation of the national governm ent (its tariff policy, rail­ ro ad policy, currency policy, etc.) it pro ­ m oted greater concentration o f incom e and w ealth, Spahr proposed that m easures be taken to reverse the tendency. H e advocated closer governm ent regulation of m onopoly and particularly of railroads, and am ending the federal constitution to perm it Congress to levy proportional and even progressive taxes on incomes and property. “T he public welfare is the suprem e law ,” he w rote, “and the h eart and conscience of the nation are bound to give effect to m easures w hich shall m ake the w ealth o f the nation synonym ous w ith the national well-being.” 12 Spahr’s estim ate represented a great ad ­ vance over T ucker’s not only in its u nder­ lying concepts and statistical techniques but also in the fact th a t it was used in an anal­ ysis of the trends in the distribution of in­ come.

2. The First Estim ates in A ustria In 1861, K arl Czoernig (1 8 0 4 -1 8 9 7 ), em inent governm ent adm inistrator and head of A ustrian statistics, prepared an estim ate of A ustria’s national incom e for 1859 based on data collected in the 1859 census— the first tru ly com prehensive census in A ustria, w hich Czoernig him self had adm inistered. H e calculated the national incom e as the net value of the principal branches o f p ro ­ duction, setting it at 3.4 billion florins ( £ 3 3 6 m illions).13 O f this total, agricul­ ture was responsible fo r two thirds and in­ dustry and mining fo r one third. Some fif­ teen years later, N eum ann Spallart carried Czoernig’s calculations forw ard on the basis of new er census data, arriving at a figure

fo r 1868 th a t was 30 per cent higher than th a t fo r 1859, and fo r 1874 at one th a t was 80 per cent higher. In 1883, R oschm an es­ tim ated A ustria-H ungary’s national income fo r th e year 1883, an d from th a t tim e on, national incom e estim ates w ere firmly es­ tablished in th at country.

3. W hy Was G erm any a Latecom er to N ational Incom e Estim ating? In G erm any, explorations in th e theory o f national incom e w ere begun in the early 1800’s an d were continued w ith vigor for the rest of the century. N o G erm an treatise on political econom y— and there w ere m any o f them published in those years— failed to discuss national incom e and to analyze its elem ents, and m any books and articles w ere exclusively devoted to th e subject.14 D espite this w idespread interest, m ost G erm an econom ists were highly skeptical of th e pos­ sibility of developing reliable estimates of n ational incom e from the existing statistical data, and hence doubted the practical value o f any such estim ates.15 T hey were ac­ quainted w ith both past and cu rren t esti­ m ates p repared in E ngland an d France, b u t did n o t respect them highly. In 1875, A dolph W agner, th e leading G erm an econom ist of the last qu arter of the century w rote, “ [they] all . . . are extrem ely uncertain; de­ spite the great im provem ents m ade in recent tim es in statistical data and techniques, they are found to be less and less usable and are, therefore, now adays less frequently at­ tem pted.” 16 D r. K leinw aechter, w riting as late as 1896, pronounced all statistical in­ vestigations of national incom e to be w orth­ less.17 F o r a long time this negative attitude on the p a rt o f G erm an econom ists tended to discourage G erm an statisticians from m ak­ ing any national incom e estimates. It may well explain why G erm any, despite the ad­ vances m ade by h er scholars in m any branches o f political econom y and statistics, was so late in beginning the preparation of national incom e estimates. T his skepticism could n o t, how ever, deter G erm an statis­ ticians indefinitely from estim ating their co untry’s national incom e. By the middle of the century some ten G erm an states had central statistical offices headed by experts, and by 1872, seventeen states had them .18 Tw o factors were responsible fo r prom pting the statisticians in these state offices to u n ­ dertake the preparation of national income

9.

135

FIRST D E V E L O P M E N T S BEYOND ENGLAND AND FRANCE

estim ates. The first o f these was the develop­ m ent o f a vigorous socialist m ovem ent in the second quarter o f th e century. L ed by Lassalle, M arx, and others, it concentrated its attack on the inequality in the distribu­ tion of incom e in G erm any under the capi­ talistic system. T he second factor was the introduction in Prussia and som e o f the other states of proportional and progressive incom e taxes. One o f the first G erm an statisticians to attem pt a national incom e calculation was G. Riimelin (1 8 1 5 -1 8 8 8 ), one of G er­ m any’s forem ost statisticians o f the tim e. In 1863 he prepared a national incom e esti­ m ate fo r W iirttem berg on the basis o f in­ com e tax statistics. N ext, E rn st Engel (1 8 2 1 -1 8 9 6 ), A dolph Soetbeer, and A dolph Sam ter— all of them statisticians of distinc­ tion— sim ilarly p repared estim ates o f n a­ tional income for Prussia, on th e basis of poll tax and incom e tax data for 1868, 1873, and 1875. It was in 1875 th a t Engel, using his estim ates fo r Prussia, developed his fam ous law o f the differences in the distribution of personal expenditures am ong the different classes of incom e recipients. In 1889, A dolph Soetbeer (1 8 1 4 -1 8 9 2 ), author o f a fam ous book on price m ove­ ments, published a series o f estim ates fo r Prussia, covering the period 1872-1887. A portion of this series is reproduced.

end o f the century to develop a single esti­ m ate fo r the entire nation. T he first over-all estim ate, finally attem pted by R. E . M ay in 1899, was adm ittedly im perfect, as it was constructed by blowing up the several state estimates to a national total.

4. A ustralia Producing Official Estim ates: A M ajor E vent in W orld Statistical D evelopm ent In 1886-87, T im othy A. Coghlan (1 8 5 6 1926), governm ent statistician o f N ew South W ales, published a singularly com prehensive and w ell-constructed estim ate of the col­ ony’s national incom e. It appeared in the governm ent’s statistical yearbook th a t C ogh­ lan had just initiated. T he m ost startling fea­ ture of this p articular estim ate was th a t it was officially prepared and published. In the next several years, Coghlan published new estim ates in the succeeding editions o f the sam e publication, extending them in 1890 into an annual series covering all th e seven colonies o f A ustralasia. T his was th e first official series o f national incom e estimates prepared and published anyw here in the w orld— an event whose significance was ap­ preciated in oth er countries only some years later. C oghlan was an engineer by training. H e

S o etb eer’s E stim a te of N a tio n a l Incom e of Prussia

1872 1875 1878 so u rc e :

P o p u la tio n (m illions)

N u m b e r of T a x p a y ers (m illions)

T o ta l Incom e (m illion m ark s)

A verage Incom e (m arks) Per T axpayer

P e r I n h a b ita n t

2 3 .8 2 4 .5 2 5 .7

8 .1 8 .3 8 .8

7 .0 7 .6 8 .1

865 919 918

293 311 323

Paul L eroy-B eaulieu, R epartition des Richesses, P aris, 1881,

T he statistical literature on the national incom e of Prussia and other G erm an states grew by leaps and bounds during the last decades of the century, and soon the G er­ m an state estimates approached, if not sur­ passed, the British and F rench estim ates in quality.19 It should be noted, however, th at they dealt with the aggregate of private in­ com es rather than w ith national incom e proper. D espite the fact th a t the nation h ad been unified in 1871 under a single federal em ­ pire, no attem pt was m ade until the very

p.

510.

entered th e public service of his native state o f N ew South W ales in 1873, and turned to statistics accidentally w hen, thirteen years later, he was appointed to the newly created post of G overnm ent Statistician. “T he trial period o f appointm ent,” says Professor A rndt, “grew into a tw enty years’ occu­ pancy of the post, during w hich he greatly developed the statistical services of the Col­ ony and m ade him self a statistician of the fro n t ran k .” 20 In m any ways he was a pio­ neer in his field, and m ade a profound im ­ p rin t on the developm ent of statistics n ot

136

THE IN C O M E OF NATIONS

only in A ustralia, b u t elsewhere as well. In 1887 Coghlan initiated the publication of the statistical yearbook o f N ew South W ales, extending it later into a statistical yearbook fo r the “Seven Colonies o f A us­ tralia”— the predecessor o f “T he Com m on­ w ealth Y ear Book o f A ustralia” o f m odem times. H is yearbook was the m odel upon w hich m any other countries built their an­ nual statistical publications in later years. But his outstanding achievem ent was the preparation and publication, in 1886-87, of an estim ate of the national incom e o f N ew South W ales as a p a rt of the official statistics of the Colony.21 Its publication in the offi­ cial yearbook signalized the attainm ent of a new status fo r national incom e estimates— th a t of officially prepared statistics. Before this, national incom e estimates w ere con­ sidered highly speculative ventures, suitable perhaps fo r private individuals to experi­ m ent with, b u t n o t fo r governm ent, whose business is to deal only w ith objective and incontrovertible facts. Coghlan broke w ith this tradition and pointed a w ay fo r other countries to follow. It is interesting to note th a t this new p ath was broken by a new country. Coghlan conceived o f national incom e as the aggregate of incom es received by p er­ sons living or holding property in the n a­ tional jurisdiction. H e subscribed to a com ­ prehensive M arshallian concept of national incom e rath er than to a restricted Sm ithian one. H e included all form s of services, ex­ cept the services of governm ent property. M ore particularly, he included incom es from pastoral, agricultural, and m ining oc­ cupations; m anufacturing; trade and ship­ ping; other m echanical and laboring pursuits (including those of railw ay men, certain em ­ ployers, etc.); incom e from trade and capi­ tal not em ployed in the preceding cate­ gories; civil service and governm ent gen­ erally; professions; personal and domestic services and food distribution and others not already listed; wages of w om en in all occupations; house rents and rents o f busi­ ness properties not em ployed in agriculture, livestock, and mining. C oghlan’s first three estimates w ere pre­ pared by the net output m ethod on the basis of census data and supplem entary in ­ form ation obtained directly from “pastoralists,” m anufacturers, and other employers, and from his ow n intim ate acquaintance with the conditions in the colony. H e p ub ­

lished few details ab o u t his sources and m ethods, b ut he kept his w orksheets in good order, an d w hen years later these were exam ined, they w ere found to contain “in his ow n h and and in m eticulous detail, every step in the calculations fo r m ost of his esti­ m ates.” 22 In his com putations of net value o f output, he lacked d ata w ith w hich to m ake p ro p er allowance fo r depreciation, in­ surance, rent, taxes, etc.; and, as h e recog­ nized, his “net value” was fa r from being com pletely net. C oghlan’s w orksheets, according to P ro ­ fessor A rn d t, reveal how he gradually re­ fined his methodology. The first estimate for 1886 was crude indeed. Incomes from primary industry were derived from his own output estimates with some ad­ justments for costs. . . . Income from trade and commerce was obtained by adding 5 per cent of the sum of exports and imports to the published net profits of banks and financial institutions; income from manufacturing by taking 70 per cent of the aggregate value of machinery and a figure of £ 1,000 per factory to represent wages and profits; wage and pro­ fessional incomes in other industries by as­ signing figures for average annual earnings (based on wage-rate data with allowance for idle days) to broad occupational groups whose sizes were obtained from revised census data of 1881.23 H e refined his later estim ates by sub­ dividing the wage an d professional groups an d giving m ore authentic earnings d ata fo r them ; by adding oth er industry groups to the original list; and by giving m ore au­ thoritative profits data. F o r his 1891 esti­ m ates Coghlan had the use o f th e im proved 1891 Census and the first Industrial Census, w hich was also held th at year. In his 1891 estim ate, Coghlan also in tro ­ duced a breakdow n of national incom e by type of incom e o r facto r share as a parallel to its breakdow n by origin in the various industrial divisions. H e did the sam e in his estim ate fo r 1894. W hen, in 1897, N ew South W ales adopted the incom e tax, C ogh­ lan began to rely on incom e tax d ata for large portions of his analysis o f national incom e by types of income. In his 1888 estimate, C oghlan introduced an estim ate of national expenditure by types of consum ption an d an estim ate o f total private savings. In the p rep aratio n o f this breakdow n, Coghlan “relied m ainly on esti­ m ates o f average daily o r w eekly consum p­ tion p er head o r p er fam ily, w hich m ay

9.

FIRST D EV ELO PM EN T S BEYOND ENGLAND AND FRANCE

have been obtained by m ore o r less form al budget inquiries or from m ere personal in­ quiries.” H e also used a variety of other data such as receipts of theaters and other am usem ent places. H e continued these esti­ m ates of expenditures for 1890, 1894, 1897, and 1900, im proving them as he w ent along, and also analyzing the various form s of sav­ ing. C oghlan’s were the first m odern estim ates to em brace all three aspects of national in­ com e— its production, distribution, and dis­ position— and to em ploy the three estim at­ ing m ethods associated w ith national in­ com e’s three phases (see C hapter 11, Sec­ tion 2 b ). T he deficiency o f his data, says A rndt, “precluded C oghlan from present­ ing his income, output, and expenditure es­ tim ates in the m odern form , as parallel and alternative statem ents of the sam e m agni­ tude. N either could it be claim ed th a t he fully appreciated the value of this procedure as a m eans of checking the reliability of each set of accounts. In other respects, how ­ ever, Coghlan m ade bold use of his statis­ tical data to reveal econom ically significant trends and relationships.” 24 F o r each one of the seven A ustralasian colonies he specu­ lated on: (a) the ratios of the values of prim ary production (agriculture and m in­ ing) and secondary production (m anufac­ turing) to total national incom e; ( b ) the ratios of national incom e to national capi­ tal; (c) per capita incom e; ( d ) the dis­ tribution o f incom e between labor and the propertied groups; (e) trends of real income expressed in constant prices; and ( /) the international com parison fo r some of these relationships. O n the whole, C oghlan’s estim ates were far in advance of any prepared up to that tim e anywhere, and blazed new trails for future national incom e statisticians and ana­ lysts to follow.

5. D evelopm ents in O ther Countries T he adoption of the incom e tax in a few Swiss cantons during the second half of the century led, in some cases, to the prep ara­ tion of national incom e estimates. Based as they were on incom e tax data, these esti­ m ates w ere calculated as the sum of indi­ vidual incomes. T his was done, for example, in N euchatel in 1874.

137

In several countries crude estimates were m ade th at attracted little attention at the tim e. In G reece, fo r example, S. Skiadas, in a book published in 1891, m ade a rough estim ate o f th e national income of th at country.25 In India, according to D r. Rao, despite the dearth of statistics, at least five crude estimates w ere published during the period 1876-1900. T hey w ere as follows: D adabhai N aoroji, 1868; Baring and Baibour, 1882; F. G . A tkinson, 1875, 1895; L o rd C urzon, 1897-98; W illiam Digby, 1899. In m ost countries the situation was still not ripe fo r the developm ent of national in ­ com e estimates. In Italy, where, tow ard the end of the century, theoretical economics again flourished, statistics w ere backw ard. Econom ists w ere m ore interested in the dis­ tribution of w ealth th an in the distribution of incom e. Follow ing the unification of Italy, M atteo Pantaleoni (1 8 5 7 -1 9 2 4 ), the lead­ ing Italian econom ist of the tim e, whose influence on the developm ent of economics and statistics extended fa r beyond his own country, published, in the 1880’s, an esti­ m ate of Italy’s w ealth and its distribution. It was only a question of tim e before sim ilar estimates of national incom e would be de­ veloped there. In a num ber of countries, native economic literature was practically nonexistent, and econom ics was studied m ostly from English, French, and G erm an works. Interest in the concept of national incom e was stim ulated by some of these works, b u t the means with w hich to execute actual estimates for the country were lacking.

6. A bsence of N ational Income Es­ tim ates in Russia Throughout M ost of the N ineteenth Century In Russia, national incom e estimation, having m ade a prom ising start tow ard the end of the eighteenth century, ceased to exist in the nineteenth century. Interest in the subject seemed to have com pletely died out. T he available evidence discloses the appearance in the country of only one esti­ m ate during th at period and then only to ­ w ards the very end of it, in 1896. T he reasons fo r this total lack of interest and activity are n ot difficult to find. First, political conditions w ere n o t favorable to any statistical and econom ic research having

138

THE IN CO M E OF NATIONS

political implications. Statistics, as a result, becam e com pletely dom inated by the G er­ m an purely descriptive school, w hich, as previously noted (C hapter 6, Section 4 b ), em phasized the collection of facts, avoided conclusions, and considered all estimates of national incom e to be thoroughly unscien­ tific, unreliable, and a com plete w aste of tim e. This school o f statistics adm irably suited the requirem ents of a reactionary bureaucratic adm inistration w hich dom i­ nated all aspects of the country’s life, in­ cluding the universities and the A cadem y of Sciences. T he political arithm etic school of statistics never recovered from the death­ blow adm inistered to it in the 1790’s by Catherine’s reversion to reaction. I t found some refuge in the V oluntary E conom ic Society, but was able to function there only w ith respect to strictly local problem s. N o t only the estimates, but even the con­ cept of national incom e soon disappeared from the pages of Russian econom ic and statistical literature o f the nineteenth cen­ tury. D uring the beginning of that century, H einrich Storch expressed him self critically regarding the national incom e estimates pu b ­ lished in Russia a few years before, declar­ ing as H upei had a few years earlier (see C hapter 6, Section 8c) that “any com pu­ tation of the value of the total annual pro ­ duction of goods and services of a country is m uch too difficult of execution and is wholly unreliable.” 26 Similarly, in his sixvolum e treatise in political econom y p ub ­ lished in 1815,27 Storch treated the concept of national incom e at considerable length, taking issue w ith A dam Sm ith’s m aterial production concept of it (see C hapter 1, Section 7 ) , but m ade no attem pt to give national incom e a statistical expression. L ater political econom ists in Russia did not even go th a t far, generally failing even to m ention the concept of national incom e in their works. Econom ics, as a theoretical sci­ ence, ceased to exist during the period of dark reaction th a t began during the middle o f the reign of A lexander I and extended throughout the entire reign of N icholas I, thus em bracing some four and a half dec­ ades (1 8 1 0 -5 5 ). Econom ics during that period becam e dom inated by the G erm an historical school, finding an escape in the explorations of the ancient past to w hich the governm ent authorities could not object. Both Q uesnay’s physiocratic doctrine and A dam Sm ith’s econom ic liberalism were de­

clared by the governm ent to be dangerous an d harm ful theories and w ere forbidden either to be taught in the universities or to be w ritten about. W ith th e com plete breakdow n of the Rus­ sian econom y and finances during the Cri­ m ean W ar and the ascendance of A lexander II to the throne in 1855, a brief era of liber­ alism ensued. T he science of econom ics came back to life and becam e a standard subject in university curriculum s. A vigorous eco­ nom ic literature developed. T hree schools o f econom ic thought began to clam or for a prim e position— the M anchester school of protectionism , the Smith-Say school of free trade and econom ic liberalism , and in the last q u arter of th e century, the M arxian school in a disguised form . E conom ic anal­ ysis concentrated on the problem s of agri­ culture and disintegration o f th e present econom y, m oney and credit, taxation and public finance, foreign trade, grow th of in­ dustry, causes of industrial crises, and, in the underground literature, the em ergence of the labor movement. T he concept of national incom e received no attention from either the econom ic his­ torians of the period 1810-55 o r from the econom ic theoreticians an d practitioners of the period 1855-1900. Professor N . C. Bunge (1 8 2 3 -1 8 9 5 ), authority on public finance and later one o f R ussia’s outstanding m inisters of finance, w ho should have been aw are o f the im portance o f national incom e estim ates in fiscal planning, devoted one short p aragraph to the subject in his book on political econom y published in 1876. A . A. T chuprov (1 8 7 4 -1 9 2 6 ) in his even m ore p opular textbook on political economy published in 1904, did n ot m ention the subject at all. T ow ard th e last q u arter o f the century, Russian statisticians finally freed themselves once m ore of the paralyzing influence o f the G erm an Conring-A chenw all-Schlotzer sta­ tistical school only to fall largely u n d er the influence of the F rench m athem atical school w hich em phasized the theory o f probability and was highly theoretical in nature. This school had m ade startling advances during th a t century in certain fields of social and econom ic inquiry through the w orks of P ierre Sim on de L aplace (1 7 4 9 -1 8 2 7 ), A dolphe Quetelet (1 7 9 6 -1 8 7 4 ), Antoine A ugustin C ournot (1 8 0 1 -1 8 7 7 ), L eon Wal­ ras (1834—1910), and oth er brilliant schol­ ars, b ut it was not yet interested in ex­

9.

FIRST D EV ELO PM EN T S BEYOND ENGLAND AND FRANCE

plorations of the national incom e. In Russia, the leaders of this school w ere Professor Y. A. Y anson, author of the Theory o f Statistics published in 1885, and A . A. T chuprov, author of a book published under the same title in 1904, neither o f w hom con­ cerned himself w ith national incom e. Statis­ tics of the “political arithm etic” school de­ veloped in the provincial zem stvo bureaus (newly instituted local self-governing eco­ nom ic organizations), and in som e of the papers presented at th e sessions o f th e V olno E konom icheskoie O bshchestvo (V oluntary Econom ic Society), b u t th e form er w ere concerned m ainly w ith local agricultural problem s and stayed aw ay from national problem s, and the latter som ehow by-passed the subject of national incom e. Reflecting this lack o f interest in th e sub­ ject of national incom e, the great m ulti­ volume Russian Encyclopaedia o f Brockhaus and E ffron, published in St. Petersburg in 1897, did not contain a single article or part of an article on the subject of national income. Its index did n o t even contain th at w ord (which in Russian reads narodny d o khod o r nazionalny d o k h o d ). T he article on “incom e” ( d o kh o d ) did not m ention “national incom e.” Finally, tow ards the end of the century, in 1897, just about a hundred years after the appearance of the first national incom e esti­ mates described above (C h ap ter 6, sections 8 -1 0 ), another national incom e estim ate m ade its appearance in Russia. Its prep ara­ tion was largely prom pted by the publica­ tion in L ondon and w ide circulation through­ out the world o f M . G . M ulhall’s D ictionary o f Statistics in w hich com parative figures of national incom e w ere given fo r som e 22 countries including Russia and in w hich the latter country was show n to have th e lowest per capita income of all civilized countries. These figures naturally intrigued Russian economists and statisticians. U nable to verify the correctness of M ulhall’s figures for the other countries, they w ere at least able to attem pt a verification o f M ulhall’s estimates for their own country. A ccord­ ingly, one of them — V. I. Pokrovsky (1838— 1915)— m ade this attem pt. Secondly, the preparation o f a national income estimate in Russia at this tim e was stim ulated also by the sharp shift that had taken place at th a t tim e in the country’s foreign trade policy, from free trad e to strong protectionism . This change, effected

139

in 1891 in the interests of Russia’s budding m anufactures, brought on immediately, in 1892-94, a tariff w ar w ith G erm any and other developm ents th a t attracted wide at­ tention both in and outside the country.28 Pokrovsky sought to appraise the merits of this new policy. To do it com petently he needed figures on the country’s national in­ com e and expenditures and the relation of its exports and im ports thereto. In the ab­ sence of authentic estimates in the country o f its national incom e he undertook in 1896 to prepare one himself. Pokrovsky was one of the leading p ro ­ fessional statisticians of Russia at the time. A n outstanding zem stvo statistician, labor­ ing in the province of Tver, he becam e eventually the chief statistician of th e St. P etersburg m unicipality and, finally, the di­ rector of statistics of the Custom s Division o f the M inistry of Finance. H e was also C hairm an o f th e statistical section of the fam ous Volno-E konom icheskoie Obshche­ stvo (V oluntary E conom ic Society) of Rus­ sia. It was w hile occupying the statistical post in the Custom s Division th a t he wrote a book on R ussia’s foreign trad e and in one of its chapters presented an estimate o f the country’s national incom e. T he lim ited p u r­ pose o f his estim ate is best revealed by the fact th a t the chapter containing it bore the title “T he R elation of E xports and Im ports to the N ational Incom e.” 29 Pokrovsky follow ed closely M ulhall’s com prehensive production concept of the national income as well as the latter’s classi­ fication o f divisions o f econom ic activity and his n et output m ethod o f estimation. H e used, however, m ore authentic and com ­ plete statistical data th an did M ulhall and he dealt w ith a later year (1894) th an did the latter. B ut he concerned him self simi­ larly only w ith the national incom e of E uro­ pean Russia, exclusive of P oland and F in ­ land, rath er th an w ith th at of the entire Em pire. H e arrived at a national income total of 7.9 billion rubles, as com pared w ith M ulhall’s estim ate o f 6.1 billion rubles for 1889. H e explained the difference partly by variations in the prices of the grain between the tw o periods, partly by the growth in population since 1889, and partly also by the fact th a t M ulhall converted rubles into sterlings at their m etallic equivalents rather th an in th eir paper values or internal p u r­ chasing powers. T aking the population in

140

THE INCO M E OF NATIONS

1894 at 100 millions, he arrived at a per capita income for that year am ounting to 79 rubles. Using M ulhall’s national income estimates fo r 13 other countries and applying to them the published figures of exports and im ports of these countries, Pokrovsky com puted fo r each as well as fo r Russia the ratio o f the im ports and exports to the national income. H e found, accordingly, th a t Russia h ad the lowest percentage ratio of all the 14 coun­ tries (4 per cent for im ports and 8 p er cent fo r exports), w ith G erm any having 14 per cent, F rance 17 p er cent, and H olland as m uch as 66 to 86 per cent. Pokrovsky’s national incom e estim ate did not attract m uch public attention in Russia at the tim e for the reason th at it was buried in a chapter o f a book devoted to the dis­ cussion of another subject and th a t its pos­ sible application to the analysis o f m any other im portant facets of the econom ic life of the nation was not pointed o u t by its author. H ow ever, it definitely laid the ground fo r the preparation nine years later of another estim ate by another scholar, w hich was presented in a m ore impressive way and attracted w ider public attention (see C hapter 10, Section I f ) .

7. The First International Colla­

tions: L evi and M ulhall In the second half of th e nineteenth cen­ tury tw o notable English economists and statisticians, Levi and M ulhall, published com parative estimates of the national in ­ comes of several countries. These estim ates attracted wide attention am ong the profes­ sional economists of E urope and A m erica, but, unfortunately, they w ere exceedingly crude and inadequately supported by ex­ planations o f the m ethods used in their preparation. In 1860, Professor L eone Levi (1 8 2 1 -1 8 8 8 ), internationally know n econo­ m ist and statistician, author o f the fam ous H istory o f British Com m erce and Wages and Earnings o f the W orking Classes, pu b ­ lished the following estimates of national incom e fo r four countries that showed a per capita incom e for the U nited K ingdom nearly twice as great as th a t for F rance, and three tim es as high as fo r Russia and A us­ tria. Strangely enough, Russia was credited in Levi’s calculation w ith a slightly higher per capita incom e than A ustria.30

Comparative Incomes C o u n tries

Po p u latio n (m illions)

Incom e (m illion £ )

U n ited K ingdom F ra n c e R ussia A u stria

29 36 60 38

600 450 400 250

P e r H e ad £ i. 21 11 6 6

— 10 13 10

In 1884, G . M. M ulhall published a D ic­ tionary o f Statistics, w hich was quickly adopted as a standard source o f statistical inform ation by all th e libraries o f conse­ quence in th e w orld.31 M ulhall presented estim ates of the national incom e fo r some eighteen countries expressed in a single cur­ rency; an excerpt from these is reproduced here. M ost o f these figures w ere prepared by M ulhall him self fro m available statistics on the gross value of agricultural and in­ dustrial outputs and other such data. H e follow ed the com prehensive production con­ cept and derived th e net value o f output fro m gross value by applying thereto cer­ tain net-gross ratios as follows: 32 “N inety p er cent of agricultural values, 90 per cent of m ining, 60 p er cent o f m anufactures. T ran sp o rt is com puted at 10 p er cent on the gross value o f agriculture, m ining and m anu­ factures; house rent, according to th e as­ sessed valuation o r the nearest estim ate; com m erce, 10 per cent on im ports and ex­ ports; shipping, 30s. per ton yearly of carry­ ing pow er; banking 5 p er cent on banking pow er; and furtherm ore an allow ance of 10 p er cent on the total o f th e preceding eight item s, to cover the earnings of do­ m estic servants, learned professions, arm y, police, civil service, etc.” M ulhall erred in assuming he could apply to all countries the same ratios of net to gross values in differ­ ent industries and identical ratios of services to m aterial goods. D e Foville, w riting at the tim e, characterized at least h alf of Mulhall’s estim ates as being “wholly arb itrary ” and deserving little credence.33 T his con­ dem nation o f them , how ever, was fa r too sweeping. D espite th e crudeness o f his m ethods, M ulhall’s figures taken as a whole w ere probably not too fa r afield. T heir pub­ lication stim ulated a great deal o f interest in the idea of international com parison of national incom e as well as in national in­ com e estim ating itself, an d was useful to th a t extent. It will be noticed th a t M ulhall attrib u ted the highest p er cap ita incom e to

9.

141

FIRST D EV ELO PM EN T S BEYOND ENGLAND AND FRANCE

A ustralia, w ith the U nited States coining second, the U nited K ingdom third, F rance sixth, G erm any tenth, and Russia last. M u lh a ll’s E stim a te of N a tio n a l E arn in g s from V arious Sources, 1880-89 C o u n try

P e r C a p ita E arn in g s ( £ )

U n ite d K ingdom F ra n c e G e rm an y R ussia A u stria Ita ly Spain P o rtu g al Sw eden N orw ay D e n m ark H olland Belgium S w itzerland

3 3 .7 2 7 .8 2 2 .2 11.5 1 5.5 12.2 1 6.5 12.1 2 2 .0 2 0 .5 3 2 .5 2 2 .6 2 8 .0 1 9 .0

E urope

1 9 .4

U n ited S ta te s C an ad a A u stra lia A rg en tin a

3 9 .0 2 6 .0 4 0 .2 2 4 .0

A verage for 18 E u ro p ea n a n d N o n -E u ro p ea n C o u n tries

2 0 .8

8. Summary T he progress m ade in economics and statistics in the nineteenth century resulted in a considerable broadening of interest in national incom e estimates. T he num ber of countries having such estimates increased from the original three— France, England, and Russia— to eight (including Norw ay, referred to in a recent publication w ithout indication, however, of the author’s nam e o r the substance of the estim ate).84 Sufficient interest in the subject h ad de­ veloped in still other countries by th e end o f the century to indicate that, in th e first two decades o r so of th e tw entieth century, the num ber of countries undertaking the preparation of incom e estimates would grow substantially.

s o u r c e : M . G . M ulhall, D ictionary o f S ta tis­ tics, 4 th ed., L ondon, 1899, p. 320.

T ab le 9-2. C ountries in W hich N a tio n a l Incom e E stim a te s W ere D eveloped for th e F irs t T im e in the N in e te en th C e n tu ry (A dditions to E n g la n d , F rance, a n d R ussia, N os. 1-3) C o u n try 4. 5. 6. 7. 8.

U n ited S ta te s A u stria G erm any A ustralasian Colonies N orw ay

N a m e of F ir s t E stim a to r

Y e a r of A ppearance of th e F irs t E stim a te

E stim a tin g M ethod

T u ck e r Czoernig R um elin, Soetbeer, M ay C oghlan

1843 1861 1863-1899 1886 1893

N e t O u tp u t N e t O u tp u t Incom e D istrib u te d N e t O u tp u t

10. UNIVERSALIZATION IN THE TWENTIETH CENTURY; RETROSPECT OF PAST DEVELOPMENT W e shall try to describe b o th the nature an d th e rate o f these changes during each o f these periods, avoiding those details p er­ taining to the developm ent of individual country estimates w hich are covered in P arts III and IV o f this volum e.1 Tow ard th e end o f this chapter, having com pleted ou r historical account of the developm ent o f national incom e estim ates, we shall try to bring it into focus by indicating the p rin ­ cipal factors th a t seem to have influenced this developm ent from the very beginning to the present time.

In the tw entieth century, the preparation and publication of annual national incom e estimates has becom e universal. Organized everywhere as a function of governm ent, national incom e estim ating has lost the casual, exceptional nature th a t characterized it during the three hundred years th a t its existence depended on the interest of p ri­ vate scholars. T he scope of national incom e estimates has been broadened to include all three aspects of national income— its pro ­ duction, its distribution, and its expenditure — as well as the related financial transactions of the various sectors o f the economy. T he statistical data and m ethods required fo r the preparation of the estimates have been tre­ mendously extended and refined, w ith the result that they have becom e m uch m ore reliable and m uch m ore useful fo r public and private purposes. A variety o f national incom e aggregates has been developed for analytical purposes. Systems of interrelated social accounts showing the incom e and outgo transactions of the various sectors of the econom y have been introduced. T ech­ niques for the projection o f national incom e estimates into the future as a basis fo r both short-run and long-range econom ic and fis­ cal planning have been perfected. N ew m ethods fo r both historical and international com parisons of national incomes have been form ulated. In addition, estimates have lost some of the air of m ystery surrounding them in the past, and have becom e understand­ able, fo r the m ost part, even to the ordinary laym an. These revolutionary changes in the status, substance, presentation, and usage of na­ tional incom e estimates have been largely precipitated by the social, economic, and po­ litical developments th a t em anated from the two great w orld w ars of the tw entieth cen­ tury. Before W orld W ar I, the developm ent was relatively slow; after th a t w ar it becam e m ore rapid; and after W orld W ar II it p ro ­ ceeded at an astounding pace.

1. The Period 1 9 0 0 —1 9 1 7 D uring the period 1900-1917, the devel­ opm ent o f national incom e estimates p ro ­ ceeded largely along the lines laid o u t dur­ ing the closing years o f th e preceding cen­ tury. A t th e same tim e, fu rth er progress in statistics m ade possible the introduction of im p o rtan t refinem ents in the construction of estim ates during this period. M oreover, the n um ber of countries having systematic national incom e estim ates was increased from eight (see C hapter 9, Section 8) to thirteen during these years (see Section e, b elo w ).

a. England, France, United States: Bowley, Stamp, Colson, and King In E ngland, Professor A . L . Bowley car­ ried fo rw ard the n ational incom e estimates th a t he h ad begun tow ard th e end o f the century, publishing new estim ates in 1904, 1907, and 1913. H e based these on income tax records, the 1901 and 1911 population censuses, the incom plete 1906 census of wages, the 1907 production census, and on inform ation on em ploym ent and wages pub­ lished by the L abor D epartm ent. In his 1904 book,2 Bowley indicated th a t the national incom e o f the U nited K ingdom h ad in­ creased from £ 1 . 4 billions in 1882 to £ 1 .6 billions in 1891 and to £ 2 billions in 1903,

N o te s begin o n p. 526. 142

10.

UNIVERSALIZATION IN THE TW E N T IE T H CENTURY

o r 38 p er cent over the w hole twenty-oneyear period as against a population increase of only 15 per cent; and th at the greatest increase— one o f 50 p er cent— had taken place in the incom e from wages. In this sam e book he m ade a plea, sim ilar to the one m ade by C olquhoun a hundred years before (C hapter 7, Section l e ) , th a t gov­ ernm ent should undertake the preparation o f national incom e estim ates. Bowley w rote: s T h ere is n o obvious reaso n w h y th e C o m ­ m issioners o f th e In la n d R ev en u e sh o u ld n o t be req u ested to p re p a re an estim ate o f th e change o f n a tio n a l incom e du rin g re ce n t years; basing it p a rtly o n th e g re at m ass o f m ate ria l w hich th ey h ave p u blished , b u t w hich th e o u t­ side p u b lic c a n n o t in te rp re t w ith such cer­ ta in ty a n d fulness o f m ean in g as th o se w ho a re beh in d the scenes; a n d p a rtly o n b o ld esti­ m ates w hich in th e ir h an d s w o u ld be fa r b e tte r th a n th e guesses w hich o th e rs a re obliged to m ake, supplem enting these by special inquiries in those directions w hich seem likely to give results. A s it is, w e are co m p letely in th e dark as reg ard s som e o f th e m o st essential d a ta fo r estim ating the incom e o f th e nation.

In subsequent papers, Professor Bowley grew even m ore critical of the existing in­ come tax statistics and advocated the crea­ tion of a central statistical departm ent that would n o t only co-ordinate all statistics es­ sential fo r the preparation o f national in­ com e estimates, but w ould prepare the esti­ m ates them selves.4 It took another thirty or m ore years, how ever, to bring this about (see Section 3a, below ). In 1914, Sir L. G . Chiozza M oney pu b ­ lished a set of estim ates covering the years 1904, 1907, and 1914 th a t differed som e­ w hat from Bowley’s estim ates in the con­ cepts and statistics em ployed.5 In 1916, Sir Josiah C. Stamp, a prom inent businessm an and econom ic analyst who had risen to in­ ternational em inence, published a com pre­ hensive book reviewing the existing official statistical sources from the point o f view of th eir suitability for national incom e esti­ m ation.6 In France, A. de Lavergne and P aul H enry, Em ile Levasseur, and C. Colson pub­ lished estimates based on th e income-distributed approach th at represented a consid­ erable im provem ent over the w ork of LeroyBeaulieu, de Foville, and other estim ators of the last quarter of the preceding century. Poincare and Caillaux, m inisters of finance, and Rene Renoult, rapporteur o f the p a r­

143

liam entary finance com m ittee in 1906 and 1907, incorporated in their reports on p ro ­ posed bills fo r the establishm ent of an in ­ com e tax, estimates of the total private incom es of the F ren ch population. O f all these estimates, probably the m ost authori­ tative w ere those prepared by Professor Colson for the years 1899-1901 and 1913. In these estimates, Colson set the national incom es fo r those years at 26.2 and 36.1 billion fran cs.7 In th e U nited States, W. I. King pub­ lished, in 1915, a survey of the distribution o f w ealth and incom e in the U nited States w hich was, in a way, an extension of Spahr’s study m ade at the close of the last century. K ing defined national incom e in com pre­ hensive term s as the aggregate net income accruing to individuals and businesses from personal services, investm ents, and other business transactions. H e developed his es­ tim ate m ainly from the decennial censuses and analyzed his totals both by the indus­ trial divisions in w hich incom es originated and by facto r shares. K ing’s calculations showed th a t per capita incom e had in­ creased from $95 in 1850 to $332 in 1910; th a t the average size of fam ily dropped from 5.6 to 4.5 persons during the same period; and th a t the average incom e per fam ily in­ creased from $535 to $1,494. In current prices the increase in p er fam ily income was slightly less th an threefold; in constant prices it was slightly m ore th an threefold.8 K ing also showed th a t the share o f wages and salaries in total incom e increased from 36 p er cent to 47 p er cent during th e same sixty-year period. K ing concluded his estim ate w ith a set of generalizations directly contrary to those advanced by Spahr, and derived m ore from his conservative philosophy th an from the figures themselves. H e found no fault with the existing social and econom ic system in the U nited States, an d w ent on record against m ost of the currently dem anded so­ cial reform s, including the dem ands of o r­ ganized lab o r for a “living w age” in which he detected a d rift tow ard socialism. The best w ay to insure A m erican prosperity and particularly the prosperity of the A m erican w age earner, he w rote, was to teach the laborer to practice birth control, and fo r the governm ent to restrict im m igration. K ing’s was one of the last national in­ com e estimates in any country to com bine the estim ating task w ith economic politics. N early all subsequent national income esti­

144

THE IN CO M E OF NATIONS

m ators lim ited themselves to the preparation of estimates, leaving the task of social, eco­ nomic, and political interpretation of their figures to other individuals.

tent by the fact that we work with very much greater intensity in Germany than they do in France. France is the land of . . . [bond­ holders], Germany the land of labour.12

b. German Estimates Coming to the Fore: May and Helfferich

It is possible that H elfferich was correct in his contention th a t the per capita incom e o f G erm any in 1912 was higher than F ran ce’s. If so, G erm an y ’s outstripping of F ran ce m ust have taken place during the previous tw enty to thirty years, w hen G er­ m any was developing h er industry at an unquestionably m uch faster rate than F rance. H elfferich also com pared G er­ m any’s national incom e w ith E ngland’s. T aking Chiozza M oney’s national income estim ate of £ 1 .7 billions (35 billion m a rk s), he arrived at the conclusion th at E n gland’s per capita incom e was 815 m arks as com pared with G erm any’s 555 m arks, indicating at the same tim e th at the m argin betw een the two countries had narrow ed. Finally, H elfferich developed a com para­ tive distribution of the G erm an national in­ com e in 1912 with its distribution in 1896, concluding th at:

F o r a tim e, national incom e estim ates in G erm any continued to be confined m ainly to m em ber states that had an incom e tax, and to very crude projections of these state­ wide figures to a national total. T he state estimates w ere rapidly improving, and their results were used extensively in curren t fis­ cal and econom ic literature. T he older economists, who had no faith in national incom e estimates, either changed their opin­ ions, as did A dolph W agner, or w ere re ­ placed in tim e by younger economists who appreciated m ore fully the potentialities of this kind of statistical exploration.9 R. E. M ay continued the projections of state esti­ m ates into national totals that he had begun in 1899, publishing estimates in 1903 and in 1909.10 T he first truly com prehensive estim ate fo r the whole of G erm any was prepared by the noted economist, Helfferich, in 1913. K arl Helfferich (1 8 7 2 -1 9 2 4 ) was an economist, banker, and statesm an. A t one tim e p ro ­ fessor at the U niversity of Berlin and author of notable w orks on m onetary theory and practice, he becam e director of the great D eutsche Bank, and later, during W orld W ar I, M inister of the Interior and ViceChancellor of the Reich. In 1913, in a book surveying the progress of the G erm an econ­ om y under W ilhelm II, he presented an estimate of the grow th of the G erm an n a­ tional income from 1896 to 1912. H e based his estim ate on the incom e tax records of the several states th a t am ong them covered three fifths of the population of the em pire. Helfferich estim ated the national incom e fo r 1896 at 21.5 billion m arks and fo r 1912 at 40 billion m arks, indicating an 80 p er cent increase in it over the sixteen-year period, or a 45 per cent increase in the per capita income during th at tim e.11 H e concluded th at: . . . France is considerably behind Germany in national income. . . . If we take the year 1908 as normal, the average income of the German people for that year would be 555 marks as compared with 514 marks for the French people. . . . As to the French people being superior to us in wealth per head this is undoubtedly counterbalanced to a great ex­

The total result was . . . a general shifting of incomes upwards; and this movement was especially heavy from the tax-free class into the income-groups from 900 to 6,000 marks. The plutocratic development, so often as­ serted, therefore does not exist. . . . That the large increase of incomes precisely in the lower tax-groups was not due to more rigorous assess­ ment, but was a real fact, can be proved espe­ cially by the development of wages.13 H e cited the fact th a t the average net wages o f m iners, after deducting their social security contributions, increased from 1888 to 1912 in the D ortm und district from 863 to 1,586 m arks and in U pper Silesia from 516 to 1,053 m arks. H e also contended that the rise in wages of the G erm an m iner dur­ ing the previous decade or so was m uch m ore rap id than the rise in wages of the British m iner. W hereas in 1900 the British m iner received 278 m arks m ore th an the G erm an, in 1912, the G erm an m iner re­ ceived 148 m arks m ore than the B ritish.14

c. Influence of Austro-Hungarian Estimates: von Fellner In A ustria-H ungary, Professor F riedrich von F ellner was the acknow ledged leader n ot only in the p reparation o f actual n a­ tional incom e estimates b ut also in the de­ velopm ent of estim ating m ethodology. H e published a series of national incom e esti­

1 0 . UNIVERSALIZATION IN THE TW E N T IE T H CENTURY

m ates during this period th a t covered the years 1900 and 1911-13 and attracted wide attention am ong continental statisticians.15 H is vigorous cham pionship of national in­ com e estimates at the meetings of the In ­ ternational Statistical Institute and at other statistical gatherings helped to stim ulate pro ­ fessional interest in the subject throughout E urope.

d. Australia, New Zealand, and Norway: Coghlan and Others T im othy A . Coghlan, G overnm ent Statis­ tician of N ew South W ales until his resig­ nation in 1904, prepared and published n a­ tional incom e estim ates first fo r th e seven colonies of A ustralasia and n ext fo r A us­ tralia and N ew Z ealand separately.16 A fter 1904 the A ustralian governm ent suspended the preparation and publication of such esti­ m ates, except fo r 1914-15 w hen it did p ub­ lish an estim ate based on the W ar Census o f individual incomes. This estim ate was generally believed to have been low because o f deliberate understatem ents of their in­ com es by m any individuals at th a t tim e. (F o r a discussion of the earlier A ustralasian estimates, see C hapter 9, Section 4, above.) In N orw ay, various private estim ates were published during the first quarter o f the century fo r the years 1898, 1906, and 1913— 24. T hey w ere m ainly based on incom e tax statistics.17

e. Five New Country Estimates D uring the period 1900-1917, systematic national incom e estim ates w ere developed in live m ore countries— Japan, Switzerland, the N etherlands, Italy, and Bulgaria— and an estim ate was prepared fo r Spain by a Frenchm an. In Japan, K. N ak am u ra p re­ p ared in 1902 an estim ate by the net output m ethod. In Switzerland, two economists, G eering and H otz, published an estim ate in a book on the Swiss econom y th a t cov­ ered the years 1895 and 1899, extending it in later editions to the years 1905 and 1913.18 In the N etherlands in 1910, P rofessor W. A. Bonger presented a national incom e estim ate for 1908. H e follow ed it w ith others during the next tw enty years until, in 1931, the C entral Statistical Office of H olland took over the preparation o f such estimates. Bonger’s estimates w ere n o t the first in the N etherlands, but we have no inform a­ tion about the earlier ones (see C hapter 22,

145

below ). In Italy, M. Santoro published an estim ate in 1911 covering the years 18601910. H is estim ate was m uch quoted at the tim e, b ut was later criticized for u nder­ stating the grow th of the Italian econ­ om y.19 In Bulgaria, Professor K iril Popoff, later head of his cou n try ’s Statistical Office, w as responsible fo r the first estimate of that country. Popoff’s estim ate covered the years 1896 and 1911.20 Spain was still w ithout an estim ate of h er own. B ut in 1917, A ndre B arthe, a F rench statistician, in a study of the national w ealth of Spain, incorporated an estim ate of th a t co untry’s national in­ com e.21

f. National Income Estimates Coming into Prominence in Russia In Russia, fo r the first tim e in m ore than a century, national incom e estim ating came into considerable prom inence during the period 1900-1918. This happened, first of all, during the Revolution of 1905-6 when the discussion of the country’s political, so­ cial, and econom ic ills and of the best means to cure them burst into the open; and sec­ ondly, during W orld W ar I w hen the coun­ try ’s econom y was subjected to terrific strains and w hen m easures h ad to be de­ vised, w ith appropriate inform ation in hand, fo r the m ore effective m obilization of the country’s fiscal and econom ic resources. T hree com prehensive estim ates were pre­ pared: a private one, in 1906, by S. N . P rokopovitch (1 8 7 1 -1 9 5 5 ); and two semi­ official ones during W orld W ar I, also by him , bu t w ith the assistance of several other statisticians. Prokopovitch was a leading Russian statistician and econom ist of the tim e and active m em ber of the fam ous V ol­ untary Econom ic Society frequently re­ ferred to above. H e was a M arxist of the so-called “revisionist” o r reform ist school. L ater, after the Com m unist Revolution of O ctober 1917, he em igrated to W estern E urope and conducted there over the suc­ ceeding m ore th an thirty years a series of critical studies of the operations of the So­ viet econom y, publishing them in G erm an, F rench, English, and other languages. P rokopovitch presented his estimate in 1906 before a session of the V oluntary E co­ nom ic Society.22 H e was greatly influenced in its preparation by the estimates published by M ulhall and Pokrovsky ten or m ore years before (see C hapter 9, sections 6 - 7 ), except that, being a M arxist, he believed national

146

THE INCOM E OF NATIONS

incom e to consist only of the fruits of m a­ terial production. Following their example, he concerned him self with the national in­ com e of E uropean Russia only, exclusive of P oland and Finland, and he used the net output m ethod of estim ation. H e dealt w ith the year 1900, thus providing a historical com parison w ith M ulhall’s calculations for 1889 and 1894 and w ith Pokrovsky’s for 1894. In his estimates, Prokopovitch used a variety of both prim ary and secondary sources, filling gaps w ith estimates based on sample data and on supposedly com parable situations. H e arrived at a national incom e total of 6.1 billion rubles and at a p er capita incom e of 63 rubles. H is estimate was m ore conservative than those of M ulhall and Pokrovsky. It is also interesting to note th at his per capita incom e figure of 63 rubles was six tim es the per capita incom e (10 rubles) estim ated by H erm ann and Radishchev m ore th an a century before (see C hapter 6, sections 8 and 9 ). H ow m uch of this six-fold increase (assum ing the cal­ culations to have been correct) was due to a rise in prices and how m uch to a rise in productivity over the century is difficult to say. Prokopovitch did not m ention H er­ m ann’s and Radishchev’s estimates. It is not even certain th a t he knew o f their existence or attributed any im portance to them. Using M ulhall’s estimates for the other countries for 1894 (given in the latter’s Industries and W ealth o f Nations, London, 1896), and converting them into rubles, Prokopovitch concluded that Russia was “the poorest of all civilized nations” and that “a Bulgarian and a Serb had one-and-ahalf times as m uch incom e as a Russian, a G erm an— three times, an Englishm an, fourand-a-half times, and an A ustralian— nearly six tim es.” Before 1905-6 such reflections on the country’s poverty w ould have scarcely passed the censor. But now, with a Revolu­ tion raging and w ith a sem i-parliam entary regim e being inaugurated, they could be published. Prokopovitch’s estim ate o f 1906 proved of greatest usefulness a decade later, during W orld W ar I w hen some of the leading statesm en and scholars of Russia had finally com e to realize that the financing o f a m od­ ern w ar was a m uch m ore form idable u n ­ dertaking and required m uch m ore system­ atic know ledge of the country’s resources, than they had ever suspected. T he tendency in Russia before the w ar was to minimize

the strains w hich a w orld w ar m ight impose on the cou n try ’s econom y. A n im portant segm ent of the Russian econom ic opinion (represented by so em inent an authority as P rofessor M. I. T ugan-B aranovsky) unw it­ tingly contributed to this optim ism by hold­ ing th at a simple agricultural econom y, such as Russia, could sustain the stresses of a great w ar better th an a m ore com plex and industrially advanced econom y. T he experi­ ence of the first year and a h alf of the w ar com pletely refuted this doctrine. It revealed R ussia’s econom y being far m ore severely strained by th e w ar th an w ere any of the others. In fact, her situation appeared wellnigh desperate. Some of the m ore progressive officials of the M inistry of F inance and leaders o f the D um a (the N ational A ssem bly) becam e particularly alarm ed in the sum m er of 1915 over the widening gap betw een governm ent expenditures and governm ent revenues. The form er had risen from som e 3 billion rubles in 1913-14, to an annual rate o f approxi­ m ately 12 billion rubles by th a t sum m er. T he revenues, on the other hand, h ad re­ m ained at about the same level o f 3 billion rubles. T o reduce this inflationary gap, these individuals proposed the im position of a com prehensive personal incom e tax. In A u ­ gust 1915 they introduced in the D u m a a bill providing for such a tax. Im m ediately a sharp debate developed in th a t body over the necessity of such a wholly new and radical im position. In estim ating the pos­ sible yields of the incom e tax, its advocates tu rn ed to the country’s statisticians fo r esti­ m ates of the total incom e of the nation and of its distribution am ong the various eco­ nom ic groups. B ut the statisticians had readily at h an d only P rokopovitch’s esti­ m ate fo r 1900. All th a t they could do was to project this estim ate to the year 1913 and n ex t to th e w ar years w ith the aid of vari­ ous indices o f population growth, changes in productivity, and price m ovem ents. They also h ad to adjust P rokopovitch’s estim ate to the vaster territory of the Russian E m ­ pire. By this process they arrived at a n a­ tional incom e total for the w hole Em pire fo r 1915 of 15 to 20 billion rubles. Using these estimates, they concluded, rath er opti­ mistically, th a t their incom e tax could p ro ­ duce 1 billion rubles of additional revenue fo r the treasury, bringing the total revenue o f th e latter to 4 billion rubles, or to about 2 0 T o 25 p er cent o f th e national income.

10.

147

UNIVERSALIZATION IN THE T W E N TIETH CENTURY

T hey thought th a t a tax burden of that p ro ­ p ortion would be bearable.23 T h e discussion of the proposed new tax was soon stim ulated considerably by the appearance in the L ondon E conom ist of a com prehensive survey of the w ar costs and w ar finance in all o f the belligerent coun­ tries including Russia, and by the publica­ tion of a Russian translation of th a t article in a current issue of the V estnik Finansov, the official journal of the Russian M inistry of F in an ce.24 In this survey, R ussia’s n a­ tional incom e was estim ated at £ 1.5 billion and at a per capita of £ 8.8. T aking the exchange rate of a pound at 10 rubles, the translators converted the foregoing figures into a 15 billion ruble national incom e and an 88 ruble per capita incom e. R ussia’s w ar­ tim e expenditures were show n to be running in excess of her national incom e by a wide m argin and draw ing heavily on her accum u­ lated capital. T he article in The E conom ist attracted considerable attention also in other belligerent countries. Inasm uch as it rep re­ sented the m ost com prehensive analysis of the effects o f a w ar on the incom e and capital of belligerent countries ever m ade since the tim e th a t G regory K ing prepared his fam ous table in the 1690’s (see C hapter 2, Section 2 e ), we are reproducing here the m ain table of this article in full.

cal year 1915-16, and the latter produced only 55 m illion rubles in the fiscal year 1916-17. W ith the overthrow of the T zarist gov­ ernm ent in F ebruary 1917 and the estab­ lishm ent o f a Republican governm ent under K erensky, the attem pts at a comprehensive fiscal reform w ere renewed, b ut the results w ere even less substantial than before, inas­ m uch as the situation of the country had in the m eanw hile fu rth er deteriorated. In con­ sidering these new m easures, however, the governm ent m ade greater use of the serv­ ices o f econom ists and statisticians and u n ­ dertook a wholly new and authentic evalua­ tion o f the country’s national income. The w ork was organized in the newly created N ational E conom ic Council and its direc­ tion was entrusted to P rofessor P rokopo­ vitch. It was com pleted w ith extraordinary speed and published before the year was over. T hen a second, even m ore com pre­ hensive survey of the same subject was u n ­ dertaken in the same agency and u nder the sam e direction. It was published in 1918 after th e N ational Econom ic Council was disbanded.25 In both these studies, the original esti­ m ate of P rokopovitch fo r 1900 was recal­ culated; a wholly original com prehensive estim ate fo r 1913 was developed; and the

T ab le 10-1. T h e C osts of th e W a r to th e E u ro p ea n N atio n s a n d Its F in an cial E ffects D u rin g the Y e ar 1915 N a tio n a l Incom e (£ )

U .K . F ran ce R ussia Ita ly Belgium a n d Serbia

N a tio n a l Savings (£ ) ----------------------------M illion P er C a p ita

P e r C en t of -N a tio n a l M illion P er C a p ita Incom e W ar C osts (£ )

M illion

P er C a p ita

2,2 5 0 1,500 1,500 850 350

4 9 .0 37 .5 8 .8 2 4 .3 2 6 .9

375 250 250 125 60

8 .2 6 .2 1 ,.5 3..6 4,.6

1,590 1,915 1,695 410 265

3 4 .6 4 7 .9 10.0 11.7 2 0 .4

71 128 13 48 76

All allied countries

6,4 5 0

2 1 .2

1,060

3. 5

5,975

19.7

93

G erm any A u stria -H u n g ary T u rk e y a n d B ulgaria

2,1 0 0 1,200 200

3 0 .9 2 4 .0 8 .0

350 200 30

5,.1 4..1 1 ,.2

2,545 1,435 210

3 7 .4 2 8.7 8 .4

121 120 105

All belligerents

9 ,9 5 0

2 2 .3

1,640

3..7

10,165

22.7

102

so urce:

The Econom ist, L ondon, D ecem ber 18, 19IS, Special Supplem ent.

D uring the second half of 1916, the in­ com e tax and an excess property tax w ere im posed in Russia but their fiscal effects w ere negligible. T he form er was budgeted fo r only some 130 m illion rubles fo r the fis-

latter estim ate was, in turn, projected into the w ar years, 1914-17. T he m aterial pro­ duction concept was adhered to throughout. T he figure fo r 1900 fo r E uropean Russia was raised slightly from 6.1 billion rubles to

THE IN CO M E OF NATIONS

148

6.6 billions, and the per capita income, from 63 rubles to 67.25 rubles. T he correspond­ ing figures fo r 1913 w ere given at 11.8 bil­ lion rubles for the national incom e total and at 101.35 rubles fo r the p er capita income. In his second estim ate, Prokopovitch re ­ duced the foregoing per capita figure to 93 rubles. T he industrial com position o f the national income during the years 1900 and 1913 was show n to have been as follows: T ab le 10-2. N a tio n a l Incom e (from M a te ria l P ro d u ctio n ) of SO Provinces of E u ro p ean R ussia, 1900 a n d 1913 A t C u rre n t Prices (m illion rubles)

A griculture F o re stry a n d Fishing M a n u fac tu rin g T ra n sp o rta tio n C o nstruction T rad e T o ta l

1900

1913

2,985 626 1,402 531 473 562

5 ,6 3 0 730 2 ,5 6 7 1,055 843 981

6 ,579

11,806

s o u r c e : S. N . Prokopovitch, Voina i N arodnoie Khosiaistvo, 2d enlarged edition, M oscow , 1918, p. 83.

N ational incom e in current prices was shown to have increased 79.4 per cent dur­ ing the period 1900-1913. Of this increase 19.1 per cent was attributed to the popu­ lation growth, 28.7 per cent to a rise in prices, and only 17.1 per cent to increase in productivity. T he estim ate also provided some inform ation on the distribution o f the national incom e in 1913, attributing 73.8 per cent of the incom e to the laboring

ernm ent. Inasm uch as the costs of govern­ m ent services were n o t included in the national incom e, the last m entioned share m ust have represented governm ent profits from the landed dom ain and from other gov­ ernm ent enterprises. F o r all of Russia, including the A siatic possessions, the n a­ tional incom e for 1913 was estim ated at 16.4 billion rubles. F ro m the point of view of the problem s of the m anagem ent of the w ar, the m ost im portant parts of the 1917-18 estim ates w ere those w hich dealt w ith the national in ­ com e and national expenditures of the years 1914-17. T hey revealed the tru ly cata­ strophic effects of the w ar on th e national econom y. Inclusive o f dam ages inflicted on the national econom y by the enem y’s occu­ pation of the Russian territory, national in­ com e com puted at constant (1 9 1 3 -1 4 ) prices w as shown to have decreased during the first three years of the w ar from 16.4 billion rubles to 12.2 billion rubles. Exclusive of the dam ages, national incom e at constant prices was shown to have decreased from 16.4 to 14.4 billion during th a t period. T he consum ption expenditures of th e civilian population, likewise com puted at constant prices, were shown to have dropped even m ore sharply from approxim ately 13 billion rubles to 7.3 billions, o r alm ost by a half. M ilitary expenditures by 1917 w ere esti­ m ated to have absorbed 49 p er cent of the n ational incom e. T he com bined civilian con­ sum ption and m ilitary expenditures w ere thus show n to have been running in 1917 some 2 billion rubles in excess of the n a­ tional incom e, thus reducing th e national capital by th a t am ount during th a t year alone.

T ab le 10-3. N a tio n a l Incom e a n d E x p en d itu re of R ussia, 1913-14 T h ro u g h 1916-17 A t C o n sta n t 1913-14 Prices (billion rubles)

so urce:

Y ear

N atio n al Incom e

C onsum ption of th e P o p u latio n

M ilita ry E xpenses

P e r C e n t R a tio of M ilita ry E xpenses to N a tio n a l Incom e

1913-14 1914-15 1915-16 1916-17

1 6.4 1 5.8 1 3.9 12.2

11.5 8 .7 7 .3

4 .3 5 .7 7.1

27 34 49

S. N . Prokopovitch, Voina i Narodnoie Khosiaistvo, M oscow , 1918, 2d e nlarged e dition, p. 174.

classes— peasants, industrial w orkers, and others— 22.4 per cent to the propertied classes, and only about 4 per cent to gov-

A t cu rren t prices national incom e was estim ated to have increased from 16.4 billions to approxim ately 35 billions, the aver-

10.

UNIVERSALIZATION IN THE TW E N T IE T H CENTURY

age price increase being estim ated at 116 per cent for the period. Productivity was estim ated to have decreased (during the pe­ riod 1913-14 to 1916—17) 10 per cent in agriculture, 50 per cent in industry, and 20 per cent for all the branches o f national production taken together. It is seen from this description th a t n a­ tional income estimates prepared in Russia in 1917 and 1918 were characterized by extraordinary com prehensiveness and pene­ tration. But it took a national catastrophe to achieve this unprecedented statistical ad­ vance. These estim ates w ere never translated into oth er languages, and they rem ained little know n to the statisticians of the w orld. Y et they represent a classic exam ple of na­ tional incom e analysis in a country in a state of alm ost com plete econom ic collapse.

2. The Extraordinary Flourishing of Estimates During the Period 1 9 1 8 -3 9 D uring the years intervening betw een the term ination of W orld W ar I and the out­ break o f W orld W ar II, the grow th o f n a­ tional incom e estimates was extraordinary. T h e w idespread interest in national incom e estimates had been stim ulated by (a ) the colossal problem s of econom ic reconstruc­ tion facing the nations vitally affected by the w ar; ( b ) the equally colossal shifts that had taken place in the relative econom ic pow er of the different nations as a result of the w ar; (c) the staggering problem s cre­ ated by th e world-w ide econom ic depression of the 1930’s; and (d ) the need fo r m obili­ zation of econom ic resources to m eet the threats of another w orld w ar. N ational incom e estim ates of this period were designed to reveal the changes th a t had taken place in the structure of th e n a­ tional economy as a consequence of W orld W ar I, o r that were taking place currently under the im pact of new technological changes and world-w ide social, economic, and political developm ents. T he issue of the distribution of national incom e h ad tem po­ rarily, at least, receded into the background, supplanted by the fa r m ore im portant issue of national econom ic survival. Accordingly, national incom e estim ates o f this period were concerned m uch m ore w ith m easuring the n atio n’s current econom ic achievements and its econom ic potential fo r both peace and w ar, than w ith determ ining the social

149

justice of the existing distribution of income. T he num ber of countries fo r w hich sys­ tem atic national incom e estimates were de­ veloped m ore th an doubled during the pe­ riod intervening betw een the two wars, in­ creasing from thirteen in 1918 to about thirty-three in 1939. W estern E urope con­ tributed m ost o f the additions to this list, being represented by Belgium, Spain, Czech­ oslovakia, Yugoslavia, D enm ark, Latvia, Poland, Ireland, Finland, and Sweden, each o f w hich took up national income estim at­ ing during this period. A m ong countries from oth er parts of the w orld in w hich n a­ tional incom e estimates were prepared dur­ ing this tim e w ere G reece, Egypt, C anada, U nion of South A frica, N ew Zealand, T urkey, Chile, M exico, British India, and Peru. a. P riva te In stitu tio n a l R ese a rc h S u p p le­ m e n tin g In d iv id u a l E ffo rt M ost of the estimates during this period continued to be prepared by individual scholars who w ere interested in the subject, unafraid of its complexities, and willing to contribute of th eir tim e and effort beyond any recom pense.26 In several countries, how ­ ever, private efforts took on a new, collec­ tive form , as institutes fo r social and eco­ nom ic studies established w ith the aid of philanthropic funds, either at the universi­ ties or independent of them , undertook re ­ search in the national incom e field. By p ro ­ viding scholars w ith financial resources for their w ork and w ith needed technical as­ sistance, these institutes eased the scholars’ task and m ade possible the preparation and publication of m uch m ore extensive, re­ liable, and m eaningful estimates than would have otherw ise resulted. In the U nited States, the Brookings Insti­ tution and especially the N ational Bureau of E conom ic Research, b oth of them founded shortly after W orld W ar I, began a series o f studies on national incom e; 27 an d the N ational Industrial C onference B oard p re­ pared its own national incom e estimates even after the governm ent began to issue official figures. In C anada, R. H . Coats, the D om inion Statistician, published in 1919 his own u n ­ official estimates. These were followed in 1925 by the inauguration of official annual series by the D om inion Bureau of Statistics. In A ustralia, in 1926, J. T. Sutcliffe pub­ lished a fairly com prehensive estimate for the fiscal y ear 1920-21 th a t he prepared

150

THE INCOM E OF NATIONS

by the net output and incom e-distributed m ethods. By extrapolating the results fo r the year, he obtained a continuous series for the net value of output from 1910-11 to 1924-25 and one for incom e distributed from 1916-17 to 1921-22. H e also esti­ m ated savings, consum ption, and the dis­ tribution of incom e by size. N ext, F . C. B enham published an estim ate in 1928 fo r 1924-25; J. G. C raw ford published some rough calculations in 1936 fo r 1928-29 to 1933-34, and was joined in 1938 by C olin C lark in a m ore com prehensive survey cov­ ering the years 1928-29 to 1937-38, w ith the figures linked to all the other previous estimates so as to provide a fairly con­ tinuous and reasonably consistent series as fa r back as 1901. Their joint w ork was published in 193 8.28 Finally, in 1945, the governm ent definitely took on the function of estim ating national income. T he C om ­ m onw ealth Census and Statistics Bureau has been publishing annual com prehensive estimates of national incom e, gross national product, the distribution of incom e by states, and other aggregates ever since. T o these, it eventually added a set of social accounts. In N ew Zealand in 1930, som e thirty years after Coghlan’s estim ates for A us­ tralia, F. B. Stephens, a N ew Zealand scholar, prepared an estim ate fo r his own country th a t related to the year 1925-26. T his was followed in 1931-32 by the in­ auguration of official annual series prepared by the Census and Statistics D epartm ent. Both Stephens and the D epartm ent’s esti­ m ates w ere prepared by the incom e dis­ tributed m ethod and w ere based largely on income tax d ata.29 In Sweden, in the 1930’s, the Institute of Social Sciences at the U niversity of Stock­ holm , aided by a Rockefeller F oundation grant, m ade a series of studies on national income and trends of wages and prices (see references in C hapter 2 4 ). In N orw ay, the University Institute of Econom ics, u nder the guidance o f Professor R agnar Frisch, published several estimates in the early 1930’s together w ith analyses of some of the theoretical and practical problem s in­ volved in their preparation (see C hapter 34, Section II-6 ) .30 In A ustria, a newly founded Institute of Business Cycle Re­ search prepared a national incom e estim ate (see C hapter 34, Section II-9 ). In H ungary, the H ungarian Institute fo r Econom ic Re­ search sponsored a national incom e study

by M. M atolczy and S. V arga in 1934—39 (see C hapter 34, Section 11-10). In Poland, the Institute of Econom ic Research gave its support to sim ilar studies by Kalecki, L andau, and others (see C hapter 34, Sec­ tion 11-12). In Canada, the Bank of N ova Scotia started in th e 1930’s the annual preparation and publication o f national in­ com e estim ates th a t w ere independent of those prep ared by the D om inion B ureau of Statistics. In Bulgaria, the N ational Bank and th e Statistical Institute fo r Econom ic R esearch at the U niversity o f Sofia con­ d ucted several national incom e studies in the 1920’s an d 1930’s (see C hapter 34, Sec­ tion 11-13). In Japan, estim ating was u nder­ taken by the M itsibishi E conom ic Research Institute and by th e Jap an E conom ic F ed ­ eration, likewise in the 1930’s (see C hapter 36, Section 12).

b. A New Group of Scholars: Colin Clark, Kuznets, Gini, and Others D u rin g th e period 1918-39, new scholars appeared on th e scene, replacing older au­ thorities an d bringing w ith them new ap­ proaches. T hey w ere Colin C lark in E ng­ land, D uge de Bernonville in F rance, Simon K uznets and M ilton G ilbert in the U nited States, R. H . C oats in C anada, P. Jostock in G erm any, M . M atolczy and S. V arga in H ungary, F . B. Stephens in N ew Zealand, C orrado G ini and B enedetto B arberi in Italy, J. T . Sutcliffe in A ustralia, F ernand B oudhuin in Belgium, E rik L indahl in Sweden, R ay n ar Frisch in N orw ay, J. B. D. D erksen in H olland, P. M ori and Julius W yler in Switzerland, V. K. R. V. R ao in India, S. H . F rankel in the U nion of South A frica, an d F rederick C. Benham in the British W est Indies an d M alaya.

c. Estimates Become Official in Nine Countries T h e m ost outstanding developm ent o f this period was the fact th at in some nine coun­ tries the responsibility fo r the preparation and publication o f annual national incom e estim ates was assumed by governm ent. A precedent fo r this had been set in N ew South W ales in the 1880’s, b u t governm ent estim ates w ere abandoned th ere in 1904 w hen T . Coghlan resigned his position as governm ent statistician of N ew South W ales. F o r a tim e thereafter, national incom e esti­ m ating in A ustralia was left to the sporadic efforts of private individuals such as J. T.

10.

151

UNIVERSALIZATION IN THE TW E N T IE T H CENTURY

Sutcliffe and F rederick Benham in 1926, an d Colin C lark and J. G . C raw ford in 1938. T h e first official estim ating agencies fol­ low ed the estim ating techniques developed by the private scholars who h ad preceded them , while at the same tim e expanding considerably the scope o f the estimates. T hey often carried the estim ates back to past years to develop com prehensive his­ torical series. T hey also deflated the figures to constant prices and introduced num erous additional analyses o f the total national in­ come, as well as m ore detailed breakdow ns o f p articular segments of it. T hese enlarge­ m ents m ade the estim ates m ore useful to governm ent, to private scholars, and to the general public.

universities as one of the standard graduate courses in economics, and basic textbooks in econom ics w ere being rew ritten to m ake national incom e one of the starting points o f discussion.

3. Universalization, Enrichment, and Standardization of Esti­ mates During and After World War II D uring W orld W ar II and the years im ­ m ediately following it, literally trem endous developm ents took place in the national in­ com e estim ating field. Some of these were: (a) w artim e fiscal and econom ic planning was now based on national income esti-

T ab le 10-4. T h e F irs t E stim a te s Placed on a n Official Basis, 1886-1939

C o u n try 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

A u stra lia C anada S oviet R ussia G e rm an y N e th erlan d s N ew Z ealand U n ited S ta te s T u rk e y Y ugoslavia S w itzerland .

Official E stim a tin g A gency G o v e rn m en t S ta tisticia n D om inion B ureau of S ta tistics C e n tra l A ccounting Office (T Z U N H U ) C e n tra l S ta tistica l Office C e n tra l B ureau of S ta tistic s C ensus a n d S ta tistics D e p a rtm e n t D e p a rtm e n t of C om m erce M in istry of N a tio n a l E conom y M in istry of F in an ce C e n tra l S ta tistica l Office

d. National Income Research Elevated to a Central Place in Economics T he enrichm ent o f national incom e esti­ m ates th a t took place during th e period 1918-39 through the com bined efforts of private research institutions and official sta­ tistical agencies enabled scholars to u n d er­ take m ore am bitious explorations o f the various phenom ena associated w ith the op­ eration of the national econom y in both peace and w ar— explorations in w hich the quantitative approach played a leading role. Studies were published of the co nsum ptionsaving ratio, capital form ation, changes in productivity, the relationship of em ploym ent to national incom e and the operation of the “m ultiplier,” the relationship of governm ent spending to national income, changes in the structure of the national incom e in w artim e, an d m any other topics of im portance. T he w hole structure of econom ics was being changed by this new analytical m ethod. N a ­ tional income analysis was being adopted by

Y e ar of Official O rganization of the A nnual E stim ates 1886 1925 1925 1929 1931 1931 1934 1935 1937 1939

m ates; ( b ) the preparation of estimates everyw here becam e a governm ent function; (c) estim ates were enriched by emphasizing national expenditures as m uch as national incom e and relating the one to the other, by em phasizing the “gross national p ro d u ct” as m uch as the “net national pro d u ct,” by developing sector accounts, and, finally, by introducing com pletely new kinds of analyses such as the input-output approach; and ( d ) national incom e concepts were standardized by international agreements. W e shall discuss each one of these develop­ m ents briefly.

a. National Income Estimates Become the Core of Governmental Fiscal and Economic Planning Before the w ar, finance ministers either dismissed national incom e estimates as the idle speculations of academ ic minds and as having no practical utility whatsoever, or, at best, referred to them merely as interest­

152

THE INCO M E OF NATIONS

ing inform ation, fo r the accuracy o f w hich they could not vouch. N ow , these same m inisters, o r their successors, used these estimates as a basis for their fiscal and eco­ nom ic program s. M oreover, in countries w here estimates had been only privately prepared up to th a t tim e, the authorities were at last persuaded th a t estimates con­ tained inform ation of crucial im portance to the nation and should be prepared by the governm ent itself because it alone possessed the resources needed to ensure their m axi­ m um reliability. In England, it took a great national em er­ gency and the com bined efforts of some of her ablest economists to bring about this change in official attitude tow ards national incom e estimates. Im m ediately after the out­ break of the w ar, in, the fall of 1939, under the influence of John M aynard Keynes (1 8 8 3 -1 9 4 6 ) and his associates, the C han­ cellor of the E xchequer finally authorized the preparation of national incom e esti­ m ates, and in A pril, 1941, built his budget around one.31 R. F . H arro d in his biog­ raphy of Keynes describes vividly how this h ap p en ed : In the month of June (1939) . . . Keynes reentered the Treasury. . . . He was not a civil servant and drew no salary . . . it was inevitable that his influence should be felt. . . . In the autumn of 1939 Lord Stamp, assisted by Henry Clay and Sir Hubert Henderson was charged with the duty of securing economic co-ordination. They had a slender staff. . . . When the Churchill Government was formed in May (1940), Mr. Arthur Greenwood, Min­ ister Without Portfolio, had the duty of co­ ordinating the work of the various Ministries concerned with our economic effort. His prin­ cipal civil servant, Mr. Francis Hemming, with wise foresight and commendable respect for economics, moved quickly and mustered a large team of professional economists. . . . They were all surprised at the large measure of agreement they found with each other on the basis of their economic science, despite po­ litical differences and adherence to rival schools of thought. Of special interest was the close associations which grew up between Mr. James Meade, a convinced Keynesian and moderate Socialist, and Professor Lionel Robbins who had been a strong individualist and not entirely convinced of Keynes’ doctrines. Mr. Austin Robinson, who had followed with interest Keynes’ study on How to pay for the War, pressed the need for official work on the kind of statistics which Keynes employed in that book, and informed Mr. Hemming that Mr. Richard Stone was the ideal man for the purpose. He was accordingly brought into the

party, and in the autumn of 1940 he and Mr. James Meade worked hard on an analysis of the national income and expenditure. Seeking further authority and guidance in their efforts, they naturally found their way to the room of Keynes, who gave them the utmost en­ couragement and help. In full collaboration with him, they got out an account that winter. Keynes pressed the importance of such statistics on the authorities. He won the support of the cautious Sir Richard Hopkins, whose instinct was to be extremely sceptical of calculations of this kind; once convinced, however, he valiantly championed them and in his turn had difficulty in persuading the Chancellor of the Exchequer that the account should be pub­ lished at the time of the 1941 Budget. This was indeed a great revolution. The Treasury had hitherto confined itself to figures for actual, known transactions. This account included es­ timates, and certain figures had to be obtained by the method of difference from other esti­ mates— all of which was very dangerous. Yet this kind of national income accounting has come to be regarded as the essential tool of any economic planning, whether of an Individ­ ualist or Socialist variety, and other nations have followed Britain in presenting such ac­ counts. The initiative must be attributed to Keynes, but for whose active interest the com­ pilation would not have been made and pub­ lished at that time. Mr. Stone felt that the 1941 paper was only a beginning and that much further work was desirable. Meanwhile Mr. Hemming’s group of economists had been divided into two quite distinct sections. It was thought to be desirable to make a sharp division between the econo­ mists who were engaged in compiling statistics relating to the war effort and those whose duty it was to comment upon them for the benefit of Cabinet Ministers; the latter, if they were to be of any use, would naturally be expected to have strong views of their own, and there was the human danger that such views might influence the form in which the statistics were presented. The former group, namely, the com­ pilers of statistics, became the Central Statis­ tical Office, under Mr. Campion; while the latter became the Economic Section of the War Cabinet Secretariat, under Professor Jewkes, and later under Professor Lionel Rob­ bins. Mr. Stone went into the Central Statis­ tical Office. He felt doubtful whether his au­ thorities there would allow him sufficient time for work upon the national income, and he approached Keynes for further support. A plan was accordingly drawn up, by which, while retaining his position in the Central Statistical Office, he was attached to Keynes as his per­ sonal assistant in the Treasury. The work on the national income went ahead. Three or four times each week Mr. Stone visited Keynes, who took a meticulous interest in every detail.32

10.

UNIVERSALIZATION IN THE TW E N T IE T H CENTURY

In countries such as the U nited States and Canada, w here official estim ates had been in vogue fo r some years, their ad­ vancem ent to a central place in governm ent w artim e fiscal and econom ic planning was effected naturally and smoothly. In the U nited States, H arold D . Sm ith, w artim e director o f the budget, aided by his chief economist, G erh ard Colm, brought about the use of national incom e estim ates as bases fo r the preparation of th e govern­ m ent budget.38 O ther econom ists, operating in o th er key econom ic agencies, likewise in­ troduced the national incom e approach to w ar production, price control, and other econom ic planning of the governm ent. T ow ard the end of the w ar, the Swedish, N orw egian, and A ustralian governm ents fol­ lowed B ritain’s exam ple and organized the official preparation of estimates. T he elevation of national incom e esti­ m ates to a central place in governm ental fiscal and econom ic planning did n o t te r­ m inate with the end of the w ar. H aving learned their usefulness in such planning during the war, governm ents th a t w ere com ­ m itted to planning fo r peacetim e reconstruc­ tion, expansion, and econom ic stabilization, continued to use national incom e estimates in m aking their postw ar plans. O ne govern­ m ent after another began to m ake projec­ tions of national incom e under different assum ptions regarding the rate of expansion of private production, consum ption, invest­ ment, governm ent expenditure, and taxation. On the other hand, in F rance, during the same period 1917-45, com paratively little progress was m ade in the field. N ational in­ come estimates continued to be prepared by individual scholars, w ithout m uch encour­ agem ent from the governm ent. D uge de Bernonville, head of the statistical research unit in the M inistry of Finance, prepared from tim e to tim e on his own responsibility national incom e estim ates w hich he even­ tually brought together in a consistent an­ nual series covering the period 1920-39. U nfortunately these estim ates w ere based on very incom plete statistics and w ere w eak in some parts. A lfred Sauvy and R. Rivet published in F ebruary, 1939, corrections of Bernonville’s estimates, raising his totals considerably.34

b. Equal Importance Attributed to Esti­ mates of National Expenditure G overnm ent expenditures, reflecting the w ar effort, expanded trem endously during

153

the w ar and modified the whole structure o f the national economy. G overnm ents as well as the general public, largely u nder the influence of the Keynesian economists, becam e aw are of the fact that national expenditure was at least as dynam ic a factor in the econom y as national income, and th a t it should be estim ated and ana­ lyzed as well. M oreover, as governm ent budgets, w ith their incom e and expenditure accounts, cam e to dom inate the operations of the national econom y, the concept of a sim ilar two-sided incom e and expenditure account of the entire national economy em erged. It becam e im p o rtan t to relate the governm ent’s incom e and expenditure to the total national incom e and expenditure. T he idea of presenting national incom e and national expenditure in a single “national account” was translated into practice alm ost imm ediately. H ere, too, credit fo r the ini­ tiation and execution of the idea belongs to the K eynesian group, fo r it was the same W hite P ap er o f A pril, 1941, th a t first gave expression to the “national income and outlay account” idea.35 U nited States n a­ tional incom e estim ators and analysts fol­ lowed suit, first w ith the preparation o f the “N atio n ’s Econom ic Budget” in January, 1943,36 ancj next, in 1947, w ith the start of the annual publication o f social accounts. In the U nited K ingdom the annual publi­ cation o f social accounts was also initiated in 1947, and in C anada it began in 1948.

c. Shift in Emphasis from Net to Gross Product Figures T he th ird im portant change ushered in by the w ar was the shift in em phasis from net to gross p roduct figures; i.e., to the inclu­ sion (o r nonexclusion) of depreciation and indirect taxes in the evaluation of the n a­ tional pro d u ct o r incom e. This change, too, was an outgrow th of the governm ent budgetm akers’ attem pt to achieve a closer connec­ tion between the governm ent’s budget and the national incom e estimates. Inasm uch as am ortization o r replacem ent of w orn-out private capital was being deferred until the postw ar period so th at the private funds and resources th a t would ordinarily have been used for th at purpose could be diverted to current w artim e production, it was im­ p ortant to present the national income figure gross of this item. Similarly, inasm uch as governm ent expenditures were m ade from revenues including indirect taxes, the calcu­ lation o f national incom e and expenditure

154

THE INCO M E OF NATIONS

had to be gross of these taxes. In no other way could governm ent incom e and expendi­ ture be accurately and consistently related to national income and expenditure. T he U nited States estim ators w orked out the “gross national product” or G N P ap­ proach in a particularly refined fashion, and were influential in its subsequent adoption by m any other countries.87 A fter the w ar, gross national product continued to play a role of equal im portance in national incom e analysis to that of net national incom e at factor cost, w ith each aggregate serving a different type o f econom ic analysis.

d. Development of “Sector Accounts”: Complete Social Accounting Approach T he developm ent o f the idea o f a double­ entry national account of incom e and ex­ penditure fo r the w hole econom y led to the subsidiary idea of constructing sim ilar double-entry accounts of incom e and ex­ penditure fo r each sector of the economy. A ctually, these tw o ideas were conceived sim ultaneously by the same group o f E ng­ lish economists— M eade, Stone, and K aldor — under K eynes’s guidance. B ut of the two ideas, in the U nited K ingdom only the first — th a t of a national account— was given im m ediate official sanction.38 These English economists m ay not have been com pletely original in the advancem ent of these ideas. Some Scandinavian and D utch economists claim to have form ulated the same ideas at about the same tim e (see C hapter 23, Sec­ tion 6, and C hapter 34, Section II-6 ). T he sector accounts suggested by M eade and Stone were four in num ber: (1 ) the private household account, including do­ mestic service, and representing largely p er­ sonal consum ption and saving; (2 ) the busi­ ness enterprise account consisting largely of undistributed profits and gross private capi­ tal form ation; (3 ) the governm ent account, representing collective consum ption and in­ vestm ent and revealing its surplus or deficit; and (4 ) the rest of the world account, showing the transactions of the domestic econom y with the outside world and the resulting positive o r negative balance. Before being able to convince their own governm ent of the advisability of m odifying the traditional presentations of national in­ com e estimates so extensively, Stone and M eade had to persuade professional opinion of the m erits of the social accounting ap­

proach. T his task was undertaken by Stone, who, during the succeeding few years, pub­ lished several articles on th e subject, had conferences w ith the U nited States and C a­ nadian official estim ators, and prepared a rep o rt in 1947 for the League of N ations’ C om m ittee of Statistical E xperts.39 T he re ­ sponse to the proposal was generally favor­ able. In 1946 the British T reasury cam e out with a W hite P aper on national incom e and expenditure th a t was based on M eade and Stone’s social accounting approach,40 and in the following year, th e U.S. D epartm ent of C om m erce published a sim ilar recast of its estim ates.41 D uring the next several years, the social accounts approach was adopted by m any oth er countries. In fact, in some countries it was tak en up w ith such enthusiasm th at national incom e estim ates as such were alm ost com pletely subm erged. T his was p ar­ ticularly tru e o f the studies published under the direction of J. R. N . Stone and M ilton G ilbert by the N ational A ccounts Research U nit o f th e O rganization fo r E uropean E co­ nom ic Cooperation (O E E C ).42 Some n a­ tional incom e analysts expressed reserva­ tions, how ever, concerning the extent of usefulness of this new type of presentation of national incom e d ata.43 T he three im portant changes in the con­ struction o f national incom e estim ates ef­ fected during the period 1 9 3 9-57— the ad­ vancem ent of the analysis of national ex­ penditures to an equal position w ith th at of national incom e; the advancem ent of gross national p ro d u ct to a position of co-ordinate im portance w ith th a t of national incom e; an d the introduction of social accounts— taken together, trem endously enriched the content and usefulness o f national incom e estimates.

e. Standardization of Concepts by Inter­ national Agreement and Example In 1944, the representatives o f the British, A m erican, and C anadian official estim ating agencies m et in W ashington and agreed on the use of certain concepts of national in­ com e an d form s of presentation in th eir n a­ tional incom e estim ates so as to m ake them m ore com parable and useful to th eir respec­ tive governm ents. This was the first interna­ tional agreem ent in this field, an d it exerted a substantial influence on the developm ent o f national incom e estim ates in other coun­ tries over the succeeding years. Finally, w ith th e organization o f the U nited N ations

10.

UNIVERSALIZATION IN THE TW E N T IE T H CENTURY

and later of the O rganisation fo r E uropean E conom ic Co-operation, and the establish­ m ent of a national incom e research unit in each of them , international standardization of concepts, m ethods, and form s of presen­ tation was fu rth e r broadened. Both national incom e research units published standard classifications of industrial divisions and standard definitions of concepts th a t w ere adopted by m any countries fo r use in their national incom e estim ates.44

f. Development of the Input-Output Approach T he developm ent of national incom e esti­ m ates gave birth, in 1941, to a new type of analysis of national production. T his was the “input-output” approach devised by W assily Leontieff of H arv ard U niversity.45 U n d er this approach, the econom y is ana­ lyzed both (a) in term s o f the physical quantities o r their values supplied by each industrial and nonindustrial division to other divisions (o u tp u t); and ( b ) in term s o f the physical quantities or their values obtained by each division from other divisions (in­ p u t). These tw o analytical series are com ­ bined in a single table, w ith the producing divisions listed in the left-hand colum n, and figures (representing th e output) running horizontally from left to right. T he same divisions are also listed as the receiving or purchasing divisions at the top of each colum n with the figures (representing the input approach) running from the top down. A lthough, in its genesis, the input-output approach was related to the national incom e and expenditure concept (being linked p a r­ ticularly to the com m odity flow approach first developed by K uznets), in its actual applica­ tion it is com pletely independent of it. It reflects the successive stages o f production in term s of the gross values of their outputs rath er than in term s of their net values or factor costs; and it calls fo r a m ore m inute subdivision of the econom y into branches of production than is custom ary in national incom e analysis. T hus, Leontieff lists alm ost fifty divisions o r sectors fo r the A m erican econom y as against the usual dozen o r so in national income estimates.

g. Universalization of Preparation of Estimates During the Postwar Period D uring W orld W ar II only five o r six ad­ ditional countries were covered by national incom e estimates, but after th e w ar the num ber of countries taking up the prep ara­

155

tion of such estim ates increased by leaps and bounds, m aking their preparation prac­ tically universal throughout the world; and practically everywhere the estimates were now official. Tw o factors have been respon­ sible fo r this practical universalization of the preparation of national incom e estimates and their organization u nder government auspices during the period following W orld W ar II: (1 ) the adoption by the U nited N ations of a rule apportioning the costs of support of the organization am ong m em ber nations on the basis of th eir relative n a­ tional incom e, w hich m ade it necessary for those of them n ot having any official esti­ m ates to develop them forthw ith; and (2) the organization by the U nited N ations of a national incom e unit th at publishes n a­ tional incom e estimates of the various coun­ tries, dem onstrates their usefulness, and stands ready to give technical assistance to any nation in the preparation of its esti­ mates. T he speed w ith w hich the countries took up national incom e estim ating after W orld W ar II was truly am azing. W ithin ten years (1 9 4 6 -5 5 ), the num ber of countries having national incom e estimates m ore than dou­ bled from about thirty-nine to m ore than eighty; and in all b ut fo u r or five of them the estimates w ere official. But even m ore startling was the develop­ m ent for the entire period of the tw entieth century here u n d er discussion: w hereas at the end of the nineteenth century only eight countries had national income estimates, by the end of 1955, eighty had them . T he growth in their num ber by periods during th a t tim e is shown in T able 10-5. T he universalization of official estimates does n o t m ean th at the day for private esti­ m ates is gone. It m erely m eans th at from now on private estimates will need to be prepared only w here governm ent, fo r some reason or other, fails to prepare one, or w here its estimates, in the opinion of p ri­ vate scholars or research institutions, are inadequate o r faulty. Even in the U nited States, there is disagreem ent between some private scholars and the official agencies as regards the correctness of the latter’s esti­ m ates; and such disagreem ents may readily develop elsewhere as well. In the case of Soviet Russia’s official estimates, which are offered w ithout any docum entation, there is a natural inclination am ong the statisticians of the free w orld to try to develop inde­ pendent estimates fo r th a t country.46

THE IN CO M E OF NATIONS

156

T ab le 10-5. C ountries in W hich N a tio n a l Incom e E stim a te s W ere F irst D eveloped D uring th e 2 0th C e n tu ry (A dditions to th e earlier eig h t: E ngland, F rance, U n ited S tates, R ussia, A u stria, G erm any, A u stralia, a n d N orw ay)

C o u n try

N am e of F ir s t E stim a to r

Y ear of A ppearance of F irs t E stim a te

E stim a tin g M eth o d

1902 1902 1910 1911 1915 1917

n e t o u tp u t incom e d istrib u te d incom e d istrib u te d n e t o u tp u t n e t o u tp u t n e t o u tp u t

incom e d istrib u te d n e t o u tp u t incom e d istrib u te d a n d n e t o u tp u t n e t o u tp u t n e t o u tp u t n e t o u tp u t n e t o u tp u t n e t o u tp u t incom e d istrib u te d incom e d istrib u te d incom e d istrib u te d

I. Period B efore W orld W ar I 9. 10. 11. 12. 13.

Ja p a n Sw itzerland N eth erlan d s Ita ly B ulgaria S pain “

K . N a k am u ra T . G eering a n d R. H o tz W . A. B onger M . S antoro K . Popoff A ndre Barthfe II. Period B etw een th e T w o W ars

14. Belgium 15. G reece 16. E g y p t

C harles C lavier M in istry of F inance I. G. L evi

1919 1921 1922

17. 18. 19. 20. 21. 22. 23. 24.

C anada Spain U nion of S outh Africa Czechoslovakia Y ugoslavia D enm ark L a tv ia N ew Zealand

1925 1925 1925 1926 1927 1930 1931 1931

25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41.

Poland Irelan d T u rk e y F in lan d Chile Sw eden M exico B ritish In d ia Peru Israel (P alestine) C hina D om inican R epublic B ritish W est Indies (Jam aica, etc.) P u e rto Rico S o u th ern R hodesia A rgentina Ceylon

D om inion B ureau of S ta tistics J . A. V andellos R . A. L ehfeldt S. N . P roko p o v itch A. V egner a n d A ssociates T . K ristensen A. C eickners C ensus a n d S ta tistics D e p a rtm e n t B. D ederko T . J . K iern an M in istry of N a tio n a l E conom y V a lte r L indberg R a u l Sim on E rik L in d ah l a n d A ssociates C e n tra l S ta tistic a l Office V. K . R . V. R ao R om ulo F errero L udw ig G ru en b au m D a K uen L ieu B rookings In stitu tio n F red erick B enham

1943 1945 1946 1946

42. 43. 44. 45. 46. 47. 48.

N o rth e rn Rhodesia R um ania A lbania L uxem bourg Indonesia B elgian Congo K en y a

49. 50. 51. 52. 53.

N y a sa lan d P a ra g u ay U ru g u a y C uba S o u th ern Rhodesia

D udley S m ith S. H . F ra n k e l B anco C e n tra l (Bunge) D e p a rtm e n t of C ensus a n d S ta tistics P hyllis D eane C e n tra l S ta tistic a l Office C e n tra l S ta tistic a l Office M in istry on E conom ic A ffairs C e n tra l B ureau of S ta tistics M . M asoin E a s t A frican S ta tistica l D e p a rtm e n t P hyllis D eane C. A. Soler a n d M . C irovic R aul O choa B anco N acional C en tral A frican S ta tistic a l Office

1932 1933 1935 1935 1935 1937 1939 1939 1939 1941 1941 1942 1942

1946 1948 1948 1948 1948 1948 1948 1948 1948 1948 1949 1949

final e x p en d itu re n e t o u tp u t n e t o u tp u t n e t o u tp u t n e t o u tp u t n e t o u tp u t n e t o u tp u t n e t o u tp u t

incom e d istrib u te d n e t o u tp u t n e t o u tp u t

gross o u tp u t

incom e d istrib u te d

n e t o u tp u t

incom e d istrib u te d

10.

C o u n try

N am e of F irs t E stim a to r

Y e ar of A ppearance of F ir s t E stim a te 1949 1949 1949 1950-52

87. M a u ritiu s

C e n tra l B an k C e n tra l B an k B anco de la R epublica C om m onw ealth C onsulting C om m ittee, L ondon; M in istry of F inance F re d e ric k C. B enham In s titu te N acional de E stad istica S ta tistic s Section M in istry of N a tio n a l P lanning (M iss P . A dy) B anco C e n tra l H . C. W allich a n d J . H . A dler U n ited N a tio n s E conom ic M ission S tich tin g P lan b u reau N a tio n a l E conom ic C ouncil C olonial Office Office of G overnm ent S ta tistic ia n T w e n tie th C e n tu ry F u n d In te rn a tio n a l B ank for R econstruction D evelo p m en t C om m issioner’s Office C e n tra l B an k U .N . Social a n d E conom ic C ouncil H . D rilier N . S. C a rey Jones S ta tistic s a n d C ensus D e p a rt­ m e n t (U .N . T echnical A ssistance) D e p t, de l ’Econom ie N a tio n ale C olonial Office E a s t A frican S ta tistica l D e p a rtm e n t B anco C e n tra l R . R. N a th a n & A ssociates H . A. M as a n d E . F . Sczepanik H . R. V an O lst (U .N . T echnical A ssistance) U n ited N atio n s A. Y . B adre B u re au of A ccounts a n d S ta tistics C e n tra l S ta tistic a l Office

88. G h a n a

M in istry of F in an ce

1956

89. T a n g a n y ik a

C olonial Office, L ondon

1957

90. Ira n 91. I ra q 92. J o rd a n

U. S. M ission H . G . F enelon A. Y . B adre

1957 1957 1957

54. 55. 56. 57.

P hilippines V enezuela Colom bia P a k ista n

58. 59. 60. 61.

Malaya Portugal Cyprus Burma

62. Guatemala 63. El Salvador 64. Bolivia 65. 66. 67. 68.

Surinam Thailand Gambia Gold Coast

69. Costa Rica 70. Nicaragua 71. British Guiana 72. Brazil 73. Iceland 74. French West Africa 75. British Honduras 76. Panama 77. H a iti 78. N igeria 79. U ganda 80. 81. 82. 83.

157

UNIVERSALIZATION IN THE TW E N T IE T H CENTURY

H o n d u ras R epublic of K orea H ong K ong E cu a d o r

84. S yria 85. L ebanon 86. T aiw a n (form erly F orm osa)

1950 1951 1951 1951

E stim atin g M ethod

n e t o u tp u t n e t o u tp u t

n e t o u tp u t

n e t o u tp u t

1951 1951 1951 1952 1952 1952 1952

n e t o u tp u t

1952 1952

n e t o u tp u t

1952

n e t o u tp u t

1953 1953

incom e d istrib u te d n e t o u tp u t

1953 1953 1953

1953 1953 1953 1954 1954 1954 1954 1955 1955 1955 1955

n e t o u tp u t

n e t o u tp u t

n e t o u tp u t

n e t o u tp u t a n d fi­ n a l e xpenditure n e t o u tp u t a n d e x p en d itu re n e t o u tp u t a n d expen d itu re expen d itu re n e t o u tp u t n e t o u tp u t

a S pain is nu m b ered in th e second section of th is ta b le because th e 1917 e stim a te w as p repared o u t­ side th e c o u n try . s o u r c e : Based, p a rtly on original research a n d p a rtly on th e U n ited N ations, Statistics of National Incom e and E xpenditure, S ta tistic a l P apers, Series H , Nos. 8 a n d 9, N ew Y ork, 1955 a n d 1956.

158

THE INCO M E OF NATIONS

4. Forces That Influenced the De­ velopment of the Past Three Hundred Years W e have now brought to an end our his­ torical account of the world-wide develop­ m ent of national incom e estimates over the past three hundred years. W e have seen how it began in the isolated efforts of one or two private scholars struggling against public indifference and the actual hostility of their governm ents, and how it eventually becam e world-wide in scope, em bracing the activities of some ninety-odd national statis­ tical agencies and innum erable governm ent statisticians. M any forces have contributed to this developm ent, but am ong them the following eleven seem to be outstanding.

a. Interest and Initiative of Individual Scholars T he forem ost factor has been the interest and initiative of individual scholars, both economists and statisticians. T hey w ere the first to realize the im portance of national in­ com e estimates as tools o f econom ic analysis. T hey bore, w ithout assistance from any out­ side source, the brunt of the w ork involved in preparing such estimates in the past. If it had not been for their efforts, there would be no national incom e estimates today. Even if private scholars no longer actually pre­ pare the estimates, at least they continue to interpret the officially prepared estimates, to furnish new ideas fo r their preparation and, if necessary, to challenge their accu­ racy.

b. Advance of Economic Theory T he second im portant influence has been the advance of econom ic theory th a t p ro ­ vided the basic concepts of national income, gave purpose and m eaning to the estimates, and illum inated their developm ent through­ out. T here is no escaping the influence that the several successive economic doctrines exercised on the concept and utilization of national incom e estimates; i.e., the neom ercantilistic or semiliberal doctrine of the sec­ ond half of the seventeenth and first half o f the eighteenth century; the physiocratic and Sm ithian doctrines of the second half of the eighteenth and first half of the nine­ teenth century; the M arxian and M arshal­ lian doctrines of the second half of the nine­ teenth and beginning of the tw entieth cen­

tury; and the K eynesian doctrine of the sec­ ond qu arter of the tw entieth century.

c. Wars, Revolutions, and Severe Eco­ nomic Crises W ars, revolutions, and severe econom ic crises, requiring governm ent m obilization of all the econom ic resources of a nation, have played an im portant role in the develop­ m ent of estimates. D uring such catastrophes and the periods of recuperation following them , national incom e estim ates achieved special im portance as tools th a t could facili­ tate the planning of such m obilization and recuperation. It is notew orthy th a t the de­ velopm ent of national incom e estimates has been particularly rap id during periods of political and econom ic stress an d on the other h and has been sluggish during periods of tranquillity. T he greatest activity has oc­ curred successively (a ) in E ngland, during the Crom w ellian Revolution and its im m e­ diate afterm ath; ( b ) in b oth E ngland and F ran ce during and after th eir w ars at the end of the seventeenth century, and again a century later during the F ren ch Revolu­ tion and the N apoleonic w ars an d the suc­ ceeding period of reconstruction, an d (c) in all countries during the two w orld w ars of the tw entieth century and th eir im m ediate afterm aths.

d. Agitation for Fiscal Reform Serious breakdow ns in existing fiscal sys­ tem s and the form ulation o f fundam ental fiscal reform s, requiring inform ation on the total national incom e, its sources, and dis­ tribution, have also played a role. N o te the fiscal purposes for w hich the estim ates of Petty, W illiam Pitt, D r. Beeke, Lowe, and P ebrer in England, and of Boisguillebert, V auban, the physiocrats, Lavoisier, Lavergne and H enry in F rance, and of Spahr in the U nited States were prepared.

e. Rapid Technological Changes in the Economy R apid technological change th a t funda­ m entally modifies the structure of the n a­ tional econom y has been an o th er factor. It has tended to heighten public interest in periodic national incom e estim ates th at re ­ veal the extent of the changes taking place in the econom y and show th eir effects on the size and distribution of the national in­ com e. T h e national incom e estim ates of the nineteenth century in E ngland and F rance

10.

UNIVERSALIZATION IN THE TW E N T IE T H CENTURY

w ere largely directed tow ard a revelation of the changes produced by the rapid growth o f industry and, in England, the decline of agriculture.

f. Class and Group Interest Conflict T h e evolution of class and group interest conflict has aroused keen interest in esti­ m ates o f the size of the national incom e an d its distribution am ong th e various groups o f the population. T he influence of this facto r is exemplified by the em phasis th a t the English, F rench, and G erm an n a­ tional incom e estim ates of the second half of the nineteenth century and the A m erican estimates of th e end of th e nineteenth and beginning of the tw entieth century placed upon the issue of distribution.

g. International Commercial and Political Rivalries International com m ercial and political com petition has operated to stim ulate inter­ est in the preparation o f estim ates fro m the very beginning. C om petition m ade it im ­ po rtan t fo r a nation to know n o t only its own econom ic strength as revealed by its national incom e b u t also th e strength o f its com petitors as revealed by estim ates of their national income. C om m ercial com petition was one of the factors prom pting English an d F rench econom ists of th e seventeenth, eighteenth, and nineteenth centuries to m ake recurrent com parisons of the British and F ren ch national incom es; and it sim ilarly prom pted the G erm an, F rench, and English estim ators of the tw entieth century to m ake like com parisons o f th e grow th o f the n a­ tional incom e of th eir countries. C om peti­ tion of political and econom ic systems has also played a part. In the eighteenth century it was the com petition betw een the m ore dem ocratic political system of E ngland and the autocratic system o f F rance; and in the tw entieth century, it has been th e com peti­ tion between the free enterprise system of the w estern w orld and th e com m unistic sys­ tem of the countries in the Soviet orbit th a t stim ulates international com parisons of the grow th of national income.

h. Improvement in Statistical Data and Methods Im provem ent in statistical data and m eth­ ods has been a continuing influence in the developm ent of estimates. T he introduction of periodic governm ent censuses o f popula­

159

tion, industry, agriculture, etc., and other periodic surveys, and the organization by governm ent of central statistical agencies have stim ulated the developm ent of national incom e estimates, particularly in the nine­ teenth and tw entieth centuries. It was n o t an accident th a t m any of the m ost im portant estim ates of th at period follow ed the taking of an im portant census. F o r example, Colqu­ h o u n ’s estimates follow ed the first census in England; M oreau de Jonnes’ followed the first agricultural and industrial censuses in F rance; T u ck er’s, in the U nited States, fol­ low ed the all-im portant census o f 1840; and C zoernig’s estim ate in A ustria followed the first general population census of th a t coun­ try. T he professional governm ent statisti­ cian, having once becom e aw are of the im­ portance of national incom e estimates, and having assumed responsibility fo r their p rep­ aration, tends, as a rule, continually to im ­ prove on his data and m ethods. If he fails to do so, he is often prom pted by public criticism— in a free country— to correct his failures.

i. Introduction of the Income Tax T he introduction of the incom e tax facili­ tated th e p reparation of national income estim ates by providing valuable data, p a r­ ticularly on the incomes of the m iddle and u pper classes. T he influence of the income tax was apparent in the preparation of the num erous estim ates of the second h alf of the nineteenth century in E ngland and G er­ m any, and, tow ards the end of th at century and at the beginning of the tw entieth cen­ tury, in France.

j. Development of Governmental Fiscal and Economic Planning T he developm ent of fiscal and economic planning by m odern governm ents fo r the prom otion o f econom ic stability, full em ­ ploym ent, and econom ic grow th has been a prim e influence in the developm ent of national incom e estimates in recent decades. T he discharge o f this responsibility requires the preparation of reliable national income estimates and even the projection of such estim ates into the future. This new factor, dating from the great depression of the 1930’s an d especially from the period of reconstruction im m ediately following W orld W ar II, is revolutionizing the field of n a­ tional incom e estimating.

160

THE INCO M E OF NATIONS

k. Influence of International Organization Last, but not least, has been the influence of international organization, especially of the U nited N ations and the O rganisation for E uropean E conom ic Co-operation through their respective national incom e research units. This new factor has been trem en­ dously im portant, and will undoubtedly con­ tinue to be so in the future.

It is difficult to p redict the fu tu re develop­ m ent of national incom e estim ates and their possible applications. T h e developm ent m ay well exceed the fondest expectations of the statisticians of today, just as the progress to date has undoubtedly exceeded the dream s of the founders of national incom e estim at­ ing— of Petty, King, an d D avenant in E ng­ land an d of Boisguillebert and V auban in F rance.

Part T w o. Theory and M ethodology

11. BASIC CONCEPTS tion, distribution, and disposition— is essen­ tial fo r any balanced appraisal of the econ­ om y’s operations. N ational incom e analysis also serves as a tool fo r determ ining the nature, size, and interrelationships of the econom y’s com plex financial transactions. These transactions, grouped together according to the various sectors of the econom y to w hich they be­ long (e.g., household sector, governm ent sector, business secto r), are called “social accounts.” N ational incom e, cutting across sector lines, serves to unify these accounts.

1. The Meaning of National Income N ational incom e is an expression, in m one­ tary term s, of the curren t achievem ents of the national econom y. T hese achievem ents m anifest themselves in the m any com m odi­ ties and services, or econom ic goods, that the econom y produces fo r the consum ption of its m em bers o r fo r the addition to their existing productive assets during a given period, usually a year. Since the national econom y is not a “closed” econom y but one th at exchanges some of its goods and serv­ ices fo r those of other econom ies, the na­ tional product of the period is adjusted fo r the net balance o f incom e derived from these international exchanges. If the econom ic goods com prising the national incom e are exam ined in the process of their creation in the various branches of the economy, relative to the services of capital and labor em ployed in them , then national incom e m easures the productivity of the econom ic system. If, on the other hand, national incom e is exam ined in its distribution phase, as a flow of m oney in­ comes from producing units to the partici­ pants in production, it appears as a m easure of the equitableness o f the existing social and econom ic order. Thirdly, if these eco­ nomic goods are exam ined in the process of their consum ption by the m em bers o f so­ ciety or of their addition to society’s capital, then national incom e appears as a m easure o f econom ic welfare. A n exam ination of all three phases of national income— produc­

a. Distinction Between Economic and Noneconomic Production; Between Economic Welfare and Welfare Generally E conom ic production covers the complex of hum an activities devoted to the creation, w ith lim ited resources, of goods and serv­ ices capable of satisfying hum an w ants and, because of their lim ited supply, having defi­ nite “econom ic value.” Econom ic production includes the ad­ m inistrative services of governm ent even though they are not sold individually to citizens and, hence, have no m arket price. Since these services entail outlays fo r labor, capital, and m aterials th a t are defrayed by the governm ent on behalf o f society, they are “purchased” collectively by the citizens through the paym ent o f taxes and other charges (fees, etc.) an d through borrowing. In this sense, they have an economic “price.” E conom ic production also includes the serv-

Notes begin on p. 529. 163

164

THE INCO M E OF NATIONS

ices of churches and other benevolent o r­ ganizations such as private universities and colleges, educational foundations, and chari­ table hospitals. T heir services, too, have an econom ic cost that is defrayed collectively by their m em bers and benefactors through voluntary contributions o r self-assessments, and in th a t sense have a “price.” Econom ic production, in national incom e accounting, also includes the current services of certain durable goods, such as land and buildings, whose values to their users can be m easured in term s of th e rents that are paid fo r them or that m ay be “im puted” to them .1 Econom ic production serves econom ic welfare, w hich can be defined objectively in term s of (a) the quantity and value of goods and services produced th a t yield satis­ factions (considered in relation to the quan­ tity and nature of available free goods); ( b ) the effectiveness of the use of these goods and services by their recipients for the satisfaction o f their w ants; (c) the pleas­ urableness, if any, of the w ork involved in their production; and ( d ) the am ount of leisure left to the producers after the com ­ pletion of their tasks. N oneconom ic production, on the other hand, consists of the production of goods and services, which, though satisfying h u ­ m an wants, do not possess “econom ic value” in the sense described above. T he values of these goods and services m ay be spiritual, psychological, social, or political. T hey may be just as great as, or even greater than, the econom ic values of the other goods, but simply belong to another sphere o f hum an life, one dom inated by another set o f hum an values. In this category belong the “hobbies” of individuals, their self-services (w hether in washing, dressing, shaving, eating, exer­ cising, o r w hat n o t), th e services rendered by m em bers of the fam ily to one another on the basis of the ties that unite them , and those exchanged by neighbors on a friendly basis. Tim e and again goods and services trad i­ tionally belonging to the noneconom ic cate­ gory cross over into the other category as the task of furnishing them is taken over by com m ercial establishments or by govern­ m ent. Com m ercial laundries, tailoring, and cleaning establishments take over the house­ w ife’s job of washing, cleaning, and m end­ ing linens and clothes. F ood is increasingly consum ed in restaurants rath e r th an at hom e. Ladies go to beauty parlors to have their h air done. W indows are cleaned by

window washers. G arbage is rem oved by m unicipal agencies. Because the scope of eco­ nom ic production widens constantly, while th a t of noneconom ic production grows n a r­ row er, it becomes difficult to m ake m ean­ ingful com parisons o f the econom ic produc­ tion of a given country over a period of tim e, or of two or m ore unequally developed countries during the sam e tim e. In any such com parison it is necessary to take into ac­ count existing differences in the ratio of econom ic to noneconom ic output. T he need for b oth econom ic and non­ econom ic production to provide a given am o u n t of hum an w elfare is determ ined largely by the availability of the free gifts of n ature. Since these are unevenly distrib­ uted over the surface of the earth, being very abundant in some regions and scarce in others, the need for both noneconom ic and econom ic production varies greatly from one region to another. T h e production of h eat by m an, w hether fo r its own use or for exchange, is necessary only to the extent th a t sun h eat is insufficient fo r bodily com ­ fort, and consequently is needed to a greater extent in n orthern th an in southern regions. Clothes an d shelter sim ilarly are required in different degrees in different climates. A sarong is sufficient in the South Sea Islands, b u t it w ould hardly do in the A rctic. A hut m ade o f bam boo sticks and palm leaves provides sufficient shelter in the tropics, but sturdier constructions em ploying stone or tim ber are required fo r shelter in northern regions. T h e construction of wells o r w ater­ w orks is necessary to the extent th a t spring w ater is unavailable o r th a t rivers and lakes are polluted. A co untry characterized by endem ic diseases requires greater m edical care th a n one subject only to epidemics. T h e contributions th a t b oth econom ic and noneconom ic production m ake to hum an w elfare can, therefore, be evaluated only in term s of w hat natu re does o r does n ot p ro ­ vide fo r hum an welfare. M oreover, in evaluating the relative con­ tributions o f econom ic and noneconom ic p roduction to hum an w elfare, one should also tak e into account the offsetting costs of the lab o r involved, m easured in term s of the w ear and te ar on m ind an d body and o f th e sacrifice of leisure. In this respect, too, conditions vary greatly from one region to another, as the m ore bountiful natural resources o f one region will yield a larger p ro d u ct fo r the same effort th an the nig­ gardly resources o f another. A given q uan­

1 1 . BASIC CO NCEPTS

tity of goods obtained w ith little effort ob­ viously contributes m ore to the individual’s w elfare than the sam e quantity obtained at great exertion. A nation m ay produce only a lim ited quantity of econom ic goods, but because of the mildness of nature, abundance of free goods, and certain institutional factors, these goods m ay go fa r tow ard satisfying its peo­ ple’s physical and other needs. M oreover, because of the bounty of nature, these goods m ay be obtained w ith an extraordinarily small am ount of exertion, thus affording people a large am ount of leisure. In addi­ tion, because o f these sam e natu ral and in­ stitutional factors, the citizens m ay value the noneconom ic goods m ore highly th an the econom ic ones. T hey m ay prefer the enjoym ent of sunlight, beautiful landscapes, play, sports, and social intercourse to the possession of m aterial w ealth. In term s of p ure econom ic w elfare, then, a people m ay exist on the edge of outright m aterial pov­ erty, yet this condition m ay be m ore than offset by their possession of noneconom ic values contributing im m easurably to their general well-being and happiness. O n the o ther hand, another people characterized by the exactly opposite set of n atu ra l and insti­ tutional conditions m ay, even w ith the same econom ic output, enjoy a m uch sm aller am ount of econom ic w elfare, and a still sm aller am ount of hum an happiness o r w el­ fare generally. It is obvious, therefore, th a t national in­ come, though constituting a fair m easure of econom ic production, is n o t an altogether adequate m easure o f econom ic w elfare and is an altogether am biguous and misleading m easure of w elfare generally. N ational incom e as a m easure o f eco­ nom ic w elfare is particularly deficient w hen one desires to com pare the econom ic w el­ fare of a people at one period w ith its eco­ nom ic welfare at another, o r to com pare the econom ic w elfare of different nations during the same period. In either com pari­ son account m ust be taken of th e differences th at m ay exist in the am ount o f effort re ­ quired to produce a given o utput and in the availability of free goods. T he difference in effort required to p ro ­ duce a given output m ay be expressed quan­ titatively, and may, therefore, be accounted fo r w hen com paring the national incom e of a country during different periods, or the incomes o f different countries during the sam e period. T hus, K uznets, in his estim ate

165

o f the grow th of the national product of the U nited States from 1870 to 1950, increased his output value figures fo r the later years by an am ount proportionate to the reduc­ tion o f the w orking hours th at took place in the U nited States during th a t period.2 B ut differences in the availability of free goods cannot be given a definite quantita­ tive expression in any intertem poral o r inter­ regional com parisons of national incomes. A ll th a t can be done is to indicate th a t the differences in econom ic w elfare between the two periods in the life of the country or betw een the tw o countries at a given tim e cannot be as great as suggested by their respective national incom e figures; or, on the contrary, th a t they m ust be even greater th an suggested by these figures, because in the one case the higher incom e is partially offset by a lesser availability o f free goods while in the other case the higher income is also attended by a greater availability of such free goods.

b. The Social Nature of the Concept T he concept of national incom e is social in natu re and differs fundam entally from the private concepts of individual, family, o r group incomes. Its concept is rooted in the general theory of econom ic production, consum ption, saving, investm ent, and eco­ nom ic w elfare, an d cannot be understood apart from it. F irst o f all, unlike private income, n a­ tional incom e is concerned alm ost entirely w ith the acquisition of goods and services. M oney plays a sm aller role in it th an in private incom e, being regarded as a circu­ lating m edium and n o t as w ealth in itself. Secondly, national incom e includes certain receipts of m oney o r of services and goods that are n ot ordinarily included in private incom e, and it excludes certain others th at are generally included. T hus, it includes em ployers’ paym ents into social security and w elfare funds fo r the benefit of their em ­ ployees, profits o f governm ent enterprises, and services of ow ner-occupied dwellings, b u t it excludes interest received on w ar bonds, social security benefits, and pensions. N atio n al incom e is, therefore, n ot a m ere aggregate o f private incom es and m ust never be confused w ith it. It is possible to derive the one aggregate from the other by elim inating certain items and adding others, and this, in fact, is often done; b u t this only serves to accentuate the im portance of clearly distinguishing between the two.

166

THE INCO M E OF NATIONS

T he problem of deciding w hich receipts of m oney or of goods and services to in ­ clude in the concept and com putation of national incom e and w hich to exclude is often difficult to resolve. G enerally, the question is decided by the estim ator on the basis of the theory of econom ic production, consum ption, saving, and investm ent th at he follows or th a t is entertained generally am ong the economists of his country at the tim e. Sometimes, however, when theory fails to give a clear-cut answer, o r m eets w ith serious practical difficulties in application, the estim ator decides on the basis o f expe­ diency. Such practical solutions of difficul­ ties or simplifications of the estim ating p ro ­ cedures m ay eventually become widely ac­ cepted and even assum e the nature of con­ vention. H ow ever, they should be recog­ nized for w hat they are, nam ely escapes from theoretical and practical difficulties. Since W orld W ar II, significant progress has been m ade on an international level tow ard the establishm ent of a com m on agreem ent on w hich items of production and consum ption to include and w hich to exclude from national income concepts and estimates. In m ost cases these agreem ents grew out of a com m unity of econom ic thinking, but in a few cases they, too, have been founded upon m ere considerations of convenience. These agreem ents have never­ theless produced m ore com parable estimates and have encouraged a great m any m ore countries to undertake the w ork o f regular annual estim ation of their national incom e.

c. National Income as a Double Flow: That of Product and That of Money Income N ational incom e is both a flow of goods and services and a flow o f money incomes. It is, therefore, called national product as often as national income. T he flow of n a­ tional incom e is circular in nature. It begins in the production units that engage capital and labor, tu rn out goods and services, and distribute m oney incom es to the participants in production. These incomes are then spent by the producers and their dependents in the acquisition of goods and services. As these goods and services are consum ed or returned to production as final investm ent goods, the capacity of the producers to engage in another round of production is being reconstituted. A t the sam e tim e, the m oney that has been spent on these goods

and services com es back to the producing units as receipts fro m sales or as invest­ m ent funds, thus rebuilding their ability to finance another ro u n d o f production and start the cycle over again. T hus the two flows— one m onetary and the oth er physical — are com pleted at the end of each round. D iagram 11-1 shows their nature and rela­ tion to each other in a graphic way. These two flows can be analyzed sepa­ rately; b u t analyzing them as com ponents o f a single or com bined “incom e-product” flow is m ore fruitful. In this flow three suc­ cessive stages or phases are discernible: (a) the production of goods and services by producing units (including governm ent and nonprofit-m aking in stitu tio n s), through the in p u t by them of capital an d labor; (b ) the distribution of incom es (wages, profits, etc.) to the factors of production an d the subse­ qu en t redistribution of these incom es by governm ent and o th er agencies (through taxes, transfer expenditures, etc.) to the n o nfactor individuals an d groups; and (c) the expenditure of incom es fo r final con­ sum ption and investm ent goods an d the consequent reproduction o r enhancem ent of th e currently used-up productive powers. D iagram 11-2 shows these three phases in the consolidated incom e-product flow. E ach such round of production, incom e distribution, and expenditure (o r consum p­ tion and investm ent) appears com pleted and closed at the ends only w hen viewed w ithin a single period. T he same rounds w hen viewed across tim e as they succeed each other, appear open at th e ends an d connecting w ith each other: each has its origins in the preceding round and, in tu rn , leads into the succeeding one. All of them tak en together form a never-ending, continuously revolving econom ic process, as illustrated by D iagram 11-3. T he successive rounds grow larger or sm aller depending on w hether a larger or sm aller am ount of resources is carried over into them from previous rounds o r injected from outside; and also on w hether these resources are used m ore efficiently o r less efficiently th an they were used before. T he long-run tendency of the rounds of production, distribution, and expenditure is to increase and for the national incom e as a w hole to grow, reflecting the underlying tendency o f the population and o f the p ro ­ ductivity o f society’s capital and labor to increase. B ut in the short ru n , these rounds an d the incom e-product flows em bodied in them m ay fluctuate considerably.

11.

BASIC CONCEPTS

167

liimplified illustrations of the Formation, Distribution, and Disposition of the National Income or Product (omitting influences of credit expansion and contraction, hoarding, etc.)

Diagram 11-1. The Two Circular Monetary and Physical Flows

Diagram 11-2. Combined Income-Product Flow and Its Three Phases

Diagram 11-3. The Economic Process Viewed Across Time: the Successive Rounds of Production, Income Distribution, and Final Expenditure (Consumption and Investment)

168

THE IN CO M E OF NATIONS

2. The Structure of National Income a. Combination of Two Types of Magnitudes T he national econom y is th e econom ic organism o f society. It is the aggregate of society’s producing and consum ing units viewed in term s of their interrelated activi­ ties. N ational incom e is the m ost unified and com prehensive expression of these op­ erations, and its structure varies directly w ith the aspect of the econom y being con­ sidered. Thus, if the national econom y is considered as an aggregate o f producing units m aking up the different branches of production, such as agriculture, mining, m anufacturing, etc., then the national in­ com e estim ate is constructed accordingly as the net output o f these branches o f produc­ tion. W hen constructed in this w ay, and com pared over a period of tim e, it m eas­ ures changes in th e contributions of these branches of production. W hen, on the other hand, the national econom y is considered as a com bination of producers of various types— some owning capital, others earning incom e entirely from their labor, still others, farm ers and the like, earning incom e partly from capital and partly from labor— then the estim ate of n a­ tional incom e is constructed accordingly as the sum o f these distributive shares. T aken over a period of tim e, it m easures the changes in these shares, reflecting either variations in the productivity o f these groups o r changes in their respective b ar­ gaining powers. Finally, w hen the national econom y is considered as a collection of consuming, saving, and investing units (individuals, families, public and private collectives), and enterprises, then national incom e is con ­ structed as an aggregate o f final products, classified as consum er durables and n on ­ durables and as net investm ent goods. Changes in the size and structure of the national incom e taken over a period of tim e in this case, m easure th e changes in the econom ic w elfare of the w hole society and o f its various groups. M oreover, n a­ tional income, considered in these term s, is related to the size of the population and is expressed as a per capita am ount. N ational incom e quantitatively expressed is always a com bination of tw o types of m agnitudes— the num ber of units o f vari­

ous types com prised in the national econ­ om y an d involved in its operations, and the values th a t are attached to these units and th eir operations. W hen constructing n a­ tional incom e estimates by aggregating net ou tput in the various branches of produc­ tion, the units are the producing units in these branches and the values are th eir net outputs. In the second presentation of n a­ tional incom e, the units are th e individual producers— p roperty owners, wage earners, etc.— and the values are th e incom es earned by them . In the th ird presentation, the units are the final goods an d services, and th e values are th eir m ark et prices (which m ay be reduced to facto r costs by an over­ all deduction of indirect tax es).

b. Importance of a Synthesized ThreePhase Treatment: Production; Distri­ bution; Disposition N ational incom e can thus be m easured in any one o f th e foregoing three ways, but if a thorough analysis o f th e econom y is the object, it should be m easured in all of them sim ultaneously. T he division o f national incom e analysis into th ree related phases is, how ever, a com paratively recent developm ent, having been introduced only a q u arter o f a century o r so ago. Before th en econom ists usually concentrated on a single phase an d treated it in isolation from th e others. T h e early British and F rench econom ists (P etty and Boisguillebert particularly) viewed national incom e prim arily from the consum ption point of view as an aggregate o f consum able goods and services. T he later eighteenthand nineteenth-century scholars (Y oung and C olquhoun in E ngland and T u ck er in the U nited States) stressed its production as­ pects, viewing it as an aggregate o f the o ut­ puts of different branches of production. T h e still later econom ic analysts (Bowley in E ngland, Helfferich and W agner in G er­ m any) em phasized its distributive aspects, treating national incom e as th e sum of dis­ tributed factor shares. Even in recent times, some economists analyzed only one o r two phases of national incom e. In E ngland, J. R. H icks, as late as the 1940’s, recognized only two phases o f national incom e— the aggregate of facto r shares as “social incom e” and the n et value o f the final products as “n et social o u tp u t.” 3 In the U nited States, C arl Shoup, w riting at about the sam e tim e, similarly distinguished only betw een national income

1 1 . BASIC CONCEPTS

as an aggregate of factor payments and national product as a sum of final product values.4 The United States Department of Commerce, as late as 1951, similarly distin­ guished only between these two phases of national income.5 One of the first to emphasize the impor­ tance of combining all three phases of na­ tional income in a synthesized analysis was Joseph Schumpeter, in an article published in Germany in 1917.6 The Reich’s Statis­ tical Office in 1932 referred to these three phases as “Wertschopfung,” “Verteilung,” and “Verwendung”; 7 and in the United States, Wesley C. Mitchell and Simon Kuznets, in a joint paper published in 1935,8 spoke of the “three definitions of national income, corresponding to the three stages at which we may choose to measure the flow of goods in the economic system: na­ tional income produced, national income paid out, and national income consumed.” In England, Meade and Stone urged a tri­ partite analysis in their schematic presenta­ tion of national income published in 1941 in terms of “net national income at factor cost,” “net national output at factor cost,” and “net national expenditure at factor cost.” 9 In France, Perroux, in his book pub­ lished in 1947, distinguished between “les trois optiques dominantes” in which national income appears as “produit,” “revenu,” and “emploi (depense).” 10 Finally, the Sta­ tistical Office of the United Nations, in all its annual publications on national income, has been defining national income in three ways as “net national product,” the “sum of distributive shares,” and “net national expenditures.” 11 The desirability of analyzing national in­ come from all three points of view now seems to be generally accepted. It is ana­ lyzed in this fashion throughout this volume. For the moment, further consideration will be given to the structure of the national income in each one of these three phases.

3. National Income as an Aggre­ gate of N et Outputs The production units in which goods and services are produced are variously classi­ fied, in different countries, according to spe­ cific branches of production. But the fol­ lowing classification into fourteen branches is most common. It represents a compro­ mise between more detailed classifications and more rudimentary ones.

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

169

Agriculture Mining Fishing Construction Manufacturing Trade Transportation, communications and other public utilities (electric, gas, etc.) Finance (banks, insurance companies, etc.) Ownership of dwellings (rented and owner-occupied) and of other real estate Service industries, trades, and arts Professions Domestics Government (administration and enter­ prises) Private nonprofit-making organizations (churches, private universities, founda­ tions, etc.)

F rom a social and econom ic po in t of view, these branches can be, and in some analyses are, com bined in the following five types of producing organizations: 1. Private business enterprises 2. Households employing domestics and often occupying their own homes 3. Government services 4. Government enterprises 5. Nonprofit-making private organizations12 T he o u tput of goods and services flowing from each b ranch of production is the sum o f the outputs of all the separate producing units in th a t branch. It is first evaluated at the m ark et prices o r costs o f the goods and services involved. T he value so com puted in each case is called “gross” because it contains m any duplicated items, such as the costs of interm ediate goods furnished by other branches of production and already evaluated once. It is called “gross” also be­ cause it is com puted w ithout deducting the depreciation or replacem ent costs o f the capital used in production. In other words, it includes the results o f the production of past years along w ith those o f the current year. T o obtain the net value of output, i.e., the value o f the contributions m ade to p ro ­ duction by the capital and labor em ployed in any one branch during any given year, it is necessary to com pute both the duplicated operating costs and capital consum ption al­ low ances and to deduct them from gross value. T h e resulting figure represents in­ com e originating in th at branch of produc­ tion during the year. All such net values represent national incom e classified accord­ ing to its origin.

170

THE INCO M E OF NATIONS

The manner in which gross values of output in the different branches of pro­ duction duplicate each other is shown in Table 11-1. In this illustration, price-in­ creasing taxes are not included because they represent a special problem requiring separate discussion.

ing units are separated into capital incomes and labor incomes. In the second, they can­ not be differentiated as easily, inasmuch as the owners act in the double capacity of workers and entrepreneurs. These incomes are mixed, being composed partly of capital income and partly of labor income.

Table 11-1. The Gross and N et Values of Output a t Different Stages of Production (Exclusive of Price-increasing Taxes)

300 800 1,400 1,900

Materials No. 2 100 120 100 100 20

Services Purchased 50 80 80 100 80

4,400

440

390

Stage of Production 1 2 3 4 5

Materials No. 1

Total

_

It will be observed in this table that the gross value of output of the five stages or branches of production amounts to 6,900, but that the corresponding net value, or “value added,” amounts to but 1,300, or less than one fifth of the gross figure. The computation of net value from gross value in this illustration would be as follows:

Net Value Added (services Capital Total Consump­ of capital Value of and labor) Production tion 50 100 120 50 50

100 200 300 250 450

300 800 1,400 1,900 2,500

370

1,300

6,900

This threefold division of incomes is peculiar to the capitalistic system, in which enterprises are organized both into corpora­ tions in which clear distinctions are main­ tained between owners of capital, managers, and labor, and into partnerships and per­ sonal proprietorships in which such distinc­ tions are not clearly maintained. Deductions of Depreciation

300 800 1,400 1,900 2,500

Deductions of Duplicated Costs 150 500 980 1,600 2,000

50 100 120 50 50

100 200 300 250 450

6,900

5,230

370

1,300

Branch of National Production

Gross Value of the Output

1 2 3 4 5 Total

This net value represents the cost of the services of capital and labor employed in these branches of production during the period. It is represented directly in the dis­ tribution phase of the national income as factor payments.

4. National Income as an Aggre­ gate of Factor Shares The factors of production— capital and labor— are either hired by the producing units, or, where the latter are constituted wholly of owners, are furnished directly by these owners themselves. In the first in­ stance, the incomes paid out by the produc-

Net Value of the Output

Not all the incomes earned by producers are paid out to them by the producing units. Some are retained or set aside for payment to them at a later time or in a dif­ ferent form. Of this nature are: (a) retained profits of corporations that are reinvested with a view to increasing their future earn­ ing power for the ultimate benefit of their stockholders; ( b ) employers’ contributions to social security, pension, and other wel­ fare funds, and to workmen’s compensation accident insurance; (c) direct pension pay­ ments to retired workers; (d ) profits of government enterprises; (e) land improve­ ments, increases in livestock, and growth of forests. The profits taxes paid by corpo­

1 1 . BASIC CO NCEPTS

rations on behalf of their stockholders and, presumably, out of their shares, are also a form of retained capital income. All incomes are not paid in cash: some are paid in kind. The housing, feeding, and other maintenance services provided to members of the armed forces and to civilian institutional employees fall into this cate­ gory, as do the farm products consumed on the farm and the services of owner-occupied dwellings. In some cases incomes are dupli­ cated and must be reduced to a “net” figure before they can be included in national in­ come. Of this nature is bank interest paid to business depositors who are also bor­ rowers. The principal types of income can be classified under the three main heads men­ tioned, as follows:

I. Labor incomes 1. Salaries and wages, including cash bonuses, commissions, etc. 2. Supplements to labor income (em­ ployers’ contributions to social security and employees’ welfare funds; and direct pension pay­ ments to retired employees) 3. Compensation in kind (mainte­ nance, free health and educational services, etc.) II. Capital incomes 1. Dividends 2. Undistributed profits of corpora­ tions before taxes 3. Interest on bonds, mortgages, and savings deposits 4. Interest earned by insurance com­ panies and credited to the insur­ ance policy reserves 5. Net interest paid out by commer­ cial banks 6. Net rents from land, buildings, etc., including imputed net rents on owner-occupied dwellings 7. Royalties 8. Profits of government enterprises III. Mixed incomes 1. Profits of the self-employed, whether taken in cash or retained in the enterprise (often referred to as profits of unincorporated enter­ prises) 2. Incomes of the self-employed in kind (farm products consumed on the farm, services of farm dwellings)

171

Nearly all incomes earned by producers, whether they come from capital, from labor, or from both, eventually become the in­ comes of households. Even employers’ con­ tributions to social security, pension, and other welfare funds can be considered as such, for they are regarded by the house­ holds as part of their total, long-run, spend­ able resources. The only incomes that do not belong in the household category are the retained profits of corporations, the profits of government enterprises, and in­ vestment incomes of nonprofit-making insti­ tutions, which are disposed of directly by the corporations, by government, and by the institutions and over which the house­ holds have no control. In the final analysis, therefore, incomes earned by producers from whatever source can be reclassified into three types: (a) in­ comes of households; ( b) retained profits of corporations and other businesses before taxes; and (c) prpfits of government enter­ prises and investment income of nonprofitmaking institutions.13 In this three-way breakdown, national income at factor cost enters the disposition phase discussed in the next section.

5. National Income as the Aggre­ gate Value of Final Products at Factor Cost and Also at Market Prices The incomes of households, undistributed business profits, the profits of government enterprises, and the investment incomes of nonprofit-making institutions are disposed of in the purchase of final products that consist of: (a ) private consumer goods and services; ( b) net private investment or capi­ tal formation; and (c) government and in­ stitutional goods and services, both current and capital.

Expenditures for consumer goods and services measure the personal or family levels of economic living or economic wel­ fare. They are classified by types of goods and services as expenditures for food, cloth­ ing, furniture, rent, travel, recreation, edu­ cation, medical care, etc. These outlays do not coincide exactly with the value of cur­ rent consumption, inasmuch as the latter includes the services of existing durable goods that were previously acquired, and does not include in full the services of newly acquired durables that will be con­

172

THE IN CO M E OF NATIONS

sumed mainly in future years. However, the purchase of consumer durable goods, other than dwellings, is generally classified as consumption expenditure, rather than as capital investment (for reasons that are more fully explained in Chapter 12, Section Id ), resulting in a disparity between current consumption outlays and the real value of current consumption. For certain analytical purposes, purchases of consumer durable goods are sometimes lumped with capital formation. N et private investment consists of the net addition to private plants and other struc­ tures, producer’s equipment, inventories, land improvements, and the growth of natural capital (livestock, forests). Expenditure on government goods and services consists of the compensation paid to government and institutional employees (and, in some estimates, interest paid on productive government loans) as well as the value of government and institutional purchases of goods and services from pri­ vate business. They are also classified ac­ cording to type as current operating ex­ penditures and capital expenditures. The first are further functionally subclassified in the case of government as expenditures for general administration, public education, health services, national defense, mainte­ nance of highways, etc.; while capital ex­ penditures are similarly subclassified by their objects as those for public buildings, schools, highways, bridges, sewer systems, hydroelectric plants, and other tangible as­ sets. In government reports, the costs of re­ placing existing government properties are lumped together with the costs of net ad­ ditions to government properties. In any computation of the disposition of national

income at factor cost for current consump­ tion and net capital formation, it is neces­ sary, therefore, to deduct from reported government capital outlays, the depreciation of government property.14 Final consumer and net investment prod­ ucts serve not only immediate but also fu­ ture needs. All final goods leave the produc­ tion sector to serve the current needs of the population or to increase their stock of wealth for greater future production and consumption. The aggregate value of these final products at factor cost coincides with the sum of all factor shares and also with the aggregate of the values added. Hence, in all three phases of national income, it is possible to obtain its computation at factor cost.15 At the same time it must be observed that the evaluation of final products at factor cost does not lend itself statistically to as detailed a breakdown by type of product as national income at market prices, and to that extent is only of limited practical value. The known values of the final products are expressed in market prices that include in­ direct taxes. It is possible to separate in­ direct taxes from factor costs in the aggre­ gate value of the final products, but not al­ ways in the values of individual products or types of products. This is so because in­ direct taxes are applied throughout the whole economic process, and their incidence at any one point in the process is uncertain. Their price-increasing effects cannot be de­ termined separately for each product with any certainty.10 Therefore, in all detailed analyses of the disposition of national in­ come by the type of final product, it is generally necessary to shift from the concept of net national income at factor cost to that

T ab le 11-2. T h e T h ree Phases of N a tio n a l Incom e a t F a c to r C ost a n d th e S tru c tu ra l E lem ents a n d

Estimating Methods Peculiar to Each Phase Production

Nature of the Aggregate Net value of output at factor cost

Distribution Disposition

Aggregate of factor shares Net value of final products a t factor cost

Structural Elements Net outputs of various branches of production at factor cost plus net in­ come from abroad Types of factor shares Types of final consumption and net investment prod­ ucts a t market prices, less indirect taxes (net of subsidies)

Method of Estimating Net output

Income distributed Expenditure for final prod­ ucts, sometimes called simply “final products” method or “consumption and investment” method (in this work often re­ ferred to as “income dis­ posed” method)

1 1 . BASIC CO NCEPTS

of net national income at market prices. The equivalence between national income and actual national expenditure on final products is obtained only when indirect taxes are added on the income side to the sum of factor shares. The practical difficulties involved in break­ ing down national income at factor cost by type of expenditure for final products in no way diminishes the great theoretical validity of this most important of national income concepts. National income at factor cost is the true objective of all national income estimates. The designation “factor cost” has been criticized because in the ac­ cepted business and accounting terminology, net profits are not a production cost. The critics, however, fail to perceive the differ­ ence between business and social account­ ing. To society at large, net entrepreneurial profits are just as much a “cost” of the prod­ uct it receives as are wages or interest.

6. Net Income from Abroad To simplify our analysis we have thus far assumed a “closed” national economy, i.e., one having no economic relations with other countries. This assumption is, of course, wholly unreal, as all countries have some economic relations with the outside world, and in some these relations are ex­ tensive. Theoretically, national income or product must always include net income earned from activities carried on abroad. Thus, according to a recently formulated standard concept, the resources of a national economy consist of labor and capital sup­ plied by normal residents of a given country whether located within or without its terri­ tory.17 The normal resident category includes commuting workers employed outside the national territory as well as members of the armed forces and national diplomatic and consular personnel stationed abroad. The labor incomes in question should be allo­ cated to the country from which the re­ cipients come, not to the country in which they work. International capital income con­ sists of: (a) profits derived from the opera­ tion of branch plants and offices abroad and from the ownership of foreign subsidiaries; (b ) dividends on the ownership of foreign stocks; (c) interest on foreign private and government bonds, foreign mortgages, and savings accounts; and (d) rents and royal­ ties from foreign properties and rights. In­

173

come from abroad includes not only the amounts received but also those that are accrued. But since such labor and capital incomes not only flow in from abroad but also flow out to other countries, only the excess of the former over the latter can be included in the national income total. Where instead of an excess there is a deficit, net income from abroad becomes a negative quantity, which reduces the home-produced income instead of increasing it. This hap­ pens when a country imports more than it exports, or when it is heavily in debt to other countries and must pay large amounts in interest and principal repayments. However, because of insufficiency of data, net national income from abroad is seldom fully accounted for in existing national in­ come estimates. The overwhelming majority of estimates take the whole domestically paid labor income without deducting from it labor incomes earned by nonresidents or adding to it labor incomes earned abroad by the country’s normal residents. To what extent the salaries of diplomats stationed abroad are included in the estimate is often not stated. Actually the amounts are often not made public. Even the addition— posi­ tive or negative— of international capital in­ come is usually unsatisfactory because of insufficient statistical data. Undistributed profits of foreign subsidiaries are especially poorly reported. Some estimates are concerned with homeproduced income only. This was true, for example, of the Turkish estimates made be­ fore World War II. Some, such as those for Italy before the war, included “unilateral” receipts, such as remittances of immigrants from abroad and reparation payments and donations, which are not factor shares and do not belong in the evaluation of the na­ tional product. Some of the prewar Italian estimates even included the entire net bal­ ance of international payments. The definition of the “resident individual” in the United States estimates comes closer to the international standard concept. The definition runs as follows: United States Government employees, civil­ ian and military, whose usual residence is in the United States are considered United States residents even when stationed abroad. Resi­ dents of foreign countries hired abroad by the United States Government are not considered United States residents. This treatment applies specifically to employees of foreign govern­ ments in the United States. In principle, foreign

174

THE IN CO M E OF NATIONS

Table 11-3. Identity of the Concept and Aggregate of the N et National Income a t Factor Cost in All of Its Three Phases Production

Distribution

Net National Product by Economic Divisions

National Income by Distributive Shares

1. Home produced income: N et output of: Agriculture Mining Manufacturing, construction, etc. Transportation Trade Finance and real estate Services of dwellings Government Professional and other services

Income of normal residents from domestic and foreign ac­ tivity and investments:

2. Plus net factor incomes from abroad ( + or —)

1. Labor income Wages Salaries Supplements Compensation in kind 2. Capital income Interest (net) Rent (net) Dividends Undistributed corporate profits Profits of governmental enterprises Profits of co-operatives, etc. 3. Income of the self-employed

National income a t factor cost ( 1 + 2 )

National income a t factor cost . ( 1 + 2 + 3)

border workers working in the United States are not regarded as United States residents. United States residents whose place of work is across the border are so regarded. As ex­ plained, however, the statistical information for implementing the definition of residents with respect to border workers is lacking. Also, there are marginal cases in which it is difficult to decide whether an individual is to be classi­ fied as a resident or as a foreigner. In practice, these cases are quantitatively unimportant.18

7. Summary National income at factor cost is theo­ retically the most important concept of na­ tional income. In fact, it is national income in its truest sense and is often, therefore, referred to without any additional qualifi­ cation as “national income.” It is the most valid concept because it measures national income in terms of the value of the current services of capital and labor, minus the con­

Disposition National Consumption and Investment Expenditure at Factor Cost 1. Expenditure of normal resi­ dents a t home and abroad for final products a t market prices (GNP): Personal expenditure on con­ sumer goods and services a t home and abroad Domestic private net capital formation Public expenditure on cur­ rent services Domestic public net capital formation Net increase of assets abroad 2. Less indirect taxes Plus subsidies 3. Less depreciation National product a t factor cost .(1 + 2 + 3)

sumption of existing capital and net of any duplicated costs. But national income at factor cost is also an ambiguous concept, inasmuch as it depends upon the nature of the country’s tax system and upon inter­ pretations of government products as either final or intermediate that are questionable. As it most generally excludes all indirect taxes, it is of little use in certain types of economic analysis. More will be said about this difficulty in Chapter 14. National income has three aspects—pro­ duction, distribution, and consumption— and it can and should be estimated and analyzed in each. But its estimation in each phase requires different statistics and differ­ ent statistical methods. If these data and methods, however, are equally accurate, they should yield identical totals for each phase. This identity of the three basic na­ tional income concepts and aggregates is brought out in Tables 11-2 and 11-3.

12. THE AREA OF PRODUCTION In Chapter 1, we traced the emergence, after a long period of evolutionary development, of two main concepts of national income: the comprehensive production concept, which defines national income or product as an aggregate of both commodities and services produced over the year; and the material production concept, which defines it as an aggregate of material goods only. A third but comparatively unimportant con­ cept exists in addition, the market produc­ tion concept, according to which national income or product consists of both material goods and services, but only to the extent that these are produced for and are dis­ tributed through the market. The main dif­ ferences between the composition of these three production concepts are shown in Table 12-1.

1. The Comprehensive Production Concept The comprehensive production concept recognizes the distinction between material goods and immaterial ones (services) as be­ ing significant in certain types of economic analysis, but considers it completely irrele­ vant to the central issue of what constitutes production and national income and what does not. Production is deemed to be a process concerned with the creation of utili­ ties having economic value, i.e., things capable of satisfying human wants and hav­ ing a determinable economic price or cost. These may be either material or immaterial in nature. Services— which may be either those of persons or of durable capital goods — are declared to be just as capable of satis-

Table 12-1. Composition of the National Income According to Three Production Concepts

Market goods Certain goods produced within the house­ hold for own or family consumption M arket services Services of dwellings Civil government and military services Profits of government enterprises Direct taxes Indirect taxes

Compre­ hensive x

Restricted Market X

X

X

0

0

X X

X

0

xc

X X

0 “ X 0 “

0 0 X X

x

xb

X 0

Restricted Material

x: Item included. 0: Item excluded. a In the only estimates based on this concept, the Kalecki and Landau estimates for Poland, the cost of public education is included. On the other hand, individual incomes are deemed to exclude personal taxes. b Certain professional services are excluded in the Matolcsy-Varga estimates for Hungary. c Only rented dwellings.

fying human wants as directly consumable material goods and, therefore, are included in production and national income. This conclusion is further supported by reference

In this chapter, each of these three con­ cepts will be analyzed and the various issues arising out of their application will be criti­ cally reviewed. Notes begin on p. 529. 175

176

THE IN C O M E OF NATIONS

to the fact that services and material goods continually interchange in the market; that the creation of both requires the investment of capital and labor; and that the satisfaction of many wants often requires the combined use of both, e.g., medical care, where the services of the surgeon are used in conjunc­ tion with such specialized material goods as instruments. Many material goods have no clearly identifiable use value without some accompanying services, and in order to ac­ count for their contribution to economic welfare, it is necessary to include the serv­ ices they implement in the national income total. The labor engaged in the production of services is deemed to be just as produc­ tive as the labor engaged in the production of material goods. Briefly, then, production is viewed as the creation of consumable things and of things facilitating their pro­ duction, whether they are material or not. National income, therefore, includes all freshly produced material goods and services that (a) are created by human labor and capital; (b) are capable of satisfying human wants directly or, as production goods, in­ directly; (c) are comparatively scarce and, therefore, need to be “economized” and have economic value; and (d) either have a definite monetary price or cost or can be given one by imputation. National income by this definition includes (a) all goods and services produced for the market by private and governmental enterprise, including serv­ ices of certain durable consumer goods, such as dwellings; (b) all services produced by government for collective use, although those of the armed forces have been gener­ ally recognized as economic production only since World War II; (c) all goods and serv­ ices produced by nonprofit-making organiza­ tions, such as churches, educational founda­ tions, etc., for the benefit of the general public or of their own members; and, finally, (d) certain goods and services produced by members of the household for their own and one another’s use, outside the market mechanism. National income includes all these goods and services net of capital con­ sumption (or depreciation) and without duplication of values. Within this comprehensive concept of national income exist certain variants re­ sulting from differences in opinion as to whether certain items of unpaid goods and services, particularly certain services ren­ dered within the household and within the government sector, should be treated as

production and hence as national income by imputing money values to them. These un­ paid goods and services may be classified into five main groups: 1. Foods and other goods produced on the farm for the farmer’s own consumption 2. Unpaid personal services of housewives and other members of the family or of broader social groups 3. Unpaid services of owner-occupied dwell­ ings 4. Unpaid services of other consumer dur­ able goods owned by households 5. Unpaid services of tangible wealth owned by governments and by benevolent or­ ganizations

We shall consider each of these items in turn.

a. Inclusion of Value of Foods and Other Goods Produced on the Farm for the Farmer's Own Consumption It is now generally recognized that food and certain other items, such as firewood, which are produced by the farmer for his own and his family’s consumption, repre­ sent the farmer’s income in kind and should be included in national income. Nearly all contemporary estimates do include them in order to obtain comparability of farm with nonfarm incomes and comparability of in­ comes of agricultural countries with those of nonagricultural ones. An additional argu­ ment for their inclusion rests on the fact that the farmer is often in a position to choose between selling his entire produce in the market and retaining a part of it for his own family use. If he chooses to sell everything, he must then buy the produce he needs out of his money income. If, in­ stead, he keeps some produce for his own consumption, he reduces his money income, but, on the other hand, also reduces his expenditures. The retained produce, there­ fore, has a definite money value to him and must be added by imputation to his money income. In some estimates these retained goods are evaluated at the wholesale prices they bring in the market, while in others they are evaluated at the retail prices that have to be paid for them. Of the two methods the latter is preferable because it makes farm income more comparable with nonfarm income. What is important in such comparisons is to obtain a meaningful value for those goods produced and consumed on the farm, and this is better accomplished by evaluating

12.

177

THE AREA OF PRODUCTION

them at retail rather than at wholesale prices. Foods and other goods produced on the farm for the farmer’s own consumption are often considerable. In Ireland in 1953 they represented 26 per cent of the total farm incom e;1 and in Canada in 1950, approxi­ mately 18 per cent.2

b. Unpaid Services of Housewives and of Other Members of the Household The classic example of unpaid personal services rendered in the family circle is the activity of the housewife. Specifically de­ fined, household work, whether accom­ plished by the housewife or other family members, consists of the preparation of meals (kitchen work and serving) and the formation and maintenance of “household capital” (cleaning, washing, mending, tailor­ ing, carpentering, painting, etc.). In giving the goods produced in the market economy the final touch that makes them ready for use, household work is the very last stage of production. It is the remnant of older and more extensive domestic production that has succumbed to the market economy in the course of its development. Household work is easily replaced in the modern econ­ omy by the paid services of commercial es­ tablishments, public services, laborsaving equipment, and paid services of domestics. Some of the increases in national income over time represent no more than this sub­ stitution of such paid services for the unpaid services of housewives. Similarly, some of the differences in the size of the national income of two countries do not necessarily indicate differences in the productivity of their labor but do reflect differences in the extent to which the unpaid services of house­ wives satisfy the wants of households. During periods of accelerated business activity and rising employment or of war­ time labor mobilization, housewives take on jobs in industry without necessarily neglect­ ing their domestic duties. They rely more heavily on outside help for the care of their home needs and pay for these services out of their earned incomes. When peace re­ turns, or in time of depression, they go back to their unpaid home duties. National in­ come, when calculated without inclusion of the unpaid services of housewives, rises in the first instance and drops in the second. The omission of unpaid services of house­ wives from national income computation distorts the picture of both long run and short run changes in the national economy

in addition to making intertemporal and interregional comparisons of national in­ come more difficult. In the face of this situation, most scholars favor, in principle, the inclusion of the un­ paid services of the housewife in national income. The difficulty, however, consists in finding a fair measure of the economic value of the housewife’s services. In a few esti­ mates that have attempted to impute a value to these services, the arbitrary assumption is made that all housewives do the same amount of work and that their services can be evaluated at the same average sum. The assumption is manifestly false. To include such a doubtful calculation in national in­ come would greatly lower the reliability of the total, and the reluctance of most esti­ mators to include this item in their compu­ tations is, therefore, understandable. Table 12-2. Imputed Value of Unpaid Domestic Services in Certain Estimates as Related to National Income

Estimate 1. Addition actually made: Hungary (Mat. Varga, 1930), including paid domestic services Italy (Vinci, 1938) Sweden (Inst. Econ. Res., 1930) » 2. Addition estimated, but not made: Finland (Lindberg, 1930) United States (Kuznets, 1939)

Per Cent Addition (with N.I. computed before addition)

10.7 27.5 20.7 19.0 26.3

a The Swedish estimators, realizing the crude­ ness of their imputation, presented the national income totals both with and without the imputed services of housewives.

For this reason only a few estimates have attempted to impute a value to the services of housewives. These few, as shown in Table 12-2, indicate that we are by no means dealing with a negligible item. The inclusion of housewives’ services can raise national income by 20 per cent or more. Household work is only the most obvious of the many activities carried on in the family that have a counterpart in the mar­ ket. The care and education of children; the “self-services” people perform in shaving themselves, dressing their own hair, acting as their own chauffeurs; gardening, and the

178

THE IN CO M E OF NATIONS

pursuit of other “do-it-yourself” hobbies, such as carpentery, painting, etc., are all of this nature. Professor Corrado Gini pro­ poses to include all such activities in na­ tional income inasmuch as they have a po­ tential market value. He even completes the list, facetiously perhaps, by adding the “matrimonial” services rendered to each other by husband and wife on the ground that these can be substituted for by the use of paid services of male and female prosti­ tutes.3 Logical consistency would demand reach­ ing out even further to include similar free services rendered outside the family, e.g., neighborly advice and co-operation versus paid professional services; the company of a friend versus that of a paid companion; volunteer care of the sick and the poor ver­ sus that by paid nurses and social workers, etc. But such a supercomprehensive concept of national income, taking all these human actions into account, would embrace the entire content of human life and would, for all practical purposes, rob the national in­ come concept of any meaning and render it useless as an expression of economic pro­ duction. c. Inclusion o f Services o f O w ner-O ccu­ pied D wellings Services of rented dwellings are tradi­ tionally included in all comprehensive na­ tional income estimates inasmuch as the renting of dwellings is considered to be no different from operating any other type of business. Depending on the phase in which national income is considered, services of rented dwellings are accounted for either by (a) including the value added in renting or the incomes paid out in renting (net rent plus mortgage interest plus wages, etc.); or (b) by including the gross or cash rent paid by the tenant as the value of the final prod­ uct (services of rented dwellings). Until a few years ago, services of owneroccupied dwellings were not as widely in­ cluded in national income. In some estimates made by the income-distributed method, they were represented only by the mortgage interest paid, and in others, were completely omitted on the ground that no money pay­ ments were made in their instance. This was true, for example, of the estimates made in the United States as late as 1945, as well as in New Zealand and Japan prior to World War II.* This difference in treatment between

rented and owner-occupied dwellings was unsupported by logic as the two performed identical economic services and generally had very similar underlying costs. Individ­ uals frequently can choose between owning and renting a dwelling and they make their decision on the basis of which course ap­ pears to them to be more advantageous. Those who have become owners often have a further choice of whether to occupy the dwelling themselves or to rent it out, and with the resulting proceeds to rent another dwelling for themselves. This interchangeability between rented and owner-occupied dwellings makes com­ pletely unwarranted any sharp differentia­ tion between their services in national in­ come accounting. The exclusion of the serv­ ices of owner-occupied dwellings is unjus­ tified on practical grounds as well as on theoretical ones, for information on the po­ tential rental values of owner-occupied homes can be just as readily obtained from census data and tax records as information on rented dwellings. The English and French economists of the seventeenth and eighteenth centuries, who treated national income primarily as a measure of consumption, were cognizant of the fact that the difference between a rented and an owner-occupied dwelling is immaterial from an economic point of view. They included the rental values of both in their estimates, but in later years their suc­ cessors abandoned this point of view. The issue was finally resolved in favor of the inclusion of the services of owneroccupied dwellings as standard procedure in 1944 at a Washington conference of rep­ resentatives of the official estimating agen­ cies of the United Kingdom, United States, and Canada.5 Home ownership was assumed to be a business producing housing services that are sold to the home owner in his capac­ ity as tenant. This decision influenced most of the subsequent estimates in the world. Only the countries subscribing to the ma­ terial production concept (Soviet Russia and her satellites and Yugoslavia) still continue to ignore this item just as they do the serv­ ices of rented dwellings. Generally, the imputed value of the serv­ ices of owner-occupied dwellings is first es­ timated as gross rent, in line with the gross rentals earned by similar dwellings in the area. Next, in more carefully constructed net output and income-distributed estimates, the owner’s expenses for maintenance, sup­

12.

THE AREA OF PRODUCTION

plies, mortgage interest (wherever possible), indirect taxes (including real estate taxes), and depreciation are deducted from this gross rental figure. The resulting imputed net rent is deemed to constitute the imputed rental income of the owner-occupier, and is included in national income in this form. Mortgage interest paid, wherever separated, is accounted for in the income-distributed estimates under the net interest category of income. On the other hand, in the expendi­ ture or final products estimates, the imputed gross rent on owner-occupied dwellings is included in national income at market prices as the value of the services in question.

d. Treatment of Services of Other Private Consumer Durables There are many other consumer durable goods in addition to dwellings involving sub­ stantial initial investments whose treatment in national income involves a special prob­ lem— motor cars, paintings, musical instru­ ments, furniture, and household appliances, etc. Should the services of these goods be treated as those of capital goods, i.e., as current services spread over the life of the goods, with a portion of their price allocated to each year of use, or should they be treated as services consumed in their totality during the year of the acquisition of these goods, i.e., as direct flows of consumer goods eval­ uated at their full purchase price in the year of acquisition? When these goods are rented, as some of them are, they are treated as capital goods and their services are in­ cluded in national income during the years of their use either at gross rentals or at net rentals; but when they are used by owners, they usually are treated as goods flowing directly to consumers. Their current services in that case are generally not included in national income. In only two estimates— both of them prepared before World War II— that for Sweden prepared by the Insti­ tute for Economic Research for 1930 and that for Italy prepared by Vinci for 1938— were the current services of such consumer durables included in national income. In the Swedish estimate this was done by an exceedingly roundabout method, while in the Italian one it was done by a crude addi­ tion. In both estimates the result added less than one half of one per cent to the total national income. In neither country was the experiment repeated. The theoretical issue, however, is by no means settled. It is still

179

a subject of discussion among national in­ come scholars and practitioners. Kuznets discussed this issue in a paper published in 1952,6 and even made some computations as to what changes might oc­ cur in the national income of the United States if the current services of such com­ modities were included. He found that net capital formation would rise on the average from an 11 per cent share of the net na­ tional income to a 13 per cent share, while the flow of goods to consumers would show an average drop from 88.3 per cent to 86.7 per cent of the national income total. Na­ tional income, computed in constant prices, would be on the average 5 to 6 per cent smaller. Kuznets admitted that his calculations were based on exceedingly debatable as­ sumptions as to the average life of con­ sumer durables (which he set at ten years) and as to their average yield (which he took to be the same as that of prime grade bonds). However, he concluded against any change in the existing treatment of con­ sumer durables on theoretical rather than practical grounds. “Individuals,” he said, “do not in fact seem to treat purchases of consumer durable commodities as invest­ ment similar to that involved in purchasing a house; nor do they, at least in the United States, have as much opportunity to choose between buying and renting consumer dur­ able commodities as they have between renting and buying residences.” 7 It is difficult to agree with Kuznets’ first postulate. Certainly, the purchases of auto­ mobiles, valuable paintings, rare books, fur coats (not even necessarily mink or sable), and the initial furnishings of a home, in­ volving substantial outlays made but once in a number of years, are treated by the average family, whether in the United States or elsewhere, as investments. The an­ nual flow of these goods to consumers is highly irregular and, therefore, is not a proper measure of their current contribu­ tions to the national economic welfare. The flow is often interrupted for a time, as in the case of war or depression; but the con­ tribution to economic welfare assured by these goods is not necessarily terminated therewith, as the existing stock of these goods continues to render services; and vice versa, when the flow is resumed upon the return of peace or business prosperity, eco­ nomic welfare may not rise as sharply inas­ much as the newly purchased goods may

180

THE IN CO M E OF NATIONS

merely be replacing the existing stock which, though worn, still renders services. Oppo­ sitely, when a country’s durable consumer goods have been largely destroyed by bomb­ ing as was the case with Germany in World War II, national income is reduced not merely by the stoppage of the flow of new consumer durables, but much more substan­ tially by the disappearance of the services of the destroyed consumer durables. Con­ sider the case of a nation, a large number of whose population has lost a roof over their heads, all furnishings, and most clothes. Is the drop in the income of such a nation measured by the mere drop in the flow of additional dwellings, furnishings and clothing, or by considerably more than that? Moreover, the replacement costs after a devastating war or other calamity may be exceptionally high because of the resulting scarcity of materials, equipment, and skilled labor. These high costs should be spread over a period of time. If assessed against the current period, they would scarcely con­ stitute a proper measure of the contribution made by the production of these goods to current economic welfare. Finally, some countries, such as Switzerland, have a large inherited wealth in the form of valuable homes, furnishings, jewelry, etc. They do not need to acquire large amounts of addi­ tional wealth to maintain a high level of economic welfare. A computation of na­ tional income that does not take into ac­ count contributions made by the services of this inherited wealth to economic welfare considerably understates the national in­ come of the country. Kuznets’ second observation, referring to the infrequency of choice between owning and renting these types of durable con­ sumer goods, is correct. But it only estab­ lishes the fact that since the renting of these goods is infrequent, the failure to account for their services during the actual time they are rendered to their owners does not create as much of an inconsistency as fail­ ure to account for the services of owneroccupied dwellings. Theoretically, the argument for including the services of consumer durables is un­ assailable, but the difficulties of doing so are immense. They are much greater than those faced in accounting for the services of owner-occupied dwellings. First of all, the variety of goods in question is colossal and information on their nature, quantity, original cost, age, useful life, present con­

dition, and, hence, value, is exceedingly scant. Secondly, most of such goods have no set of readily ascertainable rental value that could be used in the evaluation of their services, as is the case with dwellings. Even if one takes one item— automobiles— on which more information is available than on any other, it would be extremely difficult to estimate their gross rental value because of the variety of their makes, ages, and con­ ditions, and the infrequency with which they are rented. On the other hand, it may be possible at least to estimate the net rent of a number of these items by applying to their capital values an average interest rate or net rent rate. Any computation of the value of the services of these consumer durables would necessarily involve a great many question­ able assumptions, be exceedingly crude, and reduce the reliability of the national income totals. The conclusion must be reached, there­ fore, that until more information becomes available on consumer durables other than dwellings, it would be wise to exclude their services from national income. Until that time, the existing practice of including the values of these consumers’ goods directly at the time of their production and acquisi­ tion appears preferable.

e. Treatment of Services of Government Properties In the case of government properties the same issue arises. Should their services be included in national income? In the case of revenue-producing properties this is gen­ erally done, inasmuch as they are operated as business enterprises with separate operat­ ing and capital accounts, making possible the identification of the value of the serv­ ices of capital invested in them. But in the case of properties whose services are fur­ nished collectively, such as most of the pub­ lic highways, parks, bridges, sewer systems, museums, schools, public hospitals, and ad­ ministrative buildings, not to speak of mili­ tary properties, no such clear separation of capital from operating expense is generally made; or is, at best, made only in part. Gov­ ernment accounts generally are not con­ structed to permit an accurate evaluation of the current services of these properties. There are some significant exceptions to this rule, of which the 1930 Swedish esti­ mate is one. Public administrative buildings were operated there like enterprises, charg­

12.

THE AREA OF PRODUCTION

ing rent to the various governmental depart­ ments occupying them as tenants, and com­ plete business accounts were kept. Where the construction of government properties is financed by earmarked loans, information is available on the interest pay­ ments and sometimes on the annual amorti­ zation or redemption of the debt. These fig­ ures may be taken as indications of the capital yields of the properties and, with considerable reservations, as indications of the depreciation of the properties. On the other hand, in the case of properties whose costs of construction or acquisition are financed from current revenue or from un­ identified portions of general loans— which is frequently the case— no comparable in­ formation exists. Governments do not gen­ erally publish or maintain, in convenient form, information regarding the original costs of such properties, their age, deprecia­ tion, allocated debt interest and debt amorti­ zation costs, etc. Of course, government ac­ counts cover the outlay for maintenance and repair, but these are not necessarily separated from other accounts in published data. Military plant and materiel are seldom accounted for in a way that would clearly distinguish them as durable goods or iden­ tify the extent of their durability. Obsoles­ cence plays an exceptionally large role in their case, and is often difficult to estimate. All these and other factors make the iden­ tification of capital formation in govern­ ment, and the evaluation of the annual services of capital invested therein, exceed­ ingly difficult. It is not surprising, therefore, to find that national income estimators tend to evaluate the contributions of government to national income, unlike those of private business, in terms of labor only, without the services of capital. In several prewar estimates, such as those of Germany, Switzerland, and the Nether­ lands, interest on the national debt was in­ cluded in national income as a measure of the yield of capital invested in the proper­ ties of the national government. In the cur­ rent Canadian estimates the values of pub­ lic buildings, highways, ferries, and certain national inventories are ascertained directly, and interest on a proportional amount of the national debt is included in the national income as a reflection of their capital serv­ ices. In the current Swiss and Danish esti­ mates the same is done for public buildings. The new international standardization of national income estimates proposes the in­

181

clusion of imputed rent on public buildings in the national income total. This is not difficult to do in countries that operate their public buildings like enterprises, charging rent to the various governmental agencies which use them.8 All these expedients cover, at best, only a part of the value of the services of public properties. Complete evaluation would be achieved only if and when government ac­ counts are considerably improved.

f. Treatment of Services of Durable Prop­ erties of Churches and Other Benevo­ lent Organizations The same basic problem exists regarding the services of durable properties of churches and other benevolent organiza­ tions. These services are no better accounted for in national income estimates than those of government properties, yet, in a number of countries where these properties are sub­ stantial, the omission of their services under­ states national income significantly. This situation can be corrected, as the difficulties in accounting for their services are not overwhelming. For one thing, obsolescence is not as great a factor there as it is in government properties.

g. Other Issues In addition to the question of treatment of unpaid services of dwellings, consumer durable goods, the services of government properties and of institutional properties re­ viewed above, many other issues have arisen in the application of the comprehensive pro­ duction concept that require further clarifi­ cation. Among them are the treatment of intermediate services of government, taxes, subsidies, interest on government debt, and services of financial intermediaries. These problems are discussed separately in suc­ ceeding chapters. An entirely different set of issues arises under the material produc­ tion concept of national income, whether of the Smithian or Marxian variety. These will be touched upon in this chapter and in that on Soviet Russia and Yugoslavia (Chaps. 25 and 28).

2. Adam Smith’s Restricted Material Production Concept According to Adam Smith, the first econo­ mist to frame the concept of restricted ma­ terial production, productive labor was lim­

182

TH E INCO M E OF NATIONS

ited to that labor which (a) produced a marketable material product, (b) produced a commodity whose price could command a quantity of labor equal to the labor neces­ sary to produce it, and (c) which added the value of its own maintenance plus the entre­ preneur’s profit to the raw materials. All other labor was “nonproductive” and was, in fact, supported by the former. National product, therefore, consisted only of salable material goods. The labor engaged in ren­ dering services and the services of dwellings could not, Smith felt, fulfill these require­ ments and were hence excluded from pro­ duction and national income. This definition of production and national income was both inconsistent and narrow— too narrow to permit an adequate analysis of the economy’s operations. Smith erro­ neously concerned himself with wealth pri­ marily in terms of a “stock of goods,” seri­ ously neglecting its aspect as a “flow” of utilities. He did not perceive that the capac­ ity to produce services was also “wealth” and that “income” was a flow of both ma­ terial and immaterial wealth. His concept of income was affected by the then prevalent notion (and alive in some quarters even to­ day) that wealth can only be a stock of things, e.g., only an accumulation. His dis­ tinction between material goods and services from the point of view of their relative utili­ ties (the supposed durability of the first and perishability of the second) was both in­ accurate and irrelevant. He also failed to see that the production of services as well as of commodities gives rise to the creation of profits and to the circulation of entre­ preneurial capital. Even at the time he wrote there were many services being produced with hired labor and the use of capital that yielded profits to the entrepreneurs operat­ ing them, e.g., the services of innkeepers, stage coach operators, shipowners, bank­ ers, owners of boarding schools, and the like. In accordance with the first of his own two definitions— the one relating to the pro­ duction of profits— these services should have been classified as “productive,” but Smith couldn’t or wouldn’t see the incon­ sistency. By restricting the concept of production and national income to material goods alone, Adam Smith reduced the usefulness of na­ tional income estimates as measures of the economic welfare of the nation. His very admission that “unproductive labour . . .

has its value and deserves its rewards” was the best testimony of the inadequacy of his concept as a basis for measuring the eco­ nomic production and prosperity of a na­ tion. Smith’s contention that material goods are more useful than services because they are more lasting was incorrect. Durability is by no means the most important attribute of utility. Moreover, it is the durability of the effects of a thing that counts in the meas­ urement of its usefulness and not the dura­ bility of the thing itself. Some services have more lasting beneficial effects than many material goods. The effects of education may last a lifetime or may even be transmitted to succeeding generations, whereas the effects of the consumption of many material goods are only fleeting. Adam Smith also overlooked the rec­ iprocity between services and goods, where one imparts utility to the other. He stood ready to include those services in national product that helped to complete the utility of material goods, such as trade and trans­ portation, but he was not ready to include those services that constituted final utilities in themselves and in whose production ma­ terial goods were used. He included only the auxiliary material goods, accounting for them without any indication of their use— leaving them, so to speak, suspended in air. According to this notion, classroom furni­ ture and equipment would all be part of national income, but the instruction, through which these goods alone acquire utility, would not be included. Stage scenery, costumes, the curtains that separate audi­ ence from players, would be part of na­ tional income, but the performance of the actors and the work of the director and producer, which alone give meaning to these articles, would not be part of it. This ex­ clusion of services under the Smithian con­ cept made any meaningful analysis of many material goods in terms of their role in con­ sumption well-nigh impossible. Such an analysis could be completed only by tracing the expenditures of the producers for the services of the nonproducers, i.e., by ana­ lyzing the so-called “derivative incomes” and reintroducing the excluded services by the back door. But as soon as these excluded services are brought back into the analysis, and the derivative incomes are set up in parallel alignment to the so-called national income proper, the problem arises of how

12.

TH E AREA OF PRODUCTION

to com bine these two categories o f incom e in a m eaningful total and how to break down the total into categories o f expendi­ ture. Smith never acknow ledged this prob­ lem, n o r did any of the statisticians who adopted his concept as a basis fo r their esti­ m ates.9 N ational incom e restricted to m aterial production m ay be an adequate m easure o f the econom ic production and econom ic w elfare of an extrem ely prim itive society w hich is almost wholly concerned w ith the production of m aterial goods fo r th e satis­ faction of the m ost elem entary physical needs of its m em bers. B ut it is n o t an ade­ quate m easure of th e econom ic production and econom ic w elfare of an advanced so­ ciety, w hich is generally concerned w ith both the creation of services and the crea­ tion of m aterial goods, and w here both of these cater to a w ide range of hum an needs, beginning w ith the m ost rudim entary ones and ending w ith those of a m ost sophis­ ticated character. N ational incom e conceived in these n a r­ row term s is a poor tool fo r com paring the volumes of production and levels of eco­ nom ic w elfare o f two societies th a t differ widely in the degree o f their econom ic de­ velopm ent, o r of the sam e society during different periods of its developm ent. Inas­ m uch as services com e to play an everincreasing role in a developing society, na­ tional incom e restricted to m aterial p roduc­ tion m ay tend to understate, by an everwidening m argin, th e degree o f advance in econom ic production and econom ic w elfare achieved by a country over a period of tim e.10 This general proposition, however, requires some qualification, as show n in C hapter 16, Section 5b, below.

3. The Marxian Restricted Material Production Concept In the historical section (C hapter 1, Sec­ tion 8 ), the M arxian restricted m aterial pro­ duction concept was show n to be largely a replica of Sm ith’s earlier notions, differing from them only in p art of its underlying rationale and in a m ore precise and elabo­ rate application.

a. Emphasis on the Sole Productiveness of Hired Labor A ccording to Sm ith, b oth the hired w orker and his m aster are productive agents

183

and receive due rew ards fo r their contribu­ tions to production. O n the other hand, M arx attributes a productive role only to the hired w orker. H e postulates th a t labor alone is productive, denying any productive function to the entrepreneurial class. All values, he m aintains, are b ut congealed labor tim e. L ab o r reproduces the value of its own m aintenance and, in addition, cre­ ates “surplus value” for the exploiting capi­ talist who appropriates it to himself in the form of profits and n et rent. M arx does not believe th at the w orker receives his due rew ard; and does believe th at the m aster receives his incom e w ithout w arrant. The m aterial product, he declares, is created solely by the exertions of hired labor. The p roprietor has nothing w hatever to do with it. T hrough his control of the means of p ro ­ duction, he m erely appropriates to himself the value created by the hired w orker which, in all justice, belongs to the latter. T he profits (w hich in M arx ’s conception always include interest) and n et ren t obtained by the entrepreneur o r p roprietor are nothing b u t the “surplus value” created by labor over and above the value of the wages paid to it. M arx ’s definition of hired w orkers in­ cludes m ental w orkers, such as engineers, as well as m anual workers, b ut does n ot in­ clude the entrepreneur w ho manages the enterprise. T he self-employed farm er and craftsm an w orking w ithout hired help is neither productive n o r unproductive, since his w ork is altogether unrelated to the capi­ talistic process of production. H ere M arx m akes the qualification, however, th a t as fa r as the self-employed person uses capital, he m ay be creating surplus value by exploit­ ing himself. In fact, the self-employed per­ son is functionally split, being p art capitalist and p art w orker, and his wage incom e is reduced by an im puted profit.11 T he values o f all commodities, according to M arx, can be split into two m ajor parts: ( a ) the p a rt em bodying three distinct items — the cost to the entrepreneur of the wages paid to his w orkers, the cost to him of the raw m aterials and other auxiliary supplies em ployed by him in production, and the cost to him of the portion of his fixed capital consum ed in the course of production (i.e., depreciatio n ); and (6 ) the p art em bodying the surplus value created by the w orker for the capitalist and providing for him income for personal consum ption or for addition to his existing capital.

184

THE INCO M E OF NATIONS

In thus redefining the roles of w orkers and m asters in m aterial production, the bases of the product shares each receives, and the elements m aking up th e values of the com modities, M arx, in fact, redefined the substance o f national income. In his hands, the m aterial production concept of national incom e becam e a tool fo r the de­ structive analysis of the ethical foundations of capitalist society. It also becam e a con­ venient vehicle fo r the enunciation of M arx’s political program , one calling fo r the labor­ ing class’s expropriation of th e capitalist’s ownership of the m eans of production in order to end the exploitation of the w orker by the capitalist. W e are n o t concerned here w ith M arx’s political program , and are m entioning it only to provide the necessary background fo r understanding the reasons th a t prom pted him to carry on his analysis of econom ic production in capitalist society in strictly m aterial term s. It m ust be clearly u nder­ stood that, fo r M arx, economics was b u t a handm aiden of politics, and his economics was dom inated by his political beliefs and objectives. These looked to the creation of a society in w hich the workers, acting through a party representing their interests, w ould control the m eans of production and receive all the fruits thereof.

b. Clearer Distinction Between the Vari­ ous Elements of Gross and Net Na­ tional Income M arx im proved on Sm ith’s distinction between gross and net national income, and the elements m aking them up. H e broke down the gross value of national output m ore clearly than Sm ith had done into its three m ain elem ents: (a ) the costs o f raw m aterials, auxiliary supplies, and other op­ erating capital resulting from past produc­ tion and advanced by the entrepreneur (i.e., the costs of repairs, m aintenance, and re­ placem ent o f used-up inventories) and re ­ produced in the new cycle of production; (b ) the costs of the used-up portion of fixed capital (provision fo r depreciation) simi­ larly resulting from past production and re­ produced in the new cycle of production; and (c) the new ly created product or value, available fo r consum ption and fo r new in­ vestm ent, w hich alone constitutes the year’s net national incom e.12 In developing this analysis M arx uses his

ow n term inology. T hus, he refers to w hat we now call “gross value o f the o u tp u t” simply as the “gross p ro d u ct,” while calling w hat is now generally term ed “n et value of the o u tp u t” or “value added in production,” “gross incom e.” H e refrains from using the term “n et national incom e” on the ground th a t in capitalist society, the term “n et in­ com e” is used generally by the entrepreneurs to designate their share o f incom e from p ro ­ duction.13 In the several volum es of his w orks M arx uses the term “national incom e” only once, w hen he w rites: “Viewing the incom e o f th e w hole society the national incom e consists of wages plus profit, plus rent, th a t is o f the gross incom e.” 14

c. Specific Definitions of the Area of Production M arx does not add m uch to A dam Sm ith’s definition of the area of m aterial p roduc­ tion. In some respects he is m ore specific th an Smith, in others, less so. T hus, like Sm ith’s, his concept of m aterial production includes agriculture, m ining, and m anufac­ tu rin g ,15 b u t he is n o t as positive as Sm ith was about including trad e in this category. O n one hand, he m aintains th a t trad e does n o t create any use value by itself b u t m erely helps to realize the use value produced in industry; on the other hand, he observes th at trade “is im m ediately productive fo r the capitalists,” is often paid by a share in their profits, and form s a p a rt o f th e selling or exchange values of the com m odities.18 H e recognizes th a t tran sp o rtatio n of com m odi­ ties is m aterial production because it changes the location and, hence, the use value of com m odities and increases th eir exchange value, b u t he is n o t as clear as to w hether the transportation o f persons is properly a p a rt o f m aterial production. In fact, his ar­ gum ent on this lends itself to either type of interpretation.17 M arx excludes governm ent services from the field of production just as A d am Sm ith did, b ut his reasons are som ew hat different. H e excludes them not only because these services are n ot m arketable an d n o t gov­ erned by the principles of capitalistic p ro ­ duction generally, b ut also because he ques­ tions th eir utility. H e regards governm ent as a repressive agent helping th e exploiting classes to oppress the w orkers. Finally, M arx adds nothing to A dam S m ith’s distinction betw een “prim ary ” and

12.

THE AREA OF PRODUCTION

“secondary” incom es; betw een the original distribution of the national incom e and its subsequent redistribution. H e accepts Sm ith’s distinctions w ithout modification. d. A p p lic a tio n o f th e M a rxia n C o n c e p t in S o v ie t R u s s ia ’s E stim a te s T he M arxian concept of national incom e receives its fullest expression in the national incom e estimates o f Soviet Russia. T he basic w orks on Soviet Russia’s national incom e, p repared and published in 1939 and 1940, under the auspices of the Soviet A cadem y of Sciences, defined national incom e as “that p art of the social product, evaluated in m oney term s, w hich is newly created each y ear by the labor of the society and be­ comes available annually fo r consum ption and accum ulation,” adding th a t this social p ro d u ct consists of m aterial goods only.18 A m ore recent official publication of the M inistry of F inance defines the national in­ come of Soviet R ussia even m ore succinctly as “th at p art of the social product w hich is created by new ly applied labor of th e w ork­ ers engaged in m aterial production and . . . is devoted to the purpose of capital accu­ m ulation and consum ption.” 19 N ational in­ com e is calculated as th e sum of the net outputs o f agriculture, m ining, m anufactur­ ing, construction, services o f freight tran s­ portation, trade (including resta u ran ts), and, apparently, some branches of com m unica­ tions. T he last three item s are included on the ground th a t w ithout them , m aterial p ro ­ duction cannot achieve its purposes; in other words, th a t goods have no value unless they are delivered to th e ir ultim ate users. T he n et output o f these branches o f pro ­ duction is com puted as the sum o f (a ) the wages and other form s of incom e paid to their w orkers and m em bers of co-operative organizations; ( b ) social security and hous­ ing contributions and costs o f various facili­ ties furnished to producers in these branches in connection w ith their w ork, such as serv­ ices of factory clinics, nurseries, and restau­ rants as well as technical instruction and training; (c) net interest on short-term loans of the enterprises and the cost o f th eir in­ surance; ( d ) profits th a t supply new capital to the enterprises and the m eans to prom ote the w elfare of the w orkers as well as revenue to th e governm ent; and (e) th e turnover and oth er taxes paid by enterprises th a t are treated as the collective profits of the n a­ tional society.

185

O n the other hand, national incom e does n o t include the (a ) civil and m ilitary serv­ ices of governm ent; ( b ) the services and benefit paym ents of the social insurance in­ stitutions; (c) the services of passenger transportation; (d ) the services of independ­ en t physicians, dentists, teachers, artists, barbers, laundrym en, and other w orkers perform ing strictly personal services; (e) services o f domestics; and ( /) services of dwellings. T he om ission o f governm ent services and of oth er service items from the Soviet esti­ m ates tends to understate Soviet national incom e. This understatem ent, however, is not as great as m ay be expected because of the peculiar n atu re o f the Soviet econom y in w hich a w ide array of services is provided w ithin the organization of the factories en­ gaged in m aterial production, e.g., factory restaurants, factory housing, nurseries fo r children of w orking parents, etc. These services are treated as p art of m aterial p ro ­ duction an d their costs are em bodied in the price o f the p ro d u ct and included in the net o utput o f the b ranch of m aterial production involved. M oreover, u n d er the Soviet con­ cept, indirect taxes (generally excluded from national incom e in capitalist countries) are included along w ith direct taxes and may m ore th an offset the effect of the exclusion of services. Soviet economists, however, place fa r too m uch em phasis on the effects of the con­ ceptual difference existing between their estim ates and those of capitalist countries. T hose scholars who have critically exam ­ ined the Soviet estim ates generally recog­ nize th a t they greatly overstate the size of Soviet national incom e an d the rate o f its growth. T his is largely due to the use of im perfect price indexes, overemphasizing new products, and the evaluation of them in the higher prices of form er years. T he overestim ation o f Soviet national incom e on this account is fa r greater th an any u nder­ estim ation resulting from the em ployment of the m aterial production concept. As if to fu rth er com pound the error, the Soviet estim ators assume, in rath er high­ handed fashion, th at their estimates are understated by 25 or 30 p er cent in com ­ parison w ith capitalist estimates. This fig­ ure is obtained by estim ating the supposed overstatem ent of national income in coun­ tries em ploying the com prehensive produc­ tion concept. T he “overstatem ent” is com ­

186

THE INCOM E OF NATIONS

puted as the proportion that incom e from services contributes to total national income. This proportion, am ounting in some coun­ tries to 25 or 30 per cent, is then declared to m easure the degree of overstatem ent. These national incom e totals are then re­ duced by th a t proportion, and the Soviet estim ators present the result as a corrected total, com parable w ith Soviet estim ates.20 This adjustm ent is incorrect. I t overlooks the fact th at services play a m uch m ore im portant role in other countries and that their elim ination from national income affects the estim ates in their case quite dif­ ferently from Soviet Russia’s. It also ignores the offsetting overevaluation effects of cer­ tain Soviet procedures discussed above, tend­ ing to overstate Soviet Russia’s relative eco­ nomic strength. D r. Ivo Vinski, the estim ator o f Y ugo­ slavian national income, places the u nder­ statem ent of his estim ate, because of the em ploym ent o f the m aterial production concept, at the m uch m ore m oderate figure of 11 per cent (see C hapter 28, Section I I ) . It is impossible here to establish a p er­ centage understatem ent th a t w ould apply to all situations, o r even to set up a standard procedure th at w ould m ake the estimates derived under the tw o divergent concepts com parable. One thing is clear: the gap between the figures cannot be closed by simply adding the value of the excluded services to the national income totals based on the m aterial production concept. It would be necessary, on the other hand, to exclude from the same total a proportion o f the in­ cluded indirect taxes th a t would m ake the treatm ent of taxes com parable to the esti­ m ates em ploying the com prehensive con­ cept. Even then the differences due to the em ploym ent of the different concepts m ay not be com pletely elim inated, necessitating either additional adjustm ents or the em ploy­ m ent of some other reconciliation m ethods.

e. Conceptual Differences Among the M a rxists Professor Strum ilin, m em ber of the Soviet A cadem y and noted scholar, and some other prom inent Soviet economists departed from a literal interpretation of M arx’s view and took the position that governm ental, per­ sonal, and other services are just as produc­ tive as m aterial production, and should be included in national income on an equal footing.21

Professor C hernom ordik, in th e above quoted study published by the Soviet A cad­ em y o f Sciences in 1939 (see note 18, p. 5 3 1 ), though not going quite so far, ar­ gued in favor of including the services of passenger transportation in national p roduc­ tio n and national incom e.22 M arx, he said, never intended to exclude all services from national production. H e intended to exclude only services of a strictly personal nature, such as exist between a m aster and a do­ m estic servant, an d w hich are n o t subject to the rules of capitalist production, and deem ed the distinction betw een a physical good and a service to be econom ically u n ­ im portant. T he only im portant distinction according to M arx, w rote Chernom ordik, was betw een production organized under th e prevailing rules of the society an d flow­ ing through its channels an d th a t n o t so conducted, thus m aking th e form er a p art o f national production and national income, and the latter n ot a p a rt of it. Since passen­ ger transportation utilizes th e institutions of society, be it capitalist o r socialist, con­ cluded C hernom ordik, it m ust, according to M arx, be treated as a p a rt o f national p ro d ­ uct. C hernom ordik m aintained th a t neither theoretical n o r practical considerations jus­ tified excluding, fo r example, the services of so useful an undertaking as the M oscow subway. A t least one h alf of passenger trans­ portation provides services to production, in th at passengers are traveling to o r from w ork o r on specific business assignments. T he o th er half, though serving consum ption, is just as productive as consum er goods. Passenger service in 1935 engaged 600,000 w orkers, o r one fifth o f the three million w orkers then engaged in the transportation industry as a whole. A s the value of their services m ust have am ounted to at least 2 billion rubles at cu rren t prices, its ex­ clusion w ould undervalue national income alm ost 1 p er cent. W hether o r n ot this b roader concept of national production an d n ational incom e, as expounded by these several em inent M arxian authors, will in tim e gain accept­ ance in Soviet social accounting, is im pos­ sible to say. If th eir reasoning should prevail in appli­ cation to the services of passenger transpor­ tation, it w ould have to be extended to m any other fields of services, including even the services o f governm ent. It certainly appears

12.

THE AREA OF PRODUCTION

anom alous th a t the services of governm ent em ployees w ho plan and direct the nation’s “productive” effort, and who collect the taxes th a t provide the capital or operating subsidies required by the “productive” en­ terprises, should be treated as “unproduc­ tive” labor.

/. Critique of the Marxian Concept E very one of the criticism s advanced against the Sm ithian restricted m aterial pro­ duction concept o f national incom e applies w ith equal force to the M arxian concept, both as originally form ulated and as now applied in Soviet Russia and other C om ­ m unist countries. T he estim ates developed u nder this concept m easure only a portion of the true national incom e of th e country and fail to give a full picture of th e opera­ tions of its national econom y. T hey also fail to provide a true m easure o f the changes occurring over a period of tim e in the vol­ um e of production and in the level o f eco­ nom ic welfare.

4. The Restricted M arket Produc­ tion Concept T he restricted m arket production concept of national incom e is a com prom ise between the com prehensive production concept and the m aterial production concept. In com ­ m on w ith the first, it includes m arketable services in national incom e, but it refuses to include the services o f governm ent and other collectives th a t are not controlled by m arket forces. This concept uses m arketability as the m ain criterion fo r the inclusion of products in national incom e on the ground th a t only those products have an objective econom ic value th a t pass through the m arket and whose value is determ ined by the free inter­ play of the forces of supply and dem and. T he services of governm ent and other col­ lectives are said to have no objective value, because both the supply o f and the dem and for them are determ ined by political forces o r other pressures th a t m ay have nothing to do with the econom ic interests of the mem bers of the society. T he cost and utility of governm ent services to society are not equated as neatly in the political o r other collective processes as are the costs and utility of products in the m arket. T here is no assurance th a t governm ent services are

187

w orth the m oney expended upon them , and their inclusion in national income m ay give a wholly distorted picture of the tru e n a­ tional p roduct o r incom e of the society. This concept was em ployed, w ith some modifications, by L andau and K alecki in their estim ate of the national income of P oland fo r 1929; an d by M atolcsy and V arga in their estim ates fo r H ungary for 1 9 2 4 /2 5 -1 9 3 6 /3 7 .23 U n d er this concept, national incom e included all goods and serv­ ices sold in the m arket as well as farm ers’ incom es in kind. T h e latter were included because they could be evaluated in the prices of identical goods and services sold in the m arket, and also because failure to include them w ould have resulted in an understatem ent of the incom es of the rural populations com pared w ith those of urban groups. T h e services o f governm ent enter­ prises were included because they were sold in the m arket. In addition, th e Polish esti­ m ators included tw o other items th a t did not fully square up w ith th eir adopted cri­ teria: (1 ) public education, on the ground th a t education was also provided privately and, consequently, its value could be m eas­ ured in term s of the m arket value of private education, and (2 ) the value o f governm ent investm ents, w ithout any other explanation than th a t they “increase national w ealth just as m uch as do private investm ents.” T he restricted m arket production concept o f national incom e has had lim ited accept­ ance. Except fo r K alecki and L andau in Poland, and M atolcsy and V arga in H u n ­ gary, its only recent cham pion has been the G erm an scholar, O tto K raus.24 The conten­ tion th at political evaluations of services are n o t as objective as m ark et evaluations is contradicted by innum erable examples. G ov­ ernm ent decisions in a free society m ay be as rational and objective as the private de­ cisions of producers and consum ers, and sometim es m ay be m ore so. T hey take into account the long-range interests of the m em ­ bers of society often m uch m ore closely than do the private decisions of consum ers and producers meeting in the m arket. T he serv­ ices of governm ent are frequently m ore use­ ful econom ically and are w orth m ore to so­ ciety th an alternative outlays for privately produced goods and services; e.g., public education, hygiene and sanitation, as against private expenditures fo r conspicuous con­ sum ption, n ot to speak of the expenditures fo r narcotics o r vice. T he one field of gov-

188

THE INCOM E OF NATIONS

ernm ent activity, however, th a t defies ra ­ tional evaluation is w ar. But even this ac­ tivity, w asteful though it m ay seem in the long run, m ust be included in national in­ come, if the concept of national incom e is to have any reality. T he exclusion o f govern­ m ent services, from whatever viewpoint, makes no sense.

5. Conclusion It is seen from the foregoing analysis that o f the three production concepts of national incom e, the com prehensive production con­ cept is by fa r the m ost rational, b ut th a t a n u m b er o f unsolved problem s regarding the exact scope o f production to be included in national incom e are still to be dealt with.

13. FINAL VS. INTERMEDIATE PRODUCTS IN THE MARKET ECONOMY W e have already said th a t national incom e in its m ost fundam ental aspect is the aggre­ gate value of all final goods and services created by a nation over the period o f a year fo r consum ption and investm ent. This aggregate is com puted exclusive of the values of interm ediate products because these are already included in the values of the final products. T o do otherw ise would m ean counting the same things twice. H ow ­ ever, it is m uch easier to express this propo­ sition theoretically than to put it into prac­ tice, fo r it is not always easy to distinguish between final and interm ediate products. T oo often the w ay the distinction is m ade depends upon the particular goals one de­ cides are proper ends fo r econom ic activity, o r even on one’s view of the substance of life itself— subjects on w hich hum an beings tend to disagree. In the end, therefore, the distinction is usually m ade on the basis of convention rath er than on any well-reasoned philosophy of life. B ut convention is a fickle thing, and w hat is regarded as a final product at one tim e m ay be regarded as an interm ediate one at another. It is of the utm ost im portance, neverthe­ less, to try to distinguish between these two categories as rationally as possible, fo r upon this distinction, in the final analysis, de­ pends the logical consistency of the national incom e com putation. It is especially im por­ tant to do this w hen th e com putation is m ade by the net output and final products methods, and w hen the national income Notes begin on p. 531.

total is analyzed by industrial divisions and by the types of consum ption and investm ent th a t it serves. A ny erro r m ade in classifying products results either in the overvaluation o r th e undervaluation o f th e national in ­ com e and o f its various parts. W e shall consider this im p o rtan t problem o f differentiation first, as it presents itself in the m arket econom y and, second, in the succeeding chapter, as it appears in the gov­ ernm ent economy. It differs significantly in these tw o areas and calls fo r different treat­ m ent in each.

1. The Empirical Rule In th e m arket econom y, final products are distinguished from interm ediate p ro d ­ ucts by observing the flow o f com m odities and services from the prim ary o r raw m a­ terial stage to th e finished p ro d u ct stage, and by applying the follow ing em pirical rule: if the p roduct (goods o r service) is sold by one business u n it to ano th er for resale, and if its cost is incorporated in the price of the p roduct o f th a t latter unit, it is an interm ediate pro d u ct; b ut if it is sold n o t fo r resale b ut fo r personal consum ption o r fo r investm ent, it is a final product. G overnm ent enterprises are generally con­ sidered to be a p art of the m ark et economy, inasm uch as their services are sold fo r a price th a t covers all o r m ost o f th eir cost. T h eir services, therefore, should be classi­ fied as final o r interm ediate products ac­

13.

FIN A L VS. INTERM EDIATE PRODUCTS

cording to the sam e rule: nam ely, if they are sold fo r resale, they are interm ediate goods, but if they are sold fo r personal con­ sum ption o r fo r investm ent they are final products. A ccordingly, the services o f the post office sold to business and the w ater sold by a m unicipal w ater supply to ap art­ m ent houses and com m ercial and industrial buildings should be classified as interm ediate goods. O n the other hand, the services of the post office and the w ater sold to indi­ vidual hom e ow ners should be classified as final products. This m arket rule, how ever, is inapplicable to th a t portion o f the serv­ ices of governm ent enterprises th a t is not covered by the price charged fo r it, but is financed from general taxation. N o r is it applicable to the services of governm ent enterprises sold to the governm ent itself or furnished to it w ithout charge. T his portion of th e services o f governm ent enterprises should be classified as final o r interm ediate products according to th e rules applicable to the products o f governm ent generally. In the official U nited States estimates, the em pirical distinction betw een final and interm ediate goods on th e basis of the n a­ ture of their sale in the m arket is used, even w ith respect to the goods and services sold by business to governm ent and to nonprofit-m aking institutions. T he authors of these estimates explain the basis o f their classification of products, as follows: We start with the obvious fact that individ­ uals, non-profit institutions serving individuals, and general government are ultimate buyers in the sense that they do not buy for resale in the market. Accordingly, their purchases are not elements of cost in the value of other out­ put produced for the market. Hence there is a presumption that their purchases should be regarded as final products in any measure which purports to give a complete accounting of the entire output of the nation. Business organiza­ tions and government enterprises, on the other hand, are intermediaries in the sense that they produce for sale in the market. Accordingly, their purchases, to the extent used up in fur­ ther production, are included in the values of goods and services which business sells. Hence there is a presumption that such purchases are intermediate products and should not be in­ cluded separately in a measure of the value of national production.1 T o apply, as these estim ators have done, the rule of classification of products draw n from the operations of the m arket econom y to the nonm arket econom y, is a m istake, for the differences between these tw o econo­

189

mies are basic. T he fact th a t governm ent and nonprofit institutions do n o t resell the products they purchase from business does not m ake them the ultim ate consum ers of these products; n o r does it m ake these prod­ ucts, per se, final products. This approach wholly disregards the fact th at government and the nonprofit institutions are producers of services and n ot m erely “consum ers.” A ctually, the products purchased by gov­ ernm ent and these institutions from busi­ ness are interm ediate goods and are incor­ porated in the services they render. These services, in tu rn , m ay be final or interm e­ diate, depending on w hether they are fu r­ nished by these producers to individuals or to business. T he consum ers of governm ent services do n o t use the cem ent and steel the governm ent buys to build roads and bridges. T hey use the services of the roads and bridges built w ith the aid of these m a­ terials. N o r do the consum ers of th e serv­ ices o f nonprofit-m aking research founda­ tions use the chemicals purchased by the foundations in their work. T hey use the re­ sults o f the researches of these institutions. T he m arket rule of classification of p ro d ­ ucts into final and interm ediate m ay hold tru e fo r products circulating w ithin the m arket o r leaving it fo r consum ption by private households, b u t it cannot be carried over into these oth er sectors of production w ithout producing inaccurate results. In this respect the U nited States practice o f classi­ fying all purchases m ade fro m business by governm ent and nonprofit-m aking institu­ tions on the basis of the stated m arket rule is misleading. Even w ithin the m arket econom y the em pirical test of final sale vs. sale fo r re­ sale has serious flaws. Simon K uznets has pointed o ut th a t the fact th a t a good or service is sold to an individual in his ca­ pacity as a consum er and supposedly, there­ fore, will be consum ed by him in a final way, does n ot necessarily m ean th a t it will not be “resold” by him in his capacity as a producer. “M any individual consum ers,” says K uznets, “are during the current tim e u nit sellers of labor services: the food, clothing, etc., they buy fo r themselves and m em bers of th eir families may, therefore, be classified as bought fo r resale, since the rendering of labor services is contingent upon life and m inim um com fort of the w orker and his fam ily.” 2 If we should fol­ low this line of reasoning to its ultim ate conclusions, we w ould have to substitute

190

THE IN CO M E OF NATIONS

fo r the concept of national incom e th a t of a “national dividend” as a retu rn over and above the m aintenance of labor. W ithout going quite that fa r in identifying the labor­ er’s consum ption as a deductible cost of production, we may, nonetheless, point to certain items in that consum ption th a t m ay m ore properly be classified as business costs. F or example, the clothes bought by an in­ dividual m ay be used as m uch fo r his p er­ sonal consum ption as fo r his w ork; while some clothing, such as the required nurse’s uniform , used solely during w orking hours, is in reality a producer good bought for resale in the sense th a t its costs is returned to the nurse in her wage. It is a final con­ sum er good only to the extent th a t it is a substitute for the ordinary clothes one m ust w ear w hether one w orks o r not. It certainly makes little sense to call the sam e uniform s final consum ers’ goods w hen purchased by nurses themselves, and to call them inter­ m ediate producers’ goods w hen purchased fo r free distribution to the nurses by the hospitals.

2. Unresolved Theoretical Issues Even m ore profound issues th an th e one just m entioned have been raised in recent years regarding the classification of prod­ ucts into final and interm ediate. Some of the theories advanced in this connection w ould substantially expand the area of final p ro d ­ ucts and, thereby, broaden the concept of national incom e and increase its value ag­ gregate. O thers w ould do the opposite, broadening the category of interm ediate products at the expense o f final ones and, accordingly, restrict the concept o f national income and reduce its m easure. It has been suggested by some economists that the m any services m odern businesses furnish to their custom ers w ithout charge, incidentally to their m ain operations, are really final products, and that their value should be included in national income. A m ong these are cited the decorative ar­ rangem ents of m odern stores and offices th a t please the eye; their lounges and w ait­ ing room s, equipped w ith radio and televi­ sion sets, paintings, and magazines, w hich provide a variety of com forts and recrea­ tional facilities to custom ers; and businesssponsored advertising that includes attrac­ tive posters, m agazine illustrations, and tele­ vision shows brought into people’s homes for their entertainm ent and education.

a. Fisher’s and Gini’s Restrictions O n the other hand, oth er econom ists have suggested a reclassification of m any final products as interm ediate ones, and a corre­ sponding reduction in the com putation of national income. T he late Professor Irving F isher advocated restricting the concept of national incom e to currently consum able goods and services. H e w ould include both cu rren d y produced goods an d services and cu rren t services o f previously produced durable goods.3 In fact, h e w ould eval­ uate all goods in term s of the cu rren t serv­ ices p roduced by them . T he idea th at goods are m aterialized services and should be evaluated in term s of these services is not new. P rofessor Storch expressed it m ore th a n a century before.4 Professor Fisher w ould exclude from the concept an d com putation of national in­ com e all cu rren t savings an d corresponding additions to the existing stock of w ealth o n the ground th a t these savings constitute deprivations of consum ption and th at the cu rren t additions to the stock of w ealth do n ot as yet furnish any im m ediate utility. T his theory, too, is by no m eans novel, having been expressed by Jo h n S tuart M ill in his Principles a century ago.5 N ew or n ot, this notion is wholly unrealistic. T he purpose of saving in the social sense always was and always will be to increase future consum ption, and the pursuit of this objec­ tive supplies its own gratification: the antici­ pation o f larger fu tu re consum ption.6 Professor C orrado G ini criticizes F isher’s theory on the ground th a t savings an d newly produced capital goods “m ay be exchanged on th e m ark et w ith services w hich w ould be included in the reckoning of the income, an d m ay be thus exchanged fo r th eir whole value, inclusive n o t only of th eir present utility b ut also— even if duly discounted— of th eir fu tu re utility.” 7 Professor G ini w ould restrict the concept o f final products even m ore severely than any o f the w riters quoted. H e w ould deduct from national incom e “the cost incurred for feeding, clothing and lodging th e w orker, fo r keeping him in good health, and fo r his training, as well as those fo r his replace­ m ent by rearing new w orkers o f like num ber an d productive yield.” A ll these, he says, are identical to the “cost o f raw m aterials and o f the energy— physical, anim al, o r hum an — em ployed in their pro d u ctio n ,” and to the cost “in cu rred fo r the upkeep and am orti­ zation o f th e plants” th a t are ordinarily de-

13.

FINAL V S. IN TERM EDIATE PRODUCTS

ducted from the value of products. “D o we n o t do this [deduction of such costs of m aintenance of life],” he asks, “w hen esti­ m ating the services o f animals? A nd was n o t this done, is it n o t done, in valuing h um an services w here slavery prevailed o r prevails? T he change brought about in civi­ lized countries in the legal position of the w orker cannot alter the econom ic criteria to be followed in valuing his services. T he principle, in keeping w ith this point of view, of exem pting from taxation th a t quota of incom e the taxpayer requires fo r his own keep and that of his fam ily, has been ac­ cepted since ancient tim es and is, generally speaking, recognized in m odern societies, even by the T reasury.” 8 E xcept fo r its in­ elegant expression, G ini’s point is no dif­ ferent from the fam iliar proposals to sub­ stitute the concept and m easure o f a n a­ tional surplus o r dividend fo r th a t o f n a­ tional income.

b. Shoup’s, Lowe’s, and Kuznets’ Distinctions C arl Shoup, in a recent w ork dealing with an aspect o f this sam e problem , has sug­ gested a classification of consum er goods into “pu re” and “m ixed,” the first being those whose sole aim is personal gratifica­ tion, th e second being those th a t “in being consum ed, also serve to increase th e pro ­ duction of other things.” 9 B ut he did not indicate how this classification can be tran s­ lated into practical term s fo r com puting national income. A dolph Lowe, expressing ideas sim ilar to G ini’s, recently proposed a reconsideration o f the concept of national incom e along altogether novel lines. H e considers the existing com putation of net national product inconsistent as it excludes the replacem ent o f existing capital stock, thus recording the flow of capital stock a t a net figure, w hile it includes the replacem ent o f labor used up in production, thus recording th e flow of consum ption goods and services at a gross figure. H e suggests carrying th e con­ sum ption flow at a net figure also, i.e., by deducting the cost o f the replacem ent of labor energy from the aggregate value of consum er goods. H is net product o f con­ sum er goods, paralleling net capital form a­ tion, would be represented by th e am ount o f current output o f consum er goods in excess of the am ount necessary to m aintain the currently prevailing “cultural m inim um o f existence.” T he net national product,

191

under Low e’s concept, w ould m easure net changes in the level of living as well as in capital stock.10 Sim on K uznets recently pointed out th at national income, as it is com puted in highly developed countries, includes at least three item s of cost of doing business and earning a living, w hich, as they really represent in­ term ediate an d n o t final products, do not belong there. T he three items are: (a) the increased transportation and distribution costs caused by the growing complexity of the organization of production; (b ) the in­ creased costs of satisfying essential needs due to increased urbanization; and (c) the added costs of m aintenance of credit institutions, trad e organizations, and labor unions in m odern economies. K uznets would deduct these costs from national incom e in order to m ake its com putation in developed countries m ore com parable w ith the esti­ m ates fo r underdeveloped countries.11 Some o f the foregoing proposals would alter the existing concept of national in­ com e. W hatever nam e m ay be attached to this new concept, it resembles the “national dividend” referred to by A lfred M a rsh a ll12 and the “surplus incom e” o f John A . H o b ­ son 13 (if n o t th e “revenu n et” of the phys­ io crats). T he expression of these substitute concepts in practical statistical term s is com plicated by considerable m ethodological difficulty and is n o t likely to be done fo r some tim e. In the interval, it is necessary to continue operating w ith the existing con­ cept, introducing lim ited im provem ents as tim e goes on.

3. The Special Problem of Finan­ cial Intermediaries T he services of banks, insurance com pa­ nies, and other financial institutions that form a p art of the m arket econom y call for very special treatm ent in classifying their products and in com puting their value for national incom e. These institutions act as interm ediaries between possessors of capital funds and users of them . T hey render serv­ ices to b oth groups in a m anner th a t often m akes it difficult to determ ine w hich serv­ ices are rendered to w hich group and w hat is th eir exact value. T h e difficulty is in large p a rt caused by the fact that these in­ stitutions do n o t always charge specific prices fo r some o f th e services rendered to depositors, b u t reim burse themselves fo r the expense involved by withholding from de­

192

THE IN CO M E O F NATIONS

positors some or all of the interest due on their funds. This also m akes it difficult to determ ine the relation of the product and incom e originating in these institutions to the products and incom es originating in other sectors and to m easure their im pact on the total national income o r product.14 In this section we shall concern ourselves with the problem of m easurem ent and clas­ sification of the incom e and product origi­ nating in o r flowing through banking and life insurance.

a. Banks T he difficulty of evaluating and classify­ ing the incomes and products in question is particularly great in the case of banks because their operations in m odern econo­ mies touch upon those o f every other seg­ ment, and are m ore extensive th an those of any other agency short of the governm ent itself. A lm ost all the funds in the econom y flow through the banking m echanism . All business organizations o f any consequence, all governm ental units, and a large propor­ tion of private households use th e services of banks. T hey deposit their funds w ith the banks, use the banks as their disbursing agents, and borrow m oney from them . T he banks charge their depositors specific fees fo r some services, b u t fo r the m ost p art they render services to them “free,” reim ­ bursing themselves fo r the costs involved and earning their profits by withholding all or a large p art of the interest earned on the loans and investments m ade w ith their de­ positors’ funds. T he interest charged bor­ rowers is presum ed to include the true in­ terest due on the m oney plus the cost of adm inistering the loans, but the records do not show these am ounts separately. It is impossible, therefore, to determ ine from banking accounts th e value of services ren ­ dered by banks to the two groups, o r the m anner in w hich this value is distributed between them . W orst o f all, inasm uch as these accounts do not show the value of the free services rendered to depositors— which is the principal elem ent in the services of banking— they give a wholly misleading pic­ ture of the product originating in banking. Since the interest received by the banks is a factor incom e originating outside bank­ ing, and is often larger than the factor in­ comes originating in banking (wage pay­ m ents, interest paym ents to depositors, and profits), the net factor incomes originating in banking, according to the banking ac­

counts, m ay often be a negative quantity. In other w ords, according to these accounts, banking m ay appear to be a drain on n a­ tional incom e rath er th an a contribution— a result th a t would obviously m ake no sense. T his incongruity is due to the fact th a t the largest elem ent in this area of national p ro ­ duction— the services rendered to depositors — does n o t take a m onetary fo rm .15 In som e national incom e estim ates, such as the G erm an estim ates before W orld W ar II (C hapter 26, Section I ) , the p ro b ­ lem o f how to trea t banking services was avoided by com puting national incom e by the incom e-distributed m ethod alone, w ith­ o ut showing its distribution by industrial divisions o r by final products. In other esti­ m ates, such as th e official British, incom e from banking was show n w ithout correc­ tions fo r the uncharged services. I t was pre­ sented as a negative q uantity o r w as other­ wise understated. C lark an d C raw ford in th eir estim ates fo r A ustralia (fo r 1901— 193 7 /3 8 ) om itted incom e originating in banking altogether. B ut in m ost o f the re­ cent estimates, m ethods have been devel­ oped fo r correcting the m isleading nature o f the figures. T h e best m ethod so fa r developed is the one proposed by th e Com m ittee o f Statis­ tical E xperts of th e League o f N ations in 1947 and adopted soon th ereafter in the estim ates fo r the U nited States, C anada, D enm ark, and Sweden.16 U n d er this m ethod the bank accounts are reconstructed to rep­ resent a situation in w hich banks w ould be paying o ut to depositors all th e interest earned on th eir deposits, instead of m erely a p a rt thereof, and, on th e o th er hand, w ould be charging them fo r all th e services rendered to them instead o f only fo r some of them . In other w ords, b ank income w ould be increased by an im puted service charge incom e equivalent to the sum by w hich the m onetary interest received by the banks exceeds th eir m onetary interest paym ents to depositors. B ank expenditures w ould be increased by an identical am ount of im puted interest paym ents to depositors. T he profit situation o f the banks u n d er this reconstruction would rem ain unchanged. T he incom e o r p ro d u ct originating in banking, figured u n d er this m ethod, as u n d er some of the others used, is com puted as th e sum o f the wages paid an d profits earned in banking. It is split into final and interm ediate products according to the ratio o f personal deposits to business deposits, on

1 3 . FINAL VS. INTERM EDIATE PRODUCTS

the theory th a t banking services are ren­ dered prim arily to depositors and only inci­ dentally to borrow ers. This split is fu rth er supported by reference to the fact th a t the services rendered by banks to personal de­ positors becom e em bodied in final consum p­ tion, w hereas those rendered to business depositors are absorbed in the various stages of production. T h e interest— m onetary and im puted— paid o ut by banks is treated, u n d er this ap­ proach, as factor incom e originating outside banking. It is sim ilarly split into final and interm ediate products in proportion to the personal vs. business deposits. T he service charges— actual and im puted— are also split betw een the tw o groups in proportion to deposits. The im puted interest paym ents to personal depositors are treated as additional facto r incomes o f th e individuals, and na­ tional incom e is thus increased by their am ount. T he im puted service charges to personal depositors are treated as additional personal expenditures, and becom e a p art of the disposition of th e enlarged national incom e. On the other hand, the im puted in­ terest paym ents to business depositors de­ crease their net interest paym ents, b u t this decrease is offset by the identical am ount of added im puted service charges th a t in­ crease their expenses fo r interm ediate prod­ ucts. T he aggregate of business profits and the national incom e total rem ain unchanged by the im putation process. A s the U.S. D e­ p artm ent of C om m erce puts it: “A ll that occurs is a redistribution in th e industrial origin of output, in the process of w hich the share o f banking is increased and th a t of o ther industries is reduced.” 17 This m ethod o f treating th e incom e or p roduct originating in banking is fa r from perfect, and cannot be considered as a de­ finitive solution of th e problem . T he U.S. D epartm ent of C om m erce, w hich uses this m ethod in its national incom e estimates, adm its to some o f its flaws, as follows: The procedure has been criticized in general as unduly complex and more specifically, as based on certain assumptions of doubtful valid­ ity. Particular exception has been taken to the assumption that all banking services not ex­ plicitly charged for are rendered to depositors and that the borrowers of bank loans are not involved, as well as to the assumption that these services are distributed in proportion to the ownership of the volume of deposits irrespec­ tive of turnover. While these and other objections have some merit and it is hoped that a simpler and more

193

cogent solution may be found to deal with the underlying problem of measuring the value added to output by banking, it would appear that the present procedure, all things consid­ ered, is the most satisfactory devised so far. Whatever its particular limitations, it attempts to measure a real element of income and prod­ uct in the business economy and permits a sensible solution to the problem of allocating income by industries.18

b. Life Insurance Companies T he problem of accounting fo r the trans­ actions of life insurance com panies is only partially sim ilar to the problem o f doing so fo r banks. U nlike banks, w hich render service to both depositors an d borrow ers, life insurance com panies prim arily serve their policyholders. W hat services they do perform fo r borrow ers and investors are only incidental to this first function. They m ake no specific charges to policyholders fo r the services rendered to them , b u t they reim burse themselves fo r these costs, as banks do, by w ithholding from their policy­ holders a portion of the interest o r other investm ent retu rn s earned by their funds. T hey collect prem ium s and pay death claims o r pay off the “surrender value” of the poli­ cies upon cancellation. Because the services they render to policyholders are n o t re ­ corded in specific incom e and outflow items, and also because their savings and insurance functions are mixed, it is impossible on the basis of their m onetary transactions alone properly to identify and m easure the in ­ com e and pro d u ct originating in, or flow­ ing through, life insurance com panies. H ere, too, it is necessary to introduce im putations. T he currently m ost favored treatm ent is the one suggested by the League of N ations’ Com m ittee of Statistical E xperts in 1947 and already adopted by the U nited States and several other countries. U n d er this m ethod, life insurance com pa­ nies are treated as associations of individ­ uals. Receipts from prem ium s and expen­ ditures on death claims are regarded as m ere transfers of capital funds am ong them and n ot as elem ents of national income. A ny investm ent incom e w ithheld from policy­ holders is im puted as an additional credit to the policyholders’ accounts. T otal invest­ m ent incom e actually credited or im puted is placed on the expenditure side of the com pany accounts and is treated as a cu r­ rent facto r incom e of the policyholders and as p a rt o f national incom e. A t the same tim e, an incom e is im puted to the insur-

194

THE INCOM E OF NATIONS

ance com panies representing charges fo r all their services to policyholders. The im puta­ tion is com puted as the sum of the operat­ ing expenses incurred on account of these services, i.e., as the sum of wages and com ­ missions paid to employees and agents plus the costs of purchases from other firms. T he im puted charges fo r services are treated as personal consum ption expenditures of the households under the title “expense of handling life insurance.” Accordingly, the services o f life insur­ ance com panies to policyholders as well as the actual and im puted interest credits to the accounts of the policyholders are treated as final products. T he services are treated as originating in life insurance; and the ac­

tual and im puted interest, as originating in o ther industrial divisions. N ational income u n d er this m ethod is increased by the im ­ puted interest to the policyholders’ account. T his treatm ent m ay also be im perfect, but it is the best developed to date.

4. Summary These are, in the m ain, the various solu­ tions offered to the problem of how to treat financial interm ediaries in com puting n a­ tional incom e, the m ost difficult as well as the m ost controversial one in the whole com plex o f problem s surrounding th e dis­ tinction between final and interm ediate products in the m ark et economy.

14. TREATMENT OF GOVERNMENT The treatm ent of governm ent presents one of the m ost difficult problem s in national incom e estimating, revolving around consid­ eration o f the following five m ajor ques­ tions: (1 ) should governm ent services be accounted fo r as final o r interm ediate prod­ ucts, and how should they be evaluated; (2 ) how should governm ent transfer expendi­ tures be accounted for; (3 ) how should taxes be treated; (4 ) how should govern­ m ent subsidies to producers be dealt with; and (5 ) how should transactions involving social security, pension, and w elfare funds be considered. E ach one of these questions will be dis­ cussed in this chapter. T he treatm ent of governm ent enterprises as opposed to gov­ ernm ent services has already been consid­ ered in C hapter 13, Section 1, and will be m entioned here only incidentally.

1. Treatment of Government Services a. Forms of Presentation in Different Phases of the National Income G overnm ent services are represented in national incom e in different ways, depend­ Notes begin on p. 532.

ing upon th e phase of national incom e be­ ing analyzed. If national incom e is shown in its production phase, then governm ent services are represented by th eir n et output o r value added, w hich consists o f the com ­ pensation paid to governm ent employees plus, sometimes, interest on the public debt o r the portion of it representing productive debt.1 If national incom e is show n in its dis­ tribution phase, th en the incom e paid out in governm ent services is sim ilarly repre­ sented by the com pensation o f governm ent em ployees plus, sometim es, interest on the governm ent debt. Lastly, if national incom e is analyzed in its expenditure phase, gov­ ernm ent services m ay be handled in either o f th e following tw o ways: (1 ) as an aggre­ gate o f th e com pensation o f governm ent em ployees, governm ent purchases fro m busi­ ness, and, possibly, the interest on govern­ m ent debt, subdivided w ithout any break­ dow n oth er th an by the class o f govern­ m ent—national, state o r local—m aking these disbursem ents; or (2 ) as an aggregate of the expenditures m ade fo r the various functions of governm ent (national defense, internal security, social w elfare, education, an d so fo rth ).

14.

TR E A TM E N T OF G OVERNM ENT

b. Method of Evaluation T h e general rule in national incom e ac­ counting is to evaluate governm ent services — other than those o f governm ent enter­ prises— in term s of the expenditures m ade fo r them , w hether these are paid fo r from taxes, from adm inistrative charges, or from loans. T he underlying assum ption is th a t these services are w orth their costs, i.e., th a t they are produced and supplied after a p ro p er balancing by the authorities (o r by the electorate itself on given occasions) of their social advantages and social costs.2 G overnm ent services are generally very incom pletely represented and evaluated in national incom e estimates. A s already m en­ tioned, they are usually represented only by the com pensation paid to national, state, and local governm ent em ployees. G overn­ m ent purchases from business are show n as products of business, not as parts o f the p ro d u ct o f governm ent. Interest on p ro ­ ductive governm ent debt is either excluded from national incom e, being treated as a m ere transfer paym ent, o r else is included as a separate item , w ithout any functional analysis as to the type of governm ent prop­ erty o r service financed by th e debt on w hich th e interest is paid. T his type o f ac­ counting does n o t perm it an intelligent analysis of governm ent services. In the fol­ lowing discussion w e shall tre a t govern­ m ent services as com posed of both the p er­ sonal services of em ployees and the goods and services governm ent purchases from business, including n et governm ent capital form ation; and we shall m easure the value of these services by the expenditures m ade for them . In only two estim ates w ere governm ent services evaluated in term s of the taxes raised fo r their financing instead of in term s of expenditures. These w ere the estim ates for the U nited States th a t Sim on K uznets prepared in th e 1920’s and early 1930’s fo r the N ational B ureau of E conom ic Research, an d those for C anada prepared by th e D o ­ minion Bureau o f Statistics in the 1930’s an d early 1940’s.3 In these estim ates gov­ ernm ent was considered to be acting like an entrepreneur. Taxes w ere interpreted as the prices governm ent charges the com m u­ nity fo r its services, and surpluses of rev­ enue w ere treated as governm ent profits or savings, while deficits o f revenue w ere re­ garded as governm ent losses o r dissavings. On the whole, this type of evaluation o f gov­ ernm ent services was unrealistic. Professor

195

K uznets attem pted to justify it in the 1920’s o n the ground th at at th a t tim e there really existed the “neat correspondence between taxes an d governm ent p ro d u ct” that his ap­ p roach im plied. On the other hand, in the late 1930’s, when governm ent expenditures in the U nited States included a large am ount of transfer paym ents as p art of the norm al situation, and w hen governm ent deficits be­ cam e large and regular, K uznets conceded th a t this n ea t correspondence no longer existed and th a t taxes no longer accurately reflected the value o f governm ent product.4 H e then abandoned this approach in favor of the conventional evaluation of govern­ m ent services in term s of th e expenditures m ade fo r them . A few years later, the C a­ nadian B ureau did the same. A s a general rule, services are evaluated in term s of the expenditures m ade fo r them as th e only logical and realistic m ethod of m easuring th eir value.

c. Classification of Government Expenditures In national incom e estimates th at try to give a full account of governm ent opera­ tions, governm ent cu rren t and capital ex­ penditures fo r goods and services consist of the aggregate value of the current services of governm ent production factors (labor and capital) at facto r cost an d of the gov­ ernm en t’s cu rren t purchases o f goods and services from private business at m arket prices. T ran sfer paym ents, disbursements fo r goods an d services sold to consumers, and expenditures fo r the acquisition of existing assets are excluded from the n a­ tional incom e com putation. G overnm ent expenditures m ay be analyzed either by their character o r object as expenditures for personal services, purchases o f materials and supplies, interest, etc., o r on a func­ tional basis as expenditures fo r national de­ fense, internal protection, education, public health, etc. A few estimates show govern­ m ent capital form ation, subdividing it func­ tionally as expenditures fo r buildings, high­ ways, etc. M ost of them , however, com ­ bine governm ent capital form ation with pri­ vate capital form ation (see C hapter 20, Section 2 b ).

d. Treatment of Government Services Dependent on the Philosophy of Government Functions In the case of the so-called “free” serv­ ices o f governm ent, i.e., those financed from

196

THE IN CO M E O F NATIO NS

general taxation o r from loans chargeable against it, the em pirical rule used in the m arket fo r differentiating final from inter­ m ediate products is inapplicable (see C hap­ te r 13, Section 1) inasm uch as these serv­ ices are not sold. T heir placem ent in one o r the other category depends on w hat view the estim ator holds of the function o f gov­ ernm ent as it relates to business and to the econom ic goals of the society generally. T here are three possible views: (1 ) all gov­ ernm ent services are interm ediate, (2 ) all governm ent services are final, and (3 ) some governm ent services are interm ediate and some are final.

e. The View That A ll Government Services Are Intermediate Products: Matolcsy and Varga M atolcsy and V arga, authors of th e H u n ­ garian estimates fo r 1924—37, cam e nearest to expressing the notion th a t all govern­ m ent services are interm ediate products and should not be included in national income. This notion is closely associated w ith the restricted m arket production concept of n a­ tional income. T he two authors expressed their point of view as follows: We do not regard [the services provided by the State on a non-profit making basis . . . as part of the national income] . . . , our reason being that the machinery required for keeping order and securing safety, as well as the whole civil service, do not produce values in addition to the flow of consumers’ goods, but ensure only the maintenance of the present economic and social order and the maintenance of the present level of production. The values these services produce are already included in the value of consumption goods (material and im­ material). If the cost of public administration increases, the national income does not there­ fore become greater, just as it would not be­ come smaller if the cost of public administra­ tion could be reduced, provided the prices of consumers’ goods remain unchanged. . . . We do not dispute the usefulness of the public service but it seems to us that the results of its usefulness appear in the value o f the goods and services produced, and an inclusion of the cost of public services as such would mean double counting. . . . In consequence the ex­ penditure of the State and the local authorities comes under the same heading as the costs of production, and is one of the main factors determining the price level.5 T he authors w ere ready to m ake an ex­ ception in the case o f public education and include a portion of its costs in national

incom e, b u t in the end decided against it. T hey w rote: Concerning the expenses of education, it seemed to us that, as part of the population, possessing a certain standard of knowledge, dies year by year, the costs of equipping suc­ cessive generations with the same amount of knowledge are similar in character to the main­ tenance charges of capital goods and could just as little be regarded as an addition to the na­ tional income as those. In fact, education pro­ vides more than the mere maintenance of a standard of knowledge, as the population is increasing in numbers and there is also a rise in the general standard of learning. The plus is immaterial wealth and using our definition ought to be regarded as income, but was never­ theless neglected as it eludes statistical ap­ proach. T h ey also excluded other services, which, according to their ow n adm ission, consti­ tu ted an exception to their sweeping rule, explaining: We also neglected that part of the expendi­ ture of the State and the local authorities which satisfies the popular desire for luxury and en­ tertainment (public concerts, shows, etc.) al­ though it ought, however small, to have been added to the national income.® T h e authors’ sweeping proposition th at governm ent services as a rule “do n o t p ro ­ duce values in addition to th e flow o f con­ sum ers’ goods” b ut m erely ensure “the m aintenance of the present econom ic and social o rd er and . . . present level of p ro ­ duction” is both arbitrary and co n trary to fact. I t can n o t reasonably be said to apply to m odern governm ent, w hich in a thousand w ays seeks to provide necessities and con­ veniences to its citizens com plem entary to those th a t business enterprise provides or th at they provide fo r themselves. C an it be said th at public health, hospital, and clinical services th a t com bat disease, prom ote hy­ giene, an d prolong h um an life “do n o t p ro ­ duce values in addition to th e flow of con­ sum ers’ goods,” an d th a t they do n o t con­ trib u te to a “general rise in the standard of living”? O r is this true of the services of public m useum s, parks, highways, street lighting, street cleaning? Is n o t th e feeling o f security in the enjoym ent o f one’s life, provided by the various protective services o f governm ent, a good th a t is additional to th a t provided by th e flow o f consum ers’ goods? W hile the authors allow b u t a few exceptions from th eir sweeping rule, the n u m b er of such exceptions could, by their

14.

T R E A TM E N T OF GOVERNM ENT

ow n definition, be m ultiplied till the w hole structure of their proposition w ould col­ lapse. But the greatest inconsistency in this con­ cept of governm ent services as n o t being final products is the fact th a t th e authors apply it only to the personal services of governm ent em ployees. T hey trea t goods and services purchased by governm ent from business as final products, and include them in national incom e. B ut how can these p u r­ chases— police cars, fire trucks, cem ent, road-building m achinery, etc.— be final products if the governm ent services th a t they help to produce— police and fire p ro ­ tection, road m aintenance, etc.— are not supposed to be final products? M anifestly, they cannot be final products m erely be­ cause they were produced by business. They m ust render final satisfactions to qualify as such; and if they render final satisfactions through the governm ent services they help to produce, then these governm ent services themselves m ust be final products. M ore­ over, since these goods are usable in the perform ance o f governm ent services only w hen com bined w ith the services of govern­ m ent em ployees, then obviously th e serv­ ices o f governm ent em ployees add a use value to these purchased goods th a t the la t­ te r do n ot possess in themselves; and the services o f governm ent em ployees m ust then be just as m uch final products as the goods in question. It is obvious th a t the M atolcsy-V arga theory o f the lack of final utility in governm ent services m ust be re ­ jected as m ade out o f w hole cloth.

f. The View That All Government Serv­ ices Are Final Products: Gerhard Colm T he opposite theory, th a t all services of governm ent are final products, was p u t fo r­ w ard at one tim e by J. R. H icks in England. H e expounded it n o t from conviction but, as K uznets characterized it, from “despera­ tio n ” at the statistical difficulties involved in using any other approach. T hus H icks w ro te: . . . the services of police, justice and de­ fence do contribute to production, and may be thought of as used in production in the same way as power or fuel. If we decide to give its full weight to this consideration only a fraction of the output of public authorities may have to be reckoned as entering into the final prod­ uct. And in this case a deduction from our various totals equal to a large proportion of public net income must be made. . . .

197

It is, however, extremely difficult to see how much deduction should be made. The protec­ tion of life and limb is presumably a part of final output, so is the use of the roads for pleasure purposes. How do we draw the line between the values of these services and the value of those services which ought to be de­ ducted? The division seems to be entirely arbi­ trary. Consequently, if we want to measure something and not to arrive at a figure for the national income which is what it is just because we say it is, it seems better to disregard these productive utilizations of public services, and to regard them (by definition) as being reckoned entirely into final output.7 Some ten years later, however, H icks re­ versed his position, recognizing th a t only those public services should be included in national incom e w hich can be clearly iden­ tified as being final products, adm itting th at this was possible statistically. H e wrote: I have never denied that there is a distinc­ tion between those government activities which have to be regarded as a part of final output, and those which (at least in principle) are not. But I used to think that the distinction was too vague to be of much use to the statistician. Later on . . . my wife demonstrated to me that the making of a significant classification of public expenditure on these lines was a much less formidable task than I had supposed.8 Recently, G erh ard Colm took up the cudgels on behalf o f the theory th a t all governm ental services should be treated as final products, reversing the position he had taken tw enty years before th a t these services are partly final and partly interm ediate p ro d ­ ucts.9 H e changed apparently n ot because of any theoretical reconsideration of the issue, b u t fo r the same reasons of expediency that had originally m otivated H icks. Colm w rote: I have come to the conclusion that such classification of government services [into in­ termediate and final products] would introduce sources of error which may be larger than the errors resulting from leaving the intermediate services of government in the national product or national income totals. Therefore, I believe now that it is best to make no deduction of the cost services of government, directly or indi­ rectly but to include all government services in the national product and national income.10 N ational incom e so m easured w ould in­ clude n o t only paym ents to factors employed in the production of governm ent services (in addition to including governm ent purchases from business), b ut also the indirect taxes from w hich some of the governm ent services are financed. N ational incom e at factor cost

198

THE IN CO M E OF NATIONS

and net national incom e at m arket prices w ould be identical. This proposal can be fully justified theo­ retically. It entails viewing governm ent som ew hat unconventionally but n o t u n ­ reasonably as an independent agent of pro ­ duction adding distinct social values to the individually created values of the goods produced by business. T he costs of produc­ ing these social values, reflected in taxes, are incorporated in the prices of final goods along w ith the other factor costs. O nly if some such theory is advanced can this p ro ­ posal be adopted. It cannot be accepted m erely on the ground o f statistical conven­ ience, fo r it touches upon one of the m ost basic conceptual issues in the national in ­ com e field. Simon K uznets believes th a t the theory of final utility o f all governm ent services is unrealistic because it controverts th e ob­ vious fact th a t some of the expenditures of governm ent represent the “costs of opera­ tion of society.” 11 T o include all govern­ m ent services in national product, he says, would result in double counting of some of the same values. H e elaborates this point as follows: . . . the flow of services to individuals from the economy is a flow of economic goods pro­ duced and secured under conditions of internal peace, external safety, and legal protection of specific rights, and cannot include these very conditions as services. To include the latter im­ plies feasibility of national income and of a flow of services to individuals outside the basic social framework within which economic ac­ tivity takes place. There is little sense in talk­ ing of protection of life and limb as an eco­ nomic service to individuals—it is a pre-condi­ tion of such service, not a service in itself. . . . It is difficult to understand why the net product of the economy should include not only the flow of goods to ultimate consumers, but also the increased cost of government activities necessary to maintain the social fabric within which the flow is realized.12 K uznets’ criticism is pungent, but not definitive. T he contradiction th a t bothers K uznets m ay n o t be a contradiction at all. It m ay be no m ore than a m ore m eaningful interpretation of the role of governm ent in term s of a broader view of the econom ic goals o f the society. In this view the eco­ nom ic goals w ould appear as being social as well as individual; and the pursuit of both would be regarded as essential to a well-rounded life. G overnm ent services and consum er m arket goods and services w ould

be regarded as com plem entary to each other and as equally furnishing final gratifications.

g. The Theory That Government Services Are Partly Final and Partly Intermedi­ ate Products: Kuznets’ Position T h e general tren d of cu rren t thinking on this problem is th a t som e services of govern­ m ent are final products and some are in­ term ediate products. M ost of the national incom e theorists in E ngland, the U nited States, and W estern E urope seem to share this view, feeling th a t it follows the present conventional concept of th e role of govern­ m ent in society. A t one tim e, Professor K uznets refused to take any position on the question. H e w rote in 1937: “. . . no classification of governm ent activities and expenditures by business o r ultim ate destination can properly be m ade.” H e am plified this by stating th at the functions of governm ent “have such a b road reference to the needs o f society at large th a t it is difficult to say th a t they serve business o r th a t they serve individuals as m em bers of the com m unity. If a definite answ er is provided it usually results from the application of some clear-cut position in social philosophy b ut one th a t does not necessarily have general validity.” 13 In his later w orks, K uznets did take a position on the issue, concluding th a t m ost o f th e activities o f governm ent are inter­ m ediate products. In 1951 he w rote that “m ost governm ent activities are designed to preserve and m aintain the basic social fram e­ w ork and are thus a species of rep air and m aintenance w hich cannot in an d of itself produce net econom ic retu rn s” ; and th at only in tim es of a crucial w ar can the preser­ vation of the social o rd er be placed above th e econom ic goals and w elfare of individ­ uals and thus becom e, w ith all its expendi­ tures fo r arm am ents, a final product. O rdi­ narily, th e final products of governm ent, ac­ cording to him, w ould consist of “(a) direct services by governm ent to ultim ate con­ sum ers plus (b ) additions by governm ent to capital stock, i.e., to th e stock designed to provide services to fu tu re ultim ate con­ sum ers.” K uznets’ w hole approach to the distinction between final and interm ediate p roducts o f governm ent stems from the as­ sum ption (o r social philosophy) th a t “the goal of econom ic activity is to satisfy wants of individual consum ers w ho are m em bers o f the nation, present an d fu tu re.” 14 A ccording to this rath e r restricted con­

14.

TR E A T M E N T OF GO VERNM ENT

cept of the goals o f econom ic activity, K uz­ nets sets up two criteria fo r identifying final products am ong the activities o f govern­ m ent other than those in w hich governm ent acts as an enterpriser: (a ) th a t “the govern­ m ent service be available to the individual only upon his overt initiative . . . [as in the case of the] services of a governm ent hospital, available to an individual upon request,” thus specifically excluding from this group services rendered to an individual “as a m em ber of a social group w ho, as an individual, m ay be quite unaw are o f the service . . . [as exemplified by] th e serv­ ices of the state legislature, higher judiciary, the arm y and navy, etc., fo r the preservation of the social order, and thus fo r protecting and extending the position o f an individual as a m em ber o f society” ; and (b ) “th a t the services by governm ent to individuals have an analogue in the private m arkets . . . [as exemplified by public] education, . . . m edi­ cal services . . . , parks, theaters, public tourist centers, am usem ents, etc.” K uznets realizes that this criterion m ay produce odd results. T herefore he fu rth e r qualifies it by stating that the analogous private service m ust be “w idespread.” H e w ould n o t classify police protection as a final pro d u ct m erely because “some people hire bodyguards.” A t the sam e tim e, he is also aw are o f the fact th at this qualification o f w idespreadness of a private analogue m ay result in a rem oval from the final products category certain pu b ­ lic services such as public elem entary edu­ cation, w hich in certain countries have al­ m ost com pletely displaced parallel private services. In fact, he considers governm ent services, w hich have thus becom e universal, as belonging in the category o f interm ediate products, for they then becom e a p a rt of the “social fram ew ork, like free justice, free right to participate in elections, and free police protection.” 15 A ccording to K uznets’ tw o criteria, all governm ent internal and external defense activity and econom ic and social regulation are interm ediate products. H e regards the increased expenditures of governm ent in m odern times as representing, in the m ain, not a net addition to th e national product, but rath er the increased costs of producing the sam e product. These increased costs, he says, are due to the “growing com plexity of the economy and the international fric­ tions w hich inequalities in the rate o f eco­ nom ic grow th am ong nations produce.” T hey are “an increased cost o f operating

199

the econom y, the other side of the shield o f econom ic progress.” 16 K uznets does n o t indicate how, on the basis of his criteria, he w ould classify serv­ ices of public roads. These services are not m ade available “on the overt initiative of the individual,” n o r do they have any coun­ terp art in th e services offered in the m arket. Should they, therefore, be classified as inter­ m ediate products? This w ould be contrary to th e fact th a t roads are used extensively by individuals fo r pleasure travel in addi­ tion to being used fo r business travel or fo r transportation of goods. M any other services w hich, on the face of it, give final satisfactions as well as ren d er services to business w ould be placed, on the basis of his criteria, in the category of interm ediate goods. Suffice it to m ention street cleaning, street lighting, sewage disposal, etc. W ith regard to the construction activities o f governm ent, K uznets is m uch m ore lib­ eral in recognizing them as final products. In fact, he classifies all governm ent n et capi­ tal form ation, including additions to the stock of arm am ents, as final products, dis­ associating capital form ation com pletely from its im m ediate service functions. H e w ould classify a new police station as a final good even though he classifies the police service itself as an interm ediate good. K uznets defends his position on this point as follows: Even if government capital is designed for turning out intermediate products alone (e.g. armament), changes in it should be included, because additions to such stock reduce the future cost of maintaining or extending the social framework which is indispensable for operation in the future, i.e. for the future out­ put of consumers’ goods. There is no inconsist­ ency in including in the final product of gov­ ernment changes in the stock of armament, and yet excluding from final product such govern­ ment activities as are carried on by the coun­ try’s armed services; as there is no inconsist­ ency in including additions to the stock of blast furnaces in net output, and yet excluding pig iron from the flow of finished goods to the country’s ultimate consumers.17 E xcept for net governm ent capital form a­ tion, K uznets’ criteria fo r distinguishing be­ tw een final and interm ediate government products tend to understate the proportion o f final products. T his is to be expected since his criteria are based on a rath er re­ stricted. individualistically oriented concep­ tion of the goals of econom ic activity.

200

THE IN CO M E OF NATIONS

M ost national incom e theorists and esti­ m ators w ho take th e position th a t some governm ent services are final products and some interm ediate products have proceeded from m uch broader definitions of econom ic activity and th e functions of governm ent than those provided by K uznets, and have, as a result, arrived at a m uch broader cate-

governm ent are final and some interm ediate products.20 H e separated governm ent ex­ penditures into six categories. U sing the fig­ ures of the com bined federal, state, and local expenditures of the U nited States fo r 1932— the latest year for w hich com prehensive data w ere available— he obtained th e following results: 21 M illion $

Consumption expenditures (which add to the individual comfort and standard of life of the citizens, as, for instance, expenditures for education, recreation, social hygiene, and welfare; part of government enterprise deficits, and state and local interest). Political services (rendered for the political organization’s own sake, for national prestige and power, or for the protection of the social order, i.e., for national defense, internal pro­ tection, justice, part of Department of Interior, etc.). Cost services (providing means of production either to produce for the market or to carry on the public enterprise itself, such as in large part construction and maintenance of roads, streets, bridges, etc.). Transfer expenditures (veterans’ pensions, farm subsidies, tax refunding, part of government enterprise deficits, federal interest, etc.). Capital subsidies (RFC; Federal Land Bank; farm credit, etc.). Debt retirement Miscellaneous Total Government Expenditures gory of final products am ong governm ent services. T he G erm an estim ates of the 1920’s, fo r example, w hich w ere am ong the first to give expression to the theory of the m ixed na­ ture of governm ent services, classified edu­ cation, health, general welfare, and rep ara­ tions expenditures as final products, but treated m ilitary expenditures as interm ediate products.18 In the old estimates fo r Sweden prepared by the Institute of E conom ic R e­ search, half of the expenditure o f the cen­ tral governm ent was assumed to represent final products; local governm ent expendi­ tures fo r education, child w elfare, public health, poor relief, and old age pensions were treated as final products, w hile the rem aining local expenditures w ere divided equally between final and interm ediate p ro d ­ ucts.19 In the official postw ar Italian esti­ mates, public elem entary and secondary edu­ cation is allocated between final and inter­ m ediate products in varying proportions, higher technical and professional public education is treated as an interm ediate prod ­ uct, while the expenditures for the adm inis­ tration of justice, national defense, and in­ ternal security are equally divided between final and interm ediate products. G erhard Colm, in a paper published in 1937, gave theoretical justification fo r the point o f view that some of the services of

3,961 1,755 3,182 2,195 893 905 435 13,326

Colm proposed th a t indirect taxes be included in the national incom e total, i.e., added to th e facto r shares, except those w hich financed w hat he called “cost serv­ ices” an d tran sfer expenditures. U sing this approach h e recalculated K uznets’ national incom e estim ate fo r 1932 in w hich indirect taxes w ere not included. H is figure was $5 billion higher th an K uznets’ ($44.6 billion as against $39.4 b illion). R. W . N elson and D onald Jackson, in th eir attem pts to classify governm ent ac­ tivities in the U nited States as either final o r interm ediate products, treated national defense expenditures as being equally di­ vided between the tw o categories; public school expenditures as being approxim ately 98 p er cent final products; ro ad expendi­ tures as being p ro p o rtio n ate to the use of the roads by passenger cars fo r pleasure, by passenger cars fo r business, an d by com ­ m ercial vehicles. T hey arrived at th e con­ clusion th a t about two thirds o f the govern­ m ent service expenditures in 1936 w ere for final products and only one th ird w ere for interm ediate products.22 W e do n o t propose to review and appraise th e various criteria used in classifying gov­ ernm ent products as final o r interm ediate. T he im portant things to note here are, first, th a t such classification is statistically fea­ sible; and second, th a t the theory th a t gov­

1 4 . T R E A TM E N T OF G OVERNM ENT

201

ernm ent services should be separated into these tw o categories is gaining ground and finding fairly effective practical expression. A recognition of these developm ents need not, however, im ply th a t this is a final solu­ tion to th e problem .

influenced A m erican estim ators to revise their original position and to consider cor­ porate n et incom e taxes as resting on stock­ holders rath er th an on the general consumer.

2. Treatment of Taxes

T he usual m ethod of treating taxes in estim ating national incom e under the com ­ prehensive production concept is to take factor incom es gross of direct personal and direct business taxes, and to ignore indirect business taxes. T he inclusion of direct per­ sonal incom e taxes occurs naturally, inas­ m uch as personal incomes in ordinary p rac­ tice are always reported gross of personal incom e taxes, and the question of doing otherw ise never arises. B ut in the case of direct business taxes, the situation is differ­ ent. In business practice, corporate net in­ com es are reported both gross and net of these taxes, o r in accounting language both “before” and “after” taxes, and the esti­ m ator can m ake a choice about how to enter these facto r incomes. In m ost cases, this issue has been resolved in favor of entering them gross of these taxes. In the case of indirect business taxes, the situation is even m ore difficult. These taxes affect individual incom es only indirectly through their price-increasing effects, i.e., by reducing th eir purchasing power, not their size. T h e taxes themselves are n ot a p a rt of individual income. Frequently, they do not appear in the gross incomes of busi­ nesses either, being com puted separately at each sale and set aside in a separate account. A ccordingly, in some n et o u tput or incomedistributed estimates, indirect taxes are auto­ m atically excluded from the accounting, b ut in others they do appear in gross business incom e and have to be excluded by the estim ator. T he exclusion of indirect taxes from the com putation seems perfectly logical until the estim ator w ants to reconcile his national incom e total w ith his national expenditure total. T hen he discovers th a t his income and expenditures figures do n o t jibe. The differ­ ence is due to the indirect taxes th a t are excluded from the incom e figure b u t are included in the prices of purchased goods and, hence, in the expenditure figure. T he estim ator begins to w onder w hether he should n o t have included indirect taxes in his national incom e calculations in the first place. B ut how can he do this and still re­ m ain tru e to the principle th a t national in­

T here is at least as m uch controversy in national incom e theory and as m uch differ­ ence in estim ating practice regarding the treatm ent of taxes as there is regarding the treatm en t of governm ent services and ex­ penditures. In fact, some of th e issues are identical and can be solved only co-ordi­ nately w ith those in the other field.

a. Classification of Taxes Taxes are classified som ew hat differently in national incom e estim ating from fiscal practice. T hey are divided into three m ain groups: direct personal taxes; direct busi­ ness taxes; and indirect business taxes. T he first includes only personal incom e, poll, and inheritance taxes; the second, corporate n et incom e taxes; and the third, all the re ­ m aining levies th a t are collected from or through business organizations and th a t m ay be shifted to consum ers through higher prices. In other w ords, the last category in­ cludes n o t only all form s o f business taxes, except those on net business incom e, but also all consum er levies (general sales or turnover taxes, special excises, and custom s duties) and even real estate taxes. T he latter are treated as indirect levies even w hen im ­ posed on hom e owners. T he category of in­ direct business taxes is, therefore, m uch broader in national incom e accounting than in fiscal accounting. Before 1947, corporate net incom e taxes in the U nited States estim ates w ere also deem ed to be indirect taxes; and corporate n et incom es w ere included in national in­ com e n et of these taxes. D uring th a t year, pursuant to an agreem ent w ith British and C anadian officials, the A m erican estim ators reclassified corporate n et incom e taxes as direct business taxes, i.e., as taxes im posed on stockholders and w ithheld at the source.23 T hey accordingly calculated corporate net incom e gross of these taxes, i.e., before pay­ m ent of them to the T reasury. By th a t tim e these taxes were being levied at exceedingly high rates (48 p er cent) and w ere sharply reducing profits available fo r distribution to stockholders o r fo r reinvestm ent. T his fact

b. Treatment of Taxes Under the Com­ prehensive Production Concept

202

THE IN C O M E OF NATIONS

come represents the value of products at factor cost? This is the dilemma that faces every estim ator. K uznets and certain other estim ators once offered w hat seemed to be a simple solu­ tion.24 Indirect business taxes, they declared, represent paym ents by business to govern­ m ent fo r services rendered to business, i.e., fo r interm ediate products; and fo r this rea­ son should be excluded from the com puta­ tion of national incom e that represents the value o f final products only. T he figure of indirect taxes in national expenditure com ­ puted at m arket prices contains the dupli­ cated item of the cost of governm ent in ter­ m ediate products. D irect personal taxes and direct business taxes are paym ents fo r the final products of governm ent and fo r gov­ ernm ent transfer paym ents properly charge­ able against factor incomes, and should, therefore, be retained. However, this ex­ planation ignored the fact th a t the choice between levying direct or indirect taxes is generally based upon political considerations having nothing to do w ith th e question as to who is the beneficiary of the services of governm ent— business o r the general con­ sumer— o r to w hat extent it is the one or the other. It was perfectly clear th a t the neat equation supposed to exist between the revenues collected from these two categories of taxes and the expenditures of governm ent on behalf of these two groups was purely im aginary; and the theory was soon aban­ doned. In m ost estimates, indirect taxes con­ tinued to be left out o f th e com putation of national incom e at factor cost, on the ground th a t they w ere not factor paym ents. A t the same tim e, however, it was realized that national incom e inclusive o f indirect taxes (especially if com puted gross of de­ preciation allow ances) coincided closely with national expenditure for private final products and fo r governm ent services and that it, therefore, lent itself m ore readily to analysis by type of products than n a­ tional incom e exclusive of indirect taxes. In line w ith this realization, national incom e estim ators began to offer national incom e totals both exclusive and inclusive of in ­ direct taxes, w hich they called, respectively, national incom e at factor cost and national incom e at m arket prices. T he latter term was appropriate, inasm uch as the addition of indirect taxes to factor costs raised the values of the final products to the prices actually paid fo r them in the m arket. W hile

this term was technically appropriate, -it left the theoretical basis of the aggregate in question obscure. T he question of w hether o r n ot the inclusion of indirect taxes in this new aggregate represented double counting of some of the services continued to c^ll fo r an answer. Eventually, tw o different solutions to this problem were offered. T he first of these was initiated by the authors of the official G er­ m an estimates during the late 1920’s.25 They started w ith the proposition th a t governm ent services are of a m ixed n ature, some being final and some interm ediate products; and th a t governm ent expenditures consist n ot only of outlays fo r services b u t also those fo r transfer paym ents. T hey added together the expenditures fo r final services and those fo r tran sfer paym ents and charged the re­ sulting total against the sum o f direct taxes, adm inistrative revenues, profits of govern­ m ent enterprises, and n et loans. If this total exceeded the am ount of these receipts they charged the excess against indirect taxes. T he charged portion of these taxes was then added to national income. T he uncharged portion was deem ed to represent the value of the interm ediate products of governm ent. T he national incom e total ob­ tained in this w ay was consistent w ith the concept of national incom e as the value of all final products. Indirect taxes were neither indiscrim inately excluded from national in­ com e n o r indiscrim inately included. They were included to the extent necessary to give effect to the theory th a t some services o f governm ent are final and some are inter­ m ediate. A variant o f this sam e approach is im ­ plied in the Swedish p rew ar estim ate in w hich th e dom estic private an d governm ent o u tput is first m easured at m ark et prices inclusive o f indirect taxes, an d next a de­ duction is m ade fo r th e cost of governm ent services deem ed to be rendered to p roduc­ tion. In b oth th e G erm an and the Swedish approaches, taxes used to finance transfer expenditures are logically included in n a­ tional incom e. Being excluded from national incom e as a n onfactor cost, th ey m ust be included in the taxes charged against the final products. Otherwise, the source of their paym ent will n o t appear in the national in­ com e total. T ransfer paym ents thus appear in national incom e only once as taxes chargeable against th e final product. In ter­ m ediate governm ent products sim ilarly ap­ p ear in the national incom e total only once,

14.

TR E A TM E N T OF GO VERNM ENT

in the value of the final product. B ut the value of the final governm ent product itself appears twice; once in the incom e of the purchaser of this p ro d u ct (the consum er) and once again as the incom e of the factors directly or indirectly engaged in its produc­ tion. This procedure involves no “double” counting. As Colm p u t it years ago, “this sort of double counting [of final products] originates from the very essence of the econom ic exchange process” and is not “double counting” in the ordinary sense of this term . “In m y incom e,” he w rote, “the portion that I spend fo r bread is calculated; and the same am ount appears once m ore in the incom e of the producers of bread. T he only criterion involved is w hether I m ake a genuine contribution to th e social product.” 26 A n entirely different solution was offered by Colin C lark in 1932. In his estim ates fo r the U nited K ingdom he included all in­ direct taxes on the ground th a t they raise the m arket value of all com m odities and services in the same w ay as do m onopoly prices.27 A lthough he never said so, C lark’s approach implied th a t all governm ent serv­ ices were final products. F acto r incom es, u n d er this approach, w ould be increased by indirect taxes so th a t factor incom es and indirect taxes taken together w ould cover all outlays fo r private and public goods and services, w hether m ade by producers o r by nonproducers. T he outlays by nonproducers are m ade from transfer paym ents. G erh ard C olm ’s recent proposal to trea t all governm ent services as final products would similarly lead to the inclusion of all indirect taxes in national incom e. U nder both Colin C lark’s and G erh ard C olm ’s ap­ proaches the disparity betw een national in­ com e at factor cost and national incom e at m arket prices w ould disappear, and the two aggregates would be identical. T hese theoretical explorations and p rac­ tical experim entations undertaken by the G erm an, Swedish, and A m erican econo­ mists have helped to clarify the issue as to how to treat taxes in national incom e com ­ putations, but it will take m uch m ore w ork to develop a definitive solution acceptable to all.

c. Treatment of Taxes Under the Ma­ terial and Market Production Concepts In estimates based on the m aterial pro ­ duction and m arket production concepts, both direct and indirect taxes are included.

203

This is generally done for very different reasons from those determ ining their com ­ plete or partial inclusion in other types of estim ates. F irst of all, they are included, w hile expenditures fo r the services o f gov­ ernm ent em ployees are excluded. Secondly, their inclusion is n ot intended to imply th at governm ent services are regarded as final products. T he taxes are included fo r one purpose only: to account fo r the value, in one case, of m aterial products, and, in the other, o f m ark et products diverted from producers to nonproducers. This is just as true of the estimates of Soviet Russia and Y ugoslavia as of those prepared by Fellner and by M atolcsy-V arga fo r H ungary o r by K alecki-L andau fo r P oland before the war.

3. Treatment of Subsidies to Business Subsidies to business are treated in m ost estim ates as a form of negative business tax, w hich should be deducted from the sum o f business taxes before the latter are in­ cluded in national incom e. They are so treated because business receives rath er than pays them . Except fo r some special cases they are presum ed to represent governm ent com pensation to business fo r selling its goods and services below m ark et prices. It is felt th a t subsidies, w ith some exceptions, correct the otherwise resulting underpay­ m ent o f the factors of production and undervaluation of products. T here are three m ain types of subsidies. O ne type consists of governm ent purchases o f products from private producers either at o r above m arket prices, and their resale to consum ers at low er prices. Obviously all estim ates of national incom e at factor cost autom atically include these subsidies, inas­ m uch as they include the entrepreneurial profits realized w ith the aid of these sub­ sidies. A n o th er kind consists of export boun­ ties o r certificates. In D uncan’s old esti­ m ates fo r Irelan d export bounties on agri­ cultural products w ere deducted from agri­ cultural incom e as m ere incom e transfers. In the M atolcsy-V arga estimates at m arket prices,28 these subsidies financed by the sale o f certificates to the buyers of w heat and rye were treated as an indirect tax, increas­ ing the prices paid by domestic consumers, and w ere accordingly included in the n a­ tional incom e at m arket prices in that form. A third type of subsidy consists of the direct cash paym ents m ade to producers as these

204

THE INCO M E OF NATIONS

paym ents are shown in their accounts. M ainly granted to agricultural producers, they are added to the official facto r cost es­ tim ates fo r the U nited States, the U nited K ingdom , Canada, and the M atolcsy-V arga estimates for H ungary. O perational losses of governm ental enter­ prises are sometimes considered to be an­ other type of subsidy. In the U nited States, estim ated deficits of governm ent enterprises, unincorporated o r incorporated, are re ­ garded as governm ent subsidies and are not deducted from national income. U sually, however, losses of governm ental enterprises are treated in the sam e w ay as deficits of private business and are subtracted from the net totals. In the U.S.S.R. estim ates in the 1930’s both approaches w ere used. T he deficits of producer goods industries w ere treated as subsidies and w ere included in national income, w hile th e deficits of con­ sum er goods industries were regarded as deductible losses.29

4. Treatm ent of Contributions to and Paym ents from Social Se­ curity, Pension, and W elfare Funds W ith the expansion of em ployees’ pension funds, social security systems, and private w elfare institutions in m any countries, new form s of incom e paym ents have appeared. On the one hand, employers, em ployees, or both, and sometimes governm ent, contribute to these funds; on the other hand, th e em ­ ployees receive pensions and other benefits from them . W hile em ployees’ contributions are in the nature of savings o r paym ents for welfare services, and those of th e govern­ m ent (except on behalf of its own em ploy­ ees) are merely transfer paym ents, em ploy­ ers’ contributions to pension funds o r direct pension paym ents are generally treated as supplem entary o r deferred paym ents fo r la­ bor perform ed. F requently, em ployers’ con­ tributions to social security are sim ilarly interpreted as supplem entary pay to labor, b ut sometimes are treated as indirect taxes. T he m ost com m on practice is to include em ployers’ contributions to both pension and social security funds in national incom e com puted at factor cost, as supplem entary labor com pensation. Occasionally, however, the benefit paym ents from these funds are included instead. In estimates of private or individual incomes, the benefit paym ents are

the ones th a t are usually included. In so far as the em ployers’ contributions in question are treated as supplem ental lab o r incom e costs, they becom e a p art of th e value of the final product. T he benefit paym ents o u t of social security, pension, an d w elfare funds, on the other hand, are treated as transfer paym ents and are not included in the value of th e final product. L ike all transfer paym ents, they are (or should be) accounted fo r in national income only once, i.e., through the charges m ade fo r them . T he services of the personnel ad­ m inistering social security, pension, and w elfare funds, and the purchases from busi­ ness used in the adm inistration o f the funds should be treated as final products, being in the n atu re of consum er services. Actually, it is n ot always clear in specific estimates how they are treated. In a great m any estim ates all social se­ curity contributions of em ployers are ex­ plicitly included as supplem ents to labor incom e. T his is tru e of those fo r the U nited States, the revised official estimates of C anada, an d those fo r G erm any, Switzer­ land, and the N etherlands. T he official esti­ m ates fo r the U nited K ingdom and A us­ tralia exclude these social security contri­ butions as indirect taxes. T he A ustralian estim ates also exclude em ployers’ contribu­ tions to pension funds an d the direct pension paym ents themselves. T hose fo r the U nited K ingdom include pensions subject to income tax. M any other estimates, such as the old estim ates fo r Italy, Bulgaria, an d G reece, include at least pensions paid to civil servants. O nly a few estimates, e.g., the C anadian, th e revised official Swiss, and the U.S.S.R. specifically include expenditures of indus­ tries fo r plant cafeterias, housing, vacations, in-service training, an d oth er such activities carried on for the benefit of employees. M any other estimates follow ing the net o ut­ p u t m ethod do essentially the same thing by n o t deducting these expenditures from the gross value of the output. In estimates o f aggregate private incomes, em ployers’ contributions to pension and so­ cial security funds are generally excluded, b ut pension paym ents and social benefits are generally included. T his is tru e of the old estim ates o f the N ational Industrial C onfer­ ence B oard fo r the U nited States, those of the Census and Statistics D ep artm en t for N ew Zealand, and th e old estim ates for A ustria. Similarly, D uge de Bernonville in

15.

AGGREGATES, ANALYSES, AND SOCIAL ACCOUNTS

his old estimates fo r F rance excluded em ­ ployers’ contributions and included only retirem ent pensions o f civil servants and railroad employees. Relief paym ents w ere included in the N IC B estimates fo r the U nited States.

205

A s can be seen from this brief exposition, the diversity of treatm en t of social security and pension transactions in existing esti­ m ates is confusing, and in the interest of greater com parability requires thorough­ going system atization.

15. VARIOUS AGGREGATES, ANALYSES, AND SOCIAL ACCOUNTS In recent years the presentation of national incom e data has been greatly enriched by the introduction of new types o f aggregates and analyses. F irst of all, th e old aggregates have been m ore carefully defined, subclassi­ fied, and supplem ented by new aggregates, distinguished from each other in the degree of th eir grossness o r netness, the character of elements included in them , and the p u r­ poses served by them . Secondly, the m ain national incom e aggregates are generally show n in two o r all three of their basic phases— nam ely, their origin, distribution, and expenditure. Thirdly, double-entry so­ cial accounts and input-output m atrices have been developed, partly as extensions o f the traditional national incom e estim ate and partly as independent analyses of the opera­ tions o f the national econom y and of its various sectors. E ach o f these new aggre­ gates and m ethods of analysis has given rise to problem s o f its own.

1. Various A ggregates and Their Interrelations T he fact has already been pointed out th at national incom e at facto r cost is service­ able m ainly in analyzing national incom e in its production and distribution phases, but is of less value in analyzing the disposition of national incom e. In the latter phase, n a­ tional incom e consists of final products usu­ ally evaluated at m arket prices rath e r than at facto r cost. It has also been pointed out th a t the treatm ent of governm ent services and, particularly, the treatm ent o f indirect taxes is am biguous u nd er this concept be­ cause its im plication th a t indirect taxes m easure the interm ediate products of gov-

Notes begin on p. 533.

ernm ent is n ot theoretically valid (see C hap­ te r 14, Sections lb , lg , and 2 b ). This am ­ biguity lim its the usefulness of the concept o f national incom e at facto r cost n o t only fo r analyzing the disposition of national incom e or product, b u t even fo r analyzing its production and distribution. A n additional reason fo r fram ing new concepts and aggregates is th a t certain types o f econom ic analysis require incom e or produ ct aggregates th a t include nonfactor shares in addition to facto r shares, and th at otherw ise vary in th eir coverage o r in the degree of th eir grossness o r netness.

a. Nature of the Various Aggregates M ost of the additional incom e and p ro d ­ uct aggregates relate to at least two phases of national incom e o r product. T he follow­ ing schedule lists these additional aggregates together w ith the prim ary one, and shows the phases to w hich each relates.

1. 2. 3. 4. 5. 6. 7. 8.

Concept National income or product at factor cost Net national income at market prices Gross national product Private income Personal income, variant one Personal income, variant two Disposable income Aggregate of gross receipts of all enter­ prises (general turnover)

Phases to Which Related Production, distribu­ tion and, in a lim­ ited way, disposition Production and dispo­ sition

Distribution and dis­ position

Disposition Production

206

THE IN CO M E OF NATIONS

Net national income at market prices differs from national incom e o r product at factor cost in that it includes all indirect taxes, business as well as consum er excises collected by business, and excludes subsidies th a t reduce final product values. In other words, it is m ade up o f consum er goods and services (including the services of gov­ ernm ent em ployees) and net investm ent at prices actually paid fo r them . J. R. H icks has suggested that national incom e at m ar­ ket prices is especially suitable fo r the m eas­ urem ent of econom ic welfare, w hereas na­ tional incom e at factor cost is m ost appro­ priate fo r the m easurem ent of th e produc­ tivity of the econom y and especially o f its productive factors.1 This distinction is use­ ful, although it cannot be accepted w ithout some reservations.2 N et national product at m arket prices can be calculated in two w ays: (1 ) by add­ ing indirect taxes less subsidies to net n a­ tional incom e at factor cost as already m en­ tioned, o r (2 ) by com puting th e value of final goods and services (including services of governm ent em ployees) at their sales price. Since the direct approach m easures output at its final stage of readiness fo r use, it shows national product in its disposi­ tion as well as in its production phase. In fact, one w ay of estim ating net national product at m arket prices is to add up the final sales values of the various kinds o f con­ sum ption and net investm ent products. The total obtained in this w ay usually m akes pos­ sible a com parison between private con­ sum ption expenditures and net investm ent on the one hand, and public consum ption expenditures and net investm ent on the other. It also affords a basis fo r calculating national incom e in constant prices. If indirect taxes are excluded from n a­ tional income because they are deem ed to reflect the services governm ent renders to business, then their inclusion in national in­ com e at m arket prices results in doubly counting these services— once as such and once as the indirect taxes supposed to re­ flect their costs. In this particular respect, then, national incom e at m arket prices is no longer a “n et” figure. K uznets at one tim e avoided any con­ ceptual duality by including in national in­ com e at m arket prices only those govern­ m ent services rendered to consum ers th a t he felt could be m easured by personal taxes paid in peacetim e. A ccording to his defini­ tion, net national incom e at m arket prices

consisted o f the flow o f goods and services to consum ers (including personal taxes in p eacetim e); net nonw ar capital form ation; and the o u tput of w ar goods an d services net of depreciation; and was equivalent to a sim ilarly restricted concept of national incom e at facto r cost.3 Gross national product ( G N P ), w hich is the n et national p ro d u ct a t m ark et prices plus depreciation allowances (o r th e value o f cu r­ ren t capital replacem ents), is th e value o f all final goods and services produced by private enterprise an d governm ent fo r either con­ sum ption o r gross capital form ation. By per­ m itting th e study o f short-run variations in personal and public expenditures and of th eir effects on capital form ation, G N P plays an im portant role in com parisons be­ tw een private and governm ent expenditures in w artim e because depreciation reserves are th en used fo r financing w ar consum ption instead of capital replacem ents. Capital re­ placem ents are partly o r fully suspended for the duration. Stocks m ay n o t be replenished, while foreign assets m ay be liquidated at such tim e to m eet w artim e needs. T h e next fo u r associated national incom e concepts o r aggregates relate to the distribu­ tio n and disposition of national incom e. T hey exclude certain com ponents of n a­ tional incom e at fac to r cost an d include certain incom e transfers arising from a sec­ ondary distribution of incom e th a t are n o t themselves facto r shares. T h e first of these, private income, is in­ com e obtained by private individuals from any source, productive o r otherw ise, and the retained incom e o f corporations. It can be derived from net national incom e at facto r cost by m aking certain deductions and additions. T he deductions include in­ com e from governm ent enterprises and as­ sets as well as revenue surpluses, and em ­ ployers’ contributions to social security, pen­ sion, and w elfare funds. T h e additions in­ clude the vast variety o f tran sfer paym ents such as social security, pension fund, and veterans’ benefits, assistances to indigents, and interest on the unproductive national debt. O f course, these deductions an d addi­ tions are m ade only if th e item s are included in o r excluded from the original calculation of national incom e at facto r cost. Private incom e as a statistical aggregate is useful in revealing the distribution o f purchasing pow er am ong social and econom ic groups, and the disposition o f th a t purchasing pow er as betw een expenditures fo r consum er goods

15.

AGGREGATES, ANALYSES, AND SOCIAL ACCOUNTS

an d services, paym ents of direct taxes, and individual and corporate saving. Personal incom e differs from private in­ com e in that it excludes undistributed profits o f corporations, co-operatives, and sim ilar organizations. If accurately defined, it is the basic m agnitude fo r studies of th e dis­ tribution of individual and fam ily incom e as well as of individual and fam ily tax b u r­ dens. I t has tw o variants— one in w hich sal­ aries and wages are given before deduction o f employee contributions to social security and pension funds, and the other in w hich they are given after such deduction. Disposable incom e, consisting of aggre­ gate personal incom e paym ents m inus direct taxes, m easures the purchasing pow er of consum ers. It plays an im portant role in governm ent econom ic planning as well as in th e private planning o f farm ers, m an u ­ facturers, m erchants, bankers, etc. Finally, the aggregate gross receipts, or turnover value of the output, is the “gross­ est” o r least refined of all the incom e aggre­ gates developed so far. It contains the great­ est am ount o f double counting. It represents the actual financial transactions taking place in the econom y w ith the least am ount of “abstraction” o r theoretical m odification. It is closest to th e original eighteenth- and nineteenth-century concepts o f “gross value” of the national o utput and also to the con­ tem porary laym an’s concept of it. It is used particularly in countries such as F rance and Soviet Russia, having a general gross re­ ceipts o r turnover tax. It is an aggregate readily derived in these countries from the yield o f this tax, and, conversely, also serves there the practical purpose of estim ating the future yields of the tax. Finally, it also serves as a basis fo r estim ating the other, m ore refined types o f incom e aggregates. T he aggregate gross receipts concept is re ­ lated to the production phase of the national incom e and is useful in the analysis of the actual structure o f the national economy. It has been given a new lift in m odern tim es by the introduction o f the “inputo u tp u t” approach discussed fu rth e r below.

b. Interrelationships of Various Aggregates Including national incom e at factor cost, eight income concepts and aggregates are in use today. Since all are interrelated, they generally can be reconciled w ith one an­ other. Beginning w ith disposable income, it is possible by successive additions o r de­

207

ductions to derive m ost of the other aggre­ gates. T able 15-1 shows these interrela­ tionships in actual figures fo r two countries fo r the year 1945. T he grossest aggregate— th a t of the gross receipts or turnover value of the o u tput— is n ot included in this table inasm uch as it does n ot appear in the cal­ culations of these tw o countries. A m ong the estim ates th a t derive n et n a­ tional p roduct at m arket prices from n a­ tional incom e at facto r cost, there is general agreem ent on m ost points concerning m ethod. Before indirect taxes are added they are generally reduced by the subsidies paid businesses fo r the specific purpose of reducing m arket prices o r of com pensating them fo r their low-profit or no-profit prices. In the estimates fo r the U nited States, busi­ ness transfer paym ents, such as corporate gifts to nonprofit institutions and allowances fo r consum er bad debts, w hich are n o t in­ cluded in national incom e at factor cost, are added in the com putation of the n a­ tional incom e at m arket prices (see T able 15-1). In com puting gross national p ro d ­ uct from national incom e at m arket prices, some estim ates account fo r capital costs charged to cu rren t expenses and fo r bad debts. T he m ethod of com puting personal in­ com e from national incom e does n ot differ m uch from estim ate to estim ate (see Table 15-2). E m ployers’ contributions to social security funds are deducted in m ost esti­ m ates, b u t are n o t in the official British and A ustralian estimates fo r the simple reason th a t they w ere n o t included in national in ­ com e at facto r cost originally. N early all countries now exclude interest paym ents to individuals on the unproductive national debt from national income, and m ust add them in figuring personal income. In the official A ustralian and D utch esti­ m ates, th e capital incom e of charitable and life insurance institutions was deducted. In com puting personal incom e in the official A m erican and C anadian estimates, em ploy­ ees’ social security contributions are de­ ducted; in the A m erican estimates, business transfer paym ents to individuals are added. M ost of the recent estimates present all or m ost o f these seven national incom e aggregates. N ational incom e at factor cost, national incom e at m arket prices, and gross national p ro d u ct are fairly universally pre­ sented. Even in countries in w hich statistics are poor, the estim ates often give separate figures fo r direct and indirect taxes, sub-

208

THE IN CO M E OF NATIONS

Table 15-1. Interrelations Am ong Seven Income or Product Concepts

2

3

U nited States, 1945 » (billion $)

U nited Kingdom , 1945 * (million £ )

Gross national product M in u s depreciation allowances or costs of current capital replacem ents and capital costs charges to current accounts and other capital consum ption allowances

2 1 3 .1

10,191 1 0.6

1 .5

520

N ational income or product at m arket prices M in u s adjustm ent for discrepancies M in u s business transfer paym ents b M in u s indirect taxes plus su bsid ies0

2 0 1 .0

9,6 7 1 3 .1 .6 1,188

N ational incom e or product at factor cost P lu s governm ent transfer paym ents d [retirement pensions, social security benefits, veterans’ and relief benefits, and interest on national (or total) governm ent debt] and business transfer paym ents M in u s supplem ents to labor incom e (em ployers’ contributions to social security and pension funds and to em ployee welfare institutions) and profits of governm ent enterprises and from governm ent property

1 8 2 .8

1 4 .6

9 .3 e .6

8,4 8 3

891 1

3 .8 8 123

4

Private income M in u s corporate savings before taxes

188 .8 1 4 .9

9,251 900

5

Personal income, variant one M in u s em ployee contributions to social security & pension funds

173.9 2 .3 '

8,351 69 b

6

Personal income, variant two M in u s direct personal taxes Disposable incom e ‘ M in u s net savings of individuals Consumer expenditures

7

171 .6 2 0 .9 150.7 2 9 .0 121.7

8,2 8 9 1,374 6,915 1,270 5,6 4 5

a Unrevised. b Corporate gifts to nonprofit institutions and allowances for consumer bad debts, unrecovered thefts, etc. 0 For detailed specifications of indirect taxes see Chapter 14, Section 2a. d War gratuities and postwar credits of members of the armed forces, pension paym ents, unem ploy­ m ent and relief benefits, interest on the national debt, and other transfer paym ents. e Federal and state social security benefits, m ilitary pensions and disability paym ents, mustering-out paym ents to discharged members of m ilitary forces, readjustm ent, self-em ploym ent, and subsistence allowances to veterans and other m ilitary benefits; civil service and railroad em ployees’ pensions, and interest on all governm ent debt— federal, state, and local. f Em ployers’ contributions to national insurance are included in indirect taxes. e Contributions to old age and survivors’ insurance, state unem ploym ent insurance, federal unem ploy­ m ent tax, railroad retirem ent and unem ploym ent insurance, federal civil service retirem ent system , state and local em ployee retirem ent system s, cash sickness com pensation funds, and governm ent life insurance. h Contributions to national insurance not deducted from personal income in the original estim ates. 1 Contributions to old age and survivors’ insurance, state unem ploym ent insurance, railroad retire­ m ent insurance, federal civil service retirem ent system , state and local em ployee retirem ent system s, cash sickness com pensation funds, and governm ent life insurance. Contributions to private pension funds, other than railroads, and pensions paid out are included in both variants of personal incom e. ‘ W ages, salaries, interest, dividends, rent, earnings of self-em ployed persons, and transfer incom es (retirem ent pensions, social security, old age pensions, veterans’ benefits, relief) paid to individuals minus direct taxes and em ployees’ contributions to social security and pension funds. s o u r c e : Compiled from Survey o f Current B usiness, July 1947, Supplem ent, pp. 19, 20; British W hite Paper on War Finance, Federal Reserve B ulletin, July, 1946, pp. 723, 732, and 737.

15.

209

AGGREGATES, ANALYSES, AND SOCIAL ACCOUNTS

Table 15-2. Characteristics of Personal Income in Estim ates of F ive Countries U .S .“ Em ployees’ contributions to social security and pension funds Em ployers’ contributions to social security and pension funds Undistributed profits of corporations Dividends, rent, and interest Capital income of governm ent Capital income of life insurance Interest on national debt Social security and pension fund benefits M ilitary benefits and other governm ent nonfactor transfer paym ents Relief benefits Subsidies to consumers (fam ily subven­ tions, etc.) Business transfer paym ents (accident com ­ pensation, etc.)

0

U.K .b C anada0

Ireland d Switzer

X

0

Xf

0

0 0

X

X 0

0

X

0 X

0 X

0 X X

0 X X

0 X X

0

X

X 0

X

X X X

X X

X X

X X

X X

0

X

X

X

X

X

X

X

X

0 X 0

X 0 X X

X

n o t e : x: Item included.

0: Item excluded. sources:

a U .S. D epartm ent of Commerce, N ational Income, 1954 edition, Supplem ent to Survey of Current B usiness, W ashington, 1954, pp. 50-51 and 164. b Central Statistical Office, N ational Incom e and E xpenditure, 1946—1954, London, August, 1954, pp. 14, 75-76. 0 Dom inion Bureau of Statistics, N ational A ccounts, Incom e and E xpenditure, 1926—1950, O ttawa, 1952, N o. 1, pp. 34, 116-117. d Central Statistics Office, Irish Statistical Survey 1953, (pr. 2607), Dublin, 1954, pp. 50, 60. 8 O EEC, N ational A ccounts Studies, Sw itzerland, Paris, 1951, pp. 23-37, 41. 1 Em ployees’ contributions to social insurance included, but em ployees’ contributions to contributory pension funds excluded.

sidies, and business transfer paym ents, and give estimates o f personal incom e of one or the other variant, and of disposable income. Private incom e is a m uch less im portant aggregate than the others, and is often om itted. T he Statistical A ppendix at the end of this book shows that, ou t of the 87 country estim ates on record in 1957, 58 gave the country’s gross o r net dom estic output at facto r cost, 67 the national incom e at factor cost, and 60 the gross national product at m arket prices. A ll these aggregates are given in the national currencies. In alm ost every case the aggregate is broken dow n into its com ponents.

2. Social A ccounts By 1956, the system of double entry so­ cial accounts, first proposed by M eade and Stone in 1941, was adopted in a large p ro ­ portion of national incom e estimates. The N ational A ccounts R esearch U n it of the Organisation fo r E uropean E conom ic C o­

operation (O E E C ), originally headed by Stone and later by M ilton G ilbert, built its series of national incom e studies for various countries entirely upon this approach and has been particularly instrum ental in bring­ ing about its w ider adoption.4

a. Types of Accounts M ost generally, fo u r basic sector accounts and one consolidated account are set up. T he fo u r sector accounts are: the business sector; the personal o r household consum er sector; the governm ent sector; and the rest of the w orld sector. T he consolidated sector covers the com bined transactions of these sector accounts. Sometimes, as in the U nited States, a fifth sector account is set up th at covers gross saving and investment, and another consolidated account is presented, know n as the “N atio n ’s Econom ic Budget.” In the U nited States estimates, the six accounts fo r 1953 (om itting certain details) are show n in T able 15-3. It will be noticed th a t the U nited States social accounts reproduced below— and this

Table 15-3. The Six Social A ccounts of the U nited States, 1953 1. N ational Income and Product Account (billion dollars) Charges Against Gross National Product Com pensation of em ployees (wages, sala­ ries, and supplem ents) 209 .1 Incom e of incorporated enterprises and inventory valuation adjustm ent 3 8 .4 R ental income of persons 1 0.6 Corporate profits and inventory valuation adjustm ent 3 8 .5 8 .4 N e t interest N ational income

Gross N ational Product Personal consumption expenditures Gross private dom estic investm ent N e t foreign investm ent Governm ent purchases of goods and services

230 .1 5 1 .4 1 .9 8 5 .2

3 0 5 .0

Indirect business tax and nontax liability Business transfer paym ents Statistical discrepancy Less: subsidies minus current surplus of governm ent enterprises Capital consum ption allowances

3 0 .0 1 .0 1.1 - .5 2 7 .2 3 6 4 .9

T otal

T otal

3 6 4 .6

2. Business Incom e and Product Account Business Gross Product Consolidated net sales T o consumers T o governm ent T o business on capital account T o abroad Change in inventories

Charges A gainst Business Gross Product Com pensation of em ployees 169.5 Income of unincorporated enterprises and 3 8 .4 inventory valuation adjustm ent R ental income of persons 1 0.6 Corporate profits and inventory valuation 3 7 .3 adjustm ent 5 .0 N e t interest Income originating

2 61 .2

Indirect business taxes and nontax liability Business transfer paym ents Statistical discrepancy Less: subsidies m inus current surpluses of governm ent enterprises Charges against net product Capital consum ption allowances

3 0 .0 1 .0 1 .0 - .5 2 9 3 .8 2 7 .2 T otal business gross product

T otal charges against business 3 2 1 .0 gross product

3. Personal Incom e and Expenditure Account Personal consum ption expenditures: Purchases of direct services Compensation of em ployees: (dom estics, etc.) 7 .9 Interest paid 3 .0 N e t purchases from businesses N e t purchases from abroad Personal tax and nontax paym ents Personal saving

W ages and salaries and other labor income 2 0 4 .4 From business 166.0 From governm ent 3 0 .5 From households and institutions 7 .8 From rest of the world (a) Incom e of unincorporated enterprises and inventory valuation adjustm ent 38.4 Rental income of persons 10. 6 D ividends 9.4 Personal interest incom e 13.5 G overnm ent transfer paym ents 12.8 Business transfer paym ents 1 .0 Less: personal contributions for social in ­ surance • —4 .0

10.9

2 17 .1 2.1

3 6 .0 20.0

T otal personal incom e

T otal personal outlay and saving 286.1

(a) 20 millions. 210

286 .1

1 5 . AGGREGATES, ANALYSES, AND SOCIAL ACCOUNTS

211

4. Governm ent Receipts and Expenditures Account G overnm ent Expenditures Purchases of goods and services: Purchases of direct services (compen­ sation of em ployees including sup­ plements, etc.) N e t purchases from business N e t purchases from abroad Transfer paym ents N e t interest paid Subsidies minus current surplus of govern­ m ent enterprises T otal

3 1 .4 5 0 .3 3 .5 1 2.8 5 .0

G overnm ent Receipts and Deficit Personal tax and nontax receipts Corporate profits tax accruals Indirect business tax and nontax accruals Contributions for social insurance Personal 4 .0 Employer 4 .7 D eficits ( + ) or surplus ( —) on income & product transactions

36.0 21.1

30.0 8 .7

6.6

—. 5 102.5

T otal

102.5

5. R est of the World Account N e t Current P aym ents to the United States N e t paym ents of factor incom es to U .S. 1.5 W ages and salaries (a) Interest .3 D ividends .4 Branch profits .7 N e t purchases from the U nited States —3 .4 From business 2 .2 From governm ent —3 .5 From persons —2 .1 T otal (a) 20 millions.

N e t D isinvestm ent in the United States N e t disinvestm ent in the United States

T otal

- 1 .9

- 1 .9

- 1 .9

6. Gross Saving and Investm ent Account Gross Investm ent Business purchases on capital account 4 9 .9 Change in business inventories 1 .5 N e t disinvestm ent in the U nited States by rest of world —1 .9

T otal

4 9 .5

Gross Saving and Statistical Discrepancy Excess of wage accrual, over disbursements (business and governm ent) —. 1 Undistributed corporate profits (domestic) 8 .2 Corporate inventory valuation adjustm ent —1 .0 Capital consum ption allowances (business) 27.2 G overnm ent surplus or deficit on income and product transactions —6 .6 Foreign branch profits (net) .7 Personal saving 2 0 .0 Statistical discrepancy 1 .0 T otal

4 9 .5

s o u r c e : U .S. D epartm ent of Commerce, National Income, 1954 E dition, W ashington, D . C., 1954,

pp. 160-61.

212

THE IN CO M E OF NATIONS

is true of the O E EC Standard System of A ccounts and of m ost of the individual country social accounts— attem pt to deal sim ultaneously and in an interrelated m an­ n er w ith incom e and expenditure transac­ tions and w ith the supply and use of p ro d ­ ucts. T hus income is shown both as a re­ ceipt and as a charge against the product, and the expenditure both as a source of in­ com e and as a use of product. T he sim ul­ taneous presentation of so m any different phases of the same transactions in the form s of debits and credits is, to any person not fully conversant w ith accounting techniques, difficult to follow.

b. The Area of the Greatest Usefulness of the Social Accounting Approach In 1948, w hen the official U nited States estimates incorporated the social accounting approach fo r the first tim e, Professor K uz­ nets gave it only a lim ited endorsem ent. W hile acknowledging its value as a subsidi­ ary analysis of the operations of the national economy, he w arned against excessive over­ optim ism as to w hat it can accomplish. H e w ro te: There is little in the technique of the system of accounts in and of itself to help us deter­ mine the proper scope of national income and the observable flows that represent net yields and those which, from the standpoint of the national economy, represent costs; to decide upon the bases of valuation to be used, and of the significant sectors to be distinguished at any level of economic circulation.5 These basic questions can be settled w ith­ out the aid of double entry accounts, and in fact should be settled before construct­ ing any. This was readily adm itted by the chief A m erican protagonists of the social accounting approach, nam ely, M ilton G il­ bert and his associates, in their reply to Professor K uznets’ critique.6 N ational in­ come m ust not only be well defined theo­ retically, but it should also be calculated by standardized statistical methods. Only after its calculation has been com pleted by such m ethods can its elements ordinarily be apportioned am ong the various sector accounts. In the course of this apportion­ m ent it is possible to verify the correctness and com pleteness of the original com puta­ tions of these elements and of the national incom e totals themselves. M oreover, through

the construction and analysis of the sector accounts it is possible to show th e opera­ tions of the national econom y in m uch greater detail and w ith greater closeness to reality th an can be done through trad i­ tional national incom e estimates. T he pres­ entation of the national incom e data in the form of social accounts, according to K uz­ nets, is, therefore, only an extension of the traditional national incom e estim ating p roc­ ess. A s such, the social accounts are essen­ tially a derivative rath e r th an an original type of com putation, and a subsidiary rath er th an a prim ary form o f presentation of n a­ tional incom e data. T his was essentially the substance of Professor K uznets’ critical com m entary. It is n o t at all clear at this m om ent w hat exactly is the area o f greatest usefulness of the social accounting approach and w hat are the lim its beyond w hich it should n ot be extended. Professor K uznets, fo r one, de­ fined this area and these lim its rath e r n a r­ rowly, w hen he w rote in 1948: . . . [Admitting that] the contribution of the system of accounts to the solution of problems of scope, netness, and valuation in defining national income is either zero or may even be negative . . . nevertheless, the latter can con­ tribute materially to the progress of under­ standing and analysis in the field, in two direc­ tions. The first is in the clear exposition of the re­ lations between the volumes of transactions which the system of accounts details and the “netter” total (or component) that has been defined by logically antecedent canons. . . . The second is in the development of entire families of gross totals of volumes of transac­ tions, without any attempt at the “netness” that is associated with national income.7 M ilton G ilbert and his associates, on the oth er hand, conceived the applications of this new approach in m uch b ro ad er term s. T hey believed th at the social accounts could be the prim e sources fo r th e construction of national incom e estimates. T hey regarded K uznets’ restrictions on the uses o f social accounts as unnecessarily lim iting. In their reply to him they defined the advantages of the new system as follows: (1) . . . [it] is a tremendous aid in revealing the structure of the economy. (2) . . . [it] provides a powerful tool for the solution of many intricate problems that arise in the consistent formulation of national income concepts . . . [e.g., in] the treatment of

15.

AGGREGATES, ANALYSES, AND SOCIAL ACCOUNTS

financial intermediaries, non-profit institutions, imputed incom e. . . . (3 ) . . . [it is especially useful] as a peda­ gogical device for explaining the nature o f national incom e statistics and the interrelation­ ship o f the various aggregates and their com ­ ponents. . . . (4 ) . . . [it is] a great aid in defining the task o f statistical collection . . . a list o f re­ quirements obtained in this w ay provides the only assurance that the collection o f primary statistical data w ill be organized so as to yield the inform ation m ost relevant to econom ic analysis. . . . (5 ) the accounting approach facilitates the estim ation o f the various national incom e ag­ gregates and their com ponents from the avail­ able statistical m aterial . . . by m aking clear that m any items o f inform ation can be o b ­ tained from the records o f either the buyer or the seller, and hence affords flexibility in adopt­ ing estimating m ethods to available inform a­ tion.8

c. The Need for Balance Between the National Income and the Social A c­ counts Approaches Some of the adepts o f the social account­ ing approach seem to assign to it a role wholly independent from th a t o f th e esti­ m ation of national income. T hey no longer consider national incom e to be the prim e objective of an integrated statistical analysis dealing w ith the operations of the national economy, but view it as m erely one o f a num ber of pertinent item s of interest. Some of its partisans even propose to dispense w ith estimates o f national incom e com ­ pletely. The function of social accounts, in th eir judgm ent, is to reveal as closely as possible the actual “circulation and produc­ tion process” in the econom y, and this, they believe, can be done outside the national in­ com e concept. Some o f the N orw egian economists, R. F risch and O. A ukrust, for example, seem to be inclined this way. Some Swedish econom ists, w ho have en­ thusiastically em braced the social account­ ing approach, do n o t go quite th a t far. Ingvar Ohlsson, in presenting his social ac­ counts fo r Sweden, concedes th a t “national incom e statistics have been the starting point o f m odern national accounting” and th a t “national incom e statistics can easily be converted to sim ple current national ac­ counts” ; but he then proceeds to m ake n a­ tional incom e a derivative of social accounts, saying: “In theory, the curren t national ac­ counts can be said to form the basis. N a ­

213

tional incom e statistics can be considered as com puted (derived) through systematic sum m ations (extractions) of transactions (item s) in the cu rren t national accounts.” 9 T he social accountants’ underlying as­ sum ption th a t their double entry accounts are real accounts existing in the economy does n o t accord w ith the facts. T he only real account is th a t o f the governm ent sec­ tor. A ll the other sector accounts, and the consolidated account itself, are no m ore th an abstractions. T hey are the creations of the social accountant just as m uch as the national incom e estim ate is, if not m ore so. T heir substance depends entirely on w hat th e accountant decides to p u t into them or to keep o ut of them . N o t every transaction of daily life by far is included in the ac­ counts, and, on the other hand, some tran s­ actions th a t are n o t real b u t im puted are adm itted in them . W hat is to be included and w hat is n ot is determ ined by economic rath er th an accounting considerations, just as it is in the case of a national incom e estim ate. W hen F risch, in introducing his “O kosirk” system of social accounts fo r N orw ay, says, “On the one h and it appears entirely elem entary; yes, so elem entary th at one alm ost feels em barrassed to consider this seriously as a scientific problem ,” 10 the reader m ay well expect th at w hat will ap­ p ear “on th e oth er h an d ” is very com pli­ cated and th a t w hat is said to be elem entary “on the one h an d ” is n ot really so elem en­ tary after all. T he spread of th e social accounting ap­ proach during th e decade following W orld W ar II has been truly phenom enal. The approach has w on adherents particularly am ong economists trained in business ac­ counting. It has received its greatest boost from the N ational Incom e Research U nit o f the O EEC , w hich has n o t only published a w hole series of social accounting studies covering different countries, but has also given technical assistance to statistical agen­ cies of a num ber o f m em ber countries of the O E E C in the developm ent of th eir own social accounts. T he Scandinavian econo­ mists (Frisch, A ukrust, Lundberg, Ohlsson, and others) have extended the social ac­ counting approach far beyond M eade and Stone’s original pattern, while economists in a score of other countries have applied it at least in this original fram ew ork. P ro ­ fessor R. Ruggles in the U nited States pub­ lished a textbook in w hich the fram ew ork

214

THE INCO M E OF NATIONS

of social accounts was built theoretically from the double entry accounts of individ­ ual firms, households, and governm ents.11 In F rance, Frangois Perroux was chiefly responsible fo r popularizing the social ac­ counting approach.12 T here can be no doubt th at the social ac­ counting approach is here to stay. A t the same tim e, it is probably still m uch too early to define the exact nature and extent o f the contribution it is likely to m ake to national incom e analysis. In the U nited States the N ational Bureau of Econom ic Research set up in 1956 a com ­ m ittee of distinguished scholars, w ith R ay­ m ond W. G oldsm ith as chairm an, to review the status of the existing U nited States n a­ tional econom ic accounts. In a report to the U nited States B ureau o f the Budget (re­ viewed at a hearing before the Joint E co­ nom ic C om m ittee of C ongress), this com ­ m ittee recom m ended fu rth er extensions and refinem ents of the existing national incom e and product accounts and of their sector subdivisions. A bove all, it proposed th a t the three other groups of econom ic accounts “w hich hitherto have had fairly independent lives” be integrated w ith the incom e and product accounts and built around them . These are the money-flow statem ents first developed by M orris Copeland, LeontiefFs input-output tables, and the national balance sheets elaborated by Goldsm ith. T he com ­ m ittee proposed th a t all four groups of tables be prepared by the same agency— the N ational Incom e U nit of the D epart­ m ent of C om m erce (to be appropriately enlarged)— and th a t identical definitions, classifications, and valuations be used in their preparation. It proposed th a t there be established a com pletely integrated system o f national econom ic accounts covering every m ajor phase o f the operations o f the economy. T he com m ittee’s report was pub­ lished in full in the Joint Econom ic C om ­ m ittee’s docum ent The National E conom ic A ccounts o f the U nited States, O ctober, 1957, pp. 101-302. Com m enting on this report before the Joint E conom ic Com m it­ tee, M artin R. G ainsbrugh m ade the fol­ lowing illum inating observation regarding the scientific and practical value of national econom ic accounts generally in the life of any m odern society (ibid., pp. 5 0 -5 1 ): The introduction and developm ent o f an inte­ grated system o f national accounts prom ises to rank in historic significance with som e o f the more heralded inventions o f recent decades in

the fields o f the physical sciences. This growing fam ily o f incom e and product statistics is with­ out question one o f the major contributions— if not the greatest— o f the econom ic fraternity thus far in the 20th Century.

3. The N ation’s E conom ic Budget O ne type o f social account has proved its usefulness in a special way. T his is the socalled “N atio n ’s E conom ic Budget.” Both its concept and its nam e w ere apparently conceived in the U nited States B ureau of the Budget during W orld W ar II by P ro ­ fessor G erh ard C olm and D r. G rover Ensley. Basically, it is m erely an adaptation o f the G ross N ational P ro d u ct aggregate w ith a presentation o f its m oney incom e elem ents on one side and its expenditure elem ents on the other, an d w ith its sub­ division in either case by econom ic sectors and the indication fo r each sector o f its surplus o r deficit. M oreover, it is an adapta­ tion in w hich the governm ent’s incom e and expenditures are especially em phasized. This statem ent is presented annually in the U nited States and in some other countries in conjunction w ith th e governm ent’s bud­ get, and is an im portant aid in the form u­ lation o f governm ent fiscal an d econom ic policy. I t is possible, w ith the aid o f this account, to estim ate w hat effect a given change in the governm ent’s receipts and ex­ penditures m ay have upon th e receipts and expenditures of consum ers o r business, and on the total national incom e and expendi­ tu re (see C hapter 16, Section 2, on fore­ casting and pro jectio n s). T his account is reproduced in T able 15-4. In the U nited States, this “econom ic bud­ get” was originally intended as a forecast o f the anticipated operations of the econom y during the forthcom ing year. H ow ever, be­ cause o f the m any uncertainties involved in fu tu re operations o f the econom y, the fore­ casting features of this econom ic budget were eventually dropped and it was lim ited to a m ere statem ent of the operations o f the econom y during the last three com pleted years and, perhaps, o f th e annual rate of th e last qu arter of th e cu rren t year. T he original designation o f it as the “N ation’s E conom ic Budget” was dropped in the 1950’s, giving way to a less am bitious title, “T he N atio n ’s Incom e, E xpenditure, and Saving” ; and instead o f being published in the P resident’s E conom ic R eport, it now appears in an A ppendix to th e R eport.

15.

215

AGGREGATES, ANALYSES, AND SOCIAL ACCOUNTS

Table 15-4. T he N a tio n ’s Income, Expenditure, and Saving (Alias Its Econom ic Budget), U nited States, 1953-55 (billion dollars) 1953

Economic G roup Consumers: Disposable personal income Personal consum ption expenditures Personal net saving ( + ) Business: Gross retained earnings Gross private dom estic investm ent Excess of investm ent

Re­ ceipts

1954

Excess of Receipts (+ ) E x­ or Expendi­ tu res pendi­ tures (- )

Re­ ceipts

254.8

250.4

Gross national p ro d u ct 364.5

252.4 18.3

3 6 .8

3 4 .4 5 1 .4

16.8 40.9

47.2 - 1 0 .5

- 1 7 .0 -

Excess of Receipts (+ ) E x­ or Expendi­ pendi­ tures tures (- )

269.2

19.8

(- )

Re­ ceipts

236.5

230.6

International: N e t foreign investm ent Excess of receipts ( + ) or investm ent (—) G overnm ent (federal, state, an d local): T ax and nontax receipts o r accruals Less: Transfers, interest, 17.5 an d subsidies (net) 78.5 N e t receipts T otal governm ent expenditures Less: Transfers, interest, and subsidies (net) Purchases of goods and services Surplus ( + ) or defi­ c it ( —) on income an d product account 1.3 S tatistical discrepancy

1955 *

Excess of Receipts (+ ) Ex­ o r E xpendi­ pendi­ tures tures (- )

2 .0

-1 8 .5

-.3

- .3 .3

8 9 .8

98.5

20.1

21.4 77.0

69 .7 102.0

9 7.0

97.3

17.5

20.1

21.4

84.5

77.0

75.9

6.0

-

1 .3 364.5

-.8

360.5

360.5

-7 .2

1.2

-.8

.3 387.4

387.4

D etail will n o t necessarily ad d to to ta ls because of rounding. s o u r c e : Based on th e n ational income an d p ro d u ct statistics of th e D e p artm e n t of Commerce (except as noted). a Prelim inary; fo u rth q u a rte r b y Council of Econom ic Advisers. note

:

so u rc e:

Economic Report o f the President, W ashington, Jan u ary , 1956, p.170.

O n the other hand, in some of the E u ro ­ pean countries (F rance, Belgium, th e N eth ­ erlands, Sweden, e tc .), the concept of th e “N atio n ’s Econom ic Budget” as a forecast has been retained. T he forecast is prepared annually and is designated as an econom ic budget.13 (F o r F ran ce’s B udget E conom ique de la N ation see C hap. 27, Section II-4.)

4. The Input-O utput A pproach: Leontieff T h e input-output analysis o f the opera­ tions of the national econom y, devised by P rofessor Leontieff in 1941, has also m ade a deep impression on national incom e ac­ counting and analysis. In several recent

national incom e estimates, e.g., those of N orw ay and D enm ark, the input-output approach has been accorded a prom inent place.14 Professor Leontieff describes it as follows: Considered from the point of view of the input-output scheme any national economy can be described as a system of mutually inter­ related industries or—if one prefers a more abstract term— interdependent economic activi­ ties. The interrelation actually consists in the more or less steady streams of goods and serv­ ices which directly or indirectly link all the sectors of the economy to each other. These flows can be observed and described in quan­ titative terms. The whole system [as applied to the Ameri-

216

THE IN CO M E OF NATIONS

Table 15-5. Sample of Leontieff Input-O utput Approach Q uantitative Input-O utput Relations in the U nited States, 1947 a (million dollars) Industry Purchasing

Industry Producing (1) (2) (3) (4) (5) (6)

1 Agriculture and fisheries 2 Food and kindred products 3 Nonferrous m etals 4 Iron and steel 5 M otors and generators 6 M otor vehicles

Total

Agriculture Food and N onand Kindred ferrous Fisheries Products M etals

( 1)

(2 )

10,856

15,048

2 ,3 7 8

4 ,9 1 0

6

(3)

2

Iron and Steel

(4)

M otors and M otor Generators Vehicles

(5)

T otal Output

(6)

(7)

11

2 ,5 9 9 33

44,2 6 3 3 324 3 ,9 8 2

366 118

176 196

37,6 3 6 6,387 12,338

4,4 0 1

1,095 14,265

14,263

769,248

317 111

44,2 6 3

3 7 ,6 3 6

6,3 8 7

12,338

1,095

a This table reproduces a portion of the 50-industry input-output table for 1947, which is included in full in “T he Interindustry Relations Stu dy for 1947,” Review o f Econom ics and S tatistics, M ay 1952. * Less than $500,000. so u rc e :

Leontieff et al., Studies in the Structure of the A m erican Econom y, N ew York, 1953,

can economy] has been subdivided into 50 sec­ tors comprising agriculture, various extractive and manufacturing industries, electric public utilities, three kinds of transportation, trade and other types of service industries. Foreign countries are treated as a separate industry. Households and government . . . constitute the two large non-industrial sectors of the system.15 T he approach is illustrated by the fol­ lowing quotation from an all-em bracing in­ put-output table: As the author fu rth er explains, “T he headings at the top and left-hand side of the table indicate th a t each row and each corresponding colum n (except the very last one) is identified w ith a particular sector of the economy. T he figures entered along any one horizontal row show how the an­ nual 1947 output of the particular industry has been distributed am ong all the other industries, the households, and the govern­ m ent. This total output itself is shown at the end of the row, i.e., in the last righthand column. If exam ined columnwise, the same figures show the 1947 input structure of the econom y; each vertical colum n con­

p.

9.

tains the am ounts of the products of all the different branches of the econom y w hich have been absorbed by its ow n particular sector.” A lthough the figures in the table are show n in dollars, “they m ight as well have been given in physical units appropriate for the description o f the outputs of individual sectors o f the econom y— tons o f coal, bushels of w heat, ton miles of tran sp o rta­ tion, m an-hours of w ork, and so o n .” Only the total inputs “do n o t lend themselves to this kind of physical interpretation,” since “tons of coal, yards of cloth, and m an-hours o f labor cannot be added fo r any useful purpose.” 16 T h e input-output approach is n o t neces­ sarily connected w ith national incom e data. A lthough it can be connected w ith it, it represents to a large extent an independent analysis of the operations of th e national econom y. T his analysis is m ade throughout in term s of gross figures th a t are broken dow n into elements different from those involved in national incom e.

16. INTERTEMPORAL AND INTERNATIONAL COMPARISONS In this chapter we shall consider four prob­ lem s: (1 ) th a t o f com paring the national incom e of a country over tim e; (2 ) th a t of m aking international com parisons o f n a­ tional incomes; (3 ) th a t of forecasting a fu tu re national incom e either u nder given hypothetical conditions o r u nder certain ex­ pected conditions; and (4 ) th a t of m easur­ ing the effects of differences in th e distribu­ tion o f national incom e on the econom ic w elfare provided by a given size of national incom e. We shall also com pare the national incom es of various countries and the changes occurring in them . T he first two problem s have perplexed economists and statisticians fo r at least a century, but have received particular at­ tention in m odern tim es. T he universaliza­ tion of national incom e estimates, th e ac­ cum ulation of extensive historical national incom e data, and the establishm ent of agen­ cies fo r international political and econom ic co-operation have m ade such international com parisons especially im portant. T he chief difficulty present in historical com parisons is th a t the national incom es fo r different years, being com puted in current prices, are affected by changes in the price level, i.e., in the value of money. T herefore, they m ay not accurately m easure the changes in the national output and econom ic w elfare. T he difficulty in international com parisons is even greater, being rooted in the fact that the national incom es of different countries are evaluated in different currencies and are affected by differences in products, prices, cultural and clim atic conditions, and other factors responsible fo r different national re­ quirem ents and standards of econom ic w el­ fare. But while there are m ajor differences be­ tw een the problem s of intertem poral and intercountry com parisons, there are also cer­ tain points of sim ilarity between them . C om ­ parisons of the national incom es o f a given N otes begin on p. 534.

country during widely separated periods deal essentially w ith different societies. The differences in this case m ay, in some re­ spects, be as great as the differences between tw o nations. T he th ird problem — th a t of discovering the techniques to be follow ed in forecasting a futu re national incom e or calculating its probable fu tu re structure and size under given hypothetical conditions— is almost w holly new. T he practice of forecasting or projecting national incom e has itself come into being only recently. It has developed largely since W orld W ar II as a result of the expanding role of governm ent as a fac­ to r prom oting stability and econom ic growth in the economy. T he fo u rth problem — th at of accounting for the effects of differences in the distribu­ tion of national incom e on the economic w elfare provided by a given size of national incom e— is largely theoretical and has as yet been b ut little explored. T he fifth subject, involving actual com ­ parisons o f the national incom es of various countries and of th e changes occurring in them , has been relatively little explored in the past. Paucity of com parable and reliable statistics has stood in the w ay of fuller inves­ tigation. Yet, if properly executed, such com parisons can shed m uch light on the forces responsible fo r the econom ic prog­ ress of some nations and the decline or stag­ nation of others, and can either confirm or refute some of the purely theoretical or hypothetical generalizations th at have been advanced in this field. A n attem pt at such m eaningful com parisons will be m ade in Section 5 of this chapter.

1. Com parisons O ver Time N ational incom e, regardless of the con­ cept it is based on, is always calculated in the first instance at cu rren t prices. It is so calculated because it is built up from

218

THE IN CO M E OF NATIONS

current transactions, wage paym ents, p u r­ chases of goods, etc., th a t are related to current prices, and because all other cur­ rent phenom ena, such as governm ent ex­ penditures, etc., are expressed in such prices. But national incom e calculated at cur­ rent prices is not a good m easure o f the national product over tim e because the prices of different years m ay not be the same. T he figures fo r different years m ay just as m uch reflect changes in the price level and price structure from one period to another as they m ay reflect changes in the quantity of goods and services produced over that tim e. If the object of com paring the national incom e figures over tim e is to ascertain changes in the quantity (an d as fa r as possible also in the quality) of goods and services, and, hence, in the econom ic welfare of the society— and this is generally the only purpose of such com parisons— then it is necessary to elim inate the effects of price changes from these figures. In other words, it is necessary to recom pute the na­ tional incom e, and its various com ponents, in constant prices. A s m ost national income estimates today consist of a series of annual figures intended to shed a light on the changes in the size and com position of the national product over the period, such re­ com putations in constant prices are rapidly becom ing universal. T here are three m ain m ethods o f tran s­ lating national income estimates from cu r­ rent into constant prices, o r as it is com ­ monly called, of “deflating” them . T hey are: (a ) deflation of th e annual totals by a cost of living index, consisting o f w eighted re­ tail prices, o r by an index o f weighted wholesale prices— in either case expressed as a relation to the prices o f a given year taken as 100; ( b ) separate deflation o f dif­ ferent parts of the national incom e by dif­ ferent indexes specially related to them ; and (c) direct com putation o f the annual phys­ ical output of goods and services and the application to them o f the prices of the base year. W e shall discuss briefly each one of these three m ethods.1

a. Deflation of Total Current Value by an Over-all Cost of Living Index T he first m ethod, involving deflation of the national incom e totals by a single cost o f living index, is sim pler than either o f the other two as it requires but a single opera­ tion, and is, therefore, m ost com m only em ­

ployed. B ut it is also less accurate than either o f the other two m ethods. A cost of living index is an appropriate deflator for th a t p a rt of the national incom e th a t is used fo r private consum ption. B ut it is n o t an appropriate deflator fo r th e p a rt used to finance governm ent services, interest on governm ent debt, investm ent outlays, and the net incom e from , o r transactions with, the rest of the w orld. T he prices involved in these transactions are n o t included in the cost o f living indexes. W holesale price indexes are sometim es used in place of a cost of living index, because wholesale prices are m ore readily ascertained and such indexes are, therefore, m ore easily constructed. B ut they suffer from the same defect. T he m ethod of deflation by a simple index is being gradually abandoned in the m ore carefully prepared m odern estimates.

b. Deflation of Current Values of Differ­ ent Items by Different Indexes T he second m ethod, involving the appli­ cation of different deflators to different parts o f th e national incom e, is m ore difficult, b u t gives m ore reliable results. It was used, fo r exam ple, in the prew ar an d w artim e estim ates of the gross national p ro d u ct for D enm ark. U sing the prices of 1935 as a base, a cost of living index was used to de­ flate private consum ption outlays an d gov­ ernm ent purchases of goods and services; a building costs index fo r th e deflation of outlays fo r new housing and other struc­ tures; a wholesale price equipm ent index fo r the deflation o f outlays fo r m achinery; a wholesale price index of im ports fo r the deflation of the excess of im ports over ex­ ports; and a wholesale price index o f exports fo r th e deflation o f the excess o f exports over im ports. N et receipts from freight were deflated by an index of freight rates. D u r­ ing the postw ar period th e base y ear was changed to 1947 and an even m ore refined com bination o f special indexes, differenti­ ated by industrial divisions, was used. W here quantities o f in p u t and ou tp u t were know n, these w ere evaluated directly at 1947 prices. In the case of financial inter­ mediaries, governm ent services, and a few service trades, the contribution to the gross national p ro d u ct was deflated by a wage index. In the rem aining divisions, th e value added w as deflated directly b y a com bina­ tion of price indexes an d in som e cases quantity indexes.2 T his m ethod of deflation was also used

1 6 . IN TERTEM PO RAL AND INTERNATIONAL COMPARISONS

in a fairly refined form in the C anadian estim ates (C hapter 21, Section II-2 ).

c. Evaluation of Current Physical Out­ puts in Base Year Prices T he third m ethod, involving the direct evaluation o f physical outputs fo r each year in term s o f base year prices, is m ost com ­ plicated but probably also m ost reliable. It was used by M atolcsy and V arga in their national incom e estimates fo r H ungary fo r the years 1 9 2 4 /2 5 -1 9 3 7 /3 8 .3 T he average prices fo r the three-year period 1 9 2 4 /2 5 1926/27 w ere used as a base, and either the actual physical o utput o f each division of econom ic activity was com puted fo r each year, o r an index of physical outp u t was developed in relation to the know n figure of the output fo r th e base period. In th e case of agriculture, the annual physical out­ p u t was directly estim ated, the changes in it being due m ainly to variations of w eather. In com puting the volum e of production in m ining and smelting, the quantity figures of the ou tput of coal and pig iron only were used; other m ining products w ere ignored, as they were com paratively unim portant. T he quantity figures o f the tw o item s were weighted by their relative prices in the base period. Thus, the price o f pig iron being then five tim es higher th an the price of coal, the quantities of pig iron w ere w eighted with 5. T he quantities o f different m anufacturing goods w ere w eighted by the relative values-added o f these goods during the base period. These weights w ere con­ sidered as being constant fo r th e period. T his assum ption was obviously inaccurate, b ut it had to be m ade as otherw ise the com ­ putations becam e m uch too com plicated. In com puting the volum e of output in handi­ crafts and dom estic industry on w hich no direct current figures w ere available, it was assumed that this volum e changed in the same proportion as the com bined volum e of the production of agriculture, m ining and smelting, and m anufacturing. This index of growth of output was applied to the values of the base year. In the case o f transport, a volume index was developed from the tonkilom eters of all classes of freight shipped by the H ungarian state and private rail­ ways. In the case of services of com m erce, it was assumed that the changes in their volum e were directly related to the changes in the quantity of goods passing through trade. In other words, it was assum ed that the structure of the exchanges rem ained

219

unchanged during this period. T he priceincreasing effect of indirect taxes (10.4 per cent during the base year) were also as­ sum ed to have rem ained constant. These examples are sufficient to indicate the nature of the problem s involved in this m ethod of com putation of the national income at con­ stant prices. T his th ird m ethod of deflation, involving the com putation of physical outputs in base year prices, has been used in Soviet Russia fo r years. U nfortunately, little inform ation about it has been published there (see C hap­ te r 25, Section 7 ) . N o m atter how well th e deflation o f a historical series of national income estimates m ay be executed, it m ay produce reliable results fo r only a lim ited period. T he longer the series covered by a given deflator, the less reliable are th e deflation results. All such deflations to the prices of a base year are founded on the assum ption th a t the general conditions during the other years in the series are n ot m uch different from those th a t prevailed during the base year. This assum ption can hold true only if the base year is n o t too fa r rem oved from the other years. If it is, the likelihood is very great th a t such extensive technological changes have taken place during the interval, affect­ ing the natu re and quality of the products and the costs of their production, as to m ake the prices o f the base year no longer applicable to th e products and transactions o f the other years. If the base year is the initial o r m iddle year of the historical series, som e o f the products of th a t year em bodied in the index will no longer be in use during the latest years in the series o r will be used to a different degree. N ew products th at w ere n ot included in the original index will have appeared, and the weights of various curren t products will be com pletely dif­ ferent from those assigned to them in the index. T he quality of existing products will have changed and the costs o f their produc­ tion will accordingly be higher or low er than in the base year, m aking base year prices inapplicable. Because of this, it is necessary to change the base year every decade or so, substituting new indexes o r base year prices fo r th e old. D uring the 1920’s, it was custom ary in m ost countries to use 1913 prices as de­ flators, th e last year before W orld W ar I. D uring the depression o f the 1930’s, the prices o f 1924-25 or 1928-29 were so used; during W orld W ar II and the years im m edi­

220

THE IN CO M E O F NATIO NS

ately following, indexes were reconstructed to the prices of 1935-36 o r 1938-39 as a base; and during the 1950’s, they w ere re ­ built once m ore on the 1949-50 basis. T he m ore rapid the rate of technological change in an economy, the m ore frequently m ust old price indexes be replaced by new ones. This rule was violated in Soviet Russia, w here as late as 1956 national income esti­ m ates were calculated in 1926-27 prices— prices w hich are obviously com pletely out o f line w ith the new industrial structure of the country. It is not surprising, therefore, that m ost economists in capitalist countries have refused to accept the Soviet estimates of the past tw enty-odd years expressed in these prices. T heir refusal to accept them is especially justified w hen the fact is consid­ ered th a t the m anner in w hich th a t index was constructed has never been adequately explained and its relevance to later condi­ tions never dem onstrated; and that no cor­ responding figures in current prices w ith w hich the others could be com pared have been published in Soviet Russia since 1933 (see C hapter 16, Section 5a, and C hapter 25, Sections le , 7, and 11).

2. Forecasts, Projections, and Plans T he developm ent of historical series of national incom e expressed in constant prices has helped to reveal the nature and rate of changes occurring in the econom y u nder varying conditions. It has helped particu­ larly to bring to light the interrelationships of the different com ponents of the national incom e and the effects w hich, under varying conditions, a given change in one elem ent has upon the others and upon the national incom e total. W ith this inform ation in hand, it has becom e possible fo r economists and statisticians to forecast w hat the national incom e is likely to be in a given future year o r w hat it m ay then be u nder certain hypo­ thetical conditions. These conditions m ay relate to m ore or less uncontrolled interna­ tional o r domestic situations that m ay de­ velop, or to the types of policy that govern­ m ent, business, o r farm and labor groups m ay consciously choose to pursue during the given tim e. Such forecasts o r projections m ay be of the short-run variety, covering the succeeding year, or of the long-run type, covering the succeeding five or ten years. Such short- and long-range projections or

forecasts have becom e w idespread during th e decade o r so following W orld W ar II. T hey represent a logical extension of n a­ tional incom e estimates. By applying them to a b roader purpose, a new vitality and im portance have been given to national in­ com e estim ates. T hey are usually m ade in term s of the gross national p ro d u ct rath er th an n et national incom e, because o f the m ore inclusive n atu re o f th a t aggregate.

a. Forecasts: Long Range and Short Range; and Planned Goals F orecasting national incom e is an ex­ ceedingly hazardous undertaking especially if done on a long-range basis. N o m atter how elaborate the forecaster’s calculations, the accuracy of his results will depend p ri­ m arily on th e correctness of his guess re­ garding the general econom ic and political conditions th a t will prevail in the country, and, in fact, in the w orld as a w hole; and on the policies th at th e governm ent and various private groups will pursue in the face of these conditions during the period. T hese are m atters th at cannot be determ ined scientifically. T hey depend on a multiplicity of circum stances, m any of w hich are purely accidental an d can n o t be predicted. A ny chance developm ent m ay throw off a guess, no m atter how well reasoned. M ost longrange forecasts m ade by econom ists in the U nited States tow ards th e end o f W orld W ar II proved w rong. T hey anticipated a sharp decline in m ilitary expenditures, a serious lag in th e expansion o f civilian p ro ­ duction, severe unem ploym ent, a sharp de­ flation, and a low er national incom e. T he actual situation turned o ut to be exactly the reverse, an d m any equally pessimistic fo re­ casts m ade later on likewise proved wrong. T h e failure o f m ost of these long-range forecasts p u t a dam per on fu rth e r efforts along these lines.4 B ut w hile the attem pt to m ake long-range forecasts of national incom e in definitive term s is a hazardous and even futile u n ­ dertaking, attem pts to p o rtray its possible fu tu re developm ent u n d er certain clearly stated hypothetical conditions are entirely feasible and m ay be exceedingly useful. T hey m ay serve as guides to governm ent, business, and labor in charting th eir own policies fo r the future. Such projections in­ volve the construction of a n um ber of m odels o f th e gross national product, each based on different assum ptions regarding the

1 6 . IN TERTEM PO RAL AND INTERNATIONAL COMPARISONS

political and econom ic conditions th a t m ight prevail and the policies th a t m ight be fol­ lowed during the period in question; and each presenting a gross national p ro d u ct of different size and different com position. M ost feasible of all are projections of national incom e fo r the year o r tw o im m e­ diately following. These can be m ade defi­ nite, because m ajor changes in the national econom y generally take some tim e to de­ velop. Signs of their being u nder w ay can generally be discovered and the pace of their developm ent m easured before they take full shape. Estim ates can, therefore, be m ade fo r short periods ahead on the basis of existing trends. Such short-range forecasts are necessary fo r the preparation of budgetary program s and especially of revenue estimates fo r the succeeding year o r two. In planned econom ies of the m ixed as well as com pletely socialized type, and in som e partly planned capitalistic econo­ mies, long-range projections o f national in­ come play a key role in the planning process (see for India, p. 439; C hina, pp. 496, 544; Japan, p. 544).

b. Hypothetical Projections M ost valuable are th e long-range projec­ tions of a purely hypothetical nature, w hich seek to define the m axim um gross national product the econom y m ay be able to achieve over a period o f a decade or so u nder speci­ fied conditions if all the strategic groups in the econom y pursue certain stated courses o f action. Such projections are invaluable because they set definite goals fo r the econ­ omy and for its constituent groups. If these projections are com petently p repared and if their goals are accepted by all concerned, they m ay bring about concerted action th a t will insure their realization. If during the interval errors are discovered in the original calculations, o r th e conditions tu rn out to be different from those first assumed, the projections can be m odified in tim e so th a t they would continue to be useful. Policies m ay be reshaped in the light o f new pros­ pects. W e have seen in the historical section th a t the first such hypothetical projections w ere m ade by Q uesnay and his fellow physiocrats (C hapter 5, Sections lb and c ) . In the nineteenth century Lowe m ade them fo r England and F rance (C hapter 7, Sec­ tion 1-3). B ut it was the K eynesian influence th a t brought such projections into vogue in

221

the present century (C hapter 10, Section 3 a ). In developing projections of the maxi­ m um potential growth of the national in­ com e over a given fu tu re period, it is neces­ sary first of all to estim ate the rate of future increases in the general population and in th e lab o r force. Such estimates are usually based on past rates of grow th and on ap­ praisals of the probable effects upon them o f new relevant factors. Estim ates m ust also be m ade of likely changes in the num ber of w orking hours, in the productivity of the w orkers, and in their hourly earnings, so th at the total fu tu re earnings of the labor force m ay be calculated. N ext, th e annual profits and oth er incom es from capital over the years in question m ust be estim ated partly on the basis of the past relationships these incom es bore to the earnings o f labor, and p artly on th eir anticipated relation­ ships, determ ined by expected changes in the ratios of capital to labor employed, the rates of retu rn on capital, and other factors. Estim ates m ust also be m ade of probable future allowances fo r depreciation and of the probable am ount o f indirect taxes th a t governm ent will raise during those future years. T h e addition of all these am ounts will give figures fo r the gross national product fo r those fu tu re years u n d er the stated con­ ditions. Finally, this estim ate m ust be tested in term s of the possible utilization o f the projected national product in consum ption and investm ent. Tw o m ain types of such long-range p ro ­ jections have been developed in recent tim es. O ne involves the construction of dif­ feren t m odels of th e gross national income and expenditure, built upon different prem ­ ises as to the n atu re of the fiscal policy gov­ ernm ent will pursue, and th e m anner in w hich business and consum ers will behave. T he other type of long-range projection calls fo r the construction of a single p attern of the gross national incom e o r product, based on a single set of assumptions. In both ap­ proaches, the n atio n ’s econom ic budget m ay be used as a tool. In several countries, e.g., in F rance, India, and th e U nited States, such long-range projections have been de­ veloped after W orld W ar II.

c. Two Recent Projections in the United States W e shall briefly describe here only two such projections, both taken from the ex-

222

THE IN CO M E O F NATIO NS

Table 16-1. Projection of the N ation’s Econom ic Budget of the U nited States for M aximum Economic Growth, from 1953 to 1965, Prepared by the Staff of the Joint Congressional Com m ittee on the Econom ic Report (billion 1953 dollars “)

Incom es From and Expenditures for Gross National Product

Actual Calendar 1953b

Estim ated Calendar 1965

Personal Income, total disposable

250 .1

380”

Expenditures: Durable goods Nondurable goods Services

2 9 .7 118 .9 8 1 .4

50 185 122

230 .1

357

+ 2 0 .0 d

+23"

T otal expenditures Savings ( + ) Business Incomes: Corporate undistributed profits Capital consumption allowances Inventory valuation adjustm ent T otal incomes Expenditures: Residential nonfarm construction Plant and equipm ent Change in business inventories N et foreign investm ent T otal expenditures Dissavings ( —) G overn m en tB Incomes: Personal tax and nontax paym ents Business tax and nontax liabilities Contributions for social insurance Less nongross national product paym ents T otal incomes Expenditures: Major national security Public construction, civilian Schools H ighways Other public construction All other T otal expenditures

8 .9 2 7 .2 —1 .0

12* 48 0

3 5 .1

60

1 1 .9 3 8 .0 1 .5 —1 .9

16 60 3 2

4 9 .5

81

—1 4 .4

—21

3 6 .0 5 1 .1 8 .8

43 h 62 h 16 1

1 7 .4

26*

7 8 .5

95

5 2 .0 10.1 1 .7 3 .2

40 17 4 6

5 .2

• 2 3 .1

7 40

8 5 .2

97

Savings ( + ) or dissavings ( —)

—6 .6

—2 J

Statistical discrepancy ( + ) or ( —)

+ 1 .0

0

3 6 4 .9

535

Grand T otal

1 6 . IN TER T EM PO R A L AND INTERNATIONAL COMPARISONS

223

C ontinuation of T able 16-1 a R ates of com pensation for factors of production including general governm ent em ployees are assumed to change in line w ith changes in productivity. b D etail m ay not add to totals due to rounding. 0 Consistent w ith an assum ed personal income of $423 billion. d Personal savings were 8 per cent of disposable personal income. 6 Personal savings assum ed to be about 6 per cent of disposable personal income. f Consistent w ith an assum ed corporate profits of $56 billion, as follows: (in billions) Corporate profits and inventory valuation adjustm ent Inventory valuation adjustm ent Corporate profits before taxes Corporate profits tax liability Corporate profits after tax D ividends Undistributed profits

1953 $38 .5 1 .0 3 9 .4 21.1 18.3 9 .4 8 .9

1965 $56 0 56 26 30 18 12

g Includes federal, state, and local governm ent. h T ax estim ates assum e reductions in federal tax rates which would reduce total federal, state, and local revenues in 1965 by 15 to 20 per cent below the yield under present rates law a t 1965 output levels. 1 T he increases shown for these 2 item s largely reflect the projected increases in both receipts and expenditures of social insurance funds and governm ent em ployee pension funds, both federal governm ent and state and local governm ent. T he estim ates reflect so far as possible the expected influence of the m ost recent revisions in social insurance programs. » A balanced federal budget bu t a com bined state and local governm ent deficit on a goods and services basis. SOURCE: Potential Economic Growth of the United States D uring the Next Decade, M aterials Prepared for the Joint C om m ittee on the Econom ic Report b y the C om m ittee Staff, 83d Congress, 2d session, U.S. G overnment Printing Office, W ashington, D . C., 1954, p. 20.

perience of the United States—one prepared in 1952 by Gerhard Colm for 1960 and the other prepared in 1954 for the year 1965 by the Staff of the Congressional Joint Com­ mittee on the Economic Report. Gerhard Colm used the first approach.5 Starting with the gross national product of 1951, which amounted to $329 billions, he estimated that in the absence of a major war and at full employment, the American economy should be able to achieve by 1960 a gross national product, at 1951 prices, of $425 billions. He constructed six models of achievement of this size of gross national product, each based on different assumptions as to the nature of government fiscal policy and of private economic policy. Out of these he next developed a single “adjusted” model, which he pictured as the most likely possibility. The six preliminary models were based on the following assumptions: Model A High consumption—high taxes Model B High consumption—moderate taxes Model C High domestic investment— moderate taxes Model D High domestic and foreign in­ vestment—moderate taxes Model E Very low taxes

Model F

High government purchases— moderate taxes

The staff of the Congressional Joint Com­ mittee on the Economic Report prepared a single pattern projection of a possible maxi­ mum growth of the gross national product over the succeeding decade or so.6 It esti­ mated that in the absence of a war or another major change in the world and domestic situation, the gross national product, which amounted to $365 billions in 1953, might increase to $535 billions, in 1953 prices, by 1965, or by nearly 50 per cent. The projec­ tion did not necessarily assume maximum economic growth each and every year be­ tween 1953 and 1965. It assumed, however, that “any slowing up in expansion of pro­ ductive capacity in one year will be made up by a later speeding up and vice versa.” The population was estimated to increase from 160 millions to 190 millions and the labor force from 67 millions to 79 millions, of which 3 millions were assumed to be in the armed forces, 7.5 millions in the civilian government, 3 millions unemployed, 5.5 millions in agriculture, and 60 millions in private nonagricultural employment. The average annual hours per man in private

224

THE IN CO M E OF NATIONS

nonagricultural employment were assumed to decrease 0.8 per cent per year, so that by 1965 the working week will be reduced by approximately 4 hours. The output per man-hour was assumed to increase 3 per cent per year in agriculture and 2.5 per cent in nonagricultural private employment, which was somewhat below the rate of in­ crease during the immediately preceding years. Defense expenditures were expected to drop from $52 billions in 1953 to $40 billions, but public construction outlays to rise from $10.1 billions to $17 billions. Tax rates were expected to be lowered so that by 1965 they would produce 15 to 20 per cent less revenue than they would have pro­ duced under the 1954 rates. Private invest­ ment in plant and equipment was assumed to expand from an annual rate of $38 bil­ lions to one of $60 billions, and disposable income to rise from $250 billions to $380 billions, or from $5,321 per household to $6,785. The rate of personal saving was estimated to drop from 8 per cent in 1953 to 6 per cent by 1965. The full details of the projection are given in Table 16-1. The authors of this projection did not claim that a gross national product of the size indicated will necessarily be achieved. They merely contended that government, business, and the general consumer might, in the absence of a war, so arrange their actions as to achieve such a gross national product. They wrote: One purpose o f such analyses is to set forth the nature and magnitude o f the adjustments that appear needed to achieve certain objectives and to suggest the im plications for the econom y if the adjustments are not forthcom ing. The projector’s task does not necessarily include a forecast as to whether these adjustments w ill or w ill not be made. Another purpose o f econom ic projections is to provide a basis for an inter­ nally consistent econom ic program aim ed at the achievem ent o f the N ation ’s major e co ­ nom ic goals. Obviously, anyone presenting an unconditional prediction o f future econom ic developm ents, including expected changes in public and private policies, m ust expect that any variation from his predictions w ill cor­ rectly be used as a measure o f his deficiencies as a forecaster. E conom ic projecting into the future, or if one insists “forecasting,” is here to stay as long as individuals, private business, and dem ocratic governments are free to make their ow n deci­ sions. Only in an authoritarian state can w e be relieved o f this necessity; there, projections becom e com m ands.7

3. Problem s of International Com parisons Comparisons of the national incomes of various countries at a given time as meas­ ures of the relative productivity and eco­ nomic welfare of their people are even more difficult than intertemporal compari­ sons of the national income of a given country.

a. The Misleading Nature of Conversions Into a Single Currency on the Basis of Foreign Exchange Rates The first and most immediate difficulty is that the national incomes of different coun­ tries are expressed in different currencies having different purchasing powers both in international trade and at home. The tra­ ditional method of converting magnitudes expressed in different currencies into ones expressed in the currency of some one country (which may be either the home­ land of the person making the comparisons or the country that dominates international trade and exchanges at the time, namely, England years ago and the United States now) never gave satisfactory results in the past and is even less satisfactory today. First of all, the foreign exchange rates that may be applied for such conversions usually reflect the price relations only of such prod­ ucts as enter foreign trade. Secondly, these price relationships themselves are distorted by import and export duties and quotas, transportation costs, movements of capital funds, foreign exchange controls, and many other extraneous factors. Foreign exchange rates cannot be used to measure the relative prices of goods and services that circulate wholly or almost wholly within the given countries. With regard to them, the coun­ try’s currency may have a wholly different purchasing power from that which it may have over foreign goods. In some cases this purchasing power may be much greater, in others, much smaller. For example, the cur­ rency of a country whose foreign exchange rate is very low would purchase much more foodstuffs and other goods at home than it would abroad. In a country with a high foreign exchange rate the situation may be exactly the reverse, the currency having less purchasing power at home than its foreign exchange equivalent would have abroad. The second difficulty in such international comparisons is that the products repre­

1 6 . IN TERTEM PO RAL AND INTERNATIONAL COMPARISONS

sented by the national incom es are often com pletely different and satisfy vastly dif­ ferent needs determ ined by different cli­ m atic conditions and institutional factors obtaining in the countries involved. A s pointed out by K uznets, the per capita incom e of an individual Chinese o r H indu sustains him and the m em bers of his fam ily in his country, w hereas the sam e p er capita sum expressed in A m erican dollars w ould not provide the same Chinese o r H indu in the U nited States w ith enough food, clothing, and shelter to keep him and his fam ily alive. T h e p er capita incom e of the A m eri­ can expressed in dollars provides him w ith a high standard of living in the U nited States, b u t converted into Chinese or Indian currency it will provide him w ith a wholly different standard of living in these other countries. In some respects this standard will be higher, as he will be able to em ploy m ore domestic help; but in others it will be lower, as he will not be able to obtain, for any am ount o f m oney, certain com forts th at he obtains fo r little m oney in the U nited States.

225

price weights of the other— and of those of the second in the price weights of the first, and of averaging the two results. This m ethod, know n as one of “binary com pari­ sons,” has been m ost fully developed by M ilton G ilbert and Irving B. Kravis in their joint study prepared fo r the O E EC in 1954. “This m ethod,” say the authors, “involves securing appropriate quantities, prices, and values fo r as detailed a breakdow n of the gross national p ro d u ct as is possible fo r any tw o countries to be com pared, and then weighting the d ata fo r each country, w ith the weights of the first one and then the other of them . T his produces two indices of the real p roduct relationship and of th e in­ ternal purchasing pow er of the currencies.” 8 G ilbert and K ravis applied this m ethod in a com parison o f the gross national in­ com e o f the U nited States w ith the gross national incom es o f the following fo u r E u ­ ropean countries: the U nited K ingdom , France, G erm any, and Italy. T heir results, as com pared w ith those obtained by the traditional conversion into U.S. currency, w ere as follows:

Table 16-2. Comparison of Per Capita Gross National Products of Four European Countries and the U nited States, 1950, by Different M ethods Conversion by Exchange R ates

In U.S. Prices

1,810 677 634 467 292

1,810 1,136 968 785 548

In European R elative Prices

In Geometric Average

Per Capita Dollars United States United Kingdom France Germany Italy

Indexes: U .S. United States United Kingdom France Germany Italy

1,810 954 764 604 394

1,810 1,041 860 689 465

= 100

100

100

100

100

37 35 26 16

63 53 43 30

53 42 33

58 48 38 26

22

s o u r c e : M. Gilbert and I. Kravis, A n International Comparison o f N ational Products and the P u r ­ chasing Power o f Currencies, O EEC, Paris, 1954, pp. 22-23.

b. Adjusted Comparisons by GilbertKravis and by Colin Clark In recent years, attem pts have been m ade to correct the distortions in the national in­ com e figures of two o r m ore countries re ­ sulting from their being expressed in the currency of a given country com puted at the existing foreign exchange rates. T he correction involves the expression of the quantity products of one country in the

These com putations showed th at (a) the traditional m ethod of converting the gross national products of the E uropean countries into dollars by m eans of exchange rates “substantially underestim ates the position of all the E uropean countries com pared with the U nited States, and gives distorted rela­ tionships am ong the E uropean countries themselves” (understating particularly the relative position of the U nited K ingdom );

226

THE INCO M E OF NATIONS

( b ) the com parison of the products in U.S. prices overstates the relative position of the E uropean products; (c) the com parison in E uropean prices does the reverse; and ( d ) the averaging m ethod strikes a balance be­ tween the other two. T he authors concluded th a t in general the quantities of goods and services purchased in two countries will differ because of (a) differences in tastes and needs, ( b ) differ­ ences in incom e levels, and (c) differences in relative prices. A m ong the differences in tastes and needs of the five countries affect­ ing the com parisons, they m entioned “the F rench and Italian preference fo r wine, the U nited K ingdom preference fo r tea, or the smaller need fo r household fuels in Italy.” T hey also observed th a t the pattern o f a country’s expenditures “can be signifi­ cantly influenced by direct controls— includ­ ing the controls over total spending exer­ cised by the G overnm ent budget,” b u t con­ cluded th a t “so far as private expenditures are concerned, it m ay be expected th a t for the five countries being studied here, rela­ tive prices and incom e levels will be m ore im portant influences.” T hey noted th a t “as to differences in relative prices themselves, they are due in the m ain to . . . (a ) dif­ ferences in availability of natural resources; ( b ) differences in availability of m anpow er relative to capital; (c ) differences in rela­ tive efficiency in producing various sorts of goods and services; ( d ) differences in de­ m and schedules (tastes or n e e d s); w hich in tu rn affect the scale and efficiency of p ro ­ duction.” T hey adm itted th a t “the im por­ tance of these factors cannot be isolated in any systematic w ay,” but expressed the be­ lief that “keeping them [these factors] in m ind will help the reader to understand m any of the differences” brought to light by their binary m ethod of com paring the national incom es o f the countries in ques­ tion. A lthough recognizing the com plicating factors in m aking international com pari­ sons, G ilbert and K ravis have not given them due im portance. T hey have been taken to task for this failure in a review of their book by J. L. N icholson, w ho w rote that the authors seem to be saying: “we m ust ignore all these com plications, . . . o r we m ay find that we are unable to m ake any com parisons at all.” 9 J. van Y zeren, in an article published in D ecem ber, 1954 (T he N etherlands Central Bureau of Statistics, Statistical Studies, N o.

7 ) , carried the G ilbert-K ravis approach m uch fu rth er, offering three alternative m ethods fo r com paring the purchasing pow er of currencies. D r. R. C. G eary of the N ational A ccounts D ivision of the U nited N ations, in com m enting on this ar­ ticle in 1957 in an unpublished paper, offered his own m ultiple form ula fo r aver­ aging th e exchange rates of a great m any countries fo r purposes of com parison of th eir national incomes. T h e binary com parison approach is prom ­ ising, b u t m uch m ore research will have to be done in its application and in the analysis of its results before it can be accepted as offering a definitive solution to the m ethodo­ logical problem of international com pari­ sons o f national incomes.

c. Other Approaches A n o th er approach th a t has been sug­ gested by ano th er group of economists, in­ cluding Julius W yler, w ould involve the m easurem ent of the national incom es of the industrially less advanced countries in term s of th e national incom e of th e m ore ad ­ vanced country during an earlier stage of its developm ent. This proposal is based on the assum ption th at the nations share their experiences in the long ru n and th a t the technologically less advanced countries may eventually reach the sam e stages of techno­ logical advancem ent th a t the others m an­ aged to attain earlier. Sim on K uznets ex­ pressed ano th er variant o f this approach as follows: We could, with considerable labor, evaluate the products of a primitive, underdeveloped society in the scale of a more developed one: a well-trained expert could find, in an advanced country, products analogous to those in the less developed, and apply to the latter the prices of the former. This is possible because the scale of economic valuation in the developed coun­ try provides a basis for appraising an exceed­ ingly wide variety of goods. But the converse cannot be done. It would be impossible to se­ cure from the Amazon Indians any realistic appraisal of the value to them of radios, motor cars, etc., relative to that of fish or corn. By the nature of the case, a majority of the prod­ ucts of more developed economies cannot be weighed in the value system of the less devel­ oped and particularly primitive societies.10 Expressing the national incom e of a less developed country in term s o f the products o f a m ore developed one m ay prove very useful, b ut th e techniques fo r doing so have n o t yet been developed. This m ethod

1 6 . IN TERTEM PO RAL AND INTERNATIONAL COMPARISONS

has its obvious difficulties and lim itations. C olin C lark in his C onditions o f E co­ n om ic Progress (3 d ed., 1957) com pared the incomes of econom ically advanced na­ tions on the basis of an International U nit, representing the quantity of goods and serv­ ices exchangeable in the U.S. fo r $1, and o f underdeveloped countries in term s o f an O riental U nit, representing the quantity ex­ changeable in India fo r one rupee. Based on these standards, he obtained the follow ­ ing p er capita incom es fo r 1950 or there­ abouts: I.U .’s— U .S., 1,053; Switzerland, 903; C anada, 847; G erm any, 416; F rance, 389, etc.; and O .U .’s— China, u nder 100; India, 150-200; E gypt, 2 0 0 -3 0 0 ; Japan, 3 00 -4 0 0 ; Israel, 5 0 0 -7 5 0 ; Soviet Russia, 750, etc. C lark’s new approach is a vast im provem ent on his old one and opens a wide field for exploration. A s can be readily seen, fundam ental re­ search in m ethods o f international com pari­ sons o f national incom es of various coun­ tries is just beginning. H ow ever, significant developments in this field can be expected in the years im m ediately ahead.

4. Effects of Intertem poral and In­ ternational Differences in the D istribution of N ational Income Intertem poral and international com pari­ sons of national incom e are also com pli­ cated by differences in the distribution of the national incom es th a t are being com ­ pared. This com plication arises if th e com ­ parison is intended to reveal differences in the econom ic w elfare provided by the in­ com es com pared. T h at econom ic w elfare is affected by the distribution of incom e am ong the m em bers of society is indisputable. A billion-dollar incom e, the m ajor portion of w hich is concentrated in the hands of one fam ily, is n o t likely to provide the same am ount of econom ic w elfare to a nation’s people as the same billion dollars m ore equally distributed am ong its m em bers. Even if th at family w ere willing to spend its in­ com e fo r the benefit o f th e society at large, it is doubtful th a t its judgm ent w ould, in the long run, be superior to th a t o f the people themselves. W e assum e, of course, th at th e people receiving and disposing of the incom e would be governing themselves. G iven control of this incom e, they should be able to dispose of it individually and col­ lectively in a m anner th a t will give them the

227

utm ost satisfaction. On the other hand, it is n o t certain th at a com pletely equal distribu­ tion of incom e will produce as m uch eco­ nom ic w elfare as one that is m oderately u n ­ equal and th a t provides reasonably greater rew ards fo r greater individual efforts and ability. E conom ic science has n ot yet p ro ­ vided accurate m easures of the effects on econom ic w elfare of different degrees of in­ equality in the distribution of income. So long as this is true, it is impossible to m ake statistical adjustm ents in national income totals fo r this factor. B ut it is possible to introduce appropriate qualifications w hen using these figures as indicators of econom ic w elfare.11

5. A ctu al International Com pari­ sons E nough statistics are now available to m ake possible m eaningful com parisons of the various aspects of national income for some sixty-odd countries. T he actual num ­ ber of countries fo r w hich some national incom e statistics exist is greater— there are eighty-seven at present— b u t fo r m ore than tw enty of them the statistics are too general fo r elaborate com parisons. In the Statistical A ppendix the eightyseven countries are arranged by cultural and geographic groups. T his is done both for convenience an d also because there is m uch sim ilarity between the statistics of countries located in the same general region o r having th e sam e general cultural patterns. F o r the m ore exacting com parisons, how ­ ever, ventured upon in this section, a dif­ ferent arrangem ent of countries m ust be adopted, nam ely, one based on their levels of incom e m easured by their per capita in­ come. T his classification is m ore suitable for the task at h an d because differences in per capita incom e are even m ore closely related to differences in the econom ic development of countries th an are differences in cultural or geographic factors, and because they are m ore pertinent to the subject m atter. T he cultural-geographic grouping, however, will be retained as a subsidiary arrangem ent, subdividing the countries within each per capita incom e group. T he p er capita incomes used will be con­ verted from dom estic currencies into U.S. dollars at the exchange rates that prevailed during the period 1952-54. T he countries will be classified on the basis of these fig­ ures into fo u r groups. T he fo u r groups will

228

THE INCOM E OF NATIONS

em brace the so-called free or capitalist econ­ omies and will be designated as the high, upper middle, low er middle, and low per capita incom e countries. A fifth group, con­ sisting w holly of the Com m unist countries will be added. This group is set apart be­ cause the Com m unist countries em ploy a wholly different national incom e concept from that used by the rest of the w orld— nam ely, the restricted m aterial production concept— and follow different m ethods of social accounting. As a result, their national incom e statistics are n o t com parable with those of other countries. M oreover, their governm ents control both domestic prices and the foreign exchange rates of their cur­ rencies, w ith the result th a t it is impossible to convert their figures into U.S. dollars with the sam e ease and confidence as it is pos­ sible to do fo r the other countries. T he incomes of the different countries will be com pared in each of the five m ain aspects listed below. A n attem pt will be m ade to reveal the factors responsible for w hat differences in incom e there are in each o f these five respects: 1. incom e p er capita 2. industrial structure o f income 3. distribution of incom e am ong the fac­ tors of production 4. expenditure of incom e fo r various p u r­ poses 5. rate of annual grow th of incom e at constant prices T he com parisons will be m ade for the pe­ riod 1950-55. It is a sufficiently recent and sufficiently long period to disclose exist­ ing long-range trends. T he statistics are derived from the fol­ lowing sources: (a ) U nited N ations series Statistics o f N ational Incom e and E xpendi­ ture, (b ) D r. J. B. D . D erksen’s tabulation in the 3rd quarterly issue of the 1956 bulle­ tin o f the N etherlands C entral Statistical Office, and (c) the m iscellaneous other pub­ lications listed in the Statistical A ppendix and in the various country chapters in P art III and country sections o f the chapters in P art IV of this book.

a. The Relative per Capita Incomes of the Countries T he to tal incom es o f countries m easure their relative capacities to produce, con­ sume, and save fo r the future: in other words, they are a m easure o f their relative econom ic strengths in peace and w ar. T o

be able to interpret these totals, however, it is necessary to know the sizes o f the pop­ ulations producing these incomes and to break the totals down into p er capita in­ com es, fo r it is by com paring p er capita incom es th at we can best m easure the rela­ tive productivity o r econom ic w elfare of the populations involved. A t the m om ent, the only practical, although n o t altogether satis­ factory, international m easure of these per capita incom es is one expressed in U.S. dol­ lars. Such a com parison is presented in T able 16-3. In m ost cases th e figures are those given by D r. J. B. D . D erksen, form er chief of the Social A ccounts Section of the U nited N ations, in his 1956 study o f the national incom es of eighty countries.12 In Section 3 above, it was explained th at p er capita incom e figures in U.S. dollars give only a rough idea of the real differences in the productive pow ers and econom ic wel­ fare of th e different nations and m ust there­ fo re be used w ith a certain am ount of cir­ cum spection. T o this im p o rtan t w ord o f cau­ tio n it is necessary to add only one other observation, nam ely, th a t the roughness of this m easure is greatest in com parisons in­ volving countries belonging to different re­ gional an d cultural groups and differing w idely in the degree o f th eir econom ic de­ velopm ent. T h e m easure is least rough in com parisons involving countries regionally and culturally related an d sim ilarly ad ­ vanced o r sim ilarly underdeveloped eco­ nom ically. In the first case, the dollar of incom e represents about the same purchas­ ing pow er, inasm uch as th e goods produced and consum ed are the same in each and so are the prices at w hich they can be acquired. In the oth er case, the dollar o f incom e has different purchasing pow er in th e countries being com pared, because it is used to com ­ m and wholly different goods and services an d is subject to different price structures in each. This is why per capita incom e fig­ ures expressed in U.S. dollars can be used w ith m uch greater confidence in com pari­ sons o f sim ilarly developed o r sim ilarly u n ­ derdeveloped countries th an in com parisons o f highly developed w ith highly underdevel­ oped countries. F o r exam ple, one m ay well com pare th e $508 p er capita incom e figure reported fo r 1952-54 fo r W est G erm any w ith the $488 figure rep o rted fo r the same year fo r the N etherlands an d conclude that W est G erm any had a slightly higher per cap ita incom e during th a t period th an the N etherlands. On the other hand, W est G er-

1 6 . IN TER T EM PO R A L AND INTERNATIONAL COMPARISONS

229

T able 16-3. Industrial Origins of the Income of 64 Countries in 1955 or N earest Year Per

In Percentages

Income in U.S. Dollars

Agriculture, Incl. Forestry & Fishing

Industry (M fg., Constr., Transp., Trade & Finance)

Services

T otal

1,864 1,296 1,001 778 905 792 744 740 730

4 .3 10.0 2 3 .7 4 .6 9 .3 7 .7 19.5 13.1 15.2

73.5 6 8 .0 5 7 .2 7 4.1 7 3 .8 6 6 .2 6 2 .9 7 1 .8 62.5

2 2 .2 2 2 .0 19.2 2 2.5 1 6.8 2 6 .1 1 7.6 15.2 2 1 .2

100.0 100.0 100.1 101.2 9 9 .9 100.0 100.0 100.1 100.0

983

12.1

6 7 .8

2 0 .3

100.2

Finland W est G ermany Netherlands Ireland Austria Puerto Rico Argentina Uruguay Israel

548 508 488 410 361 426 391 376 487

2 3 .8 10.4 10.9 3 3 .6 14.3 15.9 18.0 2 8 .0 13.2

6 0 .6 7 2 .9 6 7 .2 4 1 .7 6 8 .6 4 9 .7 5 8 .5 4 8 .0 4 9 .9

1 5.6 16.6 2 1 .9 2 5 .8 17.5 3 4 .2 2 3.5 2 4 .0 3 6 .9

100.0 9 9 .9 100.0 100.1 100.4' 9 9 .8 100.0 100.0 100.0

Average for Group II

444

18.7

5 7 .5

2 4 .0

100.2

Italy Spain Greece Cyprus Portugal Chile Panama Colombia M exico G uatem ala1949 Jam aica Ecuador Lebanon Turkey Japan H ong K ong Union of So. Africa M auritius

306 233 204 198 175 342 334 236 192 184 152 150 247 219 184 178 284 227

2 4 .0 3 5 .2 3 5.7 2 8.5 2 8 .6 15.0 3 3 .9 4 0 .6 19.7 4 5 .5 26.5 3 9 .1 19.3 4 1 .9 2 1.8 3 .8 15.1 3 0 .0

5 9 .7 4 6 .2 4 5 .6 4 6 .4 5 3 .6 5 3 .0 3 2 .7 4 0 .2 6 0 .2 4 0 .5 4 9 .4 3 9 .1 5 5 .9 4 1 .9 5 6 .2 6 7 .7 6 0 .6 5 2 .6

16.4 18.6 19.0 2 5 .1 17.9 3 2 .5 3 3 .5 19.1 2 0.1 14.0 2 4 .2 2 2 .0 2 4 .8 16.1 2 1 .9 2 8 .6 2 4 .6 16.9

100.1 100.0 100.1 100.0 100.1 100.5 100.1 9 9 .9 100.0 100.0 100.1 100.2 100.0 9 9 .9 9 9 .9 100.0 100.3 9 9 .5

Average for Group III

225

2 8 .0

5 0.1

2 2 .0

100.0

147 142 134

3 1 .6 4 4 .4 54.1

4 4 .1 27.1 2 8 .0

2 4 .4 2 8 .5 17.8

100.1 100.0 100.0

Countries H igh per Capita Income Group (Over $700) * 1. 2. 3. 4. 5. 6. 7. 8. 9.

U nited States Canada N ew Zealand United Kingdom Luxembourg Belgium Denmark Norway France1962 Average for Group I

Upper M iddle per Capita Income Group ($350-$699) 10. 11. 12. 13. 14. 15. 16. 17. 18.

Lower M iddle per Capita Income Group ($150-$349) 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36.

Low per Capita Income Group (Below $150) 37. Brazil 38. Costa Rica 39. Honduras

THE INCOM E O F NATIONS

230

T able 16-3. Industrial Origins of the Incom e of 64 Countries in 1955 or Nearest Year ( Continued) In Percentages

Agriculture, Incl. Forestry & Fishing

Industry (M fg., Constr., Transp., Trade & Finance)

Services

T otal

Bolivia Peru Paraguay H aiti Pakistan India Burma Philippines Indonesia Taiwan Thailand South Korea E gypt Belgian Congo Nigeria K enya

122 115 107 80 66 64 46 144 88 73 71 70 122 71 64 54

5 9 .4 3 1 .2 4 8 .2 7 3 .7 5 7 .7 4 6 .2 4 2 .8 4 2 .0 5 6 .5 3 2 .9 4 4 .1 5 1 .9 3 5 .2 2 7 .6 6 6 .2 4 0 .8

19.7 5 1 .9 3 2 .1 16.3 2 1 .6 3 7 .0 4 2 .2 33 .2 28 .2 4 7 .3 3 7 .9 25 .7 3 1 .2 52.1 2 5 .9 4 0 .9

2 0 .8 16.9 19.5 10.0 2 0 .8 16.7 15.1 2 4 .8 15.5 19.9 18.1 2 2 .3 3 3 .6 2 0 .3 8 .0 18.3

9 9 .9 100.0 9 9 .8 100.0 100.1 9 9 .9 100.1 100.0 100.0 100.1 100.0 9 9 .9 100.0 100.0 100.1 100.0

Average for Group IV Average for Groups I -I V

94 339

4 6 .7 3 0 .3

3 3 .8 4 8 .6

19.5 2 1 .2

100.0 100.1

400-500

17.0 2 5 .9 2 5 .8 6 1 .6 3 4 .0

8 3 .0 74.1 7 0.2 3 8 .4

4 .0

100.0 100.0 100.0

3 3 .4 1952 3 2 .2 1963 g-v.

66.6 1952 6 1 .7 1963 g-v.

7.1

100.0 100.0 100.0 100.0

3 2 .8

6 5 .7

1 .6

100.1

Countries 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55.

Per Capita Income in U.S. Dollars

V. C om m unist Countries Upper M iddle and Lower M iddle per Capita Incom e Group 56. 57. 58. 59. 60. 61. 62. 63. 64.

Czechoslovakia Soviet Russia 1 Poland j Hungary Rumania E ast Germany Bulgaria Yugoslavia Albania A verage for Group V

200-400

200-300 177 150 265

100.0

66.0

* Breakdowns by industrial divisions are unavailable for Switzerland (per capita income $1,007), Australia ($955), and Sweden ($941). “ T otal is greater than the sum of the com ponents because of unallocable items, g.v. Gross value of the output. so u rc e s:

derived from figures given in the sources listed

m any and V enezuela h ad about the same per capita dollar incom e during th a t period, $508 and $503 respectively, yet because they are so different econom ically they can­ n o t be deem ed to be identical in productivity and in the degree of econom ic w elfare en­ joyed by their people. Even less m eaningful o r precise w ould be a com parison of the per capita dollar incom e o f tw o even m ore widely different countries, such as the U nited States and Burm a. All th a t these figures— $1,864 for the U nited

in

the Statistical Appendix.

States and $46 for B urm a in 1952-54— w ould indicate is th at the difference betw een th e productivity and econom ic w elfare of the two nations is imm ense. B ut it would be inaccurate for anyone to m aintain, on th e basis of these figures, th a t th e average A m erican is twenty-eight tim es as produc­ tive an d as well off as th e average Burmese. T he real difference in productivity and m aterial well-being m ay be m uch greater o r m uch less. H ow m uch m ore o r less we will n ot know until an international m easure

1 6 . IN TERTEM PO RAL AND INTERNATIONAL COMPARISONS

of p er capita productivity and m aterial well­ being is devised th a t takes into account and properly weighs all the pertinent factors. But even in com parisons involving coun­ tries th a t belong to the same geographic or cultural group and are sim ilarly advanced or underdeveloped econom ically, care m ust be exercised lest the figures of the per capita dollar incomes are distorted by im balances of exchange rates or price levels. E ven the per capita incom e figures o f such closely related countries as the U nited States and C anada will bear exam ination in this respect. These figures— $1,864 fo r the U nited States and $1,296 for C anada in 1952-54— suggest a difference of about 30 p er cent in the productivity and m aterial well-being of the tw o populations during th at period. A ctually, the difference m ay be som ew hat smaller, inasm uch as the C anadian dollar, even though it is officially exchanged fo r the U.S. dollar almost at par, has greater purchasing pow er in C anada than the U.S. dollar has in the U nited States. P ostw ar inflation was n o t quite as great in C anada as it was in the U nited States, and the price level there did n o t rise quite so m uch. A n o th er factor to be considered in all such com parisons is the effect w hich a difference in the age com position of the populations involved m ay have on the fig­ ures u n d er consideration. T he same per capita incom e figure fo r a country having a large proportion of children in its popula­ tion w ould represent both a greater produc­ tive capacity and a greater degree o f eco­ nom ic w elfare than would th e sam e per capita income figure fo r a country having a sm aller proportion of children in its popu­ lation. This point can be illustrated by D erksen’s com parative analysis (in the same study) o f the $792 per capita incom e o f Belgium and the $488 p er capita income of th e N etherlands in 1952-54. A p a rt of the difference, he said, was due to the fact that 39 p e r cent of the population of the N etherlands was less th an tw enty-one years old, whereas only 29 p er cent of Belgium ’s population fell in this age group. A com ­ p arison of income per gainfully em ployed person would have show n the differential in th e productivity o f the tw o populations to be considerably less than indicated by p er capita incom e figures. T h e fo u r per capita incom e groups into w hich the free economies are classified can be identified as follows:

231

1. H igh per capita incom e countries, hav­ ing in 1952-54 p er capita incomes of m ore th an $700 2. U pper m iddle per capita incom e coun­ tries w ith p er capita incomes of $350 to $699 3. L o w er m iddle per capita incom e coun­ tries w ith a range o f $150 to $349 in p er capita incomes 4. L o w per capita incom e countries with a range of per capita incomes below $150 T his division parallels som ew hat the cultural and geographical grouping followed in the Statistical A ppendix, except th at it elim inates situations th at represent an ex­ ception from the regional pattern. Thus, the high per capita incom e group em braces m ostly the English-speaking and some west­ ern E uropean countries; the upper middle p er capita incom e group, mostly w estern E uropean countries, with a slight sprinkling o f C aribbean and South A m erican countries and one N ear E astern country; the lower m iddle p e r capita incom e group contains a m iscellany of south E uropean, L atin A m eri­ can, Caribbean, and N ear, M iddle, and F ar E astern countries, and two A frican coun­ tries; while the low per capita income group consists of a m iscellany of L atin A m erican, A siatic, and especially A frican countries. T he Com m unist countries are placed in a special group. A s previously stated, their incom e figures do n o t lend themselves to easy translation into U.S. dollars because of the controlled nature of their foreign trade an d pricing system and the artificial natu re of their foreign exchange rates. T here are also other difficulties in their case, m ak­ ing the p reparation o f equally valid per capita incom e figures fo r them impossible. These difficulties can be illustrated by the examples of Poland, Soviet Russia, and China. T he p er capita national incom e of Poland, according to official estimates, was 7,800 zloty in 1955.13 T he official rate of exchange at th a t tim e was 4 zloty p er $1.00, b ut the unofficial o r black m arket rate was 70 zloty per dollar.14 A t the official rate, P oland’s p er capita national incom e in th at year w ould have been $1,450, o r next to th at of the U nited States, w hich was $1,864; at the unofficial rate it w ould have been only $111, or less th an one half o f Spain’s per capita incom e o f $233. Both results are equally unbelievable.

232

THE IN CO M E OF NATIONS

In the case of Soviet Russia, a wholly different difficulty presents itself. Since 1940 the Soviet governm ent has n o t been pub­ lishing any national incom e figures in rubles; and the figures fo r 1940— th e last published — were given in 1926-27 rubles only. T he only figures th a t the governm ent has been releasing on the subject have been index num bers of the grow th since 1915 of the national incom e at constant prices, w ith the figures fo r the year 1913 o r for 1940 being taken as a hundred. U p to 1950 the index was based on 1926-27 prices, since then, on 1950 prices. By applying to the 1940 ruble figure of the national incom e— 125,500 million rubles in 1926-27 prices— the 181 per cent increase from 1940 to 1955 shown by the index, we arrive at the na­ tional incom e total for th a t country for 1955 of 352,655 m illion rubles at 1926-27 prices. Since the population of Soviet Russia in 1955 was approxim ately 200 million, this w ould give Soviet Russia a per capita in­ com e in th a t year o f 1,760 rubles at 19 2 6 27 prices. A ssum ing th a t the results of this indirect calculation are correct (and we can­ n o t be too sure of it) , w hat am ounts do they represent in curren t rubles and how can these am ounts be converted into U.S. dollars at current prices? If Soviet statistics gave a price index showing the relation of current prices to 1926-27 prices, one could obtain the total in current rubles and do the conversion of them into U.S. dollars at some reasonable exchange rate. B ut Soviet statistics do n o t give such a price index, and it is impossible, therefore, to convert the 1926-27 ruble figures into current U.S. dollar figures. C hina’s per capita incom e is given by the E conom ic Survey fo r A sia and the Far East in 1956, at 141 yuan fo r 1955. T he u n ­ official rate of foreign exchange in 1955 was 4.45 to 5.35 yuan per $1.00. A t this rate the per capita incom e of China w ould be only $28, or less than one half o f India’s $64 p er capita incom e fo r 1955. Y et the facts would indicate th at C hina’s level of incom e could scarcely have been m uch low er than India’s before W orld W ar II and th a t since then, under the Com m unist re­ gime, it m ight even have risen above India’s incom e level. W e m ust abandon, therefore, the attem pt to arrive at dollar figures fo r the p er capita incom es of C om m unist countries by any currency conversion m ethod. Fortunately, the result we seek can be achieved by a

ro u n d ab o u t m ethod. T his is the m ethod of substitution fo r the p er capita incom e of a C om m unist country th e corresponding know n dollar per capita incom e figure o f a free econom y appearing to have a similar standard o f living, rate of capital accum ula­ tion, and rate of governm ent consum ption. T hus, fo r example, on the basis of both the prew ar and postw ar experience, Czecho­ slovakia’s level of incom e could be as­ sum ed to be com parable to W est G erm any’s, w hich in 1952-54 was $508 per capita. One w ould be justified, accordingly, to classify C zechoslovakia as belonging to the upper m iddle p er capita incom e group. O n the basis o f sim ilar prew ar an d postw ar data, E ast G erm any’s level of incom e w ould ap­ pear to be considerably below W est G er­ m an y ’s an d also substantially below Italy’s, w hich in 1952-54 was represented by a per capita incom e of $306. Soviet Russia and P o lan d ’s levels o f national incom e w ere also below W est G erm any’s, although by not quite as w ide a m argin, an d w ere probably com parable to Italy’s $306 o r A ustria’s $361 per capita incom es. H ungary, R um ania, and B ulgaria probably h ad incom e levels lying som ew here between Italy’s $306 p er capita incom e an d G reece’s $204 p er capita in­ com e. F o r Yugoslavia we have a reasonably good figure o f a per capita incom e o f $177 fo r 1952-54. A lbania, w hich is m ore back­ w ard econom ically th an Y ugoslavia, p ro b ­ ably had a som ew hat low er level o f income, approaching a $150 p er capita. T h e Com m unist countries o f A sia— C hina, N o rth K orea, M ongolia, an d V iet­ nam — judged by com parisons w ith the free econom ies of Asia, such as India, Pakistan, Burm a, and South K orea, fo r w hich national incom e figures in U.S. dollars are available, can be estim ated to have p e r capita incomes substantially below $150 and can be desig­ n ated as low p er capita incom e countries. Thus, Com m unist C hina’s level o f in­ com e, after allowing fo r th e considerable econom ic advances m ade by th a t country u n d er C om m unism , m ust have been some­ w hat higher in 1952-54 th an South K o rea’s incom e o f $70 p er capita an d In d ia’s $64. A t any rate, she belongs to th e low per capita incom e group of countries, w ith a range of p er capita incom e in 1955 still considerably below $100. T hus, by this adm ittedly ro u n d ab o u t ap­ proach, we arrive at th e follow ing classifica­ tion o f the C om m unist countries on the basis o f their per capita incom es in 1955:

1 6 . IN TER T EM PO R A L AND INTERNATIONAL COMPARISONS

Czechoslovakia

Range of per Capita Income

Group to W hich the Country Belongs

$400 to $500

Upper m iddle per capita incom e (Group II)

S oviet Russia Poland Hungary Rumania E ast G ermany

$250 to $400

Lower range of the upper middle, or higher range of the lower middle per capita income (Group II or III)

Bulgaria ] Yugoslavia f Albania J

$150 to $250

Lower m iddle per capita income (Group III)

China North Korea M ongolia Vietnam

$50 to $150

Low per capita income (Group IV)

It should be noted th a t the particular dollar figures adopted fo r the identification of the four p er capita incom e groups are based entirely on the incom e statistics of the period 1952-54, and w ould not apply to oth er periods. A different period m ay be characterized by different dollar ranges fo r the same groups. Similarly, the classification of countries in these fo u r groups relates only to the situation th a t obtained during the period in question. A substantially dif­ ferent distribution m ight be revealed during another period, some countries m oving up into a higher incom e bracket and others slipping down into a low er one.

b. Differences in the Industrial Origins of Income T he national incom e statistics published by the U nited N ations show the origin of the national incom e from eleven branches of econom ic activity: 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

agriculture, forestry and fishing m ining and quarrying m anufacturing construction electricity, gas, w ater and sanitary sources transportation, storage, and com m u­ nications wholesale and retail trade banking, insurance, and real estate ownership o f dwellings public adm inistration and defense other services (dom estics, professions, trades, hotels, restaurants, etc.)

In m ost of our presentations we have re­ duced this list to seven by consolidating

233

branches 2, 3 and 4 into one group; branches 5 and 6 into another; and 7 and 8 into still another. In the present analysis we reduce these seven by fu rth er consolidation into three as follow s: (1 ) agriculture, including forestry an d fishing, (2 ) industry and trade, including also construction, utilities, trans­ portation, com m unications, banking, insur­ ance, and real estate, and (3 ) services, in­ cluding services of dwellings, services of governm ent, and m iscellaneous other serv­ ices. T his threefold grouping is very different from the one o f so-called prim ary, second­ ary, an d tertiary production given great cur­ rency by Colin C lark in 1940 in his book The C onditions o f E conom ic Progress (pp. 3 3 7 -4 1 ). C lark’s prim ary production group is the same as ours, b ut his other two groups are different. H is secondary production is m ore lim ited th an our industrial group since it includes only m anufacturing, mining, and construction. H is tertiary production, on the oth er hand, is m uch b roader th an our service group, as it includes trade, transportation, com m unications, utilities, banking, insur­ ance, and real estate, besides public adm in­ istration, domestic, personal, and profes­ sional services. On the other hand, C lark’s tertiary production does n o t include the services o f dwellings, which, fo r reasons that are n o t m ade clear, he omits altogether from national income. W e p refer o u r grouping to C lark’s. T rade and transport, in our judgm ent, be­ long in the same group w ith m anufacturing and m ining because the last tw o cannot be conducted w ithout them and because, his­ torically, they w ere developed at the same

234

THE IN CO M E OF NATIONS

tim e. On the other hand, the other three categories of services— those of dwellings, o f governm ent, and o f domestics, profes­ sions, etc.— have independent roots and are present in both agricultural and indus­ trial economies. T able 16-3 shows th a t in low per capita incom e countries the largest portion of the incom e is derived from agriculture, a smaller portion from industry, and the smallest from services. In the low er middle per capita income group the proportion of incom e from agriculture drops to second place, while the proportion of income from indus­ try moves up to first place, and th e p ro ­ portion of incom e derived from services, though still small, is slightly larger than in the lowest per capita incom e countries. Finally, the table shows that in the upper m iddle and high p er capita incom e coun­ tries, incom e from agriculture drops to th ird place while incom e from services moves from third to second place, and in­ com e from industry becomes even m ore predom inant. In the high per capita incom e countries the role of services is not alto­ gether clear. T he proportion of their con­ tribution to incom e is higher th an in the low per capita incom e group, but, contrary to expectations, it is not as high as in the m iddle per capita incom e group. It is pos­ sible th a t this deviation from a clear-cut trend is caused by peculiar situations ob­ taining in a few countries in each group. If the influence of these exceptions is elim­ inated, a m ore consistent, if slight, rise in the proportion of service incomes w ith the rise in per capita incom es w ould appear. T he existence of a correlation between the size of th e per capita incom e of a coun­ try and the degree of developm ent of its industry is n o t surprising. It reflects the fact, know n to political economists (physio­ crats excepted) fo r m ore than three h u n ­ dred years, th a t m anufacturing and trade, taken together, are m ore productive o f in­ com e th an agriculture. This is, however, only a general proposition and does n ot hold true necessarily fo r every branch of industry and trade any m ore than it does fo r every branch o f agriculture. T here are some branches of industry that are not highly productive, and there are some that are. F o r example, th e glovemaking industry is no longer as productive as it used to be, while the production of high quality wool— a branch o f agriculture— is still a very pro ­ ductive occupation.

W ithin this general tendency of per capita incom es to rise w ith the advance of indus­ try, considerable variations as between in­ dividual countries exist. T hus, it will be noticed in T able 16-3 th a t in m ost of the high p er capita incom e countries industry an d trade supply from 68 to 74 per cent o f th e national income w ith agriculture sup­ plying only 4 to 10 p er cent, b ut th a t in tw o countries in this sam e group— N ew Z ealand and D enm ark— the share of indus­ try and trad e is only 57 to 62 p er cent, while the share of agriculture (because of its exceptional productivity th ere) was 23.7 per cent in one case an d 19.5 p er cent in the other. In the low p e r capita incom e group, the average proportion of incom e from in­ dustry is only 33.7 per cent, b ut in three countries— Brazil, Taiw an, an d th e Belgian Congo— probably because o f large foreign credits an d technical assistance, those pro­ portions are as high as 44.1, 47.3, and 52.1 p er cent respectively. T h e fact, shown in T able 16-3, th a t the proportions of incom e derived from serv­ ices, as here defined, do n ot vary altogether consistently w ith p er capita incom e is some­ w hat surprising. O ne m ight have expected greater inconsistency, especially since serv­ ice incom es have been designated in some m odern econom ic literature as “tertiary p ro ­ duction,” im plying th a t it is a superstructure characteristic of the highest and latest stage o f econom ic developm ent of society. A ctu­ ally, the theory th a t services are a m anifesta­ tion of m ore advanced econom ic develop­ m ent is only partially true. Services, as we have defined them , are also im p o rtan t in m o re prim itive o r undeveloped societies. Services m ay be cruder there, b u t just as plentiful, as, fo r exam ple, th e m aintenance o f m ultitudes of domestics in th e households of th e w ealthy and ruling families. In ad­ vanced societies, services take new er and m ore refined form s, m ultiply in volum e, and are b etter rem unerated. T h eir grow th m ay keep pace w ith the grow th of industry and trad e an d even slightly o u tru n them . T he forces responsible fo r this tren d can be pointed up m ore clearly by analyzing sep­ arately the relation of each o f the three m ain categories of service incom e— incom e fro m the services of dwellings, from govern­ m ent, an d from all o th er services— to the level of the country’s per cap ita income. It is seen from T able 16-4 th a t the p ro ­ portions of incom e obtained from th e serv­ ices o f dwellings bear a slightly reverse rela-

1 6 . IN TER T EM PO R A L AND INTERNATIONAL COMPARISONS

235

Table 16-4. Proportions of N ational Incom e Obtained from Each of the Three Categories of Services in 43 Countries in 1955 or Nearest Year In Percentages

Services of D w elling

Services of G overnm ent

Other Services (Professions, D om es­ tics, Trades, etc.)

T otal

U nited States Canada N ew Zealand U nited Kingdom Belgium Denm ark Norway Luxembourg France 1962

2 .2 5 .1 4 .6 3 .2 7 .6 4 .1 2 .6 3 .9 0 .8

11.9 7 .3 3 .3 6 .3 6 .0 9 .0 4 .0 1 0.8 1 0.9

8 .1 9 .6 11.3 13.0 12.5 4 .5 8 .6 2 .1 9 .5

2 2.2 2 2 .0 19.2 2 2.5 2 6.1 1 7.6 15.2 1 6.8 2 1 .2

Average for Group I

3 .8

7 .7

8 .8

2 0.3

3 .6 2 .7 5 .5 1 .8 6 .3

9 .1 9 .1 6 .6 9 .4 19.5

2 .9 4 .8 9 .8 1 4.6 11.1

15.6 16.6 2 1 .9 2 5 .8 3 6 .9

Countries . H igh per Capita Income Group {Over $700) 1. 2. 3. 4. 5. 6. 7. 8. 9.

U pper M iddle per Capita Income Group ($350-$699) 10. 11. 12. 13. 14.

Finland W est Germany Netherlands Ireland Israel Average for Group 11“

--

---

---

---

4 .0

10.7

8 .6

2 3 .3

1.7 2 .5 5 .9 6 .4 4 .0 8 .7 1 2.8 6 .6 4 .4 4 .9 7 .5 8 .8 3 .6 5 .0 8 .6 6 .4

10.7 7 .1 7 .4 9 .1 5 .2 1 0.9 5 .9 6 .1 5 .1 6 .3 5 .6 6 .2 8 .4 1 0.0 1 0.0

4 .0 9 .0 5 .7 9 .6 8 .7 1 2.9 1 4.8 6 .4 1 0.6 1 3.0 8 .9 9 .5 4 .1 6 .9 1 0.0 8 .1

1 6.4 18.6 1 9.0 2 5.1 17.9 3 2 .5 3 3 .5 19.1 20.1 2 4 .2 2 2 .0 2 4 .8 16.1 2 1 .9 2 8 .6 16.9

6 .1

7 .3

8 .9

2 2.3

3 .8 4 .8 9 .1 6 .4 6 .4 6 .0 4 .5

7 .1 11.4 3 .6 4 .2 4 .5 5 .8 5 .2

13.5 12.3 5 .2 10.2 8 .6 9 .0 7 .0

2 4 .4 2 8.5 17.9 2 0 .8 19.5 2 0 .8 1 6.7

Lower M iddle per C apita Income Group ($150-$349) 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 20.

Italy Spain Greece Cyprus Portugal Chile Panama Colombia Mexico Jamaica Ecuador Lebanon Turkey Japan H ong Kong M auritius Average for Group III

Low per Capita Income Group (Below $150) 31. 32. 33. 34. 35. 36. 37.

Brazil Costa Rica Honduras Bolivia Paraguay Pakistan India

“ For the other four countries in Group II— Austria, Puerto Rico, Argentina, and Uruguay— the incom e from services is not segregated into these three subdivisions.

THE INCO M E OF NATIONS

236

Table 16-4. Proportions of National Incom e Obtained from Each of the Three Categories of Services in 43 Countries in 1955 or N earest Year ( Continued) In Percentages

Services of Dw elling

Services of Governm ent

Other Services (Professions, D om es­ tics, Trades, etc.)

T otal

Burma Indonesia E gypt Belgian Congo Nigeria Kenya

3 .9 6 .5 7 .1 5 .1 .9 3 .4

9 .6 6 .5 16.1 9 .9 4 .5 11.2

1 .6 2 .5 1 0 .4 5 .3 2 .6 3 .7

15.1 15.5 3 3 .6 2 0 .3 8 .0 18.3

Average for Group IV A verage for the 43 countries in the four groups

5 .2

7 .7

7 .1

2 0 .0

5 .1

7 .9

8 .3

2 1 .3

Countries 38. 39. 40. 41. 42. 43.

s o u r c e s : derived from figures given in the sources listed in the Statistical Appendix.

tionship to the level of per capita incomes, tending to be slightly lower in the higher per capita than in the low er per capita income countries. Unless the figures are im properly com puted (representing underestim ates in high per capita income countries or over­ estim ates in low per capita incom e coun­ trie s), this m eans th a t shelter, being one of the basic necessities, requires a somewhat greater share of productive effort to secure and m aintain in prim itive than in advanced societies. Even though it becomes m uch m ore elaborate in advanced economies and incorporates elements o f com fort and even luxury, it is in com petition there w ith a m uch w ider range of w ants and consequently does not play such a significant role as a producer of income. T he table also suggests th a t social and physical factors have a great deal to do w ith the size of the contribution th a t the services of dwellings m ake to the national incom e. Thus, fo r example, in Panam a and H ong K ong, w here population is highly concentrated in small u rban areas, the incom e from services of dwellings is exceptionally high. The figures do not show any clear-cut correlation between the level of incom e of a country and the proportion of incom e from the services of governm ent. On one hand, the high per capita incom e countries can afford to m aintain a w ider range o f social services. On the other hand, depending on political circum stances, they m ay o r m ay not m aintain a com paratively large m ilitary es­ tablishm ent. F o r example, in the U nited States, both of these activities of governm ent

are considerable, w ith the result th at the proportion o f incom e from governm ent is large. In Egypt, on the other hand— a low p er capita incom e country— the proportion o f incom e from governm ent is even larger, probably because of its m aintenance of a sizable m ilitary establishm ent. In Israel— an upper m iddle per capita incom e country— th e proportion of incom e from governm ent is still larger due to its m aintenance of both extensive social services and arm ed forces. It is also possible th at some of the variations in the im portance of this incom e are due to differences in the m ethods of accounting for the services of governm ent by the different countries. T h e shares of incom e obtained from re­ m aining services, as show n in T able 16-4, seem to be even less related to levels o f in­ com e and the degree o f econom ic develop­ m ent th an are the shares from the other two categories of services. T hey do n o t vary m uch fro m group to group from the over-all average of 8.3 per cent, b ut they do vary considerably between individual countries w ithin any given group. In som e cases, the p roportion m ay be high because of extensive tourist trade, in others because of a vast church establish­ m ent. T he proportion is exceptionally high in Ireland, P anam a, and Jam aica. T he rea­ sons fo r the variations in th e proportion of this incom e are n o t easily detected. It is clear th at in o rd er to arrive at m eaningful conclusions regarding the fac­ tors affecting the shares of incom e obtained from services it is necessary to have m uch

1 6 . IN TERTEM PO RAL AND INTERNATIONAL COMPARISONS

m ore statistical d ata th an are now available. T h e accounting in C om m unist countries of th e proportions of national incom e de­ rived from different branches of production is n ot com parable w ith the accounting of them in the free o r capitalist economies. Inasm uch as the incom es from services are n ot included in the national incom e totals, the proportions reported in them fo r the incom es from industry and agriculture are higher than in the free econom ies, even though under a com parable m ethod of social accounting their relative im portance m ay be the same. Social accounting in C om m unist countries also tends to exaggerate the pro ­ portion of incom e derived from industry and correspondingly understate the proportion derived from agriculture. This results from the fact that the value of industrial produc­ tion is credited w ith the w hole am ount of the turnover tax, even though the proceeds of the latter are also used to finance the capital accum ulation and other costs of agricultural production. E ven w ith these qualifications, the C om ­ m unist countries reveal the same type of correlation between the share o f incom e derived from industry and the size o f the co untry’s per capita incom e th a t is revealed in the case of the free econom ies, nam ely, th a t w ith certain lim itations the per capita incom e rises w ith the share of incom e de­ rived from industry. T hus, it will be noted in T able 16-3 th a t Czechoslovakia, w hich has the highest proportion of incom e from industry, has the highest p er capita incom e am ong the Com m unist countries, while Soviet Russia, Poland, and E ast G erm any, w ith low er proportions of industrial incom e th an Czechoslovakia b u t higher th an Bul­ garia o r Yugoslavia, have a low er p er capita income than Czechoslovakia, but a higher one th an the latter two countries. However, it is n o t always true in C om ­ m unist countries, any m ore th an in others, th a t a rise in the proportion of incom e de­ rived from industry brings about a rise in p er capita incom e, and that, by the same token, a fall in th a t prop ortion brings about the opposite result. T he low proportion of incom e from agriculture m ay be due to backw ard m ethods of cultivation o r to crop failure. A rise in the proportion of incom e fro m agriculture, if caused by a bum per crop o r by a shift to m ore efficient cultivation o r to raising m ore productive crops, m ay bring in its w ake a rise in p er capita income. T hus, in Soviet Russia, th e sixth five-year

237

plan (1 9 5 6 -6 0 ) provides fo r a rise in the relative w eight assigned to the income from agriculture and also fo r a rise in the n a­ tional incom e p artly in consequence of the m easures to be taken to im prove the m eth­ ods o f cultivation and, particularly, to in­ crease the production o f m eat, milk, and b utter.15

c. Differences in the Distribution of In­ come Among the Factors of Production T he distribution of incom e am ong the factors of production is discussed here only as it takes place in th e free or capitalist economies, w here alone these factors are recognized. In the national income account­ ing of these countries the distribution is shown u n d er the following fo u r headings: labor incom e; m ixed capital-labor income; capital incom e (interest, rent, dividends, and undistributed corporate p rofits); and resid­ ual incom e, consisting of corporate income taxes (the proceeds of w hich belong to the governm ent) and the profits of government enterprises less interest from government and consum er debts. In the Com munist countries the distribution of incom e am ong the different groups of population is seldom shown, an d w here it is given, it follows w holly different lines. In the free economies, the distribution of incom e differs from country to country, partly w ith the degree o f econom ic develop­ m ent as indicated by th e level of the coun­ try ’s p er capita incom e, and partly with the character o f the governm ent and the nature o f the social institutions o f the country. T able 16-5 shows th e extent to w hich the distribution varies w ith the first-m entioned factor. W here it does n ot seem to do so, the influence of political and social variables can be inferred. It will be noticed th at in the underdevel­ oped, low per capita incom e free economies, the largest proportion of the national incom e — m ore th an 50 p er cent— is of the mixed capital-labor type. It is earned by proprietors, largely farm ers, w ho are both capitalists of small dim ensions and laborers. Pure labor incom e represents only 30 to 40 p er cent o f the national incom e, and pure capital incom e only about 10 per cent. In the m ore advanced economies, incom e becomes m ore differentiated. L abor incom e, in particular, rises in im portance, representing 60 to 75 p er cent of the national incom e in the econ­ omies having the highest p er capita income, b ut it includes the salaries and bonuses of

238

THE INCOM E OF NATIONS

Table 16-5. Distribution of Income Am ong the Factors of Production in 41 Countries in 1955 or N earest Year In Percentages

Countries

Labor Income

M ixed Incom e

Capital Income

Residual In com e”

6 9 .5 6 4.3 5 6 .9 6 0 .5 7 3 .2 5 9 .2 6 2 .7 5 7 .6 5 2 .9 5 4 .7 5 6 .2 5 9 .5

12.1 1 5.4 2 5 .4 2 1 .8 9 .7 2 0 .0 19.7 2 6 .6 3 0 .6

1 4 .6 15.1 11.1 13.3 1 5 .4 1 5.9 12.7 9 .1 1 5.0

3 .9 6 .0 6 .6 4 .4 1 .7 4 .9 4 .9 6 .6 1.5 1 .8

(6 1 .6 ) 6 0 .7

(2 1 .2 )

6 1 .2 6 3 .6 5 8 .1 5 1 .6 6 5 .8 5 8 .1 5 4 .0 6 2 .7

2 4 .3

T otal

I. H igh per Capita Incom e Group (Over $700) 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

U nited States Canada N ew Zealand Australia United Kingdom Switzerland Sweden Luxembourg Belgium Denmark Norway France Average for Group I : the 10 countries w ith segregated figures All 12 countries

7 .8

1 .6

100.1 100.9 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

(1 3 .0 )

(4 .2 ) 3 .8

(1 0 0 .0 ) 100.0

6 .9 3 .9 3 .8 .9 1.5

2 .4

100.0 100.0 101.0 100.9 100.0 100.0 100.0 101.8

(1 5 .6 )

(1 .8 ) 1 .9

(1 0 0 .7 ) 100.2

16.5 9 .5 9 .7 14.9

.9 1 .3 2 .5 7 .1 1 .9

2 4 .1 8 .9 21.1

3 .8 3 .1 3 .7

100.6 100.0 100.2 100.2 100.0 100.0 100.0 100.0 100.0

(1 5 .0 )

(3 .1 ) 2 .7

(1 0 0 .1 ) 100.4

13.7 7 .5

1 .4 1 .6

100.0 100.0

4 3 .5 32..0

11.8 3 1 .1

3 5 .5

II. Upper M iddle per Capita Income Group ($350-$699) 13. 14. 15. 16. 17. 18. 19. 20.

Finland W est Germany Austria Ireland Puerto Rico Argentina Uruguay Israel Average for Group II: the 3 countries w ith segregated figures All 8 countries

7 .6 3 2 .5 3 9 .1 4 7 .5 13.3

2 2 .4 4 1 .9

2 6 .0

2 0 .0 3 6 .7

(6 0 .3 ) 5 9 .4

(2 2 .2 )

3 2 .4 4 6 .5 5 3 .0 6 4 .3 4 7 .1 4 3 .7 4 7 .8 4 7 .6 4 4 .7

5 0 .8 4 2 .7

3 8 .9

III. Lower M iddle per Capita Income Group ($150-$349) 21. 22. 23. 24. 25. 26. 27. 28. 29.

Cyprus Chile Panama Cuba Colombia M exico Ecuador Japan H ong Kong Average for Group III: the 7 countries w ith segregated figures All 9 countries

4 4 .7 19.1 3 6 .1 5 6 .3 2 4 .3 4 0 .4 3 0 .5

(4 7 .2 ) 4 7 .5

(3 4 .8 )

4 1 .8 4 9 .9

43 .1 4 1 .0

5 0 .2

IV. Lower per Capita Incom e Group (Below $150) 30. Brazil 31. Honduras

1 6 . IN TERTEM PO RAL AND INTERNATIONAL COMPARISONS

239

In Percentages

Countries 32. 33. 34. 35. 36. 37. 38. 39. 40. 41.

B olivia1949 Peru Philippines Taiwan Rhodesia-Nyasaland Feder­ ation Gold Coast Belgian Congo Nigeria Kenya Uganda Average for Group I V : the 8 countries w ith segregated figures All 12 countries Average for the 27 countries with the segregated figures Average for the 41 countries

Labor Incom e 5 2 .0 3 7 .5 4 2 .4 3 2 .9 5 5 .2 2 1 .1 4 7 .7 5 .3 3 9 .3 2 2 .5

M ixed Incom e

Capital Income

Residual Income “

T otal

1 7.6 10.4

5 .4 1.5 3 .3

100.0 100.0 100.0 100.3

8 .8 1 3.0 .8

12.1 1 2.6 6 .0 7 .8

12.3

.5

9 9 .9 100.1 100.0 100.1 100.0 100.0

(1 0 .5 )

(4 .8 ) 4 .4

(1 0 0 .0 ) 100.0

(1 3 .0 )

(3 .9 ) 3 .4

(100.0) 100.1

4 8 .0 3 9 .5 5 6 .1 5 3 .7 3 2 .6 5 7 .6 3 3 .3 8 6 .2 6 0 .7 6 4 .7

(3 2 .3 ) 3 7 .3

(5 2 .4 )

(4 8 .9 ) 5 0 .6

(3 4 .2 )

5 8 .3

4 6 .1

a Corporate taxes plus income from governm ent enterprises less interest on governm ent and consumer debts. s o u r c e s : derived from figures given in the sources listed in the Statistical Appendix.

the m anagerial class, w hich are often quite substantial. C apital incom e also plays a m ore im portant role in the high per capita incom e countries, b u t n o t quite in the same degree, rising to only about 15 p er cent of the national incom e. M ixed incom e corre­ spondingly declines in im portance, falling below 20 per cent o f the total. T he fact that in the m ore advanced free economies labor incom e form s so large a proportion of national incom e, and capital incom e so sm all a one is partly the result o f the nature of capitalism and partly the result of the operation of the dem ocratic forces associated w ith its political and social aspects. It is, in fact, a result of the labor m ovem ent; of the em ergence of a large, salary-earning m anagerial class; of th e ex­ pansion of the social and cultural services of governm ent, including social insurances; as well as of the operations of the progres­ sive income and inheritance taxes, w hich tend to break up large fortunes o r to prevent their form ation, equalizing the opportunities fo r economic advancem ent. W ithin the general fram ew ork of these broad tendencies, considerable variations occur as between different countries because o f a variety of special circum stances. F o r exam ple, the distribution of incom e in Bel­

gium is quite different from the distribution in the U nited K ingdom , although Belgium’s level of incom e is also quite high. W hereas in the U nited K ingdom labor income con­ stitutes 73.2 p er cent of the national income, in Belgium it constitutes only 52.9 p er cent; and w here in the form er mixed income represents only 9.7 per cent of national in­ com e, in the latter it represents 30.6 per cent. T he differences are due to the fact th a t agriculture is m ore im portant in the Belgian econom y and th a t Belgian industry is organized u nder the personal type of proprietorship to a larger extent than British industry. In F rance, an upper middle per capita incom e country, m ixed income also represents a large share of the total national income, 31.1 p er cent, whereas capital in­ com e represents only a 7.8 per cent share. In F ran ce’s case, this is due to an even greater im portance of agriculture in the econom y and an even greater prevalence of personal proprietorships in industry. In E cuador, H ong K ong, and Peru— all lower m iddle or low per capita incom e countries— the proportions of capital income in the total national incom e are, fo r a variety of special reasons, considerably higher than is the norm al fo r th eir incom e groups.

240

THE IN CO M E OF NATIONS

In the C om m unist countries, the distribu­ tion of national incom e in the socialized sector takes the form of three shares: com ­ pensation of labor engaged in m aterial pro­ duction, including the supplem ents fo r social insurance; profits of governm ent enterprises, the bulk of w hich is appropriated by the governm ent and is used largely fo r capital accum ulation purposes; and, finally, the turnover and other taxes, the proceeds of w hich go to the state and are used fo r the advancem ent of com m on national objectives. U nfortunately, this distribution is seldom shown. In the nonsocialized sector the dis­ tribution is different, but the im portance of this sector in the Com m unist countries is generally very small and is rapidly dim inish­ ing. In every one of them , as can be seen from the following figures, the socialized sector has either already em braced the entire econom y o r is pretty close to doing so. Income from the Socialized Sector as Percentage of National Income in 1950 and in 1955, Except Where Other Years Are Specified Soviet Russia

1950 44 0 1828 * 99

Czechoslovakia Poland Hungary Rumania East Germany Bulgaria

\

78.0 70.0 70.0 62.7 55.3 68.6

1937

1955 99.99 92.0 1963 75.0 1953 81.0 1964 71.0 68.6 99.0

* Where a figure relates to a year other than 1950 or 1955, the year is indicated as a super­ script. s o u r c e : Countries of Socialism and Capitalism itt Figures, Gospolitisdat, Moscow, 1957 (in Russian).

L abor incom e is seldom broken dow n in the national income statistics of Com m unist countries by classes of recipients, i.e., as shares received by industrial workers, m em ­ bers of agricultural cooperatives, and m em ­ bers of the m anagerial class. It is impossible, therefore, to m ake any com parisons of the distribution of income not only as between C om m unist and non-C om m unist countries, but even as between one Com m unist country and another.

d. Expenditures of the Income for Dif­ ferent Objects T he expenditure of the national incom e, as noted in C hapter 11, Section l a is in

m any respects m ore conveniently analyzed in term s of the gross national p ro d u ct than o f n et facto r income. In the free, or capital­ ist economies, expenditures are broken down into the following fo u r m ain categories: private consum ption; governm ent consum p­ tion; gross domestic capital form ation; and the residual expenditures com prising net ex­ ports an d other net incom e from abroad. T he expenditures are also often shown by ch aracter of consum er an d p roducer goods being acquired. O ur analysis in th e case of the free econom ies will be lim ited to the first type of breakdow n. In the Com m unist countries the expenditures o f the national incom e, or rath e r of the m aterial product, is broken dow n into only two divisions: con­ sum ption and capital accum ulation. C on­ sum ption is seldom divided into private consum ption an d governm ent consum ption, o r fu rth e r subdivided by products. Of ne­ cessity, therefore, o u r analysis of the ex­ penditure o f the national incom e will be largely confined to the free economies. I t can be seen from the figures in Table 16-6 th a t the p attern of national expendi­ tures bears some relation to th e degree of econom ic developm ent and level of income. In all econom ies, w hether highly advanced o r underdeveloped, th e greater share of the gross n ational p roduct is used fo r private consum ption w ith the balance divided be­ tw een governm ent consum ption and gross capital form ation. B ut the relative shares of private consum ption an d capital fo rm a­ tion v ary greatly as betw een the econom ­ ically advanced high p er capita income countries an d the underdeveloped low per capita incom e countries. In the high per capita incom e countries, the shares of gross capital form ation ten d to be higher and those of private consum ption correspond­ ingly low er th an in the low p er capita in­ com e countries. T he w ealthier countries can afford to save a larger p roportion o f their incom e fo r capital replacem ent and accum u­ lation. This is clearly confirm ed by the figures. I t will be noted th a t the high an d upper m iddle p e r capita incom e countries spend ab o u t 22 p e r cent of their gross national p ro d u ct fo r gross capital form ation and only 64 and 68 per cent fo r private con­ sum ption, w hereas the low er m iddle and low p er capita income countries spend only ab o u t 16 p er cent fo r gross capital fo rm a­ tion and about 73 p er cent fo r private con­ sum ption. These tendencies are often modi-

1 6 . IN TERTEM PO RAL AND INTERNATIONAL COMPARISONS

241

Table 16-6. Expenditure of the Gross N ational Product or N ational Income in 59 Countries by O bjects, in 1955 or N earest Year In Percentages

Countries

Gross D om estic Governm ent Private Capital IConsumption Consumption Formation

N e t Export & N e t Income from Abroad

T otal

H igh per Capita Income Group (Over $700) 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

United States Canada N ew Zealand Australia United Kingdom Switzerland Sweden Luxembourg Belgium Denmark Norway France

6 3 .8 6 3 .6 6 4 .4 6 5 .4 6 7 .0 6 3 .2 6 2 .0 5 8 .3 6 8 .7 6 8 .5 6 1 .4 6 7 .2

1 7 .4 1 4 .0 1 2 .6 1 0.4 17.1 1 2.0 1 6.8 13.1 14.2 13.0 12.4 14.5

18.5 2 4 .9 26.1 2 8 .4 16.4 2 2 .4 2 1 .9 2 4 .9 16.9 17.7 29.1 18.3

.3 - 2 .5 - 3 .1 - 4 .1 - .4 2 .5 - .7 3 .7 .2 .8 - 2 .9

100.0 100.0 100.0 100.1 100.1 100.0 100.0 100.0 100.0 100.0 100.0 100.0

Average for Group I

6 4 .4

1 4.0

22.1

- .5

100.0

Finland W est Germany Netherlands Ireland Austria Venezuela Puerto Rico Argentina Israel

6 1 .6 5 6 .0 5 7 .3 7 8 .8 6 5 .6 6 7.1 8 3 .5 6 6 .9 7 8 .9

12.2 1 5.4 1 4.9 1 4.5 1 2.8 1 8 .6 1 3 .8 13.1 1 8 .6

2 4 .6 26.3 2 5 .0 15.5 2 5 .7 2 6 .4 1 4.6 19.7 2 1 .7

1 .6 2 .6 2 .9 - 8 .8 - 3 .7 - 1 2 .4 - 1 1 .9 .8 - 1 8 .4

100.0 100.3 100.1 100.0 100.4 99 .7 100.0 100.5 100.8

Average for Group II

6 8 .4

14.9

2 2 .2

- 5 .1

100.4

Italy Greece Cyprus Portugal Chile Panama Colombia M exico Guatemala Jamaica Ecuador Turkey M alaya Japan Union of South Africa M auritius

68 .1 7 6 .9 7 1 .6 7 8 .8 8 2 .0 8 0 .6 7 4.5 7 6 .6 7 6 .4 8 5 .9 7 2 .7 7 6 .2 76 .8 6 1 .5 64.2 5 1 .7

1 2.0 1 3.7 12.1 12.3 12.3 13.5 14.3 6 .4 9 .4 8 .8 13.3 11.8 13.5 10.5 12.4 18.5

2 1 .6 15.2 16.3 13.4 6 .7 11.1 12.7 16.7 14.9 1 4.0 15.8 14.6 10.5 2 6 .3 2 6 .2 2 4.1

- 1 .8 - 5 .8 - 4 .4 - .9 - 5 .0 - 1 .4 .2 - .7 - 8 .7 - 2 .0 - 2 .6 - .9 1.7 - 2 .8 5 .4

9 9 .9 100.0 100.0 100.1 100.1 100.2 100.1 9 9 .9 100.0 100.0 9 9 .8 100.0 9 9 .9 100.0 100.0 9 9.7

Average for Group III

7 3 .4

12.2

16.2

- 1 .9

9 9 .9

U pper M iddle per Capita Income Group ($350-$699) 13. 14. 15. 16. 17. 18. 19. 20. 21.

Lower M iddle per Capita Income Group [$150-$349) 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37.

THE IN CO M E OF NATIO NS

242

T able 16-6. Expenditure of the Gross N ational Product or N ational Income in 59 Countries b y Objects, in 1955 or N earest Year ( Continued) In Percentages

Countries

Gross D om estic Capital Private G overnm ent Consumption Consumption Form ation

N e t Export & N e t Income from Abroad

T otal

Low per Capita Income Group (Below $150) Dom inican Republic Brazil C osta Rica Peru Ceylon Burma Philippines Indonesia Taiwan South Korea Rhodesia-Nyasaland Feder­ ation 49. Gold Coast 50. Belgian Congo 51. Nigeria

6 7 .7 7 2 .6 7 5 .2 6 8 .5 7 7 .3 6 3 .8 8 7 .5 8 3 .6 6 6 .3 8 2 .5

12.3 1 0 .4 8 .4 8 .9 1 0 .8 1 4.3 7 .2 16.1 1 8.4 2 0 .5

2 1 .8 17.7 1 6 .0 2 1 .3 12.0 18.9 9 .0 5 .3 2 0 .0 8 .6

3 .0 - 3 .7 - 5 .3 - 4 .7 - 1 1 .6

100.0 100.0 9 9 .9 9 9 .9 100.1 100.0 100.0 9 9 .7 100.0 100.0

6 1 .2 7 2 .4 6 0 .8 8 4 .9

9 .4 9 .4 12.1 3 .0

3 4 .3 1 6 .0 2 9 .6 6 .8

- 4 .4 1 .6 - 2 .5 5 .1

100.5 9 9 .4 100.0 9 9 .8

A verage for Group IV Average for the 51 countries

7 3 .2 7 0 .4

11.5 12.9

1 6 .9 1 8 .8

- 1 .7 - 2 .1

9 9 .9 100.0

38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48.

- 1 .8 - .7 .3 1 .2

Percentages of National Income (M aterial Production Only) Consumption, P rivate and Governm ental

Capital Accum ulation (N e t Investm ent)

Communist Countries Upper and Lower M iddle per Capita Income ($150-$500) 52. 53. 54. 55. 56. 57. 58.

Czechoslovakia Soviet Russia Poland Hungary Rum ania Bulgaria Y ugoslavia

81. 75. 80. 85. 77. 75. 4 6 .8

2 1 .9

19. 25. 20. 15. 23. 25. 3 1 .3

1 00.0 1 00.0 100.0 100.0 100.0 100.0 100.0

20. 2 2 .5

100.0 100.0

Low per Capita Income 59. China Average for Group V

80. 7 7 .5

s o u r c e s : derived from figures given in the sources listed in the Statistical Appendix.

fied, however, in the underdeveloped coun­ tries by governm ental policies encouraging o r forcing a higher rate o f national saving and gross capital form ation than would otherw ise occur. T he proportion of governm ent consum p­ tion in the expenditure of the gross national product tends to be slightly higher in the case of the higher than in the case of the low er per capita incom e countries, b u t this

tendency is n o t com pletely consistent. As show n in T able 16-6, these proportions in 1955 w ere: 14 per cent o f the gross n a­ tional p ro d u ct fo r high p er capita incom e countries, 14.9 per cent fo r th e u p p er m iddle p er capita incom e group, an d 12.2 an d 11.5 p e r cent respectively fo r th e o th er two lower p er capita incom e groups. T h e reasons for this m oderate b ut n o t com pletely consistent correlation betw een the p roportion of gov­

1 6 . IN TER T EM PO R A L AND INTERNATIONAL COMPARISONS

ernm ent consum ption and the size of the per capita income are sim ilar to those re­ sponsible for the n o t com pletely consistent correlation betw een the share of national incom e contributed by governm ent and the size of the per capita national incom e (see Section 2 above). These general tendencies, however, are considerably modified by peculiar political, social, and econom ic conditions in the case o f individual countries. Thus, some u n d er­ developed countries m anage to set aside about as m uch of their m eager incomes fo r gross capital form ation as do the w ealthier countries. T hey do this generally through governm ental action u nder specially form ulated long-range econom ic develop­ m ent plans, and w ith the aid of special tax m easures and, sometim es, foreign loans and foreign technical assistance. In India, in order to achieve a high level of capital fo r­ m ation u nder the official econom ic develop­ m ent plan, the governm ent inaugurated a cam paign in 1957 to reduce the public con­ sum ption of rice and to encourage the substitution o f m ore easily produced cheaper foodstuffs. In som e underdeveloped coun­ tries—Belgian C ongo and the RhodesiaN yasaland Federation, fo r exam ple— the larger proportion of capital form ation comes from the large capital replacem ents and new investm ents m ade by a few large private or public mining and other corporations con­ trolling some of the m ost valuable resources of the country and producing m aterials for export. In some, too, a high rate of capital form ation is m aintained with the aid of in­ flationary deficit financing. V ery significant in this connection is the substantial deficit financing em ployed by a num ber of coun­ tries— Israel, V enezuela, P uerto Rico— in their international accounts, as show n in the next to the last colum n in T able 16-6. On the other hand, some high p er capita in­ com e countries set aside sm aller proportions o f their gross national product o r national incom e fo r both capital replacem ent and accum ulation than do m any low er per capita incom e countries. This was true o f the U nited Kingdom in 1950-55, largely be­ cause the country had lost m uch of its w ealth and incom e during the w ar and its people were hard pressed to m aintain even a sem blance of their accustom ed standard o f living and had little m argin of incom e left fo r saving. Similarly, the proportion of governm ent consum ption in the expenditure of the gross

243

national p ro d u ct does not always follow the general tendency noted. Thus, as shown in T able 16-6, in N orw ay, a high per capita incom e country, th e proportion of govern­ m ent consum ption in the total expenditure is only 12.4 p er cent, while in South K orea, Taiw an, and Indonesia, all low per capita incom e countries, the corresponding pro­ portion is fairly high, ranging from 16 to 20 per cent. These exceptions can be ex­ plained by the fact th at in the case of N o r­ w ay an exceptionally small m ilitary estab­ lishm ent is m aintained, while in the other three cases, exceptionally large m ilitary forces are kept. In the C om m unist countries a m uch larger share of national incom e is devoted to capital accum ulation and a m uch smaller one to consum ption th an in the free econo­ mies. But exact com parisons are difficult because of the different m ethods of social accounting used in the Com m unist coun­ tries, as pointed o ut above. T he figures in Tables 16-6 and 16-7, w hich m ust be exam ined together, show th at from 15 to 25 per cent of the national in­ com e (defined there as including only m a­ terial production) is generally set aside in the C om m unist countries fo r capital accum ula­ tion. On the oth er hand, in the free econ­ omies of com parable incom e levels only 10 to 12 per cent of the national income (which there, however, also includes services) is set aside fo r this purpose. Even in the free economies w ith higher per capita income, as is seen in T able 16-7, the proportion of capital accum ulation averages only 14 per cent of the national income. On the basis of the com prehensive production concept, ac­ cepted by all b ut the C om m unist countries, the proportions of capital accum ulation re­ ported by the C om m unist countries would have to be cut about a sixth to the level of 15 to 21 per cent of the national income. B ut even after such an adjustm ent, the p ro ­ portions of capital accum ulation fo r the C om m unist countries w ould still be con­ siderably higher than for the free economies. T he reasons fo r the m uch higher propor­ tion of capital accum ulation in the Com ­ m unist economies are m ainly to be found in the determ ination o f the Com m unist gov­ ernm ents to industrialize their countries in the quickest possible m anner and to use the m ost pow erful political and economic w eap­ ons at their disposal to achieve th at end. T hese include the heavy taxation and de­ liberate underpricing of producer goods

244

THE INCO M E OF NATIONS

(tending to keep consum ption in check and to m axim ize national saving) and the gov­ ernm ent utilization of a high percentage o f the profits o f state enterprises and co­ operatives for capital accum ulation p u r­ poses.

e. Differences in the Rate of Growth of Income and Their Underlying Causes In all the forty-tw o countries fo r w hich pertinent statistics are available, national in­ com e calculated at constant prices increased during the period 1950-55 at a considerably faster rate th an population (Table 16-7 below ). F o r the thirty-tw o free economies the increase in incom e averaged 5.3 per cent p er year as com pared w ith a population in­ crease averaging only 1.7 p er cent per year, thus exceeding the latter rate 3 to 1. F o r m ost of the W estern countries there was nothing exceptional in this develop­ m ent. T hey had experienced sim ilar gains in incom e over gains in population in the tw o preceding centuries, enabling them to enjoy a steadily rising standard of living. O ne unusual feature of the 1950-55 period was the fact that fo r m any countries, nam ely, fo r the U nited K ingdom , W est G erm any, F rance, Italy, and others, the net gain in incom e represented m ore o f an effort to regain the losses in incom e and standard of living they had sustained during W orld W ar II than an effort to achieve new gains. B ut even fo r this there were the precedents o f like efforts during other postw ar periods to regain the losses in incom e and w ealth sustained during war. A nother feature was the fact th a t fo r som e W estern countries, especially the U nited States, this net gain in incom e over population also represented an attem pt to help other less fortunately situ­ ated free economies to achieve higher stand­ ards of living fo r their people too. But fo r the underdeveloped countries of Asia, L atin A m erica, and A frica this net rise in incom e over population (even though som ew hat lower than the above-quoted aver­ age) was something new and startling. Be­ fore this tim e, as a result of their clinging to ancient m ethods of cultivation and p ro ­ duction, their incom e either rem ained u n ­ changed while their population increased or it barely kept up w ith the population in­ crease. Consequently, their standards o f liv­ ing either sank even farth er below the sus­ tenance level or, at best, rem ained the same. This was the condition th a t characterized the situation in India as late as the 1930’s.

(See R ao’s Indian estim ate in C hapter 30.) T he reversal of this dism al tren d after W orld W ar II was the result of the great social and political aw akening of these countries that started during and im m ediately following the w ar. M ore particularly, it was th e con­ sequence of th e vigorous m easures taken by the newly form ed independent governm ents to introduce and encourage popular educa­ tion, im prove m ethods of production, and develop n atu ral resources. It was also the result o f the financial and technical assist­ ance furnished to these countries through the newly created international organiza­ tions, or, directly, by the econom ically m ore advanced countries. T h e resulting net gain in the p er capita incom e of these peoples represented one of the m ost im portant and hopeful developm ents in the history of m an­ kind. E ven m ore spectacular, in som e respects, w ere the like developm ents in the C om ­ m unist countries during this sam e five-year period. F irst of all, m ost o f them recorded m uch higher rates of increase in their n a­ tional incom es th an did the free o r capitalist econom ies— increases averaging alm ost 12 p er cent p er year. Second, they achieved these increases by vastly different m ethods an d fo r vastly different objectives, using a greater am o u n t of governm ental coercion an d seeking m ore to advance th e socialist type o f organization of society th an merely to raise the standards o f living of th eir own people. T heir increases in incom e thus rep­ resented, in m ore th an one way, challenges to the so-called free o r capitalist world. Increases in real national incom e m ay be produced by any and all of th e following six m ain factors: 1. increases in population, representing additions to the co untry’s lab o r force, m aking possible a m ore effective o r­ ganization o f w ork. T his facto r oper­ ates to increase the incom e o f a coun­ try only w here the resources are abun­ d an t o r can be readily obtained from abroad; 2. increases in the proportion o f the pop­ ulation engaged in production, effected either through the introduction into th e lab o r force of w om en and other segments of the population n ot thereto­ fo re included o r th ro u g h a reduction in b irth rates and the consequent in­ crease in the p roportion of adults in the population capable of engaging in production;

1 6 . IN TER T EM PO R A L AND INTERNATIONAL COMPARISONS

3. capital investm ent, including m echani­ zation o f production increasing the productivity o f labor. These m ay com e either from the savings of the nation o r from foreign loans o r other types of inflow o f capital from abroad; 4. spread o f technical education in fa c­ tories and universities and the increase in the n um ber o f trained specialists, w hich increase the productivity of labor; 5. industrialization and other shifts in the nation’s econom ic activity fro m less productive occupations to m ore pro­ ductive ones. Such shifts m ay be achieved either through the concentra­ tion of capital and em ploym ent of lab or in m ore productive activities by private enterprises w ith the aid of gov­ ernm ent, o r directly and com pletely by governm ental action. T he fact that m anufacturing is m ore productive of incom e than agriculture is well estab­ lished. In the U nited States at this very m om ent, the farm population, com pris­ ing m ore th an 15 p er cent o f the n a­ tio n’s population, produces only a little over 4 p er cent o f the natio n ’s total incom e; 6. m ore intensive utilization o f existing plant and equipm ent through the in­ troduction o f th e three-shift system of w ork, w hich increases th e productivity of labor. O f course, nom inal increases in the na­ tional incom e m ay be produced by inflation and other types o f increases in the general price level. These nom inal increases are recorded in national incom e figures ex­ pressed in curren t prices. T o obtain the real increase it becomes necessary to recal­ culate national incom e figures at constant prices. T he figures in Table 16-7 show in col­ um ns 3 and 7 -1 0 the operation of factors 1 ,3 , and 5 as they affected the real increases in the national incom e (colum n 2) of the various countries during th e period 19 5 0 55. They do not show the roles played by factors 2, 4, and 6. These are indicated, however, in the case of som e countries in the text. T h e increase in real national incom e over population is in m ost cases prim arily the result o f gains in productivity. O ther factors, such as Nos. 2 and 6 above, could n o t have been very im portant in m ost cases during

245

the period in question, and will therefore be ignored in this analysis. T h e gains in produc­ tivity can be roughly m easured by the dif­ ference between the rate of increase in n a­ tional incom e given in colum n 2 and the rate of population increase given in colum n 3. T o m ake the gain in incom e over popula­ tion com parable as between countries, we have expressed the ratios of these two rates of increase as x : l , the 1 being the rate of increase in population and the x the ratio thereto of the rate of increase in income. These ratios are reproduced in colum n 4. T h e effect of the gain in productivity can be expressed fo r the different countries by the difference betw een x and 1, or still better as a percentage of the increase in national incom e. T hus, in the case of the U nited States, fo r example, the im portance of the gain in productivity as against the im por­ tance of the increase in population as a cause of this increase in the national income, is shown by th e relation of the figure 1.4 to the figure 1.0, in the total of 2.4. In other w ords, gains in productivity w ere respon­ sible fo r 58 p er cent o f the increase in the national incom e, while the increase in popu­ lation was responsible fo r only 42 p er cent o f it (assum ing th a t the proportion o f the labor force in th e population rem ained u n ­ changed during th e p e rio d ). F o r the U nited K ingdom , th e ratio o f the tw o rates o f in­ crease being 1 1 :1 , gains in productivity ap­ p ear to have been responsible fo r 90 per cent of the increase in the national income, and the increase in population for only 10 p e r cent of it. Colum ns 5 an d 6 show for each country the percentages o f the increase in national incom e attributable respectively to gains in population and to gains in p ro ­ ductivity. T here are several countries of the free or capitalist w orld in T able 16-7 where, due to special circum stances, the rate of increase in the national income during the period 1950-55 was exceptionally high. It was ex­ ceptionally high in W est G erm any and A ustria because of the rap id rebuilding of these econom ies from the rubble to w hich they h ad been reduced during the war. In Israel, it was exceptionally high because of the trem endous influx during this period of both population and capital from abroad. If the figures fo r these three countries are elim inated from the com putation, the rate of annual increase in national income for the upper m iddle p er capita income group w ould be only 4.7 per cent instead of the 7.2

246

THE IN CO M E OF NATIONS

per cent shown, or slightly below the aver­ age of 5.3 per cent fo r the four groups of the free economies. In the case of the high per capita incom e countries, the annual rate of increase in incom e was nearly four tim es as high as the annual rate of increase in population— 3.9 per cent as against 1 per cent— indicat­ ing th a t only about one fourth of the in­ crease in incom e was due to increases in population and th a t nearly three fourths of it cam e from gains in productivity. T he gains in productivity cam e m ainly from the high rate of capital accum ulation, averaging 14.1 per cent of the national incom e for th a t group and from technological advances generally. N o further shifts in production from agriculture to industry, to w hich any of these gains in productivity could be ascribed, took place in those high per capita incom e countries (see colum ns 8 -1 0 in T able 16-7). Incom e from industry (includ­ ing trad e) constituted 69.1 per cent of their national incom e in 1950; and 69.9 per cent — only a fraction m ore— in 1955. T he pro ­ portion of incom e from agriculture dropped slightly from 11.8 per cent in 1950 to 9.9 per cent in 1955, but this loss was picked up by an alm ost equivalent gain in the pro p o r­ tion of incom e from services, w hich rose from 19.1 p er cent to 20.2 p er cent. The expansion of the industrial sector in these economies had just about reached the m axi­ m um , and w hatever gains in productivity were effected during this period m ust have been due m ore to the shifting from less productive to m ore productive activities w ithin the industrial and agricultural sectors rather than from one to the other. A som ew hat different situation is revealed in the case of the other three income groups of the free or capitalist economies. Thus, the 5.9 per cent rate of annual increase in the national incom e at constant prices re­ corded fo r them fo r the period 1950-55— a higher rate than th a t recorded for the high p er capita incom e group— was only about three tim es as high as the 2.0 per cent rate of annual increase in their population, indi­ cating th a t only two thirds o f the increase in their incom e was due to gains in pro­ ductivity and as m uch as one third was due to increases in population. In other words, gains in productivity in their case played a som ew hat sm aller role in the increases of national incom e than it did in the case of G roup I. T he rate of their net capital ac­ cum ulation, 12.1 p er cent o f national in­

com e, w as som ew hat low er th a n th a t of th e high p er capita incom e group; and in th e case of some of them it was m ade pos­ sible only by substantial loans an d other inflows of capital from abroad. A pparently, capital investm ents in these three p er capita incom e groups were used m ostly fo r the reconstruction o r fu rth er developm ent of in ­ dustry, as can be seen from the fact th at in m ost o f them the proportion o f incom e from industry increased during this period while the proportion of incom e from agriculture declined. In the case of the low er m iddle an d low per capita incom e groups, these shifts in the proportion of incom e derived from industry an d agriculture w ere espe­ cially m arked. Thus, in G roup IV , fo r ex­ am ple, the proportion fo r industry increased from 34.8 to 36.8 per cent, w hile th e p ro ­ portion fo r agriculture declined from 47.4 to 43.5 p er cent. Industrialization was definitely a m ajo r cause of the increase in th e ir in­ com es. T his does n ot rule out, of course, any independent gains being m ade in the productivity of agriculture. In th e Com m unist countries during the period 1950-55, the rate o f increase in n a­ tional incom e was m uch greater th a n the rate of increase in free econom ies. A s shown in T able 16-7 it averaged 12 p e r cent p er y ear fo r th e nine E u ro p ean C om m unist countries as com pared w ith an average pop­ ulation increase of about 2.2 p er cent, thus being m ore th an five tim es as rap id as the increase in population. In the free economies, on the oth er hand, as noted above, the in­ crease in national incom e averaged only 5.3 p er cent p er year and was only about three tim es as rap id as the increase in population. G ains in productivity in the C om m unist countries w ere responsible fo r m ore than 80 p er cent of the increase in the national incom e, as com pared w ith less th an 70 per cent in the case o f the free economies. These rates o f increase in national income, how ever, are n o t altogether com parable, inasm uch as those fo r the Com m unist coun­ tries represent increases in m aterial produc­ tio n only, w hereas the others represent the sam e in both m aterial p roduction and serv­ ices. I t is an established fac t th a t under conditions of rapid industrialization, such as is taking place in the C om m unist coun­ tries today, m aterial production increases m ore rapidly th an services; an d u n d er m ost conditions gains in productivity are greater in m aterial production th an in services. A p a rt o f the m ore rap id increase in the n a­

1 6 . IN TER T EM PO R A L AND INTERNATIONAL COMPARISONS

tional incom e in C om m unist countries is therefore due to th e fact th a t they deal not w ith the whole national incom e b u t only w ith the m ore rapidly increasing portion of it. If com puted on the basis of the w hole national incom e inclusive of services, the rate of increase of the national incom e in their case would appear som ew hat m ore m oderate. H ow ever, even after such a dow n­ w ard readjustm ent, the rate of increase in the C om m unist countries w ould still appear m uch m ore rapid th an in the free economies. T he extraordinarily rapid increase in n a­ tional incom e in the Com m unist countries is due prim arily to the determ ined policy o f the C om m unist governm ents, as already stated, to industrialize their m ostly back­ w ard agricultural econom ies in the quickest possible m anner. T he econom ically ad­ vanced free econom ies experienced this extraordinarily rap id rate of increase in their national incom e years ago w hen they w ere passing through their industrial revolu­ tion. But the rate of increase in their n a­ tional incom e during th a t period could hardly have reached the heights now being recorded in the C om m unist countries, fo r the reason that their earlier industrial revo­ lution was not quite as concentrated. T he m ost potent instrum ents used by the Com m unist governm ents to effect this rapid industrialization bear restating. T he high proportion of national incom e set aside in these countries for capital accum ulation is the num ber one factor. T he second m ost im ­ p o rtan t factor is the channeling by the C om ­ m unist governm ents o f the newly form ed capital into industry rath er th an agriculture, and especially into the producer goods in­ dustries upon w hich the consum er goods industries depend, and in w hich and through w hich th e greatest gains in productivity can be achieved. These tw o factors taken to­ gether have w rought profound changes in the industrial structures of the C om m unist economies, as exemplified by the figures for the period 1950-55 (see T able 16-7). In every one of these countries the pro p o r­ tion of incom e from industry increased m aterially while the proportion o f incom e from agriculture declined correspondingly. M oreover, in the Com m unist countries im portant shifts took place in the internal structure of industry, w hich could n o t be shown in Table 16-7: in m any of them these shifts are dem onstrated by a rise in the p roportion of income derived from producer goods industries. This m ay be illustrated by

247

the following figures relative to the progress of industrialization in C om m unist China. A c­ cording to the official statistics of th at coun­ try, the p roportion of the gross value of in­ dustrial o u tput in the total gross value of production increased from 23.2 per cent in 1949 to 45.6 p er cent in 1956; the p ropor­ tion of the gross value of h andicraft produc­ tion increased during the sam e period from 6.1 to 9.1 per cent of th e gross value of total industrial production. On th e other hand, the proportion of the gross value of agricultural production declined during those years from 69.9 to 45.3 p er cent of the total gross value o f production. T he proportion of the gross value of the p roduc­ tion of producer goods in the total industrial production increased from 28.8 per cent in 1949 to 49.7 p er cent in 1956, while the proportion fo r consum er goods production declined from 71.2 per cent to 50.3 per cent. T he proportion of the gross value of the production of m achines in the total in­ dustrial production of C hina increased from 1.7 p er cent in 1949 to 9.8 p er cent in 1956.16 (F o r a fuller analysis of the rapid grow th of C hina’s national income under C om m unism , see E conom ic Survey o f A sia and the Far E ast, 1957, Chap. 4.) M easuring the relative contributions of agriculture and industry in term s of their respective gross values of output tends to exaggerate the contribution of industry and understate th a t of agriculture, because the gross value figures fo r industry include ex­ tensive duplications representing the costs of raw m aterials obtained from agriculture and the cost o f other interm ediate goods obtained from other branches of industry. T he gross value of agricultural production does n o t include duplications to quite the same degree. T he use o f gross values in this w ay is doubly m isleading u n d er the C om m unist type of social accounting. It exaggerates the contribution of industry and understates th at of agriculture for the addi­ tional reason th at nearly all of the tu rn ­ over tax is credited to the value of in­ dustrial production, even though its pro­ ceeds are used to advance agricultural p ro ­ duction as well and should be charged to it in due proportion. This type of m easurem ent also tends to exaggerate within the industrial sector itself the contribution m ade by the consum er goods industries as against that m ade by the p roducer goods industries, be­ cause the C om m unist system of taxation concentrates the turnover tax on the con-

Table 16-7. R ate of Annual Increase of N ational Incom e in 42 Countries and Som e of Its U nderlying Causes Per C en t A nnual Increase in

C ountries

G ain in R a te of Increase o f N a t. Incom e over P op u lation

Per C en t of Increase in N a t. Incom e D u e to

Period

N atio n a l Incom e at C on sta n t P rices *

1950-55 1950-55 1950-55 1954-55 1950-55 1950-55 1950-54 1950-55 1950-55 1950-55

4 .4 4 .8 3 .3 3 .9 3 .8 6 .2 2 .8 2 .0 3 .8 4 .4

1 .8 2 .7 0 .3 1 .0 0 .8 0 .8 0 .5 0 .8 1 .0 0 .7

2 .4 : 1 .0 1 .8 : 1 .0 1 1 .0 : 1 .0 3 .9 : 1 .0 4 .8 : 1 .0 7 .7 : 1 .0 5 .6 : 1 .0 2 .5 : 1 .0 3 .8 : 1 .0 6 .3 : 1 .0

42 56 10 26 21 14 18 40 26 16

58 44 90 74 79 86 82 60 74 84

3 .9

1 .0

3 .9 : 1 .0

26

74

18 8 23 102 101 53 166

82 92 77 -2 -1 47 -6 6

Popula tio n *

P op u ­ la tio n Increase

G ains Productiv ity

Per C ent o f N a t. Incom e U sed for C apital A ccum u­ lation in 1954-55

Per C en t o f N a t. Incom e C ontrib uted b y E ach E conom ic Sector in 1950 and 1955 t A griculture 1950 1955

Ind u stry (and Trade) 1950 1955

Services 1950 1955

I. High per Capita Income Group (Over $700) 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

U n ited S tates C anada U n ited K ingdom Sw itzerland Sw eden Luxem bourg Belgium D en m ark N orw ay France

248

A verage for Group I

1 1 .4 1 6 .3 9 .9

6 .4 1 3 .9 6 .0

4 .3 1 0 .0 4 .6

7 3 .0 6 7 .0 7 0 .1

7 3 .5 6 8 .0 7 2 .8

2 0 .4 1 9 .1 2 3 .9

2 2 .2 2 2 .0 2 2 .6

1 8 .3 8 .5 1953 1 4 .5 2 0 .0

1 1 .8 8 .7 2 1 .3 1 4 .7

9 .4 8 .2 1 9 .5 1 3 .1

7 .1 8 6 7 .9 6 2 .9 7 0 .6

7 3 .8 6 6 .6 6 2 .9 7 1 .8

1 6 .4 2 3 .4 1 5 .8 1 4 .7

1 6 .9 2 5 .2 1 7 .6 1 5 .2

1 4 .1

1 1 .8

9 .9

6 9 .1

6 9 .9

1 9 .1

2 0 .2

II. Upper M iddle per Capita Income

Group ($350-$699) 11. 12. 13. 14. 15. 16. 17.

Finland W est G erm any N etherlands Ireland A ustria Israel A rgentina

1950-55 1950-55 1950-55 1950-55 1950-55 1950-54 1950-55

A verage for Group II

5 .5 : 1 .0 1 1 .9 : 1 .0 4 .3 : 1 .0 1 . 8 : —0 .,4 8 . 0 : —0 . 02 1 .9 :1 .0 0 .6 : 1 .0

6 .0 1 1 .9 5 .1 1 .8 8 .0 1 6 .5 1 .4

1 .1 1 .1 1 .2 - 0 .4 - 0 .0 2 8 .6 2 .3

7 .2

2 .0

3 .6 : 1 .0

28

72

7 .2 7 .2 2 .6 7 .6 2 .3 2 .3 4 .2 9 .0

0 .6 1 .1 0 .7 3 .7 2 .3 2 .3 3 .2 1 .4

1 2 .0 : 1 .0 6 .5 : 1 .0 3 .7 : 1 .0 2 .0 : 1 .0 1 .0 :1 .0 1 .0 : 1 .0 1 .3 : 1 .0 6 .4 : 1 .0

8 15 27 50 100 100 77 16

92 85 73 50 0 0 23 84

5 .3

1 .9

2 .8 : 1 .0

36

64

2 4 .7 1 2 .3 1 3 .1 3 2 .8 1 6 .3 1 1 .9 *9®2 1 5 .6

2 3 .8 1 0 .4 1 0 .9 3 3 .6 1 4 .2 1 3 .2 1 8 .0

6 0 .8 6 0 .6 6 9 .7 7 2 .9 6 7 .2 6 4 .9 4 0 .1 4 0 .6 6 8 .4 6 5 .5 4 7 .6 1952 4 9 .9 6 3 .4 6 7 .1

1 4 .5 1 5 .6 1 8 .0 1 6 .6 2 2 .1 2 1 .9 2 7 .1 2 5 .8 1 8 .2 1 7 .5 4 0 . 5 1952 3 6 .9 1 7 .3 1 8 .6

1 0 .2

1 8 .1

1 7 .7

5 9 .4

6 0 .4

2 2 .5

2 1 .8

1 4 .9 6 .8

-2 .5

3 2 .4 3 3 .7 2 8 .3 4 8 .9 1 5 .0

2 4 .0 3 5 .7 2 7 .3 4 1 .9 1 5 .0

5 2 .4 4 8 .1 5 4 .9 3 3 .9 5 2 .2

5 9 .7 4 5 .3 5 4 .5 4 1 .9 5 2 .5

1 5 .2 1 8 .2 1 6 .8 1 7 .2 3 2 .8

1 6 .3 1 9 .0 1 8 .2 16.1 3 2 .5

' 4 .8 2 4 .0

3 0 .6 2 6 .2

2 6 .5 1952 2 1 .8

4 4 .9 5 5 .8

4 9 .3 1952 5 6 .2

2 4 .5 1 8.1

2 4 .2 2 1 .9

9 .6

3 0 .7

2 9 .5

4 8 .9

5 1 .4

2 0 .4

2 1 .2

1 4 .9 2 4 .2 4 .2 1 8 .4 - 1 0 . 6 1954

I II . Lower Middle per Capita Income

Group ($150-$349) 18. 19. 20. 21. 22. 23. 24. 25.

Ita ly Greece Portugal T u rk ey Chile G uatem ala Jam aica Japan A verage for Group III

1950-55 1950-55 1950-55 1950-55 1950-55 1950-55 1950-55 1950-55

IV. Low per Capita Income Group {Below $150) 26. 27. 28. 29. 30. 31. 32.

Brazil H onduras T hailand P ak istan India R hodesia-N yasaland Fed. Belgian Congo

1950-55 1950-52 1950-54 1950-54 1950-54 1950-55 1950-55

Average for G roup IV Average for G roups I- IV

7.0 3.1

2.5 3 .0 1.9

6 .1

1 .8

2 .0

3.7 7.7 6 .3

1.3 2 .7

5.1 5 .3

2 .2

1 .8

1.7

.4 :1 .0 .3 :1 .0 1 .6 :1 .0 0 .9 :1 .0 2 .8 :1 .0 2 .8 :1 .0 3 .5 :1 .0 2

2

.3 :1 .0 3 .1 :1 .0 2

42 44 63 111

58 56 37

10.5

11

35 35 29

65 65 71

26.3 21.9

43 32

57

19.6

68

1 2 .8

Per C ent A nnual Increase in

249

33. 34. 35. 36. 37. 38. 39.

Czechoslovakia Soviet Russia Poland H ungary R um ania E a s t G erm any Bulgaria

1950-55 1950-55 1950-55 1950-55 1950-55 1950-55 1948-52

1 0 .0

1 .1

13.0

1.7 2 .4

1 1.0

8.5 14.0

0 .8 2 .2

1 2 .0

1 .1

13.8

2 .0

29.0 5 6 .0 57.2 60.2 51.2

31.6 5 4 .1 1952 44.2 W54 57.5 1954 46.3 1954

46.2 27.5 30.6 19.8 33.7

44.1 28.0 1952 36. 7 1954 21.7 1954 37.0 1954

31.2

27.6

48.7

47.4 26.3

43.5 2 4.0

34.8 54.4

Agriculture 9 .0 :1 .0 7 .5 :1 .0 4 .5 :1 .0 1 0 .0 : 1 .0 6 .4 :1 .0 1 1 .0 :1 .0 2 .0 :1 .0

11

13 22 10

16 9 50

89 87 78 90 84 91 50

2 2 .0

2 5.0 25.0 25.0 23.0

4 4 .0 1948 2 8 .6 1937 38.2 1947 7 1 .0 1949

1952-55 1952-55

13.0 13.5

1952-55

1 2 .0

1 .6

1948

1 2 .0

25.0

4 9 .3

8

.0 : 1

.0

13

87

gross

Average for G roup V

76.8 gross

5 9 .5

gross

48.0

39.5

19.0 21.7

1949

11.9

2 .2

5 .3 :1 .0

18

82

gross gross

2 8 .4 ?r08S

56.0 71.4 61.8 29.0

IJ

9 1952

1 2 .2

19; ,^5 4

2 0 .1

2 0 . 8 1954

15.1

16. 7 1954

52.1

17.5

20.3

36.8 55.7

17.7

19.7 20.7

1948 1937 1947

2 0 .0

1949

65.0 74.1 70.2 3 8 .4 1953 6 6 .0

1948

1952

3 3 .3 3 1 7

19.7

B. Low per Capita Income Group (Below $150) 42. C hina

24.4

Ind u stry

35.0 25.9 29.8 6 1 .6 1953 34.0

1952

40. Yugoslavia 41. A lbania

24.8 16.5

P er C ent of N ational Income D erived from E ach of the Two Sectors of M aterial Production in 1950 and 1955 f

N ational Income Popu(M aterial production) * lation *

Communist Countries A. Upper and Lower Middle per Capita Income Groups (.$150-$S00)

1962

5 0

. 7gross

1952

6 6

. 7 gross

6 8

. 3 gross

1952

7 1

. 6gross

1949

23.2 gross 40. 5 gross 52.0

60.5

* The annual rates of increase are obtained by dividing th e per cent increase for th e period as a whole by the num ber of years in th e period; th e y are therefore slightly higher th a n actually realized, t W here a figure relates to an o th er year, such a year is indicated as a superscript. s o u r c e s : D erived from figures given in th e U nited N ations, N ational Income and Expenditure, Series H , No. 10, New Y ork, 1957, and from official docum ents of individual countries (see S ta tis­ tical A ppendix).

250

THE IN C O M E OF NATIONS

sum er goods industries. M easuring the con­ tributions of industry and agriculture in term s of the net values of their output w ould reduce both exaggerations and understate­ m ents, but it w ould not eliminate them al­ together because the bulk of the turnover tax is included in the net output of industry. A reader who is unaw are of the exaggera­ tions inherent in the gross value figures m ight be m isled by the figures fo r C hina showing an increase in the gross output of industry from 23.2 per cent of the total gross value of m aterial production in 1949 to 45.6 p er cent in 1956. H e m ight interpret them as m eaning th a t C hina’s econom y was alm ost one fourth industrialized in 1949 and th a t in a period of seven years it had becom e alm ost one half industrialized, which, of course, w ould be untrue. A lthough the shift from agricultural to industrial production has been an im portant factor in the rapid grow th of national in­ com e in C om m unist countries, in some of them it is already of dim inishing im portance. F irst of all, at some point a balance is reached between the proportions of these tw o branches of production that repre­ sents an optim um fo r the econom y under given conditions. A further developm ent of industry at the expense of agriculture m ay hu rt the economy. In fact, to obtain further substantial increases in national incom e, a reversal of the nation’s capital investm ent m ay be necessary, involving the shift of some of the investments from industry to agriculture and the expansion of agricultural activities at the expense of the industrial. Thus, it is significant th a t in Soviet Russia, under the sixth five-year econom ic plan (1 9 5 6 -6 0 ), the proportion of incom e from agriculture in the total national incom e is expected to rise in consequence of the m easures projected by the Soviet govern­ m ent to m aterially increase agricultural p ro ­ duction as a p a rt of the new policy to in­ crease consum ption and to decelerate some­ w hat the developm ent of heavy industry (see Section 5b above). I t is impossible to enum erate here all the factors th a t are responsible fo r the great increase in productivity and, hence, in the national incom e of the C om m unist coun­ tries. Suffice it to state th a t the Soviet econ­ omists attribute 80 p er cent o f the rise in Soviet national incom e to increased produc­ tivity and only 20 per cent to the increase in the num ber of workers.

T he U nited N ations E conom ic Survey o f E urope in 1955 (p. 200) m akes the follow­ ing illum inating observations regarding the effects of the high rate of capital accum ula­ tion on the rate of grow th of national in­ com e an d o u tp u t in Com m unist countries: Broadly speaking it would seem that the 25 per cent accumulation rate which has ruled in all the [European Communist countries] except Eastern Germany has resulted in the annual rate of growth of 8 to 11 per cent [in the na­ tional income] and that the capital output ratios have been between 2% and 3—in other words, that investment raises national income by 33 to 45 per cent of the amount invested. In Eastern Germany, however, the capital output ratio has undoubtedly been substantially lower —more like 1%. New capital output ratios are rather low by comparison with what is usually taken as normal in Western Europe. In part this is due to a more intensive utilization of capital in the Soviet Union and Eastern Europe (e.g., a more widespread adoption of three shift working, and shorter cold times for fur­ naces and repair times for machinery). T he th ird im portant facto r in th e rapid increase in national incom e in th e Com ­ m unist econom ies has been the considerable rise in the proportion o f gainfully em ployed persons in the total population. This has been the result of introducing into the labor force elem ents o f the population th a t were n o t previously a p art of it, chiefly women. A ccording to Soviet statistics, the proportion of w om en am ong wage earners an d salaried em ployees in Soviet R ussia increased from 27 per cent in 1929 to 38 p er cent in 1940 an d to 45 p er cent in 1955.17 This sam e high rate of accum ulation in C om m unist countries, how ever, m ay not continue fo r long. T heir econom ic plans for 1 9 5 6 -6 0 provide for a low er rate in order to im prove the consum ers’ lot (see E co­ n om ic Survey o f Europe in 1957, C hap. 1 ).

f. Summary W e have m ade com parisons of only a few o f th e m any aspects of the incom es of various countries and of the changes oc­ curring in them . These com parisons are sufficiently illustrative, how ever, o f the vast­ ness o f the field of international com pari­ son an d of the fruitfulness o f explorations th a t m ay be conducted in it. T he limits to these explorations are set only by the availability of reliable statistics and by the com petence an d im agination o f the analysts attem pting such surveys.18

17. ESTIMATING METHODS

1. Main Characteristics of the Three M ethods

only up to a certain point. Eventually he m ust resort to the income-disposed or final products m ethod. In estim ating expenditures fo r professional and other services he can use data on incom es received by those ren­ dering such services, but he m ust add to their estim ated n et earnings their outlays fo r m aterials, rent, and other producer goods and services. In estimating expendi­ tures fo r m anufactured goods he can use production data (the gross value of m anu­ facturing o u tp u t). H ow ever, to avoid du­ plication he m ust deduct the value of the interm ediate goods and services produced w ithin the industry itself. H e m ust then trace the values of the m anufactured goods as they pass through trade and transporta­ tion until they em erge as consum ption and investm ent goods. In each case he m ust add the trade m arkups o r other charges involved; elim inate the duplications of sales between businesses operating at the same stage of th e process; and m ake deductions o r addi­ tions fo r the values of the goods furnished to o r from abroad. P roduction data can be used in sim ilar fashion in estim ating ex­ penditures for nonm anufactured goods. In either case the estim ator m ust m ake a great m any m ajor adjustm ents, characteristic of the expenditure m ethod. Only w hen final sales data fo r consum er and investm ent goods are available can the estim ator meas­ ure the disposition of national p roduct di­ rectly w ithout recourse to production or incom e data.

D epending upon w hich phase o f national incom e is to be analyzed—its production, distribution, or expenditure— different data an d m ethods are required. F o r the first phase, production, the net outp u t m ethod is generally used; fo r distribution, the incom edistributed m ethod; and fo r expenditure the incom e-disposed o r final products m ethod is used. Seldom, how ever, is an estim ator con­ cerned w ith only one phase o r does he use only one m ethod, although the m an n er of com bining m ethods varies considerably from estim ate to estim ate.

a. Functional Interdependence of Methods T he way in w hich m ethods are com bined depends as m uch on the objective of the estim ator as on the statistics he has to w ork with. If the estim ator wishes to determ ine and analyze the net output o r value added o f th e various divisions of econom ic ac­ tivity, he can use either the net output or the incom e-distributed m ethod, depending upon w hich type o f data is m ore readily available— data on production o r data on income paym ents and receipts. T he incomedisposed m ethod is hardly applicable in this case as it w ould require calculating in re­ verse order from final product to com po­ nents, and then necessitate reassem bling the com ponents or “values added” by industrial divisions. In the course of such calculations, the net output m ethod would have to be used anyway, and the procedure would be m ade uselessly com plicated. W hen the estim ator’s objective is to de­ scribe the distribution o f national incom e by type of incom e, he m ust use the incomedistributed m ethod. If the estim ator’s objective is to present the disposition of national incom e by type o f final product, he can use the incomedistributed and net o utput m ethods, but

b. Choice of Methods for Different Divisions of Economic Activity T h e m ost com m on breakdow n of na­ tional incom e is by divisions of economic activity. N e t ou tp u t estim ates are autom ati­ cally broken dow n this way, but it is also used in estimates prepared by the incomedistributed and final expenditure m ethods because it points up the fundam ental struc­ ture of the econom y as well as the m anner in w hich national income is created. A nother

N o tes begin on p. 535. 251

252

THE IN CO M E OF NATIO NS

reason fo r its popularity is that the relevant statistical data are m ore generally organ­ ized by divisions of econom ic activity than by any other classification. W e, therefore, begin our description of m ethods w ith a survey of their use in estimating the in­ comes originating in the various divisions of econom ic activity. F o r incom e from agriculture, net value of output is the desired aggregate. T he net output m ethod is used in order to determ ine it because agricultural products are staples sold at standard m arket prices, and are m ade to a large extent from raw m aterials, such as seed, feed, and fertilizers, w hich originate in agriculture itself. Gross output and deductible costs can be ascertained largely from the same data. On the other hand, the factor shares in agriculture are not clearly differentiated, except w here agri­ culture is carried on by corporate enter­ prise or by co-operatives. C urrent agricul­ tural output and raw m aterials used in agri­ cultural production as well as prices are available for m ost countries fo r at least the m ain crops, but data on com ponents of agri­ cultural incom e are generally available at best only fo r small samples of farm ers’ accounts. In forestry and fishing, too, the net value of output is usually the goal, and the net output m ethod, the estim ating tool. F o r income from m ining and m anufac­ turing (including gas, pow er plants, and construction), both the incom e-distributed and net output m ethods can be used. T he first is preferable because the shares allo­ cated to the factors of production in these divisions are predom inantly m oney incom es and are recorded in business accounts and reported to tax authorities and social se­ curity agencies. In m any countries, produc­ tion censuses in these fields show the gross value of output, but they ordinarily om it small establishm ents and trades, necessi­ tating the use of the incom e-distributed m ethod, at least in their case. Similarly, the incom e-distributed and net output m ethods are alternatively used in m easuring th e contributions of trade and com m unications to national incom e. F o r m any countries, data on incom es paid out in trade are incom plete o r nonexistent, m ak­ ing the use of the net output m ethod pref­ erable. H ow ever, it m ust be applied in a special w ay because data on gross sales in trade are seldom available. T he published operational accounts of large scale transpor­

tation and com m unication enterprises (rail­ roads, telephone, telegraph, and the lik e), fo r alm ost all countries, yield figures on their gross receipts, expenditures, and “value added,” as well as on wages, debt, interest, an d n et profits. T he contributions of the services o f dwell­ ings o r oth er buildings are usually m easured by the net o u tput m ethod. In finance (banking an d insu ran ce), w here large units are th e rule, it is not difficult to obtain data on incom es paid to production factors. H ow ever, th e determ i­ nation of the tru e value added presents an intricate problem (see C hapter 13, Sec­ tio n 3 ). F o r professional, dom estic, and other services, the incom e-distributed m ethod is used exclusively. T he net o u tp u t m ethod w ould n o t be satisfactory because in their case there is either little difference between gross and net output o r it is easy to ascer­ tain directly the paym ents to production factors th at consist chiefly of labor incomes or of mixed labor-capital incomes. T he above is also tru e in the case of gov­ ernm ent services, but in a few estim ates the n et o u tput m ethod has been applied by de­ ducting from the reported gross expenditure of governm ent its purchases of goods and services from business, tran sfer paym ents, and duplications.

c. Dependence of Method on the Struc­ ture of the Economy T he availability of data, w hich is a m ost im portant facto r in determ ining the esti­ m ating m ethod, depends, in large part, u p o n the econom ic structure of the country. In agricultural countries personal incom e taxes, social security systems, and corpo­ rate organization of business are rare. C on­ sequently, the data on individual incomes essential to the incom e-distributed m ethod are either nonexistent or very m eager. M any agricultural countries, in addition to having d ata on th eir agricultural output, either take cu rren t o r periodic censuses of o u tput of th eir few branches of industrial production o r at least possess scattered statistics for some of them . T hey thus have some of the requisite data fo r the larger p o rtion of the operations of their econom y. U n d er these circum stances, estim ators fo r predom inantly agricultural countries use the net output m ethod. In industrial countries th e situation is m ore com plex. W ith elaborate fiscal systems

1 7 . ESTIM ATING M ETHODS

and records, m aterial on individual incom es is often rich, m aking possible the applica­ tion of the incom e-distributed m ethod. H ow ­ ever, as data on production, sales, consum p­ tion, and investm ent are usually to be had as well, the other tw o m ethods could also be used. B ut the la tte r types o f data are generally n o t as com plete or up to date as fiscal data. E ven such industrial countries as the U nited K ingdom and G erm any have only sporadic m anufacturing censuses. Only a few have com prehensive current statistics on retail sales and consum ption expendi­ tures. T herefore, in industrial countries, the basic estimates are often m ade by the in­ com e-distributed m ethod.

d. Statistical Short Cuts T he generally inferior character o f statis­ tics in agricultural countries renders the task of estim ating national incom e fo r these countries m ore difficult th an fo r industrial countries, and yields less detailed and, often, less reliable results. A n even m ore im por­ tant source of difficulty than the inadequacy of statistics in agricultural countries is the large proportion o f econom ic activity carried on in the household th a t escapes the price m easurem ent o f the m arket. F o r example, in ru ral Ceylon about 50 p er cent o f local produce is consum ed by the households themselves.1 Because o f these difficulties estim ators w orking w ith agricultural coun­ tries resort to short cuts inherited from ear­ lier and cruder stages of national incom e estimating. T he incom es o f wage earners, em ployers, and the self-employed outside agriculture are determ ined by m ultiplying the num ber of recipients in each group by an average incom e derived from occupa­ tional censuses and investigations of wage rates. T his procedure, in a som ew hat m ore re ­ fined form , plays a subsidiary role in m ost estimates for industrial countries. F o r ex­ ample, in the U nited States estim ates, total cash pay of dom estic servants is obtained by multiplying the estim ated num bers of such servants by appropriate average annual earnings. The sam e m ethod is used in esti­ m ating the salaries of ministers and teachers in private colleges and schools, incom es of insurance solicitors, and salaries and wages of certain other groups not covered by social security, such as agricultural, forestry, and fishery workers. T he som ew hat crude esti­ m ates fo r these groups, how ever, do not play an im portant role in the total estim ate

253

o f lab o r incom e in the U nited States, inas­ m uch as over 80 per cent of the payrolls there are covered by social security, and at least an additional 10 p e r cent are covered by reliable business accounting records. The total n et incom e of professional practi­ tioners is estim ated by m ultiplying their num ber, as reported in censuses and other reports, by an average n et income. The sam e m ethod is also used in the U nited States to estim ate total incom e of business­ m en operating w ithout incorporating, net rents of ow ners of n o nfarm dwellings, and total incom e from certain types of interest. A nother short cut, used in m ost countries fo r developing a com plete or partial series of annual incom e estimates, consists of tak ­ ing an elaborate estim ate m ade fo r a base year fo r w hich census data are available, and projecting it backw ard by m eans of a com bination of statistical indicators, am ong w hich indexes of production, prices, em ­ ploym ent, and wage rates are the m ost im ­ portant. T his m ethod is fully described in Section 3 below.

e. Classification of Estimates by Method It is difficult to classify estimates by m ethod inasm uch as m ost of them, as al­ ready indicated, are prepared by a com bi­ nation of m ethods. A ny classification m ust necessarily be m ade according to which m ethod seems to be relied upon most; but this predom inance is n o t always clearly established. T he criteria fo r the classifica­ tion of estim ates by m ethod adopted in this book are: (a ) estimates are classified as net output if the net ou tp u t m ethod is used to determ ine at least the incom e from agricul­ ture, forestry, m ining, and large scale m an u ­ facturing; (6 ) estimates are classified as in­ com e distributed whenever th at m ethod is used to determ ine incom e fo r m ost non­ agricultural production; (c) estimates are called incom e disposed or final expenditure estim ates w henever the final value o f p ro d ­ ucts available fo r consum ption and invest­ m ent is the desired aggregate, even though production and incom e data may be used to develop it. Even w hen the income-disposed o r final expenditure m ethod is hardly dis­ tinguishable from the n et output m ethod, it can be recognized by checking to see if exports are deducted from and imports are added to dom estic output. Since there are m ore predom inantly agri­ cultural th an industrial countries, and since some of the latter prepare n et output esti­

254

THE IN CO M E OF NATIONS

m ates as well as incom e-distributed and income-disposed estimates, there are m ore net output estimates than either of the other two. This num erical superiority is likely to grow less pronounced in tim e as m ore agri­ cultural countries becom e industrialized and as an increasing num ber of countries m od­ ernize their fiscal and business accounting systems and develop a m ore active interest in the distributive and the dispositional or product aspects of their national income. A n ever-growing num ber of countries (65 of them , as shown below in T able 17-1) are preparing estimates today that sim ultane­ ously cover two or all three phases of n a­ tional incom e and use two o r all three esti­ m ating methods. In these estimates, some related items appearing in different phases are built up independently from different data and by different methods, while others are based on the same data and m ethods w ith only such adjustm ents as m ay be neces­ sary to fit them to the given phase. F o r example, in the C anadian estimates incomes in cash and in kind o f dom estic servants, estim ated by the incom e-distributed m ethod, are carried over bodily into the estimates of national expenditure as the expenditures of households fo r dom estic services; and are assumed to be equal in both cases. On the other hand, the net incom es o f certain p ro ­ fessions obtained by the incom e-distributed m ethod are adjusted to a gross basis before being carried over to the expenditure side of the national incom e as the expenditures o f households fo r such services.

O ne of the m ost valuable by-products of preparing parallel estimates is the final com ­ parison and reconciliation o f their results (see, fo r exam ple, C hapter 27 on F rance, Section 1-4).

2. The Processes for A chieving A ccuracy Under Each M ethod a. Coverage and Evaluation T he goal of every national incom e esti­ m ator, w hatever his m ethod, is to ascertain the tru e national incom e o r p ro d u ct in term s o f his concept, y et the best estim ate is no m ore than an approxim ation. Im per­ fections are sometim es due to a lack of th o r­ oughness on the p art of the estim ator, but m ore often they are due to gaps and inac­ curacies in the data. N ational incom e estim ates are built up from a large num ber of item s, classified, according to the phase o f national incom e analyzed, as net outputs of various sectors o f econom ic activity; types of distributive shares (o r incom es of various types of p ro ­ d u c e r); o r final products of various kinds. Basically, th e value aggregates are th e n u ­ m erical products of two statistical dim en­ sions— quantities of com ponent units and th e values attached to each. D epending upon the phase of national incom e analyzed, the quantitative units are: enterprises, govern­ m ents, or households contributing to the net o u tp u t o r paying incomes to factors of p ro ­ duction; individuals receiving certain types o r sizes of incom e; or final products of vari-

Table 17-1. Phases of the N ational Incom e Covered in 87 Country E stim ates in 1957 A. 22 estim ates covering only one phase: a. Production: 14 estim ates— Soviet Russia, E ast Germany, Hungary, Rumania, Nicaragua, E l Sal­ vador, H aiti, Surinam, Paraguay, Pakistan, Syria, E gypt, Lebanon, China. b. Distribution: Cuba, French W est Africa, Uganda. c. Expenditure: Iceland, Dom inican Republic, Ceylon, Algeria, Morocco. B. 28 estim ates covering tw o phases (20 production, 15 distribution, and 21 expenditure): Australia, Union of South Africa, France, Switzerland, Sweden, Italy, Spain, Portugal, Poland, Czechoslovakia, Y ugoslavia, Bulgaria, Jam aica, G uatem ala, British Honduras, C osta Rica, V ene­ zuela, Bolivia, Uruguay, Turkey, India, Indonesia, V iet-N am , M alaya, South Korea, Tunisia, K enya, Rhodesia-Nyasaland. C. 37 estim ates covering all three phases: U nited States, Canada, U nited Kingdom, Ireland, N ew Zealand, W est Germany, Austria, Belgium , Netherlands, Luxembourg, Denmark, Norway, Finland, Greece, Cyprus, Puerto Rico, M exico, H onduras, Panama, British Guiana, Colombia, Ecuador, Brazil, Argentina, Peru, Chile, Israel, Burma, Thailand, Taiwan, Japan, H ong K ong, Philippines, Nigeria, Belgian Congo, M auritius, Gold Coast. Sum m ary: the 87 country estim ates contained 189 incom e aggregates: 71 relating to the dom estic output, 55 to the distribution of income, and 63 to the expenditure of the gross national product. s o u r c e : derived from the sources listed in the Statistical Appendix.

1 7 . ESTIM ATING M ETHODS

ous kinds. Correspondingly, the u nit values m ay be: the net outputs of individual enter­ prises, governm ents, o r households; o r their incom e paym ents or the incom e receipts of individual participants in production; o r the prices of individual final products. T o be reliable estim ates m ust be able to m eet two tests: (1 ) the test o f com plete coverage of all units com prising th e national p roduct or incom e, and (2 ) the test o f ac­ curate evaluation of each com ponent unit or group of units. M ore estimates suffer fro m undercoverage and undervaluation th an fro m th e opposite faults, partly because some u nit quantities and their values are im possible to m easure and are deliberately om itted o r otherw ise escape attention, and partly because esti­ m ators tend to be conservative in estim ating u nreported item s or adjusting existing data. This observation is supported by the fact th at where older national incom e estimates are revised on the basis of new er and m ore com plete statistical data and techniques, n a­ tional incom e totals are generally raised. E xam ples of such upw ard revisions include the 31 per cent increase m ade in de Bernonville’s 1938 F rench estim ate by Sauvy and Rivet, and a sim ilarly large increase m ade in Sauvy and Rivet’s estim ate by the M inistry of F inance in 1950-53 (see C hap­ te r 10, Section 3 a ). M ost overvaluations of national incom e are the result o f political considerations, patriotic sentim ent, o r even the estim ator’s own expansive personality th a t interferes w ith his objectivity. Sir W illiam Petty was accused by F rench econom ists (Boisguillebert, D utot, and others) of deliberately over­ stating the national incom e of Britain and understating that of F rance. T he Soviet esti­ m ates, according to the analysis m ade by British, A m erican, and other statisticians, contain a deliberate bias. A n overvaluation by a h alf a billion francs of F rench indus­ trial production by C haptal after the N ap o ­ leonic wars was ascribed by M oreau de Jonnes in the 1840’s to C haptal’s southern tem peram ent and natural bent fo r exaggera­ tion.

b. Extent of Estimating Involved All national incom e aggregates, as far as they are m ade up of existing data, are m ere com pilations. But in a b roader sense they are really “estimates.” F irst, the estim ator has to choose am ong various data and vari­ ous ways of com bining them . H e thus acts

255

as a co-ordinator, integrator, and arbiter. U pon his decisions largely depends the u lti­ m ate accuracy of his estimates. Secondly, he m ust often adjust his data to his concept, bridge gaps, an d correct fo r overlaps, im ­ p roper valuations, and estim ate com ponents th a t are n o t covered. W hether his w ork entails merely adjust­ ing the original data slightly, o r supplem ent­ ing them substantially, it is difficult and tim e-consum ing. In supplying missing items or m aking corrections the estim ator m ust sometim es rely on m erely representative samples and sometimes only on his general notions as to w hat the actual facts m ay be. Since a certain am ount of conjecture is always involved, the estimates of the various item s range from reasonably supported ap­ proxim ations to bold guesses.

c. Influence of the State of Economic Data T he reliability of national income esti­ m ates, i.e., the com pleteness of their cover­ age and the accuracy of the evaluation of their com ponents, depends prim arily upon the com pleteness and reliability of the n a­ tion’s statistics, and this, in turn, depends largely upon the state of the country’s eco­ nom ic developm ent. T h e better the statis­ tics, the smaller is the area requiring sup­ plem entation and the greater is the ultim ate accuracy o f th e totals. In underdeveloped countries, the range of data is narrow and their quality relatively p oor.2 M oreover, the considerable quantity of goods and services produced w ithin and fo r the household is n ot susceptible of econom ic m easurem ent, and it takes ingenuity and effort to m ake estim ates th a t will be of some value. On the other hand, if the econom y is planned, the statistics m ay be relatively good even though the country is underdeveloped. A s a rule, the econom ic statistics of highly industrialized countries are good.3 B ut in a few, such as F rance, statistics have lagged behind econom ic progress. Before W orld W ar II the F ren ch estimators, for example, struggled vainly w ith inadequate data, w ith the result th at their estimates w ere deficient. T he econom ic statistics of the U nited States are am ong the m ost ex­ tensive, and its estimates, prepared by a substantial staff of governm ent experts, are m ore elaborate th an those of almost any other country. T h e growing practice of preparing esti­ m ates of national income in its different

256

THE IN CO M E OF NATIO NS

phases, using different m ethods as well as different data fo r each, is helping to p ro ­ m ote greater reliability inasm uch as the dif­ ferent sets of independently prepared figures work as a check on each other.

d. Degree of Accuracy of Income-dis­ tributed Method T he general reliability of estim ates is also dependent upon the m ethod used. E ach m ethod o r variant is well suited for a p a r­ ticular type of data and correspondingly unsuited for another. In this and succeeding sections we will exam ine the reliability of estimates prepared by each m ethod in the light of the relative com pleteness and accu­ racy of the particular data. W e begin with the incom e-distributed m ethod because it gives m ore reliable results fo r highly indus­ trialized economies. T he incom e-distributed m ethod, as shown in C hapter 19, has two variants: the “in­ come paid o u t” variant th a t treats income at its source; and the “incom e received” variant that treats incom e at the point of its receipt. T he “incom e paid out” variant possesses certain advantages in that it relies on social security statistics and other pay­ roll data; on corporate reports of profits, dividends, and interest paym ents, and w ith­ holding taxes; and on other private and gov­ ernm ent financial reports. Because of the practical purposes these statistics are in­ tended to serve, their coverage is fairly com plete and they em body reliable valua­ tions. T hey usually need to be supplem ented, however, by estimates of salaries and wages in cash and in kind n o t covered by the social security program , as well as by esti­ m ates of unreported interest and rent pay­ m ents and earned profits. Such supplem en­ tary estimates are m ade either by m ultiply­ ing the estim ated num ber of missing recipi­ ents by an estim ated average incom e o r by using tax statistics fo r profits and interest. In both cases, the income-received variant is used. In the case of m aking supplem en­ tary estimates for farm incom e and net rent, the net output m ethod is em ployed. F o r wage and salary earners (except in local governm ent and m inor trad e s), supplem en­ ta ry estimates are usually fairly com plete, b ut they are seldom so fo r incom es from small capital investm ents and from inci­ dental occupations of the self-employed. Fortunately, these latter types of incom e are an insignificant percentage of total n a­ tional income.

U n d er the incom e-paid-out variant, the weakest evaluations are usually those of com pensation in kind, professional earn ­ ings, business profits— particularly undis­ tributed corporate profits— and the n et bal­ ance of international capital incom e. T he first tw o item s and th e last one are often poorly evaluated because of lack of data. Business profits and undistributed corporate profits derived from tax returns are poorly evaluated because of deliberate concealm ent o f profits, especially in countries w here tax adm inistration is lax and w here business ac­ counts are little publicized. T he incom e-received variant, largely de­ pen d en t on personal incom e tax returns, yields less reliable results th an the incomepaid-out variant. G enerally, incom e tax data are less com plete and reliable th an social security, payroll, or corporate data. They m ust be supplem ented by estim ates of taxexem pt incom es o r portions of incomes, and o f taxable incom es evading taxation. The failure of taxpayers to rep o rt all th eir tax­ able incom e is w idespread, though declin­ ing as tax adm inistration im proves. U n d er­ reporting is considerable n o t only fo r busi­ ness earnings b u t also fo r incom e from se­ curities, bank deposits, and silent p artn er­ ships. E ach am ount m ay be small, b ut the total is often substantial. M oreover, in ­ com es from incidental jobs frequently go unreported and, because o f th e difficulty of estim ating them , are disregarded. U n d er­ reporting o r tax evasion is usually accounted for by raising the rep o rted figures by a given percentage, and this procedure m ay involve a w ide m argin o f error. Sometimes, underreporting is left u n co r­ rected on the ground th a t it is probably just about offset by know n duplications to be contained in o th er sectors of th e estimates. In th e Brazilian estim ates fo r 1 9 4 7-49 it was impossible to exclude all intercorporate interest paym ents, yet no deductions w ere m ade because this was considered to be m ore th an offset by the significant am ount o f nondeclaration of certain interest re­ ceived (see C hapter 31, Section 3 ). A s in the case of th e incom e-paid-out m ethod, the least reliable item s in th e in ­ com e-received m ethod are com pensation in kind, earnings of the self-employed, and u n ­ distributed profits. The international balance of capital income, w hich is an additional source o f difficulty u n d er the income-paidout variant, is n ot a com ponent o f the in ­ com e received m ethod, an d its absence com ­

1 7 . ESTIM ATING M ETHODS

pensates som ew hat fo r the other com plica­ tions. T he difficulties involved in applying the income-received v ariant were revealed by the successive upw ard revisions, in the late 1940’s and early 1950’s, of the official B rit­ ish estimates of capital incom e other than rent, and of business profits and, conse­ quently, of the total national incom e. In previous years the differences betw een the totals o f gross national product and gross national expenditure in the British estimates w ere elim inated by m inor adjustm ents in individual item s of incom e or expenditure. In recent estim ates no attem pt is m ade to adjust the differences. T hey are simply shown as a residual error; i.e., presented as an unknow n item of incom e th a t m ay be either plus or minus. T he official estim ators point out that “this is purely fo r convenience of presentation and does n o t im ply th a t the estim ates of expenditure are necessarily su­ perior in accuracy to the estim ates o f in­ com e.” 4 T he size of the error m ay be seen from the follow ing figures: 5 Table 17-2

Year 1948 1949 1950 1951

Gross N ational Product (million £ ) 10,216 10,926 11,515 12,676

Residual Error (million £ ) -5 2 -6 0 49 40

e. Degree of Accuracy of the Net Out­ put Method G enerally, the net output m ethod is less reliable than the incom e-distributed m ethod, b u t in underdeveloped countries it is the only one th a t can be em ployed extensively. Because it requires tw o series o f estim ates— one of gross value of o utput and one of deductible costs— the possibilities of error are doubled. Statistics on th e gross value of output usually do n o t cover small enter­ prises, dom estic industry, or som etim es even the output o f governm ental enterprises, and are often based on prices or production val­ ues th at are n o t clearly defined. T he prices used are generally average annual prices fo r the country as a whole. Because these aver­ age prices cannot possibly reflect all the innum erable regional and seasonal variations in prices occurring during the period of a year, the evaluations are not altogether ac­ curate.

257

M ost current agricultural production sta­ tistics are restricted to m ajor crops and based on im perfect sample data. Even less com plete are the data on deductible costs. F o r these and other reasons, the estim ator is obliged to m ake his own roundabout esti­ m ates to fill the gaps in published data on the gross value of ou tp u t and also to esti­ m ate deductible costs including depreciation. T his had to be done, fo r example, in the estim ates for the British W est Indies and M alaya. Since costs could n ot be estim ated on an industry basis, the estim ator had to avoid double counting by making block de­ ductions of costs from gross total output o f all industries (see C hapter 32, Sections I and I I - l ) . These shortcom ings are accentuated in underdeveloped countries, where the statis­ tics, as already noted, are especially poor and household industry is prom inent. In the estim ates fo r K enya, fo r example, the net value of the products of A frican subsistence agriculture consum ed on the farm , estim ated on th e basis of surveys of fam ily budgets and tribal diets, resulted in an unreliable figure fo r this relatively im portant item. It ranged from 29.1 p er cent in 1947 to 21.5 per cent in 1951 of the net product totals.6 In m ore advanced countries, such as C anada, Sweden, and the N etherlands, the m ethod produces b etter results. However, D utch national incom e in the com putations m ade fo r 1921-38 was considerably lower when estim ated by th e n et output m ethod than w hen estim ated by the income-received m ethod. D espite its defects, the n et output m ethod is useful in highly industrialized economies as a subsidiary estim ating device, and in the others it is of necessity the m ain tool.

f. Degree of Accuracy of Income-dis­ posed or Final Products Method T h e incom e-disposed or final products m ethod, as m ore fully described below (C hap ter 20) has three variants: the com ­ m odity flow variant, the final sales variant, and th e fam ily budget variant. U n d er the com m odity flow variant, which traces p roduct from the raw materials to final product stage, a w ide variety of pro­ duction and distribution statistics is required and the whole process of estimating is ex­ trem ely roundabout. A t various stages of the incom e flow one m ust distinguish be­ tw een interm ediate and final products; be­ tw een final investm ent goods and final con­

258

THE IN CO M E OF NATIONS

sum er goods; between different types of purchasers such as governments, traders, and private households; and between addi­ tions to inventories and sales. In the process, m any assumptions of varying reliability m ust be m ade. A lthough the enum eration of products m ay be relatively com plete, their segregation into these several categories is exceedingly difficult and m ay easily involve considerable margins of error. This m ethod was used in estim ating the greater portion of the national expenditures in the U nited States fo r 1929-39 and 1947. It was also used, although not to the same extent, in the D utch, D anish, N orw egian, and Jap a­ nese estimates after W orld W ar II. T he final sales variant, used m ost elabo­ rately in C anada in estim ating national ex­ penditure for consum ption, is m ore m an­ ageable and dependable. It, too, requires considerable estimating, especially in deter­ m ining total final sales to consum ers, busi­ ness, and governm ent as distinguished from total interm ediate sales. It is not always easy to determ ine w hether the given sales are fo r consum ption or fo r current use in production. T herefore, national income es­ tim ates based on total final sales m ay easily be inaccurate. T he fam ily budget variant, used exten­ sively in the postw ar estim ates fo r Japan, is directly applicable only to personal in­ come. It does not furnish a basis fo r a direct estim ate of national incom e. Based on im­ perfect samples of families, and requiring adjustm ent between personal and national income, it cannot be very reliable.

g. Summary In general, the incom e-disposed o r final products m ethod is at present the least re­ liable of the three m ethods. T his was clearly recognized by the F rench estim ators who used all three of them after W orld W ar II (see C hapter 27, II-3 ). Since the various m ethods are usually com bined, their m erits and dem erits are reflected in different parts of the estimate. T he relative accuracy of an entire estimate is therefore difficult to de­ term ine. On the whole, estim ates suffer m ore from lack of reliability in evaluating the com ponents than from any failure to cover all units. In economically advanced countries, the incom e-distributed m ethod generally gives the m ost reliable estimates, especially if based on social security statistics— which m ay well become a prim e source fo r estim ating

national incom e. H ow ever, fo r estimates of farm incom e and rent, and national income generally in underdeveloped countries, the net o u tput m ethod will keep its choice place. It is significant th a t it was em ployed in the m ost recent official estim ates fo r India, w hich were developed w ith considerable care (see C hapter 30, P art I ) . F o r estim ating national incom e in its disposition phase, the theoretically m eaningful b u t practically com plicated com m odity flow v ariant cannot be used for m any countries. Even in the U nited States, where, because of its highly developed statistics, it has been applied with relative success, it has been superseded by the sim pler final sales variant, at least for years in w hich there were no com prehensive industrial censuses. Sales statistics m ay be so im proved in various countries in the fu tu re as to render the application of the final sales v ariant m ore practicable.

3. E xtrapolation and Interpolation M any national incom e estim ates are p ro ­ jected fo r a series of years from data fo r one o r a few pivotal years. T h e figures are either projected from one to oth er years, i.e., extrapolated, or are filled in between tw o separated years fo r w hich figures are available, i.e., interpolated. E xtrapolation and interpolation, w ith the aid o f subsidiary data and various technical devices, are used in estim ating the net o u tp u t o f m ajor divi­ sions o f econom ic activity, the m ajor factor incom es, o r even the national incom e total itself. E xtrapolation is m ore difficult th an inter­ polation because it deals w ith a wholly u n ­ certain past o r fu tu re tim e, and w ith bench­ m ark d ata fo r only the initial o r base year o f th a t period; w hereas interpolation p ro ­ ceeds from two base points—the initial year an d the term inal year o f a period— and, therefore, deals w ith a m uch m ore lim ited uncertainty. E xtrapolation requires the es­ tim ato r to assum e a probable rate o f increase in the fu tu re (o r past) in the various factors affecting the national incom e such as: rate of grow th of population, productivity, gov­ ernm ent activity, change in th e price level, and the probability of peace o r w ar. In ter­ polation, on the other hand, deals with know n realized facts o f the total growth of population, production, governm ent ac­ tivity, and the condition o f peace and w ar over an entire period. T h e only unknow n circum stances in the case are the annual

1 7 . ESTIM ATING M ETHODS

rates of change at w hich the total changes in these factors or conditions have taken place. In each case an assum ption m ust be m ade as to w hether the annual rate was accelerating o r decelerating in nature, and a decision m ust be m ade regarding the weights to be assigned to the initial and term inal bench-m ark data as reflections of the intervening rate of growth. T h e results of both extrapolation and interpolation becom e less reliable the longer the period covered by the estim ate. E rrors becom e larger and cum ulative as the dis­ tance lengthens from the base year in ex­ trapolation, o r betw een the two base years in interpolation, and there is no effective m ethod for checking the cum ulative errors. E xtrapolation and interpolation, although differing in their problem s, are both of five m ain types: (a ) projection of com prehen­ sive census incom e aggregates fo r a given y ear to other years on the basis o f less com plete annual data; ( b ) projection of value aggregates on the basis of reported changes in the quantities and values of com ponent o r related items; (c) projection of the net value of output on the basis of estim ated changes in gross value and in the net-gross ratio; ( d ) substitution of sup­ posedly co-ordinate series fo r unknow n data; ( e ) projection on the basis o f appraisal of business conditions o r arbitrary assumptions. Frequently, these approaches are com bined in estim ating the same item, o r they are modified. H ere we can review only some of the m ost typical and significant applications.

a. Census Figures Projected on the Basis of Less Complete Annual Data In m any countries, over-all censuses of agriculture, m anufacturing, distribution, and o ther fields of econom ic activity are taken sporadically o r regularly. F o r the other or intervening years, current statistics are avail­ able which, though less com plete than the census data, are m ore com plete and repre­ sentative than m ere samples. Indexes of the annual changes in the item s involved, de­ rived from the less com plete annual data, are applied to the census figures. F o r example, Saenz developed his annual estim ates of M exican agricultural and m anu­ facturing gross output by interpolation from decennial censuses with the aid of less com ­ plete annual data on gross value of m anu­ facturing output for the intervening years. F o r agriculture, he determ ined first the ratio

259

o f the value of the o u tput reported by the M inistry of A griculture fo r 1929 and 1939 to the value of the output in those years subsequently reported in the censuses. These ratios were taken as indicative of the extent of underreporting of the values of the out­ p u t by the M inistry of A griculture’s annual reports. N ext, Saenz took the output value figures reported by the M inistry fo r 193538 and fo r 1940-45 and m ultiplied them by the same ratio to get gross value of o ut­ put. In o rd er to obtain the net values for those years, he assum ed th at the ratios of n et to gross revealed by the censuses fo r the years 1929 and 1939 also held fo r the in­ tervening years. In C anada, 70 p er cent of the wage and salary estim ates for 1949 were taken directly from annual surveys and other annual pub­ lished statem ents, but 28 p e r cent were projected on directly related data. F or ex­ am ple, wages and salaries in agriculture w ere extrapolated from the 1941 census data on the basis of an index of the average farm wage bill (see C hapter 21, Section I-3 a ). In F rance, the 1938 value outputs reported by the trad e associations, as was required by the G erm an occupation authori­ ties, w ere extrapolated to the postw ar years by m eans o f indexes of the num ber of m anhours of em ploym ent in those branches of production and indexes of prices. Both in­ dexes were defective, the first one failing to account fo r changes in the productivity of labor, and the second being based on im perfectly developed average prices which did not take account of regional variations (see C hapter 27, Sections 1-1 and 1-3).

b. Value Aggregates Projected on the Basis of Changes in Quantities and Values of Component or Related Items Sometimes, in extrapolating or interpolat­ ing a given item of incom e or product, the only data available are those on the trend of change during the period in question of some one o r tw o com ponents of th at item, o r even of the changes only in some related items. A s an exam ple of extrapolation with the aid of data on com ponent items, we may cite two cases from British prew ar estimates. Colin C lark extrapolated aggregate wages from base year census and other data with the aid of cu rren t social security figures on the num ber of wage earners and average

260

THE INCO M E OF NATIONS

wage rates. Bowley, in his estimates, did a sim ilar extrapolation of aggregate wages on the basis of the num ber of wage earners and average wages in the base year adjusted by an index of em ploym ent and average wage rates fo r the succeeding years. E xtrapolation w ith the aid of data on related items was done in the C anadian es­ tim ates for the early 1950’s. Cash wages of domestic servants during the census year were extrapolated on a com bined index of em ploym ent in hotels and restaurants and wage rates in com m ercial laundries during the years following th e census. In some estimates, base year dwelling rentals are projected to other years by carrying forw ard the num ber of dwelling units w ith the aid of construction statistics, and applying a projection of average rents. Sometimes, pay­ rolls and gross rents are projected according to changes in the com position of the labor force or kinds of dwellings. Incom e from trade has been sim ilarly projected on the basis of sales data and estimates o f profit margins. F o r example, in the Polish esti­ mates, the earnings of artisans and small m erchants were projected from 1929 to 1933 on the basis of sales; changes in profit m argins w ere based on the variations in prices paid and received in trade.

c. Net Value of Output Projected on the Basis of Variations in Gross Value and of Either a Constant or a Variable Net-gross Ratio One of the m ost com m on m ethods of both interpolating and extrapolating the net outputs of given branches o f production is to adjust the available data on changes in gross value according to either the know n net-gross ratio of the base year o r years, o r some index of the net-gross ratio based on sample studies. Obviously, w herever such ratios have undergone substantial changes during the period, this procedure will re­ sult in substantial errors. M atolcsy and V arga used this approach, com bining changes in gross value and in the net-gross ratio, in estim ating the net output of H ungarian trade. T heir estimates o f net outp u t fo r 1928 were projected to 1924-37 by m eans of two specially con­ structed indexes: one representing the in­ crease since 1928 in the value of agricul­ tural, mining, and industrial products and im ports of finished goods entering trade;

and the other representing changes since 1928 in the ratio of net to gross o u tp u t in trade calculated on the basis of the changes in profits and payrolls o f trading corpora­ tions. In the C anadian postw ar estimates, gross revenues of unincorporated m anufacturing establishm ents w ere tabulated from census of industry returns for the years since 1946. N ext, by applying a ratio of n et to gross revenue obtained from taxation statistics, the gross revenues w ere reduced to a net basis. F o r 1926-45, th e 1946 ratio of net to gross was applied to the gross figures (see C hapter 21, Section 1-2).

d. Substitution of Supposedly Co-ordinate Series for Unknown Data T he fo u rth m ethod of extrapolation and interpolation, involving the use of sup­ posedly co-ordinate series, is one of the weakest m ethods. T he supposedly close co­ ordination between changes in the particular item s observed at one tim e m ay no longer hold tru e fo r the period u n d er consideration because of changes in technology o r other factors, w hich had occurred since. Sample projections o f this kind m ight include: (a) projection of net ou tp u t of transportation on the basis of changes in freight tonnage, passenger transportation, and sales o f gaso­ line; ( b ) projections of n et o u tput of trade on th e basis of an index of railroad freight deflated by an index o f wholesale prices; (c ) projection of total professional incom e on th e basis of taxable professional incom e; and (d ) projection of the incom es o f m usi­ cians, playwrights, and artists on the basis of changes in theater attendance as reported in adm ission tax records. Exam ples of these are fo u n d in m any o f the older estimates (e.g., M atolcsy-V arga fo r H ungary; Royal Com m ission fo r C anada; p rew ar Polish, T urkish, and Japanese estim ates; e tc .).

e. Projections Based on General Appraisal of Business Conditions or Arbitrary Assumptions T h e fifth type of projection, based on general notions of the tren d o f business conditions or on other arb itrary assumptions, is ano th er w eak m ethod, justifiable only w hen b etter techniques can n o t be used. In b o th the Belgian and Swiss p rew ar estimates, entrepreneurial incom e fo r years n o t cov­ ered by incom e tax data w ere projected on

17.

261

ESTIM ATING M ETHODS

the basis of various indexes of general busi­ ness activity. Saenz did this in his Mexican estimates for trade, and Duncan did so in his Irish estimates for the net output of small industry and income from professions.

4. R eliability of the Final O utcom e of the E stim ates a. Qualitative Appraisals National income estimators have always been concerned with ascertaining the rela­ tive accuracy of their results with a view toward achieving the greatest accuracy pos­ sible within the limits set by the available data and their own statistical skill. Ap­ praisals of this kind are essentially qualita­ tive in nature, involving the identification of the weaker elements of the estimates. After examining the way in which each part has been constructed, the same appraisal can be made of the estimate taken as a whole. Colin Clark gave expression to this type of qualitative appraisal when, in his Con­ ditions of Economic Progress, published in 1940, he arrayed the existing estimates of various countries into four groups as fol­ lows: (a) those based on accurate taxation or production census statistics; (b) those based on somewhat incomplete or deficient statistics; (c) those based on approxima­ tions; and (d) those based on very rough estimates. The Irish and Swiss estimators rated the relative accuracy of the various components in their estimates in broad qualitative cate­ gories, more or less along Colin Clark’s suggested classification. Thus, the official Irish estimates presented the following selfevaluation for 1938: 7 Table 17-3. Irish E stim ates M illion £

Per Cent of T otal

Reported figures Good estim ates (error not exceeding 5 per cent) Rough estim ates Largely guesses

6 8 .6

45

5 6 .4 2 1 .9 7 .5

36 14 5

National income

154.4

100

The Swiss report showed a breakdown of their total according to the degree of re­ liability of the various income items in 1938:8

Table 17-4. Swiss Estim ates M illions Basis of E stim ates of Francs D irect available statistics covering following per­ centages in the field: 85 per cent 3,077 50 to 85 per cent 475 less than 50 per cent 3,121 Com bination of and deriva­ tion from fragmentary 1,426 data 614 Informed guesses T otal

8 ,7 1 3 *

Per Cent of T otal

3 5 .3 5 .3 3 5 .8

16.4 7 .0 100.0

a T he final figure is larger than national income because deduction item s th a t had to be estim ated (intercorporate dividends, losses) were added.

b. Measures of Degree of Accuracy These qualitative appraisals of the degree of reliability are not sufficient in themselves. They must be supplemented by quantitative measures of the probable margin of error contained in the estimates. Users of the es­ timates, whether they be governments, pri­ vate groups, or scholars, do not generally expect them to be free of all error. They recognize that estimates can only be ap­ proximations, but in order to use them in­ telligently they must know to what extent the final totals, or the particular components they are dealing with, may be off one way or the other. It is the estimator’s responsibil­ ity to appraise the various parts of his own estimate critically and to publish the results of his appraisal for the benefit of the persons using his estimate. This final task in the preparation of estimates is, perhaps, the most difficult of all, calling for the utmost objectivity and self-discipline on the part of the estimator. So far, only a few estimators have at­ tempted appraisals of this type. For example, Simon Kuznets in his estimates for the United States (1919-38) appraised the probable error in his estimate for each in­ dustrial division, and within them for each component type of income paid out, classi­ fying the margins of error into four cate­ gories in the following ascending order: 9 1. The lowest errors ranging from 5 to 10 per cent, with an average of 7.5 per cent. If the margin appeared to be less than 5 per cent, the estimate was put into this category; 2. a range of 11 to 20 per cent, with 15 as the average;

262

THE IN CO M E OF NATIONS

Table 17-5. Ratings of Accuracy of the French National Income Estimates of 1949 R ating as to R eliability

Item s

Approxim ate Percentage W eight in the T otal

A. Elem ents of Gross N ational Expenditures Private consum ption Governm ent consumption Gross addition to fixed assets Change in inventories N e t lending to abroad

c a b d c

T otal

75 13 14 2 —4

b /c

100

B. Elem ents of Gross N ational Income Gross national income Transfers from abroad Gross national product Depreciation N e t national product Indirect taxes net of subsidies N ational income F actor income from abroad N e t dom estic product D irect taxes on corporate income G overnm ent transfers to enterprises Com pensation of governm ent em ployees Em ployers’ contributions to social security Com pensation of enterprises’ em ployees Undistributed income Income from property and entrepreneurship

b /c c b /c d c a c c c a a a a b c c

100 0

100 100

10 90 10 80 0 80 1 -4 8 6 30 5 34

C. Saving G overnm ent saving Household saving

a c

s o u r c e : Organisation for European Econom ic Co-operation, N ational Accounts Research U nit, Studies of Social Accounts, France, 1951, pp. 137-38.

3. a range of 21 to 40 per cent, with 30 as the average; and 4. the highest range of errors was 41 to 80 per cent, with 60 as the average. If the margin appeared to be more than 80 per cent, the estimate was put in this category.10 Benham, who also attempted a quantita­ tive appraisal of the margin of error in his national income total for Malaya, set it at within 10 per cent, although this figure was too low for a number of the individual items that made up the total. In some cases, quantitative appraisals of given estimates can be made by an outside research organization that specializes in sur­ veys of existing estimates and is capable of setting up estimating standards. The Na­ tional Income Unit of the Organisation for European Economic Co-operation at­ tempted, in certain instances, in its study of the national income estimates prepared

by the first official estimating agency in France in 1949 (whose estimates have im­ proved materially since that time), to de­ velop an exceedingly detailed and elaborate appraisal, indicating what it judged to be the probable margin of error for each com­ ponent element of that estimate. This ap­ praisal is shown in Table 17-5. In 1956 the Central Statistical Office of the United Kingdom published an elaborate quantitative appraisal of the relative accu­ racy of the various components of its esti­ mate of the gross national income and ex­ penditure of the country. It used three gradings: A less than 3 per cent margin of error B 3 to 10 per cent margin of error C more than 10 per cent margin of error These margins of error were said to be personal judgments, likely to be 90 per cent correct, rather than the results of any sam-

17.

263

ESTIM ATING M ETHODS

Table 17-6. Reliability of Estimates of Major Components at Current Prices— United Kingdom, 1955 Incomes: Income from em ploym ent a Incom e from self-em ploym ent Gross trading profits of companies Trading surpluses of public corporations Gross profits of other public enterprises Rent N et income from abroad Rent, dividends and interest received by persons Current grants from public authorities received by persons Expenditures: Consumers’ expenditures a Public authorities’ current expenditure Gross dom estic fixed capital formation a Value of increase in stocks and work in progress Of which— Stock appreciation Physical increase Exports of goods and services Im ports of goods and services T axes on expenditure Subsidies Gross national product

Ratings A B B A A B C B A A A B C C C A A A A A

a A detailed rating is also given for the com ponents of this estim ate. s o u r c e : Central Statistical Office, N ational Income Statistics, Sources and Methods, Her M ajesty's Stationery Office, 1956, p. 36.

pling of errors. The resulting ratings are summarized in Table 17-6.

Group A — Possessing Highest Degree of Reliability

5. Classification of E stim ates of Various Countries A ccording to Their R eliability

United States United Kingdom Canada Australia New Zealand

Without attempting to appraise all the existing eighty-odd national income esti­ mates from the point of view of their rela­ tive reliability, it is possible to do so in a very approximate way at least for a cer­ tain number of them, using two tests: (a) the quality of their underlying statistical data and (b) the relative advancement of the statistical techniques employed in their preparation. It is possible to classify them on that basis into three broad groups: group A, scoring high; group B, making a medium score; and group C, ranking low. Any attempt at a classification of the existing national income estimates into these three groups must necessarily be somewhat subjective and debatable. Nonetheless, the following classification of at least some of these existing estimates on a qualitative basis is offered herewith for illustrative purposes:

Sweden Netherlands Switzerland Norway Denmark

Group B— Possessing M edium Degree of Reliability

France Belgium Germany Italy Ireland

Finland Austria Puerto Rico Israel

Group C— Possessing Low Degree of Reliability

Spain Soviet Russia Poland Hungary Bulgaria Yugoslavia Rumania Turkey Greece

Egypt India China Malaya Japan Mexico British West Indies Brazil

264

THE IN CO M E OF NATIONS

n o t e : Since this was written, and shortly before the book went to press, the Joint Economic Committee of the United States Congress published a “Memorandum on International Statistics” prepared for it by the Office of Statistical Standards of the United States Bureau of the Budget. This memorandum contained a quality rating of the national income estimates of 64 coun­ tries. The countries were not identified by name but were referred to by numbers in each regional group. Four ratings were used: I—very good; II—good; III—fair; IV— weak. The results, reproduced in Table 17-7, are not very different from those given in the list above. Thus, it will be noted that the very good estimates (Group I) are con­ centrated in western Europe, North Amer­ ica, and Oceania, and obviously comprise the United Kingdom, the Netherlands, Sweden, the United States, Canada, Aus­ tralia, New Zealand, and some ten other countries. Group II—good estimates—is scattered among several regions and prob­ ably includes the Union of South Africa, Puerto Rico, Israel, and several others. Groups III and IV—estimates that are

merely fair or actually weak—on the other hand, are concentrated in Africa, Asia, and Latin America. The progressive improve­ ment in statistical data and techniques over the next few years should bring about a material change in this picture—a larger number of estimates appearing in Groups I and II and a smaller number in Groups III and IV. T able 17-7. Q uality R ating of N ational Accounts S tatistics in 64 Countries, by Continents, End of 1956 C ontinent T otal Africa America, North America, South Asia Europe and Oceania

All

I

II

III

IV

64

17

9

18

20

1 1 2 2 3

2 5 4 5 2

6 3 2 8 1

9 11 8 15 21

2

15

s o u r c e : U .S. Congress, Joint Econom ic Com­ m ittee, International Economic Statistics, m em o­ randum prepared by the U .S. Bureau of the Budget, Office of S tatistical Standards, 85th Con­ gress, 2d Session, Joint C om m ittee Print, W ashington, D . C., 1958.

18. THE NET OUTPUT METHOD 1. Characteristics of the M ethod The net output method, sometimes called the production method, consists, in its standard form, of three stages: (1) esti­ mating the gross value of domestic output in the various branches of production; (2) determining the costs of materials used and of services rendered in these branches by other branches as well as of annual de­ preciation on the physical plant; and (3) deducting these costs and depreciation from gross value to obtain the net value of do­ mestic output in these branches, and then adjusting the latter for net capital income from abroad. The net value of output, or more briefly “net output,” is often called the “value added” or the “income produced” originat­ Notes begin on p. 535.

ing in an economic division. Since it is identical with the sum of wages, salaries, supplementary labor incomes, interest, prof­ its, and net rent paid or accrued, it can also be calculated directly by aggregating these shares. Direct calculations are really an adaptation of the income-distributed method to the net output approach, rather than an application of the method itself.

2. Classification of Industrial D ivisions Implicit in the net output method is a classification of national income by the di­ visions of economic activity in which the income originates. Similar classifications, however, are used whenever national income

18.

THE N E T O U TPU T M ETHOD

estimates try to show the origin of the in­ come, regardless of the method by which they are prepared. Hence, in discussing the classifications of national income by eco­ nomic or, as they are commonly called, “industrial divisions,” all estimates making use of these divisions are considered. The largest and most general divisions are agriculture, manufacturing industry, transportation, trade, services of dwellings, professional and other services, and govern­ ment; but these may be variously combined with smaller divisions such as forestry, fish­ ing, mining, construction, communications, electric and other power, finance, real es­ tate; or they may be variously subdivided. Forestry is as often segregated as it is grouped with agriculture. Fishing is, as a rule, combined with agriculture, but it is treated separately in several estimates and is lumped with forestry and gardening in others. Manufacturing industry is often broad­ ened into an industrial division embracing also mining, electric power and gas plants, and construction. In the Soviet estimates, industry includes even fishing, forestry, and hunting. However, construction is often a separate category and is differently defined in the case of different countries. In some estimates this division is restricted to pri­ vate contract construction; in others it in­ cludes also public construction. In some es­ timates (United States, Norway, and oth­ ers), electric power is shown separately. Mining is as often treated separately as it is combined with manufacturing, and it is sometimes combined with metallurgy. Although trade, transportation, and finance are usually shown separately, trade is sometimes combined with transportation and in a few estimates, such as the prewar Swedish estimates, with finance. In one or two cases, such as the Matolcsy-Varga Hun­ garian estimates, freight transportation and finance were not shown as industrial divi­ sions; their incomes were partly included in the value added of the industries using their services. Services of hotels and restaurants are sometimes included in trade, but more often are put under services or, as in the postwar German estimates, are shown as a separate category. The treatment of government services varies most, sometimes including public en­ terprises and public construction, and some­ times not. National defense and other pri­

265

mary functions of government, such as internal protection and social welfare as well as general administration, are generally included under government, but public health administration, education, and con­ struction are more generally shown under the separate classifications of health, edu­ cation, and construction—the latter some­ times subdivided into private and public types. Public enterprises, particularly those pertaining to transportation, electric power and gas, and communications are generally lumped with private activities in these fields under the pertinent categories. In some es­ timates, general government is included as a separate industrial classification. In the United States estimates, the government category is very broad, covering even the contribution of government enterprises (less profits) as well as general government. The most extensive differences among in­ dustrial classifications are in the number of their subdivisions. Some, such as those used in the estimates for the United States, are exceedingly detailed, containing about sev­ enty-one divisions and subdivisions; others, such as those used in the estimates for Canada and the Netherlands, are only mod­ erately specific, containing from a dozen to a score of categories. Still other classifica­ tions, such as those used in the estimates for Soviet Russia, are highly concentrated, containing only some half dozen divisions. To make national income estimates more comparable, attempts to standardize the classification of industrial divisions were made by two important international groups: in 1948 by the Statistical Commis­ sion of the Economic and Social Council of the United Nations; and in 1952 by the National Income Unit of the Organisation for European Economic Co-operation. The UN’s classification, generally referred to as the “International Standard Industrial Classi­ fication” (ISIC),1 consisted of 10 main di­ visions. The OEEC classification increased the number of main divisions to 13 2 by breaking down the UN’s division of “com­ merce” into three main divisions: wholesale and retail trade; banking, insurance, and real estate; and ownership of dwellings. It also subdivided the UN’s two divisions of “government” and “miscellaneous services” into three: (a) public administration and defense, ( b) health and educational services, and (c) miscellaneous services, namely ho­ tels and restaurants, domestics, and other

266 T able

THE INCOM E OF NATIONS

18-1.

The Organisation for European Econom ic Co-operation’s (OEEC) Industrial Classification

1. A griculture, fo restry , a n d fishing a. A griculture b. F o re stry , h u n tin g , etc. c. Fishing 2. M ining a n d q u a rry in g 3. M a n u fac tu rin g a. Food, drink, a n d tobacco b. T extiles c. C lothing a n d footw ear d. W ood a n d cork p ro d u c ts including fu rn i­ tu re e. P ap er, prin tin g , a n d publishing / . C hem icals a n d allied trad e s g. Stone, clay a n d glass h. B asic m eta l industries i. M e tal m an u fa ctu re , engineering, a n d elec­ trical goods j . T ra n s p o rt e q u ip m e n t k. O th e r m an u fa ctu rin g 4. E le ctricity , gas a n d w a te r w orks 5. C onstru ctio n

6. Transportation and com munication a. W ater transportation b. Railroads c. Other transportation and storage d. Communications 7. W holesale and retail trade a. W holesale trade b. R etail trade 8. Banking, insurance, and real estate 9. Ownership of dwellings 10. Public adm inistration and defense 11. H ealth and educational services a. Educational services b. H ealth services 12. M iscellaneous services a. H otels and restaurants b. D om estic services c. Other services 13. N et incom e paym ents to factors of produc­ tion from the rest of the world

service industries and trades. The OEEC classification is reproduced in Table 18-1.

ments in the data that would make them fit the concepts and classifications of indus­ trial divisions employed in the estimates. For example, residential rents are generally reported only in sample studies or occa­ sional housing censuses, and current data on wholesale and retail sales are rare. On the other hand, statistics on the revenues of railroads and other transportation agencies, communication services, and government are generally ample. This was true, for ex­ ample, in the 1946-48 estimates for Nor­ way,3 where detailed data from government sources were available for railroads, com­ munication, etc. But for trade, except for transport services, the value could not be allocated directly among the various indus­ try classifications and these data were col­ lected into an “unspecified” category. For house ownership, the Norwegian population census of 1946 had to be used as a base for estimates for subsequent years. Apart from certain exceptions noted be­ low, and within the limits of their coverage, production statistics generally give fairly satisfactory valuations. They show values of output at producer prices, that is, at prices obtained by farmers, forest and mine owners, and manufacturers. Goods pro­ duced but not sold are likewise generally valued at producer prices. Additions to in­ ventories and goods in process, however, are sometimes given at cost, excluding prof­ its, or at accounting values reflecting a re­ valuation of the stocks to take account of

3. Estim ating the G ross Value of the O utput The gross value of the output of an in­ dustrial division is estimated either by com­ puting the aggregate output from production data and multiplying it by appropriate prices, or by taking from the accounts of businesses the figures of their gross receipts from sales and changes in the value of their inventories during the period of the year. The main concern under the first method is to cover the goods or services produced by each division as completely as possible and to apply appropriate prices. Production statistics for agriculture, forestry, mining, and manufacturing are not always complete, making it necessary to procure additional data. Frequently, they do not cover the out­ put of small enterprises, minor branches, or other items; e.g., home gardening, minor crops, or animal husbandry in agriculture, handicraft and household manufacturing in industry, or minor minerals in mining. Most construction statistics cover only public con­ struction of all kinds and residential build­ ing in the large cities. Under the second method, the gross value is to a large degree furnished and the esti­ mating work is simplified, but it is still neces­ sary to fill gaps in the data on sales and values of inventories and to make adjust­

18.

267

THE N E T O U TPU T METHOD

price changes occurring during the year. Revaluation of inventories gives rise to spe­ cial statistical difficulties and is not always well taken care of in the estimates. For example, only from 1948 on did the Irish estimators make provisions for the changes in the value of nonagricultural stocks aris­ ing from price movements alone. In Malaya, current agricultural, mining, and manufac­ turing production statistics are confined to quantities. Because of these many different statistical inadequacies, it is often necessary to do a considerable amount of rough es­ timating and to resort to simplified or in­ direct methods of calculation.

tion for various sizes of holding. The outlays and incomes of farmers and farm laborers in the Swiss estimates were calculated simi­ larly: average incomes were applied to total farm land for each size class of farm; and for holdings under 3 hectares (7.4 acres) not covered by the sample statistics, gross and net output were estimated separately. Basic agricultural statistics are relatively complete for only a few countries. In the majority they are fairly comprehensive only for main crops, but have wide gaps with respect to other cultures and especially with respect to animal products (meat, dairy products, hides, etc.). The official produc­ tion statistics have to be supplemented by special studies on various aspects of the farm economy. When quantities are not reported, they are estimated from average yields of the crop or output per acre cultivated, or other production unit, and from the total crop acreage or number of producing units. Sometimes, the estimators must correct the official figures on the acreage under cul­ tivation, or on the number of units in pro­ duction, which may be incomplete. Some­ times, too, the information on the average yield is somewhat crude and does not al­ ways specify the sources of the figures. Table 18-2 illustrates yield figures in ani­ mal production used in six estimates. Animal production is estimated by vari­ ous methods; makeshifts are commonly re­ sorted to when the data are imprecise. Thus, in the prewar estimates for France, the gross value of animal production was assumed to equal the value of the feed produced to sustain the livestock. In J. A. Vandellos’ estimates for Spain for 1914 and 1923 and

a. Agriculture, Including Fishing and Forestry For nearly all countries, agricultural in­ come is measured by the net output method. When the gross value of agricultural output, based on quantities and average farm prices for each product, is not given in the original statistics, it must be estimated. In the major­ ity of countries, current statistics show the gross value of agricultural output as well as the quantities of various products, but in some, information on quantities only is available, making necessary separate evalu­ ation of them by use of special price data. For the United States and apparently also for Germany before World War II, gross value was calculated by aggregating sales, the value of home consumption, and net additions to livestock and inventories. In the older Swiss estimates, the values of sales and accruals were reported, but they were based on farm accounting samples and the values were expressed in them in average amounts per hectare of land under cultiva­

Table 18-2. Average Annual Yield per U nit of Livestock Used in Certain Estim ates in Estim ating the O utput of M ilk, Eggs, and Wool Country and E stim ate Hungary, 1926-28 von Fellner M atolcsy-Varga Y ugoslavia, 1923-25

M ilk per Cow

E ggs per H en W ool per Shec

Liters

Pieces

Kilograms

1 ,3 8 1 “ 1,425 9 0 0 -1 ,0 0 0

50 70 100

1.5 4 .5 2 .5 Pounds

Pounds British India Rao, 1925-29 Rao, 1931-32 M alaya Benham, 1947-49

300 500-750 —

3 2 40

* Original figure, 1,750, reduced 18 per cent for milk fed to calves.

—

268

THE IN CO M E OF NATIONS

Table 18-3. Basis for Estimating the Gross Value of Meat Production in Selected Estimates * Basis

Estim ates

Slaughterhouse statistics (either governm entoperated or from taxes on slaughtering) plus farm slaughtering and animal exports L ivestock census plus sample data on samples and inquiries on average numbers and yield M eat inspection reports, sam ple of butchering firms, and data from common carriers

British India (prewar R ao); H ungary (prewar M atolcsy-V arga); M alaya (Benham ); P hilip­ pines (U N ); Brazil India; Norway; Y ugoslavia; Netherlands; B u l­ garia (prewar Tchackaloff) U .S.

* All estim ates are post-W orld War II and official, unless otherwise specified.

in P. Rediadis’ for Greece for 1929, the gross value of animal production was put at half of the value of the livestock. Meat production is estimated from farm­ ers’ sales or slaughtering data, the net in­ crease in livestock, or from meat consump­ tion per capita and export statistics. Table 18-3 shows, without any claim to complete­ ness, the methods used in certain estimates as far as they could be identified. Figures on slaughtering, even after adjustment for farm slaughtering, may often give a mis­ leading picture of meat production, since slaughtering may fall short of or exceed meat production, leading to a net accretion in inventories or to their depletion. Much more satisfactory are the estimates correlat­ ing slaughtering with changes in the num­ ber of livestock. Statistics on fish output are available for many countries, but are often incomplete. Matolcsy-Varga, in their prewar estimates for Hungary, had to supplement the pub­ lished statistics of the catches of the fisheries with estimates based on the total acreage of different types of fishing waters and their average fish catch per acre of water. Rao based his estimates of fish output for British India for 1925-29 on consumption statistics and for 1930-31 on the estimated average catch per fisherman. For Malaya, Benham took the official records of weight landed and average retail prices and raised the quantity by 25 per cent. Undercoverage was allowed for in the Irish estimates by taking the gross value of landed fish as the net output of fishing. Agricultural output is valued in the ma­ jority of the estimates at published or es­ timated farmers’ selling prices. In some estimates, however, the prices used for valuing-the product of agriculture include other activities. In the Indian estimates, for one, the net output of the agricultural

divisions included processing, marketing, and ancillary activities performed by the cultivator in respect of his own produce. In nearly all estimates, the same prices are applied to products consumed on the farm and to those sold. In the old Polish esti­ mates, a differentiation was made, home consumption of farmers being valued at retail rather than wholesale prices. Various price differentials had to be introduced into the estimates for Soviet Russia to account for different marketing systems for the same products. Since production for home consumption still plays an important role in predomi­ nantly agricultural countries, what value is imputed to farm home consumption de­ serves special attention. The customary im­ putation at farm prices, though consistent with measuring net agricultural output un­ derstates the real income of farmers relative to that of other economic groups, and other­ wise fails to harmonize with the concept of national income as a whole. In Ireland, subsistence farming was of considerable economic importance, and while agricultural prices were used in the computation of na­ tional income, data for farm produce and fuel consumed on farms were given in retail prices. If retail prices were used, the figure of national income would have been in­ creased by about £.16 millions (or about 3.5 per cent of national income) in 1954. Since the gross value of lumber output is reported in the majority of countries, only a few estimators make their own com­ putations. Swedish (prewar) aggregate out­ put was taken as the sum of the values of timber used in factories, used as fuel, and exported. In Yugoslavia, forestry was mostly socialized and data for products sold were taken from business accounts. In Norway, quantity, price, and transport data were good for the amount of timber cut for sale

1 8 . THE N E T O U T P U T M ETHOD

each year. However, the amount of timber and wood cut for use by the forest owners themselves (about 25 per cent of the total in 1946-48) was based on figures from a 1937 census. Where such data are not avail­ able, the estimators have to base value on average yields per acre and average prices. This was true in India, where information on the output of forest products was avail­ able for only part of the country. For the rest of the area, the yield per acre in cov­ ered areas was taken at one third for timber and at two thirds of yield in covered areas for fuel. Although the OEEC does not rec­ ommend it, Norway takes account of forest growth.

b. Mining Figures on the gross value of mining out­ put are currently published in all mining countries either separately or together with other industrial production, and are com­ monly tied closely to census enumerations. However, various estimates are based on other methods when data are not available. For example, J. Saenz, in the official Mexi­ can estimates published in 1946, did not accept the published Mexican values, but made his own estimates by a roundabout process. For each type of mineral he deter­ mined output in terms of metal content, then multiplied the content by price quota­ tions on the New York market. Deductions were made for the cost of items responsible for the differences between the New York quotations and Mexican mining prices.

c. Construction Construction estimates are based on vari­ ous sources, such as reports of building per­ mits issued, surveys of housing starts and completions, reports on the progress of pub­ lic construction, public accounts, statistics of the value of contract awards, and statis­ tics of employment, wages, and materials used. The use of building permit data is fairly widespread, especially in estimating private residential building. Sample studies were used to supplement such data in the United States and in the Netherlands. Re­ liance on censuses of the construction in­ dustry is not common because of the fre­ quently unstable composition of this indus­ try resulting from the existence of many small building contractors. While Ireland’s annual census of industrial production cov­ ers the construction industry, small con­ tractors were covered only on a sampling

269

basis. In France, the construction industry included many units of only one or two workers and others employing thousands. In the absence of production statistics covering the entire industry, France’s 1946-49 es­ timates were based on indexes of construc­ tion activity based on the number of manhours worked.

d. Manufacturing Most censuses of manufacturing or of industry in general (which often include also electric power and gas industries, min­ ing, and sometimes construction) are far from comprehensive. Whether taken an­ nually or at longer intervals, they frequently cover only enterprises employing more than a certain number of workers or utilizing a minimum motive power, or organized and registered as factories or industries receiving favored treatment from the government. The quantity and value of the output of the excluded smaller enterprises, handi­ crafts, artisans’ establishments, and/or rural industries are reported only occasionally in special studies, or the data on them are fragmentary. Estimates of net value are sometimes based on an assumed average gross or net output per worker in such establishments, but it is usually necessary to shift to the income-distributed method, i.e., to base net output on the estimated average income per producer (see Table 18-4). Another method, in countries that have developed input-output accounts, is to use the production accounts for the various industries. These accounts show the com­ position of purchases and sales by economic sectors; and, in the enterprise sector, they show payments to factors of production. These accounts were used in the estimates for Denmark, the Netherlands, and Nor­ way.'* In the French estimates prepared since 1952, the computation of the manufacturing net output was facilitated by the existence of a general tax on value added. The treatment of changes in inventories presents a special problem not only in the case of manufacturing output, but also in that of trade and other economic divisions. Serious attention has not been given this problem until recently, and even then there has been less discussion of it under the net output method than under the other two methods. The main reason for the com­ parative neglect of this problem in net out­ put estimates is that most of the countries employing this method are not advanced

270

THE INCO M E OF NATIONS

T able 18-4. Examples of M ethods of Estim ating Gross and N et O utput in M anufacturing W hen D irect D ata Are Unavailable Country E stim ates 1

E stim ating M ethod

Industries or T ypes of Establishm ent Covered

N e t Output M ethod Gross value derived Value added tax (production tax) data Ratio of net to gross value based on analysis of the operations of corresponding large scale industry Average gross output per worker from per­ sonal investigation Q uantity outp ut and prices

France

Private industry

U .S.S.R .

Small-scale handicraft and dom estic industry

M alaya (Benham)

M ost manufacturing establishm ents Certain artisans

British India (prewar Rao)

Income Distributed M ethod T ax statistics for self-em ployed and average wages for workers Average income estim ated from sample studies and number of workers T ax data plus sample industrial inquiry

Sweden (prewar, Inst. Soc. Res.) India

H andicraft and artisan

Ireland

Num ber em ployed from census of popula­ tion plus total incom e and cost of living index

Ceylon

Consensus em ploying less than three workers Cottage industries

Sm all enterprises

0 Official postwar estim ates, unless otherwise indicated.

industrially and generally lack the necessary accounting data on changes in inventories and in their values. For discussion of this problem in net output estimates, see Chapter 21, Section I-3b (Canada) and Chapter 25, Section 5b (Soviet Russia). In the various censuses of manufacturing and mining, the products actually sold are generally valued at factory or plant prices, but the stock left on hand may be valued either at the price of the goods sold, or at current cost excluding profit, or as the dif­ ference between the book value of inven­ tories at the beginning and end of the year. The problem of treatment of inventories is most easily recognized in connection with the income-distributed method because the calculation of one of the most crucial types of distributed income—profits—depends in part upon how additions to inventories are evaluated. For a fuller discussion of this problem see Chapter 19, Section I-4c and Chapter 20, Section 4. e. Transportation and C om m unication

In most cases, the net output or value added in transportation is estimated directly from the reports of the transport and com­ munication industry by summing their pay­ ments to factors of production. In only a

few estimates—those for Sweden (prewar), Soviet Russia, Norway, Malaya—is the con­ tribution of transportation to national in­ come determined from gross receipts ad­ justed for deductible costs. For the more important transportation services, gross re­ ceipts are obtained from operating accounts, and are usually readily available when a large proportion of these enterprises are operated by the government, as, for ex­ ample, in Norway, France, and Yugoslavia. However, it is usually difficult to estimate the income from the smaller units such as owner-operated taxicabs and, consequently, these services are frequently measured by multiplying their number by an average gross receipt. f. Trade

In the majority of the net output esti­ mates, net output in trade is estimated in one of several ways. One way is to apply to gross sales or receipts, as reported in censuses, tax records, or sample inquiries, an appropriate percentage representing the gross trading profit of wholesalers and re­ tailers. This gross profit is then reduced to the net value of the services by trade by deducting the cost of materials and services of other industries, taxes, and depreciation

18.

THE N E T O U TPU T M ETHOD

allowances, usually expressed as a general or specific percentage of gross receipts. This approach was followed in Soviet Russia, Mexico (1946), Ireland (for profits in trade), and in Canada (for unincorporated retail stores). When the requisite data are available for only one or two years, the figures are extrapolated to other years by an index of sales or of business activity. The second approach, used where no direct data on sales are available, is to com­ pute gross output in trade from the gross or net outputs of industries whose goods enter trade. The gross or net outputs of agriculture, forestry, mining, and manufac­ turing industries are added, after allowing for the value of the output exported or supplied to consumers directly. Then total imports and, sometimes, indirect taxes and freight charges, are added. To the resulting aggregates, representing the value of goods entering trade, the customary markups are applied, yielding gross profit from trade. From this point on, the procedure is the same as under the first approach. This pro­ cedure was followed in the estimates for Bulgaria (pre-1946) and for India (1951) as a check on the income-distributed calcu­ lation. In a few countries, altogether special methods are used. In France, the value added in trade was estimated on the basis of data derived from the records of the transactions tax. The total value of taxable transactions was adjusted for exemptions, evasion, etc.; the value of goods already included in other sectors was deducted; and the resulting figure was taken as represent­ ing the value of transactions originating in trade. In the German postwar estimates, the turnover tax provided the data for income from trade. In Yugoslavia, where trade is almost wholly socialized, the income orig­ inating in trade was computed directly from the published accounts of the enterprises. In India and Sweden (1930-34), income from trade was obtained by using the in­ come-distributed method. Wage and profits data were used, with the latter being often taken from tax statistics.

g. Services of Dwellings Even when the national income estimates as a whole are concerned mainly with the distribution phase, income from dwellings is usually estimated by the net output method. Estimates differ greatly in their coverage of dwellings, statistical sources,

271

and in their treatment of the gross value of the services of dwellings, i.e., gross rent. Most use records of taxes on real estate and rent; housing censuses reporting the number of dwellings and rents either for the entire country or at least for cities; construction statistics, which are generally incomplete; annual indexes of rents; and special surveys. Under the net output method generally, the services of dwellings are considered separately from those of industrial, com­ mercial, and other buildings. They are treated as a separate and distinct industrial division on a par with agriculture, manu­ facturing industry, mining, trade, etc. The services of agricultural, industrial, and com­ mercial buildings are accounted for in the contributions to the national income of the relevant industrial divisions. This sepa­ rate identification of the services of dwell­ ings is recommended today by both the OEEC and the UN industrial standard classifications.5 But in a few instances, be­ cause of statistical difficulties in separating the items, or for other reasons, paid or imputed rent from commercial premises, and even from industrial buildings, is in­ cluded with the services of dwellings. In these cases, the output in the particular industrial division is computed net of such rent. This is done in the estimates for Ire­ land and France. In Norway, in the 194648 estimates, scanty data led to the treat­ ment of income from the ownership of buildings or parts of buildings used by all kinds of service enterprises as a separate industrial division. Imputed rent on the equity of farmers’ houses is sometimes in­ cluded under agricultural income, but the procedure differs rather widely among the different countries. Two methods of estimating services of dwellings are common: (a) gross or net rent is ascertained directly from tax data or housing censuses, or (b) it is determined indirectly from the value of houses and the probable average percentage ratio of gross or net rent to such value, or from the number of dwellings or rooms and the average rent per dwelling or room. Housing censuses generally provide detailed data of gross rents and net rents, or of the rental values of dwellings and the costs of their opera­ tion for each class of dwelling and by na­ ture of occupancy, i.e., whether owner- or tenant-occupied. Inasmuch as housing cen­ suses, unlike tax data, are not annual in

272

THE IN CO M E OF NATIONS

nature, it becomes necessary in the compu­ tation of rent for intercensus years to esti­ mate by projection from the bench-mark years on the basis of related indexes on con­ struction, population growth, etc. Housing censuses or population censuses supplemented with special surveys are used, for example, in Canada, Australia, Norway, Ireland, and Japan; while tax data are used in the United States, France (postwar), Hawaii, and Brazil. In the United States estimates, primary data on rent income are admitted to be inadequate. Lack of gross rent and expense tabulations leads to the use of only the net rent item, which for individual landlords is based on income tax data. However, gross rent from farm prop­ erty is estimated by the United States De­ partment of Agriculture on the basis of acreage and production statistics that are based in turn on Census of Agriculture Re­ ports, crop reports, and a special sample survey made in 1936 that covered 15,000 landlords. For Sweden (prewar), where data on neither the number of dwellings nor aggregrate gross rent were available, the number of rooms was based on sample data on occupancy per room, then multiplied by an average gross rent. In the Swiss (prewar) estimates gross rent from dwellings was estimated on the basis of the total number of rooms and their average rentals derived from the 1930 sample housing census. These bench-mark figures were extrapolated back­ ward and forward by adding the rentals of the newly constructed rooms and adjusting the annual total thus obtained by an appro­ priate rent index. To obtain net rent, de­ ductions were made for rent receipts of corporations, the rental value of farmers’ homes included in farmers’ incomes, mort­ gage interest paid to banks, losses through vacancies, and housing costs. Depreciation, maintenance, and other costs were fixed at 21.5 per cent of gross rental after a study of various housing accounts. Though the same kinds of data are used also in the estimates by the income-distrib­ uted method, their treatment and combina­ tion, and, hence, the estimates themselves differ. Income from the ownership of capital rights to specific property, which is not in­ cluded elsewhere in national income, is accounted for separately. Generally, the coverage of income from rent or real estate in the income-distributed estimates varies widely, depending on the nature of the

sources and the procedures (see Chapter 19, Section I-4e).

h. Services of Financial Intermediaries The net value of the services of banks, insurance companies, and other financial institutions is generally calculated by the income-distributed method as the sum of salaries, wages, and commissions paid to the employees of these institutions, and profits earned by them. In the case of mu­ tual insurance companies, the profits largely take the form of interest credited or divi­ dends paid or accruing to the policyhold­ ers. In the case of mutual savings banks, profits take the form of interest credited to or accruing to the depositors’ accounts. The problems involved in estimating the net value of the output of financial intermedi­ aries are conceptual rather than methodo­ logical in nature, and have been discussed in detail in Chapter 13, Section 3, above. The cash receipts and payments of finan­ cial institutions are generally better recorded than those of any other business, making the statistical data for estimating the net output in their case readily obtainable. Difficulties arise only when the estimators begin to impute income payments to these institutions that are not actually made.

i. Services of Professions and Other Divi­ sions Estimated by the Income-distrib­ uted Method In estimating the net output of profes­ sional, personal, household, domestic, and governmental services, it is usually esti­ mated directly without first calculating gross value, i.e., by adding wages, salaries, and capital income. Such direct calculation is used because information on the factor in­ comes earned in this type of activity is gen­ erally far superior to that on their gross receipts and deductible costs. Occasionally, net output for these divisions is determined as a percentage ratio of gross output or ex­ penditure. For example, in the postwar (1952) French estimates for professions, value added was estimated at 75 per cent of gross receipts as determined from income tax data and special studies; in the German estimates made in 1949, the net output of service industries, such as theaters, sports, health services, etc., was estimated by tak­ ing their gross receipts as derived from the turnover tax statistics and applying to them a net-gross ratio of 25 to 30 per cent.

18.

THE N E T O U T P U T M ETHOD

4. D eductible C osts a. General Approach The data on costs to be deducted from gross value to obtain net value are usually less comprehensive than those on gross value itself. Agricultural production statis­ tics, while reporting typical food crops, hardly ever give the quantity of grain or milk fed to animals or of the seed planted. Data on mining production are usually con­ fined to gross value. Many manufacturing censuses do not give a complete breakdown of the basic costs of production. Even the best of them show only the expenses of the various industries for raw materials and fuel and power used in production. In a few countries, special surveys supply the information on certain other types of production expenses such as advertising, and legal and accounting serv­ ices. This is true, for example, of the United Kingdom, Ireland, and New Zealand. In many censuses, value added is shown gross of many interbusiness services such as in­ surance, advertising, transportation, and communication, as well as taxes and de­ preciation. This is generally done because of statistical difficulties involved in segre­ gating these items, rather than because of any conceptual reasons. As late as 1951, the Canadian Bureau of Statistics pointed out that its figures of net income originating in the various branches of the economy, prepared for the calculation of national in­ come, were generally smaller than the fig­ ures of value added published by the Bu­ reau in its other collations serving other purposes. Few censuses of trade or distribution give the ratios of profit margins, markups, or trading costs to total sales. The full census of distribution in Ireland, which is supple­ mented by annual sample censuses, gives such information. In addition, especially in the less developed countries, census opera­ tions do not usually cover the smaller in­ dustrial concerns and are limited to estab­ lishments of at least a certain size. The financial statistics of the larger transporta­ tion, communication, and financial enter­ prises, however, generally cover operating

costs more or less fully. Because of the difficulties involved in calculating deductible costs, net output in manufacturing in a number of countries is estimated directly as the aggregate of factor

273

share payments. This is done, for example, in the United States, Canada, and to a cer­ tain extent, Ireland. In several countries, the absence of cost data in manufacturing cen­ suses is corrected in part by detailed profit and loss accounts based on tax data or by sample studies covering certain industries. For example, the so-called “production ac­ counts” prepared in Denmark, the Nether­ lands, and Norway supply a considerable amount of such detailed information for various industries.

b. Costs of Materials, Services, and Taxes The procedures used in determining de­ ductible costs vary in nature as well as in degree of detail. Costs are estimated abso­ lutely or as ratios to gross value, either for each item or as a single over-all ratio that may include even depreciation. Deducting specific amounts from gross output is com­ mon in agriculture, forestry, mining, large scale manufacturing, and transportation. Costs are either estimated directly for each division as the sum of its purchases of goods and services from other divisions or they are derived from the production sta­ tistics of the industries supplying the goods and services. Agricultural costs are often set forth in considerable detail. Separate figures are given for the costs of the various feeds and seeds produced, and feed, fertilizers, and other chemicals purchased for agricultural use. Data on feed are derived from produc­ tion statistics, based on sample studies, or computed as residuals after human con­ sumption has been deducted from total pro­ duction. The cost of milk fed to calves is usually estimated as a percentage of the total output of milk. The cost of seed is determined as a percentage of the crop or as a requirement per acre. In general, the computation and enumeration of deductible costs in agriculture varies greatly from esti­ mate to estimate. In some estimates, costs are highly itemized. In those for the United States, for example, they contain as many as forty distinct items; in others, such as Lebanon, the costs of production are broken down into only three items. Among the largest items of production expense in the United States estimates in 1949 were de­ preciation, purchased feed, hired labor, pur­ chases of livestock, costs of operating motor vehicles, net rents to nonfarm landlords, and

274

THE INCOM E OF NATIONS

taxes, which together accounted for four fifths of all production expenses. In some cases, where the deductible costs cannot be computed separately for the different branches of agriculture, an over-all per­ centage deduction is made, either for the whole of agriculture, or for its two main branches, crop production and animal pro­ duction. Thus, in the estimates for Colombia, a blanket deduction of 20 per cent was made for the costs in both categories of agricul­ tural production. For large-scale manufacturing, the cost figures usually distinguish only between raw materials and supplies on the one hand, and depreciation on the other. In a few esti­ mates, the costs of fuel and electric power and light are given separately, since these are reported in manufacturing censuses. It is particularly difficult to obtain information on the costs of small-scale manufacturing units. In the less developed countries, the prevalence of small-scale units makes it diffi­ cult to cover them as part of regular census operations; and accordingly, industrial cen­ suses are limited to establishments of a cer­ tain minimum size, with the smaller units covered only on a sample basis. For ex­ ample, for small enterprises in India (1951), deductible costs were developed from sam­ ple surveys covering 50 different industries. Few estimates, as already stated, com­ pletely accounted for the various interbusi­ ness services such as advertising, insurance, postage, etc. Business taxes are deducted or not, depending on the underlying concept of government services, and on whether the national income is computed at factor cost or at market prices. In the case of other industrial divisions, percentage deductions are particularly com­ mon. Percentage deductions for the value of materials are ordinarily applied to the gross value of buildings and other construction; conversely, many countries depend for the preparation of construction estimates on the value of building materials used and other costs entering into construction. This is true in Austria, Norway, and Ceylon. Net rent on dwellings is usually derived from gross rent by making a percentage deduction for maintenance, depreciation, and other costs, as is done in India and Brazil, and for de­ preciation, as in Australia. Finally, such per­ centages for estimating deductible costs in trade, applied to the traders’ gross profit, are used in some estimates. In some national income estimates, where

deductible costs for specific industrial divi­ sions cannot be estimated even by percentage deductions, they are taken care of by an over-all percentage deduction from the total gross value of the output of all divisions. Frederick Benham, in his estimates for Malaya, had to make block deductions from the gross value of the entire national prod­ uct to account for certain raw materials and other intermediate products consumed in production.

c. Depreciation Of all the costs dealt with under the net output method, depreciation raises the most difficult conceptual and statistical problems. Depreciation is generally understood to mean the annual allowance for the con­ sumption of fixed capital through wear and tear and predictable obsolescence. This an­ nual allowance represents a charge against the current year’s production, and is gen­ erally calculated on the basis of the ex­ pected life of the capital goods and the interest rate, rather than on the actual cur­ rent capital consumption itself. While de­ preciation is an offset to capital consump­ tion over a period of time, keeping the value of the capital intact in the long run, it seldom provides for every intermediate decline in that value. Costs of maintenance and repair are generally charged to current rather than to capital accounts. Losses due to unexpected changes in the life span, in­ terest rates, and prices of capital goods are taken account of by special capital adjust­ ments. Though certain uniform accounting practices have been established, especially under the influence of tax regulations, de­ preciation rates and bookkeeping methods vary widely and are often arbitrary. While the rich literature of business ac­ counting gives the discussion of the princi­ ples and practices of depreciation a promi­ nent place, it is only recently that estimators of national income have had much to say about the problem of depreciation encoun­ tered under the net output method.6 In esti­ mating national income by the net output method and net capital formation as a part of it, an allowance for depreciation is indis­ pensable. Several of the older net output estimates frankly acknowledged defeat in trying to cover depreciation, contenting themselves with presenting a gross figure; e.g., the pre­ war Polish and Turkish estimates. A few countries, which publish figures for depre­

1 8 . THE N E T O U T P U T M ETHOD

ciation, do not compile their estimates from statistical sources, but simply deduct a flat percentage, usually based on the size of the gross national product. This was done in the postwar estimates for the Philippines, Puerto Rico, and Malaya. For most countries, the estimates of de­ preciation shown in the national income product calculations are related to original cost.7 Since the usual sources for these fig­ ures are the allowances claimed by enter­ prises in their tax returns (especially for in­ corporated enterprises), or published busi­ ness accounts, surveys, census tabulations, etc., the provisions very often reflect standard rates of depreciation fixed by law. These rates are based on tax policy considerations rather than on any measure­ ments of the actual rate of capital con­ sumption in the particular branches of the economy, and, from a national economic ac­ counting point of view, they may be wholly inaccurate. In addition, in periods of chang­ ing prices, allowances based on original cost underestimate or overestimate capital con­ sumption. Therefore, the figures taken from business accounts or tax returns are usually unrelated to the economic concept of depre­ ciation. Among countries using the original cost basis of calculation of depreciation are Australia, Canada, and Ireland, and, except for depreciation on farm property, the United States. Only a few countries now calculate de­ preciation on the total stock of capital on a current or replacement basis. Western Ger­ many uses replacement costs calculated from adjusting the depreciation originally calcu­ lated in terms of historical costs; Norway calculates it for most industrial divisions at replacement cost by direct estimation of net capital formation. Other methods are used for agricultural depreciation in the United States and in Ireland. Depreciation may be calculated either as a percentage of invested capital or as a fixed percentage of the gross value of output. The first method is more in line with accepted business practice and yields more stable annual allowances. A variant of the first method was used in the French estimates for the years 1947-51, where depreciation of the enterprise sector was computed as the annual amount “theoretically necessary to maintain the assets intact.” An inventory of assets was taken and classified according to estimated life and type for 1938. This was projected to other years on the basis

275

of the estimated net percentage change in the capital assets and the price index for that portion of capital equipment where ac­ tual maintenance expenditure was not avail­ able. The second method—taking deprecia­ tion as a fixed percentage of the gross value of output—was used in India for some rental income, for mining, and for forestry, where the depreciation percentages were taken arbitrarily. The majority of estimates follow business practices in depreciation accounting and try to approximate the values and rates set by business under the tax laws. Many coun­ tries, the United States and Canada among them, use tax data for most, or for at least part, of their depreciation estimates. Some­ times this information must be supplemented from other sources, as in the Canadian esti­ mates, where depreciation for incorporated companies was based on tax statistics for the recent period, but for years prior to 1944, they were based on a special corpora­ tion sample study. Where data are not available either for the invested capital in the industry or for the rates of depreciation, then either no allowance is made or depreciation is allowed for on the basis of related information. In the Canadian estimates, no allowance is made for forestry, hunting, trapping, or for unincorporated mining enterprises, where the amounts involved are not significant. In Ireland, the expenditure on farm machinery was treated as the allowance for deprecia­ tion in agriculture, even though it was felt that in more recent years this figure may have contained a fair element of new capi­ tal expenditure. Fundamental to any estimate of national income, whatever the method of its calcula­ tion, is the question of whether to follow the traditional accounting practice of meas­ uring depreciation as a percentage of capital or whether to calculate it as a relation to the current output of final goods and serv­ ices. An interesting departure from tradi­ tional accounting practice was made in the Swedish (prewar) estimates. Capital con­ sumption was considered to vary with the gross value of the capital services currently rendered, rather than with the capital in­ vestment. Kuznets deviated from ordinary depreciation accounting in his estimates of the American national income for 1919-38 for the following reason: In business ac­ counting, depreciation is based on the origi­ nal cost of the property, which may be

276

THE INCO M E OF NATIONS

quite different from its current replacement cost, thereby sometimes giving a distorted figure of depreciation allowances required to offset capital consumption. When prices fall, the usual depreciation accounting based on original costs tends to understate the true current net value of output by allowing too much for depreciation; when prices rise, it tends to overstate the true national income because depreciation charges fall short of actual capital consumption. These distor­ tions are similar to those brought about by the usual valuation of inventories, but the adjustments are more complicated. Kuznets used the adjusted figures calculated by Fabricant (or by his method) in which capital assets were classified by years of invest­ ment, to determine the original investment prices; and by their life span, to derive the portions to be replaced annually.8 5. Summation and Final A d ju st­

m ent of the N et Value of O utput When the net value of the output for each economic division has been computed, either by deducting the duplicating costs from gross value or by directly estimating the net values added, all values are summated. In certain cases, however, the re­ sulting total does not quite correspond to the true net value of national output, as it includes some expenses for the interbusiness services, depreciation, and indirect taxes

that the estimators have been unable to esti­ mate and eliminate. Where this is the case, the resulting aggregate represents some form of gross rather than net product. Finally, even with the adjustments indi­ cated, the total of the net outputs of all economic divisions represents only homeproduced income, not national income. It must be further adjusted for the interna­ tional balance of the country’s capital and labor income. All interest, rent, dividends, and other types of capital income forming part of the net output due foreign residents on account of their investments in the country have to be deducted from the home-produced total, and the same types of capital income derived from foreign assets added. Similarly, all labor incomes received from abroad must be added, and those paid to abroad deducted. In almost all estimates, the net balance of international capital income and, in some, also other income is adjusted for by means of an over-all deduction from or addition to total home-produced income. Separate de­ ductions from or additions to the outputs of the various economic divisions are not ordinarily made on this account. The net outputs of the various industrial divisions relate, therefore, to home-produced national income and not to the national income proper. When all the adjustments for the inter­ national account are made, the estimates of national income are complete.

19. INCOME-DISTRIBUTED METHOD The income-distributed method has two variants—income paid out and income re­ ceived. The former measures income pay­ ments in cash or in kind to capital and labor made by the enterprises, governments, or other institutions using their services, in­ cluding payments accrued but not actually transferred. The income-received variant measures the receipts of income, in cash or in kind, chiefly by individuals and often Notes begin on p. 535.

without differentiation of their productive functions as capitalists or laborers. The two variants are usually combined in the same estimate, each being applied to different parts. A good example of such combination is found in the French incomedistributed estimates prepared in 1952-53. Incomes from profits of nonfarm businesses and from dividends were estimated partly from tax data as incomes received, and

19.

IN C O M E-D IST R IB U TED M ETHOD

partly from business accounts as well as from tax data as incomes paid out. Labor incomes, in the case of industries covered by social security taxes and in the case of public enterprises and government services, were estimated on the basis of financial re­ ports and social security tax data as incomes paid out. Each variant uses data from different types of sources. The income-paid-out var­ iant relies mainly on data derived directly or indirectly from general or special finan­ cial reports of the users of labor and capital, such as the published accounts of govern­ ments, railroads, banks, insurance compa­ nies, and corporations, as well as company and institutional returns to social security, census, and tax authorities. In countries hav­ ing withholding taxes* on wages, dividends, interest, or rent, collections are an especially important source. The income-received variant is based pri­ marily on assessed personal incomes or the separately assessed incomes from wages, earnings of the self-employed, or corpora­ tion profits, and less frequently, on special income censuses. Total national income should be identical under both types of calculation, if properly carried out. In the income-paid-out variant the components are factor shares or “types of income,” such as salaries and wages, earnings of the self-employed, dividends, undistributed corporate profits, interest, and rent. In the income-received variant they generally consist of the incomes subject to the income tax plus specified exempt or unaccounted-for income or parts of in­ comes, including allowances for tax evasion. In applying both variants, a short cut is often used, i.e., an estimated average in­ come of the population or an estimated average labor income is multiplied by the number of recipients. Averages are usually confined to filling gaps as, for instance, in social security or tax data. This short cut is most commonly used for estimating in­ come for professions. It is used in Canada, Brazil, the United States, and Japan, among other countries. Typical examples of the income-paid-out variant, which is currently more common than the other variant, are the estimates for the United States, Canada, Ireland, and the Union of South Africa. Typical exam­ ples of the income-received variant are the estimates for Australia by Clark-Crawford

277

for the years 1928-29 to 1937-38, and the official pre-World War II estimates for the Netherlands.1 The totals calculated by the income-paidout variant are adjusted for the interna­ tional balance of capital income. The ad­ justments are similar to those made under the net output method, except that interest, dividends, and profits from branches re­ ceived by corporations on assets held abroad need not be added in the national income totals, as these incomes are usually included in corporate profits. The totals calculated by the income-received variant, if consist­ ently applied to personal income, do not have to be adjusted for the international balance of capital income inasmuch as they cover the incomes from abroad of residents. However, the extent of this inclusion de­ pends upon the tax regulations and inter­ national tax agreements. In any case, per­ sonal income taxes are not concerned in any way with incomes paid to foreigners. In­ terest received by government from abroad or paid to foreigners by government is ac­ counted for in connection with the supple­ mentary estimate of government capital in­ come.

7.

Income-paid-out Variant

1. Classification by T ype of Income The classification by type of payment or receipt varies according to the way in which the statistics of a nation are organized (Table 19-1). The variations are not as extensive as those to be found in the classi­ fication of economic divisions. Economists and statisticians differ less about the nature of the main distributive shares than about the main economic divisions. Accordingly, such differences in classification as obtain are primarily concerned with the degree of detail in itemizing the different types of payment or receipt. The ideal or most comprehensive classi­ fication consists of eight types of income: (1) wages and salaries (including commis­ sions); (2) supplemental labor income (so­ cial security, pension, and other contribu­ tions of employers to employees’ welfare); (3) earnings of self-employed; (4) divi­ dends; (5) undistributed corporate profits; (6) interest; (7) rent; (8) profits of gov­ ernment enterprises. A less comprehensive

278

THE INCOM E OF NATIONS

Table 19-1. Classification of Incom es by T ype of D istributive Shares in Three Selected Country Estim ates

United States Categories of Income Labor income: Wages and salaries Supplemental labor income (Income of unincorporated business) Earnings of self-employed Capital income: Dividends Undistributed corporate profits Interest Rent Government profits

x x x

x x x x 0

United Kingdom

Canada

x x x

cp or i + x x x

x

w x

r

i

- f r or

cp

c c c c

x: Separate presentation of income by the type cp: Combined with corporate profits w: Combined with wages and salaries i + r: Combined with interest and rent c: Combined with other capital income 0: Excluded s o u r c e s : United States, National Income, 1954; United Kingdom, National Income and Expenditure, 1946-1953; Canada, National Accounts, Income and Expenditure, 1936-1950.

but still logical classification consists of three types of income: labor income; mixed income (that of the self-employed); and capital income (see Chapter 11, Section 4). A few estimates, such as those of the United States, Canada, Australia, and the United Kingdom, show almost all the major types of income and their various components. For example, in the United States, interest income is divided into monetary and im­ puted interest for each of the following categories: private business, households and institutions, and government. In some esti­ mates, the classifications are exceedingly rudimentary or even crude. In the French estimates of 1952 for 1949, interest, divi­ dends, undistributed corporate profits, and income from unincorporated enterprises were lumped together (see Chapter 27, Section II-2); in New Zealand, dividends were combined with corporate savings. The simplest classification, used in the official Mexican estimates for the years 1946-50, consisted only of compensation of em­ ployees, rent and interest, and other in­ come.2 Some countries cross-classify income by type and by economic division. Such classi­ fications are likely to become more common in the future because of the usefulness of such information and its increasing availa­

bility. They are given, for example, in the estimates for the United States, the United Kingdom, and Canada.

2. Labor Income Wages and salaries consist of remunera­ tion in cash and in kind of employees who are normal residents of a given country, computed before payment of taxes and de­ duction of social security contributions. So­ cial security records, together with work­ men’s compensation statistics, play an in­ creasing role in estimating labor income. They are the most complete of any wage data, and total wages can be compiled from them currently with relatively little supple­ mentary estimating. In addition to compre­ hensive coverage and regularity of reporting in most countries, such data are very re­ liable in that they are “self-policing,” in the sense that the employees are allowed to check on their employers regarding the re­ porting of their wages. Such data are used extensively in the United States, the United Kingdom, Belgium, and the Netherlands.8 Censuses of industry or related surveys form the main basis for the estimates of labor income in Canada, Ireland, and the Union of South Africa. In Canada in 1949, approximately 70 per cent of the total labor

1 9 . IN CO M E-DISTRIBUTED M ETHOD

income estimate was taken directly from census of industry reports, annual surveys, or published statements. Industries not covered included agriculture, construction, transportation, wholesale trade, and other sectors of small importance. In Ireland, the annual census of industrial production covers industrial firms in which three or more persons are engaged. These data are supplemented by quarterly voluntary in­ quiries and sample surveys. Statistics of taxation are sometimes the main source of information on wages and salaries. Payroll tax data in Australia are available for all enterprises with a payroll in excess of a stated amount. For nonrural enterprises approximately 80 per cent of compensation is covered by this source.4 However, taxation data are usually used to supplement social security records or other types of data for estimating salaries and wages exempt from the social security tax. The figures of salaries and wages paid by the national government are derived from budgetary reports and are relatively firm. Estimates of salaries and wages paid by smaller jurisdictions are based on periodic censuses, special investigations, or fragmen­ tary statistics and generally are only ap­ proximate. Where data on aggregate wage and salary payments are lacking, incomplete, or avail­ able for certain years only, it becomes nec­ essary to estimate wages and salaries by some type of special procedure. Under these circumstances, data on the labor force are combined with those on salary and wage rates. Population censuses often furnish the principal bases for such estimates. In addi­ tion to yielding data relating to the size and composition of the labor force, such cen­ suses sometimes also provide information about employee earnings. In Brazil, for example, wages in domestic employment and in service industries were calculated by taking the number of employees as reported in the 1940 census, extrapolating it to later years, and multiplying the resulting figures for each year by the legal minimum wage adjusted for changes in the cost of living. In the United States, the average cash earn­ ings of domestic servants were based on bench-mark figures taken from the 1940 and 1950 censuses of population, which showed the frequency distribution of such persons by detailed size-of-earnings classes. This indirect method plays a subsidiary

279

role for estimating labor incomes not cov­ ered by the main source. For example, when social security and workmen’s compensation data are the main sources, the incomes of workers not covered by these systems, such as farm laborers, domestic and public serv­ ants, workers in home industry and small shops, and certain other specially exempted employees are estimated indirectly with the aid of information on their numbers, aver­ age earnings, etc. (The use of such data for interpolation and extrapolation is described in Chapter 17, Section 3a.) Whether the estimate of labor income is aggregated from several sources or taken largely from one source, it may often be incomplete and may differ in its reliability from country to country. Certain categories (incomes from incidental jobs, gratuities, and tips) are frequently excluded either be­ cause of lack of data or sheer oversight. A special estimate of tips is made in the United States, since it is believed that actual coverage under the social security laws is small. To the extent that tips are covered, it is felt that there is no duplication since they offset any exclusions of income in kind. Pay and allowances of members of the armed forces are generally shown separately from other labor income. War bonuses, pen­ sions, service grants, and similar payments are not usually included under this cate­ gory, being regarded as transfers. Income in kind is ordinarily limited to food and standard clothing. The data are taken from the reports of the military establishment and of the general budget. Compensation in kind of domestic and farm workers, hotel, restaurant, and institu­ tional employees, ship crews, and members of the armed forces is generally included in the estimates. The value of board and lodg­ ing is computed either at their cost to the employer or at the prices the employees would have to pay for like items. The OEEC recommends the inclusion of these items at an amount corresponding to the cost to the employer.5 This approach has been adopted in the United States where the value of board furnished to domestic servants is de­ rived from a standard budget developed originally in a special study for 1935-36 and supplemented by later investigations made by the National Income Division.6 Supplementary labor income is computed either as the sum of employers’ contribu­ tions to private pension funds, compulsory

280

THE INCOM E OF NATIONS

industrial insurance systems, welfare insti­ tutions, old age, and other social security insurance, or as the direct payment of so­ cial security and veterans’ benefits. In both cases the amounts are ascertained from the accounts of the agencies administering the programs, from sample accounts of enter­ prises, or annual surveys. In the United States, other labor income also includes pay of military reservists, government payments of enemy prisoners of war, merchant marine war-risk life and injury claims, directors’ fees, jury and witness fees, compensation of prison inmates, and marriage fees to jus­ tices of the peace. This miscellany of items amounted to a comparatively small sum.

3. M ixed Income This item includes income from farms and other agricultural enterprises, income of other sole proprietors and partnerships, and professional earnings. Differentiated in­ come from the ownership of land, buildings, and financial assets is usually not included under this category, but is recorded in the rent and interest shares of income. Farmers’ income is usually computed by deducting wages of agricultural labor as well as in­ terest and rent paid from the net value of agricultural output.7 Where income of nonfarm unincorpo­ rated enterprises is separately shown, nu­ merous sources are used to build up the fig­ ures. In the United States, comprehensive data existed for only 1945 and 1947. For these two years, internal revenue tabula­ tions covering sole proprietorships and part­ nerships were based on an especially large number of returns because of the compara­ tively low income exemptions and high levels of business activity. The year 1939, covered by an industrial census, also pro­ vided important bench-mark data. Profit ratios obtained from taxation and census sources were used to obtain net income from gross receipts for nearly all the larger industries. For the rest, the number of pro­ prietors was multiplied by an estimated average net income or by an average gross receipt reduced to net income by means of a profit ratio. In Canada the income of nonfarm unincorporated enterprises is simi­ larly estimated from a great variety of sources. The income of professions is generally estimated by multiplying an average income

for each kind of profession by the number of its practitioners.

4. Capital Income “Pure” capital income consists of divi­ dends, paid and imputed interest, paid and imputed net rent; undistributed profits of corporations and other private organiza­ tions; and profits earned by governmental enterprises. The addition of all dividends, interest, and rent paid out would lead to some dupli­ cation. Only the difference between the re­ ceipts and payments of these types of capital income by the entities that are both lenders and borrowers can be credited to the fac­ tors of production in national income com­ putation. Dividends received by corpora­ tions may eventually be paid out in the form of interest on bonds and savings de­ posits. Net rent is usually divided into land­ lord’s equity and mortgage interest, and the latter may finally appear as interest on mortgage bonds and savings deposits and operating costs of credit institutions. Since these different types of capital income are not estimated at the same stage, the classi­ fications are often not comparable. In many estimates adjustments for such duplications are difficult to make, since cor­ porate savings or undistributed profits are shown before the distribution of dividends and the payment of direct taxes. Only the more important corrections of possible du­ plications are generally undertaken.8 Only in the United States and in a few other countries are tax data rich enough to make elaborate adjustments for these duplications possible. The correction for duplications is gener­ ally made either for capital income taken as a whole or for the interest payments and the dividend payments taken separately. In both cases, it is done by deducting re­ ceipts from payments for each capital using and lending unit. Similarly, either of these two treatments may be used in adjusting the national income total for the net balance of foreign capital income. While dividends and interest received by governments are gener­ ally included in capital income paid, nearly all countries now exclude interest payments on the national debt from national income on the ground that they do not originate in production. Some countries, e.g., Canada, impute an interest to government-owned

1 9 . IN C O M E-D IST R IB U TED M ETHOD

productive assets and include this portion as income.

a. Dividends Dividends are computed as paid to house­ holds and to collectives (governments and nonprofit-making organizations). Intercorpo­ rate dividends are not considered, inasmuch as they are not a part of the national in­ come. Dividends paid by corporations, lim­ ited liability companies, and co-operative business organizations are usually estimated in conjunction with corporate profits. They are usually computed from three kinds of sources: reports of corporate income taxes; reports of special taxes on income from stock and other securities; or sample cor­ porate and other accounts. The third source plays a subordinate role, since it is usually available only for larger enterprises. In esti­ mating dividends, adjustments should be made for stock dividends that are not cur­ rent income and for fiscal and calendar years that do not coincide.

b. Undistributed Profits Undistributed profits (calculated after payments of dividends, interest, transfers, and direct taxes) include changes in the sums set aside by firms for future tax pay­ ments. Kuznets’ estimates for the United States for 1919-38 9 and the Dominion Bu­ reau estimates for Canada up to 1942 in­ cluded savings of unincorporated business and of government. While in some estimates net profits of governmental enterprises are shown separately, they are usually included with the undistributed profits of private companies. In almost all estimates, corpo­ rate profits are estimated mainly on the basis of corporate income tax returns. Sam­ ple accounts of corporations may be used for estimating this item or used to supple­ ment the estimates based on tax returns. They were, for example, used as the basis for estimates of corporate profits in Switzer­ land in 1944. In the United States, sample data are utilized pending the publication of tax return tabulations (which do not be­ come available for over two years after the year to which they refer); in Canada a spe­ cial sample study of corporate tax returns was used to estimate profits for the period 1926 to 1944, before complete tax data became available. Since the definition of profits for national accounting purposes normally differs from that employed by the tax authorities, vari­

281

ous adjustments in reported figures have to be made. These adjustments vary greatly, with the result that the estimates of corpo­ rate profits are not comparable among countries. The different possible adjustments include: adding back excessive depletion al­ lowances permitted in tax returns; eliminat­ ing capital gains and losses from the calcu­ lation of profits that do not measure income arising in current production; adjusting re­ ported depreciation allowances and changes in inventory values; and subtracting profits accruing to foreigners.10 Aside from the fact that adjustments of undistributed profit figures vary among esti­ mates, a more fundamental reason for the present incomparability of estimates is the fact that the underlying data reflect varying accounting practices and tax regulations. Gains or losses due to changes in prices merely deflate or inflate accounting profits. The usual business accounting may distort profits and, therefore, business savings for national income estimating purposes in three separate cases: 1. When depreciation allowances are charged against gross income, usually at the original cost of the properties, which may be quite different from their current re­ placement cost; 2. When inventories are evaluated at book prices that may be higher or lower than the actual cost entering into the value of production. This discrepancy arises espe­ cially when the “first in, first out” method of valuation is followed; 3. When realized gains or losses from the sale of capital assets, such as securities or tangible properties, are treated as addi­ tions to or deductions from current profit, and are so carried into the national income computations. To obtain true profits from the reported figures it becomes necessary to adjust the latter for some or all of these three types of distortion. The methods of calculating depreciation allowances in national income estimates were treated in the preceding chapter (Chapter 18, Section 4c). Here we may touch a bit more fully upon the meth­ ods of calculating the value of changes in inventories.

c. Changes in Inventories The practices generally followed by busi­ ness firms in accounting for changes in the value of their inventories from one year to another may be appropriate for

282

THE INCOM E OF NATIONS

business or tax purposes, but they gen­ erally are not appropriate for national in­ come estimating purposes. Business firms generally carry their inventories on their books at the prices actually paid for them, i.e., at their historical costs rather than at their replacement costs that represent their current true value. Changes in the book values of the inventories reflect, therefore, two things: (a) changes in the volume and quality of the inventories (i.e., the difference between the quantity and quality of the used-up inventories and the quantity and quality of the newly added inventories); and (b) the difference between the prices at which the used-up inventories were acquired originally and the prices at which the new inventories were acquired. The calculation of profits is affected by both these factors. This method of evaluating inventories results in a wholly misleading picture of profits in times of rapidly chang­ ing prices. When prices rise, the changes in book values of inventories calculated at original cost lag behind the rise of their replacement values, causing profits to be undervalued. In times of declining prices, the changes in book values of inventories so calculated lag behind the decline in their replacement values, resulting in an over­ valuation of profits. For purposes of national income account­ ing, inventories should be valued at their replacement values, and not at their original cost. This would make the computation of the changes in their values conform with the changes in the values of all other ele­ ments in national income that are computed at current cost. The problems of correct evaluation of inventories cannot be avoided by any shift to some type of gross evaluation of the national income, as is done in the case of depreciation, because the year-to-year changes in the values of inventories, whether caused by a change in the volume of the inventories or by a change in their prices, are generally much greater than the yearto-year changes in the consumption of fixed capital or in its replacement costs. They affect the computation of annual profits much more sharply. Among the various methods currently used for accounting purposes in valuing inventories, two may be mentioned: the First-In, First-Out (FIFO) method and the Last-In, First-Out (LIFO) method.11 Under the first method inventories are charged to

cost of sales in the order of their acquisition. If current prices exceed book cost prices, an inventory gain occurs that represents the price rise and not the change in quantity and accordingly profits must be reduced. Similarly, an inventory loss occurs when book prices exceed current prices and the inventory valuation adjustment is negative. Under the cost-or-market method, asso­ ciated with the FIFO method, year-end in­ ventories are written down by businesses if market prices are below cost prices. This practice generally necessitates a revaluation of inventory changes for national income purposes.12 The LIFO method of inventory account­ ing comes closest to national income valua­ tion procedures. Usually, revaluation of the book value of inventories is necessary only when the physical volume of inventories de­ creases. “As long as the physical volume of inventories is increasing, inventories used up represent, according to the LIFO conven­ tion of assuming that units acquired last are charged out first, current acquisitions valued at current prices.” Then “there is no differ­ ence in this case between the LIFO and national income methods of inventory valua­ tion.” 13

d. Interest The OEEC defines income from interest as including “all actual interest payments receivable by individuals and nonprofit in­ stitutions as well as imputed interest on life-insurance policies and bank deposits . . . [including] interest on government bonds. . . .” 14 The actual treatment of in­ terest income, however, often does not ac­ cord with this ideal. The estimates of interest flowing to in­ dividuals differ according to whether or not and to what extent duplicated payments are excluded, and also according to the quality and coverage of the country’s basic statistical data. Duplicated payments may be illustrated by the case of individuals who are both creditors and debtors and who, while earning interest on their investments, have to pay out of it interest on their loans, so that their interest income must be com­ puted as a net item representing the differ­ ence between their interest receipts and in­ terest payments. Interest accruing to housholds is often not shown as a component of national income. This is true particularly in those countries (United Kingdom, New Zealand, etc.) in which profits of business

1 9 . IN CO M E-DISTRIBUTED M ETHOD

are calculated before deducting interest paid.15 In these cases, the interest payments to households appear in business profits as a factor share. To record them in the in­ come of households as well would be double counting. If mortgage interest is left in net rent originating, as is done in some countries, it obviously need not be included in the computation of interest income. But if it is deducted from net rent, as is the case in the United States and Canada, it must be shown in interest income as a part of the interest received by mortgage holders. In the latter case, mortgage interest received by insurance companies, social security funds, nonprofit organizations, and personal mortgage holders has to be estimated. Mort­ gage interest receipts together with all other interest receipts must be reduced by offset­ ting interest payments so as to obtain the net interest income. Most national income estimates include interest on consumers’ debt as a factor share in the computation of national income. In the United States and other countries that regulate consumer borrowing, calculation of this item is based on the fairly compre­ hensive reports of banks and other lending institutions on the total amount of their consumer loans. In a number of countries, interest on that part of the national and local govern­ ment debt that is incurred on account of productive assets is included in national in­ come, under the category of capital or in­ vestment income, as a factor income. This treatment, adopted, for example, in the Canadian estimates, is theoretically correct. On the other hand, interest on war debts or unemployment relief debts is generally not included on the theory that it is not a factor income representing a currently ren­ dered service (for further discussion see Chapter 12, Section le ). Two methods may be used for the estima­ tion of interest. The various kinds of in­ terest payments or receipts may be distin­ guished and the amounts of each may be estimated separately, as in Canada, or the interest total may be built up from income tax data, as in the United States. While bond interest and other interest paid by business enterprises is sometimes included in censuses of activity, it is more usual to rely on tax sources for estimates of interest. The United States estimates treat income from interest in a particularly elaborate

283

manner. Hence, illustrations on the treat­ ment of this item are drawn particularly from the United States practice. In the United States estimates, the main source of the interest figures are income tax returns. Interest accruing to persons can­ not be estimated directly, because of the impossibility of obtaining such information directly from individuals. Therefore, they are measured as the payments less the re­ ceipts of relevant payer groups. The pro­ cedure followed amounts to estimating total interest paid by United States business and persons plus total interest paid to the United States by nonresidents, minus total interest received by United States business and total interest received by nonresidents from the United States. Personal interest income measures total monetary and imputed in­ terest paid to United States persons.16 Monetary interest paid by corporations, which accounted for about one third of total monetary interest paid in the United States in 1950, is based for most years almost en­ tirely on federal corporation income tax returns. Interest paid by unincorporated business, divided almost evenly between farm and nonfarm proprietors, amounted in 1950 to about 12 per cent of monetary interest paid. These and other types of interest paid are treated as follows. In general, farm interest is estimated by multiplying the amounts of different types of debt outstand­ ing by the relevant interest rates. The data are obtained from Department of Agricul­ ture estimates based on Census of Agricul­ ture bench marks and sample reports from special farm-lending agencies and commer­ cial banks. The interest paid by nonfarm unincorpo­ rated business is obtained “as the sum of industry estimates.” From 1929 to 1941, the estimates were derived by multiplying the 1939 ratio of interest paid to total re­ ceipts of all proprietorships and partner­ ships in the industry (based on tax sources) by current total receipts. Where figures on total receipts were unavailable, interest paid per partner, calculated from partnership re­ turns, was multiplied by the total number of active proprietors. For more recent years, internal revenue tabulations, itemizing in­ terest paid by industry for both partnerships and sole proprietorships, are used to estab­ lish nonfarm bench marks; the series on corporate interest paid is used for extrapo­ lation.

284

THE INCOM E OF NATIONS

For other private businesses— nonfarm landlords, mutual institutions, nonprofit or­ ganizations, such as farmers’ co-operatives —interest paid accounted for about 33 per cent of monetary interest paid in the United States in 1950. For mutual financial inter­ mediaries, various banking statistics were utilized, such as Federal Deposit Insurance Corporation data for insured banks, reports of the American Bankers Association, etc. Income tax statistics were used for nonprofit organizations. The monetary interest pay­ able by nonfarm individual property own­ ers, the largest item, was estimated mainly by applying the relevant rates to different classes of mortgage debt. Monetary interest paid by households and institutions (20 per cent of the total in the United States in 1950) included interest on personal debt and on borrowing against life insurance policies. It was based on Federal Reserve Board and life insurance data. Interest paid to United States residents by foreigners is estimated in connection with the official balance of payments estimates. The United States estimates include im­ puted interest payments that are associated principally with the transactions of financial intermediaries (see Chapter 13, Section 3). Imputed interest of commercial banks is measured as the excess of interest and divi­ dends received over interest paid on de­ posits. This is based on federal banking and deposit insurance data that are blown up to cover all commercial banks on the basis of their relative assets. Imputed interest of life insurance companies is measured by interest, dividends, and net rents received, and is treated as accruing directly to policy­ holders over and above an assumed payment covering operating expenses. Mutual finan­ cial institutions other than life insurance companies are similarly treated. For com­ mercial banks, imputed interest payments are based on data from the Federal Deposit Insurance Corporation and from the Fed­ eral Reserve System. They are measured as the excess of interest and dividends received over interest paid on demand and time de­ posits. Imputed interest payments, even in such highly elaborate estimates as those of the United States, involve many questionable assumptions and are far from reliable. Monetary interest received by business includes interest receipts of corporations, mutual financial intermediaries, unincor­ porated security and commodity brokerage firms and certain proprietors engaged in

finance, and nonprofit organizations servic­ ing business. In the United States, the first two accounted for over 98 per cent of this general category of interest in 1950. In­ terest received by foreigners from the United States is also accounted for. The sources used for these estimates are generally sim­ ilar to those listed under monetary interest paid. Imputed interest received is estimated on the basis of certain assumptions regarding the nature of the groups receiving it. Since data on the ownership of the funds by use of which the financial intermediaries ob­ tained property income is not generally available except for commercial banks, it is assumed in the United States estimates that the flows of imputed interest of all interme­ diaries except commercial banks go entirely to persons. For commercial banks, imputed interest is allocated among recipients accord­ ing to ownership of deposits, with federal banking statistics and surveys of liquid as­ sets providing the main information.

e. Rent Net rent under the income-distributed method is a very different item from the rent considered under the net output method. The latter is associated with the value of the services of dwellings, and includes in addition to the net rent from the ownership of the property, the value of the services of labor and other factors of production involved in the rendering of these services. Net rent under the incomedistributed method, on the other hand, is a capital share derived solely from the ownership of land and buildings. The coverage of rent varies. While the majority of income-distributed estimates in­ clude under net rent both cash and imputed rents due to households and private nonprofit-making institutions, rent on the own­ ership of farm and owner-occupied business buildings is generally not included under this heading. It is not so included because it is reflected in the profits of the farms and businesses as computed for national in­ come purposes. Thus, the Canadian esti­ mates include under rent due to persons, both residential and business rents, and rent on farm dwellings and buildings. Business rents include only paid rents, as imputed rents on owner-occupied business premises are implicitly included in the estimates of profits and net income. The main sources of data used in esti­

1 9 . IN C O M E-D IST R IB U TED METHOD

mating this factor share are tax returns and censuses or special surveys. Tax data in­ clude either income tax returns or records of taxes on property. Tax data are used in the United States and Brazil; census and survey data, in Canada and Australia.

II. Income-received Variant 1. General Characteristics and Application Under the income-received variant of the income-distributed method, national income is usually based on income tax data, since the alternative sources—sporadic income censuses or sample surveys—are not as com­ plete or dependable. This discussion will not be confined, however, to the problems arising out of the use of personal income tax data for estimating income; it will also touch upon those that present themselves in using other types of tax data. Tax data differ widely with respect to the tax covered, their general arrangement, and their adaptability for estimating national income. Although tax reports ordinarily give the number of taxpayers and the amounts of taxable incomes as well as figures of the tax collections themselves, occasionally they give only the latter information. In such cases, aggregate taxable incomes must be estimated indirectly from the tax collec­ tions by applying effective tax rates and making other adjustments. This procedure was used in the case of data on wage and salary taxes in the German estimates for the years 1925-31 (Chapter 26, Section I-3b). Tax data may be used in national income estimates in combination with other sources. In the Canadian estimates, gross revenues of unincorporated manufacturing establish­ ments were taken from census of industry returns, and were netted by applying a netgross ratio derived on the basis of tax data. In the United States estimates, income from unincorporated business were estimated, in part, on the basis of tax returns for 1945 and 1947 and were next interpolated and extrapolated to other years on the basis of census and other data. Tax data naturally reflect the tax system of the country. A general income tax is levied either on aggregate personal incomes or, under a system of tax schedules, on different types of income. These schedules may in­

285

clude a general tax on large incomes and withholding taxes on wages and salaries and on income from securities. Capital income in the form of corporation profits, interest, dividends, and rent is often subject to spe­ cial taxes. The appropriateness of tax data for esti­ mating national income depends largely upon the statutory tax exemptions. Only data on taxes that allow for relatively small or no exemptions are well suited. Data on other taxes either require supplementary estimates and adjustments or are altogether unusable. Personal income tax data come nearest to fulfilling these conditions, but even they fall short of the mark in most countries. It is significant that in no coun­ try are tax data the sole basis for national income estimates today. Not even in the prewar Dutch estimates (Chapter 23, Sec­ tions 3 and 6), which represent the most perfect application of the income-received variant, were income tax data the basis for much more than 58 per cent of total income (1938 figure).

2. Adjustments of Income Tax Data In trying to derive national income from income tax data, many adjustments are usu­ ally necessary. Some involve the addition of the incomes or portions of incomes that are exempt from or escape taxation, and hence are not reported; or the deduction of income that in the estimators’ opinion is a mere transfer and, hence, does not be­ long in national income. Other adjustments are more technical, entailing corrections for overlapping with other data or shifting the figures from the year of assessment of the income to the year it was earned. The exact nature of the adjustments differs with the country’s tax regulations and the nature of the national income concept. Table 19-2 shows the usual adjustments.

a. Additions to Assessed Incomes Logically, the adjustment of tax data must begin with adding back the tax exemptions deducted from assessed income in the first instance. These exemptions, in the case of the personal income tax, are usually fairly extensive, but vary greatly from country to country. They are much more restricted in the case of business or corporate profit taxes, and usually do not apply in the case

286

THE INCO M E OF NATIONS

Table 19-2. Adjustment of Personal Income Tax Data Additions to Assessed Incomes Personal exemptions Deductible expenditures Incomes below the tax exemption limit Tax evasion Employers’ social security, retirement, and welfare fund contributions Undistributed profits of corporations Capital incomes of religious and charitable organizations Capital income of government and social insurance funds Deductions from Assessed Incomes Income transfers Business losses Duplication with other statistics Other Adjustments (Additions or Deductions) in Assessed Incomes Capital gains or losses due to changes in price Revisions of assessments by subsequent audits Time lag between year of assessment and year the income was earned

of withholding (coupon) taxes on interest and dividends. Thus, the adjustment for tax exemptions is, in the main, confined to personal income taxes, and even then is appropriate only if the assessed incomes are given after exemp­ tions. Such adjustments in assessed incomes had to be made, for example, in the prewar Australian17 (Clark-Crawford) and Ger­ man estimates.18 Tax exemptions for single or married taxpayers and their dependents, chiefly chil­ dren, are given in the tax law; and the number of taxpayers is usually reported in the tax statistics. The marital status of tax­ payers and their average number of children is derived either from administrative tax data or from population censuses. For esti­ mating other tax exemptions or deductions of special expenditures from income, the data are either taken as reported (as in the case of deductible tax payments) or are specially estimated (as in the case of exempt insurance premium payments in the German estimates for 1913). Of all additions to assessed incomes, those for incomes below the tax exemption limit are ordinarily the largest. In fact, this sum often equals or exceeds the aggregate of all the other ex­ emptions that must be added back. In coun­ tries where only large and medium incomes are subject to income tax, tax data are of relatively little use for estimating purposes, and other sources, such as social security or business accounting data on salaries and wages must be used as was done in the prewar estimates for the United Kingdom,

France, and Australia (Clark-Crawford). When smaller incomes are not subject to an income tax, the exempt incomes are usu­ ally estimated by multiplying an estimated average income by the number of recipients after deducting the number of taxable re­ cipients in this group. A further correction sometimes needs to be made for the separate incomes of wives, which are assessed jointly with those of their husbands. Especially re­ fined procedures were followed in Germany before the war in estimating salaries and wages below the tax exemption limit for 1926-31. Wages in the same occupations in urban and rural regions and of juvenile and adult workers were estimated separately (see Chapter 26, Section I-3b). Aggregate incomes below a certain tax exemption limit do not bear a constant pro­ portion to assessed incomes. In years of depression and general deflation incomes fall below the exemption limit and tend to remain there; e.g., in the Netherlands in 1929, 20 per cent of assessed incomes were below the tax exemption limit; in 1932, 43 per cent; and in 1938, 48 per cent. Ob­ viously, under such conditions the deter­ mination of the average and total tax-exempt income becomes difficult, inasmuch as the decline in the previously tax-exempt income is partly offset by the addition of incomes that were formerly above the tax exemption limit. Understatement of incomes through tax evasion, with or without the connivance of the authorities, is important in the case of the earnings of self-employed persons and

1 9 . IN CO M E-DISTRIBUTED M ETHOD

capital incomes generally. Business firms may seek to understate their profits in their accounting statements for tax purposes, and security holders may not report their in­ terest and dividends in full. Moreover, small security owners often conceal their capital incomes completely. The adjustments for tax evasion, usually based on judgments of authorities, are in the form of percentage additions to assessed incomes, ranging from 10 per cent of total assessed income in the Dutch estimates for 1921-38 to 20-25 per cent of entrepreneurial and corporate in­ comes in the prewar German, and 50 per cent of the incomes of artisans and small businessmen in the prewar Hungarian esti­ mates (von Fellner). No special problems not already dis­ cussed are involved in making additions for employers’ contributions to social security, retirement, and other welfare funds; for capital incomes of government and social security funds; or for undistributed corpo­ rate profits. However, special problems do arise, in the case of corporate profits, as far as the data are derived from corporate income tax returns. Discrepancies between the profit concept used for tax purposes and for estimating national income must be taken into account (see Sections I-4a and 4b above).

b. Deductions from Assessed Incomes If transfer incomes (annuities, social se­ curity benefits, interest on the national debt) are taxable under the income tax and are included in the tax returns, their amounts must be estimated and deducted from the assessed incomes in computing national in­ come at factor cost. If they are not taxable and are not included in the returns, no adjustment need be made. Correction of the income totals afforded by tax data on business losses, however, calls for special procedures, depending on how such losses are taken care of under the income tax law. For example, if under the law the profits are reduced by a carry­ over of losses from preceding years, such offsetting items must be deducted from losses already accounted for in another way. Corrections are often necessary for du­ plications of income as between incomes reported in tax returns and those taken from other sources. In the prewar Swiss estimates duplication was eliminated be­ tween incomes from nonbusiness property

287

estimated as such and corresponding items included in the incomes of nonagricultural self-employed persons.

c. Other Adjustments to Assessed Incomes Several types of adjustment may, in some cases, involve an addition to the income totals based on tax data, and, in others, involve a deduction. Business profits as reported on tax returns, as is well known and was noted above, do not necessarily coincide with business accounting practice or with economic thinking. Under tax regu­ lations capital gains due to changes in price are considered taxable income, but capital losses due to such changes may be deducted from profits. Tax regulations define de­ ductible business costs, depreciation charges, and prescribe how inventories are to be evaluated. Consequently, business profits re­ ported on tax returns may exceed or fall short of the profits that constitute part of the current value of the output. Such dis­ tortions can be properly adjusted for only in the light of the tax regulations. When tax assessments are revised by the tax authority in later years, the original fig­ ures should be corrected. Underassessments of corporate profits are corrected in the re­ vised official United States estimates; over­ assessments of wages and salaries, in the German prewar estimates up to 1931. Since there generally is a lag between the year in which the taxed incomes are earned and that in which the taxes on them are collected, an adjustment for this lag must be made in the tax data used for estimating purposes. Thus, in the United States, there is a time lag of one to two years in the collection of corporate profits taxes, which makes necessary the extrapolation of tax data for a corresponding period for national income estimating purposes. “Such extrapo­ lation,” say the United States estimators, “has at times led to errors of significant size, particularly on an industry basis.” They reported in 1954, however, that “the situation in this respect has improved due to progress in the current reporting of cor­ porate profits through the joint surveys of the Securities Exchange Commission and the Federal Trade Commission.” These surveys enabled the estimatbrs to revise their earlier estimates of manufacturing profits since 1947.19

20. THE EXPENDITURE OR FINAL PRODUCTS METHOD independently of other estimating methods or in conjunction with them. Where it is used independently, it affords a convenient check on the accuracy of the other methods. In the United States and the United King­ dom, it was found that the two independent estimates of the gross national product— one calculated as the sum of national in­ come, indirect taxes, and depreciation, and the other calculated as the sum of national consumption and investment expenditures— were very close: in the years after World War II, the discrepancy amounted to less than 1 per cent of gross national product.6 Since the expenditure method measures national income at the disposition stage, it uses market prices rather than factor cost. However, it can be made to measure the disposition of national income at factor cost by deducting indirect taxes less subsidies from the final sales value of each product group. This procedure was followed in the British wartime and postwar estimates for each category of consumer goods.

1. General Characteristics The income-disposed or expenditure method of estimating national income is used less frequently than the net output or in­ come-distributed method. Until World War II it was employed only by Flux for Great Britain (Chapter 33, Section 2); by Landau and Kalecki for Poland (Chapter 34, Sec­ tion 11-12); by Clark and Wyler for Soviet Russia (Chapter 25, Sections 9 and 10) s1 and in one of its variant forms by Simon for Chile (Chapter 35, Section 20).2 Ele­ ments of this method were sometimes in­ troduced to complement estimates prepared by the net output method. For example, in Rao’s estimates for India (Chapter 30, Sec­ tion I-2a), the net value of certain agricul­ tural products was based on consumption data. The expenditure method was also used in estimating the net output of trade; i.e., calculating it from the value of commodities entering trade and from markups. In some estimates the income-disposed or expendi­ ture method was employed separately, after national income had been estimated by one of the two other methods, in order to show the distribution of the national expenditure between consumption and investment, or to check on the accuracy of the other estimates. In the late 1930’s and early 1940’s, new impetus was given to the analysis of national expenditures. In the United States, Simon Kuznets 3 developed the “commodity-flow” method of analysis while Wassily Leontieff 4 developed the “input-output” approach; both methods revealed expenditures within each branch of the economy, including those branches concerned with the production of final products. Richard Stone and James E. Meade in England constructed a model na­ tional income and expenditure account that was eventually accepted as a standard in a number of other countries.5 Since then, ex­ penditure analysis has been used regularly for estimating both total national product at market prices and its composition, either

2. Classification of Final Expendi­ tures or Products Total national expenditure can be ana­ lyzed either by the disposition of income— types of money expenditure, or by the dis­ position of final products—types of final products acquired by the expenditures. In the first method of analysis, the constituent elements are private expenditures on con­ sumption, expenditures for direct taxes, for contributions to nonprofit-making organiza­ tions, and private savings. In the second, the constituent elements are: (a) private consumer goods and services, (b) private investment goods, (c) government goods and services, and (d) net capital invest­ ment abroad. The second type of classifica­ tion is used far more extensively and, by the nature of the data it requires, can be com-

Notes begin on p. 536. 288

20.

THE EX PEN D ITU R E OR FINAL PRODUCTS METHOD

puted with greater ease and accuracy. In this sense it is the more important classifi­ cation and will, therefore, be considered here in some detail.7 Since the main divisions of consumption and investment are few and fairly well de­ fined, they do not differ greatly from esti­ mate to estimate. What variations there are arise mainly from different treatments of particular items.

a. Private Consumer Goods and Services Expenditures made for consumer goods and services consist of all private expendi­ ture (net of proceeds of sales of used goods) of individuals and nonprofit institutional consumers on: (a) consumer durables ex­ cept land and buildings,8 (b) nondurables, (c) services including the value of goods received in kind, and (d ) the purchase of government services by consumers. The im­ puted value of soldiers’ pay in kind is usu­ ally counted as a consumer expenditure. In the case of service industries such as hotels, restaurants, and institutions, their services are evaluated exclusive of the value of goods supplied by them, namely food, drink, etc. These goods are evaluated at retail cost and are shown separately under food, drink, etc. Rent on dwellings is usually taken gross, i.e., it includes building repairs and mainte­ nance. In estimating consumer expenditures, care is usually taken to exclude expenditure items not related to national income and to in­ clude those related to it; i.e., expenditures of foreign visitors are excluded but con­ sumption expenditures of nationals abroad are included; and inflowing remittances and gifts are deducted from total expenditures, while remittances and gifts flowing abroad are added. Consequently, aggregate con­ sumer expenditures do not exactly measure the consumption of goods and services by residents.9 Consumer expenditures are generally classified by commodities and services, with varying degrees of detail. The Japanese ex­ penditures are classified in the broad cate­ gories of food, clothing, light and fuel, dwelling cost, and others. The classifications of consumer expenditures in the official Danish estimate,10 in Benham’s estimates for Malaya (Chapter 32, Section II-2), and in those for Ceylon (Chapter 36, Section 6) and the two Rhodesias (Chapter 37, Sec­ tions 4 and 5) are fairly specific and de­ tailed. But those in the United States and

289

United Kingdom estimates are probably more elaborate than any of the others. The OEEC Standardised System recommends the following seventeen-item classification of consumer expenditures.11 1. Food (includes that consumed on farms valued at farm prices, in restaurants, hotels, schools, canteens, and clubs valued at average retail prices, and by armed forces valued at the cost to the government). 2. Alcoholic beverages (includes those con­ sumed in restaurants, hotels, canteens, and clubs valued at average retail prices). 3. Tobacco (includes that purchased in res­ taurants, hotels, canteens, and clubs valued at average retail prices). 4. Clothing (includes issues to armed forces valued at the cost to the government and deal­ ers’ margins on used clothing) and other per­ sonal equipment (jewelry, etc.). 5. Rents, rates, and water charges (includes cash and imputed rents on dwellings; repair, painting, and upkeep; and rents charged by boarding houses). 6. Fuel and light. 7. Durable household goods (furniture, kitch­ enware, refrigerators, radios, etc.). 8. Household operations (domestic services in cash and in kind, cleaning materials, electric bulbs, etc., repairs of clothing and equipment, laundry, cleaning, fire insurance, storage). 9. Personal care and health expenses (toilet articles, services of barbers and beauticians, drugs, services of physicians, dentists, nurses, hospitals, accident and health insurance). 10. Transport services (purchases of auto­ mobiles, motorcycles, and bicycles, including the markup on used goods, expenditures on tires, gasoline, parking and garaging, repairs, automobile insurance, hire of cars; railway, air­ plane, ship and bus fares; etc.). 11. Communication services (postal, tele­ phone, telegraph, etc.). 12. Recreation and entertainment (theaters; cinemas; nightclubs; restaurants, including the retail value of the food, drink, and tobacco consumed; books; newspapers and magazines; private flying; yachting, hunting, fishing; pur­ chases of games, toys, flowers; services of clubs; etc.). 13. Education and research (services of pri­ vate schools; expenditures of educational foun­ dations and nonprofit-making schools). 14. Financial services (paid and imputed bank charges, brokerage charges and invest­ ment counseling, administration cost of life in­ surance, and personal debt management). 15. Other services (religious and political activities, employment agencies, trade unions, professional associations, legal services, adver­ tisements, funeral services, etc.). 16. Personal expenditures abroad (travel, etc., other than business).

290

THE INCOM E OF NATIONS

17. Expenditures of nonresidents (whether tourists or businessmen, which are a part of the account with the rest of the world).

b. Domestic Capital Formation: Private or Combined Private and Government Depending on the organization of the country’s statistics, domestic private capital formation is shown either separately or in combination with government capital forma­ tion. Inasmuch as the methods of accounting for capital formation and financing are very different in the public sector from the pri­ vate sector, separate classification is prefer­ able. Domestic private capital formation is usually shown gross, but it can be reduced to net capital formation, and the sources for the estimates of depreciation given, as they are for the United States and Aus­ tralia.12 Generally, gross capital formation includes (a) expenditures on fixed assets by private businesses and nonprofit institutions and by government, (6) increases in in­ ventories of private business and nonprofit institutions, and (c) the increase in physical value of government stockpiles of strategic material. In the Canadian estimates, how­ ever, government investment expenditures, except for semiautonomous government en­ terprises, are included with government ex­ penditures on goods and services; and ex­ penditures for major alterations and repairs, and maintenance of producers’ fixed capital charged to current expenses, are included with gross capital formation. Only occa­ sionally are consumer durables other than dwellings included in capital formation. Private cars are included in the MatolcsyVarga Hungarian estimates and in the Aus­ tralian estimates and all consumer durables are included in the Swedish estimates up to 1934. The breakdown of private capital forma­ tion is less detailed than that of consumer expenditures for goods and services. In the Australian estimates, there is only a three­ fold breakdown: fixed capital equipment, nonfarm stocks, and farm stocks;13 and in the Canadian estimates, net changes in in­ ventories plus new residential and nonresidential construction and new machinery and equipment are shown (Chapter 21, Part II, Section lc). More detail is furnished in the United States estimates, e.g., new construction, net additions to inventories,

and producers’ equipment are separate cate­ gories, and all three are classified by type. In some estimates, e.g., Canada and United Kingdom, gross and net capital formation are shown by economic divisions. In the OEEC Standardised System of Accounts, private and government domestic capital formation are combined; while in the UN classification, there is a distinction made between the government and private sectors.14 Both classifications deal with gross rather than net capital formation. The for­ mer comprises (a) the flow of fixed asset formation of enterprises, nonprofit institu­ tions, and government (after deducting the sale of existing transportable capital goods abroad and to households) as well as (b) the value of the physical change in the in­ ventories of enterprises and general govern­ ment. The classification is as follows: 15 I. Fixed Asset Formation 1. Dwellings (value of the dwellings completed during the period as well as value of change in work in progress; including stoves, heating, plumbing, highway, and other cus­ tomary installations; excluding site except where it forms part of the value of a recorded transfer of existing items). 2. Other buildings (business, govern­ ment and of nonprofit-making in­ stitutions, excluding removable equipment and furnishings). 3. Other construction and works (roads, streets, sewers, railroad tracks, piers, tunnels, subways, bridges, dairies, sanitation and wa­ ter projects, electric transmission and telephone lines, gas mains, athletic fields, etc.). 4. Transport equipment (ships, com­ mercial motor vehicles, public serv­ ice vehicles, railway rolling stock, etc.). 5. Other equipment (power-generat­ ing machinery, agricultural ma­ chinery and implements, tractors, office machinery, equipment and furniture, industrial and mining machinery, equipment for research, schools, hospitals, etc.). II. Change in Stocks (raw materials, fin­ ished commodities, and work in prog­ ress) except that recorded in fixed asset formation above.

20.

THE E X PEN D ITU R E OR FINAL PRODUCTS M ETHOD

c. Current Government Expenditures on Goods and Services Government current and capital expendi­ tures for goods and services consist of com­ pensation of employees and net purchases from business and the rest of the world. They do not include transfer payments (to residents and foreigners), disbursements for goods and services sold to consumers, and, except in the case of the purchase of military equipment, expenditures for the acquisition of existing assets. The value of military equipment is excluded from the final ex­ penditure of the purchasing or receiving country but is included in the consumption expenditure of the selling or donor country. The explanation given by the UN for this exception is that “. . . transfers of this kind are often not included in trade statistics for reasons of military security but are included in budgetary expenditures without disclosing the amounts specifically involved.” 18 In the British estimates, however, public expendi­ tures on existing assets are included in cur­ rent government expenditures; and, there­ fore, deducted from private capital forma­ tion. In the official United States estimates, sales of government surplus consumption goods are deducted from government pur­ chases from business in order to obtain net purchases. The estimates of national expenditure made during World War II generally tried to show the war expenditure in relation to total national output. In doing so, they generally distinguished between war outlay in and outside the country, including lendlease deliveries but excluding lend-lease re­ ceipts. The general practice is to treat all defense expenditures—except the so-called “civilian defense”— as current expenditure, even though some of them may be for goods having a fairly lasting life, e.g., fortifications and battleships; and to treat civilian defense as part of gross domestic capital formation. The OEEC recommends the following classification of government current expend­ itures on goods and services: 17 By character of expenditure: 1. Wages and salaries of civilian em­ ployees—in cash and in kind, con­ tributions to social security and wel­ fare funds. 2. Pay and allowances of armed forces (including the value at cost of food

291

and clothing provided free to the armed forces). 3. Imputed rent on government build­ ings (including that on offices, schools, hospitals, etc., but not on museums, historical buildings, gov­ ernment-owned equipment, or roads and other traffic installations). 4. Purchases from businesses and abroad (military goods, such as ships, aircrafts, trucks, construction of fortifications, airfields, military camps, together with repair and maintenance; and various civilian materials, equipment and services). 5. Sales of government goods (cata­ logues, books, etc.). By purpose of expenditure: 1. General administration. 2. Defense (includes expenditure on the armed forces and in the running of defense departments, including capital expenditure for defense pur­ pose but excluding civilian defense). 3. Law courts, police, etc. 4. Health services. 5. Special social services (care of men­ tally ill, deaf, mute, and blind, de­ pendent children and aged, etc.). 6. Educational services. 7. Roads and other traffic installations (repair, maintenance, etc.). 8. Other government activities (parks, fire protection, etc.). In estimating government consumption, data are usually obtained from budgets for the larger government units, while for the smaller units, sampling is often used and the figures then are raised on the basis of popu­ lation or some other factor. In some coun­ tries, such as the United States, comprehen­ sive data are available for smaller units for census years and these data are extrapolated to other years. Only a few estimates, such as the Swedish, as was observed above, show government capital formation as a separate category of expenditure. The United States estimates, without going that far, at least show the value of all new government construction. The construction is classified according to purpose as: residential building, nonresidential building (industrial, educational, hospi­ tal and institutional, and other), military,

THE INCOM E OF NATIONS

292

highway, sewer and water, miscellaneous government enterprises, conservation and development, and all other. The United States estimates of capital formation for government enterprises show both gross acquisition of newly produced fixed assets and net change in inventories.

d. Net Capital Investment Abroad Net capital investment abroad is identical with the net balance of international pay­ ments, i.e., with the excess of the value of exports of goods and services and unilateral receipts, plus the net change in gold hold­ ings. In determining this net balance, ex­ ports of war materials, such as lend-lease shipments considered as gifts, must be offset by a unilateral payment item. Net foreign investment may be positive or negative: if negative, domestic capital formation or do­ mestic expenditures for consumption or war were financed by foreign capital, i.e., by contracting new debts or by sacrificing for­ eign assets. In the OEEC Standardised System of Ac­ counts, the unilateral receipts from and pay­ ments to abroad are not considered in com­ puting the international account part of gross national income at market prices. This account is set up as follows. Receipts from sales of goods and serv­ ices abroad (merchandise, transpor­ tation, travel, government), and Receipts of wages, salaries and prop­ erty incomes from abroad

Minus Payments for purchases of goods and services abroad (merchandise, trans­ portation, travel, government), and Payments of wages, salaries, and prop­ erty incomes to abroad.

3. Retail Sales Value and Family Budget Survey Methods The following methods can be used to estimate consumer expenditures: commodity flow; final retail sales value (of which there are two variants, retail valuation and retail sales); and family budget surveys. The com­ modity flow method, a relatively recent ap­ proach, is not limited to the estimation of consumer expenditures, but can also be used to estimate capital formation. Because the other methods are less complicated and can

be examined more quickly, the commodity flow procedure will be discussed last.

a. Retail Sales Value Method With the retail valuation variant, con­ sumer expenditures are estimated by mul­ tiplying retail prices by the quantities pur­ chased. This method was extensively used in the French postwar estimates. There, trade associations supplied the quantity informa­ tion pertaining to foodstuffs, and family budget studies furnished quantity data for many other nondurables. Estimates of con­ sumer durables were made for a base year with the use of sales proceeds figures fur­ nished by trade associations. These estimates were then extrapolated on the basis of in­ dexes of production and prices. Sweden, too, uses the retail valuation variant extensively.18 Where direct informa­ tion on the quantities of goods purchased is unavailable, the figures are derived di­ rectly from production data by subtracting exports and increases in inventories and adding imports. Retail prices are obtained from the data used in the compilation of consumers’ price indexes as well as from information reported by trade associations and price control agencies. Goods and serv­ ices purchased by government are deducted so that the resulting total relates to private consumption only. Some countries use this approach to esti­ mate expenditures for only a portion of all commodities. In the United States, for ex­ ample, the retail valuation method was used in 1939 and 1947 for approximately one eighth of total consumer expenditures, and since 1929 it has been used to estimate expenditures on passenger cars, gasoline, and oil.19 The retail sales variant differs from the retail valuation variant only in that it em­ ploys already available figures on retail sales. It has been used extensively in Canada (Chapter 21, Section Il-la ), where personal expenditures on commodities were derived from total retail sales (obtained from the merchandising census of 1941) by deduct­ ing certain items and adding others as shown below:

Deductions 1. The value of used goods sold 2. The value of retail sales for other than personal consumption 3. The value of services

2 0 . THE EX PEN D ITU R E OR FINAL PRODUCTS M ETHOD

Additions 1. The value of commodities sold for private consumption by other than retail enterprises 2. The value of imputed expenditures on commodities 3. Some taxes not included in the cen­ sus retail figure A further adjustment was made to account for net personal expenditures abroad. In other countries, the retail sales variant is rarely used as the principal method of estimation, but is used to estimate the value of certain groups of commodities. This variant is also used to make extrapolations from a base year estimate determined by some other method, as in the Polish estimates for 1933 (see Chapter 17, Section 3b). Theoretically, the final retail sales value method, with its two variants, as applied to private consumption expenditures, should yield the most accurate results, inasmuch as the final sales to ultimate consumers embody all the values added at the various stages of production and distribution of the goods involved. Unfortunately, the data on retail sales necessary for the application of these methods are, in most countries, rarely com­ plete for many categories of goods. Gen­ erally speaking, such data for passenger transportation and communication services, for finance, and for services in general are much more adequate than for commodities. Moreover, as pointed out above (Chapter 13, Sections 1 and 2), it is often difficult to distinguish between final consumption goods and intermediate production goods, inasmuch as the same goods are often used sometimes as final goods and sometimes as intermediate goods. Therefore, even in countries in which extensive retail sales sta­ tistics are available, it often becomes neces­ sary to resort to the commodity flow method in estimating some categories of final con­ sumption expenditures.

b. Family Budget Survey Method The family budget survey method is the crudest existing method of estimating na­ tional expenditure and income. It was used extensively in the days when statistics were scarce, and today is employed mostly in the underdeveloped countries, e.g., Guate­ mala, Haiti, and Korea. The Indian esti­ mates of consumer expenditures made by Mukherjee and Ghosh for 1949-50 (Chap­

293

ter 30) also were based upon surveys of family households. Japan, perhaps, is the prime example of a comparatively developed country still relying on this approach. Two kinds of Japanese surveys were used in the 1950’s to estimate consumer expenditures. One type collected income and expenditure data from farm households. The other, which began as a study of consumers’ prices, collected data pertaining to the income and expenditures of urban households. Other leading examples of past usage are the esti­ mates of gross national product made by the Polish Institute of Economic Research for 1929 and 1933; the similar estimates for Soviet Russia for 1934 and 193720 made by Colin Clark; and the survey of consumer ex­ penditures in the United States for 1935-36. The family budget survey, however, was de­ signed to show primarily the pattern of ex­ penditure as a basis for constructing indexes of retail prices. Even in those cases where it was used in constructing direct estimates of national income, the family budget ap­ proach was generally combined with other variants of the income disposed and other methods. In the Polish estimates for 1929 (see Chapter 34, Section 11-12), total consump­ tion of the urban population and consump­ tion of agricultural products by the rural population were measured by the family budget survey method. Consumption expend­ itures of the urban population were esti­ mated by first determining total national expenditures for selected goods. Per capita average expenditures of workers for cer­ tain basic foods, revealed by sample bud­ gets, were applied to the total urban popu­ lation and verified by production estimates. Then total consumer expenditures were based upon the mathematical relation be­ tween these expenditures for food and the total expenditures of the families, as shown by sample budgets of various classes of urban workers. Finally, certain items not appearing in these sample budgets were added, such as social security contributions and expenses for domestic servants and lux­ uries. Expenditures for agricultural prod­ ucts by the rural population were deter­ mined by multiplying the per capita average values of the family consumption of such products, derived from sample budgets of peasant families, by the number of rural dwellers.

294

THE IN CO M E OF NATIONS

4. Special Problems of Estimating Domestic Capital Formation

valuation at the lower of cost or market value applies generally).21

The procedures used to estimate fixed domestic capital formation differ from those used to estimate the value of the physical change in inventories. There are two main methods of estimating fixed capital forma­ tion: commodity flow (discussed in the next section) and capital expenditure. In order to estimate fixed capital formation with the latter method, data are collected from ques­ tionnaires on the purchases of capital goods, tax data, etc. The Canadian estimates are based to a great extent on this method and are discussed below (Chapter 21, Section II-lc). The methods used in estimating the value of the physical change in inventories have been discussed in the preceding chapter (Section I-4c). Only a few additional points need to be made on this subject here. Only in the case of a few commodities are figures on changes in the physical vol­ ume of inventories available. Where such is the case, it is possible, by applying to these quantity changes the pertinent prices in effect during the year, to obtain figures of the changes in inventories in monetary terms that can be incorporated in the na­ tional income total estimates at current prices without distortion of these totals. Where the large number of commodities makes the use of value statistics necessary, the data are often collected from industry censuses. In a few countries, e.g., the United States, information is obtained from tax returns. The stock data that are collected are usually the book values of inventories for the beginning and end of the year. The lack of quantity data makes it difficult to estimate the current value of the physical change in stocks. First, it is necessary to convert the value of the inventories at the beginning and end of the year into prices of the same period, and then the change in value must be expressed in average prices of the current period. The book values that are used are based on various accounting practices; however, there is a paucity of in­ formation pertaining to the practices fol­ lowed, and it must usually be assumed that the book value data are based upon the predominantly used accounting procedure (e.g., in the United Kingdom it is assumed that book values are based upon the first-in, first-out method and that the method of

5. Commodity Flow M ethod of Es­ timating Consumer Expenditures and Investment The commodity flow variant aims to de­ termine the final sales values of the prod­ ucts available for consumption or invest­ ment by tracing the accretion of values through the various stages of production and distribution. Depending upon the nature of the products and the organization of the data, the estimates are by the following divisions or some combination of them. 1. 2. 3. 4.

5. 6. 7. 8.

Manufactured products Nonmanufactured products Private construction Services of transportation, communi­ cation, finance, hotels and restaurants, amusements, and electricity (to con­ sumers) Professional, domestic, and other serv­ ices Gross rents of dwellings Government expenditures Net increase of assets abroad

The commodity flow variant is used most extensively in the United States, and is used a great deal in other countries, including Norway and Denmark.

a. Manufactured Products ( Table 20-1) Exactly as in the case of the net output method, the gross value of manufacturing production is the starting point. From it the value of goods used up or transferred in production is deducted. The difference, the value of finished products, is divided into consumer goods and producer durables. Hand in hand with this process goes a more specific classification of these two types of product. From that point on, the flow of values from manufacturing to final use is calculated by summating the intervening ad­ ditions, such as trade markups and trans­ portation charges, and by deducting the value of consumer goods added to inven­ tories. At the wholesaling stage, the aggre­ gate value of domestically produced goods is converted into the value of goods avail­ able for domestic use by adding imports and deducting exports. The typical sequences are shown separately for consumer and pro­ ducer goods in Table 20-1.

2 0 . THE EX PEN D ITU R E OR FINAL PRODUCTS M ETHOD

295

Table 20-1. Standard Pattern of Estimating the Commodity Flow of Manufactured Consumer and Producer Goods A Consumer goods 1 Manufacturing The production value of all domestically produced goods minus the production value of unfinished and investment goods equals the production value of finished domestic consumer goods, minus additions to inventories, plus manufacturers’ transportation charges (not covered by trade markups), equals purchase value of domestic consumer goods bought by (a) government, (b) whole­ salers, (c) retailers, and (d) consumers.* 2 Wholesaling The purchase value of domestically produced consumer goods plus the purchase value of im­ ported consumer goods, f minus the value of exported consumer goods, J plus markups (including indirect taxes), minus additions to inventories, plus transportation charges (not covered by trade markups) equals purchase value of consumer goods bought from wholesalers by (a) government, (b) wholesalers, (c) retailers, and (rf) consumers.* 3 Retailing The purchase value of domestic and foreign consumer goods, plus markups (including indirect taxes), minus additions to inventories equals the purchase value of domestic and foreign consumer goods bought from retailers by consumers.* B Producer durable goods B' Sales 1 Manufacturing The production value of producer finished goods minus additions to inventories, plus trans­ portation charges (not covered by trade markups) equals the purchase value of producer durables bought from manufacturers by ultimate users: government, manufacturers, private construction, trade, transportation, communications, and others; and bought from manu­ facturers for resale by wholesaling.** 2 Wholesaling The purchase value of domestic producer durables minus additions to inventories, plus the purchase value of imported producer durables, minus the sales value of exported producer durables, J plus markups (including indirect taxes), plus transportation charges (not covered by markups) equals purchase value of producer durables bought from wholesalers by ulti­ mate users: government, manufacturers, private construction, trade transoortation, com­ munications, others.** B" Inventories: Additions to inventories of 1 Consumer goods at the stage of: manufacturing wholesaling retailing 2 Producer goods at the stage of: manufacturing wholesaling

(Al) (A2) (A3) (B'l) (B'2)

* Including charities and restaurants. t Including custom duties and transportation charges not covered by markups, t Includes also direct exports of manufacturing. ** Including sales to retailers for resale (small tools and replacement items) and corresponding trade markups. Consumer expenditures for manufactured products equal the sum of the values shown under A-l-d, A-2-d, and A-3 (values of goods purchased by consumers from manu­ facturers, wholesalers, and retailers respec­ tively). The contribution of manufacturing to capital formation consists of equipment for direct use and additions to inventories. The first of these is the sum of the values shown under B'l and B'2 (the value of pro­ ducer durables purchased by ultimate users from manufacturers and the like value pur­ chases from wholesalers) adjusted for dupli­

cations. Additions to inventories are iden­ tical with the values shown under B"1 and B"2. Private domestic capital formation, as represented by manufactured products, can be shown by omitting sales to government, which are covered by government expendi­ tures.

b. Agricultural, Mining, and Other Non­ manufactured Products (Table 20-2) Thus far, only manufactured consumer and producer goods have been accounted

296

THE INCO M E OF NATIONS

for, not total consumer expenditures for commodities or total capital formation. To give a complete picture of consumer ex­ penditures for commodities, additions have to be made for agricultural, fishing, forestry, mining, and other nonmanufactured finished products sold by producers to consumers either directly or through trade. Similarly, to obtain the complete value of private do­ mestic capital formation, it is necessary to add to the value of manufactured producer goods the value of nonmanufactured con­ struction materials; the net value added in private construction of residential and other buildings; and in new transportation and communication structures and in land im­ provements; and the net increase in live­ stock, other farm inventories, and other nonmanufacturing stocks. The value of con­ struction can be computed to include ma­ terials, but then their value must be de­ ducted from the total flow of manufactured and nonmanufactured producer goods. The comprehensive summation of total com­ modity flow is shown in Table 20-2.

of hotels and restaurants the cost of raw food, beverages, and other products sold must be deducted because they have already been accounted for in the expenditures for agricultural and manufactured products. The value of the services of professions, amusements, and domestics is relatively sim­ ple to estimate. As not many goods or serv­ ices are purchased from outside, gross and net value coincide closely, and the incomedistributed method can be followed. Most of these services are rendered to consumers directly. Those rendered businesses and gov­ ernments are usually small, except in the case of legal, engineering, and a few other types of service. Consumer expenditures for services of dwellings equal gross rents, the starting point for estimating net rent.

d. Adjustments for International Accounts and Government Expenditures Total consumer expenditures for com­ modities and services are next adjusted for

Table 20-2. Standard Pattern of the Estimates of Total Commodity Flow to Private Users All consumer goods 1 Manufactured 2 Nonmanufactured 2a Directly consumed by producers 2b Bought by consumers

Total Consumer Expenditures for Commodities

Total private domestic capital formation 1 Producer manufactured durables (equipment and con­ struction materials) 2 Additions to inventories of producer and consumer manufactured goods 3 Nonmanufactured construction materials 4 Private construction and land improvement work done 5 Additions to livestock 6 Additions to inventories of farm and other nonmanu­ facturing products

Total Expenditures for Producer Durables

c. Services The next step in estimating total con­ sumer expenditures is to compute and add the value of services. The value flow need not be traced. For transportation, commu­ nications, finance, electric light and power (unless classified with manufacturing), the estimates can be derived directly from cur­ rent business accounts. For hotels and res­ taurants, census data for certain years and indexes of annual sales are available. For amusement places, box office receipts— often far from comprehensive—are used. Sales to business and government must be deducted to avoid duplications. In the case

the international exchange of services. The expenditures of the country’s residents on foreign travel and on services and goods abroad as well as gifts to foreigners are added; the corresponding expenditures of foreigners in the country, including foreign remittances, are deducted. Then net foreign investments are added. These adjustments for international accounts are based on the international balance of payments. The last step is to add government ex­ penditures on goods and services computed, as described above, from budgetary reports and samples, with such adjustments as may be necessary to avoid duplications between

20.

THE EX PEN D ITU R E OR FINAL PRODUCTS M ETHOD

federal, state, and local accounts, or the in­ clusion of government market services and of income transfers or mere purchases of existing wealth.

e. Partial and Simplified Commodity Flow Estimates From this outline it is evident that the commodity flow variant can be scientifically applied only for countries rich in business accounting knowledge and such business statistics as extensive production censuses and current sales data for producers and wholesalers, reports on inventories as well as on trade margins, freight expenses, con­ struction output, and many other items. It is not surprising, therefore, that few coun­ tries use this variant. The availability of data is an important factor that dictates the form in which the commodity flow variant is used. For example, in Norway, the type of data available makes it difficult to allocate consumers’ goods between the public and private sectors. The combined total is broken down only by type of goods and services; and from this total the estimated public con­ sumption, based on government accounts, is subtracted in order to obtain private con­ sumption expenditures.22 In India, too, a simplified version of the commodity flow method is used, because of the lack of ade­ quate data (Chapter 30, Sections II-l and 2). One short cut in applying the commodity flow variant consists of estimating gross or net national product indirectly by adding to national income indirect taxes and probable depreciation, and then estimating the ap­ portionment of the total among expendi­

297

tures. All expenditure sectors except one, which is computed as a residual, are esti­ mated. This device was applied in some estimates, including the official estimates for Ireland and the earlier estimates of the United States. Estimated as a residual were personal consumer expenditures on goods and services for Ireland, and consumer ex­ penditures for nondurable commodities and services in the earlier official United States estimates. Another short cut consists of simply esti­ mating the flow of consumer goods, distin­ guishing between foods (including tobacco) and beverages on the one hand, and other commodities, on the other hand. The quan­ tity of the various kinds of foods consumed is multiplied by an average price. Retail prices are applied to those entering retail trade, wholesale prices to those bought by eating places, and farm prices to products consumed on farms. The final value of the other goods is computed by adding imports to the value of household, manufacturing, mining, and forestry production, deducting exports, and adding distribution charges as an over-all percentage. This simplified ver­ sion was used in Clark’s and Wyler’s esti­ mates for Soviet Russia. Flux’s estimates of British national in­ come for 1907 and 1924 also involved a limited application of the commodity flow variant. He covered the commodity flow from agriculture, mining, and manufactur­ ing, roughly eliminated unfinished output, adjusted for imports and exports, and added the values accruing at the handling stage and incomes from services.

Part Three. Estimates for Selected Countries

EX PLA N A TO R Y NOTE TO PA R T TH R EE

The basic concepts of national income and the broad operational principles underlying the preparation and uses of national income estimates, reviewed in Part II of this book, control the general pattern of modern esti­ mates, but do not precisely determine their actual content or form. These are deter­ mined very largely by factors that are ex­ tremely variable in nature, such as: the nature of the economy and its problems at the moment; the degree of appreciation of the importance of national income estimates by official and unofficial bodies in the coun­ try; the state of development of its economic science and statistics; and the imagination and competence of the individuals charged with the preparation of national income estimates. National income estimates, there­ fore, differ considerably from one another in their construction, statistical bases, and degree of reliability. Each estimate is a work of both science and art, reflecting the state of statistical knowledge of the time in the country in which it is prepared and even the general intellectual and political climate of that country at that time, as well as the personalities and skills and imagination of the estimators handling the statistical ma­ terials. To understand how national income esti­ mates are made, it is not sufficient to know the general principles guiding their prepa­ ration. It is necessary also to be acquainted with the circumstances in which particular estimates are prepared, the difficulties ex­ perienced in procuring the necessary mate­ rials, and the manner, if any, in which these difficulties are being overcome. To give the reader an appreciation of the many con­ trolling factors in operation, this survey of

a number of typical estimates is being of­ fered. Each of these reviews is based upon a published account detailing the construc­ tion of the estimate.

1. Reasons for Selecting Thirteen Countries’ Estimates for Detailed Individual Review From among the eighty or more countries preparing national income estimates today, thirteen have been selected for a detailed review of their estimates. Among them are countries hig’ ly advanced industrially and some compar tively undeveloped; some that have good statistics and some that have only poor ones; some that have had a long history of estimating and some that have never prepared any estimates before. The number was limited to the most typical esti­ mates, and not necessarily to those of the largest countries. The United States and United Kingdom’s estimates are not in­ cluded in this detailed analysis for two rea­ sons: first of all, extensive comments about them are incorporated in the text of Part II; and secondly, detailed information about them can be obtained in easily available sources.

2. Distinguishing Characteristics of These Countries’ Estimates Some of the thirteen countries whose estimates are reviewed in this part of the book, namely Soviet Russia, Yugoslavia, and Turkey, have concentrated on the net output type of estimate, and one—Brazil— on the income-distributed type, but most

302

THE INCOM E OF NATIONS

have developed estimates by two or all three methods and even by two or more variants of a particular method. The tendency in national income estimating throughout the world is toward the analysis of national in­ come in all of its three phases—that of origin in the various industrial divisions, dis­ tribution of income to the factors of pro­ duction, and disposition of the national in­ come for consumption and investment—and correspondingly, toward the utilization of all three methods of national income esti­ mating (see the classification of 87 country estimates in Chapter 17, Section le ). Since World War II at least four of the foregoing thirteen countries—Canada, Swe­ den, the Netherlands, and France—have developed comprehensive sets of social or sector accounts in their estimates; and by the time this book is published, some of the others will undoubtedly be also in that list. Whether socialist Soviet Russia and Yugo­ slavia will incorporate this social accounting approach in their estimates is not clear, in­ asmuch as the sectors in their case are not the same as in capitalist countries, and are not accorded the same importance. The estimates for these thirteen countries are taken up in an order that is partly re­ gional and partly methodological. The Canadian estimates are taken up first (Chapter 21) because they are among the most advanced in the world and are prepared by all three estimating methods. They are closely related to the United States and British estimates; at the same time they embody many original features. Then the Irish estimates are reviewed (Chapter 22). They are comparatively new, and have been developed under the combined influence of the British, Canadian, and United States standards. In their latest form (1955), they similarly make use of all three estimating methods and are both modern and refined. The estimates for the Netherlands (Chap­ ter 23) are discussed next. Before the war they represented one of the finest examples of the income-distributed type of estimate. Since the war they have branched out into the other methodological approaches as well as into the social-accounts type of analysis and have generally kept pace with the con­ ceptual and methodological advances in the field the world over. The review moves on then to the Swedish estimates (Chapter 24). Before the war they represented a unique example of a net output estimate that was only a step re­

moved from the final expenditure or prod­ ucts estimate. Since the war they shifted completely to the final products method and have also developed the social-accounts presentation to the fullest degree. Most extensive treatment of all is given to Soviet Russia’s estimates (Chapter 25), which are in a class by themselves. Based on a different concept of production from most of the other estimates, they relate to a unique economy and give rise to many basic methodological issues. More space is devoted to these estimates than to those of any other country because of the impor­ tance of the country and because of the in­ teresting nature and general inaccessibility of the original materials. The German estimates come next in order (Chapter 26). The first part of the chapter deals with Germany’s prewar estimates, which had their own peculiar historical roots and represented a fairly elaborate ap­ plication of the income-distributed method. The second part of the chapter treats the postwar estimates of West Germany, which were deeply affected by the disorganization of the German economy and dislocation of her statistics resulting from the war and the subsequent partition of the country. These new estimates were prepared by the net output and final products methods and in­ volved the employment of many ingenious short cuts. The review of the French estimates (Chapter 27) is limited entirely to the post­ war estimates, which represent a complete break from their exceedingly imperfect predecessors. These new estimates went through several basic revisions in an at­ tempt to overcome the difficulties created by the disruption of the country’s economic life during the war. In their latest forms (1952-55), they lean most heavily upon the income-distributed and the final expendi­ ture or products methods and emphasize particularly the social or sector accounts. Methodologically, they are on the way to becoming as advanced as some of the bestdeveloped country estimates, although they still rest on inadequate statistics. The remaining five chapters (Chapters 28-32) deal with comparatively underde­ veloped countries. The estimates for Yugoslavia (Chapter 28) and Turkey (Chapter 29) are interest­ ing as representative of the application of sophisticated statistical techniques to inade­ quate statistical data in underdeveloped

21.

CANADA

countries. Both sets of estimates reveal the considerable advancement made in their data and methods from the prewar to the postwar period. The Yugoslavian estimates are also significant because, like Soviet Rus­ sia’s, they illustrate the application of the material production concept. The estimates for India (Chapter 30) are explored at considerable length, again be­ cause of the political importance of this country, particularly among the underde­ veloped countries of the world; the use of advanced statistical techniques influenced by British and United States examples; and the application of these techniques to the scanty statistical data available in that country. There follows an examination of the esti­ mates for Brazil (Chapter 31). It is the only South American country from among the many now preparing national income estimates whose work is reviewed in its en­ tirety in this book. Brazil also represents one of the few applications of the incomedistributed method to the preparation of an estimate for an underdeveloped country. The survey ends with a comment (Chap­ ter 32) on Benham’s estimates of national

303

income of the four islands of the British West Indies and of Malaya. His methods are discussed but no figures are presented.

3. Organization of the Country Chapters All twelve country chapters are organized on a more or less similar plan: they all open with a brief history of the country’s estimates; outline the concepts underlying them; present their general methodological approach; give for each item of the national income the statistical sources and methods employed in estimating it; and end with the presentation of the main statistical results of the estimates. Statistical tables are presented only insofar as they are necessary to exem­ plify the methodology of the estimate and its results. They are not given for the sake of their statistical information as such. A reader seeking the latest statistical informa­ tion on the national incomes of the different countries would have to turn to the publi­ cations of the United Nations or to the sta­ tistical year books of the respective coun­ tries.

21. CANADA* References 1. a. Dominion Bureau of Statistics, National Accounts, Income and Expenditure, 1926-1950, Ottawa, 1951. b. , National Accounts, Income and Expenditure, 1950-1953, Ottawa, 1954. c. , National Accounts, Income and Expenditure, 1951-1954, Ottawa, 1955.

1. History During the century before World War I, some estimates of Canadian national income were published from time to time in British economic studies, but these were based on * Acknowledgment is due to Mr. F. H. Leacy, Director of Research and Development Divi­ sion of the Dominion Bureau of Statistics for going over the draft of this chapter and making helpful suggestions. N otes begin on p. 536.

inadequate statistical information and were somewhat rough.1 More authoritative were the unofficial estimates for the years 1911 and 1918 prepared by the Dominion Statis­ tician, R. H. Coats, and published in 1919. Systematic inquiries began only in 1925 when the Dominion Bureau of Statistics published its first estimate, but this was soon followed by other estimates established on an annual basis.2 During the 1930’s the Bank of Nova Scotia published a series run­

304

THE INCOM E OF NATIONS

ning back to 1921.3 Toward the end of the decade, the Royal Commission on Dominion-Provincial Fiscal Relations published a series of estimates covering the years 192637, which was followed by a study for the years 1938-40.4 The Bank of Canada con­ tinued the Royal Commission’s series but never published the results. In 1941, after considerable study, the Dominion Bureau brought out comprehen­ sive estimates by economic divisions and types of income, carrying them back to 1919 and forward to 1938 and later to 1942. Gross national product was roughly estimated.5 In 1945, the Bureau inaugurated a new set of national accounts in line with the new approach agreed upon at a conference of representatives from the United King­ dom, Canada, and the United States in Sep­ tember, 1944 (Chapter 10, Section 3e). In collaboration with the Bank of Canada and several Dominion government departments, it published estimates of gross national product and expenditure and of personal incomes since 1938.6 In 1948, in step with the moves made by the estimating agencies of the United King­ dom and the United States, the Dominion Bureau began the publication of double­ entry social accounts. By that time it had changed the title of its reports to “National Accounts, Income and Expenditure.” Hence­ forth, in its reports, the Bureau gave greater emphasis to social accounts than to the tra­ ditional single-entry estimates; and to gross national product or expenditure than to national income proper. Since 1946 the Dominion Bureau has had the sole responsibility for publishing cur­ rent national accounts, as both the Bank of Nova Scotia and the Bank of Canada ceased preparing their own independent estimates in that year. The earlier, now obsolete, estimates of the Dominion Bureau and the Bank of Nova Scotia were made mainly by the net output method; those of the Royal Commission and the Bank of Canada mainly by the incomedistributed method. The Dominion Bureau’s estimates prepared since 1948 combine the three methods: the income-distributed (paid out) and the net output methods (the latter called “synthetic operating account method”) for estimating national income; and the final products method for estimating national ex­ penditure in particular. In 1953, the Do­

minion Bureau inaugurated quarterly esti­ mates as an addition to the annual ones. The discussion in this chapter is con­ cerned entirely with the estimates prepared since 1948, the methodology of which was explained by the Dominion Bureau in its 1951 publication (Reference No. la ). It is divided into two main parts: one dealing with the estimates of national income pre­ pared by a combination of the net output and the income-distributed (paid out) meth­ ods; and the other treating the estimates of national expenditure by the final products method.

I.

Net Output and Incomepaid-out Estimates

1. General Methodology The Dominion Bureau estimates prepared in 1948-54 were founded on the standard comprehensive production concept. Income from government was measured by the com­ pensation in cash and in kind paid to civil and military employees plus the interest on that portion of the public debt that was used to finance productive assets. Profits and losses of government enterprises were in­ cluded in the profits and losses of the busi­ ness sectors in which these enterprises were classified. Taxes on corporate net income were considered as direct taxes and conse­ quently as a part of national income. In­ direct taxes, which were not a part of na­ tional income, included all taxes deductible as expenses from the gross revenue of busi­ ness or otherwise increasing the market price of goods and services. Provincial real and personal property taxes, whether paid by individuals or by business, were consid­ ered to be indirect. This series is an instructive example of a combination of the net output and incomedistributed (income-paid-out variant) meth­ ods. The net output method—measuring net output by subtracting certain outlays from estimates of gross revenue from the sale of goods and services—was used to estimate the net income of the following unincorpo­ rated businesses: agriculture, forestry, fish­ ing, hunting and trapping, manufacturing, and retail stores; it was used also in the cal­ culation of the rent component of capital income. The net output method was applied because of the difficulties of separating capi­ tal income from labor income in unincorpo­

21.

rated business. The income from such busi­ ness was classified by its industrial origin. The income-distributed method was ap­ plied to incomes from all corporate business and from professions, trades, some other unincorporated business, domestics, govern­ ment and quasi-public institutions, as well as incomes from abroad. The incomes in these categories were classified both by types of factor shares and by industrial origin. Investment income, including corpo­ ration profits, was classified on a company basis because of the difficulties involved in allocating corporation profits and related items among the various activities of a multiplant company. In the industrial distribution of income the UN’s Standard Industrial Classification was followed for the most part. This classi­ fication consisted of the following thirteen main divisions: 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

Agriculture Forestry Fishing and trapping Mining, quarrying, and oilwells Manufacturing Construction Transportation, storage, communica­ tion, and public utility operation (later shown separately) Wholesale trade Retail trade Finance, insurance, and real estate Service Government Nonresidents

This classification differed from the Stand­ ard Industrial Classification only in the treatment of the government sector. On the one hand, all government noncommercial activity such as forestry patrol, fishing in­ spection, bridge and highway construction, health, and educational services were shown under government instead of being allo­ cated, as in the Standard Industrial Classi­ fication, to their various related industrial groups (forestry, fishing, construction, serv­ ice, etc.). On the other hand, certain gov­ ernment business enterprises, such as the Post Office and the Canadian Farm Loan Board, were assigned to communication and finance in the business sector instead of, as in the Standard Industrial Classification, to the government sector. In addition to net national income at factor cost, the estimates also showed gross national product at market prices (by add­

CANADA

305

ing indirect taxes less subsidies, deprecia­ tion allowances, and similar nonfactor busi­ ness costs); gross national expenditure; and personal income and its geographic distribu­ tion by provinces. Gross national expendi­ ture was given both in current dollars and in constant (1935-39) dollars in the 192650 series; and in constant (1949) dollars in the 1950-53 series. Part I of this chapter is divided into the following main sections: 1. Estimates of net output for certain in­ dustrial groups 2. Estimates of income paid out by cer­ tain industrial groups and of invest­ ment income 3. Distribution of personal income by provinces 4. Gross national product and gross na­ tional expenditure

2. N et Output Portion of the E stim ates7 Agriculture was the most important in­ dustry in which income was measured using the net output method. Prepared by the Agriculture Division of the Bureau, the esti­ mates of total gross farm income by type of product sold were based on the government’s extensive statistical coverage of the agricul­ ture sector of the economy. Food and forest products grown and consumed on farms were valued at alternative market prices. The imputed values of gross rents on owneroccupied farm dwellings were included in “income in kind” received by farmers. Gross farm income also included the value of the physical change in inventories of grain and livestock computed at year-end prices. This change was added or subtracted as inven­ tory stocks had been increased or depleted. Outlays deducted from this gross farm income figure included taxes on real estate, gross rents on farm land, labor costs, in­ terest on farm debt, feed and seed, repairs and depreciation, and similar farm oper­ ating expenses. After deduction of these expenses, government payments to farmers on Wheat Acreage Reduction account and under the Prairie Farm Assistance Act and Prairie Farm Income Plan were added. The resulting figure was called “net income of farm operators from farming operations.” Three further adjustments were made to this total in accordance with the national income definition:

306

THE INCO M E OF NATIONS

First, imputed rents on owner-occupied farm dwellings and corporation profits in agriculture were deducted, since these were included in investment income. Second, payments from the government under the Prairie Farm Assistance Act and Prairie Farm Income Plan were de­ ducted, since these were regarded as transfer payments. Payments for wheat acreage reduc­ tion were considered a subsidy and, therefore, were not excluded. These adjustments result in the aggregate called “net income received by farm operators from farm production.” The third, and final, adjustment accounted for the undistributed earnings of the Canadian Co-operative Wheat Producers and the Cana­ dian Wheat Board and for inventory valuation adjustments which bring the earnings figures into agreement with national income figures. The aggregate used in the national income tabu­ lation was the “accrued net income of farm operators from farm production.” The relation­ ship between the totals and the various adjust­ ments are shown in Table 21-1.

deduction of miscellaneous expenses, adjust­ ments were made to eliminate wages and salaries paid to employed guides and trap­ pers, which were included elsewhere in the estimates, and sales off fur farms. The census of industry returns since 1946 were the source of data on the gross rev­ enues of unincorporated manufacturing establishments. A “net to gross” ratio, ob­ tained from “Taxation Statistics,” published annually by the Department of National Revenue, was used to reduce the gross fig­ ures to net income. This was projected back to 1926 by extrapolating the 1946 figure of the gross value of production in unincor­ porated manufacturing on the basis of changes in the total gross value of produc­ tion of selected industries. Finally, these figures were netted by the 1946 ratio of “net to gross” income.

Table 21-1. Accrued Net Income of Farm Operators from Farm Production, 1952-54 Reconciliation Statement (million dollars) 1952 1954 1,929 1,130 Net income of farm operators from farming operations -42 -43 Deduct: Amounts included in investment income -5 -2 Deduct: Transfer payments under Prairie Farm Assistance Act 1,882 1,085 Equals: Net income received by farm operators from farm production -31 -27 Adjustment on grain transactions 1,851 1,058 Accrued net income of farm operators from farm production so u rc e :

Reference No. 1c, p. 52.

For fishing, estimates of gross revenue were taken to be the “value of fish caught and landed” as compiled by the Fisheries Section of the Bureau. Depreciation was calculated at the rates allowed under the income tax. Repair costs were taken as an arbitrary percentage of the capital value of vessels, boats, and other equipment. Fuel costs for 1941 were estimated on the basis of a percentage of sales of tax-exempt gaso­ line in Nova Scotia, and were adjusted up­ ward by the ratio of the tonnage of power boats used in fishing in Canada to the cor­ responding tonnage in Nova Scotia. The resulting estimation was then projected to other years on a composite index of gaso­ line prices and the tonnage of power boats employed. Miscellaneous expenses were taken as a percentage of gross revenue. The gross value of Canadian fur produc­ tion was taken to be the gross income of hunting and trapping as reported annually by the Agriculture Division. In addition to

Total retail sales were derived from sur­ veys conducted by the Merchandising and Services Section of the Bureau for the years 1941, 1944, 1945, 1946, and 1948. These surveys provided data that were used to cal­ culate a weighted average ratio of net profits (before withdrawals) to net sales. Estimates of net sales for 1926-50 were based on bench-mark data from the 1931 and 1941 decennial censuses projected to other years according to the trend of total retail sales. The “sales to profit” ratios were applied to the data to obtain net income.

3. Income-paid-out Parts of the Estimates a. Labor Income 8 In determining wages and salaries paid out in each division, the Bureau used, in the main, annual censuses of industry tabula­ tions of aggregate payrolls, annual surveys,

21.

and the published statements of government and business. Where coverage was incom­ plete, these sources were supplemented with directly related data. The various interpola­ tion techniques used were based on decen­ nial bench-mark material and indexes. The Bureau analyzed the wage and salary esti­ mates in 1949 in regard to sources of data, concluding that approximately 70 per cent of the total labor income estimate was taken directly from annual surveys and published statements, 28 per cent was projected by means of directly related data, and 2 per cent was estimated indirectly. For the years prior to 1941 the proportion obtained from direct annual surveys was smaller, making the estimates much less reliable. The various kinds of labor income were estimated as follows: Annual surveys and censuses of industry conducted by the Bureau provided the informa­ tion for estimates of labor income in manufac­ turing, mining, transportation, communication, and public utilities. Adjustments were made to eliminate the earnings of the self-employed which were included elsewhere. Wages and salaries figures of federal and provisional gov­ ernment employees were obtained from the Public Accounts. In finance, insurance and real estate estimates were based on data from the Canadian Bankers’ Association, from various sections of the Bureau, and from published re­ ports of companies. Directly related data were used for estimates of labor income in agriculture, fishing, hunting and trapping, construction, water transport, and wholesale trade. Wages and salaries in agricul­ ture were projected from the 1941 census bench mark by an index of the average farm wage bill. For 1926-41, the decennial census bench marks were interpolated by a combined index of hired labor force and farm wage rates. This bench-mark projection method was used for fishing, in which the estimates for intercensal years were based on the trend of the total value of fish caught and landed; for hunting and trap­ ping, in which projections were based on an index of “wild life” fur production as reported to the Bureau; and for wholesale and retail trade, in which bench marks from the census of merchandising and services were projected by employment and wage indexes prior to 1941. The estimates of wages and salaries in con­ struction were based on the 1946 figure of ag­ gregate payrolls reported in the census of con­ struction, which was blown up by the ratio of the Labor Force Survey count of paid workers in private construction to the census figure of employment in this industry. The resulting bench-mark figure was extrapolated forward by the index of aggregate payrolls in construction.

CANADA

307

The same method was used for the estimates for the years 1941-45. Earlier estimates were extrapolated back from the 1941 census bench mark by indexes of wages and salaries of con­ tractors reported to the annual census of con­ struction. Those estimates based on indirect evidence included wages and salaries paid in religious service, journalism, accounting, art, and music that were calculated from census material and various indexes of employment and payrolls. For certain groups the cost-of-living index was the basis for extrapolation. The estimating procedure was similar for the wages and salaries paid for legal service and for domestic service. For legal service the 1926-1941 estimations were based on project­ ing bench marks obtained from the 1931 and 1941 population censuses by an index of net income of lawyers. The 1941 bench mark was extrapolated to current years by an index based on estimates of the number of practicing law­ yers and average weekly earnings in finance. For domestics, the figures for earlier years were projected by a combined index of employ­ ment in hotels and restaurants and wage rates in laundries. To cash income was added the imputed value of board and other allowances based on census data and special surveys; this was estimated separately for industries in which they were important. Wages and salaries were estimated on a “gross” basis, i.e., be­ fore taxes, employee contributions to unem­ ployment insurance, etc. Employers’ contri­ butions to the various social security schemes were included and estimated separately by industry from surveys, financial statements of government units, and other sources. The payments made on labor account for the several industrial groups were then summed to give total wages, salaries, and supplementary labor income. Military Pay and Allowances were shown separately in the national income tables. Payments to members of the armed forces included remuneration in cash and in kind. Data on cash pay were furnished by the Department of National Defense and, for the earlier years, from the Public Accounts. Income in kind was calculated by apply­ ing average military strengths of the three armed services to estimates of man-year costs for food and clothing. b. Capital In c o m e9 Capital income (in Canadian terminology, investment income) consists primarily of corporation profits, interest, net rents, and profits of government business enterprises.

308

THE INCO M E OF NATIONS

Corporation profits, taken before pay­ ment of corporate net income taxes (which are classified as direct taxes), were esti­ mated from “Taxation Statistics,” published by the Department of National Revenue beginning with 1944. This publication was based on information from corporation in­ come tax returns as they were received by the Department. The data were checked to ensure that all large companies were in­ cluded; and it was believed that understate­ ment of results owing to omission of com­ panies was well under 1 per cent. Certain adjustments were made in the figures to make them correspond to national income definitions. Corporate losses were deducted and depletion allowances and charitable contributions were added back to profits. The profits series was adjusted for the time lag between the taxation and business ac­ counting year. The Department of National Revenue supplied the figures for profits re­ funded to the government on negotiation of war contracts; these were eliminated from the profits series and also from the govern­ ment expenditure series. Dividends paid to non-Canadians were deducted. Estimates of profits for the years 1926 to 1944 were based on a special sample study of tax re­ turns of approximately 1,000 companies conducted in the Department of National Revenue. The sample was constructed as follows: For nonmanufacturing corporations, the list of all corporations existing in 1946 as given in “Taxation Statistics” for that year, and for manufacturing corporations (accounting for more than half of corporate profits), the list of such companies existing in 1926 and in 1946 was used. The total list was divided into twenty-three industrial groups. Companies from each group were selected for the sample. Larger corporations were recorded completely. The profits of the smaller concerns were inflated to the total of all corporations by comparing the sales of the companies in the sample for the year 1946 with the sales of all corporations for the same year, separately for each industry group, and by groups related to the size of the company. The results for groups of all sizes were added to obtain the industry totals. The profits estimates were considered gen­ erally good. However, except for manufactur­ ing concerns, the results for the early years were not as accurate as those for later years because the sample of nonmanufacturing cor­ porations was limited to the list of companies in operation in 1946 and, therefore, did not reflect the profits of the companies that went out of existence prior to that year. However,

adjustments designed to overcome this short­ coming were made in the series. The sample made possible an industrial breakdown of profits, depletion charges, bond interest paid and received, and other items as well as estimates of bad debts; and it improved the reliability of estimates of gross domestic investment. Interest income included Canadian bond, mortgage, and deposit interest received by persons (defined as individuals, noncommer­ cial institutions, estates, and trust funds), and investment income of life insurance companies, plus some miscellaneous income; but did not include interest and net rents paid to corporations and government busi­ ness enterprises, as these were included im­ plicitly with the profits of these institutions. The figures were obtained as follows: Interest received by individuals on corporate and other nonregistered bonds was ascertained for 1937 from a sample study of the RowellSirois Commission on ownership certification; and interest on registered government bonds was supplied by various government depart­ ments. The 1937 bench mark was projected for­ ward to 1938 and back to 1926 by an index of total interest paid. The index included interest on provincial, municipal, and federal direct and guaranteed debt and funded corporation debt, less total interest paid to nonresidents. A wide variety of government sources was used in the construction of this index. For later years, in­ terest received by individuals on federal gov­ ernment direct debt was estimated by applying an average rate to the average holdings of indi­ viduals, while all other bond interest receipts by individuals were based on similar projec­ tions as described above. Mortgage interest accruing to persons—con­ sidered among the least satisfactory estimates in investment income—was estimated sepa­ rately for farm and nonfarm. To total indi­ vidual holdings in each segment, an average rate of interest was applied.10 Deposit interest paid to individual depositors was calculated from 1943 on by subtracting from total deposit interest, interest paid to non­ residents, and to corporate bodies in Canada. Figures for earlier years were obtained by pro­ jection on the basis of changes in interest rates and Canadian interest-bearing deposits. The Bank of Canada provided these estimates as well as a valuation of banking services provided to individuals without charge. This latter figure was added to deposit interest. Estimates of net rent received by per­ sons were calculated separately for residen­ tial nonfarm rents, nonresidential nonfarm rents, and residential and nonresidential

21.

farm rents. The estimating methods were similar for the three groups. First, total gross rents were estimated; second, amounts received by business corporations and gov­ ernments were subtracted to arrive at gross rents received by persons; third, expenses were deducted to obtain net rents received by persons. The details of the procedure were as follows: Gross residential nonfarm rents were esti­ mated by multiplying the number of nonfarm dwellings, obtained from the Central Mortgage and Housing Corporation for 1926-1949, by an estimated average net figure. The number of dwellings in 1950 was based on estimates in the monthly publication of the Bureau, “New Residential Construction.” Average rents were estimated from the 1941 housing census and projected back to 1926 and forward to 1946 by the rent component of the cost-of-living index. From 1947 to 1950 these estimates were based on special surveys conducted by the Bureau. Amounts received by government agencies were subtracted from the total of gross resi­ dential nonfarm rents. The 1941 housing census provided data on heating costs included in rents, taxes, repairs, insurance, and mortgage interest paid. Average fuel costs per dwelling were estimated by dividing the expenditures of consumers on fuel by the number of dwellings in Canada. Depreciation was taken as equal to 15 per cent of urban gross rents paid and im­ puted, and 20 per cent of rural nonfarm rents, paid and imputed. These deductible costs were taken as a percentage of total gross residential rents, paid and imputed, to obtain net rents received by persons. Gross nonresidential nonfarm rent received by persons was estimated from “Taxation Sta­ tistics,” from the special corporation sample study for 1926-44, and from other govern­ ment information; rents paid by unincorporated business were estimated on the basis of incom­ plete information. In general bench marks de­ rived from special surveys or censuses were projected by industrial groups. Rents received by government and incorporated businesses were subtracted. Expenses for this group were estimated on the basis of the experience of life insurance companies since more adequate data were lacking. Residential farm rents were estimated at 5 per cent of the value of owned and rented farm dwellings. Mortgage interest paid was deducted. Nonresidential farm rents were based on data on rents “in kind” and in cash given in the 1921, 1931, and 1941 decennial censuses and in the 1936 prairie census. Figures for intercensal years were estimated by projecting census data by the trends of values of farm crops and land.

CANADA

309

Special procedures were used also in esti­ mating the following capital incomes: The information on the capital income re­ ceived by life insurance companies, consisting of interest, dividends, and net rents was ob­ tained from reports of the Superintendent of Insurance. Interest on policy loans was excluded because it was treated as interest on consumer debt. Income accruing to non-Canadian policy­ holders was excluded on the basis of the ratio of the liabilities outside Canada to total liabili­ ties of Canadian companies. Net profits of agri­ cultural and other co-operatives and of mutual nonlife insurance companies not included else­ where were estimated from “Taxation Statis­ tics” and other government sources. Government capital income consisted of profits of government enterprises and of in­ terest on government loans and advances and on investments of pension, social security, and other public funds. The figures were ob­ tained from the Public Accounts of the fed­ eral and provincial governments, the Bank of Canada, “Statistical Summary,” city financial statements, and other government sources.

Certain adjustments were necessary so as to make certain categories of capital in­ come accord with national income defini­ tions. The portion of interest on the public debt classified as a “transfer” was excluded. The “transfer” portion of interest was esti­ mated as follows: total interest on the public debt, both provincial and federal, was ob­ tained from the Public Accounts of the various authorities. An estimate of interest on debt used to finance productive assets was deducted. This portion of interest was estimated by applying the average interest rate on outstanding debt of the various gov­ ernments to the total value of productive assets (excluding those of government busi­ ness enterprises). The data regarding the value of productive assets, compiled from the Public Accounts and government re­ ports, were incomplete and not classified in the required manner. For these reasons, this procedure was considered arbitrary. At the municipal level, the “transfer” portion was defined to include only interest on relief debt. Interest on consumer debt was deducted. This was estimated by applying a rate of interest to the total amount of consumer debt outstanding (excluding policy loans). Among the sources used to estimate the amount of consumer debt outstanding were the decennial census, reports of the Super­ intendent of Insurance for Canada, and other government reports.

310

THE INCOM E OF NATIONS

Estimates of dividends and interest paid to nonresidents that had been excluded in the various capital income categories in­ cluded a portion retained by the federal government in the form of “withholding taxes.” The figures were taken from the federal government’s Public Accounts and were added back. Finally, an inventory valuation adjustment was made in corporation profits from pri­ vate grain dealing. Since inventory changes for grain in commercial channels on the expenditure side of the national accounts were valued at the current value of the physical change rather than the change in the reported book value, a corresponding inventory valuation adjustment was neces­ sary on the income side of the accounts as corporation profits and net income of un­ incorporated business are usually computed on the basis of changes in book values of inventories. c. Income of the Self-employed11 Income of the self-employed other than those engaged in agriculture and allied trades, manufacturing, and retail trade (esti­ mated by the net output method as described above), which together form the total of unincorporated business, was estimated sep­ arately for each economic group by the income-paid-out method. For the most part, this income was estimated by multiplying average net income by the number of pro­ prietors in each group and by extrapolating bench-mark data back to 1926: Net professional income of independent prac­ titioners, including doctors, dentists, nurses, lawyers, and engineers, was estimated, in gen­ eral, by multiplying average net income by the number of independent practitioners and pro­ jecting these bench-mark estimates to earlier years on the basis of related data. Average net income of independent doctors and dentists for certain years was obtained from special sur­ veys of these professions conducted by the Na­ tional Income Section of the Bureau. For years after 1946, income information was obtained from “Taxation Statistics.” The number of inde­ pendent doctors for earlier years was taken from the decennial censuses, and more recently from annual surveys conducted by the Department of National Health and Welfare. A combina­ tion of special surveys, census data, and infor­ mation supplied by various governmental units formed the basis of estimates for other profes­ sional net income. Since income information was of a fragmentary nature, projections for some groups were based on directly related series. Thus, the net income of lawyers, esti­

mated for 1946-48 from a special survey, was projected to 1950 and back to 1938 on the trend of average net income in the medical pro­ fession and from taxation data. The net income of engineers, architects, designers, chemists, and metallurgists was extrapolated from a 1941 census bench mark to 1945 and to 1938 on the basis of an unweighted average index of indus­ trial production and construction activity. In mining the number of proprietors was ob­ tained from the decennial censuses and pro­ jected to other years on the basis of Canadian gold prices in the case of prospectors and placer miners, and of nonmetallic mineral prices for other mining proprietors. “Taxation Statistics” provided bench-mark estimates of net income for unincorporated wholesale trade, finance, insurance, and real estate. Projections were made by the index of wholesale sales for the wholesale trade and by the index of total values of shares traded on the stock exchanges for stock and bond dealers. For insurance and real estate income, use was made of special surveys by the Bureau’s Mer­ chandising and Services Section. Both the quarterly Labor Force Survey and returns made to the Bureau’s Construction Sec­ tion were the bases for estimating construction income for 1946. This figure was projected on the basis of various indexes of wages and on the estimated number of proprietors. Prior to 1941 the 1931 and 1941 census data on the number of proprietors were interpolated on the basis of indexes of volume and wage rates in construction. Data from the Public Finance and Trans­ portation Division and from the Labor Force Survey provided the basis for income estimates since 1947 in transportation and communica­ tion. Projections to other years were made by indexes of employment and wages and census bench-mark estimates. For repair service for 1946-50, the Labor Force Survey was the basis for estimates of numbers of employers. Average net income was taken as two thirds of the average net income of unincorporated retail trade proprietors. The 1931 and 1941 bench-mark estimations were adjusted to other years on the basis of related series for the various types of repair service. Services of miscellaneous trades (such as per­ sonal, recreational, business services, and board­ ing and lodging) were calculated by sub-groups. For personal services (barbering and hairdress­ ing, dyeing, cleaning and pressing, laundries, un­ dertaking, hotel and tourist camps, restaurants, cafes and taverns), the method consisted largely of multiplying the number of proprietors ob­ tained from the census of population by the average net income based on the retail services census. Projections were then made by various related indexes such as the index of barbers’ fees from the cost-of-living index and gross annual receipts of power cleaning and dyeing plants, as reported to the Merchandising and Services

311

2 1 . CANADA

Bureau, tourist expenditures in Canada from the Balance of Payments Section, and on other related indexes. In recreational service, the bench-mark estimates were projected by an index of receipts (excluding taxes) of motion picture theatres. Net income of accountants (business service) was based on a special sur­ vey conducted in 1944 by the Merchandising and Services Section and was projected to other years on the basis of the index of industrial production. The Bureau’s “Family Expenditure Survey, 1947-1949,” was the basis for estimates of net income from boarding and lodging. One half of the gross revenue in these years was allowed for expenses and the resulting figures of net income were projected to other years by the trend of restaurant sales. National income distributed by factor shares is shown in Table 21-2. Since most of the factor incomes were estimated on an industry basis, the Canadian statistics also give the distribution of national income by industrial divisions, as is shown in Table 21-3. The major adjustments that had to be made in classifying the shares of the fac­ tors on an industrial basis were for capital income.

4. Geographical Distribution of Personal Income 12 The distribution of personal income was estimated separately for each of the prov­ inces. Table 21-4 shows this distribution for 1952 and 1954. The following proce­ dures were used: Salaries, wages, and supplementary labor income for 1938-50 were determined primarily

T able 21-3. N ational Income, by Industrial D ivisions, 1952 and 1954

(million dollars) Business: Agriculture Forestry Fishing and trapping Mining, quarrying, and oilwells Manufacturing Construction Transportation Storage Communication Public utility operation Trade: Wholesale Retail Finance, insurance, and real estate Service Government a Nonresidents Total

1952

1954

2,095 361 67

1,279 338 66

662 656 5,396 5,456 1,025 ■ 1,233 1,200 1,205 47 48 363 317 386 433 920 1,658

927 1,716

1,389 1,422 1,644 -268

1,708 1,598 2,041 -288

18,326

18,774

a Includes government nonbusiness transactions only. Income originating in government business enterprises is included in the appropriate industrial groups with the business sector. s o u r c e : Reference No. 1c, p . 31. from tabulations by provinces of figures sub­ mitted to the Dominion Bureau by various in­ dustries and from census material. When com­ prehensive data were lacking for earlier years, allocation was made in proportion to the shares established in the 1921 and 1931 censuses and those obtaining in 1938. Income received by farm operators from farm production was es-

Table 21-2. National Income by Distributive Shares and Gross National Product, 1952 and 1954 (million dollars) Wages, salaries, and supplementary labor income Military pay and allowances Investment income Net income of unincorporated business: Accrued net income of farm operators from farm production Net income of nonfarm unincorporated business Net national income at factor cost Indirect taxes less subsidies Depreciation allowances and similar business costs Residual error of estimate Gross national product at market prices so u rc e :

Reference No. le, p. 16.

1952 10,868 270 3,763

1954 11,989 367 3,715

1,851 1,574

1,058 1,645

18,326

18,774

2,714 2,120 95

2,914 2,511 -158

23,255

24,041

312

THE IN CO M E OF NATIONS

Table 21-4. Geographical Distribution of Personal Income, 1952 and 1954

5. Gross National Product and Gross National Expenditure 13

(million dollars) Newfoundland a Prince Edward Island Nova Scotia New Brunswick Quebec Ontario Manitoba Saskatchewan Alberta British Columbia Foreign countries b Personal bad debts (not dis­ tributed by province) Canada

1952 229 71 564 408 4,153 6,729 913 1,184 1,253 1,663 22

1954 254 68 617 435 4,617 7,418 898 761 1,197 1,856 24

25

28

17,214

18,173

“ Provisional. b Includes receipts of income of Canadians temporarily abroad, including pay and allow­ ances of Canadian armed forces abroad. so u rc e :

Reference No. 1c, p. 36.

timated annually on the basis of material sup­ plied by the Bureau’s Agricultural Division. Net income of nonfarm unincorporated busi­ ness was distributed on the basis of annual production and sales data prepared by the Bureau, 1941 and 1946 census data, and sur­ veys of net income in certain professional groups. For 1926-37, information from the reports of the Rowell-Sirois Commission was used to obtain the provincial breakdown in a number of industries. Dividends, bond interest, and mortgage interest received by persons were obtained from “Taxation Statistics” for cur­ rent years. For earlier years projections of the percentage distributions prevailing in selected years were used. Rental income was based on the Rowell-Sirois data, government sources, and on the family expenditure survey of ex­ penses. The “Annual Report of the Superin­ tendent of Insurance” was the basis for esti­ mating investment income of life insurance companies and fraternal societies for the years 1936-50. For the period 1926-35, the per­ centage distribution for the provinces was as­ sumed to be the same as for 1936. Other items, such as the investment income of industrial pension funds, profits of nonlife mutual in­ surance companies, etc., was prorated on the basis of the distribution of the total of the above components. Data on transfer payments were obtained from the national and the provincial govern­ ment. Charitable contributions from corpora­ tions were distributed on the basis of the pop­ ulation of the various provinces.

From national income at factor cost, gross national product at market prices was derived. Information on indirect taxes net of subsidies, required for this computation, was available in reports of various Royal Commissions and in federal, provincial, and local financial accounts. All estimates dis­ cussed in the immediately succeeding sec­ tions are expressed in 1935-39 prices. Depreciation charges included special business reserves, losses by fire, accidental damage to business and residential prop­ erty, capital expenditures chargeable to cur­ rent account, and bad debt allowances (less recoveries), but not depletion. Depreciation allowances and similar business costs were based on business current accounting al­ lowances for depreciation, obsolescence, and amortization. This item was estimated by four basic methods. First, the method of consolidation of individual accounts was used for estimates of corporation deprecia­ tion, the claim portion of business insur­ ance, bad debt allowances, and depreciation of certain government business enterprises. Second, the projection from bench-mark data on related series was employed for un­ incorporated manufacturing and retail trade. Third, the ratio method, using directly re­ lated data, was used for most municipal public utilities, certain unincorporated busi­ ness groups, such as agriculture, fishing, construction, wholesale trade, finance, insur­ ance, and real estate, elements of the service group, and the bulk of the estimate of “capital outlay charged to current expense.” Fourth, the ratio method, using indirect data, covered the remainder of unincorpo­ rated business, such as transportation, com­ munication, and certain of the service in­ dustries. For taxable corporations, depreciation charges were estimated since 1944 from “Taxation Statistics”; estimates for other years were obtained from the special cor­ poration sample study. For certain unincor­ porated business, information on deprecia­ tion, obsolescence, and amortization was inadequate and no allowances were made for these items. This was the case for for­ estry, hunting and trapping, and unincorpo­ rated mining enterprises. This omission was not considered serious since the amount of capital employed in these sectors was not

21.

significant. Depreciation on residential build­ ings was estimated in connection with net output. The derivation of the gross national product from the national income at factor cost is shown in Table 21-2 (p. 311). Gross national product was shown as the sum of personal expenditures on consumer goods and services; government expenditure on goods and services for war and nonwar purposes; gross private domestic investment; and net private investment abroad. For the method of computation of these shares of national expenditure see Part II, below.

II. Estimates of National Expenditure by the Final Products Method The estimates of national expenditure prepared by the Dominion Bureau of Sta­ tistics by the final products method are sub­ sidiary to its estimates of national income by the other two methods, and are con­ cerned with the disposition of the gross na­ tional product rather than with that of the national income. The computations for the years 1926-53 were made first in current prices and next in constant prices (see Table 21-5).

1. Gross National Expenditure at Current Prices a. Consumer Expenditures14 Expenditures on consumer goods were estimated by the r*tail sales method. An estimate of retail sales for 1930 was de­ rived from the 1931 census of merchandis­ ing and was adjusted for nonpersonal pur­ chases, sales of secondhand goods, under­ coverage, etc., in the same way as discussed below for 1941. Adjustments for 1931-40 were made on the basis of a specially con­ structed index of retail sales. For 1926-30, estimates of total retail sales were based on statistics of production, imports, and ex­ ports, which were used to extrapolate the 1930 bench-mark figure of personal expendi­ ture on goods back to 1926. Sales of commodities in retail trade for 1941-50, obtained from the 1941 merchan­ dising census and from annual surveys con­ ducted by the Merchandising and Services Section of the Bureau, were adjusted for undercoverage, for retail sales by manufac­

CANADA

313

turing and catering establishments obtained from various sources, and for retail sales by wholesalers that were projected from the merchandising census by a weighted index of retail sales. Retail sales for 1951-53 were similarly estimated from the 1951 merchandising census. Nonpersonal purchases at retail (including total retail sales of building and construction materials, hardware, office and store equip­ ment, business purchases of gasoline, new com­ mercial and passenger vehicles, the value of meals and beverages charged to business and government expense accounts, and similar items) generally were estimated for 1941 by commodity groups and were eliminated. Pro­ jection to other years was made mainly with the aid of specific indexes of retail sales pre­ pared by the Dominion Bureau. For certain expenditures such as fuel charge­ able to landlords’ fuel cost, as explained by the Bureau, “the projection was based on the trend of domestic consumption of certain types of fuel. The business consumption of gasoline, oil and grease, and other affiliated automobile services was projected on the trend of selected retail sales and the weighted index of motor vehicle registration; for others such as the sales of new, commercial and passenger vehicles for business use, the total of new commercial vehi­ cles as well as 20 per cent of new passenger vehicles were assumed for nonpersonal use ex­ cept during the war years.” Sales of secondhand merchandise were ex­ cluded from the totals; only the markup value of the used goods was included. In the case of automobiles, a markup value of 20 per cent for used car dealers and a varying markup for automotive dealers, estimated from the mer­ chandising census, was generally used. This markup was applied to total used automobile sales that was obtained as a residual—i.e., by deducting from the total receipts of the auto­ mobile dealers their receipts from the sale of new vehicles, gasoline, oil and grease, and re­ pairs and services—-and also from information collected by the Merchandising and Service Sec­ tion from used car dealers. Sales and markups for other used goods were also estimated from the census and were projected on the basis of the general retail sales index. Receipts for repairs and other services were deducted since they were included in the serv­ ices category. Retail sales taxes paid separately by the purchasers were added, as were goods purchased by consumers from outside of retail channels. Consumer expenditure for services (pas­ senger transportation, communication, elec­ tricity and gas, laundry, dry cleaning) and expenditures of private nonprofit institutions were estimated mainly on the basis of an­

314

THE INCOM E OF NATIONS

nual gross receipts reported in the census of industry, annual surveys, or published reports of the Bureau and other govern­ ment agencies. These items were adjusted for services rendered to business. Arbitrary allowances had to be made in those cases where sales to business or govern­ ment form an important part of total revenue (e.g., railway transportation, post office). Ex­ penditures on jewelry and auto repairs, funeral and burial expenses, and taxicabs were calcu­ lated from bench-mark estimates obtained from the 1931 and 1941 censuses of merchandising and services projected to other years by ap­ propriate indexes. Gross receipts of commer­ cial recreation establishments were obtained from the 1941 census of services and were projected on the trend of receipts of motion picture theatres, adjusted to include taxes. Per­ sonal expenditure on repairs, including those on shoes, furniture, radios, and appliances were estimated from census bench marks and were extrapolated according to the trend of total retail sales. Gross rent, wages and salaries of domes­ tics, net personal expenditure for board and lodging, professional services, banking serv­ ices, commercial and trade school instruc­ tion, and expenditures of religious organi­ zations were taken, with some modifica­ tions, from the national income estimates prepared by the income-distributed method. Consumer expenditures for goods and services included those in kind, such as farm produce consumed on the farm, board and living allowances, food and clothing issued to the armed services, and gross im­ puted rent of owner-occupied houses. These expenditure figures were based on the esti­ mates used for the income side of the ac­ counts. Finally, net consumer expenditures abroad (by tourists and through private remittances and gifts) were adjusted by in­ cluding those made abroad out of Canadians’ incomes and excluding those made in Canada out of foreigners’ incomes. For the war years, allowance was also made for the expenditures of Canadian service men abroad and of foreign troops in Canada. b. Government E xpenditures15 Government expenditure on goods and services included gross investment by gov­ ernment other than for semiautonomous government enterprises, but excluded trans­ fer payments such as family allowances, old age pensions, interest, purchase of existing

capital assets, international capital trans­ actions of a purely financial nature, and subsidies. For 1926-37, the estimates were based on Royal Commission data on public finance; for the years after 1937, on bud­ getary accounts and direct information from government authorities adjusted, where nec­ essary, from the fiscal year (April 1 to March 31) to the calendar year basis. c. Gross Domestic Investm ent16 Gross private domestic investment com­ prised all expenditure for new durable assets in Canada and net changes in inventories of private and semiautonomous enterprises. Durable assets consisted of new buildings, engineering construction, and other new producers’ equipment. Major alterations were included as well as expenditures for repair and maintenance of producers’ fixed capital charged to current expenses. For the series’ gross domestic investment in new construction and new machinery and equip­ ment, the main source was “Private and Public Investment in Canada, 1926-1951,” published by the Department of Trade and Commerce. Estimates of housing construc­ tion were taken from “Residential Real Es­ tate in Canada.” 17 The Bureau’s capital ex­ penditure survey of the larger firms in pri­ vate business, collected annually since 1941, also provided information on the investment outlays of various institutions. Special gov­ ernment studies were used for earlier years. The annual census of industry furnished data on changes in the value of manufac­ turing inventories for all years from 192650. Information for trade was given in the special corporation study, decennial census material, monthly indexes of wholesale in­ ventories of nondurable goods, “Taxation Statistics”; and, since 1945, in the Bureau’s operating cost survey. Inventory data for mining, lodging, construction, public utili­ ties, transportation, storage, communica­ tion, and services were ascertained from various sources (census of industry returns, annual reports, annual reports compiled by the Bureau, “Taxation Statistics,” the spe­ cial corporation survey, etc.). Changes in inventories of grain and livestock held on farms were estimated by the Bureau from sample data collected annually. Changes in stocks of grain held in commercial chan­ nels, including holdings of the Canadian Wheat Board, were calculated separately for each grade of wheat by multiplying the quantity change in inventories by the aver­

315

2 1 . CANADA

age price paid to producers. All data neces­ sary for these calculations were obtained from “Canadian Grain Statistics” published by the Agriculture Division of the Bureau.

d. Net Investment A broad18 Net investment abroad was derived from the annual Canadian balance of payments with certain adjustments.

2. Gross National Expenditure at Constant P rices19 a. The Nature of the Constant Dollar Series and Implicit Price Indexes The Gross National Expenditure series was deflated to remove the effects of price changes from the current dollar series in order to obtain an estimate of the growth of real production. Price changes were re­ moved by dividing the current dollar values by appropriate price indexes. The 1926-50 series were expressed in constant dollars of the base period 1935-39 (see Table 21-5); the 1950-53 series used 1949 as the base period. The deflation procedure involved three steps. First, a breakdown of the value series (current dollar series) was made in as fine

an item detail as possible. Second, each of the detailed value series was divided by an index specifically selected or specially con­ structed to match the contents of the indi­ vidual value components, thus obtaining a constant dollar series. The third step con­ sisted of summing the detailed items ex­ pressed in constant dollars. The constant dollar estimates contained certain imperfections. A number of cur­ rent dollar value components were divided by an index of costs rather than by one of market prices. Thus, lacking market prices for items such as new buildings and domes­ tically produced machinery and equipment, wage rates and raw materials were combined in order to construct a deflator. This pro­ cedure assumes that market prices moved in the same manner as raw materials prices and wage rates. In addition the available statistics did not permit adjustments to be made for changing productivity, except in the case of residential construction since 1941. Other factors for which it was not possible to allow in the indexes were chang­ ing profit margins, bargain sales, concealed discounts, and premiums. The constant dol­ lar expenditures on government direct serv­ ices (mainly wages and salaries) also as­ sumes constant productivity. Quality changes

Table 21-5. Gross National Product of Canada in Current Dollars and in Constant (1935-39) Dollars, 1952 and 1954 * Current Dollars (million dollars) Personal expenditure on consumer goods and services Government expenditure on goods and services Gross domestic investment New residential construction New nonresidential construction New machinery and equipment Change in inventories Exports of goods and services Deduct: Imports of goods and services Residual error of estimate Adjusting entry Gross national expenditure

Constant Dollars (million dollars)

1952

1954

1952

1954

14,366

15,676

7,382

7,998

4,245

4,361

2,149

2,063

786 1,554 9,916 310 5,573 -5,400 -95

1,166 1,676 1,711 -280 5,136 -5,562 157

284 699 969 220 2,448 -2,426 -47 -1

407 726 849 -120 2,344 --2,508 77 -111

23,255

24,041

11,677

11,725

: price indexes from 1951 to 1954 were converted to a 1935-39 base, using 1949 as the overlap between old and new series. They were then divided into the current dollar figures, to give the constant (1935-39) series. This procedure yields year-to-year movements in the constant dollar series that are identical with those expressed in 1949 dollars, except for a small residual difference produced by the change in the major group weighting pattern. s o u r c e : Reference No. 1c, p p . 16, 50.

316

THE INCOM E OF NATIONS

that occur in commodities over the years cannot be accounted for completely. The special type of quality change that was ac­ counted for was an improvement in “group quality.” Groups of commodities treated sta­ tistically as individual products frequently change in composition. If the proportion of more expensive, better-quality clothing pur­ chased in 1950, were higher, for example, than in 1946, the average quality of the group “clothing” may be said to have in­ creased and this would be reflected by a rise in its current dollar value. When this higher value is divided by an index of prices of specified individual clothing items, the quality measurement will contain an in­ crease because of this improvement in “group quality.” Direct volume measure­ ments will not account for this improvement and in this case deflation methods are su­ perior. The implicit price deflators (see Table 21-5) are currently weighted, i.e., they have changing weights in accordance with changes in the patterns of constant dol­ lar expenditures in successive years and each index is, therefore, only comparable with the base period.

b. Deflation of Personal Expenditure on Consumer Goods and Services For this category, which accounted for about two thirds of gross national expendi­ ture, a relatively larger amount of price data was available for deflation purposes. The majority of the subgroups and a number of individual items of the cost-of-living in­ dex were used in various combinations to deflate some fifty individual categories of commodities and services. However, in some cases special proce­ dures were used. Special price indexes were constructed in the cases of jewelry and tour­ ist expenditures. Various sources of infor­ mation were used, including direct corre­ spondence with manufacturers, and annual catalogues of large mail-order houses and jewelry stores.

c. Deflation of Government Expenditure on Goods and Services The categories utilized for the govern­ ment sector were federal wages and salaries, provincial and municipal wages and salaries, military pay and allowances, interest on the “productive” portion of the government debt, defense expenditures, investment ex­ penditures on machinery and equipment,

building and engineering construction, and a residual consisting of war materials and of routine expenses. For salary and wage estimates base period expenditures were extrapolated on the basis of indexes of the number of civil servants and teachers. Military pay and allowances were adjusted on the basis of a volume in­ dicator consisting of the number of mem­ bers of the armed forces, adjusted for changes in composition of ranks. Interest on the “productive” portion of government debt was deflated, using a mov­ ing average of interest rates and construc­ tion costs. Special indexes were constructed to deflate defense and routine expenses.

d. Deflation of Gross Domestic Investment This group consisted of residential con­ struction, nonresidential building construc­ tion, engineering construction, machinery and equipment, and inventories. For each of the three groups of residential construction—new nonfarm construction, farm construction, and conversions—indexes of residential building materials and wage rates were combined in appropriate propor­ tions to develop deflators. A productivity factor was applied to non­ farm residential construction to account for changes in efficiency from 1941 onward. Changes in profit ratios were not accounted for as studies indicated a relatively stable profit ratio on construction except in periods of subnormal economic activity. Nonresidential building construction and engineering construction were deflated by specially constructed combined indexes of building material prices and wage rates in the construction industry. The lack of data prevented making profit and productivity adjustments. The agricultural component of the ma­ chinery and equipment group was deflated by a price index for farm machinery pub­ lished by the Bureau’s Prices Section. For nonagricultural machinery and equipment, the deflator was a combination of prices of materials used and wage rates. Imports of machinery and equipment were deflated by the price index for this group, as published by the United States Department of Com­ merce; adjustments were made for exchange rates and import duties, to arrive at an in­ dex of prices paid by Canadian users. The constant dollar series of farm inven­

2 1 . CANADA

tories and grain in commercial channels were arrived at on the basis of the physical data for these groups multiplied by base period prices. Business inventories were classified separately by groups, and each group was deflated separately. Each indus­ try or trade type was deflated separately within the manufacturing wholesale and re­ tail groups, while a single composite price index was applied in construction, logging, public utilities, and the remaining groups. The general procedure in deflating busi­ ness inventories involved dividing the yearend inventories for each industrial classi­ fication by price indexes appropriate to the contents of inventories, the turnover periods, and the predominant method of inventory accounting. This results in inventories ex­ pressed in base period dollars, and the yearto-year change in them is a measure of physical change for each quarter. The cur­ rent value of the physical change was ob­ tained by multiplying the physical change by average quarterly prices. The difference between the current value of the physical change and the original book value change is called “inventory valuation adjustment.” The sources for the index used to deflate business inventories came from the Prices Section and the Industry and Merchandising Division of the Bureau. The turnover peri­ ods for nonmanufacturing groups were ar­ bitrarily estimated. The inventory valuation adjustment was deducted from gross na­

317

tional expenditure so that the value change of gross national expenditure may be sep­ arated into its price and quantity com­ ponents.

e. Deflation of Exports and Imports Merchandise exports and imports were each deflated by specially constructed cur­ rently weighted price indexes. Current weights were necessary because of the shifts in postwar patterns of trade. The Interna­ tional Trade Division of the Bureau pro­ vided the annual deflator. The primary source of information for quarterly esti­ mates was the publication, “Trade of Canada.” Expenditures of foreigners in Canada were deflated by an index based upon the prices of goods and services nor­ mally purchased by tourists in Canada. The tourist and travel expenditures of Canadians abroad were deflated by the prices in the United States of goods and services pur­ chased by tourists, with adjustments for ex­ change rates in foreign countries. A similar procedure was followed in deflating expend­ itures by Canadian servicemen abroad, ex­ cept that a combination of United Kingdom, French, and German prices were used. Per­ sonal remittances, received and paid, were deflated by the consumer expenditure price deflator. The net nonmonetary gold series was based on price and exchange rate data. Special deflators were constructed to deflate freight and shipping debits and credits.

22. IRELAND * References 1. “National Income of the Irish Free State,” Economic Journal, March, 1933, pp. 74-87; “National Expenditure of the Population of the Irish Free State in 1926,” Journal o f the Statistical and Social Inquiry Society of Ireland, 86th Sess., vol. 43, 1932-33, pp. 91-100. 2. G. A. Duncan, “National Income of the Irish Free State,” Majority Report o f the Com­ mission of Inquiry into Banking, Currency, and Credit, Dublin, 1938, pp. 426-57. 3. G. A. Duncan, “Social Income of the Irish Free State, 1926-38,” Journal of the Statistical and Social Inquiry Society of Ireland, 93d Sess., 1939-40, pp. 1-16; and “Social Income of Eire, 1938-40,” ibid., 94th Sess., 1940-41, pp. 140-41. 4. Ministry of Finance, Ireland, National Income and Expenditure, 1938-1944. Report pre­ sented to The Oireachtas (Parliament) by the Minister for Finance, Dublin, March, 1946 (Pr. 7356). 5. Central Statistics Office, Irish Statistical Survey, 1951-52, Dublin, 1953 (Pr. 1983), ibid., Irish Statistical Survey, 1953, Dublin, 1954 (Pr. 2607), hereafter referred to as Irish Sta­ tistical Survey, 1953; also Irish Statistical Survey, 1954 (Pr. 3052). 6. R. C. Geary, “Statistical Methods used in Ireland in the Compilation of National Income, National Expenditure, Capital Formation, and Other Cognate Matters with Some Observa­ tions on Analogous Problems in Other Countries,” for discussion at the meeting of the Inter­ national Association for Research in Income and Wealth at Royaumont, August, 1951 (Mimeo.).

1. History

2. C oncepts and General M ethods

T he first estim ate of th e national income of the Irish F ree State was prepared in 1933 by T. J. K iernan of the N ational University o f Ireland and covered the year 1926. In 1938, under the auspices of a G overnm ent Commission of Inquiry into Banking, C ur­ rency, and Credit, G . A. D uncan o f the U niversity of D ublin m ade m ore com pre­ hensive estimates fo r 1929 and 1931-35. L ater he carried his estimates back to 1926 and forw ard to 1940. In 1946 an official estimate of national incom e and expendi­ ture for 1938-44 was prepared in accord­ ance w ith the new approach adopted by agreem ent am ong several English-speaking countries. T he official estimates are prepared at present by the C entral Statistics Office and are published annually in the Irish Statistical Survey. We discuss only the official estimates.

T he estim ators adopted the com prehen­ sive concept of national income. Irish n a­ tional incom e was estim ated in the m ain by the n et o u tput m ethod, b ut the incomedistributed and final expenditure m ethods were also em ployed fo r certain elem ents as well as fo r a cross-check against the m ain net o u tput estim ates. F or example, profits of m anufacturing industries w ere estim ated by relating them to sales data obtained from retu rn s furnished to the Revenue Com m is­ sioners. T he estim ates w ere classified b oth by facto r shares and by incom e originating in th e various sectors. T able 22-1 shows Irish national incom e distributed by facto r shares; T able 22-2 presents the national incom e by sectors o f origin.

* Acknowledgment is due to Mr. R. C. Geary, formerly Director, Central Statistics Office, Ireland, for reviewing the draft of this chapter and offering helpful suggestions.

a. Agriculture, Including Fishing

3. N et O utput and Income-distributed Estim ates

Notes begin on p. 537. 318

T he agriculture estimates w ere developed by the traditional net o u tp u t m ethod. The

2 2 . IRELAND

319

Table 22-1. National Income of Ireland by Factor Shares for Selected Years, 1938, 1949, 1953, 1954 (million £) Income from agriculture, forestry, fishing: 1. Profits (including income of farmers and mem­ bers of their families) 2. Wages and salaries Nonagricultural income: Profits, professional earnings, interest, dividends, and income from lands and buildings: Domestic 3. Public and private companies’ profit before tax 4. Other profits, professional earnings, interest, etc. 5. Income from buildings (personal and business rent) 6. Rent element in land annuities Foreign 7. Net inflow of profits, etc., from the rest of the world Wages, salaries, pensions: Domestic 8. Wages, salaries, pensions 8a. Employers’ contributions to social insurance Foreign 9. N et inflow of wages, salaries, pensions from the rest of the world Other 10. Net inflow of emigrants’ remittances, etc., from the rest of the world 11. Less provision for stock appreciation 12. Total national income 13. Domestic 14. Foreign

1938

1949

1953

1954

28.0 * 6.7 *

84.1 16.1

115.16 16.4

116 19

11.4

32.0

39.0

42

19.7

27.6

30.8

31

7.0 2.2

8.0 2.2

8.0 2.2

8 2

6.3

11.2

13.6

14

68.7 1.2

142.7 2.1

191.7 2.8

197 3

3.0

4.0

4.8

5

3.0 (a)

9.7 - 0 .8

10.7 - 3 .1

11 —

157.2 144.9 12.3

338.9 314.0 24.9

438.7 409.6 29.1

448 418 30

( a ) Not a v a i l a b l e . * Revised. s o u r c e s : Irish Statistical ( f o r 1949 e s t i m a t e s ) .

Survey, 1954,

p.

47

gross value of agricultural production was taken as the gross receipts from the sale of agricultural produce, tim ber, and turf. This estim ate of the sales of off-farm products (excluding interfarm transactions) together w ith the on-the-farm consum ption of prod­ uce, yielded an estim ate of the gross value of agricultural output. Estim ates of the consum ption of produce by persons in farm households w ere based partly on a sam ple inquiry m ade at the tim e of the estim ation of crop yields. E ach N ovem ber fo r several years p rio r to 1955, 18,000 farm s were sam pled and inform a­ tion was obtained as to the weekly consum p­ tion of eggs, butter, milk, potatoes, and cab­ bages, and the annual consum ption of poul­ try in the farm household.1 T he N ovem ber consum ption was assumed to be typical of

(fo r e s tim a te s

1938, 1953,

and

1954);

and

1953, p. 49

average annual consum ption of some prod­ ucts. T he consum ption of some products could be checked from other sources. Thus, the annual production of m ilk less estimates of m ilk consum ed by persons, m ilk sent to cream eries, and m ilk fed to livestock should give the same result fo r the output of m ilk used to m ake farm butter as the sample inquiry indicates. Statistics on tu rf draw n from bogs and on pigs slaughtered fo r use in farm households were also ob­ tained in the annual agricultural enum era­ tion. T he output of produce consumed in farm households is valued at the prices that farm ers receive fo r sim ilar goods sold on th e m arket.2 A lthough estimates of farm produce sales were generally firm er th an those for prod­ ucts consum ed on the farm , they varied con­

320

THE INCOM E OF NATIONS

Table 22-2. Net National Product of Ireland a t Factor Cost by Sectors of Origin, for Selected Years, 1938, 1949, 1953, 1954 (million £) Sector Agricultural: Agriculture, forestry, and fishing Nonagricultural: Industry Distribution and transport Other domestic (including personal and busi­ ness rent) Public administration and defense Less provision for stock appreciation Rest of world: Emigrants’ remittances Other Total net national product a t factor cost

1938

1949

1953

1954

36.9

102.4

134.2

137

96.4

78.0 53.5

106.5 60.4

111 62

11.6 (a)

56.1 24.8 - 0 .8

68.9 36.5 - 3 .1

71 37 —

3.0 9.3

9.7 15.2

10.7 18.4

11 19

157.2

338.9

438.7

448

(a) Not available. so u rce:

1949

Irish Statistical Survey, 1954, p. 48

(fo r

1938, 1953,

and

1954

e s tim a te s ),

and 1953, p. 50

(fo r

e s tim a te s ).

siderably in their reliability, w ith th e esti­ mates fo r m ajor products being usually bet­ te r than those fo r m inor products. Those estimates considered firm w ere the output of milk supplied to cream eries for the m anu­ facture of butter and cheese, statistics of sales of pigs, purchases of cash crops by the processing industry, estimates of cattle out­ put, and cattle exported. H om e slaughter­ ings were derived from 100 p er cent returns of slaughtered hides from hides m erchants. O utput of sheep and lam bs was estim ated in a sim ilar m anner. O ther anim al products were estim ated as follows: M ilk output figures are obtained at the sum­ mer census when farmers are asked to state the quantity of milk obtained “yesterday” from their cows. Since individual farmers do not know the annual output of most of their prod­ ucts, farmers are merely asked to state the production of eggs and milk for the day before the enumerator calls. These figures are grossed and seasonally adjusted to give an annual aver­ age yield per cow and production estimates are made on the basis of these average yields and the number of cows. A further adjustment is made from Department of Agriculture data on milk delivered to creamery establishments. Egg production statistics are obtained at the annual summer census of livestock. Informa­ tion on yields for one day—again the day be­ fore the enumerator calls— are aggregated and seasonally adjusted to give the total annual production of eggs. The number of eggs used for hatching, estimated from the number of young fowl in June, is deducted to arrive at

the output. A second estimate, similar in method to the first, is made in January from a sample enumeration of livestock covering one quarter of the agricultural holdings. The third method of estimating output is based on sta­ tistics of farm consumption of eggs derived from the November sample inquiry and data on sales off the farm obtained from informa­ tion supplied by wholesalers to the Department of Agriculture. This is assumed to represent 80 per cent of total farm sales. Two separate estimates for poultry produc­ tion are made. The first is derived from esti­ mated annual hatchings based on the numbers of young birds on June 1 after allowance for mortality and stock replacement. The second estimating method was based on farm house­ hold consumption figures obtained prior to 1955 at the November sample inquiry, plus sales off farms that are estimated from data collected at the Winter sample enumeration of livestock. The output for smaller types of poul­ try (ducks and ordinary fowl) is based on the change in numbers of these birds on farms as shown at the winter and summer livestock enumerations. The different methods of esti­ mating poultry output may give different re­ sults in which case one method, or an average of a number of methods, is chosen. Estim ates of o u tput of potatoes consist o f the consum ption on farm s, derived prior to 1955 from th e N ovem ber sam ple in ­ quiry; 3 consum ption by the nonagricultural population; and the am ount going into in­ dustry and exports. G ross ou tp u t estimates include paym ents in cash an d k ind to farm em ployees. A gri­

2 2 . IRELAND

cultural wages are based on estim ates of the num ber of wage earners taken from an annual enum eration and an average m ini­ m um rate is applied. Estim ates of feeding stuffs, fertilizers, seeds, rents, and oth er expenses are based p artly on total enum erations and partly o n special surveys. T he expenditure on farm m achinery has been treated, in the absence o f any m ore appropriate figure, as the allowance fo r depreciation in agriculture, although in recent years this figure m ay contain a fair elem ent o f new capital ex­ penditure. T he annual figures fo r the ex­ penditure on farm m achinery were sm oothed before being used as estim ates fo r deprecia­ tion. These item s are deducted from the figure for gross o utput to arrive at the net value added by the agricultural sector. T he product of fishing is entered at its gross value (the total catch as la n d ed ), rath er than its n et value. T he failure to deduct purchases from other industries is justified on the ground th a t the volum e of the catch is underestim ated anyw ay and also on the ground th a t the entire product is not very large. F o r the distribution of incom e by factor shares, agricultural incom e was divided into wages, salaries, and profits. Profits w ere es­ tim ated as a residual after deducting agricul­ tu ral wages and salaries from the net out­ p u t estimates.

b. Manufacturing Estim ates o f the contribution of the m anu­ facturing industry to national incom e w ere derived largely from the A nnual Census of P roduction, w hich contained d ata on gross value of output and costs. T he Census has practically com plete industrial scope, ex­ tending to all concerns in w hich three or m ore persons are em ployed. Inform ation from the Census as to total em ployee rem u­ neration was supplem ented w ith voluntary quarterly inquiries o f o utput of concerns producing transportable goods. T he net out­ p u t was estim ated largely as the sum of lab o r incomes and profits, i.e., by the in­ com e-distributed m ethod, as follows: Comparison of the Census of Production with the Census of Population showed that persons engaged in “excluded” industrial em­ ployment—concerns with less than three per­ sons employed—numbered less than one fifth of persons engaged, but their net output was only one tenth of the total of all industries, large and small. For many of these smaller

321

concerns net output was obtained by the direct method. Estimates of employee remuneration in “excluded” concerns were based on a volun­ tary sample inquiry that yielded information on remuneration per employee and a percent­ age distribution by industrial status by main industrial groups. These groups consisted mainly of small builders, motor repair shops, boot repairers, blacksmiths, confectioners, dress­ makers, and tailors. Average employee re­ muneration in small concerns was applied to the number of workers in Census of Population years. The data were extrapolated and inter­ polated for noncensus years. Members of pro­ prietors’ families working in the business were given an imputed income equivalent to the average wage-earner remuneration. Data on the profits of the larger concerns, those having taxable profits of £2,000 or more, were obtained from the Revenue Commission­ ers. Information on taxable profits, sales, pur­ chases, stocks, wages and salaries, distributed and undistributed profits, etc., were segregated on an industrial basis; and, for each group, profits are taken as a percentage of sales plus other income. Total profits for the larger in­ dustrial concerns were then obtained by apply­ ing these percentages to the corresponding in­ dustry group totals of sales plus other income as given in the Census of Production. To obtain a figure for profits for the smaller concerns excluded from the Census of Produc­ tion, the average remuneration of salaried em­ ployees in each group reported in the voluntary sample inquiry was applied to the estimated number of proprietors in that group on the ground that there was correspondence between such profits and such salaries within each group.

c. Trade and Transport Before inform ation becam e available from the annual Census of D istribution begun in 1951, net o u tput in the distribu­ tion sector was derived from various sources. It was obtained indirectly in p art by first estim ating gross output, and directly in p art from paid-out factor shares. F rom external trad e data, the census of production, and agricultural output statistics, the total supply of goods available was know n. By deducting the cost of m aterials or goods used in fu r­ th er production and fo r export, a residue was obtained th a t represented goods enter­ ing distribution channels. Average retail prices w ere applied to the quantities of prin­ cipal foods and other fairly standardized goods. W here retail prices were n o t avail­ able, percentage additions ( derived from the Price C ontrol A uthority) were applied to give th e final selling value. A djustm ents w ere m ade fo r overestim ation owing to re­ duced price sales to hotels, public assistance

322

THE INCOM E OF NATIONS

bodies, etc., and for inclusion of transpor­ tation costs. A sample of the annual ac­ counts furnished to the revenue authorities yielded data on profits, wages, salaries, and changes in stocks. A dditions to value added in trade were m ade to cover the distribu­ tion of raw m aterials to industry. In 1951 a change was m ade in the gen­ eral estim ating procedure. A s M r. G eary puts it:

rep air an d m aintenance w ere taken from the Census of P roduction and from inform a­ tion on the sm aller building concerns. H ow ­ ever, data on costs an d gross rents were n ot sufficient to indicate the tren d in n et rent and, consequently, for the years following 1946, net ren t was kept constant. R ents on business quarters were taken from the rev­ enue figures. T he incom e from oth er services was estim ated as follows:

A full Census of Distribution was taken in 1951 and a sample census has been taken every year since then. Wages and salaries, sales and changes in stocks are estimated from these sample inquiries. For the concerns included in the sample, figures for sales, wages and salaries, profits, etc. are abstracted from revenue returns. The sample is divided, on the basis of sales, into three size groups and, for individual industries in each group, taxable profits are expressed as a percentage of sales plus other income. Total profits in each size group are then obtained by applying these percentages to the sales in the corresponding size group obtained from the Census of Distribution returns, and aggregating the results.

Employment exchanges furnished average wage rates for many service employees. Aver­ age wages were weighted by data obtained from the Census of Population. Payments in kind were estimated for one year and adjusted for subsequent years by the cost-of-living index. Tax receipts furnished most of the other data on services such as hotels and restaurants, en­ tertainment, etc. Profits of hotels and restaurants were esti­ mated from data on wages and salaries as obtained from the Revenue Commissioners. These were calculated as a percentage of net output and the residual is the profits estimate in this sector. Similar methods giving profits as a residual were used for other services. For entertainments other than theaters and movies, net output was estimated using tax and other data; and, with information from the Census of Population, an allocation between wages and salaries and profits was made. Certain mis­ cellaneous services had to be dealt with arbi­ trarily; however, the incomes involved were small. Since 1951, wages and salaries and profits in hotels and restaurants are now estimated from information obtained in the Census of Distri­ bution and from returns furnished to the Rev­ enue Commissioners. Similar methods are used for some other services. The civil service census provided informa­ tion on the wages and salaries of government employees. Special returns gave data for the Army and police force. Local government in­ formation was obtained from annual inquiries. The questionnaire method was used to ob­ tain data on employee remuneration and on numbers employed in banks and insurance and the Irish Hospitals Trust. Profits of banks were obtained from their published accounts, and deposit interest and wages and salaries were added to give net output. Profits of insurance companies were obtained from published state­ ments and by calculating changes in various reserves. The Department of Education provided data for estimating the net output of teachers. Other professional occupations were covered by data from the Revenue Commissioners and were computed on the basis of average remunera­ tion and the number of professional people in each group.

Incom e contributed by transport services was obtained from statistics of earnings of employees in a certain week in each year as reported to the governm ent. In some transport concerns, earnings w ere estim ated as a percentage of net receipts from returns furnished by these concerns. N et output for taxicabs was estim ated on the basis of an average per vehicle applied to the num bers registered; of this net output 50 per cent was allocated to wages and salaries and 50 per cent to profits. Profits of rail com panies w ere obtained from their published accounts after receipt of subsidies. Profits of other road passenger transport w ere estim ated as 50 per cent of net output, w hich was obtained from fu r­ nished accounts. F o r other road m erchan­ dise transport concerns, profits w ere taken as the difference between total receipts and expenditure.

d. Other Services A gricultural rent was taken as p art of the land purchase annuities paid to the L and Commission. This was included in the net output of agriculture. Inform ation as to the ownership of private dwellings was estimated from the detailed inform ation on rents ob­ tained in conjunction w ith the 1946 Census of Population. Figures for expenditure on

22.

e. Rest of the World N ational product was adjusted by the pensions and allowances paid by the British governm ent to ex-servicem en residing in Ireland less the paym ent of pensions by the Irish governm ent to persons outside the country. This inform ation is obtained from governm ent records. D ata on the receipts from and paym ents to abroad (interest, dividends, etc.) w ere supplied by the R ev­ enue Com missioners. Rem ittances of em igrants, unlike other country estimates, are included in the n a­ tional income on the theory th a t they are sim ilar to investm ents abroad.

4. Final Product or Expenditure Estim ates G ross dom estic capital form ation was estim ated directly by com bining capital goods produced at hom e together w ith changes in stocks in industry obtained from the Census of P roduction, and capital im ­ ports from trade im port statistics. Repairs and consum ers’ durables were excluded from hom e-produced capital form ation. This was considered a fairly firm estimate. Estim ates of changes of stocks in industry w ere obtained from the Census of P roduc­ tion. They were available fo r individual in­ dustries and w ere separated into (a) stocks of raw materials, ( b ) w ork in progress, and (c) stocks o f finished goods.4 T hose for trade were available since 1951 in the an­ nual Census of D istribution. Inform ation with regard to changes in stocks, therefore, was now fairly com plete. D epreciation was calculated as the am ount of the annual expenditure in the agricultural sector fo r agricultural m achin­ ery, sm oothed over a num ber of years. W hile this item m ight have included a sig­ nificant am ount of new capital expenditure in addition to capital replacem ent, it was used in the absence of any m ore appropriate figure. F or nonagricultural sectors, deprecia­ tion was derived from tax statistics, i.e., taken as the allow ance reported on the tax returns fo r wear and te ar and obsolescence. Since depreciation was based on tax allow­ ances, it was related to original rather than replacem ent cost. D epreciation in the pub­ lic sector was taken as the sum of the pay­ m ents of the central governm ent to th e N a­ tional L oan Sinking Fund, L and Bonds re­

IRELAND

323

deem ed, am ortization on the Local Loans F und and the loan repaym ents of local au­ thorities (other than the Local Loans F u n d ), plus an allow ance fo r the estim ated value o f accidental dam age to fixed capital. D irect estim ation of personal expenditure on consum ers’ goods and services was m ade fo r the period 1938-44 on the basis of data from the Census of Production, which gave d ata on consum ption goods entering trade, and statem ent of agricultural output and external trade expressed at retail prices. These w ere supplem ented by estimates of produce consum ed on the farm at agricul­ tural prices and of estimates of expenditures on services. These estimates were larger than the total retail sales estimates obtained from sample investigations. F o r this reason estimates of personal consum ption expendi­ tures were taken as the residual in the table of net national expenditure at factor cost fo r the national income accounts. These residual estimates are about 10 p er cent below the estim ates obtained by the direct m ethod. T he n et expenditure by public authorities on curren t goods and services at m arket prices was estim ated as the residual of ex­ penditure after deducting expenditure allo­ cated to other categories. T he estim ating o f net foreign investm ent was com plicated by the lack of exchange control between Ireland and G reat Britain and by the invisible transactions th at form a large proportion of the Irish Balance of Paym ents. C apital and cu rren t debit items exceeded credit items by approxim ately £ 7 - 8 millions a year. This unknow n credit item was taken to consist m ainly of over­ valuation of im ports, undervaluation of ex­ ports, fees fo r bloodstock breeding, and smuggling of cattle, tobacco, etc. This item was allocated am ong agricultural and dis­ tribution profits. A llow ance fo r smuggling was m ade in estim ating personal consum p­ tion expenditure. F rom a sample inquiry, estimates of net receipts from visitors’ expenditures were m ade. In addition, sam ple investigations into the valuation of im ports and exports w ere undertaken (see T able 22-3). T otal savings w ere defined as the differ­ ence between n et capital form ation at hom e and disinvestm ent abroad, i.e., by deducting from domestic capital form ation the portion financed by draw ing on external assets. T hus, total savings were a residual item.

324

THE INCOM E OF NATIONS

Table 22-3. Gross and N et National Expenditure of Ireland, for Selected Years, 1938, 1949, 1953, 1954 (million £)

Personal expenditure on current goods and services a t market prices Net expenditure by public authorities on cur­ rent goods and services at market prices Gross domestic capital formation: Value of physical increase in stocks and work in progress New producers’ capital goods brought into use: Imported Home produced Net foreign investment Gross national expenditure a t market prices Less provision for depreciation Net national expenditure a t market prices Plus subsidies Less indirect taxes Net national expenditure (product) a t factor cost—-national income

1938

1949

1953

1954

141.5

306.2

375.8

393

22.9

43.7

69.8

69

(a)

- 1 .5

3.6

-5

3.1 14.0 + 2 .0

12.7 33.9 - 9 .7

22.2 52.1 - 7 .0

21 53 -6

183.5 - 4 .2

385.3 - 9 .0

516.5 - 1 4 .0

525 -1 5

179.3 + 2 .0 - 2 4 .1

376.3 + 15.1 - 5 2 .5

502.5 + 10.2 - 7 4 .0

510 + 13 -7 5

157.2

338.9

438.7

448

(a) Included in personal expenditure. so u rc e :

Irish Statistical Survey, 1954,

p.

49

(fo r

D irect estimates w ere m ade of its two com ponents, i.e., undistributed corporate profits and savings by public authorities. The form er item was obtained from p ub­ lished revenue statistics on profits of com ­ panies paying the corporate profits tax, plus the estim ated taxable profits of com panies not so assessed. A sample inquiry by the Revenue A uthority, classified as to corporate and unincorporated enterprises, gave an es­

1938, 1953,

and

1954),

and

1953,

p.

51

(fo r

1949).

tim ate of undistributed profits as a percent­ age o f total profit. T his was applied to total corporate profits. Public authorities saving was th e residue of governm ent receipts, ex­ cluding borrow ing, after expenditure on subsidies, transfers, and cu rren t goods and services. Personal saving was the residual item obtained fro m total savings after de­ ducting undistributed corporate profits and public authorities saving.

23. THE NETHERLANDS * References 1. a. Centraal Bureau voor de Statistiek, Onderzoek naar den invloed van het arbeidsloon op den kostprijs der producten, The Hague, 1933. b. “Enkele Berekeningen over het Nationale Inkomen van Nederland” (Some Computations of the National Income of the Netherlands), De Nederlandsche Conjunctuur, Speciale Onderzoekingen, No. 2, The Hague, 1939. (The estimates were prepared by Dr. J. B. D. Derksen.) c. Ph. J. Idenburg and J. B. D. Derksen, “Methods of Evaluation of the National Income of the Netherlands, 1921-1938,” Revue de Vlnstitut International de Statistique, vol. 8, 1940, issues 1 and 2, pp. 13-31. d. J. B. D. Derksen, “H et geproduceerde en het uitbetaalde nationale inkomen en de inkomens betaald aan natuurlijke personen (1921-1938),” Maandschrift van het Centraal Bureau voor de Statistiek, vol. 36, March-April, 1941. 2. a. Centraal Bureau voor de Statistiek, Nationale Boekhounding (National Accounting), The Hague, September, 1944 (Mimeo.). b. The Netherlands Central Bureau of Statistics, National Accounts o f the Netherlands, 1946 and 1947, and ibid., 1948-1949, The Hague, 1952. c. Centraal Bureau voor de Statistiek, H et Nationale Inkomen van Nederland 1921-1939, Monografieen van de Nederlandse Conjunctuur No. 7, Utrecht, 1948. (The National Income of the Netherlands, 1921-1939.) d. Centraal Bureau voor de Statistiek, De Nationale Jaarrekeningen: doeleinden, problemen, resultaten (National accounts, objectives, problems, and results), Monografieen van de Nederlandse Conjunctuur No. 8, Utrecht, 1950. e. Centraal Bureau voor Statistiek, Nationale Jaarrekeningen 1948-1950, Utrecht, 1953 (N a­ tional Accounts 1948-1950). 3. Organisation for European Economic Co-operation, National Accounts Studies, Nether­ lands, Paris, 1941, and Netherlands, Annex, Paris, 1951.

1. History Professor W. A. Bonger introduced sys­ tem atic national incom e estim ates in the N etherlands in 1910 w ith the publication of figures for 1908 based on taxable incom e and other data. H e follow ed this study after W orld W ar I w ith investigations of the n a­ tional incom e fo r 1912-20.1 H is lead was follow ed by several individual scholars,2 a governm ental com m ission on econom ic pol­ icy in 1929,3 and, finally, beginning in 1933, by the C entral Statistical Office.4 T he latter’s estimates, prepared on a continuous * Acknowledgment is due to Dr. J. B. D. Derksen, formerly of the Central Bureau of Statistics of the Netherlands and later chief of the National Income Unit of the United N a­ tions, for reviewing this chapter and supply­ ing a note on the developments during the period 1940-55. Notes begin on p. 537. 325

basis, covered the years 1921-38 and were based on the n et o u tput as well as the incom e-distributed m ethods. In the later official studies, national incom e figures were carried backw ard to 1900 and w ere also broken dow n by individual provinces. They included estim ates of capital form ation, com prehensive analyses o f national product and expenditure, interrelated social ac­ counts, and projections of current gross national p ro d u ct into the fu tu re in th e form o f long-range “econom ic budgets.” 5 T oday, th e N etherlands is one of the few countries having consistent national income series covering a long period and prepared by independent methods. T he present study concerns itself mainly w ith th e official estim ates prepared during the 1930’s chiefly by the income-distributed m ethod. In the last section a brief summ ary

326

THE INCOM E OF NATIONS

is given of the principal changes m ade in the m ethodology and subject m atter of the estimates prepared during the period 1940 55, in w hich the social accounting approach is em phasized.

th e State an d several cities; th e educational statistics o f the C entral Statistical Office; reports o f religious institutions; and the Statistical A b stra ct o f the N etherland In ­ dies.

2. Concepts and Sources

3. Incom e-distributed Estim ates

N ational income is defined as the net value o f all goods produced and all services rendered in the country in a year, including net incom e from abroad. T his value is iden­ tical with the total of labor and capital in ­ come, including net ren t on ow ner-occupied houses. Interest on the national debt has been included on the assum ption that the loans represent tangible capital assets. U n ­ em ploym ent benefits have been deducted as transfer income, but n o t public relief w ork paym ents, retirem ent, pension, and social security benefits. T he net balance of international incom e consists of capital in ­ come, transfers of pensions, and private remittances. F o r the estim ate of national incom e by the incom e-distributed m ethod, the m ain source of statistical data was the income tax statistics published by the C entral Sta­ tistical Office; while official production sta­ tistics for agriculture, m ining, and industry were the basic sources of th e estimates by the net output m ethod. Since these sources provided only p art of the required data in either case, it was necessary to tu rn to m any special statistical publications and studies fo r supplem entary inform ation. A m ong them, the following w ere of particular im ­ portance: T he Census o f Population and Occupation; the wage statistics of the N a ­ tional W orkm en’s C om pensation B oard ( Ongevallenstatistiek van de R ijksverzekerings Bank)-, the collections of corporation profit and loss accounts in D r. Sternheim ’s K roniek and Van Oss E ffektenboek; th e es­ tim ates o f the balance of paym ents; and the investigation of the C entral Statistical Office on the structure of the costs of production (O. Bakker, O nderzoek naar den invloed van het arbeidsloon op den kostprijs der producten). Special inform ation covering certain industrial divisions was supplied by the official statistics of constructed dwell­ ings; the figures on aw arded contracts 6 p ub ­ lished in the weekly paper, D e A annem er (C o n trac to r); the reports of building so­ cieties; the official statistics of fishing (Jaarverslag der Visscherij); of banking and in­ surance; o f enterprises in transportation and com m unications; the financial accounts of

W hile th e national incom e was estim ated according to b oth n et ou tp u t an d incomedistributed m ethods, special em phasis was p laced u p o n the second m ethod. T his was because the incom e tax statistics were m uch m ore reliable th an the production and cost statistics.7 T he figures fo r 1900-20 were p rep ared exclusively by this m ethod. H ow ­ ever, since incom es from agriculture and wages an d salaries w ere n o t taxable prior to 1914, estim ates fo r these tw o im portant com ponents h ad to be prepared by other m ethods. L ike all incom e estimates th a t hinge on incom e tax statistics, those fo r the N ether­ lands required certain additions to the re­ po rted taxable incomes, as well as certain deductions therefrom . T he n atu re of these adjustm ents are show n in T able 23-1. Table 23-1. National Income of the Netherlands by the Income-distributed Method, 1938 (million guilders) Income assessed for income tax 3,010 Additions: Exemptions for children 221 Incomes below tax exemption limit 1,229 Other exemptions from personal income tax: Contributions to trade unions 11 Premiums on life insurance 116 Other tax exempt income: Government and employers’ social se­ curity contributiona 149 Undistributed corporate profits 240 Capital income of insurance companies 115 Profits of governmental enterprises 56 Income of charities 60 Tax evasion 301 Deductions: Time lag in dividend payments —7 Transferred incomes b —348 National Income

5,153

a Governmental contributions to Old Age In­ surance excluded. b Unemployment benefits, dismissal compensa­ tions to government employees, old age pensions and annuities, and other social security benefits. s o u r c e

:

Reference N o.

1

d.

23.

THE NETHERLANDS

a. Taxable Incom es T axable income included all incom es ex­ ceeding the tax exem ption lim it of 800 guilders fo r single and m arried persons w ith­ out offspring, and the higher lim it fo r fam i­ lies with children. F o r the purpose of the incom e estimates, the assum ption was m ade that the tax levied in the year of assessment referred to incomes earned one year earlier. F o r instance, the statistics fo r the fiscal year of assessment 1936—37 w ere used to esti­ m ate national incom e in 1935.

b. T a x-exem p t Incom es; E xem p tio n s for Children T he tax law granted credits in th e form of exemptions, depending upon the size of the fam ily incom e and the num ber of chil­ dren. In the tax statistics these exem ptions w ere n ot subtracted, b u t the incom es that fell short o f th e tax exem ption lim it o f 800 guilders, after deduction of these credits, were simply om itted. T o adjust fo r this loophole, the assum p­ tion was m ade th a t the distribution of fam i­ lies, according to the num ber o f children, in each income bracket betw een 800 and 3,000 guilders was the sam e as show n by the 1930 Census of Population. T he num ­ ber of families thus calculated was m ul­ tiplied by an average incom e determ ined in accordance w ith th e n um ber o f children an d the corresponding credits as provided by the tax law. In this way, the total am ount of incom es over 800 guilders was obtained.8 Incomes below tax exemption limit. T he incom e recipients of this im portant class consisted of those gainfully em ployed (in ­ cluding tem porarily unem ployed) 9 and those n ot gainfully em ployed. T he incom e of the form er category was estim ated by deducting the num ber of incom es over 800 guilders from the total num ber of persons occupied, and m ultiplying the rem ainder by an assumed average incom e. A djustm ents w ere m ade fo r m arried w om en in occupa­ tions because their income was already in­ cluded in the joint tax retu rn o f husband and wife. T he 1920 and 1930 Census data on occu­ pations were projected to the other years on the assum ption of the same am ount of increase from year to year, and an average of 600 guilders per incom e recipient (based on some fragm entary data on wages of young laborers, incomes of sm all farm ers, unem ploym ent benefits, and wages earned in dom estic service) was applied throughout.

327

T he incom es in the other category, nam ely, those of persons living exclusively on pensions or returns from small property, w ere determ ined in a sim ilar way by ob­ taining figures from various sources on the num ber of incom e recipients and average incom es.10 T he estim ated am ount varied only slightly, between 50 and 60 million guilders, fo r the different years.

Other exemptions from personal income tax. Since the D utch incom e tax allowed deductions from personal incom e fo r con­ tributions to trad e unions, dividends paid by life insurance com panies to policyholders and by consum er co-operatives to their m em bers, and fo r incomes from forestry, it was necessary to adjust fo r the omission o f these items. H ow ever, adjustm ent was m ade only fo r the first and m ost im portant of these items, nam ely, contributions to trade unions. A special adjustm ent was necessary be­ cause of a peculiar feature of the Revenue A ct th a t included annuities and pensions in taxable incom e b u t allowed a deduction fo r prem ium s below a certain sum. T he cor­ rection consisted o f deducting the pensions and adding back the prem ium s. It covered the state pension funds, the life insurance com panies, railroad and m ine w orkers’ pen­ sion funds, an d resulted in a net addition to taxable income. Other tax-exempt income. Because of the heavy dependence of the governm ent on personal incom e taxes, governm ental and em ployers’ contributions to retirem ent pen ­ sion funds and social security institutions, undistributed profits of corporations, and capital incom e of insurance institutions were excluded, as w ere profits of governm ent enterprises and incom e of charitable and religious institutions (including foundations, orphanages, churches, m onasteries). W hile governm ental and em ployers’ re­ tirem ent pension funds and social security contributions, returns of funds invested by life insurance com panies, pension and social insurance funds, and incom e from governm ent-ow ned assets w ere reported, the other incom es had to be estim ated. T he estimates covering private pension funds, charities, and m ortm ain were rath er rough. T he in­ com es o f charities w ere determ ined from scanty data on their property ownership given in poor relief statistics, death duty figures, and various oth er sources. “Care has been taken n o t to overestim ate this item, as p a rt of the incom e earned by some of

328

THE INCOM E OF NATIONS

these institutions will be redistributed and appear later in the personal income statis­ tics.” 11 Similarly, in estim ating govern­ m ental and em ployers’ contributions to pen­ sion and social security funds, a deduction was m ade fo r adm inistrative expenses. T he estimates of undistributed corporate profits were m ade separately fo r corpora­ tions operating in the N etherlands and fo r those registered in th a t country b u t operat­ ing abroad. F o r estimates o f incom e fro m domestic corporations, the only data available were the tax rate, the statistics on the proceeds of the taxes on dividends and “tantiem es,” and th e profit and loss accounts of three hundred large corporations. F rom these tax data the total am ount of dividends was de­ term ined, w hile from the com pany accounts sample figures w ere ascertained of dividends and corporate savings and their relation to each other. T hen the proportion of divi­ dends to savings in this sam ple was applied to total dividends.12 W ith regard to the corporations operating abroad, account could be taken only of those active in the D utch E ast Indies. D ata on their undistributed profits since 1925 were taken from a sample of corporation balance sheets and fo r the earlier years from D e K roniek and Van Oss E ffektenboek.

c. Tax Evasion In accordance w ith previous estimates by Professor Bonger, tax evasion was fixed at 10 per cent of the assessed income. H ow ­ ever, fo r the years 1915-17 a higher p er­ centage was used. It was also assum ed that tax evasion was m uch higher in the years 1900-13.

d. Correction for Time Lag in Dividend Payments Since the concept of national incom e adopted referred to the years in w hich the income was produced, w hile dividends were shown for the years in w hich they were paid out, all dividends were shifted back a year. This correction was applied only to divi­ dends of domestic corporations, estim ated on the basis of the statistics o f dividend taxes m entioned above.

e. Transfer Payments U nem ploym ent benefits, dismissal com ­ pensation to governm ent employees, social security benefits, and pensions and annuities included in taxable and nontaxed incom e w ere deducted.

4. Subsidiary N et O utput Estim ates S eparate estim ates fo r a n um ber of in­ dustrial divisions w ere m ade by the n et o ut­ p u t m ethod. H ow ever, only fo r agriculture, fishing, electric light, pow er and gas, and certain m anufacturing subdivisions was the n et value o f the ou tp u t directly determ ined fro m gross value. F o r the o th er m anufac­ turing subdivisions, an indirect m ethod, de­ scribed below, was used. T he incom e of all o th er industrial divisions taken in the ag­ gregate w as p u t down as the total of their incom e paym ents. T h e breakdow n by industrial divisions was given as follows: A griculture and horticulture Fishing M ining Electric light and pow er, an d gas Construction M anufacturing T rade, hotels, and restaurants Banking Insurance Services of dwellings and com m ercial buildings Shipping O ther transportation and com m unications G overnm ental services Professions and personal services Ecclesiastical services F orestry was om itted as playing an insig­ nificant role in the economy.

a. Agriculture and Fishing A lthough estim ates of agricultural incom e (p lan t an d anim al p roduction) w ere being p repared by various agencies o f the D e­ p artm en t o f A griculture, they had to be adjusted by the C entral Statistical Office for certain cost item s n o t taken into account. These cost item s w ere derived from an official rep o rt fo r the period 1 9 2 3 /2 4 -1 9 3 0 / 31, and extrapolated fo r later years on the basis of official data fo r sam ple farm s. These item s included depreciation, m aintenance, certain indirect taxes for the m aintenance of w aterw ays an d the “polders,” and other overhead charges. A djustm ents w ere also m ade fo r the annual increase in the num ber o f livestock. F o r lack o f any direct inform ation con­ cerning the contribution o f horticulture to n ational incom e, th e estim ates w ere m ade in a rath e r roundabout w ay by sum m ing up th e value o f vegetables an d fru it sold in

23.

THE NETHERLANDS

dom estic m arkets and exported as reported by D epartm ent of A griculture. T hese rather rough figures were com pared w ith the co r­ responding expenditures on vegetables and fru it in fam ily budgets fo r A m sterdam in 1923-24 and 1934-35 and fo r various other tow ns in 1935-36. T he agreem ent w ith re ­ gard to the rate of increase in consum ption was fair. As these estim ates referred to final sales values, costs of production w ere de­ ducted. These costs w ere determ ined on the basis of the investigation on th e costs o f production in horticulture based on data obtained from samples of farm s. A djustm ent fo r changes in these costs in other years were m ade in the light of changes in the wage bills. W hile the value of the catch in fishing, including that directly sold in E ngland, was reported in official statistics, data on the expenses of various types of fishing vessels were available only fo r one year. T heir ratio to wages was applied to the know n wage bill fo r all years.

b. Mining, Public Utilities, and Construction Official statistics reported the net value of the output in the governm ent ow ned coal m ines as well as the value added p er w orker. This value was applied to the private coal m ines, for w hich only the total num ber of em ployees and the total volum e of p roduc­ tion w ere know n. Com plete statistics w ere available fo r electric light and pow er plants, w hile the estimates relating to gas w orks w ere based on data for the largest factories, covering 40 per cent of total production. Since n o direct inform ation on the value contributed by construction to national in­ com e was available, various d ata and m eth­ ods w ere com bined fo r m easuring this value. F o r this purpose distinction was m ade be­ tween residential buildings, other buildings, public works, and m aintenance and repair work. T he estimates fo r residential buildings w ere based on the annual statistics of w ork­ ers em ployed in construction and th e data on the average net construction costs (value added) per w orker and per dwelling unit.13 F o r the com putation of the average net con­ struction costs per dwelling unit, data on co-operative and w orker housing covering the years 1927-30 were extrapolated by the index of construction costs. It was assumed, furtherm ore, that the cost of dwelling units

329

other th an fo r w orkers was one and one h alf tim es th at of w orkers’ dwellings. P ri­ vate statistics o f construction contracts, pub­ lished in the weekly, D e A annem er (The C o n tra cto r), w ere the m ain source for the estim ates of the value of new nonresidential buildings. Since these statistics were incom ­ plete because they excluded construction w ork done directly by builders and by pub­ lic bodies, it was assum ed th at the n et value of new factories as reported in D e A a n ­ nem er constituted 10 per cent of the total. D ata on expenditures fo r public works (roads, b ridges), including maintenance, were derived from financial reports of the national governm ent and local authorities. Finally expenditures fo r m aintenance and repair w ere added in accordance w ith the estim ates m ade in a special study (see N ote 6 ).

c. Manufacturing T he m ain sources fo r measuring the net value of m anufacturing production were the yearly production statistics th at showed output and sales, purchases of raw materials, services, and oth er items, from w hich the gross value added fo r the large plants in selected industries could be derived. Statis­ tics of the labor force in each industry were published by the W orkm en’s Com pensation Board, an d in some special studies and re­ ports. F o r the om itted enterprises o r in ­ dustrial divisions (rayon, electric bulbs, to ­ bacco, and dairy p ro d u cts), representative business account data w ere taken. N ext, all data w ere grouped separately for the con­ sum ption goods and production goods indus­ tries. T hen, fo r these two industrial divi­ sions, th e gross value added p er w orker was com puted and applied to the respective total n um ber of workers. T hus, th e com puted average “gross value added per w orker” had to be applied to approxim ately 10-15 per cent of all w orkers employed in indus­ tries fo r w hich direct production data were n o t available. F o r the m ore recent years the basis of the estim ate could be extended. In the consum ption goods industries the gross value added according to th e annual census of industries was reported for plants covering 29 p er cent of all w orkers in 1921, and for 38 per cent in 1938.14 F o r the production goods industries, this percentage was about 75 per cent throughout the years. T he deduction fo r m aintenance and de­ preciation (included in the gross value

330

THE INCOM E OF NATIONS

added) was com puted from sample data on business accounts on the basis of (a) the proportion th at depreciation allowances bore to profits and (b ) the differential between gross value added and wages paid out. T he results so obtained w ere checked against the findings of the special investigations on the structure of costs.15 A dditional data on depreciation, collected in the com plete in­ dustrial census held during the F irst W orld W ar, were also used.

insurance com panies to national income was considered to be the sum o f salary and wage paym ents, and distributed and undistributed profits. These data w ere de­ rived from annual reports of insurance cor­ porations and social insurance institutions. T he annual reports o f the official Insurance C ham ber provided data fo r all life insurance com panies.

d. Trade

A lthough this division also included rent paid on com m ercial prem ises (shops, board­ ing houses, e tc .), it was identified under the heading of services of dwellings. R ent from dwellings, including that on ow ner-occupied dwellings, is assessed fo r personal tax on rents and reported in the tax statistics. These data, how ever, had to be adjusted fo r tax exem ptions, w hich varied w ith the size of the tow n, and fo r understatem ent allowed by law relative to ren t o f furnished apart­ m ents. T he am ount of the tax exemptions and data on the percentage of tax-exem pt ren t in the total ren t fo r A m sterdam and T he H ague provided the required correc­ tion. F o r m aintenance and repair, a deduc­ tion of 10 per cent was m ade based on the judgm ent o f researchers in this field (see N ote 6 ).

This division included wholesale and re­ tail trade, hotels, restaurants, and boarding houses. Estim ates of incom e originating in trade were based on annual statistics on sales and operating costs of a sample of wholesale and retail traders in im portant branches, and the statistics o f wages insured for w orkm en’s com pensation. F rom these sources and the occupational censuses, the average annual earnings per w orker w ere com puted and the same averages were ap­ plied to the num ber of entrepreneurs.16

e. Banking and Insurance T he estim ate was rath er com plete as it included the whole banking system except for small loan banks and associations. T he net value of the services rendered in bank­ ing was considered equal to the aggregate of salaries, wages, dividends paid out, and undistributed profits. D ata on labor incomes were derived from the official statistics of the W orkm en’s C om pensation B oard and from the yearly accounts o f the large banks. The latter source, as well as official statis­ tics, also provided inform ation on profits fo r 1930 and later years, w hile fo r the earlier years, an extrapolation was m ade ac­ cording to the sample data. T he estim ators w ere aw are th a t the m eas­ urem ent of the net output o f banks involved serious duplications. In fact, it was assumed that com m ercial banks rendered services free of charge to the banks’ debtors (i.e., other industries), and that, therefore, in­ terest paid by borrow ers did not form p art of the net output of these industries, but contributed paym ents for banking services. T herefore, it was thought that the duplica­ tion m ight be offset approxim ately by as­ suming th a t all interest received by banks on short-term investm ent at hom e or abroad represented paym ents for bank services and not a factor incom e paym ent.17 As in banking, the contribution of the

f. Services of Dwellings

g. Shipping F ro m the annual reports of the large ship­ ping com panies, data on dividends, undis­ tributed profits, and interest w ere compiled, w hile figures on wages and salaries were taken from the statistics on accident in­ surance fo r shipping. F o r the inland navi­ gation fleet (approxim ately 20,000 ships, total tonnage 4 million tons) very fragm en­ tary data based on reports for some large enterprises had to be used.

h. Other Transportation and Communications W ages, salaries, interest paid out, and n et profits of the railroads and trolley car systems in the three large cities (A m ster­ dam , R otterdam , and T h e H ague) and of the public com m unication services (post offices, telegraph and telephone services) w ere reported in yearly financial statem ents. In o rder to estim ate th e incom es derived fro m o th er street-car services, the ratio of value added to gross receipts in the three covered cities was applied to the gross re­ ceipts o f other street-car enterprises in the

23.

THE NETHERLANDS

country as published by the C entral Statis­ tical Office. F o r other transportation units, such as taxi, auto repair shops, delivery services, and air transportation, the wage bill reported by the W orkm en’s C om pensa­ tion B oard was assum ed to account for the same share in the net output as fo r the railways.

i. Government Services Available statistics and budgetary docu­ m ents provided m ost of the d ata on in­ com e from public services, including edu­ cation b ut excluding the industrial activities of the public bodies. F o r towns, other than the fo u r largest cities, an estim ate was m ade using the statistics o f pension funds, based on the assum ption th a t 50 p er cent of the wages and salaries covered w ere paid to industrial workers. T o the wage and salary bill the interest on public loans was added.

j. Other Services T he num ber of persons gainfully occupied in professions, personal and dom estic serv­ ices was m ultiplied by an average salary th a t was based on fragm entary inform ation. Publications o f religious bodies provided the data on salaries paid out fo r ecclesiastical services. R ough estim ates w ere required in order to m ake adjustm ents fo r understate­ m ent and to include small congregations. Interest on debts of churches was added.18

k. Net Income from Abroad These incom es w ere reported in the offi­ cial statistics of balance of paym ents which, however, were incom plete fo r the years prior to 1930. T he figures fo r 1936 in T able 23-2 show the com ponents o f this item : Table 23-2. The Balance of Payments Items in the National Income of the Netherlands, 1936 (million guilders) Credit 195 Interest and dividends Profits of colonial enter­ 54 prises Expenses in the Nether­ lands of colonial enter­ 22 prises Pensions and salaries of 19 colonial governments 4 Remittances Total s o u r c e

:

Debit Balance 120 75 54 22 1 21

18 -1 7

-------

—

-------

294

97

197

Reference No. 16.

331

5. B reakdown of N ational Income

by T ypes of Incom e Payments; G ross Capital Formation A s a rule, inform ation on wages and salaries, m ainly taken from the statistics of th e W orkm en’s C om pensation Board, was already em bodied in the incom e estimates fo r the industrial divisions and merely had to be brought together. T he same was true o f data on pensions th a t w ere fully available in official reports. U ndistributed profits and rent fro m dwellings h ad similarly been es­ tim ated in the com putation of the national incom e total. A dditional estimates had to be m ade only fo r agricultural ren t and dividends and interest paid out. By adding up the items m entioned and deducting them from national incom e, entrepreneurial in­ com e was com puted. D ata on ren t paid by farm ers w ere de­ rived from an official report, while divi­ dends and interest were roughly estim ated on the basis of the dividend tax statistics of the balance of paym ents, of bonds regis­ tered w ith the A m sterdam Stock Exchange, an d the reports of financial institutions. M ortgage interest was included w ith rent and not w ith interest. These rath er tentative estimates were m ade for the years 1913, 1929, 1934, and 1938, except fo r the wage and salary bill fo r w hich th e survey started with the year 1903.19 A rough estim ate of gross capital fo rm a ­ tion was m ade by com bining th e values of gross ou tp u t in the construction and m etal industries, adjusted fo r exports and im ports.

6. D evelopm ents Since 1940 D uring W orld W ar II, despite the G er­ m an occupation, w ork on national income estimates in the N etherlands continued u n ­ abated, and in the decade im m ediately fol­ lowing the w ar progressed rapidly. T he es­ tim ates kept pace w ith the newest develop­ m ents in the statistically m ost advanced countries and, in some respects, even an­ ticipated them . D r. D erksen, who had done m uch of the pioneering w ork on the D utch estimates before the w ar, sum m arized these developm ents in a statem ent reproduced on th e next page. T he official national incom e studies, pub­ lished in 1939 and 1940, proved very valu­ able for the analysis of current economic problem s, such as the treatm ent of war

332

THE INCO M E OF NATIONS

finance and expenditure, and they also p ro ­ vided quantitative inform ation on im portant structural questions. Using the results o f the national incom e calculations as a basis, it becam e possible to provide answers to such questions as the im portance to the N ether­ lands econom y of ocean-going shipping or to analyze the grow th of the m anufacturing industry since the F irst W orld W ar. A study on the contribution of the N etherlands E ast Indies to the national incom e of the N ether­ lands, prepared in 1941 and published in 1945, drew world-wide attention.20 To m eet the rapidly increasing dem and fo r estimates of national incom e, related aggregates, and com ponents, it becam e necessary to w iden the statistical basis of the estimates and to im prove the conceptual fram ew ork. Says D r. D erksen: In 1941, the Netherlands Central Bureau of Statistics made an attempt to construct a sys­ tem of national accounts for the year 1938. This study was stimulated by two articles by Mr. Ed. van Cleeff in the Dutch monthly De Economist.21 Mr. Van Cleeff, who also initiated the term “national book-keeping,” used ficti­ tious figures and his system of tables was differ­ ent from that adopted later by the Central Bureau of Statistics. Owing to the war, the results could not be published, but they were circulated in mimeographed form and were later incorporated in the first postwar publica­ tions on the national accounts of the Nether­ lands.22 After the war, when contacts with the Anglo-American countries could be re-estab­ lished, it was found that similar ideas had been developed in the United Kingdom and other countries. However, whereas in the N eth­ erlands the work was concerned mainly with the statistical problems of estimating the vari­ ous flows in the national accounts, the work in Great Britain was concerned mainly with the development of concepts and definitions. In the present D utch system the follow­ ing sectors and accounts are distinguished (see T able 23-3) : business enterprises, gov­ ernm ent, households, collective incom e re ­ cipients (i.e., life insurance com panies, pen­ sion funds, and social insurance fu n d s), foreign economies, and the com bined capi­ tal account. O ne characteristic o f the pres­ ent D utch national accounting system is that “collective incom e recipients” are treated as a separate sector, the reason being their growing im portance as providers of funds on the capital m arket. (In the international standards now generally used, the group is consolidated w ith the “households sec­ tor,” w ith the exception of governm ent

pension and social insurance funds, w hich are consolidated w ith “general govern­ m ent.” ) T he accounts fo r “business enter­ prises” in th e D u tch system are derived from accounts set up for the principal divi­ sions o f econom ic activity: agriculture, m anufacturing industry and mining, trans­ p o rt and com m unications, etc. Thus, at an early stage, it was necessary to construct input-output tables, and these tables, even if n o t always published, have been a per­ m anent p a rt of the w orksheets. D r. D erksen fu rth er notes: Another characteristic of the Dutch national accounting system is that balance sheets for the various sectors and for the economy as a whole were introduced at an early stage, so as to provide a method for checking the esti­ mates of savings and changes in physical assets by sectors. During and immediately after the war, the introduction of rationing, exchange controls, and various other measures resulted in the collection of a great deal of information by various government agencies which proved useful for national accounting purposes. As a result, it has been possible to improve the na­ tional income estimates for 1938 and previous years. The revised official estimates for 1921— 1939 were published in 1948.23 This study also presents the breakdown by distributive shares for the entire period 1921-1939, as well as index numbers of real national income at mar­ ket prices and at factor cost, and of real per capita income for the period 1900-1939. The indexes of income in constant prices have been obtained by deflating current price estimates by the cost-of-living index. Various efforts have been made in the post­ war period to expand the system by adding detailed estimates of private consumption ex­ penditure in current and in constant prices. Work has also continued on detailed estimates of public and private fixed capital formation and of changes in stocks; the general govern­ ment accounts have been re-arranged to agree with the economic category classification re­ quired for the national accounts. The accounts of local authorities (eleven provinces and ap­ proximately 1,000 municipalities) have been compiled on the basis of a uniform system of reporting (since 1923), which has now been revised in order to provide the data required for the economic character classification in the national accounts. The methods used for ex­ pressing estimates of gross national product and gross national expenditure have been im­ proved. In general, the basis for the postwar estimates of gross domestic product by indus­ trial origin are more accurate and more com­ plete than before the war because the coverage of the basic industrial statistics has been ex­ panded. Moreover, the work on national ac­ counting has favorably affected the develop-

23.

333

THE NETHERLANDS

(billion guilders) 1938

1946

1950

1954

5.5 - .5 (a) - .5

10.7 - 1 .1 .5 - 1 .0

19.1 - 2 .4 .3 - 1 .7

25.7 - 3 .1 .2 - 2 .2

4.5 .4

9.1 .2

15.4 .3

20.6 .5

4.9

9.3

15.7

21.1

2.5

5.5

8.4

11.5

2.1 .2 (a) .2 - .1

3.8 (a) .1 .2 - .3

5.9 .7 .8 .1 - .2

8.7 (b) .9 .2 - .2

National income III. Industrial origin of net domestic product Agriculture, forestry, fishing Mining Manufacturing Construction Transportation, communication, utilities Wholesale and retail trade Ownership of dwellings f Public administration and defense Other services

4.9

9.3

15.7

21.1

.3 .1 1.0 .2 .5 .5 .6 .3 1.0

1.2 .2 2.2 .5 .9 1.1 .6 1.1 1.4

Net domestic product IV. Expenditure on gross domestic product Private consumption expenditure General government consumption expenditure and gross fixed capital formation Gross fixed capital formation of public and private enterprises Increase in stocks Exports of goods and services Less imports of goods and services

4.5

9.1

15.4

20.6

4.4

8.1

12.7

15.7

.7

2.2

2.8

4.3

.5 + .1 1.5 - 1 .7

1.4 + .4 1.4 - 2 .8

3.6 + 1.6 7.7 - 9 .1

4.9 1.0 12.6 - 1 2 .7

5.5

10.7

19.1

25.7

I. Relationship among main aggregates Gross domestic product a t market prices Less indirect taxes Plus subsidies Less depreciation Net domestic product a t factor cost Plus net factor income payments from abroad •\T_j- -J.... f i ■ ..pi. . i* _ ..I I'ici nauuiiai , uuuul a 4-t lcxv-- lui tuoi ui iiauuuai income II. Distribution of the national income * Compensation of employees Income from unincorporated enterprises, rent and interest, dividends Savings of corporations Direct taxes on corporations General government income Less interest on public and consumers’ debt

Total gross domestic expenditure

2.2 .3 4.9 .9 1.21 2 -1 J .6 1.1 2.1

2.6 .5 7.5 1.2 4.0 1.0 1.4 2.5

ail may not add to totals because of rounding. (a) Less than 50 million guilders. (b) For 1952 and 1954 saving of corporations is included in income from unincorporated enterprises. * Components do not add to total in Part IV since they are not adjusted for bank interest. f Includes cash and imputed land rent and cash rent of nonresidential buildings. s o u r c e : United Nations Statistical Office, Statistics of National Income and Expenditure, Series H, No. 8, September, 1955, pp. 49, 87, 106, 114; Population: 1919-38 from Centraal Bureau voor de Sta­ tistiek, Jaarcijfers voor Nederlands, 1947-1950, Utrecht, 1951, p. 3; for 1952-54, United Nations, Statis­ tical Yearbook, 1954, New York, 1954, p. 30.

334

THE INCO M E OF NATIONS

ment of the basic economic and financial sta­ tistics because it has resulted in a better co­ ordination of the concepts, definitions and classifications used in various fields. Work has started on the preparation of quarterly national accounts and a quarterly index of national in­ come in constant prices. T he income tax statistics no longer p ro ­ vide a basis for the national incom e esti­ mates as they did in the period before 1940. D uring the w ar the incom e tax system in the N etherlands was drastically changed, and the pay w ithholding m ethod of taxation

was introduced w ith respect to wages and salaries, dividends, etc. As a result, the p re­ w ar system o f com piling incom e tax statis­ tics, based on decentralized reporting by the heads of district offices of internal revenue, h ad to be abandoned. N ew m ethods fo r com piling incom e tax statistics, w hich are in p a rt based on m odern sam pling m ethods, have been introduced. “It is too early yet to state w hether these m ethods will yield statistics w hich can be used fo r the p rep ara­ tion o f independent estimates of national incom e.”

24. SWEDEN

References 1. Erik Lindahl, Einar Dahlgren, and Karin Kock, National Income in Sweden, 1860-1930, in two parts, University of Stockholm, Institute for Social Sciences, London, 1937. 2. a. E. Dahlgren, Produktionsstatistisk uppskattning etc. (Estimate of the Swedish National Income 1930-1934 by the Production Statistical M ethod), Stockholm, 1936. b. ------ , P. M . med vessa berakningar etc. (Certain Calculations of the Size of the National Income), Bihang till riksdagens protokoll, Stockholm, 1941. 3. Konjunkturinstitut, National Bokforing, 1946-1951 (National Book-keeping), Meddelanden, Stockholm, 1951 (with English summary). 4. Organisation for Economic Co-operation, National Accounts Studies, Sweden, Paris, 1953. 5. Ingvar Ohlsson, On National Accounting, Konjunkturinstitut, Stockholm, 1953. 6. Erik Lundberg, “Recent Experiences in the Use of Social Accounting in Sweden,” Paper for the Meeting of the International Association for Research in Income and Wealth in Cam­ bridge, Aug. 27-Sept. 3, 1949 (Mimeographed), Konjunkturinstitut, Stockholm.

1. History T he first systematic inquiry into the Swedish national incom e was m ade by E rik Lindahl, E inar D ahlgren, and K arin K ock for the Institute of Social Sciences of the University of Stockholm as p art of a broad survey of the Swedish national economy. T he estimates were published in 1937 and covered, year by year, the period 18601930, a longer span of years th an those of alm ost any other country.1 These estimates were continued, under the auspices of the M inistry of Finance, by M r. D ahlgren fo r 1931-34, and in a less elaborate way, by M r. Lindahl fo r 1936-39. In connection w ith these incom e studies, Notes begin on p. 538.

estim ates of consum ption outlay and capital form ation w ere also m ade. In 1946, the Institute fo r E conom ic Re­ search (K onjunkturinstitut) began the reg­ ular periodic publication of estim ates o f the gross national expenditure an d gross n a­ tional incom e and o f the social accounts of Sweden. T he first of these national incom e estim ates was m ade for th e years 1943-46 and fo r 1938/39 as a prew ar year fo r com ­ parison. All calculations w ere m ade under th e direction of Ingvar Ohlsson. In 1947, a special u n it was set up w ithin the govern­ m ent adm inistration know n as th e “N ational Budget D elegation” to prepare an annual n ational econom ic budget. These budgets were published in the spring o f 1948 and

24.

1952 in the publications of the K onjunk­ tu rinstitut.2 T his chapter will concern itself prim arily w ith the estimates prepared before W orld W ar II. A brief note at the end of the chapter indicates th e principal innovations introduced after W orld W ar II.

I. Prewar Estimates 1. C oncepts in the Prewar Estim ates A broader concept of national incom e was used in the Swedish prew ar estim ates th an in those o f m ost other countries of the tim e. In the 1860-1930 estimates, the concept included, in addition to the output of the various divisions of econom ic ac­ tivity producing com m odities and services, the unpaid dom estic w ork of housewives, as well as services of household goods and m otor cars. H ow ever, national incom e was estim ated both inclusive and exclusive of the unpaid dom estic services of housewives, and only the latter aggregate was used in the m ore com prehensive analyses. In the succeeding estim ates, how ever, the concept was m ade narrow er by excluding, first, in those of 1931-34, the services of housewives and next, in those of 1936-39, the services of household goods and m otor cars. In estim ating net value from gross value indirect taxes were included. But, on the other hand, the costs of governm ental services rendered to business w ere deducted. As net capital incom e from abroad was not taken into account, theoretically, the esti­ mates related to hom e-produced incom e only. Actually, how ever, since this item played an insignificant role in the econom y at the tim e, the result was practically iden­ tical w ith national income.

2. General Estim ating M ethods

a. P eculiar C haracteristics T he prew ar estim ates w ere consistently based on the net o utput m ethod. T he net value of output was com puted from gross value for all divisions except handicrafts, hotels and restaurants, professions, govern­ m ent, domestic service, and, fo r the years 1930-34, trade. F o r these divisions, it was com puted directly as the aggregate of dis­ trib u ted incomes.

SW EDEN

335

In com puting the n et from gross value fo r each division a unique m ethod was fol­ lowed. First, the costs of raw materials and services of other divisions used in produc­ tion were com puted in each case from the production statistics of the other divisions, and, so fa r as the division supplied its ow n raw m aterials, from its own produc­ tion statistics. They w ere n ot com puted, as is usually done, from the financial statistics of the enterprises in the given division. For the costs of raw m aterials obtained from abroad, im port statistics were consulted. Changes in inventories of raw materials from one year to another w ere disregarded. T he second peculiar feature of the Swe­ dish estimates was th at the annual deprecia­ tion of durable p roducer goods other than buildings was calculated fo r each division as an aggregate o f the am ortization install­ m ents on the capital goods added during each of the past years. T he am ortization in­ stallm ents were calculated during the year of the addition of these goods. This m ethod of calculating depreciation took m ore closely into account the variations in the age com ­ position of the stock from year to year than the traditional m ethod of calculating depre­ ciation on the basis of the total existing stock of capital goods and their estimated life. But, on the other hand, this m ethod w as also m uch m ore com plicated than the traditional one (fo r further description see Section 2c below ). Thirdly, the classification of the output by divisions of econom ic activity departed in some respects from conventional lines. T he outputs of the m ain divisions were com puted gross of the services of finance and governm ent and were reduced to a net figure only in the aggregate. T hey were also calculated gross of the services of wholesale trade in producer goods. T he estimates were m ade separately for the following divisions of economic activity: A griculture, forestry, fishing, and hunting M anufacturing, mining, and handicraft C onstruction T ransportation and com m unications T rade, finance (banking and insurance), and professions Services of consum er durable goods Dwellings O ther consum er durable goods (m otor cars, furniture, and other household goods)

336

THE IN CO M E OF NATIONS

Table 24-1. Computation of the Value of N et O utput in the Swedish Income Estimates, 1930 (million kronor) A. Determination of Internal Supply According to Its Use Use of Internal Supply

Division of Economic Activity Agriculture, forestry Mining, manufacturing Construction Transportation, communications Trade, finance, professions Government Unspecified items b Total industries

Excess of Imports Over Exports 116 84

Gross Output 1,870 5,853 967

Producer Goods Internal Supply 1,986 5,937 967

Consumer Goods and Services 741 3,092

Nondurable Goods and Durable Services < Other ° 14 1,231 709 2,136 967

1,008

-345

663

393

270

1,964 684

-5 0

1,742 396

437

1,914 684 437

172 288 437

242

12,588

6,364

12,346

723

4,534

967

11,621 B. Allocation of Nondurable Producer Goods and Services to the Divisions of Economic Activity in Which They Are Used in Production Divisions in Which Goods and Services Are Used in Production Divisions in Which AgriGoods and Services culture, Are Produced Forestry Agriculture, forestry 450 Mining, manufacturing 205 Construction Transportation, com­ munications 15 Trade, finance, pro­ fessions Government Unspecified items'3 8 Total industries

Mining, Manu­ facturing 780 1,273

678

109

Con­ struetion 1 434

Trans­ portation, Communi­ cations

Trade, Finance, Profes­ sions

Govern­ ment

72

124

28

Total 1,231 2,136

16

70

14

270

98

172“ 288° 437

46

178

153 2,315

481

266

194

140

4,534

C. Gross and Net Output by Divisions of Economic Activity

Division of Economic Activity Agriculture, forestry Mining, manufacturing Construction Transportation, communications Trade, finance, professions Government Services of consumer durable goods Residential buildings Other durable goods Paid domestic services Joint debit items National income

Gross Output 1,870 5,853 967 1,008 1,964 684

Goods and Services Used Up 678 2,315 481 266 d 194 140

Depreciation and Maintenance Stocks, Equipment 97 149 2 94 446

Buildings 43 87 68 26° 121f

781 1,020 140

699

Net Output 1,052 3,302 484 580 1,726 518 660 321 140

460 14,287

4,534

1,085

345

8,323

24.

D om estic services P aid domestic services Services of housewives G overnm ental services

b. Gross Value of Output and Deductions for Materials and Services T he com putation o f the net value of o u t­ p ut in the first four divisions began w ith the calculation of gross value. T he excess value of im ports over exports in each group of goods was added to the gross value figures. In m aking this addition care was taken to m ake the im port and export values fully com parable w ith product values.3 By adding this excess of im ports over exports, the value of the internal supply of goods in the division was obtained. N ext— and here is w here the peculiarity of the Swedish estim ating m ethod really lay— this internal supply was broken down into producer goods and services and con­ sum er goods and services, w ith a fu rther subdivision of each into nondurable goods and services and durable goods. D urable goods were deem ed to be all goods having a life in excess of a year. T he nondurable p roducer goods and services produced by each division w ere fu rth e r subclassified in accordance w ith the divisions to w hich they w ere supplied. Finally, from the gross value of output of each division the values of all nondurable p roducer goods and services w ere deducted, w hether supplied by th a t division fo r itself or secured from other divisions. A n exception was m ade, how ever, fo r the nondurable producer services of governm ent and finance. These could n o t be allocated to the various divisions, individually, and h ad to be charged as a joint debt item against the com bined outputs o f all divi­ sions. In the end, after these several deduc­ tions w ere m ade, a value fo r the o utput of the several divisions was obtained th a t was free from duplications.

SW EDEN

337

T his general estim ating process is illus­ trated in Table 24-1 w ith the aid of the fig­ ures fo r 1930 covering the com putation of the value of m anufacturing and mining o ut­ put. It will be noticed th at the gross value of th a t ou tp u t am ounted to 5,853 million kronor an d the excess of im ports over ex­ ports to 84 m illion kronor. The internal supply, therefore, aggregated 5,937 million kronor. O f this supply, 3,092 million kronor were consum er goods, 709 millions durable producer goods, and 2,136 millions non­ durable producer goods. O f the last aggre­ gate, 1,273 m illion k ronor were used up in the division itself, w hile 863 millions were supplied to agriculture, construction, trans­ portation, com m unication, trade, finance, professions, and governm ent. In addition to using up 1,273 m illion kro n o r of its own nondurable producer goods, the m anufac­ turing and m ining division consum ed 1,042 m illion k ronor of such goods and services produced by other divisions. By deducting the value of consum er nondurable producer goods an d services from the gross value of the ou tput of this division (5,853 million k ro n o r), a figure of 3,538 million kronor was obtained fo r the n et value of the output of m ining and m anufacturing.

c. Deduction of Depreciation H aving thus obtained, fo r each division, the value of its o u tput net of the value of nondurable p roducer goods and services consum ed, the estim ators next calculated the deductions fo r depreciation allowances. This was done fo r all estimates beginning w ith the year 1896. F o r p rio r years depre­ ciation allowances w ere calculated fo r all the divisions com bined and were deducted from the aggregate value of th eir output. Depreciation of buildings used in production was calculated in the conventional way as fixed percentages of the aggregate capital values. In the case of buildings used in manufacturing and

N OTES TO TA BLE 24-1: ° Producer and consumer durable goods. b Imports of certain nonhome-grown raw materials (mainly cotton in manufacturing and gold) and raw materials the value of which was directly deducted from the gross output (mainly shipping and government). With regard to shipping, depreciation charges are included with cost of raw materials. c Joint debit items. d Including depreciation allowances in shipping. e Including depreciation of stocks and equipment in government made jointly with that for trade, finance, and professions (Reference No. 1, Part II, p. 515). 1 No cost items other than maintenance and depreciation allowances have been deducted from gross rent (Reference No. 1, P art I, p. 222). s o u r c e : Compiled from Reference No. 1, Part I, pp. 61, 79, 87, 141, 151, 157, 173, 197, 199, 211, 219, 225, 229; Part II, pp. 141, 303, 305, 307, 309, 409, 422, 427, 431, 435, 441, 445, 447, 453, 455, 461, 463, 469, 471, 514, 557, 569.

338

THE INCOM E OF NATIONS

mining, the values are taken as given in fire insurance valuations, to which a computed value of noninsured plant was added in each case, and for which the rate of depreciation was fixed at 2 per cent. The maintenance costs of these buildings were estimated at 1.5 per cent of the same values. In the transportation and communications industries, depreciation and maintenance costs were taken directly from the financial reports. No matter how these items were calculated in the first instance, they were eventually translated into stated percent­ ages of the gross income or value of the out­ put in each division. Thus they are estimated for 1914-30 (in some cases 1901-30 or 191130) for agriculture and forestry at 5 per cent of the gross income, for manufacturing and mining at 1.75 per cent of the gross value of the output, for state railways (for tracks as well as buildings) at 9 per cent of gross in­ come, and for private railways (also for tracks as well as buildings) at 14 per cent.4 For durable producer goods other than buildings depreciation allowances were calcu­ lated indirectly from the gross values of the services rendered by them, instead of directly from their aggregate estimated capital values. The gross values of the services were computed for each year’s addition of such durable goods separately, and were calculated in each case as an annuity earnable over the lives of the goods at 5 per cent interest on the capital invested in them. The goods were classified into three groups, according to their length of life, of 5, 10, and 20 years. All such goods used in agricul­ ture, manufacturing, and mining, and in the building and construction industries were as­ sumed to have a life of 10 years, while those used in commerce and government were as­ sumed to have lives of 5, 10, or 20 years, depending on their character. The annuities, or gross values of the services, were calculated for each year and each age group in such a way as to take account of the entire series of the past annual additions of such goods still rendering service at the time. Adjustments in these values were made in accordance with price changes. Next, the percentage ratios of the deprecia­ tion allowances to the gross values of the serv­ ices of the durable producer goods were cal­ culated for each age group in each industrial division. These allowances were calculated as annual amortization installments making it possible, at an interest rate of 5 per cent, to retire the capital value of the goods during their estimated lives. Inasmuch as the calcu­ lations of both the values of the services and the depreciation allowances are based on the same valuations of the goods, the same esti­ mates of their lives, and the same interest rates, these percentages were constant in character. The percentages thus computed were as fol­ lows: 5

Annual Amortization as Percentage of Annual Gross Value Producer Durable Goods of the Services of Other Than Buildings the Goods Agricultural, 10-year life 78.8 Manufacturing and mining, generally 10-year life 75.7 Construction, generally 10year life 74.8 Commercial and govern­ mental, 5-year life 85.8 Commercial and govern­ mental, 10-year life 75.2 Commercial and govern­ mental, 20-year life 51.7 For railways, depreciation and maintenance costs of the durable producer goods other than buildings were computed directly from their financial records. The combined depreciation and maintenance costs for rolling stock were estimated for 1921-30 at 14 per cent of the gross income for the state railways and at 10 per cent for the private railways.

d. Net Value of Output T he figures thus com puted fo r each of the fo u r divisions m entioned approached the n et value of their ou tp u t as closely as possible. T he only items of duplicating n a­ tu re still contained in them w ere those rep­ resenting th e costs of the services o f govern­ m ent and finance rendered to these divi­ sions. It has already been said th a t it was impossible to estim ate the producer services o f governm ent and finance by recipient di­ visions. T he duplication could be elim inated only from the total value of the o u tput of all divisions taken together, by deducting the estim ated aggregate value of these serv­ ices. T he resulting residue was the net value of the entire national output— the n et n a­ tional income. T h e w ay in w hich this n et value was ob­ tained is show n in P art C o f T able 24-1. T he entire process m ay be recapitulated w ith the aid o f the unnum bered table on p. 339 (figures are fo r 1930). In th e estim ates fo r 1861—95 cruder m ethods w ere used. N o deduction fo r de­ p reciation on buildings w as m ade fo r any division owing to the lack o f reliable data on the o u tp u t of the construction industry. D epreciation and m aintenance w ere de­ ducted as a joint debit item fro m th e aggre­ gate value of the o u tput of all divisions. T he n et incom e from the construction in­ dustry was n o t included in national incom e.

2 4 . SW EDEN

M illion Kronor

Gross value of output Minus Nondurable producer goods and services used up in production: a. Charged to each division b. Joint debit item (services of government and finance) Depreciation and maintenance of stocks and equipment Depreciation and maintenance of buildings Equals Net value of output (national income)

14,287

4,074 360 1,085 345 8,323

In order to afford com parability w ith the figures for the m ore recent years, th e series were projected on the sam e basis to 1930. As a result, fo r the years 1896-1930, two series were available: one prepared in the usual way, including the net incom e from construction, and one excluding it. Both ap­ proaches are show n in T able 24-2.

3. Estim ates by Industrial D ivisions H aving described the general m ethods used in estimating the net value of output from gross value figures, it is now possible to examine in some detail th e procedures em ployed in preparing th e estimates for each division (see T able 24-4).

a. Agriculture, Forestry, Fishing, and Hunting T he value of the gross output of agricul­ ture was ascertained by m ultiplying the

339

quantities of the various products by ap­ propriate producers’ prices as follows: In computing the value of gross vegetal pro­ duction, the value of certain raw materials produced and used up in agriculture such as hay, straw, cattle feed, and pasturage was ex­ cluded from the value of gross output. Meat production consisted of “quantity of meat that corresponds to normal breeding in the year, and which in one form or another is obtained or could have been obtained from the stock without changing the number of ani­ mals in stock.” 6 This natural increase of live­ stock included domestic consumption, exports, and positive or negative increase of stocks. It was calculated as a percentage of the number of animals in each year after careful examina­ tion of existing studies on this matter. The production ratios increased from 1861 on, and for 1930 were fixed at the following percent­ ages of stock: 7 Cows: Young cattle: Lambs: Pigs: Poultry:

15 per cent of stock of cows 75 per cent of stock of bulls 70 per cent of stock of ewes 150 per cent of stock of pigs 50 per cent of stock of hens

The existing statistics permitted the com­ putation of annual stocks with interpolations necessary for only a few years. By multiplying these stocks by the production ratios the quan­ tity of meat, in number of animals, was ob­ tained. For 1861-1912 this number was mul­ tiplied by the average farm price per animal. For the estimates after 1912 this method was refined by estimating the average live weight per animal and by applying an average price per kilo live weight. M ilk production was estimated on the basis of an average quantity of milk produced per cow. For 1913-30 these averages were derived from the statistics of the milk control asso­ ciations. For 1860-1912 they were estimated

Table 24-2. Comparison of the Two Alternatives in the Swedish Income Estimates, 1930' (million kronor)

National income before deduction of joint debit items Excluding construction Construction Total Minus Joint debit items Banking, insurance, & government Raw materials & services for construction b National income

First Approach (Excluding Construction)

Second Approach (Including Construction)

8,644

8,299 483 8,782

8,644 459 441 7,744

459 8,323

“ Excluding services of housewives. b 481 million kronor as reported in Table 24-1, Part B, minus the value of imported building materials, so u rce :

Reference No. 1, P art I, pp. 235 and 237.

THE INCOM E OF NATIONS

340

indirectly on the basis of a close correlation shown in control association statistics between food units consumed per cow and the average milk yield, by ascertaining the total quantity of cattle feed and converting it into food units.8 The trend of annual figures thus obtained, ex­ pressed in percentages of the average feed con­ sumption from 1913 to 1927, was assumed to correspond to the trend of average milk pro­ duction per cow.9 The total volume of milk produced was multiplied by the average pro­ ducers’ price. The value of egg production was computed by multiplying the average production per hen by the number of hens and then by average prices. Similarly, the value of fruit production was based on average yields per fruit tree. The estimate of net output from market gardening was mainly based on estimates of acreage and average yields made by N. Sonesson of the Academy of Agriculture.10 To determine the production of fishing in 1902-30 official statis­ tics were employed, while sample data (for the county of Bohus) were used for earlier years. A s neither annual production figures nor current price data were available fo r the com putation of the value of gross output in forestry, the estim ate was m ade on the basis of tim ber consum ed by sawmills, by the pulp industry, in charcoal m anufacturing, and as firewood. T he export balance was added to this aggregate.11 These quantities were m ultiplied by appropriate prices. T he internal supply of agricultural and forest products (Table 24-1, P art A ) was individually allocated to the following uses: to agriculture (seeds and fo rag e), to m an­ ufacturing (alcohol distilleries, breweries, dairy factories, sawmills, pulp, paper, etc.), and to hum an consum ption. M oreover, p art of m eat production was allocated to pro ­ ducer durable goods (increase of livestock). These allocations w ere based on existing estimates or statistics,12 o r were estimated on assumed average consum ption figures, as in the case of potatoes and w ood fuel.

F ro m the gross value of o u tput a deduc­ tion was m ade fo r the value of included agricultural raw m aterials (corn and roots fo r seed and cattle feed, and m ilk fo r cattle feed, e tc .), raw m aterials and services re­ ceived from other industries (including costs o f transportation) or im ported, and deple­ tion of stock.

b. Manufacturing, Mining, and Handicraft T he value of the gross o u tp u t in m anu­ facturing and m ining was com puted from yearly industrial statistics and other sources as follows: The most important industries for which continuous data were not available in the in­ dustrial statistics for the period 1861-1930 were mining, sawmills and planing mills, flour mills, dairy factories, and alcohol distilleries. The value of their gross output was estimated according to the following methods. Since data on mining products before 1896 were restricted to quantity output, their value was derived by applying adjusted current export prices.13 The gross output of the sawmill industry was com­ puted for 1861-95 by adding the value of ex­ ports and the estimated domestic use of timber (planing, housebuilding, and handicraft). The gross output of flour milling for the entire pe­ riod was computed by deducting from the in­ ternal grain supply the estimated quantity of grain used as seed, feed, and industrial raw material. The remainder, representing human consumption, was multiplied by the average grain price, and to the resultant value the es­ timated cost of milling was added. The gross value of output of the dairy industries before 1890 was assumed to be equal to the value of exports.14 The statistical base of the estimate after 1860 was enlarged by more complete statistical in­ formation on production. The greatest exten­ sion of the base occurred in 1896 when flour mills, sawmills, planing mills, iron works, and other industries were included in the industrial statistics. This is indicated in Table 24-3.

Table 24-3. Composition of Gross O utput of Mining, Manufacturing, and Handicraft (percentage of total)

1861 1895 1896 1910 1930 s o u r c e

:

Data Derived from Other Data Industrial Statistics on Industries 24.0 58.7 44.0 40.5 75.0 10.9 78.7 7.4 83.9 1.5

Reference N o. 1, Part I, pp. 96-97.

Handicraft 17.3 15.5 14.1 13.9 14.6

Total 100 100 100 100 100

2 4 . SW EDEN

Com plicated com putations w ere required to allocate the dom estic supply of m ining an d m anufacturing products to their uses.15 First, the internal supply (dom estic produc­ tion plus net exports) was classified into nine subdivisions.16 W ithin each o f these nine divisions allocation was then m ade either fo r groups of products o r by indi­ vidual products. In m any cases the nature o f the product indicated its use. H ow ­ ever, m ost frequently the value of a specific p ro d u ct had to be split up on the basis of sketchy inform ation in order to obtain its allocation to different uses. F o r instance, tinw are was allocated to its uses in the fol­ lowing w ay: durable goods: agriculture 10 p er cent, m anufacturing 10 per cent, trade 10 per cent, consum ption 20 per cent; n on­ durable goods: m anufacturing 40 p er cent, construction 10 per cent.17 F ro m the gross value of output, a deduc­ tion was m ade for the cost of raw m aterials, supplies, services, and depreciation allow­ ances, estim ated in accordance w ith the m ethods discussed and in Section 2c below. T h e net o utput in handicraft was ascer­ tained by m ultiplying the num ber of crafts­ m en by an average annual incom e. The average income was assum ed to be the same as th at of industrial w orkers in correspond­ ing trades. I t was assum ed th a t the value o f gross output was twice these incomes.

c. Construction A s no statistics on the value of gross out­ p ut in construction were available directly, special com putations had to be m ade from d ata on the value of m aterials and services produced fo r the construction industry by other industries. Based on experiences of the cost structure in residential building and public works, this value was doubled to in­ clude wages and salaries paid in construc­ tion and an estim ated am ount was added for the annual yield of equipm ent capital invested in construction. This total aggre­ gate corresponded to the gross value of con­ struction. N et value added in construction consisted of wages and interest on equip­ m ent.18 However, correlation of the results w ith the allowances fo r depreciation on buildings revealed the inaccuracy of the data prior to 1896. Consequently, only for 1896-1930 were the figures included in the national income. T he restriction necessitated preparing an alternate national income

341

series, as has already been m entioned (see Section 3 b ).

d. Trade, Finance, and Professions T he gross value of the output o f each of these divisions, including also hotels and restaurants, was estim ated separately. N et value, how ever, could n o t be com puted separately, inasm uch as the costs of m ate­ rials could be treated only as a lum p sum chargeable against all the divisions taken together. T he gross value of services o f trade in consum er goods 19 was com puted by esti­ m ating first the difference between the value of production at factory prices and the value at retail prices (distribution m arg in ). T rans­ portation charges w ere subtracted from this difference and sales taxes added.20 T he distri­ bution m argin was estim ated as a percentage addition to the production value, differen­ tiated fo r agricultural products and fo r the nine divisions of m anufacturing and mining (see note 16, p. 5 3 9 ). T he m ain sources fo r this com putation w ere the price statistics and data collected by the Bureau o f Busi­ ness Research at the Stockholm University of C om m erce and various com m ittee re­ ports. T he distribution m argin ranged from 15 per cent (new spapers) to 100 per cent (firewood, b o o k s). D ata on deductible trans­ p o rt charges were derived from the alloca­ tion o f the gross o u tput of transportation (see Section 3e below ). T he estimates o f gross value o f services of trade m ade by D ahlgren for 1930-34 de­ p arted from this m ethod insofar as they w ere m ade for 1930 on the basis of income tax assessment (raised by 50 per ce n t), and for the subsequent years by m eans of index num bers of the value of im portant consum p­ tion goods trad ed .21 This change in m ethod is chiefly responsible fo r the reduction of national incom e, com pared with the earlier com putation. H ow ever, in the estimates for 1936-39 the n et ou tp u t m ethod o f estim at­ ing was reverted to. T he gross income earned in trade was com puted as a per­ centage addition to the production value of goods entering domestic trade, distinguish­ ing fo u r groups of goods (agricultural prod­ ucts, m anufactured foodstuffs, products of the oth er large-scale industries, and those of sm all industry and h andicraft) and assum­ ing over-all additions in each case. A deduc­ tion was m ade fo r general costs, and indi­ rect taxes as well as n et output of export

342

THE INCOM E OF NATIONS

trade were added. Incom es from the other activities were estim ated as follows: Income from hotels and restaurants was computed according to the income-distributed method. The 1870-1930 decennial censuses provided data on the number of entrepreneurs and employees, while special inquiries in 1914— 15 and 1929-30 furnished information on wages and salaries. Entrepreneurial earnings were re­ ported by the 1920 census. The number of in­ come recipients was extrapolated along a straight line, average wages and salaries were extrapolated by the index of agricultural wages, and entrepreneurial earnings were extrapolated on the basis of the cost of living index. The cost items added to obtain the value of gross output were rent, fuel, and equipment. The value of foods and beverages purchased was not added since it was allocated directly to consumption. The value of gross output in banking and insurance was said to equal salaries, fees, com­ missions, pensions paid out, other administra­ tion costs, rent, and depreciation on equipment. It was computed from official statistics and private reports. In some cases special extrapo­ lations were necessary for the earliest years. The income of persons gainfully occupied in professions was also estimated by the income-distributed method. Data on their num­ ber were derived from the decennial censuses and were obtained for intercensus years by linear interpolation. Per capita income figures for several professions were provided by the 1920 census; for the other years they were estimated on the basis of an appropriate in­ come index. In order to obtain the value of gross output, an addition of 25 per cent was made. T he larger portion of the value of gross output of these divisions (trade, hotels and restaurants, professions, a portion of the gross output of life, old age, accident and health insurance) was allocated to consum er goods while the rem ainder of the gross out­ put in insurance and the entire gross output in banking was debited against all industries as a joint debit item. F rom the com bined gross values of the services of these divisions deductions w ere m ade for raw m aterials from industry (pro­ duction less exports), transport and com ­ m unication services, and im ports o f raw m aterials.

e. Transportation and Communications This division includes foreign shipping; coastal and inland w aterw ay shipping; rail­ way, tram w ay, taxi, and bus traffic; postal, telegraph, and telephone services. In con­ trast to the estim ate fo r the other divisions,

deductible costs were estim ated directly fro m reports of the enterprises. Official sta­ tistics of railways, post office, telegraph, and telephone services provided the required data except fo r the earliest years fo r which extrapolations were m ade: Special estimates were made for gross and net income of shipping, taxi, and bus traffic. The gross receipts of foreign shipping before 1873 were computed on the basis of average receipts per registered ton of ingoing and out­ going vessels, derived from Norwegian sta­ tistics. The cost items for the years before 1913 were estimated according to a percentage of the gross output shown in special statistics for 1913. For the years after 1913 the various components of the net output (pay of sea­ men, board and lodging on board ship, inter­ est, and profits) were computed from various sources of information. In order to determine the gross receipts of taxi and bus traffic, the number of vehicles was first estimated and then the average cost items (fuel, other supplies, maintenance, and depreciation), interest, prof­ its, and wages per automobile were computed from various sources (including direct infor­ mation from persons engaged in this industry). The value of gross output in transportation and communications was allocated to the fol­ lowing uses: the total for taxis, buses, and tramways to consumption, the total for coastal shipping, railways, postal, telegraph, and tele­ phone apportioned between consumption and industrial divisions according to special infor­ mation on the distribution of freight, stamps sold, or other receipts. This information, how­ ever, was for selected years only, and its ap­ plication to other years involved some arbi­ trary assumptions.

f. Services of Consumer Durable Goods T h e gross value o f the services o f dwell­ ings was based on the actual o r im puted rentals of dwellings. T h e rentals w ere com ­ puted separately fo r different localities 22 by m ultiplying the n um ber of existing room s b y an average ren t p er room . T h e num ber o f room s was estim ated according to an as­ sum ed density of occupation per 100 room s. D ata on average re n t were available for la ter years (fo r Stockholm since 1894, for th e w hole country since 1913), and these figures w ere extrapolated fo r the earlier years by th e cost of building index. F rom this gross value o f the services o r rentals, depreciation and m aintenance costs w ere de­ ducted. These were roughly estim ated to be 20 p er cent of the rentals (w ith the m ain­ tenance costs assum ed to be about double th e depreciation ch a rg es).23 In this w ay the

24.

net value of the services of dwellings was obtained. T h e net value o f the services of other durable consum er goods, such as furniture, household utensils, and autom obiles, in­ cluded in all the estim ates p rio r to 1935, was calculated in a w ay sim ilar to that em ­ ployed in estim ating the value of the serv­ ices o f durable producer goods. F irst the gross value of these services was calculated fo r each year on the basis of th e value of th e annual additions o f such goods. This gross value was calculated fo r each year as the sum of the annuities earnable during the life of the goods by the capital invested in past additions of them still rendering service at the tim e: The goods were classified into three age groups by the class of industry producing them. All the goods produced in a given industry were assumed to have the same average life. The classification used was as follows: 24 Types of Goods and Class of Industry Household china and glass (Class 2) Bedding and carpets (Class 6) Bicycles, motor cars, and metal goods (Class 1) Leather goods and furs (Class 7) Linoleum (Class 8) Furniture (Class 3) Books (Class 4)

Estimated Life (years) 5 5 10 10 10 20 20

Next, depreciation on the goods was calcu­ lated as a fixed percentage of the gross value of their services. For goods estimated to have a 5-year life, the ratio was fixed at 86.3 per cent; for those deemed to have a 10-year life at 74.8 per cent; and for those estimated to have a 20-year life at 59.4 per cent.25 By de­ ducting depreciation from the gross values of the services the net value of the services of the goods is obtained. For example, for 1930 the gross value of the services of durable consumer goods of 5-year duration (other than build­ ings) was estimated at current prices at 67.6 billion kronor, and the amortization at 58.3 billion kronor, yielding 9.3 billion kronor for net value of the services of these goods. The value of the stock of these durable goods at the beginning of the year was estimated at 185.2 billion kronor and the value added to this stock during the year was 69.8 billion kronor.26 The methods just described for estimating the net value of the services of durable con­ sumer goods other than dwellings were used for the last time in the national income esti­ mates for 1930. The results thus obtained

SWEDEN

343

were reproduced without change in the na­ tional income estimates of 1931-34.

g. Governmental Services T his division covers the services of cen­ tral and local adm inistration, including edu­ cational, ecclesiastical, and health services, inasm uch as th e activities o f governm ent in agriculture, m anufacturing, construction, and trade have already been dealt with in the com putation of the gross output of these divisions. T he n et ou tp u t of services of public adm inistration was said to be equal to the salaries, wages, and pensions paid out. By adding to this am ount the value of supplies an d services purchased from other industries and the m aintenance and depre­ ciation charges fo r public plants and build­ ings, the gross value of services in this divi­ sion was obtained. T his value was allocated to public and private consum ption and p ro ­ duction as a joint debit item . H alf of the value of the gross o u tput of the central gov­ ernm ent was assum ed to equal services ren ­ dered to consum ers. All expenses of local governm ent fo r ecclesiastical purposes, poor relief, child w elfare, public health, and old age pension insurance w ere attributed to consum ption, while the rem aining portion of th e gross o u tput of local services was divided equally into services to consumption and services to industries.27

h. Domestic Services T he net ou tp u t of the services of both paid dom estic servants and unpaid domestic w orkers (housewives, etc.) was derived by m ultiplying their num ber by an average wage paid dom estic servants. T he num ber was ascertained from census data, adjusted for agricultural w ork done by farm servants. These figures were derived fo r intercensus years by linear interpolation. T he results of a special inquiry on wages in cash and in kind fo r 1928 were extrapolated fo r other years on the basis of a wage index (on fe­ m ale farm servants) derived from the A gri­ cultural Statistics.

4. Consum ption and Capital Formation In the 1861-1930 series, the values of nondurable consum er goods produced in the various divisions w ere added together to show consum ption outlays. Similarly, the values of durable producer and durable con­ sum er goods were com bined to obtain gross

344

THE INCOM E OF NATIONS

Table 24-4. National Income of Sweden, 1861-1930 8 and 1936-1939 b Population, millions National income in current prices, including unpaid services of housewives, billion kronor Cost-of-living index, 1929 = 100 National income a t 1929 prices, billion kronor National income a t current prices, million kronor Agriculture, etc. Manufacturing, mining, and crafts Construction Transportation and communication Trade, finance, and professions Services of dwellings Services of other construc­ tion durables0 Domestic services, paid Services of housewives Government services Total Less services rendered by finance and government to production Less domestic raw materials and services used in con­ struction Home-produced national income

1861 3.9

1880 4.6

1900 5.1

1913 5.6

1920 5.9

1930 6.1

.8 51

1.4 53

2.2 52

3.9 60

12.8 160

9.3 97

1.5

2.6

4.3

6.6

8.0

9.6

287

504

623

824

2,515

1,095

1,028

1,157

89

182

554

1,179

3,921

3,389

4,104 667

5,248 1,019

11

53

118

244

789

648

563

711

110 101

214 130

386 191

676 293

2,663 531

1,7521 781 2,545

3,017

7 30 140 27

19 35 235 57

50 40 360 87

121 70 695 155

402 235 2,365 620

321 140. 1,580 518

580

813

802

1,429

2,409

4,257

14,041

10,224

9,487

-1 7

-3 9

-8 0

-1 5 6

-6 2 3

-4 5 9

-3 8 0

-5

-2 7

-9 1

-1 7 6

-5 8 1

-4 4 1

780

1,363

2,238

3,925

12,837

9,324

1936b

9,107

1939'

11,510

“ Reference No. 1, Part I, pp. 218-19, 224-25 and 234-35; Part II, pp. 4-5. Although the income totals refer to home-produced income, they are practically identical with national income, inasmuch as the net income from abroad was consistently small during the years in question. Cost-of-living index derived from the same source (p. 247). Net output of construction is not included. No deductions are made for depreciation on buildings. The value of raw materials and services contributed by domestic industries to construction is deducted. b The 1936-39 figures are not comparable with the 1861-1930 series. They include construction, but not services of housewives. They are derived from Reference No. 2b, p. 2. 0 Including services of household goods (furniture, implements) and motor cars owned by consumers

capital form ation. N o estim ates w ere m ade of net capital form ation, but a general state­ m ent was m ade that during th e period 1926-30 net capital form ation probably am ounted to 900 million k ronor a year, or 11 per cent of the national incom e exclu­ sive of the services of housewives.28 T he estimates of gross capital form ation for the period before 1896 excluded con­ struction. In 1941 an attem pt was m ade to measure net capital form ation and total n a­ tional expenditure fo r the years 1936-39.29 T he resulting figures are given in T able 24-5.

II. Postwar Estimates A fter W orld W ar II substantial changes w ere introduced in the Swedish estimates, w hich h ad now becom e official, and were published annually by the Institute of E co­ nom ic R esearch (K o n ju n k tu rin stitu t). They w ere prepared by the final products m ethod rath e r th an by the n et o u tp u t m ethod and em phasized the gross national p ro d u ct rath er th an the national incom e. M oreover, the social accounting approach was given a dom inant role in th e presentation of the

2 4 . SW EDEN

Table 24-5. Consumption and Capital Formation in Sweden, 1936-39 (million kronor) 1936

1939

Consumption Agricultural products Manufactured products Transportation services Trade services Services of finance Services of dwellings Other private services a Government services

930 3,626 366 1,353 55 875 300 802

1,051 4,717 436 1,722 63 960 300 1,533

Total consumption Domestic capital formation

8,307 1,231

10,782 2,042

Total 9,538 Net exports of goods & services + 16

12,824 -3 8 0

Consumption and total capital formation 9,554

12,444

Gross Capital Formation Buildings Means of transportation Machines and stocks Total Depreciation Buildings Means of transportation Machines and stocks Total Net Capital Formation Buildings Means of transportation Machines and stocks Total

1,050 267 637

1,717 350 1,009

1,954

3,076

265 76 382

304 125 605

723

1,034

785 191 255

1,413 225 404

1,231

2,042

“ Hotels, restaurants, and professions. so u rc e :

Reference No. 2b,

pp.

4-5.

estimates. W hile this new er series extended back only to 1938-39, w ork was u nder w ay by 1953 that w ould carry these estimates still further back, m aking them com parable con­ ceptually and m ethodologically w ith those covering the postw ar years.

1. Concepts and Basic M ethods In the Institute’s estim ates, gross national p ro d u ct at m arket prices includes a con­ cept of gross asset form ation th a t is exceed­ ingly com prehensive and includes all sorts of m aintenance and small repairs. M ilitary expenditures on capital goods such as build­

345

ings o r guns are treated as investm ent ex­ penditures. In the 1951 publication of the Institute (R eference N o. 3 ), net investment w as presented. Previously, only gross figures w ere given th at contained repairs and m ain­ tenance. T he figures fo r depreciation, how ­ ever, are given only as a fixed percentage o f the gross national p ro d u ct and are re­ garded as unsatisfactory. A ccordingly, the new er series no longer deals w ith national incom e at facto r cost. It is concerned en­ tirely w ith gross national product. G overn­ m en t consum ption excludes an im puted rental value fo r adm inistrative buildings used by the governm ent. T he m ain m ethod of calculating national incom e is from the final expenditure side, since the basic data lend themselves best to this m ethod. These estimates are presented regularly in the form of a “balance of re­ sources” 30 th a t is com posed of the following item s: governm ent gross investment, private gross investm ent, changes in inventories, governm ent consum ption, private consum p­ tion, and exports. T he estimates m ade from the incom e side, based m ainly on incom e tax statistics, were in all cases low er th an the corresponding estimates m ade from the expenditure side. T he reasons for this discrepancy, am ount­ ing to betw een 3 and 5 p er cent of gross national product, was due “in p art to statis­ tical errors the nature of w hich cannot be specified and in p a rt to certain specific sources of error, such as the right of enter­ prises to distribute their provisions fo r de­ preciation over tim e at will, tax evasion, and th e fact th a t the tax statistics do not cover incom es below K r. 600 a year.” 31 W hile efforts w ere being m ade to obtain m ore com plete coverage of income, the esti­ m ates based on expenditures were believed to be m uch m ore reliable and any discrep­ ancies betw een the two w ere added to those obtained from data on incomes. N o industrial classification of the gross national p roduct was available at the tim e o f this writing.

2. E stim ates from the Expenditure Side 32 U nder the new approach, private con­ sum ption was estim ated, fo r the m ost part, by m ultiplying quantities consum ed during a year by th at y ear’s average prices. This was used fo r the greater p art of the con­

346

THE INCOM E OF NATIONS

sum ption of foods and beverages, housing, fuel and light, clothing, and, am ong durable household goods, m otor vehicles. Q uantity figures were based on production data plus im ports; exports and stock increases were deducted to obtain quantities consumed. Price data were available partly from the Price C ontrol A gency and from different trading organizations. F o r calculating the expenditure on housing, d ata on the total num ber of dwellings and total rents (actual and im puted) w ere available fo r 1945 as a result of a census of dwellings undertaken by the Social Board. T hose fo r other years were derived from the 1945 total value by cum ulating the product of the num ber of newly constructed dwellings and their aver­ age rent. N o figures were available fo r quan­ tities or for changes in stocks of durable household goods. Values of the supply of these goods were estim ated fo r 1938-39 and for 1946 by m eans of production and fo r­ eign trade statistics. T rade m argins were added to obtain the retail value, which, in turn, was taken as the value of consum p­ tion. F o r subsequent years, the values for 1946 w ere extrapolated by m eans of statis­ tics of retail trade turnover com piled by the Social Board. T he expenditures fo r private education and research were believed to be small. Public consum ption consists of all cur­ rent expenditures (excluding gross invest­ m ents) on goods and services by the cen­ tral governm ent and local authorities. T he figures given fo r the central governm ent were obtained m ainly from the A ccountantG eneral. Com plete figures of the finances of local governm ent w ere available up to 1949. F o r 1950 data were obtained from the Cen­ tral Bureau of Statistics. Inform ation for 1951 and 1952 was obtained by a special sample investigation m ade by the K onjunk­ turinstitut and the C entral Bureau o f Sta­ tistics, w hich solicited inform ation relating to wages and salaries, purchases of goods, and planned investments. Gross dom estic asset form ation, including all expenditure for building and construc­ tion, purchases of other capital goods, and all expenditure fo r their m aintenance and repair was shown separately fo r private in­ vestm ent and fo r public investm ent. W here data w ere available, the m ethod of calcula­ tion consisted of multiplying average prices by quantity figures. In 1950, the proportion estim ated in this w ay was only about 25 per

cent (as against 67 per cent fo r private con­ su m ption). This m ethod was used prim arily fo r the value of new dwellings. Inform ation was available relating to the num ber of com ­ pleted dwellings and dwellings u nder con­ struction. A verage building costs were de­ rived from the statistics of perm itted con­ struction. A n investigation by the Social B oard yielded inform ation on the m ainte­ nance and repair of dwellings. Similar data on quantities and prices were available for transport equipm ent. Capital expenditures of industry were de­ rived from sam ple statistics o f the B oard of T rade, w hich included the value o f repair and m aintenance. Projections o f these sam ­ ple data fo r the whole econom y were m ade on the basis o f the ratio between the total num ber of w orkers in different branches and the n um ber of w orkers in the reporting enterprises. T he sample figures were thus increased by approxim ately 20 per cent. E xpenditure on agricultural m achinery was estim ated on the basis of sam ple inquiries. Inform ation on the value o f capital expendi­ tu re by public authorities was obtained from governm ent trading departm ents, w orking budget figures, etc. F o r the rest of capital form ation, various sources were used. Changes in inventories w ere shown, even though the recording o f stocks was incom ­ plete. T he estimates fo r 1938-39 w ere based on an assessment o f the increase of im ports during July 1938 to June 1939 over the period Jan u ary 1936 to June 1938. T he am ount by w hich the average im ports in 1938-39 exceeded those of 1936-38 was treated as an increase in stocks. T he 1946 figure was unsatisfactory, being based on a variety of quantitative data m ultiplied by corresponding average m arket prices. Only since the beginning of 1947 have such sta­ tistics been im proved. N e t purchases o f goods and services fro m abroad and n et incom e paym ents fro m abroad w ere taken from the official statis­ tics of the B oard of T rade. T he value of freight and other services w ere obtained fro m the C entral B ank of Sweden.

3. E stim ates from the Incom e Side A s was m entioned above, the estim ate of the gross national p roduct fro m the incom e side was incom plete. Its com parison with th e m ore com plete estim ate from th e ex-

24.

SW EDEN

penditure side showed a discrepancy o r re­ sidual th a t was credited to th e household account for 1938-39. Since this residual subsequently showed large variations from year to year, a p a rt of it, the regular re ­ sidual, was assum ed to m ove in proportion to the incom e of households and private nonprofit institutions, and was credited to the household account. T he rem aining part, term ed the “irregular” residual, was treated as undistributed saving accruing to house­ holds and enterprises. It was not possible to divide this regular residual betw een wages and salaries and incom e from property and entrepreneurship. It is obvious th a t the estim ate of the gross national product from the incom e side plays a purely subsidiary role in the new er Swedish estimates, and the estim ates are no longer prepared to any significant extent by the n et output m ethod. T he m ain approach is th at of the final products or expenditure method. T he organization of the figures u nder the new estim ates prepared fo r the O E E C is shown in Tables 24-6 and 24-7, one cover­ ing the gross national product from the ex­ penditure side and the other the gross and n et national incom e from th e incom e-dis­ tributed side. T his series is not com parable with those of the prew ar period. Tables 24-6 and 24-7 below conform to the O E E C ’s classification o f incom e-distrib­ uted estimates and of expenditure o r final products estimates. T he tables presented in the Swedish publications are not exactly

347

com parable to m ost estimates of gross n a­ tional product, since the calculation of n a­ tional incom e and gross national product w ere attem pted to show how they could be fitted, w ith certain additions, into a national accounts scheme. T hus, the “balance of re­ sources” th at is built up from the accounts of households, governm ent, etc., shows the supply an d use of the total product avail­ able to the Swedish people on opposite sides. T he supply side (debit) consists, accord­ ingly, o f gross national p roduct plus im ­ ports; the use side of the account (credit) is com posed of governm ent gross investment plus private gross investment, plus changes in inventories, plus governm ent consum p­ tion, plus private consum ption, plus ex­ ports.33 Table 24-8 shows this presentation in the Swedish official publication of 1954. T he Swedish estim ators publish two dif­ ferent national accounts systems and meas­ ures of gross national product. T he first is intended to be used fo r analysis of in p u t and productivity, while the second is prim arily designed to serve business-cycle analysis. In the first, depreciation estimates were m ade. T he second, and less com plex one, is based on a simple set of social accounts. D epre­ ciation, fo r example, is taken at w hat busi­ ness firms themselves have w ritten down and not, as in alternative 1, at some am ount w hich can be considered to correspond to real w ear and tear.34 G ross national product estim ates based on these alternatives are shown in T able 24-9.

Table 24-6. Gross National Expenditure of Sweden, 1938/39 to 1951 (billion kronor)

Private expenditure on goods and services Government current expenditure on goods and services Asset formation: Private gross fixed asset formation Public gross fixed asset formation Change in stocks Net borrowing and net transfers from the rest of the world ( + or —) Gross national product a t market prices

1938/39

1946

1949

1950

1951

8.4

15.0

18.1

19.5

21.9

1.2

2.8

3.7

4.0

4.9

1.5 .4 .3

2.9 .9 .1

2.9 1.6 .1

3.6 1.7 - .2

(a)

- .1

+ .5

+ .2

4.2 2.2 1.4* + .9

---

---

---

---

--- -

11.8

21.5

26.9

28.7

35.5

(a) Less than .1 billion. * This figure is considered unreliable because large price increases during 1951 made calculation of average prices extremely difficult. s o u r c e

:

Reference No. 4, p. 17.

THE INCOM E OF NATIONS

348

Table 24-7. Gross National Income or Product of Sweden, 1938/39 to 1951 (billion kronor) 1938/39 Compensation of employees Income from farming Professional, independent traders’, and partnership income Interest, rent, dividends, and corporate grants to households and private non­ profit institutions Less: interest on the public debt Undistributed corporate income (public and private) * Government income from property and entrepreneurship Regular residual f Irregular residual f National income Indirect taxes less subsidies Net national product at market prices Depreciation and other operating expenses Gross national product at market prices

1946

1949 14.9 1.7

1950

1951 18.9 1.9

5.8 .7

11.3 1.3

.7

1.7

1.8

1.9

2.6

.8

1.0 - .4

1.0 - .4

1.1 - .4

1.2 - .4

(a)

(a)

.6 .9 ± z 3.4 ± x t

.6 1.1 ± z 5.3 zb x |

- .2

15.6 1.8

1.3 ± x

2 . 2 dz X

.3 .9 ± z —-

.5 .7 ± z (a)

10.3 ± y .8 11.1 dby

18.5 ± y 1.7 20.2 ± y

23.2 ± y 2.0 25.2 ± y

25.0 ± y 2.0 26.9 ± y

31.0 ± y 2.4 33.4 ± y

.7 ± y

1.3 ± y

1.6 ± y

1.7 ± y

2.1 ± y

11.8

21.5

2.9 ± x .6 .9 - .1

db Z

26.9

28.7

35.5

(a) Not given. * As no estimates of corporate saving are available for 1950 and 1951, corporate saving for these years forms a part of the regular residual. The residual (see Reference No. 4, p. 16) is dependent on errors in the estimation of depreciation and corporate saving which are denoted by y and x respectively. Thus x and y represent an unspecified error in the figures. f See Note *; z = x + y. J Includes corporate saving. s o u r c e : Reference No. 4, pp. 18-19.

Table 24-8. “ Balance of Resources,” Sweden, 1938/39 to 1953 (billion kronor) Supply Gross national product at market prices Imports of goods Decrease of stocks, etc. Total supply Use Private domestic gross investment Public domestic gross investment Increase in stocks, etc. Export of goods and net services Private consumption Public consumption Total use

1938/39 1946 23.3 12.8 2.3 3.4 — — ----------15.0 26.7

1949

1950

1951

1953

29.3 4.3 — -----33.6

31.5 6.1 .1 -----37.6

38.5 9.2 — -----47.7

43.4 8.2 .4 -----51.9

5.5 3.0

7.3 5.7 8.5 25.1 5.4 51.9

2.2 .9 .3 2.2 8.5 1.0

4.2 1.8 .1 3.3 15.1 2.2

4.8 2.8 .1 4.8 18.1 3.0

6.3 19.6 3.2

6.6 3.8 1.0 10.1 22.2 4.0

15.0

26.7

33.6

37.6

47.7

—

—

s o u r c e : National Bokforing for Sverige, 1938/39 och 1946-1953 (National Book-keeping for Sweden, 1938/39 and 1946-1953), Konjunkturinstitut, Stockholm, 1954, p. 3.

2 4 . SW EDEN

349

Table 24-9. Conceptual Differences Between GNP Totals, Sweden, 1938/39 to 1953 (billion kronor)

Gross national product a t market price Less: Indirect taxes Plus: Subsidies Gross national product a t factor cost, Alternative I Less: Repair and maintenance by enterprises Gross national product at factor cost, Alternative II so u rc e :

1938/39 12.8 .9 .1 12.0 .9 11.2

1946 23.3 2.0 .4 21.6 1.6 20.0

1949 29.3 2.4 .4 27.3 2.3 25.0

1950 31.5 2.5 .6 29.5 2.4 27.1

1951 38.5 2.8 .4 36.1 3.0 33.1

1953 43.4 3.4 .4 40.4 3.3 37.1

National Bokforing for Sverige, 1938/39 och 1946-1953, Konjunkturinstitut, Stockholm,

1954, p. 4.

25. SOVIET RUSSIA References 1. V. Katz, Narodny Dokhod, U.S.S.R. (National Income), Moscow, 1932. 2. D. I. Chernomordik et al., Narodny Dokhod, U.S.S.R. (National Income), Akademia Nauk, Institut Economiki, Moscow, 1939, prepared in collaboration with A. I. Pashkoff, A. B. Bolgov, M. D. Kolganoff, and D. I. Moskvin (hereinafter referred to as “Cherno­ mordik”). 3. M. B. Kolganoff et al., Narodny Dokhod, U.S.S.R. (National Income), Akademia Nauk, Institut Economiki, Moscow, 1940 (hereinafter referred to as “Kolganoff”). 4. I. M. Krasnolobov, Planirovanie i Utchet Narodnago Dokhoda (National Income Planning and Accounting), Moscow, 1940. 5. A. I. Petrof, Narodny Dokhod, U.S.S.R. (National Income), Obschestvo Raspostr. Politich. i Nauchnych Znanij, Moscow, 1949. 6. Bolshaia Sovietskaia Enziklopedia (The Great Soviet Encyclopaedia, “National Income”), 2e izd., Moscow, 1954, vol. 29, “Nazionalny Dokhod,” p. 302. 7. S. N. Prokopovitch, (a) Economic Bulletins published from 1928 to 1938 in Prague; (b) Memorandum 3, The National Income of the U.S.S.R., Birmingham University; Bureau of Russian Economic Research, November, 1931; (c) Quarterly Bulletin o f Soviet Russian Economics, Geneva, 1939-1941 (especially numbers 1-2 and 7 ); (d ) Russlands Volkswirtschaft unter den Soviets, Zurich, 1944, Chapter 6; and (e) Histoire Economique de I’U.S.S.R., Paris, 1952, Chapter 13. 8. Colin Clark, (a) A Critique o f Russian Statistics, Macmillan, London, 1939; (b ) his article in Review of Economic Statistics, Cambridge, Mass., vol. 29, 1947, pp. 213-217; (c) Review of Economic Progress, Brisbane, Aust., vol. 1, 1949, Nos. 2-3, pp. 1-10; (d) Similar studies were made by M. Polanyi, The Contempt o f Freedom, 1940, and A. J. Brown, Bulletin of International News, The Economic War Effort, the U.S.S.R., March 4, 1944. 9. Paul Studenski, “Methods of Estimating National Income in Soviet Russia,” Studies in Income and Wealth, vol. 8, National Bureau of Economic Research, New York, 1946, pp. 195-234. 10. Paul Studenski and Julius Wyler, “National Income Estimates of Soviet Russia, Their Dis­ tinguishing Characteristics and Problems,” American Economic Review, Proceedings, vol. 37, 1947, pp. 596-610. 11. Julius Wyler, (a) “The National Income of Soviet Russia,” Social Research (New Y ork), vol. 13, 1946, pp. 502-518; (b) “Die Schatzung des Soviet Russischen Volkseinkommens,” Schweizerische Zeitschrift fur Volkswirtschaft und Statistik (Zurich), Nos. 5 and 6, 1937. N otes begin on p. 539.

350

THE INCO M E OF NATIONS

12. U.S.S.R., Central Statistical Office (C.S.U.), Narodnoie Khosiaistvo v 1956 godu, Moscow, 1957 (in Russian). 13. A. Strukof, “The Planning of the National Income in U.S.R.R.,” Planovoie Khosiaistvo, Moscow, August, 1957 (in Russian).

1. History a. Early Beginnings Followed by a Cen­ tury of Inactivity A s described in C hapter 6, the first n a­ tional incom e estimates in Russia were p re­ pared by several private scholars towards the end of the eighteenth century. These estimates were based on A dam Sm ith’s m a­ terial production concept, w ere developed w ith very inadequate statistical data, and were, consequently, som ew hat crude. F o r m ore than a century thereafter, due to u nfa­ vorable political conditions and correspond­ ingly unfavorable state of Russian statis­ tical thought, no other estim ates were devel­ oped in Russia (see C hapter 9, Section 6 ). The only im portant developm ent in the field that took place there during th a t century, was purely theoretical in nature and, m ore­ over, was lim ited to the first quarter of that century; it consisted of the writings on the concept of national incom e by H einrich Storch.1 H einrich Storch (C hapter 6, Section 4 c ), m em ber of the Russian A cadem y of Sci­ ences, and gifted economist, whose writings enjoyed wide circulation throughout W est­ ern Europe, was a follow er of A dam Sm ith in all respects except in his concept of the nature of production and o f the substance of national incom e. H e criticized Sm ith’s restriction of the concept of production to the creation o f m aterial goods and of the concept of national incom e to the incomes derived from the latter source only. H e m aintained th a t the rendering o f services contributes just as m uch to econom ic wel­ fare and is just as m uch a source of n a­ tional incom e as the creation of m aterial goods. H e w rote persuasively and interest­ ingly, and his plea fo r a broader concept of production and of national income had considerable effect on contem porary thought on these subjects in F rance, G erm any, and England. But, as already stated (C hapter 9, Section 6) he did not believe th a t national income could be m easured statistically w ith any degree of reliability; and he did not attem pt to m easure it either fo r Russia or for any other country.

D uring the period of severe political re ­ action, em bracing the greater p a rt of the reign of A lexander I and all of the reign of N icholas I and extending from about the year 1810 to the y ear 1855, econom ic lit­ erature in Russia practically ceased to exist; an d statistics becam e com pletely dom inated by the G erm an descriptive school, turning into a m ere cataloging of facts and record­ ing of figures. W ith the ascension of A lex­ ander I I in 1855 cam e a period of some­ w hat m ore enlightened o r less obscurantist political policies, lasting through the reigns o f his im m ediate successor and, beyond that, to the end of the century. A fairly vigorous econom ic literature developed during this period despite the barriers o f official censor­ ship. It concerned itself w ith the problem s of the em ancipation of the peasants from serfdom and, later, after em ancipation, with the m eans by w hich the peasants could be helped to m aintain themselves on their small landholdings; as well as w ith the problem s o f fiscal and m onetary reform , of tariff pol­ icy, and of the rise of m odern industrialism w ith its atten d an t periodic crises, unem ploy­ m ent, and labor m ovem ent. Statistics during this period freed themselves from the G er­ m an influence an d becam e m ore interpreta­ tive an d m eaningful, especially as practiced in th e Z em stv o organizations (local selfgoverning ru ral econom ic b odies). B ut even during this period of som ew hat lessened po­ litical reaction, no inquiry was initiated into the co u n try ’s national income. T h e political an d social conditions of the period did n o t generate any interest in the subject. Both the econom ic an d the statistical literature of this period continued to ignore this subject (see C hapter 9, Section 6 ) .2 It is, therefore, not surprising to find that m ost of the estim ates of nineteenth century R ussia’s national incom e w ere m ade by for­ eigners— Leone Levi, distinguished English econom ist, statistician, an d jurist, in the 1860’s and G . M. M ulhall in the 1880’s— and w ere devoid of m uch, if any, detail o r supporting evidence; an d th a t only one estim ate was prepared by a Russian statisti­ cian— V. I. Pokrovsky— an d th en only to ­ w ards the very end of th at century, in 1897.3

25.

SOVIET RUSSIA

b. Prokopovitch’s Estimates of 1906 and 1917 Finally, w ith the institution in Russia of sem iparliam entarism and of greater freedom of the press after th e Revolution of 1905, some interest in the m easurem ent of the na­ tional incom e developed in the country; and this interest becam e especially intense du r­ ing W orld W ar I in connection w ith the form ulation of new fiscal m easures by the governm ent. T hree national incom e esti­ m ates by S. N . P rokopovitch o r u n d er his direction were prepared (as described in C hapter 10, Section I f ) , one in 1906 fo r the year 1900, w hich was p riv a te ;4 and the others, in 1917 and 1918, fo r the years 1 9 1 3 -1 4 through 1916-17, w hich were semiofficial.5

c. Official Soviet Estimates During the Period of Their Free Publication In the m iddle 1920’s, soon after the coun­ try’s political and econom ic stabilization, the Soviet governm ent began, in conjunction with the preparation of its long-range eco­ nom ic planning, the preparation and pub­ lication of national incom e estimates. In these estim ates,6 national incom e was cal­ culated by the net output m ethod and was shown according to its origin in the several divisions o f m aterial production. It was also broken down into the portion created in the so-called “private” sector of the econom y (independent farm ing, private trading, and h andicraft production) th a t was then still in existence, and the portion created in the so-called “collectivized” sector, then being rapidly expanded at the expense of the p ri­ vate sector. T he national incom e was m eas­ u red both in current and in constant prices so as to reflect both its current com position and its growth. T he constant prices were at first taken as of the year 1913, so as to represent conditions existing prior to W orld W ar I. Beginning in 1928, however, follow ­ ing the consum m ation of an extensive re­ form and stabilization of the m onetary and price systems, the national incom e com pu­ tations in constant prices w ere changed to the prices of 1926-27. Some interest also began to be displayed during the twenties in the distributional as­ pects o f the national incom e; and its divi­ sion into wages, profits of enterprises, and taxes, and into consum ption expenditures and capital accum ulation was shown. T he incom e paym ents to individuals w ere broken

351

dow n into shares of various social classes, such as industrial workers, employees, petty tradesm en, independent farm ers, K olhoz m em bers, and the like. N ational income estimates were then being co-ordinated with the national budget, w ith the unified finan­ cial econom ic plan, and w ith the five-year program of capital investments. D uring this period the estimates becam e increasingly inform ative and were quoted in governm ental budgets and other docu­ m ents of prim e im portance. T hey w ere rap­ idly developing into an im portant tool of financial, economic, and social planning and analysis.

d. Shift to a Policy of Secrecy Sometime after 1931, however, a sharp change in the scope, form , and content of the Soviet governm ent’s national income es­ tim ates took place. T he fu rth er estimating of national incom e in cu rren t prices was suddenly abandoned and only the estimating in 1926-27 constant prices was continued. This had the effect of cutting down m a­ terially the inform ational content of the estimates. T heir value was further reduced by the sim ultaneous suspension of the fu r­ th er preparation and publication of the uni­ fied index o f prices. T his index was found by governm ent authorities to place undue w eight on prices in the private as distin­ guished from the governm ent m arket, to account inadequately fo r the prices of new goods, and otherw ise to give a distorted pic­ ture of the real changes in the price level and a w rong basis on w hich to form ulate governm ental price and m onetary policies. T he subsequent shift to active m ilitary pre­ paredness and, finally, the outbreak of the w ar undoubtedly m ust have added further reasons against the publication, if not prepa­ ration, of extensive statistical data on the n ature of national production and on the com position of national income. W ith the discontinuance o f estimates of national income in current prices, the esti­ m ates in constant prices lost their contact w ith reality. T he estimates becam e com ­ pletely divorced from all current financial transactions of society, all of which are ex­ pressed in cu rren t prices. T hey could no longer be related to the national budget, the unified financial econom ic plan, and the fiveyear plan, or any parts thereof. N ational incom e could n o t be com pared with the ag­ gregate cu rren t wage paym ents, profits of enterprises, taxes, governm ent expenditures

352

THE INCOM E OF NATIONS

for m ilitary and social purposes, consum er expenditures, replacem ents of and additions to existing capital, and other published data on the various aspects of the financial and econom ic life of the national com m unity. N ational incom e could no longer be broken up into any of these com ponents o r any other distributional shares. A ll of such breakdowns th a t h ad been started in the late twenties were now given up. N ational income estimates becam e a m ere index num ­ ber of the grow th of m aterial production and a very im perfect and abstract index of it, at that. T he size of the national incom e expressed in 1926-27 prices lost any sig­ nificance taken by itself. N ational incom e estimates becam e of very little use in finan­ cial, economic, and social planning and, hence, of slight interest to either the gov­ ernm ent o r the com m unity at large. This is the status of national incom e esti­ m ates in Soviet R ussia today (1 9 5 7 ). The estimates have very little inform ative value. U nlike their role in capitalist countries, w here national incom e estimates are used extensively in fiscal and econom ic planning, in Soviet Russia, w here planning is used on a m uch larger scale, national incom e estimates seem to play very little part. Some Soviet economists in th e 1930’s recognized the unsatisfactory condition of the existing national incom e estimates and seemed to wish to see it corrected. They urged the resum ption of the earlier calcula­ tions of national incom e in current prices and its breakdow n by distributive shares. T hey also advocated renewal of the w ork on a unified price index w hich w ould reflect changes in the purchasing pow er o f the ruble. T hey em phasized th e im portant role that national incom e estimates could and should be playing in social and economic planning. T hey stressed the fact that the very core of th a t planning m ust be the planning of the national incom e itself in all of its m ajor distributional as well as productional aspects, and th a t this required the preparation of com prehensive national income estim ates.7 U p to the present tim e (1957) these dissents from the traditional approach have not had any effect on the officially decreed conceptual approaches. N o r do the published national income sta­ tistics go beyond furnishing national incom e totals in 1926-27 prices. N o breakdow ns of the totals by industrial divisions are given. F requently, even the totals are om itted, and only the index of grow th of the total since

1926—27 is cited. T his is all th a t the latest official statistical handbook does (see Centr. Statist. U pravlenie, N arodnoie K hosiaistvo S S S R v 1956 godu, M oscow, 1957, p. 4 2 ).

e. Critical Attitudes Toward Soviet Estimates Abroad F ro m th e very outset th e Soviet estimates w ere regarded w ith suspicion abroad be­ cause they indicated a rate o f grow th o f the co untry’s national incom e th a t was unbe­ lievably high. Independent calculations m ade by several expatriate Russian econom ists as well as by non-Russians confirm ed this sus­ picion, suggesting a m uch low er rate of grow th of the co untry’s national income. Independent calculations of the Soviet n a­ tional incom e becam e even m ore im portant w hen the Soviet governm ent ceased publish­ ing any details fo r its b ro ad and enigmatic figures. Especially notew orthy am ong these private estim ates were those of Prokopovitch published over a period of m ore th an thirty years in Prague, G eneva, and B irm in g h am ;8 o f Colin C lark published in E ngland and A u stralia;9 A bram Bergson,10 N au m Jasny,11 Julius W yler,12 and P aul B a ra n 13 in the U nited States; an d J. Rosen in Sw itzerland.14 Some of the independent estim ates pub­ lished outside Soviet Russia w ere m ade at constant British pound sterling prices (Colin C lark’s ) ; some at constant A m erican dollar prices (C lark ’s fo r certain years, W yler’s and B aran’s ) ; some at cu rren t Soviet ruble prices (B ergson’s, B aran’s, an d R osen’s ); som e at constant Soviet ruble prices (Jasn y ’s ) ; and C lark’s latest— in I.U .’s.

2. C oncept of N ational Income T h e Soviet concept of national incom e is founded on K arl M arx’s idea of it as “that p a rt o f th e [material] product, evaluated in m o n etary term s, w hich is newly created each y ear by th e labor o f th e society and becom es available annually fo r consum ption an d accum ulation.” 16 A 1955 official pub­ lication, designed fo r use in college courses on Soviet finance, defined the national in­ com e in Soviet Russia even m ore explicitly as “th a t share of the social p ro d u ct w hich is produced by newly applied labor of w orkers engaged in m aterial production and used to the end of socialist accum ulation an d consum ption.” 18 Incom es from gov­ ernm ental and personal services as well as services o f dwellings are n o t considered to

2 5 . SOVIET RUSSIA

be a p a rt of the national incom e. T hey are treated as “derivative” incom e, resulting from the redistribution o f the “prim ary in­ com e.” T he Soviet concept of national incom e appears to be m uch narrow er theoretically th an the concept em ployed in capitalistic countries, but actually the m argin o f dif­ ference is not very great. F irst of all, the exclusion of governm ental services is coun­ terbalanced in large p a rt by the m ore com ­ plete inclusion of taxes. W hereas estimates in the capitalist countries as a rule include only direct o r personal taxes, the Soviet esti­ mates include all indirect taxes as well, re ­ sembling in this respect the “national in­ com e at m arket prices” concept o f the W estern economists. Secondly, not all non­ governm ental services are excluded from the Soviet national incom e. Some are in­ cluded u nder another nam e. F o r example, the services of com m unal restaurants are included u nder retail trade, w hile legal serv­ ices to business are included in the general overhead of enterprises. Thirdly, some of the excluded services, such as hotels, laun­ dries, cleaning, hairdressing, and the like, are m uch less developed in Soviet Russia th an they are in m ore industrialized coun­ tries, and their exclusion from the national incom e in the Russian estim ates is, there­ fore, of relatively little consequence. H ow m uch th e Soviet national incom e is underestim ated on account of the use o f the narrow er concept is difficult to say. One Soviet economist, reversing the com parison, charged (in 1940) that other countries, em ­ ploying the broader concept, overestim ate their national incom e on th a t score by be­ tween 25 and 40 p er cent.17 H ow ever, the author in question failed to account fo r the m ore com plete inclusion of taxes in the Soviet estimates and otherw ise seemed to overstate the difference betw een th e tw o types of calculation. A nother Soviet econo­ mist, Professor Petrof, in a m ore recent publication, estim ated th a t by including in­ comes from services the U nited States esti­ m ates “exaggerated their country’s national incom e about 15 p er cent.” 18 It is interest­ ing to note th a t the authors of th e official Y ugoslavian estimates, w hich also em ployed the m aterial production concept, recently estim ated that it w ould be necessary to raise th eir figures by about 15 p er cent in order to relate them to the concept o f production em ployed in capitalist countries. In the ab­ sence o f detailed data it is impossible to

353

m easure exactly the degree to w hich the Soviet calculations are short of the W estern ones because of th eir narrow er conceptual base. B ut the probability is th a t it is not very great.19 In Soviet Russia, the national income is conceived strictly as the n et value of the national output. It is divided into two parts: national incom e distributed to individuals engaged in m aterial production in th e form o f wages paid; and the collective share of the national incom e consisting of profits of enterprises, n et short-term interest, social levies, and taxes. T he profits of enterprises and the turnover taxes are the socialist counterparts o f private profits in capitalist society an d form a larger share of the n a­ tional incom e th a n do profits in capitalist societies. In addition to the prim ary incomes of individuals th at are a p art of the national income, there are the so-called “derivative incom es” of individuals th at are not, di­ rectly, a p a rt of the national income. These incom es are earned by state employees, em ­ ployees of social institutions, and other p er­ sons rendering personal services. They, too, take the form of wages and are paid out of the taxes, social levies, and prim ary incom es o f individuals engaged in m aterial produc­ tion. T hey are treated as parts, n ot o f the original distribution of the national income, but of its subsequent redistribution. W hereas in the capitalist countries using the com prehensive production concept n a­ tional incom e norm ally is som ew hat in excess of the total incom e paym ents to individuals, inasm uch as it includes undistributed profits, in Soviet Russia, w ith its restricted national incom e concept, the reverse m ay readily hold true. A t least during the 1930’s, the sum o f individual prim ary an d derivative incom es was estim ated by one Soviet econo­ m ist to be 10 per cent or m ore in excess of the national incom e to tal.20 H ow these ag­ gregates w ould com pare at th e present time, it is impossible to tell.

3. Statistical Sources T he official estim ates reviewed here were prepared by the central planning and ac­ counting agencies of the governm ent, the G osplan, and, later, by the T Z U N H U . The d ata w ith w hich gross and n et values of output could be com puted were reasonably com plete fo r industry, transportation, and

354

THE INCOM E OF NATIONS

trade, where adm inistrative organization and financial control w ere highly centralized. They were som ew hat less adequate for agri­ culture and construction, w here controls and, hence, records were fa r less centralized. T he data fo r the socialized sector of the econom y cam e from the annual reports of the state trusts, farm s, and co-operatives and from the consolidated reports prepared by the T Z U N H U itself. T he d ata necessary to com pute the net value of output of the nonsocialized sector, i.e., of independent artisans and farm ers, were obtained during the 1930’s from industrial, agricultural, and other censuses, incom e tax statistics, and sample studies. A ccounting and production records w ere so com prehensive and so highly centralized u nder the Soviet system of so­ cialized ownership and central control of econom ic enterprises as to m ake the job of estimating the national incom e, from one point of view, easier than in capitalist economies.

4. D ivision by Econom ic A ctivity M aterial production was deem ed to con­ sist o f six branches: (1 ) agriculture, (2 ) industry, (3 ) construction, (4 ) freight trans­ portation, (5 ) retail trade, and (6 ) miscel­ laneous. As in the statistics of other coun­ tries, the line between agriculture and in ­ dustry was flexible. Owing to the progres­ sive industrialization and specialization of production, certain types of econom ic ac­ tivity form erly classified u nder agriculture, such as butter and cheese m aking, spinning yarn, milling grain, lum bering, and prepar­ ing fuel wood, w ere classified, in the m ore recent production statistics and national in­ come estimates of the 1930’s, under indus­ try. Slaughtering and production of m eat, hides, feathers, bristles, and sim ilar com ­ modities w ere classified u n d er agriculture. There was a growing opinion, however, that these, too, should be included under indus­ try. M ining, fishing, and hunting w ere in ­ cluded under industry. T he line draw n betw een industry and construction was sim ilarly indistinct in some instances. W hen, as in the case of the instal­ lation of m achinery in a factory, the con­ struction was done by an enterprise w ith its own labor, it was classified under industrial production. W hen, on the other hand, it was perform ed by an outside organization spe­ cializing in construction work, it was classi­ fied u nder construction. Theoretically, re ­

pairs, w hether done by special repair shops or by the regular employees of the enter­ prise itself, were classified u nder industrial production, w hile replacem ents of equip­ m ent w ere classified under construction. In practice, how ever, w ork called “repairs” was often so com prehensive as to am ount to capital replacem ents. T he Chernom ordik study advocated crediting such repairs to the construction industry.

5. E stim ates at Current Prices T he n et value of o u tput was estim ated fo r each branch o f the econom y either by calculating gross value and then deducting certain cost item s representing duplications, o r by aggregating directly the various com ­ ponents o f n et value. T he first m ethod was used chiefly fo r ag­ riculture, the construction industry, freight, transportation, and trade. G ross value was com puted in one o f three ways: (1 ) by adding reported receipts from sales, (2 ) by adding all recorded expenses, profits, and tax paym ents, or (3 ) by m ultiplying the re ­ p orted n um ber o f units o f physical output by th eir prices. F o r retail trade, the gross value was calculated as the proceeds o f a stated m arkup o r “gross profit” on the wholesale price of goods. Since prices were fixed by central authority, they varied less an d w ere m ore easily ascertained th an in econom ies w here prices are determ ined by m arket forces. H ow ever, th e fact th a t they varied in som e branches o f the econom y (chiefly agriculture) according to the type of p urchaser som ew hat com plicated the use o f price data. M oreover, the effectiveness of Soviet prices as m easures o f econom ic value was reduced by the fact th a t they were determ ined as m uch by political as by econom ic decisions. T he deductions from gross value necessary to avoid duplications in com puting n et value w ere m ainly for expenses fo r raw m aterials and other com ­ m odities and services purchased from other enterprises; auxiliary supplies and m aterials; purchased fuel and electric energy; repairs (done o u tsid e); and am ortization o f plant and equipm ent. T he second m ethod— direct aggregation of the paym ents o f wages, social security and oth er social levies, n et short­ term interest, profits an d taxes—was used fo r all large-scale industries, inasm uch as th eir accounts provided direct inform ation on all these items.

25.

SOVIET RUSSIA

a. Income from Agriculture

355

The m ain division o f agricultural output was between plant and anim al production. T he C hernom ordik study recom m ended a ninefold division of plant productions: (1 ) grain, (2 ) technical cultures (cotton, flax, sugar cane, tobacco, e tc .), (3 ) feed cultures (hay, e tc .), (4 ) potatoes, (5 ) vegetables, (6 ) fruits and berries, (7 ) citrus fruits, (8) tea, (9 ) decorative and protective plants (flowers, trees, etc.).

N et agricultural outp u t in the socialized sector of the econom y was estim ated on the basis of the annual reports of the sovkhoz (state farm s) and ko lkh o z (collective farm s) and of the tractor organizations by first com ­ puting gross value and then deducting cer­ tain costs. Since agricultural output was subject to several price systems, com putation of gross value was difficult. T he prices of those prod­ ucts agricultural enterprises were obliged to sell th e state were exceedingly low. Sales to consum er co-operatives and to the People’s Com m issariat fo r the P rocurem ent of A gri­ cultural Products, w ith the single exception of grain, were at slightly higher prices, while those sales m ade by the kolkhoz in the open m arket were at m uch higher, so-called “free prices.” Finally, products resold by th e state to individual consum ers as well as to co­ operatives and other organizations, being subject to the turnover tax, w ere the highest priced. In com puting the value o f agricul­ tural output, it was necessary, therefore, for the estim ators to determ ine w hat portion of each com m odity was sold to w hom . A s the resulting valuations gave a distorted picture of the relative im portance of each category of enterprise, the C hernom ordik study p ro ­ posed th at the value of each product be com puted at a single w eighted price, irre­ spective of who produced it o r w here it was delivered.

Anim al production included the usual range of products. The inclusion of slaughtering and the sale of animals for slaughter under animal production was criticized in some Soviet sta­ tistical works as yielding highly misleading figures, as the volume of slaughtering done in any year bore no necessary relationship to the natural reproduction of the animals. It was pointed out that an increase in slaugh­ tering at the expense of the accumulation of livestock or fowl really meant an increase in capital consumption, not in production: for example, that the great increase in slaughter­ ing, 1929-33 (for political reasons), consti­ tuted a wanton destruction of capital, not an increase in production (Table 25-1). On the other hand, the decrease in slaughtering during the succeeding several years meant that live­ stock was being built up, not that animal pro­ duction decreased. These particular statisticians advocated, therefore, a reclassification of animal produc­ tion to include only the breeding of livestock and fowl (i.e., the replacement of slaughtered or deceased animals and the addition to and improvement of the stock of the living) and the production of milk, eggs, wool, and other

Table 2S-1. Turnover of Livestock in Selected Years, 1927-36 (percentages of total at beginning of year) Number at Beginning of Year

Natural Reproduction

1927-28 1929-30 1931-32 1933-34 1935-36

100.0 100.0 100.0 100.0 100.0

42.7 37.5 39.0 47.0 44.2

1927-28 1933-34 1935-36 1927-28 1933-34 1935-36

Natural Death

Slaughter

Number at End of Year

9.0 8.7 10.2 8.4 6.5

30.0 50.6 44.0 28.1 22.9

103.7 78.2 84.8 110.5 114.8

100.0 100.0 100.0

Sheep and Goats 52.2 13.1 51.2 12.3 62.0 9.5

34.1 35.5 32.4

105.0 103.4 120.1

100.0 100.0 100.0

169.2 173.6 180.0

110.8 94.0 107.2

112.0 144.6 134.8

Cattle

Pigs 46.4 35.0 38.0

Reference No. 2, p. 104; for more detailed reproduction see Reference No. 9.

356

THE IN CO M E OF NATIONS

products of living animals, claiming that it would afford a better basis for measuring changes in annual animal production and for distinguishing between capital replacements and income in that sphere. A gricultural production was classified fu r­ ther according to type of ow nership into production on sovkhoz and on kolkhoz; do­ mestic production by kolkhoz m em bers; hom e gardening by u rban employees and workers; and production by independent farm ers. T he gross value of agricultural produc­ tion was com puted as total finished produc­ tion plus any increase, o r m inus any de­ crease in unfinished production (the value of the plowing, fertilizing, sowing, and other w ork done for the next year’s c ro p ). Total unfinished production as well as increases or decreases in it from year to year in m any branches of Soviet agriculture w ere rath er substantial. Some Soviet economists m ain­ tained that the inclusion of this variable item was inadvisable, inasm uch as it intro­ duced a considerable am ount of conjecture into the com putation of the value of the output w ith the possibility of either exag­ gerating or understating the actual produc­ tion.21 D eductions from gross value, m ade to ob­ tain net value, covered such item s as seed, feed, depreciation (o r am ortization), chem ­ ical fertilizers, insecticides, fuel, lubrica­ tion, other industrial products, repairs, etc., and other production expenses. T he C herno­ m ordik study, however, m aintained that this classification failed to distinguish adequately between various item s of agricultural ex­ pense.22 In its stead, th e following m ore elaborate grouping was suggested: deprecia­ tion on buildings, m achines, and livestock; current repairs; fuel and lubrication; seed; feed; fertilizer; insecticides; miscellaneous expenses. These cost item s w ere estim ated as follows: Depreciation on buildings, machinery, and livestock, usually based on the estimated life of the property, was computed at a fixed per­ centage of the original investment.23 On the ground that the gross underestimation of the lives of machines tended to promote careless­ ness in their use, failure to repair them prop­ erly, and made users scrap them prematurely, the Chernomordik study advocated lower amor­ tization rates. It maintained that in a socialist economy there could be no such thing as obso­ lescence, and that the life of machinery should be estimated more liberally. Instead of being scrapped, machinery of old design was trans­

ferred to less developed sectors of the economy. Before 1930 it was customary to estimate the life of a tractor at approximately five years and to compute its annual depreciation at 18-20 per cent of its original cost. Beginning with that year, however, its life was estimated in terms of hours of actual use instead of years of potential use. The life of a tractor operating on chains was fixed at 12,000 hours of actual utilization; of one operating on wheels, at 8,000. Differences in the intensity of the utili­ zation of different kinds of tractors or even of the same tractors by different farms were thus recognized. During the 1930’s the TZUNHU computed depreciation on livestock only for animals em­ ployed in farm work, such as horses and oxen. This was an error in the opinion of Cherno­ mordik and his associates, who felt that the breeder animal is capital, too. They believed that the cost of acquiring and reproducing breeders necessary to maintain the existing level should likewise be taken care of through depreciation reserves.24 The other deductible costs were estimated in the following way: Repairs on agricultural tractors included socalled capital repairs, i.e., actual replacement of worn-out parts, as well as current repairs proper. On other agricultural machinery, how­ ever, after 1938, current repairs alone were classified as such, because allowance for capital repairs in the actual accounting of the sovkhoz and kolkhoz farms appeared to lead to a neglect on their part of ordinary repairs and to a tend­ ency to replace old parts with new on the slightest pretext, with a consequent waste of capital funds. Outlays for fuel and lubricants were com­ puted from the annual reports of the sovkhoz and kolkhoz to their central administrative or­ ganizations and from other reports. Those for seed were computed for the sovkhoz and kolkhoz from their annual records. For inde­ pendent farmers they were estimated 10 per cent higher so as to take account of the less efficient methods of seeding by hand. Outlays for feed, larger than for any other agricultural materials, were also calculated fairly accurately for the sovkhoz and kolkhoz, but for other farms could be estimated only on the basis of sample investigations. The cost of fertilizers was usually included among current expenses and deducted from gross value of output. Ac­ cording to the Chernomordik study, however, such a procedure understated the value of agri­ cultural production. A substantial part of this outlay was said to represent capital investment, inasmuch as many chemical fertilizers tended to improve the fertility of the soil permanently. Such capital investment increased from year to year, and raised the efficiency of agricultural production continually. Any such outlay, ac­ cording to Chernomordik, should, therefore, be treated as a part of net value, not as a de­ ductible expense; the only proper deduction

2 5 . SOVIET RUSSIA

would be for the annual depreciation on the fertilizer. The kolkhoz contributions in kind for the maintenance of the agricultural tractor stations servicing them, included in the net value of agricultural product as a whole, covered only part of the maintenance costs; the other part was covered by state subsidies. T he net value o f agicultural o u tp u t re ­ m aining after these several deductions was com prised of labor com pensation, social security and other contributions to the w elfare of producers, wage paym ents to ko lkh o z m em bers, paym ents to trac to r sta-

357

output; in th e m etal products industries, as m uch as 17.7 p er cent.25 Changes in the am ount of such unfinished production from year to year, however, seemed to have been relatively small. It was adm itted th at M arx, in his discussion o f national income, re­ ferred to finished production alone, and his neglect of unfinished production was explained on the ground th at he was con­ cerned w ith discovering the basic laws of capitalist and socialist production, n ot with th e m ethodology of com puting th eir net value, or national income. A s show n in T able 25-2, the gross value

Table 25-2. Items of Gross Value T hat Were: A P art of N et Value Salaries and wages, basic and supplementary Contributions to social security, workers’ welfare, dissemination of technical information, and training of new workers Interest on loans Insurance Municipal taxes and other charges Profits General turnover tax

Not a Part of Net Value Raw materials Supplies and equipment Purchased fuel Purchased electric light and power Purchased transportation and delivery services Travel, postal service, telegraph, telephone, fines, breakage, losses, etc. Depreciation

tions, profits, interest, taxes, reserves against crop failures, and other item s m ore fully described in Sections 6a and 6b below.

of industrial ou tp u t was divided into those item s that, u nder the Soviet concept, form a p a rt o f n et value and those th at do not. M ost o f th e item s represent production costs. A m ong them , as show n in Table 25-3, raw m aterials m ade up 53 p er cent fo r the year 1935 in all industries com bined; sup­ plies, fuel, electric pow er, 10 p er cent; am or­ tization, 3 p er cent; labor costs (including social security and w elfare contributions), 27 per cent; and miscellaneous items, 7 per cent. T h e highest ratio of raw m aterials to total expense was in the food industry, the lowest in the electric pow er industry. Con­ versely, the highest ratio of labor cost was in oil extraction, the lowest in the fats, soap, and perfum ery industries. T h e ratio of am ortization to total expense was highest in the electric pow er industry; lowest in the food and in the fats, soap, and perfum ery industries. In calculating m ost of the items com ­ prising n et value, no special methodological problem s w ere encountered. Some contro­ versy arose, how ever, over the treatm ent of payroll contributions fo r the m aintenance of factory clinics, factory schools, clubs, etc. P rofessor Strum ilin held th a t only the part th a t is spent fo r labor on these facilities should be included in the n et value of the product o f the given industry. T he part

' b. Income from Large-scale Industry Before 1937, the official practice was to confine the concept of m aterial production in industry to w ork perform ed in the fac­ tory, and to include in both gross and net value only items pertinent to such produc­ tion. D uring th a t year, the concept o f m a­ terial production was broadened and the m ethod of com puting both gross and net value modified to cover also costs, profits, and tax paym ents connected w ith th e com ­ m ercial activity of the industrial enterprises carried on outside the factory— accounting, planning, financing, storing, selling, delivery, research, and several other such functions. T he category “business costs” was substi­ tu ted fo r the narrow er “factory costs.” In accordance w ith th e practice follow ed in m ost countries, Soviet estim ates o f both th e gross and net value of industrial output to o k account not only of the year’s finished o u tp u t b ut also of the increases o r decreases in th e unfinished output. T otal unfinished production was substantial in m any branches of th e Soviet economy. F o r example, in 1935, fo r all large-scale industries com ­ bined, it represented 4.8 p er cent o f total

THE IN CO M E OF NATIONS

358

Table 25-3. Production Expenses in Soviet Industry, Excluding Taxes, 1935 (percentages of total) All ProConIndustries ducer sumer Combined Goods Goods Raw materials and basic goods Auxiliary supplies and materials Fuel Electric power Salaries and wages Supplementary labor compensation Social contributions Amortization Miscellaneous Total so u rc e :

Reference No. 2,

p.

Electrie Power

Steel and TexMining Iron tiles

Garments

2

24

74

76

21

16 1 4 48

9 14 2 30

4 1 1 12

1 0.2 0.4 15

1 3 1 3

3 5 16 15

6 6 6 11

3 5 5 8

1 3 1 2

1 4 0.3 2

100

100

100

100

100

0.4

53

34

74

6 3 1 21

7 5 2 30

4 2 1 11

5 33

2 4 3 7

3 5 4 10

100

100

—

100

Food 80 3 2 0.3 7 1 2 1 3 100

63. For more detailed reproduction see Reference No. 9.

spent fo r m aterials, he w rote, is accounted for in the net value of the product of other industries.26 T he authors of the C herno­ m ordik study, on the other hand, held that the then existing practice of covering the full am ount was correct, inasm uch as the entire cost of m aintaining these facilities was borne by the given industry itself. They held th at the inclusion of the full am ount was essential in com puting net value regard­ less of w hich concept of production was used in estim ating national incom e— the concept of m aterial production alone or that of com bined “m aterial” and “im m aterial” production.27 Interest costs w ere regarded as a contri­ bution of the socially ow ned capital to pro­ duction and w ere com puted from the in­ terest contributed by enterprises on their short term loans from the G osbank, less the interest received by them on their bank de­ posits. Interest costs were of considerable im portance in consum er goods industries, especially in the food industry, w here p u r­ chases of raw m aterials during certain sea­ sons w ere so large th at they could not be financed entirely from operating capital. Insurance costs related only to insurance taken out by co-operatives and handicraft industries and artisans. T he properties and operations of state enterprises w ere not insured. Provision was m ade in the state budget fo r their losses from fire or other contingencies. In estim ating the net value of the output of producer goods industries, state subsidies

w ere n ot deducted, for they w ere considered to be a redistribution of the profits earned by industry as a whole and paid into the budget, the p ro d u cer goods industries get­ ting back the profits they w ould have earned h ad they been perm itted to charge full norm al prices to the consum er goods indus­ tries. T his inclusion o f subsidies m ay well constitute double counting, inasm uch as the profits of the consum er goods industries from w hich they w ere paid h ad already been included in the national incom e total.

c. Income from Small-scale Handicraft and Domestic Industry T he gross and n et value of the ou tp u t of small-scale industrial undertakings could not be calculated as precisely as those o f large scale, because their operations w ere reported in less detail. T he estimates h ad to be based on th e gross value as given in th e biennial industrial censuses, and on ratios o f n et to gross value com puted from an analysis of the operations of the corresponding largescale industries. T he censuses covered p rac­ tically th e entire socialized sector of the econom y, nam ely: all industrial establish­ m ents having electric or o th er pow er-driven m otors or em ploying three o r m ore persons; all w orkshops of an industrial character, w hether independent or subsidiary, m ain­ tained in conjunction w ith agriculture, con­ struction, transportation, trading enterprises, o r educational institutions, including black­ sm ith shops, carpentry shops, etc.; all subsidiary w orkshops of the ko lkh o z and

25.

SOVIET RUSSIA

sovkh o z engaged in the prim ary processing o f agricultural goods, such as mills, cream ­ eries, churneries, and bakeries. Only estab­ lishm ents perform ing personal services, such as barber shops, private laundries, and pri­ vate restaurants, as well as independent w orkers and artisans w ere n o t covered. T he net value of artisans’ o utput was estim ated from th eir incom e tax paym ents. Their incom e, considered to be equivalent to the n et value o f their product, was ad­ justed to take care of exem pt incomes. T he n et value of the dom estic industrial ou tp u t of kolkhoz m em bers and independ­ ent farmers-—fuel wood, peat, building m a­ terials, m ushroom s, w ine, fish, and miscel­ laneous handicrafts— was estim ated on the basis of sample studies of fam ily budgets and production.

d. Income from Construction Before 1936, no separate construction in­ dustry existed in Soviet Russia. E ach indus­ trial organization constructed its own plant, organizing and training a labor force and acquiring m aterials and equipm ent. U pon com pleting th e project, it transferred some em ployees to regular production w ork and laid off others, and disposed o f th e equip­ m ent in various ways. In short, the carefully built-up organization and stock of technical equipm ent was dispersed instead o f being used on another sim ilar undertaking else­ w here— obviously a w asteful and inefficient procedure. In 1936 a far-reaching reform

359

was instituted, looking to the organization of a perm anent construction industry, m ore or less centrally directed an d provided w ith perm anent equipm ent and labor force. It was charged w ith the responsibility of doing the construction w ork fo r all state enter­ prises and industrial and trade co-operatives, but n o t fo r the kolkh o z and artisans’ co­ operatives, o r independent farm ers. T he or­ ganization o f the construction industry was fu rth e r im proved in 1939 by bringing all construction agencies u nder the direction of the People’s Com m issariat of Construction. T he com putation of the n et value of an­ nual construction was thereby greatly sim­ plified. M ost estimates w ere based upon the annual accounting and statistical reports of the construction industry and of the kolkho z. T he only estimates th at still con­ tinued to be rough, based as they were on fragm entary data, w ere those covering the construction activities of independent farm ­ ers and artisans’ co-operatives, which, in aggregate value, were relatively small. C onstruction included construction proper, the erection of industrial, agricultural, resi­ dential, public, cultural, com m ercial, and other buildings, as well as other structures, such as roads, railroad tracks, blast fu r­ naces, oil pipelines, telephone lines, w ater supplies, an d dam s; installation of m a­ chinery and oth er perm anent fixtures; grad­ ing, landscaping, and oth er land im prove­ m ents. T he calculations fo r construction

Table 25-4. Components of Construction Costs, 1933-36 Actual Cost to the Construction Agency or to the Undertaking Doing Its Own Construction (exclusive of profits and of the costs of equipment installed in buildings) 1933 Labor (basic and supplementary pay) Materials, total Basic costs Transportation and storage Auxiliary supplies and implements Fuel, steam, gas, electricity Depreciation and rent for use of equipment Limited or conditional outlays Administrative overhead Contributions to social security, workers’] welfare, and training [ Expenses for municipal services to workers J Other outlays Other expenditures

:

Reference No.

2, p . 168.

1936 23.45 52.50 40.09 10.08 1.10 1.23 1.55 19.17 8.11

8.49

9.58

8.78

6.96

2.36 3.41

2.51 2.73

4.07 2.01

4.10 3.33

100.0

Total s o u r c e

24.68 46.77 35.72 8.14 1.35 1.56 2.47 22.67 11.82

1934 1935 (percentages) 26.18 24.95 46.41 49.13 34.74 35.22 8.90 10.45 1.49 1.23 1.28 2.23 2.63 1.64 22.05 22.27 9.42 9.96

100.0

100.0

100.0

360

THE IN CO M E OF NATIONS

done by state agencies were m ade as fol­ lows: The gross value of construction for state agencies was calculated in one of two ways, depending on whether the work was performed for another enterprise or for one’s own. If the former, it was calculated at the contract price, which included, besides the actual costs of con­ struction, a predetermined profit and the turn­ over tax. If the latter, it was computed simply by aggregating all the costs involved. In both, the costs of the machinery and other fixtures installed in finished structures, as well as the increase or decrease in the year’s unfinished structures, were included. The gross value of road construction included also the imputed value of the labor rendered without pay by kolkhoz members and independent farmers (a minimum of six days a year) and of the serv­ ices of the horses, tractors, and other equip­ ment lent by them to the government without charge (Table 25-4). The net value of construction is the residual left after deducting the costs of raw materials and auxiliary supplies such as fuel and electric power (including the costs of their transporta­ tion or transmission), costs of machinery and fixtures involved in the installations, deprecia­ tion and repairs on the equipment of the con­ struction agency, as well as the costs of the scaffolding. As in the case of industry gen­ erally, the net value was comprised of the basic and supplementary pay; contributions to social security, the housing fund, workers’ welfare (medical aid, etc.), outlays for technical in­ struction, and for training of new workers; and, in the case of work done under contract, profit and the turnover tax. The accretion to the capital of the construc­ tion agencies performing work under contract was computed either by deducting expenses from the contract prices or by adding the value of economies realized in executing the financial plan and the profits from the execution of spe­ cific construction projects. The process followed in calculating the net value of construction is shown in the accom­ panying table. However, the net output of

only a portion of total construction work, ag­ gregating in the late 1930’s some 5-6 billion rubles, could be calculated from actual account­ ing records. For the rest of the construction work done for or by state agencies, net value was calculated by applying to gross value the estimated average ratio of net to gross value, or the estimated average ratios of specific expenses to gross value as derived from the accounting records. The total was understandably some­ what rough. The Chernomordik study urged the adoption of more refined calculations using dif­ ferent ratios for different types of construction. It also called attention to inaccuracies resulting from the fact that in construction done under contract, the ratios were based on figures in­ cluding profit, while in other construction they were based on figures embodying no profit. T he n et value of construction done by the ko lkh o z and th eir m em bers, the vari­ ous co-operative organizations, independent farm ers, and w orkers an d em ployees in cities was m ore difficult to calculate: Certain peculiarities in the organization of this type of construction, however, created spe­ cial problems. First of all, the kolkhoz annual reports gave data for only certain items of con­ struction costs. From these figures it was im­ possible to arrive at either the gross or net value of construction. For example, informa­ tion was given on the monetary outlays for materials and labor, but not on the value of home-produced materials. Under these circumstances, net value had to be based on several somewhat doubtful assump­ tions. Gross value was the sum of the reported monetary outlays for construction and the im­ puted value of the labor of kolkhoz members. Costs of material, deductible from gross value, were calculated by applying to gross value the net-gross ratios common in sovkhoz and other state agricultural agency construction. One weakness of this method was that the imputed value of the labor contributed by kolkhoz members to their common construction work was based on their shares of the kolkhoz’s earnings during the year of the construction. Deductions for:

Types of Construction Construction proper Installation Improvement and irrigation of land Landscaping Machinery, fixtures, etc. Geological and other surveys, and other planning work

Cost of the Work to the Ordering Agency

...

Costs of Materials (Basic and Auxiliary), Fixtures, etc.

Depre­ ciation (Amorti­ zation)

Net Value

25.

SOVIET RUSSIA

As a result, the value of the construction itself depended on the kolkhoz’s current earnings, whereas it really was an independent item. This weakness was corrected during the late 1930’s by calculating the labor contributions of kolkhoz members to construction in terms of the wage rates paid construction workers on the sovkhoz. It was even more difficult to estimate the net value of the construction performed by kolkhoz members and urban workmen and employees for their own use. As data were not collected by any single agency, they were exceedingly fragmentary. The state insurance organization had some data on the value of agricultural buildings, but reliable estimates of the volume and value of new construction could not be made; for at least a part of the changes in the annual reported aggregate value of buildings was due to changes in the valuation of existing buildings. Sample studies by the TZUNHU af­ forded some basis for a rough estimate of the volume and value of urban construction by individuals for their own use. The costs de­ ductible from gross values were roughly esti­ mated by the TZUNHU in 1934 at the follow­ ing ratios to gross value: for construction by kolkhoz members, 59.7 per cent; for construc­ tion by farmers, 56 per cent; and for that by urban workmen and employees, 55 per cent.

e. Income from Freight Transportation M ost of the m echanized transportation in Soviet Russia was organized in large stateowned enterprises u n d er various com m is­ sariats— railroad, w aterw ay, and the like— and com prised the transportation industry of the country. In addition, each industrial, trade, or agricultural enterprise m aintained transportation services to m ove its own goods and w orkers from shop to shop w ithin its own walls, or from its sources o f supply to its plant, o r to its stores o r custom ers outside. These transportation services were classified under agricultural, industrial, or other production, and not as an activity of the transportation industry as such. T he latter included only services perform ed by the specialized agencies fo r pay. In the opinion of P rofessor C hernom ordik and his associates, this basis o f distinction was ap­ propriate to a capitalist, b u t n o t to a so­ cialist econom y.28 T he function of transpor­ tation is to m ove goods from enterprise to enterprise or from enterprise to consum er. A ny such moving of goods, even if done in the enterprise’s own vehicles, should be classified under transportation. Only the m oving of goods w ithin the confines of an enterprise (e.g., from the w arehouse to the shop, or from one floor of a plant to an other) need not be so classified.

361

T he value of th e services of passenger transportation, as already noted, was om itted from national incom e on the ground th a t it is not a p art of m aterial production. T ransportation was classified into railway, waterway, autom obile (tru ck and b u s), air­ plane, street railway, an d horse wagon. T ransportation of oil o r gas by pipeline and the transm ission o f electric energy were credited to other industrial categories: The gross value of transportation services was calculated largely from the gross earnings of the transportation industries. Included were fines imposed on shippers for violations of rail­ road rules (aggregating in 1935 more than 200 million rubles at current prices). Fines imposed for holding up empty railroad cars in expecta­ tion of their ultimate loading (aggregating one half of total receipts from fines) represented, according to Chernomordik, a charge for a productive service, corresponding to a charge to shippers for the rent of railroad cars. It was doubted, however, that other fines could be considered to represent a receipt from a pro­ ductive service. N et value is the residual left after deducting from gross value expenses for fuel and other materials and supplies, costs of repairs, and de­ preciation (amortization). Difficulties were in­ volved in computing only the last two items. Under the official practice in vogue during the 1930’s, only the cost o f the materials used in connection with repairs done by the trans­ portation agency itself was deducted from gross earnings. The cost of the labor was included in the net value of transportation services. On the other hand, the entire expense of repairs done outside in special repair shops, including that for labor, was deducted. In the opinion of Pro­ fessor Chernomordik, this would have been correct had the labor been accounted for in the net value of the output of the repair shops. Somehow, the outlay for such labor was not considered a part of material production during the ’30’s. Consequently, national income was understated by several hundred million rubles.29 In estim ating th e depreciation (am ortiza­ tion) o f railroad plant and equipm ent it was difficult to differentiate between capital replacem ents, repairs, and additions to capi­ tal. Theoretically, am ortization covered the wholesale replacem ent of w orn-out parts, as w hen rails w ere replaced along substantial sections of the track o r locomotives were com pletely renovated or replaced. On the other hand, repairs were supposed to cover only replacem ents of m inor w orn-out or missing parts, such as a single rail o r tie, o r the readjustm ent of usable parts. T heo­ retically, additions to capital included new

362

THE IN CO M E OF NATIONS

structures or the replacem ent of old by larger and better new structures, as when a light rail was replaced by a heavy, o r a small locom otive by a m ore pow erful one. In practice, however, these outlays over­ lapped to such an extent th a t it was often impossible to tell in w hich of the three cate­ gories a given outlay belonged: A technical conference held in the late 1930’s by the TZUNHU decided that more than half of the sums spent in nearly all categories of railroad outlays for materials belonged under capital replacement, i.e., amortization. The only exceptions were outlays for fuel, where re­ placement represented only 17 per cent of the expenditure, and outlays for lighting materials, grease, twine, bags, canvas, books, and paper, where no capital replacement was involved. In twenty types of outlay for material, capital re­ placement represented 75-95 per cent, and in eleven, 60 per cent. By using the reported fig­ ures of the proportion of capital replacement in each type of railroad outlay, total railroad expenditures for capital replacements year by year could be approximated. The rate of depreciation varied so greatly that the adoption of any uniform amortization rate for the capital as a whole was scarcely possible. It was estimated, for example, that the railroad bed has a life of some 200 years; and depreciation would amount to no more than 0.5 per cent. Tunnels, according to some authors, have a life of 100 years; stone bridges, 75 years; steel frames, 50 years; rails, 25 years; sleepers, 10 years; brick buildings, 75 years; and frame buildings, 30 years. Other authors gave these structures shorter lives, suggesting rates of annual depreciation of 1.8 per cent for stone bridges, 2.2 per cent for stone buildings, and 5.1 per cent for frame buildings. Even for the same type of capital structure, the rate of depreciation varied from plant to plant and from time to time according to the care given to the structure by the management. For example, certain improvements introduced during the 1930’s in the methods of repairing and main­ taining locomotives were found to have greatly lengthened their lives. During the 1930’s, depreciation on railroads was commonly fixed at only 1 per cent of the capital. Accordingly, allocations for replace­ ments were limited to this amount. As a result, rails, sleepers, and superstructures during some of these years were not generally replaced; and the railroad properties deteriorated consider­ ably. During the 1930’s allowances for amorti­ zation were liberalized. In 1936, they amounted to 2 per cent of gross earnings for motor trucks, to 9 per cent for railroads and to an even higher ratio for water transportation. Outlays for automobile tires were charged entirely to depreciation; the life of the tires being fixed at 7,000 miles for those used on gravel roads and

36,000 for those used on cement roads. The annual rate of depreciation on motor vehicles was fixed at 15 per cent. Soviet statisticians contended, however, th at since railroad outlays for replacem ents during the 1930’s contained substantial net additions to capital, because superior m ate­ rials an d equipm ent had been used, the val­ uation o f transportation services was some­ w hat too low.

/. Income from Retail Trade Soviet estim ators em phasized the dynam ic role o f trad e in the economy. By facilitat­ ing exchanges am ong the enterprises them ­ selves, and betw een them and consum ers, trad e was said to intensify production, ac­ celerate capital accum ulation, and tie the agricultural and industrial sectors m ore closely together. T he term “trad e” was used to denote re­ tail trad e alm ost exclusively. W holesale trad e was n o t organized as a separate in­ dustrial division. E ach industrial enterprise, syndicate, o r corporation carried on its own w holesale trad e and accounted fo r its opera­ tions in the value of its own output. T he net value of the services of retail trade was the residual left after deducting certain item s of business expense from the realized value of the m arkup or gross profit o f the trading agencies as it was annually reported. T he deductions covered the costs of th e trading com m issariats fo r central ad­ m inistration, presum ably on th e ground that it constituted a governm ent activity, as well as fo r m aterials and services purchased from other enterprises, an d depreciation. T he values of th e trading activities o f the state and co-operative organizations were com puted separately. State organizations carried on about three tim es as m uch trad e during the 1930’s as the co-operatives (see T able 25-5).

6. D istribution and R edistribution of N ational Income In Soviet economics, five shares o r types of prim ary incom e were distinguished: 1. T h e incom e of the state, com prised in the m ain of the profits of its enterprises and of taxes and other social levies on eco­ nom ic undertakings. 2. Com pensation of em ployees o f stateow ned an d co-operative undertakings. 3. Profits and collective funds of the

25.

363

SOVIET RUSSIA

T able 25-5. N e t Value of the Services of Trade, 1936

Million Rubles (current prices)

Income from markup (on sales ag­ gregating 80 billion rubles for the state trading organizations and 36 billion rubles for the co­ operatives)

Percentage of Total

Total

State

Co-operative

Total

State

Co-operative

15,831

11,980

3,851

100.0

100.0

100.0

3,552

2,841

711

22.5

23.7

18.2

2,271 10,008

1,697 7,442

574 2,566

14.4 63.1

14.0 62.3

14.7 67.1

D eductions

Expenses for operation of trucks and wagons; rent; repairs; office supplies; amortization; loss, theft, and spoilage of goods; packing; advertising, display, trade conferences, etc. Payments to other industries for services rendered (transporta­ tion costs; travel; postage, tele­ graph and telephone; municipal charges for water, light, heat, etc.; protection of property, etc. Net value of trade services so u rc e :

Reference No. 2,

p.

270.

ko lkh o z and of the artisans’ and consum er (trad e) co-operatives. 4. W ages of ko lkh o z m em bers and a r­ tisans’ co-operatives. 5. Incom es of independent farm ers and artisans. These five shares w ere produced by five types of enterprises: (1 ) state-ow ned (agri­ cultural, industrial, construction, transpor­ tation, and trad in g ); (2 ) kolkhoz; (3 ) trade co-operatives; (4 ) artisans’ co-operatives; (5 ) independent farm ers and artisans. T he first and third shares represent the collective income of citizens; the other three, the private incom e of those citizens partici­ pating in production. E xcept fo r w hat was taken away by taxation and other deduc­ tions, private incom e was available fo r p ri­ vate spending and saving. a. In c o m e D istrib u te d b y S ta te -o w n e d E n terp rises T he incom e distributed by Soviet R us­ sia’s state-owned enterprises constituted the greater p art of national incom e, inasm uch as the state ow ned practically all industrial construction, transportation, and financial enterprises, the greater portion of trading organizations, and som e very substantial farm s. It was m uch broader in scope than the incom e distributed by state enterprises in capitalist economies. O ne p a rt w ent to

the state, the other to labor. T he form er consisted mainly of profits of the enter­ prises; net interest of state banks; turnover and certain other taxes; and payroll contri­ butions to social security and other welfare funds. Profits o f state-ow ned enterprises served a double objective: to provide new invest­ m ent capital fo r the expansion of produc­ tion and to provide funds for w orkers’ w elfare. T heir subsidiary function was to prom ote p ro p er business accounting and econom ic m anagem ent within the enter­ prises. O ne p a rt rem ained in the enterprise and was reinvested o r spent on w orkers’ w elfare; the oth er was transm itted to the state, w hich distributed it in the form of budgetary subsidies am ong industries oper­ ating at a loss or earning only small profits. These deductions, in addition to producing revenue fo r the state and providing means fo r the equalization of capital resources of enterprises, helped the state to enforce its controls over the enterprises and the use of their funds. T he profits were usually planned in advance as a specific percentage addition to business costs, the ratio varying from industry to industry depending upon the character of the industry or enterprise, its cost structure, and the price policy of the governm ent. A high ratio did not always m ean a high rate of profit, or a low ratio

THE INCOM E OF NATIONS

364

a low rate; for in an industry employing little or no raw m aterials o r only inexpensive raw materials, the ratio, all other conditions being the same, w ould necessarily be higher than in one using m any raw materials, or expensive kinds of raw m aterials. T he general tendency of the Soviet eco­ nom ic authorities, according to their state­ ments, was, as far as possible, to fix prices high enough to cover costs and to yield sufficient profits to the industries involved to cover their capital requirem ents and to enable them substantially to im prove their w orkers’ w elfare. H ow ever, fo r several in­ dustries prices could not be raised to such a level even during the m ore prosperous 1930’s, not to speak of the ’20’s. Conse­ quently, these industries, especially producer goods, were still fa r from self-supporting; e.g., the lum ber industry h ad a deficit of 10.6 per cent in 1937; the cem ent industry, 7.3 per cent; the alum inum and the building supply industries, varying deficits. M ost con-

collectively fo r investm ent an d capital accu­ m ulation purposes as well as fo r various social and other purposes designated in the state budget and econom ic plan; and a p o r­ tion to be distributed am ong the producers as th eir private incom es to be spent on con­ sum ption or to be saved. A ccording to Soviet w riters, the turnover tax represented the state’s profit from socialist production to be used fo r the collective benefit of the com ­ m unity.30 T he tax functioned as an instrum ent of over-all control over production and the use of funds by enterprises. It w as also a price regulator helping to keep supply and dem and in line.31 T he rate during the 1930’s varied from 1 per cent of the final price (o f w hich it is a p a rt) of some products to m ore th an 90 per cent o f others. W here the tax am ounted to 90 per cent o f the final retail price, it really increased th e price of the good to the consum ers tenfold. Its rapid grow th and w ider application after the tax

Table 25-6. Receipts from the Turnover Tax (million rubles) Total receipts Receipts from industry Percentage receipts from industry are of total so u rc e :

Reference No. 2,

1931 11,672 9,349

1932 20,514 15,445

1933 27,060 19,751

1934 37,596 21,969

1935 52,167 29,194

1936 65,762 39,249

1937 76,795 46,819

80.1

75.3

73.0

58.4

56.0

59.7

61.0

p.

85.

sum er goods industries, on the contrary, had surpluses. Interest represented the value of the serv­ ices of finance (the services of personnel, profits, etc.) and, as already stated, was com puted as the difference betw een interest earned and interest paid. T he turnover tax was the state’s principal fiscal source as well as an im portant price regulator and instrum ent of general eco­ nomic control. O rdinarily, it financed al­ m ost 80 per cent of the national budget (77.4 per cent in 1935, 78.4 in 1937), was paid by all state and co-operative enter­ prises th a t sold their ow n products, and was levied at only one stage. It was used first to insure a supply of new capital to the various sectors of the national economy; second, to provide fo r th e costs of general adm inistration, national defense, education, and other social and cultural needs. T hrough it the state divided the total social product, or national incom e, into a portion to be used

reform of 1930 can readily be seen in Table 25-6. O ther taxes and the paym ents to or on account of labor were treated as follow s: Before 1930, state-owned industries as well as other types of undertakings (not to mention individuals) were subject to the income tax. After the tax reform of that year state enter­ prises were exempted and became subject to the turnover tax instead. But the income tax, as a business or production tax, continued to be imposed on co-operative organizations, handi­ craft industries, and individual enterprises. The land rent tax was levied at specific rates per square meter of land, depending on how the community was classified as well as on how the land was used. Communities were classified by such factors as size, social and occupational characteristics of the population, degree of de­ velopment of industry and commerce, location with reference to transportation, administra­ tive importance, and other factors determining the rentability of the land. This tax must be distinguished from the rents charged by local authorities for the exploitation of peat lands and for the collection of fuel wood in munici­

2 5 . SOVIET RUSSIA

pally controlled forests. The tax on buildings was laid at a fixed percentage of the cost of buildings or of their insurance valuation. The rates were somewhat higher for industrial than for residential buildings. They varied also with the type of industrial (or residential) occupant. The fishing levy, applied to the socialized fish­ ing agencies, was based on the catch or number of aquatic animals killed. The basic pay of employees of state-owned enterprises consisted of compensation, at pre­ scribed rates, for regular and overtime work, as well as of bonuses for especially hazardous, strenuous, or unhealthful work, long service, possession of certain qualifications, and the performance of certain civic duties; bonuses for the fulfillment by the entire enterprise of a des­ ignated production program, the reduction of waste, breakage, and other defects in produc­ tion, and the prolongation of the life of the machinery; and premiums for efficiency and extra pay to experienced workers for instruct­ ing apprentices. Supplementary pay of employees of stateowned enterprises was not based directly on the nature of their jobs. It was pay given for time off granted minors for school attendance, rest, and recreation, time lost by mothers in nursing their babies, and time taken by shop leaders and other workers to attend to party activities or civic duties; also vacation pay, allowances for lodgings, reimbursement for mu­ nicipal taxes and other charges to workers not living in factory houses, street-car or railway commutation, and costs of various services and goods furnished workers gratis, such as meals in factory restaurants or laundering of workers’ personal effects. The latter services or goods were distinguished from those furnished work­ ers for use in the course of their work, such as working clothes or aprons, soap, towels, foods neutralizing the effects of industrial poi­ soning, duty laundry, etc., which were treated as ordinary expenses of production deductible from gross value. In 1935, the supplementary pay in money or in kind averaged approxi­ mately 10 per cent of the basic pay. The payroll contributions by state-owned en­ terprises for the collective benefit of workers were made either in cash or in kind, and con­ sisted chiefly of contributions to the centralized state social security fund, the industry-wide housing fund, and to the enterprises’ own wel­ fare institutions. The social security contribu­ tions were fixed at varying percentages of the payroll, depending upon the character of the industry or trade. For example, for artists they were fixed in 1936 at 3.7 per cent; for certain categories of chemical workers they were as high as 107 per cent. Most of the contributions by enterprises to their own welfare activities were consolidated in 1936 into the “Director’s Fund,” fixed at 4 per cent of the authorized or planned profit and at 50 per cent of any excess profit. At least

365

one half of the income from the fund was to be expended each year on housing for workers. The rest could be spent, under the authority of the Director, for cultural or other needs of the workers, such as clinics, clubs, libraries, cafe­ terias, and laundries, as well as to improve operation and reward special efficiency. Before 1936, each enterprise provided for the main­ tenance of its own factory schools, nurseries, and kindergartens by a special contribution of one quarter of 1 per cent of its payroll. It pro­ vided also for the dissemination of its own technical information among workers and train­ ing specialists as well as rank and file workers by means of special contributions fixed as a certain percentage of its payroll. During that year, however, the responsibility for financing schools and nurseries and some of the costs of technical training were transferred to the state budget. In 1935 all these contributions to social security and workers’ welfare funds and for workers’ training together averaged, for all in­ dustries combined, approximately 20 per cent of the basic and supplementary pay (see Table 25-3).

b. Income Distributed by Kolkhoz and Other Enterprises T he incom e produced by the kolkhoz was distributed in the form of paym ents to state tracto r stations, largely in kind; a 10-15 p er cent reserve of seeds and feed set aside against crop failures; additions to the capital fund, am ounting usually to 10-20 per cent o f the m oney incom e and used to acquire agricultural m achinery, building materials, and other m eans of production; adm inistra­ tive expenses, lim ited as a rule to 2 per cent of the total income; funds fo r cultural p u r­ poses, such as the acquisition of radios and the training of leaders; relief fund fo r the benefit of disabled workers, the aged and sick, and soldiers’ dependents, and to m ain­ tain nursery schools and care fo r orphans (likewise to a large extent in k in d ); wage paym ents to ko lkh o z m em bers based on the num ber of days of lab o r perform ed; and the agricultural incom e tax levied on gross m onetary incom e of the ko lkh o z■ This tax was levied to enforce control over the finan­ cial operations of th e ko lkhoz as m uch as to raise revenue. In addition, ko lkhoz m em ­ bers received incom e from their own farm s and household occupations. Incom e p ro ­ duced by other enterprises was distributed as follows: The income produced by the artisans’ co­ operatives, unlike that produced by the kolkhoz, was distributed mainly in cash. The major part was divided into wages of co-operative mem­

366

THE INCOM E OF NATIONS

bers, fixed on a piece rate basis and in accord­ ance with the grade of work done, and profits. The latter were divided, in turn, into income tax payments to the state, additions to the capital fund, amounting usually to 30 per cent, and the sums allocated to the improvement of the housing and living conditions of the mem­ bers, as well as to any other purposes approved by the general assembly of the co-operative. The income produced by trade co-operatives was distributed in the form of wages and profits —from which the income tax was paid. Little explanation was given in the official documents as to the distribution of the income produced by independent farmers and artisans, a small item.

c. Redistribution of National Income The foregoing “prim ary” incomes dis­ tributed am ong participants in m aterial pro ­ duction w ere deem ed to be redistributed in part as various types of “derivative” income to citizens not participating in m aterial p ro ­ duction. T he redistribution was being ef­ fected, as the Kolganoff study pointed out, through the channels of governm ent finance, the sale of personal services in the m arket,

and the activities of voluntary social organi­ zations.32 A substantial p art of th e so-called p ri­ m ary incom e was redistributed through gov­ ernm ent finance channels, nam ely through the levy of graduated incom e taxes and cultural and housing levies on individuals (ku l-zh il-sb o r); the issue o f governm ent loans; an d governm ent paym ents fo r com ­ pensation and m aintenance of its civilian and m ilitary personnels, stipends to students, pensions, social security benefits, interest on governm ent loans, prizes in governm ent lotteries, insurance benefits, and interest on deposits in state banks. D erivative income fro m private personal service (literary, artistic, medical, domestic, e tc .), played a relatively m in o r role in the Soviet economy. Its origin was n o t developed in the estimates, n o r was m u ch inform ation given on the “derivative” incom e o f voluntary social or­ ganizations (C om m unist party, trad e unions, and w orkers’ clubs) and their redistribution to th eir paid w orkers and m em bers. Despite

Table 25-7. National Income, U.S.S.R., by Recipient Groups Prior to Complete Socialization, 1928-31 Million Rubles (current prices)

Percentages of Total

1930

1931

1928

1931

16,604 13,094 7,113 5,067 914

19,487 15,752 8,848 5,819 1,085

39.9 28.6 14.7 11.6 2.3

1929 42.2 32.2 16.7 12.9 2.6

1930

11,934 8,563 4,406 3,467 690

1929 13,486 10,285 5,325 4,122 838

40.3 31.8 17.3 12.3 2.2

38.6 31.2 17.5 11.5 2.1

Artisans and handicraftsmen In co-operatives Not in co-operatives

1,807 562 1,245

1,916 771 1,145

2,192 1,113 1,079

2,590 1,669 921

6.0 1.9 4.2

6.0 2.4 3.6

5.3 2.7 2.6

5.1 3.3 1.8

Entrepreneurs Industrial Small merchants Other

1,102 150 652 300

794 89 489 216

792 79 555 158

588 52 406 130

3.7 0.5 2.2 1.0

2.5 0.3 1.5 0.7

1.9 0.2 1.3 0.4

1.2 0.1 0.8 0.3

462

491

526

557

1.5

1.5

1.3

1.1

B. Agricultural Workers In socialized sector In nonsocialized sector

15,502 536 242 294

14,999 625 343 282

19,594 844 657 187

22,183 1,307 1,195 112

51.8 1.8 0.8 1.0

47.0 2.0 1.1 0.9

47.6 2.1 1.6 0.5

43.9 2.6 2.4 0.2

Kolkhoz members Independent producers Kulaks (larger proprietors)

260 13,306 1,400

614 12,460 1,300

4,921 13,329 500

10,432 10,154 290

0.9 44.5 4.7

1.9 39.0 4.1

12.0 32.4 1.2

20.7 20.1 0.6

2,462

3,454

4,967

8,823

8.2

10.8

12.1

17.5

29,898

31,939

41,165

50,493

100.0

100.0

100.0

100.0

1928 A. Nonagricultural Factory workers and employees Factory workers Employees Others

All others

C. Profits of socialized sector Total national income s o u r c e : Reference N o. 1, p . 99.

25.

SOVIET RUSSIA

their im portance, derivative incom es were only sketchily indicated. d . R e c o n c ilia tio n o f D istrib u tio n a n d R e d istrib u tio n o f In c o m e If one w ere to add together all prim ary an d derivative individual incom es as well as the collective incom e, the sum w ould ex­ ceed the com puted national incom e. T he difference w ould be due to the double counting of the derivative incom es— once as such and once as p a rt of the prim ary incom es.33 So far, the relation betw een the distribu­ tion and redistribution of national incom e has n o t been dem onstrated statistically. T he last published statistics on the distribution of national incom e w ere fo r the early 1930’s and did not bear an official stam p (see T able 2 5 -7 ). In none of the official studies quoted was there a statistical table of the redistribu­ tion fo r any year.

7. Com putation of N ational

Incom e at Constant Prices T he conversion of the national incom e in current prices into national incom e in constant prices was m ade in a unique m an­ ner. Special indexes o f wholesale o r retail prices o r both w ere constructed fo r each b ranch of production and these w ere used to deflate the net outp u t in each branch. T he general objective seem ed to be to at­ tain as nearly as possible a physical m easure o f current net output independent of price changes. In the pursuit of this objective, the m ethods em ployed produced widely diver­ gent and frequently peculiar results.34 A ltogether, three m ain m ethods of con­ version into constant prices w ere em ployed: one know n as the “dynam ics of gross out­ p u t” ; another, called the “index of the price o f the product” ; and the third, described as the “direct com putation o f net o utput in constant prices.” In all three m ethods, the constant prices w ere taken as of the year 1926-27. a. T h e “D yn a m ic s o f G ro ss O u tp u t” M e th o d U n d er the m ethod of the dynam ics o f gross output, the increase in the net value of the output in constant prices was m eas­ ured by the increase in gross value in such prices. The latter was com puted by m ul­

367

tiplying th e cu rren t quantity of the output by the price of th e u n it of th a t output in the base year; after this was done, the ratio of net to gross value effective in the base year was applied to the figure. T he resulting sum was taken to represent the net value of the current output in constant prices. F o r example, if 10,000,000 units are p ro ­ duced in the x year, the price of the unit in the base year is 10 rubles, and th e ratio of net to gross value of th a t p roduct in the base year was 60 p er cent, then the gross value fo r the cu rren t year in constant prices is set at 100,000,000 rubles, and the net value in constant prices at 60,000,000 rubles. This m ethod was first introduced in the national incom e com putation for the years 1925-30. It was still used in the 1940’s in the case of certain branches of production, such as freight transport, retail trade (in p a r t) , and others. In the case of transport, it was applied in a simplified form . T he net value of the services perform ed in carrying a ton of freight fo r the distance of a kilom eter in 1926-27 was taken as a start­ ing point. This sum was th en multiplied by the num ber of ton-kilom eters of service p er­ form ed in the cu rren t year, and th e result was deem ed to represent the net value of the service or incom e of freight transport in the cu rren t year at constant prices (see C hapter 16, Section l c ) . b. T h e “I n d e x o f the P rice o f th e P ro d ­ u c t” M e th o d U nder the second m ethod, index o f the price o f the product, the n et value of the outpu t was com puted first in current prices. It was next deflated by the price index of the finished product. F o r example, if the price of the p roduct doubled, the n et value of th e o u tp u t in the course of its conversion into constant prices w ould be cut in half. T he m ethod of deflation could give correct results only if the ratio of n et to gross value of the given ou tp u t rem ained the same. But this was n o t the usual case. As a rule, changes in prices affected the net-gross ratio of ou tp u t values. This m ust have occurred frequently in Soviet Russia at th a t time, w ith the result th a t this type of conversion could n ot be correct. This m ethod was used as an alternate m ethod in preparing the 1925-30 estimates, b u t no clear statem ent was m ade as to the branches of production in w hich it was applied.

368

THE INCOM E OF NATIONS

c. D irect C o m p u ta tio n o f N e t V a lu e in C o n sta n t P rices U nder the third m ethod, o r the direct com putation o f net value in constant prices, the net value was com puted as the difference between the gross value of the output in base year prices and the costs of materials in base year prices. This m ethod was used extensively in the com putation of income from agriculture and apparently in certain other cases as well. d. D isto rtio n s U nder A ll T h re e M e th o d s All three m ethods o f estim ating net out­ put in constant prices led to a distorted pic­ ture of the national incom e according to its origin in the various branches of production and, probably, to an overestim ation of its aggregate to an unknow n degree.35 M ore­ over, certain departures from the applica­ tion of constant prices were even m ore re ­ sponsible fo r a possible exaggeration of the national incom e expressed in these prices. Several Soviet economists have pointed out that in im portant instances the estimates rested n o t upon 1926-27 prices, but upon the generally higher prices of the current year, o r of some other later year than 192627. This was the case fo r building construc­ tion, a p art of retail trade, and those seg­ m ents of industry th a t produced articles not produced in 1926-27. In the building industry, base year prices scarcely existed. E ach new construction project had its peculiar features, and hence costs. Accordingly, conversion of net out­ put into constant prices u nder any of the m ethods described could hardly be done in its case.36 T he net output fo r that branch of production was estim ated, therefore, as the difference between the gross value of the construction w ork done at current prices and the m aterial costs likewise incurred at current prices. T he income from retail trade was com ­ puted partly at current and partly at con­ stant prices. First, the gross value of goods entering trade was com puted at current prices. N ext, by applying the m arkup in effect in 1926-27 the gross value of the services of retail trade was com puted. This was next reduced to net value, by applying the ratio of net to gross value in trade in effect in 1926-27. Finally, the excise taxes levied on retail sales were added at the then existing percentage rates. This m ethod m ust have exaggerated the net output in trade in

tw o ways. F irst of all, it took the wholesale prices of the cu rren t year, w hich w ere gen­ erally m uch higher th an those of 1926-27, as a starting point. T hey w ere generally higher n o t only because of th e higher wage and price level b u t also because they in ­ cluded packing, transportation, and delivery costs, w hereas the early prices were strictly factory prices. Secondly, the 1926-27 ratio of n et to gross value was generally higher th an the cu rren t ratio because n et value in 1926-27 included the excise taxes then levied on retail trade. One of the greatest sources of the over­ estim ation of th e Soviet national incom e probably was the m ethod of com puting the net o u tput in constant prices fo r industries producing articles n o t produced in 1926-27. In this case, prices had to be used fo r the years in w hich the articles w ere first placed on the m arket. N o t only did these prices reflect the generally higher wage costs of those later years, b u t they often reflected the higher initial production costs of those articles. E ven though the prices and costs o f these products w ere m aterially reduced, as the goods w ere p u t into mass produc­ tion, their net value continued to be evalu­ ated in their original high prices and costs, thus producing a perpetually exaggerated figure of n et value o r income (Tables 25-8 and 25-9). A review of the m ethods of com puting the Soviet national incom e in constant prices m akes it plain that, in one w ay or another, they all unw ittingly tended to dis­ tort, and largely to exaggerate, the growth of the real national incom e of the country and of its several com ponent parts. Some of the indicated distortions, such as those in retail trade, could probably be easily corrected, b u t some w ere inherent in the rap id change of th e econom y and p ro b ­ ably could n ot have been avoided no m atter w hat system of price devaluation was used. T he leading Soviet statisticians of the late 1930’s seem ed to appreciate th e im perfec­ tions in the m ethods of price deflation they used in com puting the national incom e, b ut they did n o t reach any agreem ent as to how these im perfections might be corrected. One authority saw a solution in a shift to the estim ating of national incom e in term s of evaluating consum er goods and new ly added capital goods in constant prices. H e stopped short o f recom m ending it, how ever, because in his opinion it w ould m ake it difficult to break dow n the national incom e total into

25.

SOVIET RUSSIA

369

Table 25-8. National Income, U.S.S.R., by Divisions of Economic Activity, 1913-37

Agri­ culture

1913 1928 1930 1931 1932 1933 1934 1935 1936 1937

9.1 9.4 8.7 8.6 8.4 9.2 9.4 9.6 14.9

Industry (Including Lumber and Fishing)

6.7 8.7 14.8 17.4 20.6 23.0 27.9 34.8 46.1 50.9

Construction

Transpor­ tation

Trade (Including Communal Restaurants)

Billion Rubles (1926-27 prices ■) 1.2 1.2 1.5 1.6 3.8 2.0 6.2 2.1 6.2 2.5 5.6 2.6 6.7 3.0 8.7 3.8 12.7 4.7 12.5 5.2

Miscel­ laneous

Total

1.8 3.0 4.4 5.4 6.8 7.0 7.8 8.6°

1.0 0.8 1.3 1.2 1.0 1.1 1.0 1 .0 ”

11.8°

1 .0 “

8.6 12.0 12.6 13.2 15.0 14.5 13.9 12.9

4.8 3.2 3.7 2.9 2.2 2.2 1.9 1.5

100 100 100 100 100 100 100 100

12.2

1.0

100

8.1 13.7 10.0 9.5

0.6 0.4 0.4 2.9

100 100 100 100

21.0 25.0 35.0 40.9 45.5 48.5 55.8 66.5 86.0 96.3

Percentages of Total (1926-27 prices b) 1913 1928 1930 1931 1932 1933 1934 1935 1936 1937

43.3 37.6 24.9 21.0 18.4 18.9 16.8 14.5 15.4

31.9 34.8 42.2 42.6 45.3 47.4 50.0 52.1 54.6 53.5

1913 .925-26 1929 1937

48.0 41.4 35.1 25.7

34.9 35.6 41.7 53.1

5.7 6.0 10.9 15.2 13.6 11.5 12.0 13.2 14.8 12.5

5.7 6.4 5.7 5.1 5.5 5.5 5.4 5.8 4.2 5.4

Percentages of Total (current prices d) 4.1 4.3 5.4 3.5 6.1 6.7 5.8 3.0

a Agriculture of U.S.S.R., 1935, Yearbook, p. 190; Socialist Construction in U.S.S.R., Yearbook 1936, p. 30; U.S.S.R., Country of Socialism, 1936, pp. 7-10. b Reference No. 3, pp. 63, 65, and 77. 0 Approximate figures. d Reference No. 3, pp. 63, 65, and 77. p arts contributed by the different branches o f production. A nother authority suggested the conversion of national incom e at cu r­ ren t prices into national incom e at constant prices with the aid of a unified price index, the preparation of w hich, he said, should be resum ed at once.87

8. O verstatem ents D ue to Other than M ethodological Factors O n top of m ethodological factors, changes in the structure of the Soviet econom y con­ tributed to overstatem ents of the grow th of the Soviet national incom e. M uch of the reported increase in th a t incom e m erely represented a m ore com plete accounting of production previously carried on in the

farm ers’ households o r otherwise. I t was, therefore, only a paper increase. F o r ex­ am ple, in the course o f the breakdow n of the old ru ral household economy, a consid­ erable am ount of production carried on by m em bers of the fam ily at hom e fo r fam ily consum ption was taken over by the kolkhoz or was organized and classified as industrial production, and the w ork of millions of w orkers n o t previously accounted fo r in n a­ tional incom e suddenly was recorded. W ork­ ers’ meals taken at the fam ily table were replaced to an increasing degree by meals taken at the factory canteens or com m unal restaurants, and the services rendered in supplying them becam e, u nder the category of retail trade, a p art o f the national in­ come.

THE INCOM E OF NATIONS

370

Table 25-9. National Income, Total and Per Capita, U.S.S.R., 1913-52

National Income (1926-27 prices) Population (000) * 1913 1917 1920 1921 1924-25 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1949 1950 1951 1952 1953 1954 1955 1956

140,000 131,000 143,400 146,000 148,570 151,389 153,846 155,810 157,397 158,131 158,662 159,603 160,660 162,330 165,124 168,587 170,467 175,000

Total (million rubles b) 21,000 16,800 10,500 9,400 16,800 21,700 23,000 25,000 28,900 35,000 40,900 45,500 48,500 55,800 66,500 86,000“ 96,300 105,000 115,000 125,500 177,000d

Indexes of Growth 1929: 100 73 58 36 33 58 75 80 87 100 121 142 157 168 193 230 298“ 333 363 398 434

1913: 100d 100 71 40 38 —

103 110 119 138 167 195 217 231 266 317 410 459 500 —

611 834 1,003 1,126 1,249 1,367“ 1,533“ 1,716“ 1,922“

Per Capita (rubles) 150 80 117 149 155 165 188 225 260 288 306 350 414 530 583 623 675 721

a All population figures are as of the middle of the year. Those for 1913 and 1920 are from Statistisches Handbuch der Weltwirtschaft, Statistisches Reichsamt, 1936, p. 251. The censuses of December, 1926 and January, 1939 reported a population of 147,028 million and 170,467 million, respectively (see League of Nations, Statistical Year Book, 1932-33, p. 21; ibid., 1940-41, p. 16). The population for 1926 was roughly estimated on the basis of the census data; figures for 1927-38 are derived from estimates prepared by Frank Lorimer in his book, The Population of the Soviet Union: History and Prospects, 1946, pp. 30 and 135. The population for 1940 is a rough estimate. b Reference No. 3, p. 53. The figure for the national income for 1939 is arbitrarily fixed as the middle between the reported 1938 and 1940 figures. The figure for 1940 is from the report of N. Vosnesensky, Chairman of Gosplan, Economic Plan for 1941 (18th Congress of the Communist Party, February, 1941). “The reason for the great jump in the figure from 1935 to 1936 is not clear. It may have been due to some unexplained change in the method of computation of the 1936 national income. d Reference No. 6, p. 302. e Reference No. 12, p. 42. The em ploym ent of the m aterial produc­ tion concept of national incom e also tended to exaggerate the rate of grow th of the Soviet national income. T he exaggeration resulted from the fact that the output of the m aterial production sector of the econ­ omy, w ith w hich national incom e coincides under this concept, increased in Soviet Russia m ore rapidly than the output of the entire economy, i.e., than the national in­ come conceived in broader term s. It m ust be rem em bered that the m aterial production

sector was being expanded in Soviet Russia at the expense of all other sectors. T he extraordinary em phasis on capital accum u­ lation in Soviet econom ic policy in peace­ tim e was responsible fo r that. T he rate of grow th of production of the com bined m a­ terial and im m aterial sectors of the economy, i.e., of the national incom e m ore broadly conceived, therefore, m ust have been con­ siderably low er th an the rate of grow th of its m aterial production sector alone. In m ost capitalist economies, by contrast, the

2 5 . SOVIET RUSSIA

services sector norm ally increased m ore rapidly than the m aterial production sector. M uch of the reported rapid grow th of the Soviet national incom e (confined to m a­ terial production) m erely represented a transfer of econom ic activity from one sec­ tor of the econom y to another, rath er than a n et increase in the activity of th e econ­ om y taken as a whole. N onetheless, the in­ crease was unquestionably very rapid. Julius W yler sought to identify the m is­ leading elements in the Soviet estim ates for the period of the 1930’s (Table 25-9 above) as fo llo w s:38 The figures showing a sharp decline in the country’s national income during World War I and the subsequent years of revolution and military communism and equally sharp recovery of it during the years 1924-1927 seems to con­ form with reality. Similarly plausible appears the estimated annual increase in the national income for the period 1937-1940, of approxi­ mately 10 per cent. But the alleged five-fold rise in the national income for the period 1928— 1940 is wholly unbelievable. It would have meant that within the brief compass of 12 years, Russia rose from deepest poverty to the position of one of the wealthiest nations of the world, for it would have placed Soviet Russia’s per capita national income in 1940 only 25 per cent below that of the United States. All known observations and numerous statistical data con­ trovert completely any such relationship be­ tween these two national incomes.

9. Colin Clark’s R ecom putations of the Soviet N ational Income Colin C lark’s estimates o f the Russian national income published in 1939 rep re­ sented the first system atic attem pt to de­ velop independent estim ates fo r th a t country that would be com parable w ith those de­ veloped fo r the U nited K ingdom and other countries. His estim ates covered th e years 1913, 1928, 1934, and 1937 and shed an interesting sidelight on some aspects of the Soviet econom y.89 C lark’s objective was to convert th e Soviet national income in current rubles into pounds at prices prevailing in 1934. N eces­ sarily, he had to use as his basic concept the n et national product o r expenditure at m arket prices. T o the figures of incom e derived from m aterial production, to w hich the Soviet estimates w ere restricted, he added the value of the services o f rented dwellings and farm ers’ hom es as well as the

371

value of th e services o f governm ent, p ro ­ fessions, domestics, and other independent workers. C lark’s estim ates were based m ainly on statistical yearbooks, international statistical abstracts, and various studies on Russian econom ic conditions published in English or G erm an. F o r his revaluation of the Russian n a­ tional incom e in British prices, Clark first com puted the value of consum ption and in­ vestm ent in current rubles, employing m ethods of estim ation for the years 1913 and 1928 different from those for 1934 and 1937. F o r the earlier years, he m ade use of Prokopovitch’s estim ates of the net value of m aterial production, adding to them his valuations of the om itted services, indirect taxes, and depreciation. N ext, he broke up this gross national p roduct at m arket prices into such com ponent parts as food consum p­ tion, expenditures for rent, consum ption of other goods and services, and gross invest­ m ent. F o r the years 1934 and 1937, how ­ ever, C lark did n ot have the benefit of any such initial estimates, and he himself had to com pute the values of the various in­ dividual types of consum ption and invest­ m ent in cu rren t rubles on the basis of in­ come, sales, budgetary and production data. In doing so, he distinguished between agri­ cultural and u rb an populations, and listed governm ent expenditures separately. C lark’s re-evaluations of the Soviet n a­ tional incom e in British prices were round­ about and som ew hat arbitrary and were susceptible of errors. H is results were strik­ ingly below the Soviet estimates. Thus, for example, fo r 1934, C lark showed a per capita incom e in constant prices th a t was about the same as th at in 1918 and in 1928, while fo r 1937 he showed one only 40 per cent higher than in those years. The corre­ sponding per capita figures published by the Soviet governm ent for 1934 were m ore than 100 per cent higher, and for 1937 about 300 per cent higher th an in 1918 and 1928 (see T able 25-10). In Conditions o f E conom ic Progress (3d ed., 19 5 7 ), C lark makes another re-evalua­ tion of Soviet national income; this tim e in the International U nits (I.U .’s) he devel­ oped (see C hapter 16, Section 3 b ). H e uses new data and extends his calculations to 1953, arriving at the following annual per­ centage rates of grow th of the Soviet na­ tional incom e per m an-hour of labor: 0.7

THE IN CO M E OF NATIONS

372

Table 25-10. National Income, U.S.S.R., 1913, 1928, 1934, 1937 Estimates by Colin Clark

National Income a t British Market Prices in 1934

1913 1928 1934 1937

Population (millions a) 140.0 150.0 159.6 165.1

Pounds (millions) 2,803 2,840 3,299 4,637

1928: 100 98 100 116 163

Per Capita (pounds a) 20.0 19.0 20.6 28.1

a Clark’s population figures were replaced by those given in Table 25-9, and his per capita income figures have accordingly been recomputed. so u rc e :

Reference No. 8a, pp. 41, 47.

in 1913-28; 1.6 in 1928-38; and 2.0 in 1938-53. H e concluded: “Though acceler­ ating, the rate of growth is still below that of a good m any other countries” (p. 250).

10. Estim ates by W yler C lark’s figures were criticized by several English, U nited States, and other economists as being too low.40 T he extraordinary in ­ dustrial preparedness dem onstrated by So­ viet Russia at the outbreak of th e w ar was pointed out as confirming the inaccuracy of C lark’s estimates. A recent estim ate by Julius W yler m ade by the expenditure or final products m ethod 41 showed a doubling in the Soviet real national incom e from 1928 to 1940, w hich was just about a midway figure between the Soviet and C lark’s esti­ mates. In 1928, according to W yler’s com ­ putations, 80 p er cent of the Soviet national income was devoted to personal consum p­ tion, 7 p er cent to capital form ation, 3.6 per cent to national defense, and 9.1 per cent to other governm ent expenditures. In 1940, these proportions were found by him to be as follows: 42.1 per cent fo r personal consum ption, 17.1 per cent fo r capital fo r­ m ation (including outlays fo r the w ar in ­ dustries), 26.9 per cent for m ilitary expendi­ tures, and 13.3 per cent fo r other govern­ m ent expenditures. D espite the rapid in­ dustrialization of the country, the per capita consum ption, according to W yler’s esti­ mates, not only did not increase from 1928 to 1937 in Soviet Russia, b u t actually de­ clined. T he gain in per capita national in­ come fo r the period was realized entirely in capital form ation.42

11. Corrections by Bergson and A ssociates In 1953 an d 1954, P rofessor A bram Bergson of the Russian Institute of C olum ­ bia U niversity and H ans H eym ann an d Oleg H oeffding of the R and R esearch C orpora­ tion published three related w orks presenting a recalculation of the Soviet national incom e fo r 1928, 1937, and 1940-48, in “adjusted rubles” and “adjusted factor costs.” 43 The adjustm ent aim ed to elim inate some of the elements of the Soviet pricing system th at distort the calculations. M ore particularly, the adjustm ent involved the elim ination of the following n o nfactor costs th a t tend to increase o r to decrease the value o f the Soviet output: (a ) the turnover tax and o th er indirect taxes (included in Soviet n a­ tional incom e to ta ls); ( b ) subsidies to in­ dustry from the governm ent budget; and (c) planned profits, o r differentials in profit m argins as betw een p roducer and consum er industries th a t operate as subsidies in the one case and as indirect taxes in the other. T hey m ade no correction fo r price inflation n o r fo r any of the other distortions of the Soviet price indexes pointed o ut above. T he authors reached th e results of T able 25-11 which, a t best, represent only a very in ­ com plete correction of the Soviet estimates. Bergson and H eym ann’s revaluation p ro ­ duced substantial differences in th e alloca­ tion of the gross national p ro d u ct from those show n in the Soviet official figures. Thus, the share of household consum ption in the total gross national p ro d u ct in 1948 ap­ peared low er th an in th e Soviet estimates (45.2 per cent as against 58 per c e n t); the share of com m unal services, national civil

2 5 . SOVIET RUSSIA

Table 25-11. Gross National Product of U.S.S.R. According to Bergson-Heymann-Hoeffding44 (billion rubles, current values)

1928 1937 1940 1944 1948

In Established Prices

In Adjusted Prices

32.6 291.8 458.0 487.4 811.2

30.0 223.9 366.1 408.5 616.7

adm inistration, and defense appeared higher (29.2 p er cent as against 2 3 .5 ); as did the share of gross investm ent (25.6 per cent as against 18.5 p er c e n t).46

12. Prospects for the Official Preparation of a M oney-Flow Balance Sheet and of Social A ccounts In recent years Soviet statisticians have become aw are of the need fo r the construc­ tion of social accounts and a money-flow balance sheet em bracing all sectors o f the economy, i.e., taking in not only the tran s­ actions of the m aterial production sector, but those of the personal, institutional, and governm ent sectors also, and taking in not only the so-called “prim ary” distribution of the national incom e, b u t its redistribution as well. Articles on th e subject have ap­ peared in the leading econom ic journals, for example, A. S trukof’s in the Planovoie Khosiaistov, July, 1957, and N . M argolin’s in the August, 1957, issue of the sam e pe­ riodical. T he Course in E conom ic Statistics, edited by Professor A . I. P etro f (2 d ed., M oscow, 1954), devotes eight chapters to the structure of financial and m aterial bal­ ance sheet of the econom y and o f its vari­ ous com ponent social accounts. T he AllSoviet C onference of Statisticians, held in M oscow in June, 1957, devoted a consid­ erable portion of its m eetings to this sub­ ject, w ith V. Sobol, P rofessor Strum ilin, Professor Petrof, and V. K atz leading the discussion. F o r a sum m ary o f these m eet­

373

ings see Voprosy E konom iki, September, 1957. A ll these discussions, however, centered on the theoretical problem s involved in the construction of such accounts and on the uses to w hich they could be put. N one cited any figures, because no figures had been m ade available. T he CSU (C entral Statis­ tical Office) subm itted to the conference fo r discussion an outline of a proposed bal­ ance sheet, fisting the items to be covered, b u t giving no figures. Some of the discus­ sants deplored the absence of figures, con­ tending th a t w ithout them it was difficult to m ake a com petent appraisal o f the project. T he stum bling block in the way of prepa­ ration and publication of the proposed ac­ counts seemed to be the same th at h ad pre­ vented th e preparation and publication of com prehensive national incom e statistics over the past q u arter century, namely, the Soviet governm ent’s reluctance to m ake public com plete inform ation on the coun­ try ’s econom ic operations.

13. Summary T he official published estimates for Soviet Russia evidence th at considerable w ork by Soviet statisticians has been done in th e field; th a t a large body of statistics necessary for fair estim ation of the country’s national in­ com e are available in governm ent offices; and th a t the technical com petence of the Soviet statisticians is considerable. But the figures so fa r published are com pletely use­ less. It is to be hoped th a t the veil of m ys­ tery surrounding the structure and growth of the Soviet national income will be lifted in the years to com e, and th a t well-docu­ m ented figures will be released by the au­ thorities. T he estimates prepared by non-Soviet statisticians are founded on inadequate sta­ tistics and are highly speculative in nature. T hey are a poor substitute fo r the homeproduced, official, objectively prepared fig­ ures th a t are yet to come.

26. GERMANY*

References 1. Statistisches Reichsamt, Das Deutsche Volkseinkommen vor und nach dem Kriege, Einzelschriften zur Statistik des Deutschen Reichs, Berlin, 1932, No. 24. (Hereinafter referred to as “Volkseinkommen 1932”). 2. Institut fur Konjunkturforschung, Kapitalbildung und Investitionen in der Deutschen Volkswirtschaft 1924-1928, Vierteljahreshefte, Sonderheft, 22, Berlin. 3. P. Jostock, “Volkseinkommen und Volksvermogen,” Die Statistik in Deutschland, Festgabe fiir Fr. Zahn, ed. by F. Burgdorfer, 1940, pp. 766-74. 4. , Die Berechnung des Volkseinkommens und ihr Erkenntniswert, Stuttgart & Berlin, 1941 (“Computation of National Income and Its Use,” a commentary on the methodology and the changes made in it in 1932-40). 5. Statistisches Bundesamt, “Das Volkseinkommen und Sozialprodukt, 1936 and 1948,” Statistische Berichte, September 22, 1949, Wiesbaden. 6. Hildegarde Bartels, Otto Schorry, H. H. Raabe, and others, articles on national income and on its estimating methods in Wirtschaft und Statistik, April, 1951, Heft 4; June, 1951, Heft 6; October, 1951, Heft 10; January, 1952, Heft 1; May, 1952, Heft 5; August, 1952, Heft 8; February, 1953, Heft 2; April, 1953, Heft 4; April, 1955, Heft 4.

1. H istory N ational incom e estim ating in G erm any began in her individual states long before her political unification. In 1805, on the eve of Prussia’s fatal involvem ent in a w ar w ith N apoleon, an estim ate of her national income was published there by Leopold K rug, but the national incom e in that esti­ m ate was conceived rath er narrow ly as the net output of agriculture, including fishing and forestry, and incom e from the export of goods. In 1846, C. F. W . D ieterici p re­ pared an estim ate fo r Prussia based on the per capita consum ption of her population. In the 1860’s and ’70’s national income studies covering W iirttem berg, Prussia, and Saxony were m ade by the well-known statis­ ticians Rum elin, Engel, and Soetbeer, as well as by Samter. These investigations were based on income tax statistics and only oc­ casionally included an estim ate of the entire * Acknowledgment is due to Dr. P. Jostock for going over the draft of Part I of this chap­ ter and making many helpful suggestions for its amplifications; and to Dr. Otto Schorry of the West German statistical office for a similar helpful review of Part II. Notes begin on p. 541.

374

national incom e fo r these states. T his was tru e fo r m ost of the other incom e studies th a t appeared m ainly fo r th e state o f Prussia during the succeeding two o r three decades. A q u arter of a century after the founda­ tion of the G erm an E m pire (1 8 7 1 ), the first elaborate estim ates of the G erm an n a­ tional incom e w ere m ade. T hey w ere p re­ p ared by R. E . M ay for th e years 1895, 1900, and 1907. M ore com prehensive esti­ m ates fo r 1888-1913 were m ade by K arl H elfferich ju st before the o u tbreak o f W orld W ar I. T he fact th a t the estim ation of national incom e started later in G erm any th an in either E ngland o r F ran ce is som ew hat diffi­ cult to explain, inasm uch as econom ic th eo ry in th a t country had been concerned w ith the concept of individual an d national incom e fo r m any years and interest in sta­ tistics h ad developed there fairly early. The theoretical discussion dealt particularly with th e distinction betw een the national income as a social pro d u ct and national incom e as the sum of private incom es, some econo­ mists leaning tow ards the fo rm er concept (R um elin, Schall, R obert M eyer, and o th­ ers) w hile others favored the latter defini-

2 6 . GERM ANY

tion of it (H erm ann, Roscher, Schmoller, and o th ers). T he skeptical attitude tow ards the practical value of national incom e esti­ m ates that was characteristic of G erm an economics of the period, as discussed in C hapter 9, probably had som ething to do w ith this lag in the initiation of national income estim ates in G erm any. Some of the leading figures in G erm an econom ics— D iehl, fo r exam ple— even as late as the 1920’s contested the scientific significance of the concept.1 A fter W orld W ar I, the problem s of reparations and unification of the tax system created greater interest in national incom e estimates. T he T reaty o f Versailles stipu­ lated th a t G erm an reparations should be increased if the tax burden in the Allied countries becam e heavier than th a t in G er­ m any.2 F urtherm ore, the rapid increase of governm ental expenditures and taxes and the expansion of central governm ent and tax adm inistration led to internal contro­ versies about the im pact of the national tax burden on the distribution and grow th of national income, and on capital form ation. U nder these circum stances, estim ates of n a­ tional income becam e necessary. W hen m onetary stabilization in 1924 restored the possibility o f m aking com parable annual statistical com putations, national incom e es­ tim ates becam e especially num erous. T he basis of these new estim ates was greatly im proved because a com prehensive federal incom e tax had been introduced in the meanwhile. M ost of these estim ates w ere private in character (K uczinski, Elsass, Rogowski, K nopp, and o th e rs). In 1926-27 the Institute fo r Business Cycle Research (In stitu t fu r K onjun k tu rfo rsch u n g ), w hich was closely connected w ith the Statistical Office of the Reich (Statistisches R eichsam t) began the publication of periodic estimates of incomes from salaries and wages. In 1932, the Statistical Office of the Reich published a com prehensive volum e providing a con­ tinuous series of national incom e estimates for 1891-1913 and 1925—31.3 It continued this series annually th ereafter until 1938, m aking substantial m ethodological im prove­ m ents in the estimates from tim e to time. These estimates w ere prepared by the incom e-distributed m ethod. D uring W orld W ar II all publication of national income estimates was suspended. A fter the w ar the preparation of national incom e estimates was resum ed in W est G er­ m any u nder the auspices of its central sta­

375

tistical office (Statistisches B undesam t). These new estimates, in contrast to the old, were prepared by the n et output method. Since the territory of G erm any was now partitioned, these estimates concerned them ­ selves only w ith the territory of W est G er­ m any and Berlin. P art I of this chapter is concerned with the prew ar official estimates and P art II w ith the postw ar estimates.

I. Prewar Estimates by Incomedistributed. Method 1. Concept of National Income in the Prewar Estimates N ational incom e was considered in its three different aspects as (a) the net value of all goods and services produced, ( b ) the sum of incomes derived from production, and (c) the total value of current consum p­ tion and net capital form ation. However, it was n ot broken down by economic divi­ sions and was analyzed in detail only in the second and th ird phases. Such a breakdown by econom ic divisions was m ade only in an unofficial study published in 1933.4 N ational incom e was com puted from the sum of private incomes by deducting trans­ fer paym ents (w ith the exceptions noted below) and adding corporate savings, profits of governm ent enterprises, and em ployers’ social security contributions. The nonde­ ducted transfer paym ents included interest on civil debt; from 1925 to 1938 also in­ terest on w ar debt; and civil pensions, which w ere considered as deferred pay. Salaries of civil servants were taken net of pension contributions. M ilitary pay and m ainte­ nance w ere included. A djustm ent fo r net capital incom e from abroad was m ade, but reparation paym ents were n o t deducted from the hom e-produced income. These paym ents w ere considered as the p art of national incom e n ot available for domestic expenditure. Indirect taxes were included to the extent th at they were used together w ith direct taxes to finance governmental services to consum ers, governm ent capital form ation, and governm ent transfer pay­ ments. N ational incom e was accordingly com puted neither com pletely at factor cost nor com pletely at m arket prices, b u t on a m ixed basis. It was closer to factor costs th an to m arket prices. (F o r a similar con­

376

THE INCOM E OF NATIONS

ceptual treatm ent see C hapter 24 on the Swedish estim ates.)

2. Estimates for 1891-1913 Considerable effort was devoted in the 1920’s, to the preparation of estimates for 1913. They w ere deemed im portant for measuring the effects of the w ar and in­ flation on the national econom y and also as a basis for carrying the estim ates backw ard for some ten or m ore years in a consistent series. In preparing the estimates fo r 1913, m ain reliance was placed on the incom e tax sta­ tistics of the several states th a t had fairly well-developed incom e taxes, such as P rus­ sia, Saxony, W iirttem berg, and Baden. These state incom e taxes w ere assessed on the total incom e of individuals and corpora­ tions and were different from the later fed­ eral income tax, w hich was im posed on dif­ ferent types of incom e separately. T he esti­ mates developed for these four states were next blown up to a total representing the entire national territory. T he figures fo r 1891-1912 w ere devel­ oped from the Prussian income tax statistics for those years. First, the estim ates were prepared fo r those years fo r the territory of Prussia. N ext, these figures w ere m ul­ tiplied fo r each year by the ratio th e P rus­ sian national incom e bore to the national income of G erm any in 1913. T he assum p­ tion of an unchanging ratio for the entire period was obviously incorrect. However, there was no way in w hich changes in that ratio could be m easured.

3. Estimates for 1925-31 a. Statistical Sources T he estimates m ade for the years 1925-31 were based on uniform federal tax statistics and other rich supplem entary statistical m a­ terial, and were m uch m ore elaborate and accurate than those m ade fo r th e period before W orld W ar I. T able 26-1 illustrates, in 1928 figures, the m anner in w hich vari­ ous items w ere com bined to arrive at the various types of distributive shares and, finally, at a national incom e total. It can be seen th a t the items derived from the federal incom e tax statistics accounted for 64 per cent of the national income. T he federal tax was com posed of the following four schedules:

1. A withholding tax (Lohnsteuer) on sala­ ries, royalties, and pensions received by individuals in excess of the tax exemption 1,200 reichsmarks in 1925, other sums in later years. The tax exemption limit was deducted from each taxed income. Tax­ payers could claim a tax credit for the tax paid at the source on capital income, wages, and salaries. From each income 960 reichsmarks plus special exemptions could be deducted. In addition, special deductions were made for married per­ sons, dependents, costs incurred in earn­ ing the income, and interest on debt. 2. Income tax based on returns filed by the taxpayers (Veranlagte Einkommenssteuer), levied on all incomes exceeding 1,300 reichsmarks from independent ac­ tivities (agriculture, industry, trade, and professions), from rent and annuities. This tax was also levied on all incomes from labor and from capital if the total income exceeded 8,000 reichsmarks. 3. Tax on capital income (Kapitalertragssteuer). This tax was levied at the source on dividends and interest on securities. 4. Corporation income tax (Koerperschaftssteuer) levied upon the profits of corpo­ rations, limited liability companies, co­ operatives, mutual insurance institutions, autonomous estates of private law, and on enterprises and other entities of public law. In the original estimates the statistics of the first tax schedule were used to com pute th e bulk of wages, salaries, pensions. R ec­ ords of the second schedule also provided data on higher salaries, entrepreneurial in­ com e, and rent. T he fo u rth schedule was the basis of the com putation of incom e fig­ ures fo r lim ited liability com panies, of other juridical persons, and of corporate business savings. Statistics of the tax on capital in ­ com e w ere n o t em ployed because other sta­ tistical sources w ere available and because th e tax on bond interest was abolished in 1931. A fter 1931, fo r th e estim ate of the greater p a rt o f wages and salaries, social security statistics replaced those of th e first tax schedule. T hese social security statistics, showing the num ber of w eekly o r m onthly em ployees’ contributions, w ere applied back to 1929 in a revision of the first estimates. This new source was supplem ented by bud­ getary statistics relating to th e rem uneration of governm ental em ployees and workers. D ata on income from dividends w ere de­ rived from the statistics of corporations th at covered all corporations w ith capital stock o f one m illion reichsm arks and m ore, as

377

2 6 . GERM ANY

Table 26-1. Method of Estimating the National Income of Germany, 1928 (million reichsmarks)

Labor income Wages, salaries, pensions, etc., subject to withholding tax 32,282 Wages, salaries, and pensions taxed according to returns filed by the taxpayers (Table 26-2) 2,933 Wages, salaries, and pensions below tax exemption limit 8,872 Incomes from subsidiary occupations 645 Total wages, salaries, pensions 44,731 Employers’ social security contributions 2,250 Total labor income

46,981

Farmers’ and other entrepreneurs’ income Farmers’ income Income taxed according to returns filed by the taxpayers (Table 26-2) Allowances for family members working on the farm Allowances for understatement and evasion Gardening Incomes below tax exemption limit Total income from farming, forestry, and gardening Income of entrepreneurs in manufacturing, trade, and professions Income taxed according to returns filed by the taxpayers Allowances for understatement and evasion Income below tax exemption limit Total income of entrepreneurs in manufacturing, trade, and professions Total entrepreneurial income

"

2,045 1,250 1,050 500 1,021 5,816 9,936 1,990 261 12,187 18,003

Capital income Rent taxed according to returns filed by the taxpayers Dividends of corporations Dividends of limited liability companies Reported undistributed profits of corporations and other organizations Allowances for unreported undistributed profits (tax evasion) Bond interest Interest on savings deposits Interest on mortgages held by individuals Net income from governmentally owned enterprises, securities, and real estate Total Less income from intercorporate and governmental security holdings Less net capital income on securities from abroad

836 820 434 831 477 785 775 400 2,483 7,841 325 100

Total capital income Additions of taxes and fees

7,411 2,978

National income so urce

:

Reference No. 1, Volkseinkommen, 1932,

well as all corporations listed on G erm an stock exchanges. H ow ever, after 1931, the statistics of the corporation incom e tax were substituted fo r this source. Interest on securities and m ortgage bonds was com ­ puted from the reported statistics of the am ount of outstanding bonds, interest rates, an d certain types of interest paym ents. In ­ terest on savings deposits was estim ated on th e basis of statistics of com m ercial and sav­ ings banks. D ata concerning incom es of publicly owned enterprises, taxes, and social security benefits w ere derived from the F i­

75,373 pp.

31-38; 40-55; 60; 87-88.

nance Statistics o f the Reich (Reichsfinanzstatistik). T he 1907, 1925, and 1933 population censuses, unem ploym ent data, and wage statistics provided inform ation for the cal­ culation of incom es below the tax exemp­ tion limit. M any other statistical sources w ere used to estim ate m inor items.

b. Labor Income F o r 1926 and 1928, wages and salaries subject to the w ithholding tax were ascer­ tained from the tax statistics that showed

378

THE INCOM E OF NATIONS

their am ounts before tax exem ption. F or 1927 and 1929-31, wages and salaries were extrapolated on the basis of the proceeds of the withholding tax. This projection involved special difficulties, mainly because (1) tax proceeds included those of the wages and salaries subject to the tax under schedule 2; (2) the original tax rate (10 per cent) was changed after 1928; and (3) the proceeds referred only to the income exceeding the tax exemptions. The receipts from land re­ served to retired farmers (so-called “Altenteile”) were included with capital incomes or incomes of self-employed. To the wages and salaries subject to the withholding tax, those assessed on returns filed by the taxpayers, and those reported in the tax statistics were added. The exempt salaries and wages were estimated by multiplying the number of exempt recipi­ ents by an average yearly wage or salary. The number of such exempt recipients was deter­ mined first by extrapolating the yearly labor force reported in the 1925 Occupational Census and then by deducting the number of taxpayers in that labor force whose income was already accounted for; and, after 1928, by deducting the number of the unemployed as well. The residual number (the tax-exempt workers) were classified into the following groups: (a) agri­ cultural laborers; (b) domestic servants; (c) home workers; (d ) apprentices; (e) other juve­ niles below 18; (/) unskilled workers and re­ cipients of pensions. The average annual in­ come of each group was estimated on the basis of wage and employment statistics. Similarly, wages and salaries derived from odd jobs were estimated by multiplying the number of persons engaged in such odd occupations, as reported in the census, by a certain average income. The aggregate of social security contributions was added as supplemental labor income, and the corresponding benefit payments were deducted.

c. Agricultural Entrepreneurial Income: Farm Income Farmers’ incom e was being assessed under the tax law after deduction of in­ come of fam ily m em bers— grown-up chil­ dren and, in part, the aged— w orking on the farm . M oreover, the incom e reported took only partial allowance fo r the value of agri­ cultural products consum ed on the farm and fo r im puted rent of farm ers’ dwellings. The net product value of hom e gardening of in­ dustrial w orkers was not included at all. Allowances for these omissions were m ade, therefore, in the income estimates, and, in addition, the incomes below the tax exemp­ tion lim it w ere com puted. T he deductions for incom e of fam ily m em bers w orking on the farm were estim ated on the basis of the 1925

O ccupational Census and an estim ated aver­ age income of 750 reichsm arks. T he under­ statem ent an d evasion, difficult to estimate, were roughly estim ated at one billion reichs­ m arks ( R m ) .5 T he income below the tax exem ption lim it and th a t from hom e garden­ ing by industrial an d other w orkers were com puted on th e basis of 1925 O ccupational Census data on the num ber of persons en­ gaged in agriculture and an assum ed aver­ age incom e (1,000 reichsm arks in the first case, and 250 reichsm arks in the second). T o check the additions to farm incom e for the above factors, the value an d com posi­ tion of the net ou tp u t was roughly estim ated. T h e above-m entioned com putations w ere m ade only fo r 1928. Figures fo r the other years in the period 1925-29 w ere roughly estim ated by adding a constant am ount to assessed and tax-exem pt incom es. B ut for later years, they w ere apparently determ ined according to variations in the value of gross o u tput and farm ers’ expenditures.

d. Nonagricultural Entrepreneurial Income This item was show n in the yearly tax statistics (T able 26-3, taxes based on returns filed by the taxpayer) separately fo r farm ­ ers, entrepreneurs in m anufacturing and trade, and professionals. E ntrepreneurial incom e below the tax exem ption .lim it was estim ated in the same m anner as tax-exem pt wages and salaries. T h e average incom e was assum ed to be 1,000 reichsm arks fo r some years and other appropriate sums fo r others. A m ounts o f incom e assessed fo r ta x p u r­ poses were raised by stated percentages (20 in some years) to take account o f u nder­ valuation and evasion, partially offset by business losses n o t considered in the tax assessment. U pw ard revisions m ade later proved th a t these estimates w ere based on som ew hat shaky grounds.6

e. Capital Income T he items of capital incom e included in the estim ate are show n in T able 26-1. N et rents received by individuals w ere reported in the incom e tax statistics based on returns filed by individuals (schedule 2 ) . D ata on dividends were obtained from the statistics of corporate profits. U nreported dividends of sm aller corporations w ere om itted as being negligible. F o r the years 1925-28, dividends paid by lim ited liability companies were assum ed to be 65 per cent o f the net profits reported for the corporation income

2 6 . GERM ANY

tax (schedule 4 ) and fo r th e year 1929 to be 70 per cent. U ndistributed profits of corporations and lim ited liability com panies were calculated as percentages of the total n et profits re ­ ported for the corporation incom e tax (schedule 4 ). This ratio varied at first from 20 to 25 per cent fo r corporations,7 and from 30 to 35 per cent fo r lim ited liability com panies. By 1940 m uch higher rates were used (60 to 70 per cent fo r corporations). Finally, in order to account fo r tax evasion, a fu rth er addition of 10 to 20 per cent of the reported net profits was m ade. In de­ term ining the latter percentages, allowances were m ade fo r business losses. The other elements of national income were estimated as follows: The statistics of outstand­ ing public, industrial, and mortgage bonds pro­ vided the basis for the computation of bond interest. Interest paid on savings bank deposits was shown in current statistics. But with regard to co-operative savings banks and funds, only the amount of savings was known. No basis existed for separating individual from business deposits with commercial banks. The indi­ vidual deposits were estimated roughly at one third to one half of all domestic accounts, with government and foreign accounts excluded.8 The interest rate on savings in co-operative banks was assumed to be somewhat higher than that of savings banks, while the current interest rate on bank deposits was applied to the savings of individuals in commercial banks.9 The interest on mortgages held by individuals also had to be approximated. The amount of old preinflation mortgages was calculated by deducting from the total the amount of those mortgages held by institutions (banks, insur­ ance companies, social security institutions). However, with regard to the new postinflation mortgages, only the amount held by institutions was known. The holdings of individuals were roughly estimated by assuming that the rate of increase from 1924 to 1930 was half of that shown by the institutional holdings and was one quarter of it in the later years.10 The offi­ cially fixed interest rates were applied to the preinflation mortgages, while assumed rates were applied to the new mortgages.11 From the financial reports of the national, state, and local governments, figures were taken on the net profits of government-owned rail­ roads, post offices, telegraph and telephone offices; of public utilities generally; and on interest and dividends on security holdings (in­ cluding those of social security funds); and rent received from real estate. Adjustments were made for duplications of interest and dividends received by corporations and government, and for net capital income from abroad.12 These adjustments were based

379

on corporate balance sheets, financial statistics, and the balance of payments.

/. Taxes and Fees T he final com ponent of the G erm an n a­ tional incom e was the paym ent of taxes and fees insofar as they w ere used to finance final consum er services of government. T he determ ination of the am ount of taxes and fees used to finance such services in­ volved considerable difficulties, inasm uch as th e taxes w ere n o t earm arked fo r the financ­ ing of such final services, n o r were the ex­ penditures fo r such services clearly iden­ tifiable. T h e procedure follow ed in deter­ m ining the am ount of taxes to be added to the com putation of the national income (nam ely, indirect taxes, inasm uch as direct taxes were included in the factor shares), is shown in Table 26-2, and was explained m ore fully in C hapter 14, Section 2b. Since the com putation o f all consum er paym ents to the governm ent fell short by 2,978 m illion m arks of the value of the services and transfer incom e obtained from the governm ent, the difference m ust have been covered from indirect or business cost taxes. Obviously, business taxes m ust be equal to this tax difference plus the cost of the interm ediate services of th e governm ent.

4. Estimates for 1 9 3 2 -3 8 13 A fter 1931 the bases of the G erm an n a­ tional incom e estimates underw ent substan­ tial changes. H ow ever, the inform ation given about the new sources and m ethods used was by far less detailed and com plete th an the description in the official report fo r 1925—31. T h e changes mainly involved the estim ates o f salaries and wages, incom e of farm ers, and corporate profits: Salaries and wages after 1929 were estimated on the basis of the quarterly social security statistics. The data related to the number of weekly contributions of workers and monthly contributions of salaried employees, broken down according to the size of weekly wages or monthly salaries. The arithmetical average of the upper and lower limit of each income bracket multiplied by the corresponding number of contributors leads easily to the determination of the total income aggregate. The average in­ comes in the highest and in the lowest brackets, however, were computed partly from other sources and partly with the aid of the Pareto curve. To avoid duplications, the wages of in­ dependent workers, insured on a voluntary basis, for part of this period, were eliminated. To the

Table 26-2. Method of Estimating the Amount of Indirect Taxes to Be Included in the German Income Estimates, 1928-29 (million reichsmarks) Governmental Expenditures on Behalf of Consumers Services Education 2,685 General welfare 1,392 Health and housing 677 Total services 4,754 Reparations 2,159 Capital formation 2,519 Income transfers * 3,800 Total expenditures

Payments to Government by Consumers Taxes 3,722 Income taxes'1 451 Property taxes 139 Other taxes Total taxes 4,312 Fees and contributions 981 New domestic public debts 2,699 Government profits ° 2,262

13,232

10,254

Total payments Difference—to be included in national income

2,978 d

a War pensions, relief payments, and government contributions to unemployment insurance institutions. b Including corporate profit tax. 0 Including income from governmentally owned securities and real estate, and restitution to govern­ ment by industry. d Actually, the amount was 2,652 million reichsmarks, inasmuch as certain subsidies aggregating 326 million reichsmarks were deducted from this total. source:

Reference No. 1, Volkseinkommen, 1932,

pp.

56-59.

Table 26-3. National Income of Germany, 1913-38 a 1932

1938

64.9 45.2 78.3 57.7

6 8 .4 “ 82.1° 81.6 97.7

76.0

45.2 e

82.1 •

33.8

43.0

25.7

43.0

5.7

5.5

3.7

6.4

10.9 1.2 .5

11.8 3.3 .9

6.0 2.3 .8

15.9 3.0 1.2

5.6

9.2

9.4

7.9

43.6

57.6

73.6

47.8

77.4

Undistributed profits of corporations and other 1.2 organizations N et income of public enterprises and from public assets 1.1 .5 Social security contributions of employers .1 Addition for taxes and fees Adjustment for duplications - .8

.9 1.5 1.3 2.6 - 3 .8

.9 2.5 2.4 3.6 - 7 .0

- .5 1.0 1.7 2.6 - 7 .5

3.9 1.5 2.7 2.3 - 5 .7

Population, millions b National income in current prices 0 (billion marks) Cost of living, 1929 = 100 d National income in constant prices (1929)c (billion marks)

59.7 45.7 64.9 70.4

1913

1925 62.4 60.0 92.1 65.1

1929 64.0 76.0 100.0 75.9

National income by types in current prices (billion marks)

45.7

60.0

Wages and salaries 20.7 Entrepreneurial income: agriculture, forestry, fishing 5.7 trade, manufacturing and other industry, and 9.2 professions 5.7 Dividends and interest .9 Rent, urban Social security and relief payments, war pensions and 1.4 civil service pensions Total income of individuals

s 1913 territory includes Alsace Lorraine, former Prussian province of Posen, and former parts of Prussian provinces of Silesia, East Prussia, and the Saar district. Territory after World War I excluded the above, but in 1935-39 included the Saar. b Compiled from the Statistisches Reichsamt: Statistisches Jahrbuch fu r das Deutsche Reich, 1938, p. 9; Statistisches Handbuch der Weltwirtschaft, 1936, p. 36 (figures for Saar district). 0 Compiled from Reference No. 1, Volkseinkommen, 1932, p. 69. Wirtschaft und Statistik, 1939, p. 705; Statistisches Jahrbuch fu r das Deutsche Reich , 1938, pp. 559-60; 1941/42, p. 605. d Statistisches Jahrbuch fu r das Deutsche Reich, 1938, p. 331; W irtschaft und Statistik, 1939, p. 20. e Figures for 1932 and 1938 from Dr. Jostock. 380

2 6 . GERM ANY

incomes of the insured workers and employees were added the salaries and wages of the other workers and employees. The income of civil servants and employees of public enterprises were computed on the basis of the Federal Financial Statistics and other reports. The bien­ nial income tax statistics continued to be used for determining the salaries of top personnel not covered by the social security system. Farmers’ income was estimated by the net output method by resorting to improved esti­ mates of the gross value of output and cost expenditures in agriculture.14 The estimate of income derived from forestry was rather rough. For the estimates of corporate undistributed income, the corporation income tax statistics were used instead of the special statistics of corporate profits. This change in source was justified by the fact that in 1936 the latter sta­ tistics covered only 42 per cent of all profits subject to corporate income tax as compared with between 83 and 86 per cent in 1928-29.15

381

In adjusting national income for capital in­ come due to foreigners, apparently only the amounts actually transferred were deducted.16 T he national incom e estim ates fo r 193638 were provisional. Estim ates for 1939-40 were published in the Statistical Year Book, 1941/42; b u t no other estimates were pub­ lished through the rest of th e w ar period. T he few private estimates published abroad during th a t period were of questionable reliability.17 Table 26-3 shows the calcula­ tions fo r the period 1913-38.

5. Estimates by Size of Individual Incomes T he G erm an statisticians in the various states w ere originally m uch m ore interested in determ ining the distribution of individual

Table 26-4. Unadjusted Incomes of Individuals in Germany Classified by Income Classes, 1913-32“ Number of Income Recipients in Each Class0 Income Classes b (thousand reichsmarks) Over .0 to 1.2 1.2 to 3.0 3.0 to 5.0 5.0 to 8.0 8.0 to 12.0 12.0 to 16.0 “ 16.0 to 25.0 “ 25.0 to 50.0 “ 50.0 to 100.0 “ 100.0 All classes

1913

1928

1932

thousands 11,219 10,044 1,241 580 212 64 87 59 30 14

per cent 47.6 42.6 5.3 2.5 0.9 0.3 0.4 0.2 0.1 0.1

thousands 15,831 9,941 1,991 779 251 86 73 42 12 5

per cent 54.6 34.3 6.9 2.7 0.9 0.3 0.2 0.1 0.0 0.0

thousands 16,643 7,065 1,551 475 130 42 33 18 5 2

per cent 64.1 27.2 6.0 1.8 0.5 0.2 0.1 0.1 0.0 0.0

23,550

100.0

29,011

100.0

25,964

100.0

Aggregate Incomes of Individuals in Each Class Over

.0 to 1.2 to 3.0 to 5.0 to 8.0 to 12.0 to 16.0 to 25.0 to 50.0 to 100.0

1.2 3.0 5.0 8.0 12.0 16.0 25.0 50.0 100.0

All classes

millions Rm per cent millions Rm per cent millions Rm per cent 8,348 22.5 14,728 27.9 12,662 33.0 13,801 37.2 17,549 33.2 13,296 34.7 3,858 10.4 7,508 14.2 5,844 15.3 7.0 9.0 2,877 2,597 4,776 7.5 1,447 3.9 2,409 1,234 3.2 4.5 631 1,184 2.2 1.7 555 1.5 1,150 3.1 1,419 2.7 628 1.6 1,298 3.5 1,417 580 2.7 1.5 1,410 3.8 832 1.6 314 0.8 2,560 6.9 1,034 2.0 352 0.9 37,100

100.0

52,856

100.0

38,342

100.0

a Statistisches Reichsamt, Wirtschaft mid Statistik, 1939, p. 660. Incomes exclude nonassessed capital income, income from subsidiary occupations, war pensions, social security benefits, and relief payments. No account is taken of certain tax exemptions, understatements, and evasions in reporting entrepre­ neurial incomes. b Corresponding income classes for 1913 are 25-40 per cent lower. 0 Income recipients, exclusive of helping family members, but inclusive of recipients of pensions.

THE INCOM E OF NATIONS

382

Table 26-5. Income of Individuals in Germany by Types and Classes, 1928 *

Income of Individuals (percentages of total in each class) Entrepreneurial Income Income Classes b (thousand reichsmarks) Over .0 to 1.2 1.2 to 3.0 3.0 to 5.0 5.0 to 8.0 8.0 to 12.0 “ 12.0 to 16.0 16.0 to 25.0 “ 25.0 to 50.0 “ 50.0 to 100.0 “ 100.0 All classes

Wages and Salaries 68.8 65.7 65.0 58.0 47.6 35.7 30.3 25.0 21.5 15.4 61.3

Agri­ culture and Forestry 5.0 13.1 7.1 5.0 4.3 3.4 3.1 3.1 3.4 5.2

Trade, Industry, Dividends and Pro­ and fessions Interest * 1.2 2.4 13.3 19.3 6.1 24.8 9.8 3.8 40.5 49.8 6.5 8.3 53.7 55.6 11.5 55.0 15.5 55.8 20.0

8.0

16.8

3.8

Rent * 1.2 1.2 1.5 2.8 3.6 3.6 3.8 3.5 2.3 1.1

Social Security and Relief Payments, War Total Pensions0 Percentage 100.0 25.0 100.0 4.3 100.0 1.3 100.0 0.9 100.0 1.0 1.0 100.0 1.0 100.0 1.0 100.0 100.0 1.1 100.0 1.3 9.0

100.0

a Reference No. 1, Volkseinkommen, 1932, p. 113. Incomes are adjusted for nonassessed capital income> income from subsidiary occupations, war pensions, social security benefits, and relief payments. b Income classes refer to incomes before adjustments (see Note a). 0 Old age pensions of civil servants are included with wages and salaries. * Separate figures not available; included with other types of income payments in the same income class. incomes by size of individual incomes than in calculating the total national income. Even when, in the present century, they finally took up the estim ation of the national income (as distinguished from the aggregate of private incom es), they continued their original interest in incom e distribution. Their series of incom e distribution in the nineteenth and tw entieth centuries were among the m ost developed o f any country of the time. D uring the late 1920’s and the 1930’s two different official estimates o f incom e distribution by size of individual incomes were prepared. One of them was built di­ rectly from income tax statistics and fol­ lowed the tax classification of income. It em braced taxed incomes plus those below the personal exem ption lim it in the year 1913 and in every alternate year of the period 1926-36. It excluded social benefits, nonassessed capital income, incom es from subsidiary occupations, and m ade no ad­ justm ents fo r understatem ents of income on tax returns or for other tax evasions or for certain tax exemptions. T he second estim ate was developed from the com putation of the national income itself and followed the latter’s classification of incomes by type throughout; but it was prepared only fo r the year 1928. The m ain

difficulty in the developm ent of this second estim ate was the allocation of certain esti­ m ated totals am ong different categories of incom e. A ll understatem ents of incom e were distributed in equal proportions am ong the different classes of assessed incom e. Missing capital incom es were classified p roportion­ ately even am ong the incom e classes of 1,200 to 8,000 m arks. W ar pensions were allocated in p roportion to the num ber of taxpayers. Relief paym ents w ere included in the low est incom e group. Some of the results of these com putations are given in Tables 26-4 and 26-5.

II. Postwar Net Output Estimates Soon after its establishm ent in 1948, the statistical office of W est G erm any (Stat­ istisches B undesam t), resum ed the w ork of estim ating the G erm an national income, previously conducted by the R eich’s Statis­ tical Office and interrupted by G erm any’s collapse during the war. In th e p rep aratio n of these estimates, m ain reliance was placed on the n et output m ethod rath er than, as by the R eich’s Sta­ tistical Office, on th e incom e-distributed m ethod. T his shift was probably m ade for

2 6 . GERM ANY

383

two reasons: first, because o f the disorgan­ ized conditions of the country at the time, production data, how ever incom plete, w ere m ore reliable for national incom e m easure­ m ent th an fiscal data; and secondly, it was m uch m ore im portant to m easure the re­ covery of production in the various branches o f the econom y than to m easure the distri­ bution of incom e am ong different groups o f producers. T h e estimates w ere concerned w ith the national incom e o f the W est G erm an te rri­ tory including the City of Berlin, and they had to be prepared in a roundabout process. First, an estimate had to be prepared for the com bined territory of the form er Reich fo r a prew ar year th a t could be used as a base from w hich to m ake the new calcu­ lations (just as 25 years before, in estim ating the national incom e of G erm any after W orld W ar I, an estim ate had to be p re­ p ared first fo r the year 1913 as a b ase). In this case, the year 1936 was selected as a base because: (a ) it was the last com ­ paratively “norm al” year; and ( b ) especially rich production data w ere available fo r th a t y ear from several sources including an in­ dustrial census held in 1936, and a series of special studies initiated by the R eich’s Statistical Office in 1936 covering the opera­ tions of the turnover tax and the rise in production costs. These data w ere easily extrapolated from 1935 and 1937 to 1936. T he second step in the procedure was the determ ination fo r 1936 of the proportion of this national income total allocable to the W est G erm an and Berlin territories. T he th ird and final step was the extrapolation of the 1936 figure to the year 1948 w ith the aid o f appropriate production and price indexes.

as given by the R eich’s Statistical Office (4 billion R m ). Thus, the net output of agriculture was p u t down at 8.2 billion Rm. T he Industrial Census of 1935 gave an “over-all value added” fo r the m anufactur­ ing industry of 34 billions, b ut this figure included taxes and a num ber of other ex­ traneous costs representing the contribution of services by other branches of the econ­ omy, such as rent, advertising, insurance, travel, mailing, etc. I t was necessary, there­ fore, to calculate these items and to deduct them from the foregoing total. T he statistics of the turnover tax as well as of the busi­ ness and property taxes showed total pay­ m ents by the m anufacturing industry in 1936 o f 2 billion Rm . T he other extraneous item s w ere estim ated indirectly. Using as a guide the fact that, according to the tu rn ­ over tax statistics, the corresponding de­ ductible costs in the case of handicrafts and retail and wholesale trade am ounted to be­ tw een 4 and 10 per cent of gross sales, the estim ators settled on a 5 p er cent p ropor­ tion for m anufacturing, or 3.4 billion Rm. A fter deducting these two figures (2 billion and 3.4 billion) from the foregoing total of 31.4 billion, a figure of 26 billion Rm was obtained as a reasonable estimate for the net value of o u tp u t in m anufacturing. T he turnover tax statistics fo r 1935 gave com prehensive inform ation on the gross values of the output of handicrafts, retail and wholesale trade, and hotels. These gross values w ere next extrapolated to 1936. N et values w ere derived by applying the net to gross ratios developed fo r these branches of the econom y by the R eich’s Statistical Office fo r 1937. T he n et outputs o f other indus­ trial divisions were estim ated as follows:

1. Estimates for the Former Reich for 1936 by the N et Output Method

The net value of the transportation and com­ munication services in 1936 was ascertained either directly from published reports or indi­ rectly from other data. For the Post Office and the railroads, fairly complete direct information was available. For overseas shipping, direct in­ formation available for lines representing about 40 per cent of the business was used in calcu­ lating the total. For inland water transportation and street railways, the turnover tax statistics supplied the gross values. To these gross value figures fixed ratios of net to gross were then applied. For inland transportation, the over­ seas shipping’s ratio of net to gross (30 per cent) was adopted. In the case of motor trans­ port a 50 per cent ratio was adopted which was somewhat lower than the 55 to 65 per cent ratio found to exist in the case of railroads. For banks, the net value was calculated from

In preparing the estimates, the standard classification of industrial divisions recom ­ m ended by the statistical u nit of the U nited N ations was very closely followed. T he gross value of the output in agricul­ ture. was estim ated on the basis of figures prepared for 1936 by the Econom ic Research Institute (Institut fu r K onjunkturforschung) of the gross sales of farm products (9.1 bil­ lion R m ) and of the value of farm ers’ con­ sum ption of their own products (3.1 bil­ lion R m ). The deductible item s w ere taken

384

THE INCO M E OF NATIONS

Table 26-6. National Income of the German Reich, and the Share Allocated to the West German Territories, 1936, by Industrial Divisions, Official Estimates 1948 (billion Rm at 1936 prices) For the Combined Territory of the Reich per cent Agriculture 8.2 (13) Manufacturing 26.0 (41) Handicrafts 7.1 (ID Retail trade 2.7 (4) Hotels and restaurants 1.0 (2) Wholesale trade 3.1 (5) Transportation and communication 5.4 (8) Banks 0.9 (1) Insurance 0.4 (1) Housing 2.2 (3) Government services, including social security 5.3 (8) Free professions 1.0 (2) Domestic and other services 0.8 (1) Total home-produced income Net interest from abroad Net national product a t factor cost Indirect taxes and fees Net national product a t market prices Depreciation allowances Gross national product source:

64.1 - 0 .5 63.6 10.8 74.4 7.0 81.4

(100)

For the United States and British Zones 3.9 13.8 3.3 1.3 0.5 1.7 , 2.9 0.4 0.2 1.1 2.6 0.5 0.4

For the French Zone 0.9 1.8 0.6 0.2 0.1 0.1 0 .4 0.1

For Berlin —

0.1 0.4 0.1 —

2.2 0.9 0.4 0.1 0.6 0.5 0.3 0.1 0.2 0.3 0.1 0.1

4.8

5.9

—

32.6 - 0 .2 32.4 5.5 37.9 3.6

4 .8 0.8 5.6 0.5

5.9 1.0 6.9 0.6

41.5

6.1

7.5

—

—

Reference No. 5, pp. 9, 12.

the accounts of a sample representing about 27 per cent of the business. For insurance, the net value was taken directly from published re­ ports. Rental incomes from dwellings were taken as reported by the Reich’s Statistical Office (980 million R m ), but to this figure was added interest on mortgages and other debts (1.2 billion Rm ). Government services were evaluated from official reports, as the amount of the compen­ sation in cash and in kind paid to government employees. Interest on the public debt was not included, being treated as a transfer payment. The net value of the services of professions was calculated by applying a 70 per cent net to gross ratio to their gross receipts reported in the 1936 turnover tax statistics. The value of domestic services was estimated by (a) multi­ plying the reported number of persons engaged in these occupations in 1936 (1.4 million) by an average income of 700 Rm and (6) cutting the resulting total (1 billion Rm) in half. This cut was made on the ground that a large pro­ portion of these services had already been ac­ counted for in the estimates of the net output of agriculture, commerce, handicrafts, and hotels. The net value of other personal services, such as theaters, sports, health services, etc., was estimated by taking their gross receipts as de­

rived from the turnover tax statistics and ap­ plying a net-gross ratio of 25 to 30 per cent to them. N et capital income from abroad was taken directly from the published international balance of payments statement. T he sum m ation of all these item s p ro ­ duced a total national incom e at factor cost fo r the Reich territo ry fo r 1936, at 1936 prices, of 63.6 billion R m as show n in T able 26-6.

2. Computation of West Germany’s Share of the Reich’s National Income in 1936 In calculating the portion o f the R eich’s n ational incom e at factor cost allocable to W est G erm any, separate calculations were m ade fo r th e British-U nited States zone, the F ren ch zone, and Berlin. M oreover, the incom e portions w ere com puted fo r each b ran ch of the econom y separately, using a distinct m easuring rod in each case. Thus, the incom e from agriculture was m easured on the basis of the ratio o f the value of

2 6 . GERM ANY

agricultural property in the area to the total value of such property in the Reich. F o r some other divisions the proportions of turnover tax collections w ere used. In still other cases the proportions of producers or o f units of equipm ent w ere used as m easur­ ing rods. F or the B ritish-U nited States zone the proportions of the national total ranged from a low of 36 p er cent in th e case of street railways to a high of 94 per cent in the case of overseas shipping. T he net result of this allocation was to give the British-U nited States zone and Berlin larger shares of the national incom e total than their shares of the national popu­ lation total w hile producing exactly reverse results for the F rench zone.

49.1 6.3 7.7

National Income Shares (per cent) 50.9 9.3 7.4

63.1

67.6

Population Shares (per cent) British-United States zone Berlin French zone Total so urce

:

Reference No. 5, p. 14.

By applying these proportions to the n a­ tional income totals fo r 1936 (see T able 26 -6 ), the estim ators obtained the national incom e figures fo r these territories fo r that year. U n d er this allocation, the per capita n a­ tional incom e of 1936, at 1936 prices, varied as follows: Per Capita Per Capita 1936 1936 N.I. a t N.I. at Market Factor Cost, Rm Prices, Rm British-United States 1,141 975 zone 1,063 910 French zone 1,613 1,380 West Berlin Combined 1,100 940 Reich Territory

3. Extrapolation of the Estimate for 1936, by the N et Output Method, to 1948 T he 1936 national incom e at factor cost o f each zone was next extrapolated to 1948 on the basis of special production and price indexes. These indexes w ere com puted sepa­

385

rately fo r each branch of the economy. In applying the production indexes (which deal w ith the physical volum e of p ro duction), the assum ption was m ade th at the ratio of net value to gross value rem ained unchanged over the period, i.e., th a t the net value de­ veloped at the same rate as the volume of production. T his assum ption had to be m ade, fo r there was no w ay of ascertaining w hat changes in the ratios of net to gross value had actually taken place during that tim e. T he indexes used and the results of their application are show n in Table 26-7. T he com putations showed the national product at m arket prices fo r the BritishU nited States zone fo r the h alf year 1948 to be 27.3 billion R m in 1948 prices, a sum 44 p er cent higher th an th at fo r the cor­ responding period of 1936 in 1936 prices. W hen this figure was deflated, however, to a constant price basis, th e real national incom e of the zone was found to have dropped 16 p er cent from the 1936 level; and the per capita real national incom e to have dropped 29 p er cent from th a t level and 42 p er cent from the 1938 level. T he structure of the national income changed substantially from 1936 to 1946, as show n in T able 26-8.

4. Subsidiary Estimates of Distri­ bution and Expenditure of Na­ tional Income F rom the estimates of national incom e at factor cost com puted by the net output m ethod, two subsidiary estimates were next developed w ith the aid of additional data: one of the income distributed; and one of th e expenditure fo r final products at m arket prices. T he original incom e-distributed estimates fo r the Reich fo r 1936 m ade by the old R eich’s Statistical Office were accepted with certain adjustm ents fo r duplicated transfer incom es (T able 2 6 -9 ). These old estimates gave regional subtotals fo r the income dis­ tributed as well as national totals.18 T he structure of the national expenditure of the Reich in 1936 was developed by first estim ating n et capital form ation and expenditures fo r governm ent services. These tw o item s were deducted from the total of the national incom e at m arket prices, with the rem ainder representing private expendi-

Table 26-7. Extrapolation of National Income of United States-British Zone of Germany

Half Year 1936 to Half Year 1948 (billions Rm)

Agriculture Manufacturing Handicraft Retail trade Wholesale trade Hotels and restaurants Transport and communications Postal services Railroads Streetcars Steamship lines Highway transport Banks Insurance Housing Government services Free professions Domestic and other services Rest of the world Net national product at factor cost Indirect taxes Net national product at market prices Depreciation Foreign aid Gross national product so urce:

1936 Half Year (billion Rm) 1.9 6.9 1.6 0.7 0.8 0.3

1948, Half Year ----------------------------------------------------------------In 1936 Prices In 1948 Prices ----------------------------------- --------------------------100 Prod. 100 Price Index = 1936 Billion Rm Index = 1936 Billion K 80 1.6 155 2.4 70 4.9 204 10.0 90 1.5 180 2.7 66 0.4 230 1.0 75 0.6 185 1.2 50 0.1 140 0.2

0.4 0.8 0.0 0.1 0.1 0.2 0.1 0.6 1.3 0.3 0.2 - 0 .1

0 115 215 60 150 100 100 80

16.2 2.7 -----18.9

(85) -----

1.8

85

120

0.3 0.9 0.1 0.1 0.1 0.2 0.1 0.4 1.9 0.3 0.3

0 150 120 200 200 140 130 100 120 110 100

0.6 1.4 0.1 0.1 0.3 0.3 0.1 0.4 2.3 0.3 0.3

13.8 2.1

(172) (172)

23.7 3.6

---

---

15.9

20.7

27.3

1.5 0.9

3.1 1.3

205 145

18.3

31.7

Reference No. 5, p. 18.

Table 26-8. Structure of the Net National Product at Market Prices of 1936 in the Combined West German Area by Economic Divisions in 1936 and 1948 (Without foreign aid) (In percentages of the total) Percentages Half of 1936 Agriculture 10.0 Manufacturing 36.5 Handicraft 8.5 Trade, including hotels and restaurants 9.5 Transportation, communications, banking, and insurance 9.0 Housing 3.2 Government services 6.9 2.1 Domestic, professional, and other services Indirect taxes 14.3 100.0 sou rce:

Reference No. 5, p. 18. 386

Half of 1948 in 1936 Prices 10.1 30.8 9.4 6.9 11.3 2.5 12.0 3.8 13.2 100.0

2 6 . GERM ANY

387

Table 26-9. National Income of Former German Reich and West Germany, 1936, by the Income-distributed Method a t 1936 Prices (billions Rm)

Agriculture and forestry Commerce, manufacturing, and handicraft Salaries and wages Capital income Rent Pensions and benefits Aggregate private incomes Retained corporate profits Profits of public enterprises Employers’ contribution to social security Total Deduction for double counting of gov­ ernment transfer expenditures National income a t factor cost

Former Reich 5.8 10.6

35.3 2.71

West Germany (Incl. Berlin) 3.5 “ 8.4 * 23.2 9p . o5 Z

1-0/

7.8

2.5 b

63.2 2.3 1.31 2.4J

40.1

69.2

42.0

1.9

- 5 .6 63.6

42.0

“ Including retained profits. b Net of duplications of government transfer payments. s o u r c e : Reference No. 5, pp. 8, 12.

Table 26-10. National Expenditure of Former German Reich and West Germany, 1936 (1936 prices) (billion Rm)

ss national product Less depreciation national income a t market prices Less net capital formation and Less expenditure for government services ate consumption Food Clothing Household goods, vehicles, etc. Fuel and light Personal care and health expense Housing Transportation and communication Recreation Insurance Domestic services Other personal services so u rc e:

Reference No. S, pp. 8, 12.

Former Reich Territories per cent per cent 81.4 - 7 .0 74.4 100

West Germany (Incl. Berlin) 55.1 - 4 .7 50.4

- 8 .0 - 1 3 .2 53.2

(11) (18) (71)

100

35.4

22.6 5.0 7.1 1.1 2.0 7.0 2.0 0.5 1.6 1.0 3.3

(30) (7) (10) (1) (3) (9) (3) (1) (2) (1) (4)

(42) (9) (13) (2) (4) (13) (4) (1) (3) (2) (6)

14.6 3.3 5.2 0.7 1.3 4.8 1.4 0.3 1.0 0.7 2.1

- 1 5 .0 —

388

THE IN CO M E OF NATIONS

ture fo r consum ption. This residual was next broken down by types of consum er expendi­ tures on the basis of turnover tax and con­ sum er excise tax data. T he com position of the national expenditure is shown in Table 26-10. T he estim ate of W est G erm any’s share of the total national expenditure in 1936 was based on separate breakdow ns of its share of each type of expenditure w ith the aid of a different m easuring rod in each case as explained previously (Section 2 ) . In the case of Berlin the share of the total national expenditure (10.8 per cent) was substan­ tially greater than its share of the national population (6.3 per ce n t). In the case of the British-U nited States and the F rench zones of W est G erm any, the shares of the expenditures and of the populations w ere about the same. Both the incom e-distributed and final ex­ penditure estim ates fo r W est G erm any fo r 1936 were next extrapolated to 1948 with the aid of appropriate indexes.

5. Subsequent Improvement of Estimating Methods T he estim ating m ethods were fu rth er im ­ proved after 1949. T he n et o u tput m ethod continued to play a m ajor role, b u t sub­ sidiary estim ates by the incom e-distributed and the final expenditure m ethods were also prepared. Tables 26-11, 26-12, and 26-13 show the changes in the national incom e by industrial origin and by types of expenditure in consistent series fo r the period 1925 to 1953. A sim ilar incom e-distributed series apparently could n o t be developed. In 1954 the efforts o f the N ational Incom e U nit of the Statistical Office w ere concentrated on developing a fundam ental estim ating m eth­ odology th a t w ould enable it to estim ate the national incom e on the basis of direct, cur­ ren t statistical data instead o f by extrapola­ tion of the 1936 estimates. This new m eth­ odology and its statistical results, however, w ere n ot yet com pleted and m ade public at the tim e of this w riting (1 9 5 6 ).

Table 26-11. N et National Income at Factor Cost and a t M arket Prices, and Gross National Product of West Germany, 1925-54 Population (millions) Totals, a t current prices (billion marks) National income Net national product at market prices Gross national product a t market prices Totals, a t 1936 prices (billion marks) National income N et national product a t market prices Gross national product a t market prices Per capita income, a t current prices (thousand marks) National income Net national product a t market prices Gross national product a t market prices Per capita income, at 1936 prices (thousand marks) National income Net national product a t market prices Gross national product a t market prices

1925

1929

1932

35.8

36.7

34

1953 49.0

1954

1939

1949

37.3

1936 38.4

39.5

46.8

43

25

38

52

63

104

112

38

48

30

44

60

73

124

134

42

52

34

48

65

80

134

145

29

34

26

38

51

39

56

61

33

38

32

44

59

44

67

72

35

42

35

48

63

47

72

77

1.0

1.2

0.7

1.0

1.3

1.3

2.1

2.

1.1

1.3

0.8

1.1

1.5

1.6

2.5

2.

1. 2

1.4

0.9

1.3

1.6

1.7

2.7

2.9

0.8

0.9

0.7

1.0

1.3

0.8

1.1

1.2

0.9

1.0

0.8

1.1

1.5

0.9

1.4

1.5

1.0

1.1

0.9

1.3

1.6

1.0

1.5

1.6

* Preliminary. s o u r c e : Statistiques et Etudes Financieres, Ministere des Finances, January, 1955, p. 105; Wirtschaft und Statistik, Heft 4, April, 1955, pp. 191, 192. Population estimates from U.S. Department of Com­ merce, Bureau of the Census, Population of the Federal Republic of Germany and West Berlin, p. 58.

2 6 . GERM ANY

389

Table 26-12. National Income and Gross National Product of West Germany, 1951, 1953, by Industrial Origin (billion Deutsche marks) Current Prices

1936 Prices

1951 10.0 0.9 40.2 5.8 4.1 3.3 4.9 0.9 6.7 0.9 0.4 1.4 7.1 0.9

1953

1951

10.3 1.0 46.3 7.2 4.7 3.8 5.9 1.1 8.0 1.1 0.5 1.6 9.4 1.0

5.1 0.4 20.1 2.7 2.2 2.2 2.4 0.6 4.5 0.6 0.3 1.3 5.1 0.7

1953 5.4 0.3 23.6 3.4 2.3 2.5 2.8 0.8 5.1 0.6 0.3 1.4 5.7 0.8

1.4 1.1 —

1.3 1.2 - 0 .6

1.0 0.8 —

0.9 0.8 - 0 .4

N et national product a t factor cost Indirect taxes Subsidies

89.9 16.3 - 0 .6

103.0 20.4 0.1

49.7 — —

56.3

Net national product a t market prices Depreciation

105.5 8.1

124.1 10.2

58.8 3.9

66.8 4.8

113.6

134.3

62.7

71.6

Agriculture Forestry Manufacturing (excluding construction) Construction Handicraft (excluding construction) Retail trade Wholesale trade Hotels, restaurants Transportation Banking Private insurance Housing Government Liberal professions Payments of services to the occupying authorities Domestic and other services Income from abroad

Gross national product so u rce:

—

—

Wirtschaft und Statistik, April, 1955.

6. Estimates Prepared in 1 9 5 6 -5 7 A fter these tables and text w ere set in type, the publication of new W est G erm an estim ates based on direct statistics has com e to the author’s attention (see W irtschaft und Statistik, M arch, 1957; reproduced in

the U nited N ations Yearbook o f N ational A cco u n ts Statistics, 1 9 5 7 ). These estimates are expressed in both current and 1954 con­ stant prices and cover the years 1950-56. This inform ation has com e too late to be incorporated here. Publication o f complete double entry accounts is prom ised in time.

THE INCO M E OF NATIONS

390

Table 26-13. Gross National Expenditure a t M arket Prices, West Germany, 1936-53 (billion Deutsche marks) Current Prices

1951

1953

1954*

26.8

11.7

13.0

14.0

11.9 11.5 6.0 3.0

12.7 11.9 6.5 3.7

3.5 4.9 2.8 1.7

4.7 6.0 3.3 2.0

5.2 6.3 3.6 2.2

2.3 4.9 3.1 3.6 0.9

3.0 5.8 4.0 4.3 1.0

3.2 6.2 4.5 4.7 1.0

1.6 4.5 1.8 2.5 0.6

2.1 5.1 2.3 3.0 0.7

2.3 5.4 2.6 3.3 0.7

0.3

0.3

0.3

0.2

0.2

0.2

65.1

75.7

81.7

35.9

42.5

45.9

4.7 1.2 7.0

4.6 1.6 9.2

4.6 1.7 9.9

2.4 0.5 4.9

2.3 0.7 5.5

2.3 0.7 5.7

5.5

7.1

7.8

3.5

4.2

4.7

18.4

22.5

24.0

11.3

12.7

13.5

22.5

27.8

31.0

10.7

12.6

14.1

5.6

4.1

+ 4 .5

2.8

2.0

+ 2 .1

—

+ 0 .1

+ 0 .3

—

28.1

32.0

35.4

13.6

14.6

16.3

111.6

130.1

141.1

60.7

69.8

75.6

2.0

4.2

+ 4 .2

+ 2 .0

1.8

+ 1.8

113.6

134.3

145.3

62.7

71.6

77.4

65 18

57 16

56 17

56 17

60 21

62 22

57 18

59 18

59 17

17

27

27

27

19

16

25

23

24

100

100

100

100

100

100

100

100

100

1936 1949 Private consumption: Food Alcoholic beverages, coffee, tea, tobacco Clothing Furniture, household goods Fuel, lighting Personal care and health expenses Housing Transportation Education and recreation Domestic services Other services, including banks and insurance Total private consumption Public consumption: Expenses of the occupation authorities: Current expenses Investment Expenditure for personnel Expenditure for goods and services Total public consumption Capital formation: Equipment Inventories In the home economy In free ports and bonded warehouses Total capital formation Total goods and services used in Germany Exports minus imports plus net in­ come from Soviet zone, etc. Gross national product a t market prices Divisions of gross national expend­ iture in percentages of total Private consumption 60 Public consumption 21 Capital formation including bal­ ance of trade, etc., and net in­ come from abroad 19 Gross national product

100

1951

1953

1954

22.0

24.9

9.7 10.8 5.2 2.4

(a) less than 0.1 (42 millions). * Preliminary. so u rce:

Constant 1936 Prices

Wirtschaft und Statistik, April, 1955, pp. 195-224.

1936 1949

+ (a) + 0 .1

27. FRANCE

References 1. Commissariat General du Plan de Modernisation et d’Equipement, Estimation du Revenu National Frangais, Paris, March, 1947. (Published also in the Etudes et Conjoncture, Paris, March to April, 1947.) 2. Organisation for European Economic Co-operation. National Accounts Research Unit, N a­ tional Accounts Studies, France, Paris, 1952 (hereafter referred to as OEEC). 3. Ministere des Finances et des Affaires Economiques, “Rapport sur les Comptes Economiques de la Nation,” rendu au Parlement, Jan. 23, 1951, Statistiques et Etudes Financieres No. 27, March, 1951. 4. a. Ministere des Finances, Service des Etudes Economiques et Financieres, Les Comptes Economiques de la Nation pour I’Annee 1949 (Economic Accounts of the N ation), Paris, 1952. b. , Les Comptes Economiques de la Nation pour I’Annee 1950, Paris, 1953. c. , Comptes Provisoires de la Nation pour I’Annee 1953 (Provisional Accounts of the N ation), Paris, 1954. d. , Rapport sur les Comptes de la Nation, Paris, 1955; vol. I, Comptes 1945-55, vol. II, Methodes (also the Comptes and the Budgets Economiques, published in 1956-58). 5. Ministere des Finances, Statistiques et Etudes Financieres, No. 40-41, April-M ay, 1952; also Supplement, Finances Frangaises, No. 20, 1953, Part II, “Le Revenu National de la France.” 6. LEconom ie, Journal, June 18, 1953, “Les Comptes de la Nation . . . 1953” (hereinafter referred to as L ’Economie, June 18, 1953).

1. Recent History Before W orld W ar II, national incom e estimates in F rance were prepared privately by individual scholars and w ere based on exceedingly crude and inadequate statistical data. O utstanding am ong them w ere those m ade by D uge de Bernonville, D irector of F ren ch Statistics, covering the decades of th e 1920’s and 1930’s, and based on the incom e-distributed m ethod. T hey w ere theo­ retically interesting, b u t w ere based on sta­ tistics of very p o o r quality. D uring th e G er­ m an occupation, research in the national incom e field was alm ost suspended. A fter liberation, w ork on national incom e esti­ m ates was not only revived, but was at last placed on a firm official foundation. T he “C om m issariat du P lan de M odernisation” (Com m ission on the M odernization of the * Acknowledgment is due to Mr. C. Gruson of the Office of Economic Studies of the Min­ istry of Finance of France for having reviewed this chapter in its preliminary form and having made helpful suggestions. Notes begin on p. 541. 391

E conom y ), recognizing the im portance of national incom e estim ates as tools o f eco­ nom ic planning, p repared some in 1947-50. In 1951, the M inistry of Finance took over this w ork, establishing a special economic and financial research u n it ( “Service des Etudes E conom iques et Financieres,” here­ inafter referred to as S E E F ) to carry it on along w ith other new associated studies. These additional tasks included the prepa­ ration o f com prehensive social accounts, as well as o f an annual econom ic budget that is both an econom ic forecast an d an eco­ nom ic plan for the nation. T he national in­ com e estimates were prepared independently by the net o u tput m ethod and, in a sub­ sidiary way, by the final products method, w ith additional analyses by th e income-dis­ tributed m ethod. T he estimates were based on 1938 d ata projected to subsequent years by production indexes and price indexes. In 1952, substantial changes in the m eth­ ods of estim ating the national incom e and constructing the social accounts were intro­ duced. In applying both th e net output

392

THE INCOME OF NATIONS

m ethod and the incom e-distributed m ethod, a shift was m ade from 1938 figures to cur­ rent data. Only in the case of the final p ro d ­ ucts m ethod were the 1938 data and pro ­ duction and price indexes used as a base. B ut even in the latter case considerable re­ finements were introduced. In 1955 still fu rth er com prehensive changes in the construction of the estimates w ere m ade. These new estim ates provided m ore detailed social accounts and carried the calculations backw ard to the year 1949 on the basis of uniform classifications of industrial divisions and types o f incom e. M oreover, they em phasized particularly the final expenditure or products approach while retaining the incom e-distributed m ethod, but dispensing w ith the net o utput type o f esti­ mate. In the developm ent of these new ap ­ proaches, the research u nit of the M inistry of Finance had the active collaboration of the Institut N ational de la Statistique.*

.

2 C oncepts' National incom e ( “revenu national” ) is defined as the aggregate of incom es received by habitual residents of the country from their productive efforts at hom e and abroad. It is distinguished from the home-produced, incom e ( “revenu interieur” ), w hich includes factor incomes earned in the country by nonresidents but excludes factor incom es re ­ ceived by F rench residents from abroad. Accordingly, national incom e consists of the labor incomes paid to F rench residents by private and public enterprises, govern­ ment, and social security institutions; capital and entrepreneurial incom es received by F rench residents o r retained and invested by enterprises fo r their benefit; direct busi­ ness net income taxes; and profits o f public enterprises less interest on the public debt. Corresponding to this m onetary concept o f national incom e as an aggregate o f dis­ tributive shares are the tw o product con­ cepts: (a ) the net national product at factor * n o te : After this chapter was written, the new report prepared by the Research Unit in co-operation with the Institute of Statistics and Economics, presenting the results of these new approaches, became available. It came too late, however, to be made use of in this text. It provided a homogeneous series for the years 1949-55 based on current production and other data. The social accounting approach largely displaced the traditional single-entry national estimate. The series 1949-55 were deflated at 1952 prices. (Reference No. Ad.)

cost ( “p ro d u it national n et au cout des facteurs” ) , representing the net values added o f the different branches of the enterprise sector plus the value added in th e public sector, adjusted fo r the n et balance of the international facto r account; an d (£>) the net national p roduct at m arket prices, i.e., th e value of final private an d public goods an d services less depreciation, b ut inclusive o f indirect taxes, adjusted fo r im ports and exports. Finally, there is the gross national p ro d u ct ( “p ro d u it national b ru t” ) , consist­ ing o f the gross dom estic enterprise product ( “production interieure b ru te” ) , n et factor incom es from abroad, and com pensation of public em ployees. T he interrelationships of these aggregates and th eir elem ents are show n in D iagram 27-1. T h e national econom y is deem ed to con­ sist of three m ain sectors: (a ) the enterprise sector, private and public, w hich em braces only the production o f goods and services sold fo r a price and is called, m uch too broadly, the “production sector” ( “Secteur de la P roduction” ) ; ( b ) the public sector com posed solely of the services of govern­ m ent and social security em ployees th a t are furnished free (i.e., paid fo r by general com pulsory levies); and (c ) the rest o f the w orld sector ( “secteur etranger et outrem er” ) . F o r som e purposes additional sectors are recognized, such as the sector o f capital flow s ( “secteur, m arche des capitaux” ) com ­ prised of the m ovem ent of banking funds and o f personal properties an d a sector of private households ( “secteur des particuliers” ) w hich em braces individuals and cer­ tain groups in their capacities o f consum ers an d savers. T his chapter is concerned alm ost exclu­ sively w ith th e estim ates p repared during the period 1947-55 and is divided into two parts. P a rt I deals w ith th e estimates pre­ p ared during the years 1947-51 by the “C om m issariat du P lan de M odernisation” ; and P art II treats those p repared in 195255 by th e Econom ic and F inancial Research U nit of the M inistry o f Finance.

I. Estimates Prepared During

1947-51 T h e “C om m issariat du P lan de M oderni­ sation,” in preparing its estim ates o f F rench national incom e in 1947-51, used all three estim ating m ethods. In applying the net o ut­ p u t an d final products m ethods, the esti­

27.

393

FRANCE

Diagram 27-1. Principal Elements of the French National Accounts, 1949 (billion francs) Uses of products

Sources of products

Exports

Gross domestic investment

Private consumption

Public consumption

1,040

1,630

5,720

1,170

Imports

Gross domestic enterprise production (final goods and services) a t market prices

Compensation of public employees

1,100

7,800

660

Gross national product

Gross national product (gross domestic enterprise production + compensation of public employees + net factor income from abroad) 8,480

Derivation of national income

sou rce:

Indirect taxes

Depreciation

National income a t factor cost

970

870

6,640

Reference No. 46,

pp. 2

and

8.

m ators used, to a large degree, the figures fo r the 1938 year, extrapolating them to the postw ar years w ith the aid of a variety of indexes. Only in the case of certain elements o f the national incom e were they able to obtain current data. In the preparation of estimates by the incom e-distributed m ethod, on the other hand, they were able to develop their estim ates to a m uch greater degree from direct current inform ation.

1. Estimates by the N et Output M ethod 2 T he n et output estim ates prepared in 1947-51 for th e years 1946-50 concerned themselves w ith the net national incom e at factor cost and, as already stated, w ere p re­ pared largely by first estim ating the elem ents of the national incom e originating in the various branches of national production in 1938 a t 1938 prices and then extrapolating them a t th e same prices to later years with the aid of various indexes. T he figures for 1938 w ere derived partly from the censuses of the 1930’s w ith the aid of extrapolations, and partly from various official surveys con­

ducted during the G erm an occupation in w hich the econom ic activity of 1938 was taken as a bench m ark. A m ong the indexes used fo r the extrapolation of these figures w ere those of population; of the num ber of wage earners engaged in given industries or occupations; num ber of m an-pow er hours used in production; average wages; am ounts of dividend paym ents; and prices. Only in the case of some branches of economic activity were direct current data on the gross and net values of outputs obtained. T his was tru e of agriculture, transporta­ tion, banking, insurance, professions, do­ mestics, public adm inistration, and social security. F o r m ost branches of the economy, the gross value of the o u tput was com puted first. T hen deductions were m ade fo r the costs o f producer goods and services, in­ direct taxes (less subsidies) and, insofar as ■practicable, fo r depreciation. B ut fo r some branches, the value added, net o f these two items, was com puted directly by adding the factor paym ents. D epreciation in m ost cases was com puted at theoretical am ounts, w ith­ o ut regard to the am ounts reported on tax

394

THE INCOME OF NATIONS

returns or carried on the books of the firms. M ore will be said about this later. T he branches of the enterprise sector were classified into eight m ain groups: 1. 2. 3. 4. 5. 6. 7. 8.

Power T ransport A griculture Industry— basic m aterials, construction, public works, and m anufacturing T rade Services (professional, domestic, etc.) Banks and insurance com panies Services of dwellings

T he sum of the net outputs of these eight branches of econom ic activity constituted the net product of the enterprise sector at factor cost. T he addition to it o f indirect taxes net of subsidies— an easily calculated item— converted it into a figure of the net product of this sector at m arket prices; while the fu rther addition of depreciation converted it into the gross product of this sector. T able 27-1 shows the m ain elements in the calculation of the net product. Depreciation as an elem ent of the gross product of the enterprise sector was com ­ puted as the annual am ount “theoretically

necessary to m aintain the assets intact.” A n inventory of these assets, w ith their esti­ m ated values, was taken in 1938. T he assets were classified in the m ain into three cate­ gories: 3 (a ) buildings, the life of w hich was assum ed to be 70 years; ( b ) fixed equip­ m ent, such as bridges and ports, assum ed to have 50 years of life; and (c ) transport equipm ent, cranes, elevators, m achinery, and tools w ith an estim ated life of 15 years. The figure of theoretical depreciation fo r 1938 in 1938 prices was next projected to other years by the estim ated n et percentage change in the capital assets an d price index, except in the case o f railways an d certain other special equipm ent fo r w hich the actual m ain­ tenance expenditure fo r those years was taken. This theoretical depreciation was dis­ tinguished from “am ortization,” w hich rep­ resented the actual provisions fo r deprecia­ tion m ade on th e books o f the business concerns o r allowed by the law on the tax returns. T h e dom estic national incom e a t factor cost was obtained by adding to the net do­ m estic p roduct of th e enterprise sector at facto r cost, th e com pensation in cash and in kind of governm ental civil an d m ilitary em ployees and of social security employees.

Table 27-1. Net National Product of France a t Factor Cost and a t M arket Prices, by Branches of Economic Activity, 1938-49 1951 Estimate (billion 1938 francs) Enterprise Sector 1. Power 2. Transport 3. Agriculture 4. Primary materials 5. Construction and public works 6. Manufacturing 7. Trade 8. Services 9. Banks and insurance 10. Services of dwellings 11. N et enterprise product a t factor costs 12. Indirect taxes net of subsidies 13. Net enterprise product at market prices II. Public Sector 14. Services of public employees (including social security) 15. Net national product a t factor cost 16. Net national product at market prices

1938

1946

1949

9.0 27.0 68.5 22.0 17.0 81.0 49.0 25.0 6.0 13.0

10.0 30.5 57.5 17.5 18.5 58.5 36.0 23.0 4.0 11.0

12.0 36.0 68.0 27.5 25.5 83.0 48.0 26.0 6.5 12.0

317.5 29.5

266.0 16.0

344.0 33.0

347.0

282.0

377.5

20.0

31.0

(a)

337.5 367.0

297.5 313.0

(a) (a)

(a) Figure in 1938 francs not available. sou rce:

Rearranged from References No. 1 and No. 5, Supplement, p. 633.

27. This item was taken from governm ent and social security accounts. Some difficulties w ere involved in estim ating the com pensa­ tion of local governm ent em ployees because of the incom plete nature o f the data relat­ ing thereto and also in estim ating the in­ com e in kind of all categories of public employees. T o obtain the national incom e at factor cost, the foregoing item was adjusted fo r the net factor incom e from abroad— an item th at m ay be either positive o r negative. It was com puted from the current figures of exports and im ports and of exchanges of so-called invisible item s, such as tourist ex­ penditures, shipping services, etc. T he fig­ ures of exports and im ports w ere rath e r re ­ liable. T hose on invisible item s, on the other hand, w ere n o t com plete o r altogether ac­ curate.

2. Estimates by the Income-distrib­ uted M eth o d 4 In the preparation o f the first postw ar series of F rench estimates, the incomedistributed m ethod played a subsidiary role. T he figures w ere derived from curren t fiscal and business accounting data: The incomes from salaries and wages paid by enterprises were calculated on the basis of the statistics of the Ministry of Labor. These, in turn, were based on quarterly sample in­ quiries into average wage rates in different trades and occupations and on the estimated number of persons engaged in them. As stated in the OEEC report, however, these figures were not very reliable inasmuch as (a) the enter­ prises covered by the inquiries were not neces­ sarily representative of the particular trades; (b) not all trades and occupations were covered in the inquiries; (c) the occupational classifica­ tions used in the inquiries did not correspond to the classifications used in the population census; and (d) the population censuses were taken infrequently (1936, 1946) and did not contain sufficient details on various trades and occupations and on the positions of the work­ ers in them. Social security data, which might have given more reliable figures on salaries and wages, were not ready for use at that time. Incomes from interest and dividends were estimated on the basis of information obtained from statistics of dividends paid, records of foreign exchange control, fiscal data, govern­ ment and public enterprise accounts, and sta­ tistics of issues of government securities. Net rents were derived from gross rents as estimated under the net output method (serv­ ices of housing) after deductions of deprecia­

FRANCE

395

tion and property taxes. Income from agri­ cultural land was estimated on the basis of the net output analysis of the value and costs of agricultural production. “Except for the figures derived from the ac­ counts of government and of public enterprise,” says the OEEC report, “the estimates of in­ come from property (interest, dividends, net rent, and income from agricultural land) are only approximate.” 5 Fortunately, incomes from property were comparatively small and, there­ fore, any inaccuracies in these estimates could not affect the national income total too seri­ ously. Undistributed corporate profits were esti­ mated on the basis of profits tax statistics. They were computed net of depreciation as reported on the tax returns. No attempt was made to adjust the figure for the difference between fiscal depreciation and the generally larger theoretical depreciation. The accuracy of the computation of undistributed profits was, therefore, questioned by the OEEC report. It was believed that this item of income was over­ stated. Entrepreneurial income was calculated as the residual item in the gross domestic enterprise product (as adjusted for net income from abroad, public debt interest, and government transfers to enterprises) after deducting all other component incomes and charges. As a residual item, the estimate of entrepreneurial income was, therefore, affected by all the errors made in estimating the value of the total enter­ prise product and of all its other components. According to the OEEC report, it was, there­ fore, the weakest item in the estimates. The components were next added, as shown in Table 27-2. Personal income, consisting of factor and nonfactor paym ents to individuals, was esti­ m ated on the basis of business and govern­ m ent accounts and of the international bal­ ance of paym ents statem ent. Its disposition was shown in term s of direct taxes, contri­ butions to social security, private consum p­ tion expenditures, and savings. T he weakest elements in this estim ate w ere the last two items.

3. Estimates by the Final Expendi­ ture or Products Method 6 In the F ren ch postw ar estimates, par­ ticular em phasis was placed on the final products or expenditure approach. Even though the estimates prepared by this m ethod were n ot as precise as those by the value-added and incom e-distributed m eth­ ods, they did serve as checks on the other calculations as well as provide im portant

396

THE INCOME OF NATIONS

Table 27-2. National Income of France by Factor Shares, 1938-50 1951 Estimate (billion current francs) 1938

1946

1947

1948

1949

1950

180 150 30

1,125 859 266

1,676 1,307 369

2,796 2,226 570

3,408 2,678 730

3,755 2,840 915

Income from property and entrepreneurship 180 16 a. Undistributed profits before taxes 50 b. Interest, dividends, rent, etc. 11 c. Income from property 123 d. Entrepreneurial income -2 0 e. Interest on public debt

1,471 173 51 47 1,242 -4 2

1,627 151 65 57 1,402 -4 8

2,634 377 23 164 2,138 -6 8

3,165 330 122 222 2,574 -8 3

3,568 640 177 235 2,624 -1 0 8

360

2,596

3,303

5,430

6,573

7,323

Compensation of employees a. in private and public enterprise b. in government (civil and military)

Total sou rce:

Rearranged from: Reference No. 2, pp. 139-40.

inform ation on national expenditures. T he m ain objective of these estim ates was to evaluate the gross national product of France, classified by final products at m arket prices. W ith the aid of data obtained from the other two estimates, this aggre­ gate was next reduced to net national in­ come at factor cost. First, we will consider the m ethod of esti­ m ating the final product of the enterprise sector; next the addition of the product of the public sector and net factor incomes from abroad necessary to obtain gross na­ tional product; and finally, the com parison and reconciliation of these estim ates with the net output and incom e-distributed esti­ mates. a. F in a l P ro d u c t o f th e E n te rp rise S ec to r T he final product of the enterprise sector consists of: (a) all com pletely finished do­ mestically produced consum ption and pro ­ duction goods and services requiring no fu r­ th er processing and ( b ) all m aterials, raw o r in process, w hich are added to inven­ tories or are exported during the period. They were evaluated at m arket prices. T heir total value was next reduced by the value of interm ediate goods and services im ported during the period. T he final goods and serv­ ices were classified into ten groups: 1. 2. 3. 4. 5. 6.

Food O ther consum ers’ nondurables Services of dwellings Clothes, textiles, and leather goods O ther consum ers’ durables Producers’ m echanical equipm ent and transport equipm ent 7. C onstruction and public w orks

8. D efense equipm ent 9. Changes in stocks of unfinished goods 10. Exports of unfinished goods For most of the foregoing groups, estimates were first prepared for 1938 at the prices of that year on the basis of comprehensive sur­ veys conducted during the occupation period. They were next extrapolated to later years with the aid of production and price indexes. Only for foods and for a few types of manufactured goods was it possible to base the estimates on direct current production and sales data. Ex­ ports of unfinished goods were also estimated on the basis of direct current data. For most types of food, the quantities pro­ duced or sold were obtained from trade organi­ zations and multiplied by their respective average prices prevailing during the period. The valuation was necessarily crude, inasmuch as it was impossible in computing average prices to take account of, and to weight accu­ rately, all the variations in prices as between different regions of the country and periods of the year. The volume of output of other consumers’ nondurables, such as tobacco, fuel, recreation, medical services, travel, etc., was estimated in some cases on the basis of reported production figures and in others on the basis of studies of family budgets and government expenditures. In some instances, 1938 output was taken and extrapolated to the current years by means of production indexes, and was evaluated in 1938 prices deflated into current prices by means of price indexes. Services of dwellings were evaluated at gross rents in effect in 1938 (which included taxes and maintenance costs) and in the case of owner-occupied dwellings, at corresponding im­ puted rents. These rents were next adjusted to current conditions by a rent-price index, after a further adjustment for changes in the number of dwellings resulting from war damages, re­

27. construction and new construction. It was ex­ plained in the OEEC report that “since the number of dwellings and the rent scales have changed relatively little since the war, because of the official rent control, this method gives a fairly reliable estimate of the official value of housing services for the immediate postwar years.” It was admitted, however, that this type of computation did not take account of the “incomparably higher prices paid by all subtenants” or of the “exorbitant droits de reprise which are generally paid by the new tenant to the previous tenant, or to the house agent,” neither of which are subject to rent controls.7 The gross values of output of clothes, tex­ tiles, and leather goods were computed from the figures of sales in 1938 reported at that time by trade organizations. These were next adjusted for changes in output since 1938 with the aid of so-called “activity indexes” (showing the number of working hours used in produc­ tion each year) prepared by the Commission for the Modernization of the Textile Industry and further adjusted to current prices by the use of price indexes. In the case of some goods, such as fabrics, yarns and ribbons, linen goods, shoes, gloves, etc., separate price indexes were used, while in the case of others (clothing, tailors’ services, hosiery, etc.), a common clothing industry activity index and a common clothing price index (weighted for each com­ ponent trade) were applied. The reliability of this part of the estimate was declared by the OEEC report to be subject to the following possible errors: “ (a) the unreliability of sales proceeds declarations, ( b ) the incompleteness of the activity indexes, which do not take changes in productivity into account, and (c) the uncertainty of price indexes, which increase the disadvantages of the average price concepts . . . particularly in the case of goods and serv­ ices of very different and varying quality.” 8 The value of other consumer durables, such as furniture, pottery and other household uten­ sils and equipment, clothes, watches and jew­ elry, private cars, bicycles, radios, etc., was esti­ mated by methods very similar to those em­ ployed in the case of the clothing group. How­ ever, an additional complication presented itself in these cases, in that it was impossible for most of them to determine and segregate the proportion employed in business, as against that employed for personal consumption. Only in the case of automobiles and lorries was it pos­ sible to do so. As a consequence, this group of goods was probably overestimated. The value of producers’ mechanical and transport equipment— an item of gross capital investment—was estimated on the basis of the values reported for the year 1938 by the com­ pulsory trading organizations established by the German occupation government. Inasmuch as these organizations allocated raw materials to

FRANCE

397

producers, their information was generally ac­ curate. These reported values were extrapolated to later years, in some cases by means of the activity indexes developed by the Commission for the Modernization of the Agricultural and Machinery Tool Industries and in others on the basis of the allocation of steel and power to these industries as well as on pertinent produc­ tion indexes. Finally, these values, expressed in 1938 prices, were converted into current prices with the aid of price indexes. Because of the crudeness of the activity and price indexes em­ ployed in the preparation of these figures, they were considered to be of low reliability. The value of private construction and public works—similarly a component of gross capital investment—was also derived from the 1938 figures, extrapolated and deflated as in the cases just described. The OEEC report said they were likely to include costs of repairs and re­ placements of all kinds that should not be in­ cluded among final products. The value of defense equipment was taken directly from gov­ ernment reports and represented one of the most reliable components of the estimate. Next, changes in the amount and value of inventories of unfinished goods were estimated. These figures were especially weak, as they were not based on any direct information but were calculated as a certain fixed proportion of the output of finished production goods in each year. These figures were next corrected in accordance with prevailing business opinion as to what changes in current inventories had most likely taken place during the year. T he sum o f all the foregoing items rep­ resented the domestically produced final product of the enterprise sector. This total was next adjusted fo r exports and im ports o f unfinished goods. T h e value o f exported raw materials and unfinished products, not included am ong any o f the foregoing cate­ gories o f finished products (e.g., hides, raw textiles, various metals, tim ber, coal, oil, etc.), was established on the basis o f cu r­ rent foreign trad e statistics. T he resulting estim ate was considered to be fairly accu­ rate, allowing fo r the usual m argin o f un­ derstatem ents of exports in countries having exchange controls. T h e value o f im ported unfinished products, a deductible item , was sim ilarly derived from foreign trad e statis­ tics and was taken c.i.f. (cost, insurance, and freig h t). E xcept fo r possible errors of classification, it was deem ed to be fairly correct. Finally all the foregoing items were added up (except fo r the last one th a t was de­ ducted) to obtain “the gross product of enterprises.”

398

THE INCOME OF NATIONS

Table 27-3. French Gross National Product by Final Products, 1938-49 1951 Estimate (billion current francs) 1. 2. 3. 4. 5. 6. 7.

Item Food Consumers’ nondurables Services other than housing Housing (gross rents) Clothing, textiles and leather Other consumers’ durables Armament

8. Domestic production of consumption goods and services (items 1 to 7) 9. 10. 11. 12.

Agricultural equipment Industrial equipment Transport equipment Building and public works

1938 137 391 651 28 46 31 10

1946 1,125

1947 1,635

1948 2,118 (522 1872 64 980 563 50

1949 2,500 750 1,130 150 1,180 670 70

526

705

58 410 224 27

43 * 465 295 40

365

2,370

3,183

5,169

6,450

133

83 292 277 535

360 330 630

110

15 | 12 25

112 166

| 36 1175 126 241

13. Domestic production of capital goods (items 9 to 12)

57

411

578

1,187

1,430

14. Domestic production of finished goods and services (items 8 to 13)

413

2,781

3,761

6,356

7,880

12 1

28 57

43 13

138 80

210

17. Domestic production of final goods (items 14 to 16) 18. Less intermediate goods and services imported

426 -2 9

2,866 -1 2 2

3,717 -214

6,574 -5 3 1

8,170 -5 8 0

19. “ Production nationale brute” (gross product of enter­ prises) (items 17 and 18)

397

2,744

3,603

6,043

7,590

20. Depreciation of government assets 21. Government services 22. Net income from the rest of the world

5 30 8

38 266 8

45 367 3

85 570 41

95 730 1

23. Gross national product (items 19 to 22)

440

3,056

4,018

6,739

8,416

5!

15. Intermediate goods exported 16. Intermediate goods stocked

80

8 This abnormally low figure can probably be explained by the omission of real estate taxes; the neces­ sary adjustment could not, however, be made in this report. s o u r c e : Reference No. 2, p p . 141-42. b. G ross N a tio n a l P ro d u c t a nd E x p e n d itu re T o obtain the gross national product as generally defined, tw o item s had to be added to th e “gross product of the enterprises” : (a) gross incom e from the public sector, defined as com pensation of governm ent and social security employees and depreciation of governm ent assets, and ( b ) net factor incomes originating in the rest of the world. T he final structure o f this com putation is presented in Table 27-3.

4. Degree of Reliability of the 1947—51 Estimates N ext, com parisons w ere m ade of the results obtained by the final expenditure

m ethod w ith th e results obtained by the value-added (o r n et o u tp u t) m ethod. It was fo u n d th a t these com parisons w ere only p artly valid and could serve only as p a r­ tial cross-checks o f th e correctness o f the figures because (a ) “fo r some goods and trades . . . [the two m ethods] cover exactly th e sam e ground and use the same sources of inform ation” 9 and (b ) th e tw o m ethods account fo r depreciation in a very different m anner. T heoretically, it m ay seem th a t if the figures u n d er the tw o m ethods w ere cor­ rectly com puted the difference betw een them should equal the sum of indirect taxes (less subsidies) and depreciation, inasm uch as the net ou tp u t figure excluded these items while the final product figure included them. B ut actually this was n o t the case because

27. depreciation u nder th e net output m ethod was n o t com pletely deducted in all cases. W herever the values added w ere obtained by adding up th e facto r shares directly, it was “fiscal depreciation” rath e r th an “theo­ retical depreciation” th a t was excluded from the total. A s fiscal depreciation was gen­ erally sm aller th an theoretical depreciation (defined as the annual am ount of provision “theoretically necessary to m aintain the as­ sets in tact” ) , the resulting net o utput total did include some depreciation and was not com pletely net. A fter exam ining this difficulty, the O E EC rep o rt concluded th at “w hen the sources of inform ation used are different, a com parison of the results of the two m ethods does p ro ­ vide at least a rough check, th e m argin of erro r being equal to the difference between total theoretical and to tal fiscal deprecia­ tion. Since in any case this difference does n o t exceed 10 per cent of gross enterprises’ product, it can be taken th a t to the extent th at th e sources of inform ation used in the

399

FRANCE

two m ethods are n ot affected by the same factors of deform ation, acting in a similar direction, it should be possible to obtain the gross p roduct of enterprises w ithin a margin of accuracy of some 10 per cent. . . 10 T he O E EC rep o rt m ade the following pertinent observation regarding the relative reliability of the value-added (n et output) and final products estimates: “G enerally speaking, the values-added m ethod is con­ sidered to be the m ore accurate. T he con­ cept of final goods cannot be determined statistically w ith any degree of certainty. F urtherm ore, m ore is know n of variations in the production of the interm ediate sec­ tors, w hich generally consist of highly con­ centrated enterprises often belonging to the nationalized sector, th an of variations in the final sectors.” 11 T he O E E C study attem pted to grade the relative reliability of different com ponents of these estimates separately, ranking them as follows: a, fairly accurate estim ate; b, approxim ate; c, tentative; or d, bad esti-

Table 27-4. Relative Reliability of the Components of the French Estimates for 1949 as Prepared in 1950-51

Items

Rating as to Reliability

Approximate Per Cent Weight in the Total Gross National Product

b/c

100

A. Elements of Gross National Income Gross national income Transfers from abroad Gross national product Depreciation N et national product a t market prices Indirect taxes net of subsidies

b/c d c a

National income a t factor cost Factor income from abroad

c c

80 0

N et domestic product Direct taxes on corporate income Government transfers to enterprises Compensation of government employees Employers’ contributions to social security Compensation of enterprises’ employees Undistributed income Income from property and entrepreneurship

c a a

80 1 —4

B. Elements of Gross National Expenditure Private consumption Government consumption Gross addition to fixed assets Change in inventories Net lending to abroad Total C. Elements of Saving Government saving Household saving

e

0

100 10 90 10

a

8

a b c c

6

30 5 34

75 13 14 2

-4 100

400

THE INCOME OF NATIONS

mate. Table 27-4 shows this grading system in application.

II. Estimates Prepared in

1952-55 T he estimates prepared in 1952-55 by the Econom ic and Financial R esearch U nit in the M inistry o f Finance (in co-operation w ith the N ational Institute of Statistics) at first relied on all three estim ating m ethods, b u t eventually were developed prim arily by the incom e-distributed and final products m ethods. Accordingly, the discussion in this p art of the chapter is divided into four sec­ tions: (1 ) estim ates prepared in 1952-53 by the net output m ethod; (2 ) those p re­ pared during the same period by the incomedistributed m ethod; (3 ) those developed during the same period by the final expendi­ ture o r products m ethod; and (4 ) further im provem ents of estimates in 1954 and 1955.

1. Estimates Prepared by the Net Output M eth o d 12 T he net output estimates prepared in 1952-53 differed from the earlier ones in two m ajor respects: (1 ) as already pointed out, they were based on current fiscal data instead of on projections of the 1938 esti­ mates, and were consequently m uch m ore accurate; and (2 ) they w ere prepared in two stages: first, by calculating the value of the national output at m arket prices; and second, by converting this aggregate into a value at factor cost, i.e., free o f indirect taxes (less subsidies). T he F rench economists use the term “value added” in the sense in w hich this term is em ployed in the F rench tax law, i.e., as including the am ortization of capital equipm ent. Thus, the estim ate fo r 1949 spe­ cifically s a y s :13 . . the value added of an enterprise equals the difference between the value of the goods and services which it has sold or has added to its inventories and the value of the goods and services w hich it has bought during the given year, all capital equipm ent goods being excluded from the purchases . . . it is also equal to the sum of the incomes and indirect taxes derived from the enterprise after deduction of the subsidies received.” It is in this sense that the term is used in this section.

T he value added was com puted mainly on the basis of the records of assessments an d collections of tw o indirect taxes— the production tax (w hich was transform ed in 1954 into “the value-added tax ” ) and the transactions (o r tu rn o v er) tax. In some cases subsidiary inform ation was also ob­ tained from the records of th e assessments an d collections of direct taxes, particularly incom e taxes. T he enterprise sector was divided accord­ ing to th e type of data available fo r esti­ m ating purposes into three subsectors: (1 ) private industrial, agricultural, and com m er­ cial activity covered by the production tax; (2 ) other branches of production n ot cov­ ered by either the production or the trans­ actions tax and (3 ) trade, covered by the transactions tax. a. V a lu e A d d e d in P riva te In d u stria l a n d A g ric u ltu ra l P r o d u c tio n 14 A fter 1948, the records of the production tax furnished rich data on values added, inasm uch as the tax was levied on the values added at different stages of production in a firm rath e r th an on its total value added, as before. A s a consequence, it provided a fairly consistent p icture of the structure of the value added in the econom y. T he total assessed value subject to the tax was 5,100 billion francs in 1949. In estim ating from it the value added, account had to be taken of certain differences betw een the con­ cept o f assessed value added and the con­ cept of econom ic value added; and the fol­ lowing adjustm ents w ere m ade:

a.

b. c. d. e.

Billion Deductions from Assessed Value Francs Nondeductible expenses of produc­ tion under the old production tax, including outlays for equipment and materials not physically used up in the production process (estimated on the basis of a representative sample a t 9 per cent of the gross sales) —950 Purchase of coal, oil and other ma­ terials by the transport industry —170 Consumption of goods and beverages a t hotels, restaurants, and bars by their owners and employees —130 Costs of certain financial and com­ mercial services not deducted from assessed value —135 Imports assessed under the tax —570 -1 ,9 5 5

27.

/.

g.

h. i. j.

FRANCE

401

Billion Additions to Assessed Value Francs Wholly exempt value added Bread +110 Milk, eggs, and other foods +430 Consumption by farmers of own foods +310 Partially exempt value added in util­ ity services (gas, electricity, water), publishing business (books, news­ papers, etc.), naval construction, and services of artisans +315 Tax-exempt exports +780 Difference between assessed and real value added of the period due to the time lag in assessments +20 Difference between changes in as­ sessed value of inventories and changes in their real value during the year +90

w as calculated at 13,530 billion francs. To this total were added the am ounts of exempt transactions, corrections for end-of-year lag in assessments and collections of taxes, transactions of the Saar region, and unde­ clared transactions representing tax evasion (estim ated at 8 per cent of the to ta l); and on the other hand, a deduction was m ade for double counting. T he total value of transac­ tions in trade, thus adjusted, am ounted to 17,480 billion francs. T he gross value origi­ nating in trade was com puted from this figure by deducting the value of goods and services entering, o r furnished to, trade. The net value added in trad e was in tu rn com ­ p uted from the gross value (4,420 billion francs) by identifying its elements and cal­ culating them directly as follows:

+ 2,055

a. N et profits in trade, according to income tax statistics, at 8 per cent of the sum; b. Salaries of employees in trade, according to the same source, at between 6 and 8 per cent; c. Taxes, amortization, and maintenance of equipment, at 6 per cent.

T he total deductions and additions for 1949, as seen above, nearly canceled each other out. T he corrected value added was 5,200 billion francs. b. V a lu e A d d e d in O th er B ra n ch e s o f P ro d u ctio n 15 F o r the branches of production n o t cov­ ered by the production tax, such as public enterprise, liberal professions, ow nership of real estate, and certain financial businesses and domestic services, the value added had to be estim ated on the basis of other fiscal data. T h e resulting figures of value added in these activities fo r 1949 w ere as follows:

Public enterprises Liberal professions Ownership of real property Domestic services Interest earned by enterprises on their in­ vestments in obligations of other busi­ nesses Total

Billion Francs 375 100 320 55 20 870

c. V a lu e A d d e d in T r a d e 16 T he value added in trade was estim ated on the basis of data derived from the rec­ ords of the transactions tax. T he transac­ tions tax in 1949 was levied in m ost cases at the rate of 1 per cent; on com paratively few types of transactions at 1.8 per cent; and on still few er types at 2.75 per cent. T he total value of taxable transactions

The total value added in trade was accord­ ingly estimated at 21 per cent of the gross value originating in trade, or at 928 billion francs (4,420 X 21)/100. d. C o rrectio n s fo r T a x E v a s io n 1'' T he estim ators conceded th a t the extent o f tax evasion and frau d varied considerably from industry to industry, from location to location, and from firm to firm; and th at it defied exact calculation fo r individual in­ dustries. B ut on the other hand, they be­ lieved it possible to estim ate the extent of tax evasion fo r the enterprise sector of the econom y taken as a whole. O ne im portant facto r facilitating this task was the very m ultiplicity of the F rench taxes, w hich af­ forded different approaches to the correct evaluation o f the same taxable object. The general procedure in estimating tax evasion involved two stages: first, it was as­ sumed that there was no evasion in the pay­ ments of the value-added tax, and the total transactions were calculated from the base of the tax; second, it was conceded that there were evasions of the transactions taxes, and their extent was determined by comparing the as­ sessed transactions with the transactions esti­ mates from the value-added tax. A deduction was made for evasion under the income tax, and a correction was made to take account of the relatively negligible evasion of the valueadded tax that initially was assumed to be nil. Evasion under the value-added (payrolls, pur­

402

THE INCOME OF NATIONS

chases of materials, etc.) were comparatively large and subject to close fiscal control. The actual evasion under the value-added and income taxes was determined from the records of reassessments by the tax auditors. A survey of 40,000 taxpayers revealed under­ statements of value added ranging, on the aver­ age, from 4 to 5 per cent.18 e. V a lu e A d d e d in the Saar a n d in the P u b lic a n d Social S ec u rity S ec to rs T he value added in the enterprise sector of the Saar region and the incom e from the public and social security sectors were calculated directly from the d ata given in official reports. T he incom e originating in the public sector, as already stated, was deem ed to consist only of com pensation in cash and in kind paid to governm ent and social security employees.

.

2 Estimates Prepared by the Income-distributed. M ethod 19 T he estimates prepared in 1952-53 by the incom e-distributed m ethod w ere based on direct current fiscal data. In the case of incomes paid out by private enterprise, they were based on the statistics of personal in­ come and profits taxes, and in the case of those paid by public enterprises and by the public sector, they w ere based on govern­ m ent accounts. Salaries and wages were estimated mainly on the basis of the income taxes, taking into account the differences between the periods for which they were assessed and the time in which they were collected, as well as the deductions of pension contributions. The payrolls in public enterprises, however, were taken directly from public accounts. Employers’ contributions to social security, family allowances, and acci­ dent insurance were calculated from social se­ curity accounts. The figures on indirect taxes were taken from official reports. Data on amortization were available for public enterprises, for most of the power in­ dustries, and for certain of the transport in­ dustries. For other industries, use was made of the results of a special study of existing amortization provisions. This study by the Bu­ reau of Direct Taxes, made in 1949, covered 6,803 enterprises having total gross sales of 1,482 billion francs and payrolls of 213 billion francs. Their amortization deductible for tax purposes in 1949 amounted to 34 billion francs. On the basis of this figure, amortization pro­ visions for the entire private enterprise sector including private buildings were calculated at 480 billion francs. Provisions for business losses

and other charges were estimated on the basis of tax data. The costs of maintenance o f plant and equipment were estimated for certain en­ terprises at 33 per cent of amortization and for residential buildings at 25 per cent of gross income. Incomes o f farmers could not be estimated from assessed incomes because most farm in­ comes were either specifically exempt from the income tax or otherwise escaped payment of the tax. They were estimated on the basis of special studies of farm incomes conducted by the National Statistical Institute. Incomes from ownership o f land and build­ ings related mainly to those derived from in­ dividual ownership. Those of corporations were included only if they had not been accounted for already in the profits of the corporations. These estimates were largely derived from the records of the assessed values of urban and rural properties. Profits o f industrial, commercial, and pro­ fessional businesses were estimated on the basis of income tax and profit tax records, with due allowance made for profits not included in the assessments. Among the latter were profits be­ low the taxable amounts; profits exempted from these taxes for a variety of reasons, such as desire to encourage investments of profits; profits absorbed by previous losses; profits of tax exempt public enterprises; and concealed profits. The aggregate of all these profits had to be reduced by the amount of losses sustained by some firms. The gross income originating in the Saar region was calculated from the official documents of that region. A ll the foregoing item s w ere then added, and from the resulting total (7,462 billion francs) a fu rth er deduction was m ade for governm ent subsidies and fo r incomes re­ ceived from abroad. T he resulting total (7,237 billions) represented the gross n a­ tional p ro d u ct o f the enterprise sector as estim ated by the incom e-distributed m ethod. By adding incom e fro m the public sector and elim inating indirect taxes n et of subsi­ dies an d provisions fo r depreciation, n et national incom e at facto r cost was obtained.

3. Estimates Prepared by the Final Expenditure or Products M ethod 20 T h e estim ates m ade by th e expenditure m ethod were considerably im proved in 1952 and 1953. C learer distinctions were draw n betw een final and interm ediate goods, better production and price indexes were used in extrapolation, and m ore direct cu rren t data w ere used. T he extrapolations w ere still

27. m ade from the 1938 totals, but steps were taken tow ards developing enough direct cu r­ ren t data fo r 1950 to m ake possible an even­ tual shift to the latter year as a base for extrapolation. It was realized, how ever, that the infrequency of the production censuses and the inadequacy of the annual consum p­ tion and sales data w ould necessitate the use o f extrapolation fo r a long tim e to com e in estim ating gross national product and national incom e by the expenditure m ethod. First, the earlier estimates fo r 1949 w ere revised. T he figures expressed in 1938 francs w ere converted into curren t francs and foods w ere evaluated directly from current data. T able 27-5 shows the production and price indexes th a t w ere em ployed.

403

FRANCE

H aving estim ated the gross product of enterprises for th at year by the final product m ethod at 7,609 billion francs, the esti­ m ators proceeded to correct certain com po­ nents of the total th at were obviously in­ accurate. Using current fiscal and produc­ tion data, they m ade the following correc­ tions, reducing the foregoing am ount by 186 billion francs: a. The market value of investment goods (equipment, construction, and public works) was raised 119 billion francs. b. The following values were added: 82 bil­ lion for values added to imported goods in grading, repacking, and otherwise pre­ paring them for marketing; 247 billions for exports of raw materials; and 60 bil­

Table 27-5. Final Products of the Enterprise Sector of France at Market Prices, 1949 1952-53 Estimates (billion francs) 1949

1938 Products in 1938 Prices

Production Products Index in 1938 1938 = 100 Francs

Price Index 1938 = 100

Consumer goods and services 1. Foods 2. Industrial consumer goods a. Nondurable b. Clothing c. Other

40.2 47.5 31.1

110 92 125

44.2 43.7 38.8

1,675 2,560 1,775

Total 3. Services Dwellings Other

25 64.7

92 112

23 72.5

500 1,600

Total services Total consumer goods and services Investment goods 1. Equipment a. Agricultural b. Industrial and commercial c. Transport d. Armaments

222.2

208.5 5 14.9 12.4 10

102 148 1,700 41

5.1 22.1 20.7 4

1,690 1,680 1,460 1,750

Total equipment 2. Construction and public works Total investment goods Total final products source:

Reference No. 4a,

p.

19.

25

149

37.3

1,700

Products in Current Francs

Adjusted Value in Current Francs

2,549

2,520

740 1,120 690 2,550

2,412

115 1,157

115 1,114

1,272

1,229

6,371

6,161

85 371 302 70

85 363 300 70

828

818

630

630

67.3

89.2

1,458

1,448

275.8

311.4

7,829

7,609

404

THE INCOME OF NATIONS

lions for additions to inventories of raw materials. c. Deductions were made of 592 billion francs for imports of raw materials; 48 billions for foreign freight charges on imports; 24 billions for values added to domestic products by foreign services; and 30 billions for purchases from the public domain. T o this corrected sum of 7,423 billion francs an addition of 150 billion francs was m ade for incom e originating in the Saar region, making a total gross national product of enterprises of 7,573 billion francs.21 This total was next com pared w ith the one of 7,510 billion francs developed by the value-added m ethod and the one of 7,237 billion francs arrived at by the incomedistributed m ethod. T he last figure was found to be m uch too low. Some of its com ponents w ere derived from data belong­ ing to 1948, a year when incom es and prices were lower, rath er th an to 1949. T he figure also did not include the profits of taxexem pt enterprises, particularly of agricul­ tural co-operatives. T he choice appeared to be between the two other totals. O f these, the one based on the value added o r net output m ethod, e.g., 7,510 billion francs, appeared m ore reliable. Its accuracy was supported by the statistics of gross sales in some 99 industrial divisions, w hich the Bu­ reau of the M inistry of Finance had inde­

pendently developed fo r 1949. Although these statistics w ere still incom plete, cover­ ing only 60 per cent of the firms and 70 p er cent of the sales, they were nonetheless m ost illum inating, particularly w hen the 99 divisions were consolidated into 10, thus re­ ducing the possibility of error. In addition, a variety of other statistical sources were used fo r the verification of these figures: fo r agriculture, special studies of the N a­ tional Institute of Statistics and of Econom ic Studies and various fiscal reports on agri­ cultural sales and agricultural incom es; for pow er and transport, reports of pertinent trade associations; fo r others, a variety of sources bearing on payrolls, taxes, profits, m arkups, percentages of purchases, etc. T he total sales o f these 10 divisions of the enter­ prise sector, on the basis of these data, ap­ peared to aggregate 17,480 billion francs and their purchases, 9,980 billion francs. T he difference of 7,500 billion francs repre­ sented “gross p roduct of the enterprises.” A ccordingly, the estim ators settled on the figure of 7,500 billion francs as the value of th e gross p roduct of enterprises (including those in the Saar region) fo r 1949. T his gross p roduct was next broken down by industrial divisions and by types of in­ com e paym ents and charges w ith the aid of corrected data obtained by the value-added an d incom e-distributed methods, as shown in T able 27-6 below.

Table 27-6. Gross Product of Enterprises of France in 1949 by Industrial Divisions and Types of Income Payments and Charges (billion francs) Indirect Taxes 1. Agriculture, and agricul­ tural food industries 2. Power 3. Transport 4. Mechanical industry 5. Chemicals 6. Textiles 7. Construction and public works 8. Miscellaneous industries 9. Commerce and services 10. Public enterprises Total, France 11. Saar Total, France and Saar sou rce:

Salaries Social Amorti­ and Wages Security zation

Mainte­ nance

Profits

Total

100 120 65 200 60 145

250 129 245 390 65 220

48 42 72 100 16 56

83 62 40 114 46 71

34 20 25 35 10 18

880 4 -6 0 251 73 236

1,395 377 387 1,090 270 746

65 45 353 6

216 105 440 150

55 27 96 8

48 26 93 27

14 7 27 10

122 54 907 174

520 264 1,916 375

1,159 23

2,210 61

520 13

610 11

200 4

2,641 48

7,340 160

1,182

2,271

533

621

204

2,689

7,500

Reference No. 4a, p. 26.

27.

405

FRANCE

T he final steps in the revision of the esti­ m ates fo r 1949 consisted of (a ) calculating gross national incom e by adding incom e from the public sector and net incom es from abroad to the product of the enterprise sector; ( b ) breaking down gross national incom e into net national incom e at factor cost and n et national incom e at m arket

prices w ith the aid of the n et output and incom e-distributed estimates and (c) seg­ regating the com ponents of gross national incom e by type of expenditure into con­ sum ption and investment. These final sum ­ m ations and breakdow ns are shown in T able 27-7. H aving revised the earlier estimates for

Table 27-7. French Gross and N et National Income, 1949, by Types of Payments, Charges, and Final Expenditures 1952-53 Estimates Billion Francs I. Net and gross national income Income of factors, originating in the national territory a. Salaries and wages Enterprises Public sector b. Contributions of employees to social security c. Incomes from property and enterprises State (net of interest on public debt) Individuals Rest of the world d. Undistributed business incomes e. Direct taxes on enterprises

2,271 661 554 81 2,227 —16 363 362

Total internal factor income Income of factors originating abroad a. Salaries b. Incomes from property and enterprises

6,503

National income a t factor costs Indirect taxes Subsidies

6,539 1,146 —191

Net national product a t market prices Amortization and other provisions for maintenance of capital

7,494 825

Gross national product a t market prices Transfer to rest of the world

8,319 —10 *

Gross national income received II. Gross national expenditure Private consumption Current purchases of goods and services by individuals Public consumption (services of government and social security employees) a. Compensation of employees in the national territory b. Compensation of employees abroad and in the colonies

36

8,309 5,542 564 118

Total public consumption Gross domestic capital formation a. of enterprises b. of state c. increase in inventories

1,380 134 300

Total gross domestic capital formation Net income from abroad

1,814 —43

Total gross national expenditure

8,309

a Does not include foreign aid received as gifts. sou rce:

Reference No. 4a, p. 82.

996

406

THE INCOME OF NATIONS

1949, largely on the basis of 1949 statistical data, the estim ators proceeded w ith the preparation of a consistent series fo r the years 1950-53, partly by extrapolating the 1949 bench-m ark figures and partly by de­ riving them from current data. W e are not reproducing here these series inasm uch as they were subsequently replaced by new series shown in the section below.

4. Further Improvements of Esti­ mates in 1954 and 1955 22 In 1953 and 1954 further im provem ents in the estim ating m ethods were m ade and some of the previously developed figures w ere revised. T he estim ate of gross national product ( “produit national b ru t” ) fo r 1949 was raised in the 1954 publication from the 8,319 billion francs previously calculated to 8,480 billion francs, and national incom e at factor cost was sim ilarly raised from 6,539 billions to 6,640 billions.23 T he principal changes w ere in the following item s: (a ) the value of certain finished products, p ar­ ticularly textiles, was substantially raised; ( b ) certain exports erroneously classified before as finished products were now classi­ fied as interm ediate products; (c ) certain goods form erly classified as finished in the additions to inventories were now reclassi­ fied as raw m aterials and unfinished goods. Similar recalculations of the distributed in­ comes resulted in (a) increases in the esti­ m ates of salaries and wages to take account m ore adequately of em ployers underreport­ ing their payrolls on tax returns; ( b ) addi­ tions to salaries of certain social charges; and (c) reclassification of profits of tobacco and m atch m onopolies as indirect taxes. T ow ard the end of 1955, the Research U nit of the M inistry of Finance, in collabo­ ration with the N ational Institute of Statis­ tics and Econom ics, finished revising its postw ar estimates and published the results in a com prehensive report.24 This report presented a com pletely new, consistent series of estimates covering the years 1949-55 that

w ere based on cu rren t production, sales, fiscal, and accounting data. T his new series was p repared almost w holly by the final products and incomedistributed m ethods. T he net ou tp u t m ethod, w hich had been used extensively in the pre­ vious series, was no longer relied upon to the same degree. N ational incom e and gross n a­ tional p ro d u ct w ere analyzed by distributive shares, by types of expenditures and p ro d ­ ucts, an d by econom ic sectors, b u t no longer by industrial divisions of origin. D ouble­ entry social accounts w ere strongly em pha­ sized in all presentations. In fact, sector accounts w ere used as the basis of th e esti­ m ates of national incom e and p ro d u ct ag­ gregates. T he series covering 1949-55 was given both in cu rren t prices and in 1952 constant prices. T hree sets of figures w ere now p ro ­ vided: (a) prelim inary ones fo r the most recent two o r three years; (b) definitive ones, replacing th e prelim inary figures two o r three years later; and (c) prospective figures covering the forthcom ing year and related to the n atio n ’s “econom ic budget.” T able 27-8 reproduces this new series at cu rren t prices. T h e figures in constant prices are n o t reproduced here as they do not differ m uch from th e others. T h e publication o f the 1955 rep o rt seems to have com pleted the basic reorganization o f national incom e accounting procedures begun in 1947. It was a m ajor undertaking, and w ith its com pletion the F ren ch esti­ m ates have becom e, m ethodologically speak­ ing, at least, n ear equals to those of the acknow ledged leaders in the field; e.g., the U nited States, the U nited K ingdom , and C anada. T hey still rest, how ever, on incom ­ plete statistics. Since 1955, greater em phasis has been placed on projecting the accounts into the fu tu re and on preparing an annual econom ic budget fo r the nation fo r the succeeding year. In b oth the 1956 and 1957 reports, the entire second volum e was devoted to the presentation of this “econom ic budget.”

27.

407

FRANCE

Table 27-8. National Income and Gross National Product of France at Current Prices, 1949-55 1955 Estimates (thousands of billion francs) 1949 Production by enterprise sector Services of domestics Services of government Other elements of production

7.5 .1 .7 —a

1951 10.6 .1 1.1 —a

1953 12.9 .2 1.4 —a

1955 14.7 .2 1.5 —a

----

----

----

----

8.4 - .9

11.9 - 1 .2

14.5 - 1 .5

16.4 - 1 .5

7.5 - 1 .1

10.6 - 1 .8

13.0 - 2 .3

14.8 - 2 .4

6.4

8.8

10.8

12.4

3.6

5.1

6.4

7.4

2.6 .3 .1 —a

3.4 .4 .1 —a

4.0 .5 .1 —a

4.5 .6 .1

6.4 - .4 6.0 1.5 .2

8.8 - .6 8.2 2.4 .2

10.8 - .8 10.0 3.1 .1

12.4 - .8 11.5 3.2

Net national expenditure Depreciation

7.7 .9

10.8 1.2

13.2 1.5

14.8 1.5

Gross national expenditure Consumption of households Consumption of government Gross capital formation

8.6 5.7 1.1 1.8

12.1 8.1 1.8 2.2

14.7 9.8 2.4 2.5

Gross national product a t market prices Less depreciation Net national product a t market prices Less indirect taxes N et national product or income at factor cost Distribution and expenditure of national income, net and gross Labor income Net property incomes a. belonging to households b. belonging to societies c. belonging to government Transfers Total national income Less direct taxes Disposable income Taxes Transfers

a = less than 50 billion francs. so u rce:

Reference No. 4 d, vol. I,

p p . 6 7 -6 9 .

—a

—

16.4 11.0

2.4 3.0

28. YUGOSLAVIA

References 1. Dresdner Bank, Die wirtschaftlichen Krafte der Welt, Dresden, 1930. 2. V. M. Djuricic, M. D. Tosic, A. Vegner, P. Rudcenko, M. R. Djordjevii Nasa Narodna Priveda i Nacionalni Prihoda (Our National Economy and National Income), Beograd, 1930. 3. Privredni Bilten Zavoda Unapredjenje Spoljne Trgovine, No. 59, Beograd, 1938. 4. a. A. Vegner, “Das Volkseinkommen Jugoslaviens,” Weltwirtschaftliches Archiv, 1927, vol. 26, pp. 282-93. b. Otto von Franges, “Die Treibenden Krafte der weltwirtschaftlichen Strukturwandlungen in lugoslavien,” Weltwirtschaftliches Archiv, vol. 46, 1938, p. 337. c. Statistique General de l’Etat, La Yugoslavie par les Chiffres, Beograd, 1937. d. Annuaire Statistique, Royaume de Yugoslavie, 1938-39, p. 485. 5. Savezni Zavod za Statistika i Evidenciju, Methodologija za Obracun Narodnog Dohotka u 1952 Godini (Statistical Office, Methodology o f Computation o f National Incom e), Beo­ grad, 1953. 6. Statistical Yearbook of the Federal Peoples’ Republic o f Yugoslavia, 1954, Beograd, 1954, p. 109. 7. Dr. Ivo Vinski, “Valorizacja Narodnog Dohotka” (Evaluation of the National Income), Statisticke Revija, Beograd, No. 3-4, 1954. 8. , Upotreba Narodnog Dohotka 1953-1954 (Disposition of the National Income), Economski Pregled, Zagreb, 1956. 9. R. F. Byrnes, Yugoslavia Under the Communists, New York, 1957.

1. History T he first estimates of the national incom e for Yugoslavia w ere m ade by the econo­ mists of the D resdner Bank in G erm any during the late 1920’s. They covered the years 1913 and 1921-28, but w ere n o t ac­ com panied by any detailed explanations of their conceptual and methodological bases. F ar m ore com prehensive and systematic estimates were prepared at about the same tim e by a group of Yugoslavian economists — D juricic, Tosic, V egner, R udcenko, and Djordjevic— covering the average year of the period of 1923-25. In the 1930’s the Yugoslavian M inistry of Finance prepared official estimates fo r the years 1929, 1931, and 1935-38, using the D juricic et al. esti­ mates as a basis. * Acknowledgment is due to Dr. Ivo Vinski of the Ekonomski Institut, Beograd, for re­ viewing the draft of this chapter and making many valuable suggestions. Notes begin on p. 542, 408

D uring W orld W ar II, w hen Yugoslavia was occupied by th e G erm an and Italian armies, all w ork on the co untry’s national incom e estimates was suspended. M oreover, the records o f the previous official estimates were destroyed during Belgrade’s bom bard­ m ent. Fortunately, substantial descriptions of these estimates, w ritten by several of their authors, had appeared before the war in G erm an statistical publications.1 In 1937, following the re-establishm ent of Yugoslavia as an independent state w ith an enlarged territory, th e newly created F ederal Com ­ m ittee on a Five Y ear P lan fo r Econom ic D evelopm ent resum ed the preparation of national incom e estimates, publishing fig­ ures fo r the years 1947-51. In 1952, the Federal Statistical Office took over the preparation of the estim ates as a regular activity. Since then national incom e esti­ m ates have been appearing annually.

2 8 . YUGOSLAVIA

I. Prewar Estimates (Djuricic et al. and Official) 1. Concepts and Methods T he prew ar estim ators were heavily in­ fluenced by the theories and statistical ap­ proaches developed by F riedrich von Fellner in his national income estimates fo r H u n ­ gary (see C hapter 34, Section 10). Follow ­ ing his example, they adopted A dam S m ith’s concept of national income as the net value of m aterial production only, excluding from their national incom e com putations the value of the services of governm ent, professions, domestics, and dwellings. M oreover, they dealt w ith the “hom e-produced incom e” only, m aking no adjustm ents fo r the net in­ come received from abroad. In D r. V inski’s opinion,2 these estimates w ould have to be raised about 15 per cent to p u t them on a p ar w ith those based on the com prehensive production concept em ployed in m ost of the capitalistic countries. T he n et o utput m ethod was used fo r the bulk o f the estimates. Only in the case of handicraft production, transportation, com ­ m unications, and finance was th e incomedistributed m ethod em ployed. T he com puta­ tion was m ade at factor cost. N ine divisions of econom ic activity w ere distinguished: agriculture, forestry, m ining, construction, m anufacturing, transportation, com m unica­ tions, trade, and finance. Statistical materials were taken from pro­ duction and population censuses and re­ ports and from published business accounts. T he values w ere com puted fo r each year at current prices and were not com parable, inasm uch as the prices in 1938 w ere roughly one th ird low er than in 1925. T he series was converted into 1929 prices, but the estimates fo r 1925, even at constant prices, seemed to be too high: a. Agriculture, Mining, and Construction The value of vegetal production was esti­ mated by using official figures on physical out­ put, increased by 15 per cent to take account of understatements, and decreased by the amount used as seed and fodder; and multiply­ ing them by the average farm price paid at the most representative market for each product.3 It was assumed that 10 per cent of the potato crop, 30 per cent of the corn, and 90 per cent of the hay and fodder were fed to animals.4 No information was given about the sources and methods of estimating the quantity of

409

meat produced (cattle, pigs, sheep, goats, poul­ try). It seems, however, that no slaughtering statistics were available at the time. An aver­ age live weight of meat was assumed of 300 to 400 kilograms per cow and 80 to 100 kilo­ grams per pig. The average market price was reduced 15 per cent to take account of the fact that cattle was also traded outside the markets at lower prices. The quantitative pro­ duction of eggs, milk, wool, and honey was estimated on the basis of an average yearly yield per unit, such as 100 eggs per hen; 9001,000 liters of milk per cow; 100 liters per goat and 40 liters per sheep; 2.5 kilograms of wool per sheep, and 1 kilogram per goat; 3-7 kilo­ grams of honey per beehive.5 In all cases ap­ propriate average prices were applied. Further estimates were made on the basis of available statistics for the value of the output of wine, silk, and coastal fishing. From the total value of the agricultural in­ come thus obtained, the following cost items were deducted: (1) depreciation and amortiza­ tion of buildings, 6.5 per cent of their capital value; (2) depreciation and amortization of equipment, 15 per cent of its capital value, and (3) amounts paid for fertilizers, seed pur­ chased, and insecticides.6 Since the official statistics for forestry prod­ ucts referred to marketed cut timber only, in­ dependent estimate of forest products was made on the basis of 22.5 million cubic meters, or 3.2 cubic meters per hectare of forest land, of which 60 per cent was attributed to firewood. The production of the mining of soft coal, iron, and copper was reported in official sta­ tistics. From the gross value of production the value of materials used was deducted. Construction included the value of new dwellings, factories, commercial buildings, rail­ ways, roads, and bridges. Since, according to the 1921 Census of Population, each family in­ habited a house, the number of dwelling units was calculated in proportion to the increase of population. However, an adjustment was made for a lag in construction activity on ac­ count of lack of capital. Then the number of new dwellings was multiplied by an assumed average cost of building construction to obtain the value added. b. Manufacturing This industrial division covered: (a) agri­ cultural domestic industry, (b) artisans, (c) factory industry. The agricultural domestic in­ dustry was still of great importance in Yugo­ slavia at the time. It embraced, among other things, cheesemaking, distilling of alcohol, weaving of cloth and carpets, and the produc­ tion of implements, furniture, cabinets, and other wood products, either for home con­ sumption or for the market. There was no in­ dication given of the methods of estimating the value of the output of domestic industry. The net value of production of the artisans was

410

THE INCO M E OF NATIONS

estimated according to the income-distributed method by adding up the estimated earnings of the masters and workers. The gross value of the factory output, employing 260,000 work­ ers, was estimated by the Chamber of Labor at 15 billion dinars for 1924. Assuming a profit of 10 per cent and an average wage of 15,000 dinars per worker, a total income of 5.4 billion dinars was obtained for private industry, to which the net output of public enterprises (450 million dinars) was added.7 c. Transportation and Communications Profits and wage bills of railroad and navi­ gation enterprises were reported in current statistics, while the net output of cartage and other transportation activities was estimated in a rough and unspecified way. The salaries and wages paid by the State post, telegraph, and

telephone offices were published in official documents. d. Trade and Finance The net value of the services of retail trade was estimated at two thirds of the combined gross values of agriculture, mining, and manu­ facturing products and of imported goods enter­ ing wholesale trade. The profits were said to be 5 per cent of the purchase values in whole­ sale trade, and 15 per cent of those in retail trade.8 Then the salaries and wages paid were added. The wage bills and profits of the na­ tional bank, commercial banks, savings banks, and postal savings, as well as of insurance com­ panies were taken directly from available sta­ tistics. The results of the prewar estimates are shown in Tables 28-1 and 28-2.

Table 28-1. National Income of Yugoslavia by Divisions of Economic Activity a t Current Prices (billion dinars)

Agriculture, vegetal Agriculture, animal Agriculture, other, including fishing and hunting Forestry Mining Rural domestic industry Manufacturing Handicrafts Trade Banking and insurance Railroad, water, and motor transportation Carting Post, telegraph, and telephone Construction Miscellaneous Total

1925

1929

1931

1938

18.6 12.9 0.1 3.9 0.9 4.8 5.8 7.8 8.3 0.7 2.4 2.0 0.3 0.8 —

18.6 12.7 0.3 3.9 0.9 4.8 5.8 7.6 8.5 0.7 2.5 2.0 0.3 0.8 —

9.1 5.8 0.1 2.6 0.9 3.6 3.5 5.2 6.5 0.6 2.1 0.7 0.3 0.7 —

11.5 7.5 0.2 2.5 1.2 3.0 7.1 3.7 3.8 1.4 2.4 0.6 0.4 1.0 —

---

---

---

---

69.6

69.6

42.0

46.8

Reference No. 4a, p. 292.

Table 28-2. The National Income of Yugoslavia, 1923-25 to 1937 National Income Current Prices

1923-25 1929 1932 1937

Adjusted National Income 1929 Prices

Population (millions)

Total (billion dinars)

Per Capita (million dinars)

Total (billion dinars)

Per Capita (million dinars)

12.6 13.6 14.2 15.172

69.6 69.0 32.0 44.5

5.5 5.1 2.3 2.9

69.0 41.8 68.4

5.1 2.9 4.5

CostofLiving Index (1929: 100)

100.0

76.6 65.1

s o u r c e : Royaume de Yugoslavie, Statistique General de l’E tat, Annuaire Statistique, 1938-39, p . I l l ; Reference No. 4 a, p . 293; Reference No. 46, p . 337; Statistical Year Book of the League of Nations, 1941-42, p . 198. (Index number of the Chamber of Labor for three cities.)

2 8 . YUGOSLAVIA

11. Postwar Estimates The postw ar estimates w ere based on the M arxian concept of national incom e as de­ veloped in the 1940 publication of th e Soviet A cadem y of Sciences (see C hapter 2 5 ). U nder this concept national incom e in­ cluded only the net value of m aterial p ro ­ duction. It excluded the services of bank­ ing, insurance, professions, artisans, dom es­ tics, dwellings, and governm ent. T he serv­ ices of passenger transportation were in­ cluded, contrary to the standard m aterial p roduction concept, only because they could n o t be separated statistically fro m the serv­ ices of freight transportation th a t concep­ tually were deem ed to be a p a rt o f the n a­ tional incom e. T o p u t th e postw ar estimates on a p ar w ith those based on the com pre­ hensive production concept used in the W estern countries, according to D r. Vinski, they w ould have to be raised 11 p er cent. N ational incom e was com puted, just as in p rew ar estimates, as dom estically pro ­ duced incom e, i.e., w ithout any adjustm ent for the net incom e received from abroad. T he estim ates related to national incom e at m arket prices, rath er th a n at facto r cost. T he total included indirect taxes less sub­ sidies, b ut was net of depreciation. T he estim ates w ere prepared by the net output m ethod w ith the following classifica­ tion of industrial divisions in w hich the national incom e originates: agriculture and fishing; forestry; industry, electricity, m in­ ing and quarrying; construction; transpor­ tation and com m unications; trade; handi­ crafts; and miscellaneous. A ll com putations were m ade in current prices. T he estim ates fo r 1947-51 w ere com puted on the basis o f the official prices applied in 1947 in the socialized sector. These prices w ere m uch low er th an those paid in the free m arket w here m ost o f the products of the individual farm ers, repre­ senting the bulk of the country’s agricul­ tural produce, w ere being sold. T he national income was to that extent underestim ated. F o r 1952 and 1953, on the other hand, a m uch higher price level was used, inasm uch as by th at tim e the low er official price sys­ tem was abandoned and, m oreover, the country’s currency was devaluated from an exchange rate of 50 dinars per U.S. dollar to one of 300 dinars per dollar. T he estimates for 1947-51 were also dis­ torted by the continually rising rates of the turnover tax used to subsidize various groups

411

of products. A s a result of these differences in the price levels and in the values of the currency, the national incom e series for 1947-51 and those for 1952-53 were not com parable. T heir greatest incom parability, however, was with the prew ar estimates, w hich were based on different concepts of national incom e and w hich also related to a m uch sm aller territory and population. T he territorial gains after the w ar included the form er Italian peninsula of Istria, the cities o f Fium e and Zara, as well as three A driatic islands, representing a population gain of 500,000 in all. In 1954 the Y ugoslavian Econom ic Insti­ tute inaugurated a study designed to develop an entirely new series of national incom e estimates covering the years both before and after the w ar, w hich w ould be ex­ pressed in constant prices of a consistent n ature and w ould be based on single con­ cept and apply to the same territory and population. This new series, however, was n o t ready at the tim e of this writing. Sufficient data w ere available after the w ar fo r the com putation of the incomes originating in the state enterprises and col­ lectives m aking up the socialized sector, as these were required to m aintain com plete accounts of the gross values of their p ro ­ duction, their expenditures fo r raw m ate­ rials, depreciation, etc., and of the net value of their output. B ut fo r the private sector em bracing agriculture, trades, and handi­ crafts, the data w ere m uch less satisfactory. These data w ere prepared fo r general sta­ tistical purposes by governm ent-appointed district com m ittees consisting of persons n o t generally skilled in the collection and organization of statistics.

a. Agriculture, Fishing, and Forestry T he net output of the private sector of agriculture and fishing, w hich com prised the bulk of the agricultural output, was estim ated on the basis of production and price data. F o r the socialized sector—state farm s and co-operatives— the incom e was taken directly from the accounts of these enterprises. T he following description deals only with the estimating m ethods used in the private sector. The value of vegetal production was esti­ mated on the basis of the official statistics of physical output (with due allowance for un­ derreporting) multiplied by average prices paid to farmers by wholesalers. Deductions were then made for the quantity used for seed and

412

THE INCOM E OF NATIONS

fodder, as estimated by the local committees. For example, in the case of the hay crop, 90 per cent was allowed as food for animals on the farm. Meat production was estimated by the local committees in a rather complicated manner. On the basis of agricultural censuses, the an­ nual increase in the number of livestock was calculated for each category. For cattle, sheep, and pigs the net increase in numbers was mul­ tiplied in each case by the weighted average of live animals and was then multiplied by the average prices of live weight paid to farmers by wholesalers. The average weight was as­ certained from slaughtering data. The infor­ mation on average prices paid to farmers was taken from the reports of the commercial wholesalers. The annual increase in the num­ ber of horses was multiplied by the average price of a one-year colt. For chickens, turkeys, geese, and ducks the annual increase was taken as reported in the agricultural census, with an adjustment for the probable number perishing during the year. The value of animal products, consisting of milk, eggs, wool, etc., was calculated on the basis of assumed national average yield per animal. The physical output of fruit and grapes and of wine, cider, and their other by-products was reported in the current agricultural sta­ tistics. The production of honey was estimated by the local committees on the basis of an assumed average yield per beehive. The annual value of fishing was estimated on the basis of the catch in fresh weight. For mari­ time fishing the data were fairly complete, but for river and lake fishing the data were clearly inadequate. From the total gross value of vegetal, animal, and fishery production in the private sector, deductions were made for the value of the vegetal products used in animal production as well as for the costs of fertilizers and other chemicals, veterinary services, etc., and de­ preciation. As data on the amount of capital employed in the private sector of agriculture were not available, depreciation was calculated at a suggested rate of 3 per cent of the gross value of the output less all other costs. Incomes from subsidiary occupations of the agricultural population such as the production of firewood, cartage, and home production (spinning, weaving, cheesemaking, etc.) were added to its foregoing basic incomes. These subsidiary incomes had decreased considerably since prewar times. Although farmer’s produc­ tion for own consumption was included in farm income, it was not estimated separately. T he net value of the output of agricul­ ture and fishing was given in a single total. It was n o t broken down by the various branches of agriculture and fishing because of the impossibility of allocating deductible costs on such a basis.

T he n et output of forestry, w hich be­ longed mainly in the socialized sector, was taken directly from the accounts of the enterprises. On the oth er hand, the values o f the firewood collected by the farm ers for th eir own consum ption was com puted on the basis o f a special survey m ade by the F ederal Statistical Office in 1952.

b. M ining, M anufacturing, and Electrical Industry A ll m ining (including quarrying and oil ex tractio n ), the production and distribution of electricity, and m anufacturing belonged in the socialized sector. T he data on the value of their n et o u tp u t w ere given directly in th e accounts of these enterprises. Before 1954, depreciation in this sector was calcu­ lated as a percentage of the value of the output, since no data were available on the value of the invested capital. In the 1953 census, however, com prehensive inform a­ tion was collected on the am ount of fixed capital in these enterprises expressed in 1953 prices, m aking possible the calculation of the annual am ortization of this capital, on a replacem ent basis, at 1953 prices for all th e subsequent years. I t m ust be noted th at the relative proportion of the incom e from m ining and m anufacturing appeared to increase considerably in recent years (42 to 47 per cent in 1952 and 1953 as com pared w ith 37 per cent in 1950 and 20 per cent before th e w a r). T his increase was partly the result of a governm ent policy allowing a higher price level fo r industrial products th an for agricultural ones, and to th a t extent the increase was exaggerated.

c. Trades and H andicrafts F o r this branch of the econom y reliable data w ere available only fo r th a t smaller p art of it w hich belonged in the socialized sector, i.e., fo r state and co-operative enter­ prises. F o r the larger portion, w hich be­ longed in the private sector, th e data were n o t satisfactory. T he census gave the n um ­ bers of artisans and their em ployees and th e records of the tu rnover tax gave some inform ation on the value o f th eir output. B ut their net o u tput was com puted by ap­ plying to their gross o u tp u t th e proportions o f n et to gross obtaining in th e correspond­ ing occupations in the socialized sector— a rath e r hazardous procedure. W ith the ad­ vance of industrialization, the im portance o f this branch of the national economy

2 8 . YUGOSLAVIA

steadily declined over the years, in both absolute and relative am ounts.

d. Transportation and C om m unications T he bulk of this activity, em bracing rail­ ways; posts, telegraph and telephone; m ari­ tim e and inland w ater transportation; and m otor and road transport, belonged in the socialized sector. Statistics on the incom e originating there were taken directly from the accounts of the enterprises. O n the other hand, the m ajor p art of cartage, af­ fording im portant subsidiary incom e fo r the ru ral population, especially in the transpor­ tation of tim ber and building m aterials, was already covered in the incom e fro m agri­ culture.

e. C onstruction This branch of the econom y was con­ cerned w ith all types of construction and m aintenance (except railw ay m aintenance, w hich was included in tran sp o rta tio n ), and was largely socialized, being conducted by large-scale state enterprises. B ut m ore re ­ cently, especially since 1952, a considerable increase in private construction activity took place, especially in the m atter of residential construction in rural areas. F o r the social­ ized enterprise, the net value of construc­ tion was taken from the balance sheets of the enterprises. A ccount was taken not only of the net value of the new construction finished during th e year but also o f the

413

increase in construction w ork in progress. A djustm ent was m ade for the n et o u tput of private residential construction.

/. Trade This branch included n o t only wholesale and retail trade, b ut also hotels and res­ taurants. W ith the exception of the u n o r­ ganized trade in secondhand goods it was alm ost wholly socialized. T he incom e origi­ nating in trade was, therefore, com puted directly from the published accounts of the enterprises. T he postw ar estimates at cur­ ren t prices are show n in T able 28-3. They refer to the national incom e at m arket prices. In 1954, D r. Vinski m ade an ex­ ploratory com putation of the national in­ com e at factor cost fo r the year 1953-54. T his was the first postw ar com putation of its kind, all others having been m ade at m arket prices. H e arrived at a figure of 780 billion dinars at facto r cost as against 895 billions at m arket prices, a difference of 13 p er cent. T h e au th o r calculated the tax burden in Yugoslavia in 1954 at 27 per cent of the national incom e (including in factor incom e, the interest on fixed and working capital as well as social security benefits), com pared w ith a tax burden of 45 p er cent in 1949, when extensive governm ent con­ trols over the econom y were m aintained, and a tax burden of 23 to 33 per cent dur­ ing the prew ar period.9

Table 28-3. National Income of Yugoslavia at Market Prices by Industrial Origin, 1947-53 (billion dinars) Industry, electricity, mining, and quarrying Agriculture and fishing Forestry Construction Transportation and communications Commerce Handicraft Not allocated items

1947

1948

1949

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

132.9

189.7

233.1

1950 75.9 58.2 4.5 15.7 20.8 16.7 16.3

1951 74.2 82.4 3.4 13.0 20.6 18.3 18.8

208.0

230.8

1952 402.6 217.0 3.8 63.5 42.7 56.3 41.2 2 6 .8 b 853.8

1953 430.3 310.0 12.0 74.3 47.0 76.8 48.7 23.4 b 1,022.6

a D ata not available. b Contains interest on working capital of enterprises (except of the private sector, i. e., excluding mainly agriculture and handicraft) as well as the balance originating from the different levels of foreign exchange rates for exports and imports. source:

Reference No. 6, p. 109.

29. TURKEY *

References 1. T. C. Ekonomi Bakanligi Konjonktur Servisi (Bureau of Economic Research), Serie 1. No. 2, Milli Geliri (National Income), Ankara, 1935 (in Turkish). 2. T. C. Ekonomi Bakanligi Konjonktur Servisi (Bureau of Economic Research), Serie 1. No. 7, Turkiye Milli Geliri (National Income of Turkey), Ankara, 1937 (in Turkish). 3. T. C. Basbakanlik Istatistik Genel Mudurlugu, Turkiye Milli Geliri 1943 ve 1944 (National Income of Turkey 1943 and 1944), Publication 275/124, Ankara, 1947 (in Turkish). 4. a. Sefik Bilkur, National Income of Turkey, Turkish Republic Central Statistical Office, Publication 289/127, Ankara, 1949; also b. A paper by the same author, “National Income Estimates of Turkey and New Measures to Improve Reliability,” presented before the Meeting of the National Association for R e­ search in Income and Wealth, August 25-September 7, 1951. 5. a. Republic of Turkey, Central Statistical Office, National Income Unit, National Income of Turkey, First Provisional Estimate, Ankara, 1952. b . , Turkiye Milli Geliri 1938, 1949-1951 (National Income of Turkey, etc.), Ankara, 1954 (in Turkish). c . , Turkiye Milli Geliri 1938, 1948-1951 (National Income of Turkey, etc.), (Statis­ tical tables only, with titles in English), Ankara, 1955.

1. H isto ry T he need for national incom e estimates was first publicly recognized in T urkey in 1928. D uring th a t year Prim e M inister Inonu requested the H igh Econom ic Coun­ cil to prepare such an estimate. T he Council m ade some studies of the theoretical bases fo r such an estim ate, but concluded that because of the lack of necessary statistics, no actual calculation could be m ade at the tim e. In 1929, M r. Inonu requested Camille Jacquart, G eneral D irector of Statistics (who had designed the first m odern census in T urkey) to prepare such an estimate. A pparently M r. Jacquart m ade some crude estim ates based on his personal judgm ents * Acknowledgment is due to Dr. Resat Aktan of the University of Ankara and also of the National Income Study Unit of the Prime Min­ istry of Turkey for reviewing and correcting the preliminary draft of this chapter and supplying supplementary material covering the estimates prepared in 1951-54 and explaining the meth­ ods used for deflating the series at constant prices; and also to Mr. Server Sayar of the Central Bank of Turkey for preparing for the author an English translation of the 1937 pub­ lication (Reference No. 2). Notes begin on p. 542. 414

rath e r th an on any statistics, b u t these esti­ m ates w ere never published.1 T he Bureau of E conom ic Research of the T urkish M inistry of Econom y, established in 1935, prepared in th a t same year the first estim ate of T urkish national incom e.2 T his pioneer work, w hich covered the years 1927/28 and 1 9 3 3 /3 4 3 was unavoidably rough because of th e paucity of statistical data, b u t shortly afterw ard, w ith the devel­ opm ent of m ore adequate statistics, m ore elaborate estimates covering 1933/34 to 1935/36 were undertaken with the collabo­ ration of several governm ent departm ents. T hese later estimates, prepared u nder the direction of D r. F ran z E ppenstein, together w ith explanations of m ethods used, were published in 1937.4 In 1947, the Central Statistical Office (established during the 1920’s) published estim ates fo r the years 1 9 4 2 -4 4 .5 A ll th e official estimates p repared during the period 1935-47 appeared sporadically. D uring the years 1948-51 no estim ates were p repared by any official agencies, but Messrs. Bilkur, Eldem , an d Inan, all of w hom w ere officials, p rep ared several esti­ m ates on their ow n responsibility.6

29.

In 1950 a special research u nit was set up within the C entral Statistical Office of the Prim e M inistry to prepare national in­ com e estimates as well as to conduct other research projects. H ow ever, it soon becam e obvious th a t the u n it could n o t sim ultane­ ously conduct three m ajor censuses (re­ quired by law to be held every ten years) and prepare adequate estim ates of national incom e. T herefore, a separate N ational In­ com e Study U nit was established in 1951 fo r the express purpose o f preparing such estimates. T he first estim ate prepared by this new u nit u nder the chairm anship o f M r. Bilkur appeared in 1952.7 In 1955, further im provem ents in estim ating m ethods were introduced, but inform ation on them was not available at the m om ent of this writing. T he discussion in this chapter is divided into two parts: P art I concentrates on the earlier governm ent estimates published in 1937, w ith occasional m ention of the second official estim ates published in 1947 and also of those privately m ade by Bilkur fo r the y ear 1942-43. P art II treats the estimates prepared since 1951.

I. Estimates Prepared During the Period 1935-50 1. Concepts A ll the estim ates prepared during the period 1935-50, except the one by Eldem , dealt w ith the gross national product avail­ able to the country, rather than w ith the national incom e at factor cost o r at m arket prices. T hey included in this total profits o f governm ent monopolies and indirect taxes, and they m ade no deduction fo r de­ preciation. N o r did they m ake any adjust­ m ent fo r the net international balance of capital incom e. On that basis, th e totals in the B ureau’s and the Statistical Office’s es­ tim ates and in Bilkur’s resem bled homeproduced national income at m arket prices. H ow ever, since no deduction was m ade for depreciation, this aggregate resem bled even m ore closely hom e-produced gross national incom e at m arket prices.8 O n th e other hand, E ldem in the private estim ate prepared in 1948 fo r selected years o f the period 1929-45 used a concept ap­ proaching th a t of the national incom e at f a c to r cost.

TURKEY

415

2. Estimating Methods T he net ou tp u t m ethod was applied for the preponderant p a rt of the estimates. Only incom e from banking, insurance, transpor­ tation, governm ental, and other services, in­ cluding professions, was m easured by the incom e-distributed m ethod. Separate esti­ m ates were m ade fo r each division of eco­ nom ic activity. Profits of governm ent m onopolies and indirect taxes w ere included in the n et out­ put of trade, thus giving national incom e at m arket prices. Since they were shown sepa­ rately, it was possible to deduct them from the latter total to get a national incom e aggregate closer to national income at factor cost (Table 29-1).

a. Agriculture T he estimates for agriculture, w hich is the leading econom ic activity in Turkey, were fairly elaborate. F o r each m ajor and m any m inor agricultural products, the quan­ tities produced, as ascertained from p ro ­ duction statistics o r estimates, w ere m ulti­ plied by the appropriate farm prices, prices received at slaughterhouses, or, in the case of exported fruit, average export prices after a certain percentage deduction. As, by and large, direct price data were scarce, collateral inform ation w ith appropriate ad­ justm ents had to be resorted to. T he gross value of vegetal production was com puted, in accordance w ith the usual p at­ tern, on the basis of quantity and appro­ priate price data fo r a long list of crops. F o r undercoverage of vegetal production an over-all addition o f 10 per cent was m ade. Because specific inform ation on de­ ductible costs was lacking, an over-all ratio o f 30 per cent of the gross value of vegetal production was assumed to represent all cost allowances, including probably depre­ ciation.9 M ore com plicated was the estim ate of the value of anim al output, and especially, of m eat and m ilk production. Since the slaugh­ tering statistics covered only the m eat con­ sum ption of the urban population, an elabo­ rate estimate of total m eat consum ption had to be m ade. This investigation, made w ith the aid o f m any governm ent agencies and experts, being quite original, deserves special consideration: Yearly returns on livestock, required for tax purposes, were adjusted for understatement. By combining these statistics with yearly re­

416

THE INCO M E OF NATIONS

production rates and the balance of foreign animal trade, the number of slaughtered ani­ mals was determined. The natural reproduction was elaborately estimated by taking account of all factors involved such as births, stillbirths, deaths, and replacement requirements, the un­ derlying assumption being based on the ex­ periences of experts. Then, average meat out­ put (meat, fats, and bones) in kilograms per head of mature and young animals was ascer­ tained for the period 1925-35 and applied to the number of animals slaughtered. Finally, the quantity data thus obtained were multiplied by the average prices received at slaughter­ houses. To the resulting figure the value of exported animals was added. These estimates were made separately for sheep, mohair, and goats, distinguished by sex and age classes. In the case of beef and buffalo meat a simplified procedure was used. Their respective proportions in total meat consump­ tion were considered equal to the proportions of the animals involved in the total weight of animals slaughtered at the urban slaughter­ houses. Tables 29-1 and 29-2 show in concrete terms the major steps taken in determining consump­ tion of mutton and lamb in 1934-35.10 The annual value of milk production was estimated by applying an average milk yield to the number of ewes, female goats, and cows, and multiplying the quantity by an average price. The average milk output per animal, ascertained from special investigations of the Statistical Office, excluded milk fed to lambs, young goats, and calves. Then the total value of milk production was raised by 10 per cent to allow for the additional value of milk prod­ ucts, such as cheese, butter, and yoghurt. For minor components of animal products such as hides, wool, hair, poultry, and honey, for which no direct quantity data were avail­

able, similar techniques were used. First, the average yield per animal was determined and then the total production was multiplied by an average price. For instance, in order to esti­ mate poultry production, it was assumed that an average farm has 15 hens of which 4 brood hens produce 6 live chickens each, per year. T he estim ates of the value of o u tput of game and fishing w ere based on the appro­ priate returns of th e production tax. This tax was levied at a rate of 12 per cent of gross output. In order to take account of understatem ent, additions of 20 p er cent for fishing and 100 per cent fo r game were m ade to the production value indicated by the tax. F ro m the gross value of total anim al pro­ duction, a blanket 15 per cent was deducted fo r costs of production. In the Central Statistical Office series for the years 1942-44 the 10 per cent of the total gross value of all livestock products except m eat was deducted as representing costs of production of these products. H ow ­ ever, in cases of incom es derived from fish­ ing, games and n et increase in the num ber of livestock no production cost was con­ sidered.11

b. Forestry T he value o f cut tim ber, firewood, and charcoal products was calculated by apply­ ing to the total ou tp u t of these m aterials th e average prices obtained by the state forest adm inistration. D eductible costs were assum ed to account fo r 15 per cent of the gross value of production.

Table 29-1. Production of Lambs in Turkey, 1933-35 a (thousand head) Stocks and Production Ewes at the beginning of the year b Natural deaths (10 per cent of this stock) 0 Ewes living through the year Sterile ewes (15 per cent of this stock) Lambs born (1 per ewe) Stillbirths (5 per cent of born lambs) Live-born lambs Natural deaths during the year (10 per cent of this stock) Lamb supply for replacement and consumption d Females Males

1933-34

1934-35

12,200 2,440 9,760 1,464 8,296 415 7,881

11,800 1,180 10,620 1,593 9,027 451 8,576

788 7,093 3,547 3,546

858 7,718 3,859 3,859

a Reference No. 2, p. 53. b Yearly data on livestock were derived by adjusting tax statistics for undercoverage. c Normal mortality increased to 20 per cent in 1933-34. d Females and males are each 50 per cent of the total.

29.

TURKEY

417

Table 29-2. Consumption of M utton and Lamb in Turkey, 1934-35 ; (thousands) Supply, Replacement Requirements, and Consumption

Female

Lambs born in previous year Slaughtered in the previous year (J^ of stock) Carried over from previous year Required for replacement of sheep Loss through natural death Slaughtered Normal carryover for lamb slaughtering Normal slaughter of lambs born in the year (J4) Normal slaughter of sheep (J^ of average stock) b Normal supply for slaughter Adjustment for changes in sheep stocks 0 Consumption after adjustment for foreign trade and accurate age distribution Lambs Sheep Average meat yield per head (in kilograms) Lambs Sheep Total meat (lamb and mutton) consumption (in tons)

3,547 3,547 1,180 1,770 597 1,770 2,367 -1 ,3 0 0

Male 3,546 1,773 1,773

Total 7,093 1,773 5,320

59 133 1,581 1,930 133 3,644 -6 5

1,239 1,903 2,178 1,930 1,903 6,011 -1 ,3 6 5 2,609 1,403 9.0 21.0 198.1

° Reference No. 2, pp. 54-55 and 65. b This ratio was based on the average reproduction period of a sheep of six years. 0 The increase in stocks is equal to a decrease in consumption. In the 1942-44 series in order to cover the cutting and transporting activities of the people engaged in forestry an additional in­ com e of the sam e m agnitude w ith the value of total forestry output was added, and no production costs w ere deducted.12 In his 1942/43 estimate, M r. B ilkur re ­ lied, in general, on the same sources as the official estim ates under discussion. H ow ever, the im position of a tax in kind on agricul­ tural produce during the early years of the Second W orld W ar affected, in his opinion, the reliability of production figures in that sector. P roduction was so understated that according to official figures grain produc­ tion, fo r example, declined from an ap­ proxim ate eight million tons in 1938 to five m illion tons in 1943, a year of exceptionally fine growing w eather. A s a result, Bilkur turned to indirect m ethods fo r his estimate. H e first com puted the average production figures by com m odity for the years 1937-39, and the average price per com m odity for the base year 1935/36. H e next m ultiplied these prices by appropriate coefficients of price increase to arrive at estim ated 1942/43 prices, m ultiplying these in tu rn by the ad­ justed production figures to arrive at the value o f output. F o r those com m odities not covered in the price indexes upon w hich the

coefficients were based, a proportionate rise was assum ed.13

c. M anufacturing, M ining, and O ther Industry T his broad division included m anufactur­ ing, mining, electric light and pow er, gas, and construction. It em braced large enter­ prises, small firms, and independent m anual w orkers (artisans and craftsm en). It should be stressed, however, th a t the 1927 Census of M anufacturing, w hich was the basis of the estimates, related to enterprises in cities only. The im portant ru ral carpet weaving and other h andicraft industries were omitted. G enerally the coverage of the Census was inadequate, as the incom plete results of the 1935 Population Census and other studies indicate. T he Bureau took the 1927 Census results on industrial em ploym ent as the starting point of its estimates. This em ploym ent figure was first raised about 50 per cent for undercoverage. T hen the resulting figure was projected to the years of the estimates in proportion to the population growth. H ow ­ ever, the special encouragem ent given by the State to industry during the years 1927-33 was acknowledged by allowing for an in­

418

THE INCOM E OF NATIONS

crease som ew hat higher than the general population growth. T hen the em ploym ent figures in the years 1933-34 to 1935-36 were split into w orkers in protected indus­ tries and w orkers in other industries. T he data on gross value added and em ­ ploym ent in the protected industries for 1933 were directly available. T he Bureau assumed th a t the gross value added per w orker in the nonprotected p art of industry was one fifth of that in the protected p a rt.14 By m ultiplying the resulting figure by the num ber of w orkers in nonprotected indus­ try the Bureau obtained the total gross value added in it. The estimates fo r 1934 and 1935 were roughly approxim ated on the basis o f the increases in output and prices. T he estimates for 1942-44 by the Cen­ tral Statistical Office w ere m ade som ew hat differently. M anufacturing industries were divided into four groups. (a) Industries owned wholly or partly by the State. (b) Electric and gas plants owned by munici­ palities. (c) Private industries competing with State industries. (d) Other private industries. For estimating net incomes of group (a) the balance sheets and other reports of these enter­ prises were used after making certain adjust­ ments. In case of group (b) use was made of avail­ able fairly complete sets of information re­ garding income and expense accounts of a few municipal plants and of the outputs for the rest, after appropriate adjustments. In estimating the income of group (c) a calculation was made first of the percentage share of output of each particular private in­ dustry in the total output of the same kind; and then comparable incomes were assigned to each subsector on the basis of incomes de­ rived in the State-owned counterparts. In case of group (d ), first the total number employed in these industries was determined by extrapolating 1935 population census figures for the total industrial employment; second the known numbers employed in the groups (a), (b), and (c) were deducted from this num­ ber; thirdly the average annual value created per worker in these industries was assumed to be half of the value created per person in the comparable State industries; and, finally, the number employed was multiplied by the aver­ age value created per person, thus producing the total net income for group (d ).15

d. Trade, Finance, and Transportation In the absence of direct data on income created by com m ercial activity, an indirect m ethod was used by the Bureau. T he distri­ bution sector, including trade proper, tran s­ portation, banks, and insurance companies, was taken as a whole. It was assum ed rather roughly th a t the value added therein, plus profits of governm ent monopolies and busi­ ness cost taxes,16 represented certain p er­ centages of the value added by agriculture, forestry, and industry during the years cov­ ered by the estim ates.17 Incom es in tran s­ portation, banking, and insurance as well as profits o f governm ent m onopolies and taxes w ere estim ated separately, m aking it pos­ sible to com pute the o u tput of trade proper. T h e separate estim ates fo r the other divi­ sions w ere m ade as follows: Banking and insurance. Compensation of employees and profits of the banking institu­ tions, which are to a great extent State-owned in Turkey, were ascertained from the financial reports of their institutions for 1935 and were assumed to remain constant in the other years. The net output in insurance was measured by premium payments. Transportation. Salaries, wages, interest, and profits of railroads, street cars at Istanbul and Smyrna, and shipping lines were reported in current statistics. To these incomes a flat amount of 15 million pounds was added each year to take account of the net output in truck transportation, on the one hand, and other road transportation, such as buses, taxis, horsedriven cars, etc., on the other. This estimate was in the nature of a guess. T h e C entral Statistical Office’s estimates of incom e from trad e fo r the years 1942-44 differed both conceptually and m ethodo­ logically from the B ureau’s estimates. V alue added by trade p ro p er was considered as including transportation and also possibly banking and insurance services. T he total turnover values o f goods com ing from agri­ cultural, industrial, and foreign trade chan­ nels was calculated first. C ertain m arkups were applied to them w hich w ere calculated as the differences betw een the farm , plant, o r C IF prices and the w holesale o r retail prices of the same type of goods. F ro m the m arkup values, deduction was m ade fo r cer­ tain costs including those of coal used by the State railroads, seaways, and also p ri­ vate ship lines.18

29.

e. Services o f Dwellings, G overnm ent, Professions, a nd Others T he value of the services of dwellings was com puted from the returns of the buildings tax: This tax was assessed at a rate of 12 per cent on the rentals derived from rented and owner-occupied houses, after deduction of 20 per cent for costs. By applying the aforemen­ tioned rate to tax proceeds of 6.5 million liras in 1933/34, the net rental was calculated at 67.5 millions. Inasmuch, however, as some dwellings were tax exempt and, on the other hand, the figures included assessments on com­ mercial and industrial premises, the net rental was taken as 60 million liras for 1933/34, 61.1 millions for 1934/35 and 62.2 millions for 1935/36 on the assumption that construction increased at no greater rate than population and that rents remained the same. TTie esti­ mates for the years 1942-44 were made in the same way. However, they included also rental incomes from government buildings.19 Bilkur’s 1942/43 estimates were based on the tax in­ formation plus the results of a special survey. The statistical m aterial covering the in­ com es of governm ent em ployees, profes­ sions, and dom estic servants was inade­ quate. Summary budget studies led to the conclu­ sion that the salaries paid by the central gov­ ernment added up to 60 million liras. Equally rough were the estimates of the salaries and wages paid by municipalities and the incomes of professions (called “free trades people”) 20 and domestic servants. It seems that the esti­ mates aimed at obtaining the flat amount of 100 million liras for the whole composite divi­ sion in 1933-34. This sum was projected to the other years in proportion to the increase in population. The 1942-44 estimates included in this category also the compensation of other employees, calculated on the basis of the oc­ cupational census and assumed average rates of pay.

II. Estimates Prepared in 1951-54 21

TURKEY

419

gates. These aggregates were: domestic gross output, gross national product adjusted for international transactions, national income at m arket prices, and national income at factor cost. T he estimates covered the years 1938 and 1948-54, and were presented at both current and constant prices. Their gen­ eral setup was as follows: 1. Agriculture, forestry, and fishing-hunting a. Agriculture b. Forestry c. Fishing and hunting 2. Industry a. Mining b. Manufacturing c. Electricity, gas, and water 3. Construction 4. Trade 5. Financial institutions 6. Communications 7. Professions and services 8. Dwellings 9. Governmental services Total: Domestic income 10. Plus: Rest of the world account National income (Net national prod­ uct at factor cost) 11. Plus: Indirect business taxes Net national product at market prices 12. Plus: Depreciation allowances Gross national product at market prices T he U nit was fortunate in being able to use better and m ore com plete statistical data than had been available to its predecessors. A m ong these the 1950 censuses of A gricul­ ture, Business, and Population are p articu­ larly w orth m entioning.22 T he U nit also undertook a few small surveys of its own in order to fill in gaps in the data, and it applied m ore refined estim ating m ethods than those used before. Some of its m ost im portant innovations em ployed in the vari­ ous segments of the estim ates are indicated below. T he results are given in T able 29-3.

2. Estimates by Industrial Divisions

1. General Approach

a. Agriculture and Forestry

T h e estimates prepared by th e N ational Incom e Study U nit of the C entral Statistical Office in 1951-54 contained m any im prove­ m ents over the previous estim ates. T hey follow ed closely the O E EC ’s standards of classification of industrial divisions and its suggested setup of production income aggre­

A very detailed subclassification was m ade in order to assure com plete coverage in agriculture. M inor crops, small anim al pro­ duction, and such subsidiary activities as the preparation of food o r clothing on the farm were covered. In earlier estimates some of the latter item s w ere omitted.

THE INCO M E OF NATIONS

420

Table 29-3. National Income of Turkey a t Factor Cost, by Divisions of Economic Activity, a t Current Prices, for Selected Years (millions Turkish lira) Econ. Res. Bureau Series

National Income Unit Series

Central Statistical Office Series

1954

1948

Preliminary

775.0 204.4 16.4 182.2

4,691.4 926.3 78.2 828.1

5,943.9 2,032.7 178.2 1,800.3

5.8 76.7 167.5 30.4 94.0 67.2 170.1 52.5 1,637.8 - 3 .7

20.0 280.3 960.6 129.4 406.3 325.3 877.6 237.5 8,834.7 - 2 0 .0

53.6 814.3 1,576.0 383.6 1,066.7 589.2 1,448.4 452.6 1,306.8 - 3 4 .1

1,634.1 230.4

8,814.7 913.8

14,272.7 1,696.4

Net national product a t market prices Depreciation

1,864.5 88.5

9,728.5 338.5

15,963.7 704.4

Gross national product

1 ,953.0

1 0 ,0 6 7 .0

1 6 ,6 7 3 .5

1935-36 1. Agriculture, forestry, hunting, and fishing 435.0 2. Industry 215.0 a. Mining b. Manufacturing c. Electricity, gas, and water * 3. Construction industry 4. Trade 5. Financial institutions 514.5 6. Communications 7. Professions and services 103.6 8. Government services 62.2 9. Dwellings 1,330.3 Total domestic income b 10. Rest of world Total national income Indirect taxes

1942

1944

3,710.0 558.3

3,451.3 1,520.4 66.4 1,434.9 19.1 * 2,301.4

* 1,652.4

t t 318.9

t t

365.5

130.0 129.5 6,369.2 7,768.6 1.0 .6 » 6,369.8

7,769.6

1938

* Included in industrial incomes. t Except insurance income in certain cases, not explicitly considered, t Included in the trade income figures. a Represents only royalty incomes from Mosul (Iraq) oils. b Not considered. source:

Reference No. 5c, p. 4.

In valuing agricultural production two kinds of seasonal price averages were used. F o r m arketed products an elaborate doubly weighted average was developed; fo r prod ­ ucts kept on the farm simple w eighted aver­ ages of seasonal prices were used. A fter com pleting the evaluation at curren t prices, the com putation was repeated at 1948 prices. In earlier com putations a rough and incom ­ plete index of prices was used fo r this pur­ pose. In those being developed currently, 1951 prices are used as a base. The net increase in all animal stock (and not only in some, as before) was computed in a slightly different manner. Different price averages were used for animals kept on the farm and for those sold, making unnecessary a separate estimate for consumed meat. Deductible costs of vegetal and animal pro­

duction were itemized on the basis of available information, estimates of experts, and other data; those of forestry output were derived from reports of the Forest Administration; and those of fishing were estimated. Depreciation allowances for agricultural buildings and machinery used in production were estimated separately for each category and for each year. At one point, figuring de­ preciation of work animals was also consid­ ered, but later on, at the suggestion of Mr. Milton Gilbert of the OEEC, it was abandoned. F or milk, wool, hair, honey, cocoons, etc., quantity and price figures were available. For fish production, the Meat and Fish Organiza­ tion provided fairly accurate data. No basic changes were made in the old methods of esti­ mating the number, value, and yearly increase in poultry on the farm. Evaluations were made of formerly omitted farm production of flour, bread, dried fruit, etc.

2 9 . TURKEY

In order to compute the net output in for­ estry, deductions from gross value were made for purchases from other industries; deprecia­ tion of plant and equipment; and the consump­ tion of forest capital, calculated as the excess of the value of timber cut over the value of new timber grown.

b. Manufacturing and Other Industry T he industrial classification in the U n it’s estimates included w ater distribution in cities and towns, but excluded construction. Some p art of the o utput of State enterprises in­ cluded trading activity. W hile the net output m ethod was used fo r private industries, the incom e-distributed m ethod was used for State industries. F o r the 1950 estim ates busi­ ness census data w ere used. T he procedures were as follows: State enterprises: Value added was computed as the sum of salaries, wages, and other social benefit payments (including retirement fund contributions of enterprises) plus their profits, gifts and transfer payments, business taxes paid, and depreciation allowances. The concept of State enterprises was interpreted broadly to include provincial and municipal undertakings, including some urban public utilities (elec­ tricity or water distribution or bus service), legally operated as corporations. Information about them was collected from their reports and especially from the yearly reports of the High Auditing Board—an agency attached to the Prime Ministry and operated under the auspices of the Grand National Assembly. Private industries: Value added was com­ puted as total receipts from sales less the value of raw materials, supplies, and certain other expenditures. Depreciation allowances, where not actually reported, were estimated from the value of the fixed assets. To get the contribu­ tion to national income, indirect business taxes were deducted. For years other than 1950, data were not available except for private mining activity, and a questionnaire was sent to 30 large representative firms to get information on their yearly production, sales, etc. An index was then built that could be applied to the 1950 census figures for large industries. For small industries a different method of conversion was used: output per man was assumed to be un­ changed from 1950, with changes in the value of the output being attributed solely to price or wage increases; net output was extrapolated on the basis of a combined index derived from the separate indexes of the costs of raw ma­ terials, wages, and wholesale prices.

c. Construction A s a first approxim ation, value added in construction was com puted by taking the

421

num ber of gainfully em ployed in different branches of it as given in the Population Census, and m ultiplying it by an average estim ated incom e in each category. U nder another parallel approach, deem ed to be m ore reliable, the num ber of buildings con­ structed each year was determ ined from building perm its, building tax data, and other sources. These were grouped in classes, and an average value fo r each class was esti­ m ated. C ertain percentages of these values w ere then taken as representative of value added.

d. Trade, Finance, and Transportation F o r trade, the was mainly used, against another com m odity flow m ethod:

incom e-distributed m ethod b u t its results w ere checked estim ate prepared by the variant o f the net output

a. By the income-distributed method: The numbers of people engaged in various kinds of trade in different sizes of communities were estimated separately as entrepreneurs, employ­ ees, etc. In addition to the population censuses, income tax returns, labor statistics, and other sources were used for this purpose. Average incomes of entrepreneurs in the various groups were determined from income tax returns. Average incomes of employees in trade were partly determined from wage statistics, and partly guessed. The basic computation was made for the year 1950 with indexes used in other years to convert the 1950 data. b. By the commodity flow variant o f net output method: The quantities and values of goods entering wholesale trade from agricul­ ture, industry (mining included), and foreign trade were determined; next the proportions entering retail trade were computed, and finally, consumer expenditures were determined. Profit margins in wholesale and retail trade were esti­ mated by deducting from wholesale and retail prices in different commodity groups the pro­ portionate costs of goods entering trade. These profit margins were then further refined by ap­ propriate deductions, and the resulting ratios, after being applied to the gross figures, yielded income in trade. The figure obtained by the commodity flow method did not agree very closely with that obtained by the income-distributed method. Consequently, the average of the two figures was used in the Unit’s first publication. Later on, however, the figure arrived at by the first method alone was used. Because the official reports of the banks and insurance com panies were n ot uniform , making them unsatisfactory for estimating purposes, the U nit sent special question-

422

THE INCO M E OF NATIONS

naires to all these institutions requesting data on salaries, wages, social paym ents, profits, direct and indirect business taxes paid, grants, and aids. D ata on incom es of agricultural co-operatives were collected from the records of the Bank of A gricul­ ture, w hich supervises them ; on those of loan associations, from the M inistry of C om m erce and Econom y records; and those on insurance brokers, from incom e tax re ­ turns. Incom es of state and m unicipal transpor­ tation and com m unication services w ere esti­ m ated w ithout difficulty on the basis of offi­ cial reports and special questionnaires. Those of the private agencies in the same field w ere not as easily calculated. T he proce­ dures followed for the main groups in this field were as follows: Income originating in private sea transpor­ tation was estimated on the basis of a special survey covering the number and tonnage of vessels, and their business volume and financial operations. For motor vehicles, the statistics of the licensing agencies provided information on their number and character, while a sample survey made by the Unit provided data on their average earnings. Vehicles belonging to indus­

trial or commercial establishments were ex­ cluded from this computation. Total earnings in transportation by animal-drawn carriages (excluding those in villages) was estimated on the basis of a rough calculation of their num­ ber and an assumed average earning per vehicle determined from a sample survey. The number of porters was determined from population censuses, and their average incomes were esti­ mated by comparison with incomes of common laborers. The 1950 business census was used to obtain the 1950 income created in travel and shipping agencies, including foreign air­ lines operating in Turkey; and the same figure was used for subsequent years.

e. Services (Professions, Dwellings, Gov­ ernment) and Rest of the World T he incom e created in the rem aining di­ visions of the econom y and in its transac­ tions w ith other countries was likewise esti­ m ated by better m ethods th an h ad been used previously. Professions and services: Population census figures adjusted on the basis of annual income tax returns and other sources were used to de­ termine the number involved. The yearly aver­ age incomes were estimated on the basis of income tax records and comparison with earn-

Table 29-4. National Income of Turkey for Selected Years at Current and Constant Prices, Totals and Per Capita National Income at Current Prices

Years

Popu­ lation

Total (millions Turkish lira)

Cost-ofLiving Index Per Capita (Turkish (1933-34 = 100) lira)

National Income at Constant Prices Total (millions Turkish lira)

Per Capita (Turkish lira)

(1933 prices) A. Ministry of Economics Series: 1933-34 15,491 1934-35 15,821 1935-36 16,158

1,149.7 1,250.1 1,330.3

73.56 78.54 82.10

100.0

103.9 108.6 General Wholesale Price Index (1938 = 100)

B. Central Statistical Office Series: 18,203 1942 18,396 1943 18,592 1944

6,369.8 8,286.0 7,769.6

349 .19 449,.06 416. 33

339.6 590.1 458.9

1.149.7 1.202.7 1,224.9

73.56 75.57 75.61

(1938 prices) 1,875.7 1,404.2 1,693.1

102. 82 76. 10 90. 72

(1948 prices) National Income Unit 1938 1948 1950 1952 1954 (Preliminary) source:

Series: 17,016 20,049 20,947 22,160 23,441

Reference No. 5c, p p . 4, 24.

1,634.1 8,814.2 8,964.2 12,424.1 14,272.7

96 440 428 561 609

— — — —

—

7,348.5 8,814.7 9,098.3 11,393.0 11,487.7

432 440 434 514 490

29.

ings in comparable government positions. In the case of domestic servants, income in kind was also estimated and added to the total. Dwellings: The buildings tax covered most of the buildings in cities and towns. In order to allow for exemptions, an adjustment was made in the figure obtained from the tax rec­ ords. Buildings in villages were not well covered in tax data; so an estimate of them was made from population figures. For village dwellings, an average rent was estimated. Because of rent control, the buildings tax figures did not re­ flect actual rents, and an adjustment was made for those years in which such control was in effect. Government services: Incomes from both national and local government were all esti­ mated from government accounts with adjust­ ments for payments in kind. Retirement allow­ ances and travel allowances were excluded. All computations except for one year were made on a fiscal year basis (March to February). Rest o f the world: Incomes obtained from abroad by Turkish citizens and the incomes of foreigners residing in Turkey were partly esti­ mated and partly taken directly from the finan­ cial statistics of the Ministry of Finance.

TURKEY

423

b.

c. d. e. /.

III. Methods of Deflation of the Income Series T he th ree official national incom e series com puted at current prices were deflated into a series at constant prices by a variety of m ethods (see T able 29-4). 1. Bureau of Economic Research Series (1933-34 to 1935-36): This series was deflated by means of an index of con­ sumer expenditures. This index had but a limited coverage. It was based on un­ weighted retail prices of twenty-four com­ modities— all of them necessities—in twenty cities. Rent was not included in the index. 2. Central Statistical Office Series ( 194244): In deflating this series an index of wholesale prices was used. This index was prepared by the Bureau of Economic Re­ search, Ministry of Economy and cov­ ered about one hundred items which in­ cluded food materials, feed, industrial raw materials, and some semimanufactured goods. The prices were weighted. It was a fairly good index. 3. National Income Study Unit Series (1938, 1948-54): In deflating this series differ­ ent methods were used for each sector of the economy: a. Agriculture, forestry and fishing: the value of output in different years at constant prices was derived by mul­

g.

h.

i.

tiplying production figures in each year by the average prices of 1948. In case of production costs a similar method was used, i.e., average 1948 prices of cost items were applied to inputs of various years. This method was laborious, but it provided a rather satisfactory deflated series for this branch of the economy which was as large as all the other branches com­ bined. Industry: a composite index was used of the number of people engaged in small industries and handicrafts and of the amounts of output of basic government industries and mines and also of large-scale private industry. Construction: a special index was used based on the number of people engaged in construction industries. Trade: a composite index was used made of weighted averages of whole­ sale prices and cost-of-living indexes. Financial institutions: wholesale prices index was used. Communications: for rail, sea and air transportation, indexes were used that were based on volume of cargo and persons transported; for bus and street­ car transportation, an index based on number of fares; for other motor ve­ hicles, an index based on number of vehicles in use; for income of porters, an index based on their numbers; for postal, telephone and telegraph com­ munications, an index based on the rates effective during different years. Services: individual income levels dur­ ing the postwar years were assumed to have remained unchanged. The only deflation made was for the year 1938, and this was done by attributing to it postwar income levels. Dwellings: incomes from dwellings during the postwar years when govern­ ment rent control measures were in effect were adjusted by raising by a cer­ tain proportion the total rents com­ puted on the basis of old assessment values of the buildings tax. Similar ad­ justments were made in the rent in­ comes of 1938. Government: a special index was pre­ pared on the basis of the average change in the bulk of government wages and salaries toward the end of the war—the only time when these wages and salaries were substantially raised. This index was applied to the 1938 government-paid wages and sal­ aries. In order to deflate the payments other than wages and salaries, the in­ dex of general wholesale prices was used.

30. INDIA

References 1. V. K. R. V. Rao, A n Essay on India’s National Income, 1925-29, London, 1939. 2. , The National Income of British India, 1931-32, London, 1944. 3. National Income Committee, Department of Economic Affairs, Ministry of Finance, First Report of the National Income Committee, April, 1951. 4. , Final Report of the National Income Committee, February, 1954. 5. M. Mukherjee and A. K. Ghosh, “The Pattern of Income and Expenditure in the Indian Union: A Tentative Study.” Bulletin of the International Statistical Institute, vol. XXXIII, Part III, December, 1951. 6. United Nations, Economic Survey of Asia and the Far East, 1954-57 issues.

1. History

problem o f the p rep aratio n of com prehen­ sive estimates fo r th e In d ian U nion. The C om m ittee invited P rofessor Sim on K uznets of the U nited States, M r. J. R. N . Stone of E ngland, and D r. J. B. D . D erksen o f the N ational Incom e U n it of th e U nited N a­ tions to com e to India to aid them in their studies o f the problem . A s a result of these studies an d conferences, a N ational Incom e U nit was created in the M inistry of Finance, an d th e Com m ittee’s F irst R eport contain­ ing national incom e estim ates fo r 1948-49 was published in 1951. In the Com m ittee’s F in al R eport o f 1954, containing estimates fo r the years 1948-49, 1949-50, and 195051, a recom m endation was m ade th a t the N ational Incom e U n it present national in­ com e estimates in the form of an annual W hite P aper analogous to the W hite P aper o n national incom e in th e U nited K ingdom . T h e N ational Incom e U n it (hereinafter referred to as N IU ) , w hich prep ared the first official estim ate in 1951 u n d er the aus­ pices of th e N ational Incom e Com m ittee, follow ed very largely the m ethod adopted by R ao in his 1931-32 w ork. Its estimates likewise com bined the p ro d u ct approach w ith the incom e approach, “except th a t they have perhaps a sm aller m easure o f inaccu­ racy th an R ao’s, m ainly because of the availability . . . of m ore d ata.” 4 In 1953-54 two m em bers of the staff of the N IU , M. M ukherjee an d A. K. G hosh, p rep ared an estim ate o f consum er expendi­ tures and capital form ation, chiefly by the income-disposed or final expenditure m ethod.

D espite a dearth of statistical data, esti­ m ates of the national income o f British India appeared at various intervals since 1876. By the end o f the 1920’s ten such estimates h ad been published.1 In the 1930’s, w ith the im provem ent in statistical sources, the estimates becam e m ore elaborate. One series was prepared by G . F indlay Shirras, the noted British scholar in public finance stationed in India, and another by the Indian economist, V. K. R. V. R ao.2 T he form er covered the years 1921-22 to 1932-33 and 1937-38 to 1941-42, while the latter was for the average of th e five years 1925-26 to 1929-30 and fo r 1931-32. M ost of these estim ates took account only of the British Provinces o f British India. Because of data they did n o t include the Indian States. Burm a, w hich was severed from British India and was set up as an independent adm inistrative u n it in 1937, however, was included in both S hirras’ and R ao’s estimates. In the latter, it was treated separately fo r 1931-32.3 Both estimates con­ veyed significant inform ation on the social and econom ic peculiarities o f India, b u t this w as especially true of R ao’s tw o volumes. In 1949, soon after the form ation of the new governm ent of the independent state of India, a N ational Incom e Com m ittee was appointed. T he Com m ittee, com posed of representatives of the M inistry o f Finance and of several Indian econom ic and statis­ tical institutes, was set up to consider the Notes begin on p. 543. 424

30.

T he present chapter consists of two parts, one discussing the Rao and the official esti­ m ates by the net output m ethod; and the other, the unofficial, M ukherjee-G hosh final expenditure or product estimates.

2. Concepts In b o th his estimates, Rao adhered to the concept of national income as the net value o f goods and services produced at factor cost and adjusted fo r the foreign income balance. B ut he did not treat the latter item the same in both estimates. W hile in the first estim ate he deducted the n et outflow of capital incom e and rem ittances from the n et value o f goods and services produced, in the second estim ate he deducted not only the net outflow of capital incom e but also oth er services (m ainly shipping).5 In his first estim ate he om itted net rent from dwellings and incomes from auxiliary occupations, w hile in the second he included these items. In both estim ates he ignored interest derived from “unproductive” gov­ ernm ent debt. T he N IU ’s treatm ent of governm ent is sim ilar to th a t of the U nited States: govern­ m ent adm inistration is taken as the sum of wages and salaries paid by governm ent, and governm ent enterprise is classified w ith in­ dustry and treated on the same basis as other industries. In the N IU ’s F irst R eport, land revenue tax was treated as an indirect tax on incom e; in the Final R eport this item was treated as a direct tax. Interest on con­ sum er debt h ad been treated as the im puted incom e o f indigenous m oneylenders for their banking services; in the later R eport this was regarded as a transfer and om itted.

I. Net Output Estimates 1. General Methods and Sources a. Methods in Rao’s Two Estimates In the 1925-29 estimates, the net output m ethod was applied to agriculture, forestry, mining, factories, handw eaving, trade, ship­ ping, and road transportation, while in the estim ates for 1931-32 its use was lim ited to agriculture, forestry, and m ining only. In b oth cases the incomes shown fo r individual divisions fell short of their true net output or incom e produced. In the 1931-32 esti­ mates R ao took the aggregate reported in­ comes assessed under the incom e tax (which

INDIA

425

covered incom es in excess of 1,000 rupees derived from other sources th an agriculture) and added to them the n et output of the various branches of production w ith adjust­ m ents for th e duplications between these two sets o f figures. A djustm ent was m ade in some cases, such as mining, by deducting from the assessed incomes the am ounts know n to be included in th e net output of a given division; and in others by deducting from the n et ou tp u t of a division the esti­ m ated sum of assessed incomes in th a t divi­ sion. In o rd er to m ake these adjustm ents, incom e tax payers w ere classified by occu­ pation and by source of th eir capital in­ com es. T he largest duplication of capital incom es was in the interest on agricultural debt. Rao assum ed th at two thirds of the am ount of the agricultural debt (9 billion rupees) reported by the C entral Banking E nquiry represented bona fide agricultural debt. By applying thereto an interest rate of 9 p er cent, he arrived at a total interest on such debt of 540 m illion rupees, but re­ duced this sum to 370 m illion rupees to take account of tax evasion, overestimates of interest paym ents, and like items. In addition to including services from dwellings, previously om itted by him , R ao’s 1931-32 estimates contained one other im­ provem ent. T hey took account o f the in­ com es of individuals engaged in “auxiliary occupations,” nam ely, w orking dependents and people having supplem entary jobs in m anufacturing, trade, transportation, and professions. H e assum ed such incom es to am ount to 25 per cent in the first case and 33 per cent in the second, of the incomes of persons fully occupied in these industrial divisions. N either the incom e totals n o r the figures fo r the individual divisions are com parable in R ao’s two estimates (show n in Table 30-1) fo r two reasons: (a ) in 1925-29 the value o f raw m aterials im ported fo r agri­ culture and m anufacturing was included in the totals fo r these divisions and debited against the total incom e, while in th e 1931— 32 estimates these values w ere excluded from the gross value of the output in each division; ( b ) in the 1925-29 estimates no special item fo r incomes assessed under in­ com e tax was given, b ut in those for 1931-32 this item included incomes from all divisions except agriculture, forestry, and mining. O nly a few com prehensive statistical sources were available for R ao’s estimates

THE IN CO M E OF NATIONS

426

Table 30-1. Rao’s Two Estimates of National Income of British India for Average of Years 1925-29 and of 1931-32 1925-29 Million Division of Economic Activity Rupees 14,696 Agriculture a 139 Forestry 252 Mining Industry b 2,549 900 Trade and finance 690 Transportation 860 Public services 416 Professions 0 540 Domestic services

Total home-produced 21,042 income Deductions: 275 Imports of raw material Net capital income flowing 267 abroad Remittances by foreigners in 180 India 20,320 National income d

Per Cent 69.8 0.6 1.2 12.1 4.3 3.3 4.1 2.0 2.6

100.0

1931-32 Assessed Incomes and Division Million Rupees of Economic Activity Income assessed under income tax 0 2,161 Not-assessed incomes: 8,982 Agriculture f 92 Forestry 180 Mining 8 2,100 Industry h 1,233 Trade and finance 774 Services of dwellings Transportation e 329 58 Communications 6 636 Public services ‘ Professions 416 325 Domestic services Total home-produced 17,286 income Net income flowing abroad 1 -3 9 6

National income

Per Cent 12.5 52.0 0.5 1.0 12.2 7.1 4.5 1.9 0.3 3.7 2.4 1.9 100.0

16,890

a Including fishing and hunting. b Including electric light, power and gas, manufacturing, and construction. 0 By mistake Rao included 333 million rupees in his final table instead of 420 million as estimated on p. 147 (before deduction of cost items). d See Note c. The original rounded-off total is 20,230 million rupees (Reference No. 1, p. 155). e Including net deficit of railways and post offices for India, not including Burma. The total income of all public employees in transportation, communications, and administration was split up by deducting for each category the pro rata amount of assessed income. f Including fishing and hunting. 6 Assessed incomes in mining are included with this division. h Including manufacturing, electric light, power and gas, and construction. ' Including pensions and addition for interest on “unproductive” debt not considered elsewhere (see also Note g). j See Section 7, below. so urces:

Reference No. 1, pp. 147, 153-55;

and

of the national incom e of India and Burm a. R ao relied principally on the 1921 and 1931 Censuses of O ccupation (w hich provided the bases for the breakdow n of incom e re­ cipients by divisions o f econom ic activity) and on a variety of other official sources. W here com prehensive statistical data were lacking, Rao used the results of fragm en­ tary studies. Personal investigations played a greater role in his estimates than perhaps in any other m odern incom e estim ate. H e collected data fo r m any items by m eans of questionnaires, but these regrettably yielded only few returns and incom plete answers.

Reference No. 2,

pp.

161-62, 178-86.

b. Methods of Official NIU Estimates In the official estim ates the n et output m ethod was used fo r agriculture, animal husbandry, forestry, fishing, m ining, factory establishm ents, and fo r rental incomes. The incom e-distributed m ethod was used for all the rem aining sectors. T h e classification of the national incom e by divisions of eco­ nom ic activity did n ot differ m aterially from D r. R ao’s, except th a t th e new estimates w ere presented in finer breakdow ns. Unless otherw ise specified, the account o f the offi­ cial estim ates is based on the Final Report

30.

INDIA

427

Table 30-2. Official Estimates of National Income of the Indian Union by Industrial Origin, 1948-49 and 1950-51

Agriculture, animal husbandry, and auxiliary activities a Forestry Fishing Mining Factory establishments Small enterprises Communications (post, telegraph, and telephone) Railways Organized banking and insurance Other commerce and transport Professions and liberal arts Government services (administration) Domestic service House property N et domestic product a t factor cost N et earned income from abroad N et national output a t factor cost = national income

1948-49 (million rupees)

1950-51 (million rupees)

1950-51 Per Cent

47.8 0.7 0.4 0.7 5.5 9.1 0.4 1.8 0.7 14.0 4.7 4.3 1.3 4.1

1948-49 Per Cent 48.1 0.7 0.3 0.7 6.4 10.0 0.3 2.0 0.6 15.6 5.0 4.6 1.4 4.5

41.6 0.6 0.3 0.6 5.5 8.7 0.3 1.7 0.5 13.5 4.3 4.0 1.2 3.9 86.7 - 0 .2 86.5

95.5 - 0 .2 95.3

100.2 - 0 .2 100.0

100.2 - 0 .2 100.0

50.2 0.7 0.4 0.7 5.8 9.6 0.4 1.9 0.7 14.7 4.7 4.5 1.4 4.3

" These include processing, marketing, and auxiliary activities performed by the cultivator in respect of his own produce. so u rc e :

Reference No.

4, p . 106.

o f the N ational Incom e C om m ittee (R efer­ ence No. 4 ). T he estimates fo r 1948-49 and 1950-51 are show n in T able 30-2. Since it was necessary to derive the in­ come originating in some branches of the econom y by the incom e-distributed m ethod, estim ates w ere m ade of the total w orking force and its distribution am ong different occupations. T he population estimates for 1948-49, 1949-50, and 1950-51 were based on interpolations from the 1941 and 1951 Censuses of Population. Estim ates o f the factory labor force were also available fo r 1949 and 1950 from a Sample Survey of M anufacturing Industries. Statistics o f em ­ ploym ent in tea, rubber, and coffee planta­ tions were taken from governm ent reports; the same source provided inform ation on em ploym ent of railway, postal, and tele­ graph workers. In the fifteen years after R ao m ade his estimates, and especially after the form ation o f an independent state in 1950, m any im ­ provem ents were introduced in Indian sta­ tistics. T he N IU estimates w ere based on the regular official statistics of th e Central and State G overnm ents in addition to vari­ ous unpublished materials from the M inis­ tries and States, and the accuracy and com ­ prehensiveness of the estimates w ere closely

linked to the accuracy and com prehensive­ ness of the official statistics. T he Final R eport of the N IU was based, for the m ost part, on the same m ethods and sources underlying the estimates of its First R eport. T he estimates m ade in the Final R eport differed from those in the F irst Re­ po rt because of certain conceptual changes referred to above and because fuller use was m ade of statistical m aterial and new data. H ow ever, the estim ates for 1948-49, as published in both reports, differed less th an 1 per cent. T able 30-3 com pares the two estimates. W hile the Final R eport presents figures fo r three years, it should be noted th a t only those fo r 1950-51 w ere independent; those fo r the other two years w ere partly deriva­ tives. A t attem pt was m ade in the F irst Report to set up a fram ew ork for social accounts. Because data were unavailable, particularly fo r depreciation, consum ers’ expenditure, private capital form ation, and private sav­ ing, th e national accounts tables were dropped in the Final Report. However, sub­ sequent studies by some of the mem bers of the N IU provided independent estimates of certain social accounts. M ost im portant of these were the estimates o f G ross N ational Expenditure.

THE INCO M E OF NATIONS

428

Table 30-3. National Income by Industrial Origin for 1948-49 as Given in the First Report and in the Final Report (billion rupees) First Report 41.5 15.0 17.0 4 .6 4.5 4.7

Final Report 42.5 14.8 16.0 4 .0 3.9 5.5

N et domestic product a t factor cost Net earned income from abroad

87.3 —0.2

86.7 —0.2

Net national output a t factor cost = national income

87.1

86.5

Agriculture, forestry, and fishing Mining, manufacturing, and hand trades Commerce, transport, and communication Government services House property Other services (professions, domestics, etc.)

so urce:

Reference No. 4,

pp.

147-48.

2. Agriculture, Forestry, Fishing, and Mining a. Rao’s Estimates Separate estimates were made for the gross value of the output of crops other than fodder crops, livestock, fishing, and hunting; and for the various deductible costs, such as waste, seed, cattle, feed, and maintenance of farm and fishing equipment. Rao’s basic sources were the yearly agricul­ tural crop statistics that included informa­ tion on area under cultivation, normal (long run) and current yield per acre for the main crops, and harvest prices: Rao calculated the output of the main crops,6 accounting for approximately 80 per cent of cultivated acreage, by taking for each crop (a) the amount of acreage devoted to its cul­ tivation, (b) the estimated normal yield per acre based on long-time experience, and (c) the estimated current yield per acre expressed as a percentage of normal yield. The resulting figures were adjusted for the cultivated acreage not covered by official statistics and for under­ estimation of yields, e.g., as in the case of cot­ ton. The quantity output was then multiplied by the average harvest price, except for tea, coffee, castor, indigo, and rubber, to which wholesale prices were applied. For minor crops,7 only the acreage under cultivation was known, not the yields. Rao estimated the aver­ age yields per acre for each crop and province separately, from scant data furnished by in­ dividual village studies.

The estimates of the output of animal husbandry revealed certain peculiarities of India’s social and economic life. The reli­ gious laws of the Hindus prohibit meat con­

sumption, and cows are held mainly for re­ ligious purposes, manure production, and as draught animals. Consequently, the output of milk and meat in that country, which contains the largest cattle stock in the world, was small. The estimate of this output proved rather difficult, as the only avail­ able statistics were provided by the official Census of Livestock made at five-year inter­ vals. The methods used for estimating the various animal products were unique: In order to estimate meat production in 1925-29, a ratio of the number of animals slaughtered at slaughterhouses in various towns of the Bombay Presidency 8 to the correspond­ ing livestock of the Presidency in 1925 was applied to the livestock of the whole Indian territory covered. This computation was made separately for cattle and for goats and sheep, and an average price was assumed in each case.9 For the 1931-32 estimate, Rao related the reported number of cattle, goats, and sheep slaughtered in nearly 200 towns (ascertained by a personal inquiry) to the population of these towns; and calculated the number for the total urban population on the basis of these ratios, except that he increased it by 5 per cent for understatement. He multiplied the numbers by average meat weight, for cattle and for sheep and goats. Rural consumption of mutton was also estimated on the basis of consumption per head ascertained from village studies by the Punjab Board of Economic Enquiry. “To this we may add the figure of 50 million pounds for beef, on the assumption that on the average one head of cattle is slaughtered per year in each of the 500,000 villages in British India, mostly in celebration of the Mohammedan ‘Id Festival.’ ” 10 The results were checked against the older estimates made by Gaitonde, as well as against a study of the average life and mor­

429

3 0 . INDIA

tality rates of the animals made by the Imperial Institute of Veterinary Research. Finally, the quantities estimated were multiplied by a spe­ cially calculated average price. These estimates of the value of meat pro­ duced were supplemented by rough estimates of the increase of livestock in 1931-32 based on the census data and average prices recorded by the All-India Village Industries Associations, and on estimates of production of poultry, game, hides, skins, bones, and horns. “The she-buffalo rather than the cow is the m;7A:-producing animal of India.” 11 The aver­ age milk yield per head was low, as Rao’s aver­ ages showed. For 1925-29, he applied a milk production figure of 300 pounds per head to the total number of cows. For 1931-32, he used an average of 500-750 pounds for between 30 and 40 per cent of the cow stock, while the milk yield per buffalo was assumed by him to be 1,500 pounds in 1925-29 and 1,500 to 2,000 pounds in 1931-32. These estimates were made by averaging whatever sketchy figures were obtained. Information was more abundant for the second estimate, having been obtained in part through personal inquiries in various vil­ lages. A corresponding estimate for goat milk was made. Then the total estimate of milk production was checked against estimates of total milk consumption based on available frag­ mentary published materials and personal in­ vestigations on milk consumption per head. The results seemed to agree.12 As a final step in the calculation, the figure of the milk pro­ duced was multiplied by an average price, which for the 1931-32 estimate was carefully determined as the weighted average of urban and rural prices. Wool production was estimated on the basis of an average clip per sheep (1925-29, 3 pounds; 1931-32, 2 pounds, as estimated by the Royal Commission on Agriculture) multiplied by the average wholesale price. The output of fishing was estimated for 192529 from budget studies, while for 1931-32 the

value of the catch per fisherman was derived from various official documents and multiplied by the number of operatives. In the 1931-32 study rough estimates for raw silk and honey were included. Cost items deducted from the gross value of crops and livestock are shown in Table 30-4. W astage in T able 30-4 related to destruc­ tion o f crops by insects, rats, wild animals, m oisture, and deterioration. T he figures for both wastage and seed w ere based on offi­ cial estimates, especially those of the Fam ine Com mission. Feedstuffs of livestock covered only grains and oil cakes, since fodder crops w ere n o t included in th e gross output, and their values w ere determ ined from figures of their average consum ption p er head of livestock. A gricultural im plem ents are prim i­ tive an d poor in India. T hey consist m ainly of plows, spades, levelers, and w ater bags. T he annual cost of m aintaining these im ­ plem ents was derived from figures on the prices and life o f plows, then the result­ ing figure was doubled to account fo r the use of other implements. C urrent official statistics on tim ber and firewood production and exports w ere avail­ able, as well as estim ates on the quantity and value of domestic consum ption of forest products. These figures w ere adjusted to in­ clude only British India. In 1925-29 the estim ate o f forestry included shellac con­ sum ed and exported, and it was brought dow n to n et values by a flat deduction of 10 per cent. T he annual m ining statistics provided Rao w ith figures on the quantity and the value of the principal m inerals m ined. Only in the incom e estim ates fo r 1925-29 was a deduction of 10 per cent m ade fo r depre­ ciation and upkeep.

Table 30-4. Cost Item Deductions from the Gross Value of the Agricultural Output 1925-29 Estimate Millions of Rupees Wastage Seeds Feedstuff of livestock Repair and depreciation of farm equipment Repair of fishing equipment

7.5 per cent of crop value 2.5 per cent of crop value 50 lbs. per cow and bullock

949.5 316.5 624.5

—

0

—

0

1931-32 Estimate Millions of Rupees

3 per cent of crop value 235 6 per cent of crop value 470 — 600 10 per cent of value of farm equipment 64 10 per cent of gross value *

13

“ While Rao mentions a rate of 15 per cent (Reference No. 2, p. 117), a check of his computation leads to the ratio of 10 per cent. so urces:

Reference No. 1, p p . 150-51; Reference No. 2, p p . 75-81, 117,

430

THE IN CO M E O F NATIONS

b. Official NIU Estimates T he N IU ’s estimates of the quantities and values of crops were based on m ore sub­ stantial official data than R ao’s had been. T he average yields continued to be based on eye estim ation by the village patwaris (revenue officers). It should be noted that the official inform ation on the total acreages devoted to various crops was very im perfect even at the tim e of the estimates of the F inal R eport inasm uch as 24 per cent of India’s land area was still unsurveyed. The patwaris’ estimates of the average yields for any current period were generally calculated by them as a ratio to an average yield taken over a period of years. According to NIU es­ timators, this method involved heavy reliance on the personal judgment of the patwari and was probably responsible for substantial under­ statements of the actual yields in good years and overstatements in lean years. The crops were valued for the most part at wholesale prices reported for the important market cen­ ters of the country. Home consumption was similarly evaluated at these prices. The net values of the crops continued to be calculated, as they had been in Rao’s estimates, on the basis of varied information obtained from official and unofficial studies as to the ratios of the costs of seed, manure, repairs, etc., to the gross values of the crops. For example, the costs of repairs and maintenance of equipment was estimated at 20 per cent of the value of the equipment, which figure was obtained, in turn, from the National Sample Survey.13 The foregoing calculation covered the reporting area of the Union. To cover the nonreporting areas (areas not covered by data from the Ministry of Food and Agriculture), which amounted to about 25 per cent of the total land area, only 4 per cent was added to the net value of output of the reporting areas because the nonreporting area contained a great deal of waste land. Estimates of the value of livestock and live­ stock products in the Final Report were ob­ tained from the quinquennial Livestock Cen­ suses of 1945 and 1951, which covered more than 90 per cent of India’s area.14 The estimates for 1948-49, 1949-50, and 1950-51 were de­ rived by linear interpolation between 1945 and 1951. The quantities and prices of animal products were obtained from the official Mar­ keting and Inspection Reports. The deductions necessary to obtain the net values of these products were estimated on the basis of a variety of official and unofficial information and were found to equal 47 per cent of their aggregate gross value.15 H unting incom e was obtained by taking the num ber o f persons engaged from the

census and applying to it an arbitrary rate of earnings. F o r fishing, relevant data on production, price, and value w ere taken from a 1949 M arketing R eport on fish. N et value was obtained by a m ore or less arbi­ trary deduction from gross value of 5 per cent to cover m aterials an d depreciation. T he results were checked against incom e and consum ption data and w ere found to be som ew hat understated. T he ou tp u t of forest products was based o n official inform ation fo r rep o rted areas, extended to cover u n reported areas on the basis o f one th ird the yield p er acre of w oodland for tim ber and two thirds the yield fo r fuel of th a t in the reported sec­ tions. N et value was estim ated by deducting 10 per cent in the F irst R eport and 5 per cent in the Final R ep o rt fo r costs an d de­ preciation. G ross value of m ineral production was easily estim ated inasm uch as official data of recent date relating to this field of activity w ere satisfactory. T he ou tp u t was valued at pit-head prices an d deductible costs were calculated on the basis of th e balance sheets o f m ining com panies as reported to the Re­ serve B ank of India.

3. Manufacturing and Handicrafts a. R ao’s Estimates W hile in his estim ates of the n et o u tput of m anufacturing and handicrafts fo r 1925-29 R ao com bined th e n et output, the final ex­ penditure, and th e incom e-distributed m eth­ ods, in those fo r 1 9 3 1-32 he applied the latter m ethod only. In his earlier estim ates, he had few data w ith w hich to w ork, even fo r the large-scale o r facto ry industry th a t covered only ap­ proxim ately 28 p er cent o f the entire indus­ trial o u tput and em ployed 13 per cent of all industrial w orkers. F o r his later estimates, how ever, he had the help o f the 1931 census an d of wage statistics, b ut even these had to be supplem ented by special question­ naires sent to a sam ple o f representative large-scale enterprises. T h e returns to this questionnaire, how ever, w ere poor, w ith only 113 o ut of 8,145 replying. T h e bulk o f industrial activity in India is carried on by sm all enterprises and selfem ployed persons, such as artisans in handloom weaving, clothm aking, woodwork, m etalw ork, ceram ics, food, building con­ struction, cleaning, w ashing, etc., and last,

30.

b ut n ot least, the socially and econom ically im portant occupation o f the Indian barber, who, as Rao says, is frequently capable of using “his knife in a fashion th a t would excite the jealousy of a surgeon.” 16 H e esti­ m ated their earnings on the basis of the national and provincial wage censuses. b. O fficial N I U E stim a te s T he official N IU estimates for the indus­ trial sector w ere based on occupation and earnings statistics that had been greatly im ­ proved during the war. O ne of the im por­ ta n t sources of d ata on average wages in different industries was the annual returns collected by regional chief factory inspectors u nder the P aym ent of W ages A ct. But even these returns om itted extra paym ents, did n o t distinguish adequately betw een different occupations, and were otherw ise incom plete in their coverage and uneven in their re­ sults. O ther im portant sources fo r estim at­ ing the value of the output w ere the Census of M anufactures, w hich covered alm ost half of all industries com paratively com pletely, and the Sam ple Survey of M anufacturing Industries (S S M I), w hich covered the w hole range b ut used only samples fo r each indus­ try. Since the sample survey did not give inform ation on depreciation, this item was developed from depreciation rates fo r cov­ ered industries given in the Census. N et output o f factory establishm ents fo r 1948-49 was estim ated from the 1949-50 figures by appropriate production and price indexes. T he N ational Incom e U n it experienced the same difficulty as Rao had in estim ating incom e from sm all enterprises, a substan­ tial portion of w hich are carried on in the household. E ven the Census o f O ccupations did n ot furnish adequate inform ation on this sector o f the econom y: By subtracting factory employment from the reported total number of persons occupied in industry, the number engaged in small enter­ prises was obtained, a procedure which did not allow for unemployment. Next, the aver­ age income of these workers was estimated on the basis of a fairly wide collection of case studies and some localized sample surveys, and adjustment was made for data not relating to the period under consideration. When data per­ tained to an earlier period, they were inflated by an index calculated on the basis of the rise in average earnings of the factory workers dur­ ing the corresponding period. Incomes relating to days, weeks, or months were converted into annual earnings by multipliers of 300, 52, and 12 respectively. Since these multipliers were

INDIA

431

somewhat arbitrary, as the average number of working days per year was not accurately known for small-scale industries, the NIU in­ dicated that this procedure may have introduced considerable error into the annual earnings es­ timates. Coverage was not complete in regard to the various States and wherever actual fig­ ures of earnings were not available earning figures in similar industries were used.

4. Trade, Finance, and Transportation a. R a o ’s E stim a te s In R ao’s 1925-29 estim ate, incom e from trade was estim ated in accordance w ith the net output m ethod by com puting the value of goods handled by trade and applying an average m arkup. Rao assum ed th at one half of the gross value of agricultural produce, tw o thirds of the industrial output, and all forest and m ineral products entered trade, and th a t 10 per cent of the total value thus obtained was the gross profit in trade. H e allowed a 10 per cent deduction fo r upkeep and depreciation, taking the balance as net output in trade. H e gave no figure fo r in­ com e from finance. In R ao’s 1931-32 esti­ m ate, income from trad e was calculated, in accordance with the incom e-distributed m ethod, by assuming th a t it was equivalent to the average earnings of factory workers, and m ultiplying the figure involved by the num ber of incom e tax-exem pt persons en­ gaged in trad e and finance. Incom e from transportation was estim ated by R ao as follows: For 1925-29 wages and salaries paid by railroads, which are State owned, were esti­ mated at 50 per cent of gross receipts, while in the 1931-32 study they were computed from budget estimate figures adjusted for incomes assessed under income tax. Interest was not taken into account, partly because most of the capital stock of the railroads was foreign-owned and partly because, in the 1931-32 estimates, interest was included in assessed incomes. The railway deficit in the latter year was deducted. Gross earnings of water transportation in 192529 were put at 5 per cent of the value of merchandise imported, exported, and trans­ ported by coastal vessels. An addition of 10 per cent was made for inland water transpor­ tation and about 25 per cent of total earnings was deducted for fuel, repair, and depreciation. The gross earnings from road transportation were estimated in Rao’s 1925-29 estimates at 2 per cent of the sum of one half of the value of the agricultural produce and two thirds of the industrial output, reduced by 10 per cent

432

THE INCO M E OF NATIONS

for repairs and depreciation. For 1931-32 he estimated it by taking the number of persons employed in this type of transportation, as re­ ported in the Census of 1931 and multiplying it by the average wages of skilled and unskilled workers in factory occupations.

b. Official NIU Estimates T he N IU estim ated the incom es from trade, finance, and transportation by the incom e-distributed m ethod, using census data and various scattered inform ation on earnings in com parable occupations. F or banks, insurance, and co-operative societies, direct published business data on their in­ comes were available. N et incom e from in­ surance was com puted as the sum of com ­ pensation of em ployees plus operating sur­ plus. Im putation o f income fo r banks and insurance com panies was m ade on the basis of a sample analysis of balance sheets and profit and loss accounts of banking com ­ panies, carried on by the Reserve Bank of India. Im puted values of business insurance prem ium s paid by different sectors of pro ­ ductive activity were estim ated and deducted from the net value contributed by the rele­ vant sectors.

on the statistics of the number of houses (ob­ tained from housing censuses and by interpola­ tion for the intervening years) and from in­ formation obtained from 91 municipalities on the average annual rental incomes per house. A deduction of 10.75 per cent of gross rental was allowed for repairs on the basis of sample returns covering 28 municipalities; and the reported collections of taxes on houses were also deducted. For rural housing, the National Sample Sur­ vey gave preliminary all-India data regarding the average value of rural houses, their age distribution, and the annual repair expenditure per household. These data together with census data on the number of rural houses were used for the imputation of income from rural house property. The number of occupied rural houses was estimated by linear interpolation between the figures of the 1941 and 1951 censuses. The average current value of a rural house in 195051 was obtained from the NSS and was ap­ plied to other years with adjustments based on wholesale food price indexes. As income from rural property is almost entirely an im­ puted item, the rental value of rural houses was estimated on the basis of rural interest rates. An average of 10 per cent was used to obtain the gross rental of rural houses from the estimated gross value thereof. The NSS provided data on depreciation, cost of repairs, and maintenance.

5. Services of Dwellings a. Rao’s Estimates T he income from the services of dwellings was determ ined separately fo r tow ns and fo r villages fo r 1931-32. T he num ber of houses by areas was reported in the 1931 census. A nnual net ren t paid in Bom bay, C alcutta, and M adras, and the proceeds of the real property tax in the m unicipalities were available in curren t reports, while the average tax rate on rent was ascertained by a personal inquiry. F rom these elements, the rent paid in cities was calculated and an assum ed average rent was applied to rural houses, w hich are mostly huts.17 As the “annual value” m eant net rent, no ad­ justm ent was necessary for the m aintenance and depreciation.

b. Official NIU Estimates O n the basis of census data, th e N IU di­ vided all house property in the country into two classes, u rban and rural: For urban housing, data on rents available from municipal tax records were not relied upon to any large degree because the results might not be comparable with those obtain­ able for rural regions. Main reliance was placed

6. Government, Professions, and Domestic Service a. Rao’s Estimates In R ao’s first estim ates, incom e from gov­ ernm ent included only wages an d salaries paid by provincial and central adm inistra­ tions as com piled from budgetary reports. In his second study, he also included salaries of local civil servants (estim ated on the basis of personal investigation), pensions, an d the interest paym ent on domestically held “productive” debt. A djustm ents were m ade fo r incomes assessed fo r the incom c tax. T he salaries of em ployees of the mail, telegraph, and telephone services w ere ap­ parently included w ith those o f public serv­ ants in the 1925-29 estim ates. In the second study, however, they w ere show n separately. Incom es from professions w ere estim ated by R ao fo r 1925-29 on the basis of income tax data and by assuming, som ew hat arbi­ trarily, th a t exempt incom es w ere six times the size of taxed incom es and by deducting 1 p er cent from the total fo r repair and m aintenance of office equipm ent. In his later estim ate, R ao simply applied an appropriate

30.

assum ed average incom e to the num ber of persons in each profession. In b oth of his estimates, R ao calculated the incom e of dom estic servants by m ulti­ plying the reported num ber of servants by an average incom e, in the first case set “after careful consideration,” and in the second, decided on th e basis o f special inquiries. H e reduced the num ber of servants by 10 per cent in order to elim inate the servants engaged in the business w ork o f their em ­ ployers.

b. Official NIU Estimates T h e n et value contributed by governm ent was divided into tw o sections: general ad­ m inistration and business enterprises. G ov­ ernm ent business enterprises (railroads, post office, etc.) w ere included along w ith p ri­ vate business in the appropriate sectors. W ages and salaries w ere estim ated in some cases as fixed ratios of reported governm ent expenditures. T hus, out of the total govern­ m ent expenditure on construction, wages and salaries w ere taken at 33.3 p er cent on the advice o f the Chief E ngineer o f the C en­ tral Public W orks D epartm ent; fo r all other works, wages and salaries were arbitrarily taken to be 50 p er cent of total expendi­ tu re.18 Estim ates had to be m ade w here governm ent sources of inform ation did not clearly distinguish wages and salaries from purchases o f m aterials o r other m iscella­ neous expenditure. Budget data of all levels of governm ent w ere the prim e source of inform ation. E stim ating the value added by professions and dom estic services presented difficulties to the N IU since, apart from th e census statistics, there was no regular inform ation bearing on the em ploym ent situation in these occupations. Dividing the num ber of gainfully em ployed in this sector between u rb an and ru ra l areas was also difficult to do. N evertheless, an estim ate was m ade on the basis of estim ated average earnings in each category. T he figures thus arrived at were com pared w ith the estim ated value of the net output of these sectors derived from an analysis of the som ew hat m ore com pre­ hensive, but not particularly representative, d ata available on consum er expenditures.

7. N et Income from Abroad a. Rao’s Estimates Because India was a debtor country at the tim e R ao ’s two estim ates w ere m ade,

INDIA

433

and incom e was flowing o ut of India to a greater extent th an it was com ing in, the national incom e was reduced accordingly. In the 1925-29 estimates, the interest re ­ ceived on holdings o f foreign capital was n o t included in national incom e. C orre­ spondingly, only dividends from foreign p ri­ vate industrial investm ents in India (esti­ m ated in a special study at 8 p er cent of £ 2 5 0 millions in v e stm e n t)19 and rem it­ tances by foreigners (roughly estim ated) were deducted from the total. In the estimates fo r 1931-32 R ao derived the figure to be deducted from the official figures on the balance o f paym ents in the following w a y :20 Credit Items Merchandise Gold Total Debit Items Dividends and interest (net) Other services (net) Capital movement

Millions of Rupees 223 580 803 Millions of Rupees 349 179 275

Total

803

T he difference between the export of m er­ chandise and gold (803 million rupees) and the outflow of capital (275 m illion ru p ees), nam ely, 528 m illion rupees, corresponded to net paym ents of dividends, interest, and outlays fo r services, m ainly shipping. This am ount was reduced by 25 per cent in order to exclude th e In d ian States and th e re­ m ainder, 396 m illion rupees, was deducted from the national income.

b. Official NIU Estimates T he N IU ’s deduction fo r net income from abroad was based on balance of paym ents inform ation com piled by the Reserve Bank of India.

8. Reliability of the Estimates Rao estim ated his m argin of erro r for the national incom e total at only ± 6 per cent, b u t he m ay have been overly optimistic. His error margins fo r individual categories of incom e ranged from 10 p er cent for live­ stock products to 20 per cent for fishing and hunting, and 15 per cent to 20 per cent for incomes n ot subject to incom e tax. T he N IU , w hich h ad a larger am ount of

THE INCO M E OF NATIONS

434

data on w hich to base its estimates, placed its probable over-all m argin of error at 10 per cent, w ith the lowest possible m argins being attributed to large-scale industries and a 33.3 per cent m argin fo r small enter­ prises as well as certain other sectors, as show n in T able 30-5. Table 30-S. Possible Margins of Error in NIU Estimates, 1948-49 Per Cent 1. Agriculture, animal husbandry, and 20.0 auxiliary activities 25.0 2. Forestry 25.0 3. Fishery 10.0 4. Mining 5. Factory establishments 10.0 33.3 6. Small enterprises 7. Communications (post, telegraph, 10.0 and telephone) 10.0 8. Railways 10.0 9. Organized banking and insurance 33.3 10. Other commerce and transport 33.3 11. Professions and liberal arts 12. Government services (administra­ 10.0 tion) 33.3 13. Domestic services 33.3 14. House property so urce

:

Reference No. 4, p. 146.

II. Final Expenditure or Product Estimates (Mukherjee and Ghosh) T he estimates of national incom e and expenditure prepared by M ukherjee and G hosh related to the year 1949-50. Because of the lim ited nature of the available sta­ tistical data, th e authors used various short cuts, some of w hich are particularly inter­ esting. T he following account, based on the estim ators’ own published statem ents, gives a fair idea of the nature of the problem s th at faced them and the m ethods they em ­ ployed to solve them.

1. Consumer Expenditures M ukherjee and G hosh estim ated consum er expenditures on the basis of sam ple surveys m ade by the N ational Incom e C om m ittee of the consum er expenditures of representa­ tive fam ily households and other sources. The survey yielded fairly com prehensive data on the average expenditures per house­

hold o f the ru ra l population, w hich was then m ultiplied by the num ber of such households to obtain a total. B ut the calcu­ lation for the u rb an classes was m uch m ore difficult, inasm uch as the available statistics covered only the u rban w orking and middle classes, and even w ith respect to these, were old an d had to be brought to date w ith the help of cost-of-living indexes. T here were no statistics fo r the o th er u rb an groups, som e o f w hich were considerably poorer and others considerably rich er th an those surveyed. T he estim ators calculated con­ sum er expenditures o f the u rb an working an d m iddle classes on the basis of their household budgets in representative urban centers, developing fo r each a weighted average fo r the co untry as a w hole. N ext, on the basis of occupational statistics and scanty data on the earnings of salaried em ­ ployees and wage earners in factories, and even scantier d ata on the standards of living of other u rb an groups, the estim ators arbi­ trarily adopted the ratio of 3 to 1 for the n um ber of w orking class to m iddle class household budgets. O n the basis of this ratio, they calculated a single w eighted aver­ age of household expenditure fo r th e urban classes and m ultiplied it by th e estim ated n um ber o f u rban households. F o r b oth rural and u rban households they showed th e p ro ­ portion of food to nonfood expenditures. F o r th e ru ral this proportion was 66 to 34 p er cent, while fo r th e u rb an it was only 53 to 47 p er cent.

2. Gross Domestic Investment M ukherjee and G hosh m ade tentative calculations of th e investm ent expenditures fo r 1949-50, using a m odified com m odity flow approach. T hey could n o t use it w ith­ out m odification because the m anufacturing census covered only 29 groups of industries an d left o ut a substantial n um ber of im por­ ta n t capital goods industries, thus m aking it im possible to segregate products in each case into final and interm ediate ones. T hey em­ ployed, therefore, the follow ing short cut: they assum ed th at the n et im ports an d do­ m estic production of capital goods during the year represented the capital form ation of th a t year. A dm itting th a t such an approxim a­ tion would not be true w ith regard to indi­ vidual firms im porting o r producing these goods, they held th a t it “w ould n ot be incor­ rect fo r th e country as a w hole . . . the lag

3 0 . INDIA

between im port and actual use w ould be com pensated by inventory changes. . . .” 21 T hey departed from the pure com m odity flow m ethod in estim ating directly the value of construction and building activities and the value of governm ent capital outlay. They reconciled the figure o f the form er to the com m odity flow approach by isolating and om itting all building m aterials from the im ­ p ort and production statistics o f capital goods, thus avoiding double counting. As to governm ent capital outlay, reconciling the estim ate o f its value to the com m odity flow approach and avoiding double counting was m ore difficult, but the offsetting advantage was th a t the figure, taken from governm ent reports, was m ore com plete and accurate th an any th a t could be obtained by the other m ethod. The figures of net imports of capital goods were taken at c.i.f. prices at ports or other places of entry from the fairly comprehensive statistics on imports that are rather elaborately broken down by types of products. They were taken net of all building material, railway stores and equipment, postal and telegraph equipment, electrical goods, and all metals, etc., inasmuch as capital formation on government account and for construction were estimated directly. An over-all allowance of 2.5 per cent was made for clearing, forwarding, and hauling charges; 5 per cent for import duties; and 5 per cent on the f.o.b. value for the distributive margin. In the case of heavy machinery, an additional 10 per cent was allowed for the cost of in­ stallation. The gross value of domestic output of capital goods in the factory establishments sector was taken from the Sample Survey of Manufac­ turing Industries for 1949. Among these goods were included: sewing machines, 50 per cent of which were treated as capital goods in in­ dustry and 50 per cent as consumer goods; producer gas plants, general and electrical en­ gineering (manufacturing only, exclusive of re­ pairs, of which 67 per cent was treated as capital goods and only 33 per cent as consumer goods), automobile and coach building (trucks and commercial vehicles only), shipbuilding, aircraft repair and manufacture of textile ma­ chinery. An over-all allowance of 20 per cent for all distribution charges was made. Ma­ terials purchased by the railways were de­ ducted. Finally, the value of steel allocated to private industry (other than steel processing industries) for the fabrication of replacements of machinery and other structures was added to the estimate of domestic output of capital goods. On the other hand, the gross value of output of capital goods in the small enterprises and

435

households sector of the economy, for which the statistics were much less complete, was es­ timated on the basis of the working force en­ gaged in the following small-scale productive activities: a. Excavation, well-sinking, etc. b. Smelting or iron ore; blacksmiths, etc., of which 80 per cent was treated as capi­ tal goods work. c. Sawyers, carpenters, joiners, etc., of which only 25 per cent was treated as capital formation, inasmuch as construc­ tion had been independently estimated. d. Coach building, etc. e. Carriage painters, upholsterers, etc. The authors stated that a good portion of these expenditures could be expected to relate to capital formation in agriculture, as in the case of excavation and digging of land, build­ ing, well-sinking, etc., repair and fabrication of country-made iron and wooden plows and other agricultural tools and implements, bullock carts, etc. Therefore, these expenditures would include repairs, and because a good proportion of them were nonmonetary in nature, could not be traced to any visible source of monetary savings. Urban construction was estimated on the basis of the allocation of iron, steel, and cement for building purposes during the year 1949-50, and on the application of the ratios of iron and steel costs and cement costs to total construc­ tion costs in residential and nonresidential struc­ tures separately. These ratios were obtained from the records of large construction com­ panies as well as the government public works department. The ratios adopted for iron and steel costs to total costs were as follows: in private residential buildings, 5 per cent; in government residential buildings, 6.5 per cent; and all nonresidential structures, such as fac­ tories, etc., 25 per cent. The results thus ob­ tained were checked against direct informa­ tion on government construction given in gov­ ernment budgets and were found to be reason­ ably accurate. On the other hand, rural construction was estimated on the basis of the preliminary re­ sults of the National Sample Survey. Urban construction was also estimated independently on the basis of the records in sample munici­ palities of the value of the building permits issued, including new construction, additions, and repairs, but the resulting figure was found to be an underestimate because the sample of municipalities did not include the larger urban areas in proper proportion. Government capital expenditures—including both those of administrative departments and those of government enterprises reported in the budgets—were independently estimated, but in order to reconcile them with the over-all estimates based on the commodity flow ap­

436

THE INCO M E OF NATIONS

proach, all government purchases of both im­ ported and domestically produced machinery and equipment were omitted from the calcula­ tions. Furthermore, the authors of the esti­ mates excluded all developmental expenditures by administrative departments that were being financed from the general revenue, such as re­ search projects and grow-more-food campaigns, even though a considerable portion of such ex­ penditures, such as those of the grow-morefood campaign, appeared to be of a capital nature. They also excluded all inventories of administrative departments, all purchases of existing private assets that were transfers, all purely financial capital payments, such as commuted pensions, and finally, all civil works establishment expenditures met out of the gen­ eral budget; while, on the other hand, they included all construction financed out of the revenue budget (subject to the deductions, al­ ready stated, of items already accounted for under the general commodity flow approach). T he foregoing item s taken together com ­ prised gross domestic capital form ation, ex­ cept fo r one rem aining item , inventory changes. F o r governm ent enterprises in­ cluded in the budget, this item was esti­ m ated from the budget statem ents of their borrow ing accounts. F o r the private sector, estimates could be developed only fo r the stocks held in m anufacturing on the basis of data given in the Census of M anufac­ tures and in the Sample Survey of M anu­ facturing Industries, and for trading stocks in cotton and jute on the basis of the re­ ports of the Indian Central C otton C om ­ m ittee and the Indian Jute Mills A ssocia­ tion. A ll other trading stocks being u n ­ know n, it was deem ed best to ignore changes in inventories in the private sector alto­ gether, i.e., to consider them to be zero and to include only inventory changes of gov­ ernm ent enterprises. Similarly, changes in livestock w ere ignored because of the ex­ ceedingly incom plete and crude nature of the Livestock Censuses of 1946 and 1951, and the questionable nature of any possible interpolation of these figures fo r the year 1949-50. T he balance o f the current transactions with the rest o f the world was calculated w ithout m uch difficulty from the current balance o f paym ents statistics of the D epart­ m ent of Research and Statistics o f the Re­ serve Bank of India. This balance, w hich happened to be negative in 1949-50, was deducted from the adjusted total of gross dom estic capital form ation. T he calculation th a t finally em erged appears in T able 30-6.

3. N et Capital Formation Depreciation, w hich had to be deducted from gross capital form ation to obtain net capital form ation, was particularly difficult to estim ate because of the lack of data and because of conceptual problem s. T he depre­ ciation rates allowed by the C entral B oard o f Revenue for incom e tax purposes were fo u n d to be largely inapplicable fo r social accounting purposes, inasm uch as they took no account of the extra w ear and tear of m achinery during w ar an d also perm itted the accum ulation of depreciation fo r non­ replacem ent of capital. M oreover, such in­ com e tax data on depreciation covered only a p a rt of the private sector. T he pertinent inform ation for governm ent enterprises was available in governm ent budgets. F o r agri­ culture the First R ep o rt of the N ational In ­ com e C om m ittee contained estimates of the depreciation of agricultural im plem ents, bul­ lock carts, and other equipm ent. D eprecia­ tio n fo r factory establishm ents was devel­ o ped from the Sample Survey data, on the basis o f depreciation rates given in the Census of M anufactures fo r 29 groups of industries. F o r m ining, some conventional estim ates w ere m ade on the basis of balance sheets and profit and loss accounts o f a sam ple o f joint stock com panies. F o r small enterprises, depreciation was roughly esti­ m ated as a percentage o f the gross value of their o u tput on the basis o f such rela­ tionships existing in factory industry and in agriculture, inasm uch as data on their capital assets w ere unavailable. F o r tran s­ p o rt, the N ational Incom e C om m ittee’s re­ p o rt supplied estim ates of depreciation and repairs o f all vehicles, including airplanes fo r civil aviation. D epreciation of buildings in the private sector, such as trade, bank­ ing, and insurance, was om itted, because all construction activity h ad been estim ated in ­ dependently o f th e various sectors served by it. N ext, both gross and n et capital fo rm a­ tion w ere allocated to th e different sectors o f dom estic production. T he allocation was largely arbitrary. Item s 3 and 4 in T able 30-6, after deducting capital purchases of governm ent, w ere allocated m ainly to largescale industry including m ining, and to pri­ vate transport. Item 5 was allocated to gov­ ernm ent. Item 6 was attributed largely to agriculture and sm all enterprises. Table

30.

INDIA

437

Table 30-6. Gross Capital Formation of Ind'a, 1949-50 Rs. Crores

1. Rural Construction: new outlay maintenance

164 73 237

2. Urban Construction: a. Government (i) residential: new maintenance (ii) nonresidential: new maintenance

21 7 25 4 57

b. Private (i) residential: new maintenance (ii) nonresidential: new maintenance

92 28 15 6 141

3. a. Imports of capital goods other than construction material and material used by govern­ ment (e.g., telegraph and telephone equipment) b. Markup to cover distribution costs

124 27 151

4. a. Domestic output (large scale) of capital goods b. Steel allocated to private industry not enumerated in 4a c. Markup to cover distribution costs in respect of 4a

64 11 13 88

5. Investment by government administrative departments; railways, post and telegraph; other government departments and projects; exclusive of construction and purchase of machinery, etc., already enumerated in 3a and 4a: new maintenance

88 39 127

6. Gross value of output of a capital nature, of small enterprises a. excavation, well sinking, etc. b. blacksmithy, etc. c. carpenters, etc. d. coach building, etc. e. carriage painting, etc. *

60 99 96 8 3 266

7. Gross (product flow) estimate of capital formation a t home (exclusive of inventory adjust­ ments) 1,067 8. Add inventory accumulation a. in state trading schemes b. with private sector

—22 n.a.* 1,045

9. Deduct deficit on current account of foreign transactions 10. Gross domestic capital formation during 1949-50 * Not available. so u rc e:

Reference No. 4, p p . 64-65.

—23 1,022

438

THE INCO M E OF NATIONS

Table 30-7. Sectoral Breakdown of Domestic Capital Formation in India, 1949-50 Rs. Crores Net Capital Formation

Sector Government 1. Construction 2. Administrative departments (including multipurpose river schemes) 3. Railways 4. Other government enterprises 5. Inventory changes in state trading schemes 6. Total government

Depreciation and Maintenance

Gross Capital Formation

46

11

57

59 31 60 -2 2

22 29 3 —

81 60 63 -2 2

174

65

239

Private 7. Construction: Urban 8. Construction: Rural 9. Agriculture, animal husbandry, forestry, fishery, and cottage industry 10. Mining, factory establishments, small-scale industry, and trade and transport 11. Inventory changes

107 164

34 73

141 237

28 92 * 53 n.a.

217 81 * 120 —

255 173 n.a.

12. Total private

362

444

506

13. Total goverment and private capital formation a t home

536

509

1,045

* Depreciation: Factory establishments Rs. 38 crores; Mining Rs. 1 crore (both a t Central Board of Revenue rates); Private Transport (NIU estimate) Rs. 42 crores. Since CBR depreciation rates are low, depreciation allowance of Rs. 39 crores has been arbitrarily doubled to Rs. 78 crores. Net capital formation taken as residual. so u rce:

Reference No. 5, p. 66.

30-7 shows net capital form ation at hom e thus calculated for 1949-50. T he N ational Incom e Com m ittee’s net output estim ates fo r agriculture and other prim ary production and its incom e-distributed estimates for other sectors gave a total national figure of approxim ately Rs. 9,100 crores. M ukherjee and G hosh’s estimates of Rs. 1,045 crores for gross dom estic capital form ation and of Rs. 536 crores fo r net do­ mestic capital form ation w orked out, there­ fore, as 11.5 per cent and 6 per cent of the national incom e. N et capital form ation in governm ent am ounted to nearly one third of total national net capital form ation. “If the alm ost entirely im puted items of rural construction and investm ent in agriculture, etc., are om itted,” w rote the estim ators, “capital form ation on governm ent account will be seen to be of the order of 52 per cent of total net investment, i.e., m ore than half o f total net investm ent represented by a dem and fo r savings in m oney term s.”

4. Savings Available for N et Capital Formation As a final step in th eir estim ates o f the disposition of In d ia’s national incom e, M uk­ herjee and G hosh developed the analysis, reproduced in T able 30-8, o f the sources of available m onetary savings fo r net capi­ tal form ation. T he total savings available fo r capital form ation appeared to be considerably less th an the authors’ estim ate of n et capital for­ m ation (Rs. 536 cro res). H ow ever, they de­ ducted ru ral construction and investm ent in agriculture, etc., from th e latter, since these w ere im puted or entirely self-financed. The estim ators then arrived at a figure o f Rs. 374 crores for net capital form ation at­ tributable to m onetary savings; and, because of the closeness of this figure to the figure fo r available savings o f Rs. 382 crores, they felt th a t their net dom estic capital form a­ tion estim ate was fairly accurate.

3 0 . INDIA

It will be observed that th e sum of con­ sum er expenditures obtained by th e authors fo r 1949-50 (Rs. 8,710 cro res), net do­ mestic capital form ation other than agricul­ ture and rural construction (Rs. 374 cro res),

439

and negative net foreign capital form ation (Rs. 23 crores) approxim ates very closely the figure of Rs. 9,100 crores of net n a­ tional incom e obtained by the net output and incom e-distributed methods.

Table 30-8. Savings Available for Capital Formation in India, 1949-50 Rs. Crores 1. Surplus on current account of government administration 2. Operating surplus of government enterprises (railways, post and tele­ graph, etc.) 3. Small savings (postal, cash and national savings certificates, P.O. savings bank accounts, etc.) 4. Provident fund contributions of central state government employees 5. Life insurance premiums 6. New share-capital and increase in deposits of co-operative societies, etc. 7. New issues of equity subscribed by the public 8. a. undistributed profits of corporate industry 0 b. estimated investible profits of unincorporated enterprises 9. Borrowing from abroad

99 27 30 15 41 18 ‘ l4 b 70 60 23 d 382

10. Total

“ Share-capital of co-operative societies Rs. 6 crores; increase in deposits and loans of co-operative societies Rs. 12 crores. Debentures of land mortgage banks have been excluded as these are largely sub­ scribed by provincial governments. b Total issues sanctioned by the controller of capital issues were roughly Rs. 26 crores of which roughly Rs. 12 crores were bonus shares. 0 Undistributed profits of corporate enterprise have been estimated as follows: Rs. Lakhs 1. Corporate income assessed to income tax in 1949-50 2. Less income and super taxes in 1949-50

18,560 —8,014

3. Net income 4. Less dividends in 1949-50

10,546 —3,695

5. Undistributed profits in 1949-50

6,851

d This item is not counted as savings. The total is computed without it. so u rce:

Reference No. 5, p. 67.

5. Projections Under the Second Five-Year Plan In d ia’s second five-year plan aims at an annual increase in national incom e of 5 per cent, com pared to an estim ated population increase of 1.4 per cent; and at a step-up in

public and private net investm ent from 7.3 p er cent of national incom e in 1955-56 to 10.7 per cent by 1960-61. T o achieve these goals both fiscal and m onetary measures and direct state controls are to be used (U nited N ations, E conom ic Survey o f Asia and the Far East, 1957, C hapter 3 ) .22

31. BRAZIL

Reference 1. “Estimativa da Renda Nacional do Brasil, 1947-1949,” Revista Brasileira de Economia, September, 1951, No. 3, pp. 216 (entire num ber), including summaries in English and French (see also the September issues, 1952-55, for later estimates). T he Brazilian estim ate fo r 1947-49, the first com prehensive and truly systematic esti­ m ate for that country, was prepared by the Brazilian Econom ic Institute. It was calcu­ lated by the incom e-distributed m ethod, ex­ cept fo r the income from agriculture, w hich was estim ated by the net output m ethod. T he underlying concept was th a t of national incom e at factor cost. T he national incom e was estim ated ac­ cordingly as the total of the following six types of incom e: (1 ) com pensation of labor in all branches of the econom y except agri­ culture; (2 ) profits, except in agriculture; (3 ) interest; (4 ) rent; (5 ) incom e from agriculture; and (6 ) net incom e paid abroad (see T able 31-1). T he later estimates (1 9 5 0 -5 4 ) gave also breakdow ns of the aggregates by industrial origin and by final expenditure.

em ployees w ere ascertained from govern­ m ent reports. T hey represented the m ost reliable p a rt of the entire estim ate. In all com putations of salaries and wages, rem u­ neration in kind (food, lodging, etc.) was included. E m ployers’ contributions to social security w ere added as supplem ents to labor incom e. T he labor incomes of self-em ployed w ork­ ers in various occupations, including those of th eir indirectly rem unerated fam ily m em ­ bers, w ere estim ated in a m ore roundabout fashion. F irst of all, their num bers reported in the 1940 Census w ere extrapolated in the sam e m an n er as in the case of domestics. N ext, on th e basis of sam ple investigations, ratios w ere developed, on the basis of a 1948 survey, o f the average earnings of the self-employed to th e average wages of w orkers in the same occupations. These ratios w ere then applied to the know n aver­ age wages paid in these occupations in the o th er years. T he resulting estim ates of average earnings of the self-employed in those years were then m ultiplied by their estim ated num bers. In the case of occupa­ tions fo r w hich average earnings could not be developed in this way, the average earn­ ings of oth er sim ilar occupations w ere sub­ stituted. In the case of professions, the Census did n o t report the num ber of persons engaged in them separately fro m the n um ber of salaried persons in the sam e occupations, w hether it be education, health services, or adm inistration. It was necessary, therefore, to estim ate the num ber o f professionals for each year on the basis of ratios developed in a special survey in 1945 of the num ber o f such persons to the num ber of salaried em ployees in the sam e occupations.

1. Labor Income W ages and salaries in em ploym ents cov­ ered by Social Security were estim ated on the basis of the current collections of the social security taxes and of the ratios of taxable to nontaxable portions of the wages and salaries revealed by the 1948 Survey of W ages. W ages in dom estic em ploym ents and in service industries were calculated by taking the num ber of employees as reported in the 1940 Census, extrapolating it to later years, and multiplying the resulting num ­ bers fo r each year by the legal m inim um wage adjusted fo r changes in the cost of living. A m inim um rath er than an average wage was taken because no inform ation was available as to the actual wages paid in these em ployments. W ages and salaries paid to governm ent 440

3 1 . BRAZIL

T he im puted labor incom e of proprietors in various types of business other than agri­ culture was taken as reported by proprietors in th eir tax returns u n d er the category of labor incom e (as subject to low er tax rate s).

2. Profits C orp o rate profits were com puted before corporate taxes and were estim ated on the basis of tax data. T hey were fu rth er broken up fo r certain years into corporate profits taxes, dividends, and retained profits. Profits of unincorporated enterprises w ere com ­ puted after deductions of the im puted labor incomes of their owners. T hey w ere like­ wise estim ated from tax data. A djustm ents were m ade fo r the underreporting o f profits and also fo r differences between the defini­ tions of profits under the incom e tax and the national incom e concept of profits.

3. Interest and Rent Incom e from interest was estim ated from incom e tax data. Interest on the public debt was excluded as a transfer paym ent. I t was impossible to elim inate all intercorporate interest paym ents, but it was believed th a t any resulting double counting was probably m ore th an offset by the underreporting of interest incom e. R ental incom e was calcu­ lated on the basis of property tax data. N et ren t was calculated by deducting 10 per cent from the gross rent.

4. Income from Agriculture A gricultural incom e was com puted by the usual net output m ethod. L ack of data m ade it impossible to segregate it into wages, profits, interest, and entrepreneurial income. T he totals obtained included, there­ fore, the incom es of hired help, of selfem ployed w orkers, and corporate and unin­ corporated enterprises. It was acknow ledged th at the figures obtained in this p a rt of the estim ate were, fo r the m ost part, approxi­ m ations based on insufficient statistical m a­ terial. T h e gross values of 29 m ost im portant crops (bananas, cocoa, coffee, tea, pine­ apples, alfalfa, cotton, rice, potatoes, yam s, sugar cane, etc.) were obtained from the yearly published official estimates. T he sum total of these values was next raised 10 per

441

cent to account for the gross values of some 23 additional m inor crops. N et value was obtained from these figures on the basis of special studies of the costs of agricultural production. These costs, covering seeds, fer­ tilizers, insecticides, etc., am ounted in the aggregate to approxim ately 10 per cent of th e gross value. D epreciation was n o t in­ cluded in these costs. H ow ever, the am ount of capital em ployed in Brazilian agriculture is small and, therefore, this omission was not considered to be im portant. N o account was taken of the costs of erosion and de­ pletion of the soil due to im proper utiliza­ tion. T he gross value of anim al production was com puted by estim ating separately the value of m eat production, the accretion to exist­ ing livestock, and the values of oth er animal products. T he production of m eat (beef, pork, lamb, and goat) was estim ated on the basis of reports o f governm ent-inspected slaughter­ houses. T h e figures did n ot include slaugh­ tering for own consum ption and to th a t ex­ tent were underestim ated. On the basis of reported price data fo r 1947 and 1948, n a­ tional w eighted average prices were devel­ oped for each one of the foregoing cate­ gories of meats. N o price data being avail­ able fo r 1949, the 1948 figures w ere used fo r that year. T he value of poultry m eat was definitely underestim ated as it was based only on data on poultry killed at govern­ m ent slaughterhouses. Changes in livestock w ere com puted only fo r (a ) the m ajor categories of animals raised for m eat production— cows, calves, pigs, sheep, and goats— and ( b ) w ork ani­ m als— horses, donkeys, and mules. T he esti­ m ates were based on official figures fo r 1947 and 1948; and fo r 1949, on the basis of the 1948 statistics and of the reproduction co­ efficient obtained from the 1940 Census. M ilk production was obtained by m ulti­ plying the num ber of cows by the annual production p er cow in 1939 as derived from the 1940 Census of A griculture. This figure was m ultiplied by the prices prevailing in the m ost im portant dairying states (M inas G erais, Sao Paulo, and Rio de Jan eiro ). Egg production was estim ated by taking the average annual production per hen derived from the 1940 Census of A griculture and applying it to the num ber of hens estim ated to exist in 1947-49. T he value of this pro-

442

THE IN CO M E OF NATIONS

duction was calculated on the basis of the average retail price of eggs in the capitals of various states and the relation of this retail price to the prices paid to producers in the four m ost im portant egg producing states (M inas Gerais, Rio G rande do Sul, Santa C atarina, P a ra n a ). It was adm itted th at the prices used in the evaluation of both m ilk and eggs w ere based on incom ­ plete data and m ight have contained a sub­ stantial m argin of error. T he deductions for the costs of anim al production to obtain its net value included only two items— alfalfa and corn consum ed by livestock. Because of lack of inform ation other costs were not subtracted. It was con­ ceded, therefore, that the net value of anim al production was overestimated. T he value of hides and products obtained by hunting, w hich were com paratively m inor items, was estim ated on the basis o f the re ­ ported exports of hunting products. The export values contained the costs of tanning and other processing and to that extent were overestimates. On the other hand, fo r lack of inform ation the value of hunting fo r consum ption in Brazil was not estim ated. It was believed th a t the omission of the latter item just about offset the overvaluation of

the exported products. T he value of fishing was derived directly from published statis­ tics o f the M inistry. T he value of the production of some 19 im portant extractive plant products, such as rubber, nuts, jute, liana, wax, etc., was ob­ tained from official statistics. T he value of tim ber production fo r 1947 was available in the official reports. F o r 1948 and 1949 the Instituto N acional do Pinho (N ational Institute of Pine W ood) published the tim ­ ber production figures fo r the fo u r southern states. T he national totals for these years w ere derived fro m these figures on the basis of the proportion th a t the production of these states bore to the total national production in 1947 as revealed by the 1947 reports. T he average price o f tim ber re­ ported by the M inistry of A griculture for 1947 was extrapolated to the other years on the basis of wholesale prices fo r pine wood in Rio de Janeiro and Sao Paulo and of the average value of the exported tim ber. It was fu rth e r assumed th a t prices paid to p ro ­ ducers vary concom itantly w ith wholesale prices. T he quantity o f firew ood and charcoal produced in 1947 was available in official statistics. Estim ates fo r 1948 and 1949 were

Table 31-1. National Income of Brazil 1947-49 (million Cr$) I. Compensation of labor, except agriculture 1. Wages and salaries a. Private b. Government 2. Supplement to wages and salaries 3. Income of own account workers 4. Income from professions 5. Other labor income II. Profits, except agriculture 6. Corporate profits before taxation a. Corporate profits tax b. Dividends c. Other income payments d. Undistributed profits 7. Other profits III. Interests IV. Rent V. Agriculture VI. Net income paid abroad a VII. National income at factor costb

1947

1948

1949

63,460 41,302 31,097 10,205 1,215 12,251 3,710 4,982 15,557 6,373 928

72,769 46,251 34,602 11,649 1,358 14,207 3,863 7,090 15,534 6,494 937 2,813

9,184 1,262 4,284 42,331 -6 6 2

9,040 1,519 5,501 50,773 -1 ,1 6 6

87,437 55,007 41,291 13,715 1,723 16,733 3,978 9,995 18,010 8,907 1,253 3,210 1,130 3,314 9,103 1,775 7,148 57,238 -1 ,1 6 0

126,233

144,930

170,447

a The surplus of net income accruing abroad over net income actually paid abroad amounts to 1,319 million Cr$ in 1949. b Components do not always add to totals because of rounding. so u rc e:

Reference No. 1, p . 140.

32.

BRITISH W EST INDIES AND MALAYA

based on the reported production in those years in the State of M inas G erais, the greatest single producer of them , and the assum ption that the 1947 share of the pro ­ duction of that state in the total rem ained unchanged. T he average prices paid for these products in 1947 were available in official statistics. It was assum ed th a t their average price in the country as a w hole during the other years varied in the same proportion as their reported prices did du r­ ing these years in M inas G erais.

5. N et Income from Abroad T he balance of incom e on international account was a negative item . It was com ­ puted as the excess o f dividends, branch profits, interest, and the investm ent incom e

443

paid to foreigners over the am ounts received from these same sources from abroad. The inform ation was based on exchange control data issued by the Banco do Brasil. The figures referred to am ounts actually tran s­ ferred in foreign exchange and did n o t in ­ clude profits of branches and dividends of subsidiaries n ot rem itted abroad b ut rein­ vested in Brazil. T otal incom e accruing to nonresidents was derived from incom e tax statistics, but unrem itted am ounts w ere ex­ cluded. A separate estim ate of this item was added, however, in order to define the am ount of reinvested earnings of nonresi­ dents. T he figure of such incom e was a negative item, m ore m oney being paid out to abroad th an was received from abroad. T his was tru e of both dividends (th e larger figure) and interest (the smaller su m ).

32. BRITISH WEST INDIES AND MALAYA References 1. Frederick Benham, The National Income of Jamaica, 1942, Development and Welfare in the West Indies, Bulletin No. 5. ------ , The National Income of St. Vincent, 1942, ibid., No. 8. ------ , The National Income of Barbados, 1942, ibid., No. 9. ------ , The National Income of Granada, 1942, ibid., No. 12. ------ , The National Income of Malaya, 1947—49, Singapore, 1951. ------ , “The Estimation of National Product in Underdeveloped Countries,” Bulletin of the International Statistical Institute, vol. XXXIII, Part III, December, 1951. 2. A. P. Thorne, Size, Structure, and Growth of the Economy o f Jamaica, Institute of Social and Economic Research, Kingston, 1955. 3. International Bank for Reconstruction and Development, The Economic Development of Malaya, Johns Hopkins University Press, Baltimore, 1955.

1. Importance of Benham’s Estimates F rederick Benham , onetim e Econom ic A dviser to the C om ptroller for D evelop­ m ent and W elfare in the W est Indies and later Econom ic Adviser to the Com m is­ sioner G eneral for the U nited K ingdom in South E ast A sia (Singapore), prepared in 1942-43 national incom e estim ates fo r the fo u r colonies of the British W est Indies, Jam aica, Barbados, G ranada, and St. V in­ N otes begin on p. 544.

cent, and in 1950 a national income esti­ m ate for M alaya fo r the years 1947-49. These estimates w ere distinguished by the application of m odern national incom e con­ cepts and estim ating techniques to under­ developed countries possessing only the scantiest statistics. T he various short cuts em ployed by Benham are suggestive of w hat m ay be done u nder such circum stances by skilled technicians. F o r this reason Benh am ’s estimates are reviewed here in fair detail. T heir statistical results, however, are

444

THE IN CO M E OF NATIO NS

not reproduced here, as the reader will readily find them in the original sources cited above. In the 1950’s, the preparation of national income estimates in Jam aica was taken over by the Institute of Social and Econom ic Research, and in M alaya by the Registrar o f Statistics. In both countries the figures prepared by Benham were carried forw ard to later years.

I. Estimates for British West Indies B enham defined national incom e as the net value of the output of the economy, i.e., as the net value of hom e production (after deduction of depreciation, indirect taxes on domestic output, and duplicated costs) plus net incom e from abroad. The statistics he had at his disposal w ere exceed­ ingly scant. Population censuses in the four colonies had been held infrequently and such as had been held were n o t detailed; and no agricultural and m anufacturing cen­ suses had been held in any of them . D e­ tailed statistics w ere restricted to prices, foreign trade, governm ent, banking and in­ surance operations, utilities, production or sales of tobacco, cigarettes, and certain foodstuffs, and foreign trade. In Jam aica, however, the production and consum ption statistics were m ore am ple th an in other colonies, particularly as related to m anu­ facturing, w hich was m ore im portant there than in the other colonies. A s a result Benham had to rely in a great portion of his estim ates on his personal knowledge of the conditions in the islands, personal in­ quiries, inform ation supplied by govern­ m ental agencies, extrem ely roundabout processes, and short cuts. All the three m ethods— net output, in­ com e distributed, and final expenditure— were used by him, but the first was relied upon most. In fact, he aim ed at determ in­ ing the aggregate net output at each stage of production including distribution, de­ ducting as far as possible at each stage the m aterials and supplies used and other dupli­ cating costs. Because of the underdeveloped nature o f the economies and inadequacy of their statistics, he classified the divisions of econom ic activity som ew hat differently from the m ethod generally used fo r m ore ad­ vanced economies. Thus, he classified do­

m estically produced goods into three cate­ gories: exported products, including stocks awaiting export; u n m anufactured foodstuffs; an d m anufactured foodstuffs, including elec­ tric light and pow er and gas. M ost of the o th er divisions— construction, finance, serv­ ices of dwellings, transportation and com ­ m unications, governm ent services, domestic services and professions, personal services, an d trades— conform ed m ore or less with the general standard, except th a t th eir con­ tents w ere n o t always the same. T rad e as a division was represented only by the “m arketing of im ports” inasm uch as the services of m arketing of dom estic products, w hether fo r internal use or fo r export, were accounted fo r in the value of the products (w hich was taken dow n at retail prices). Be­ cause of the lim itations of statistics the value of im ported raw m aterials was de­ ducted from the total value of the domestic net o u tput in the over-all am ount, rather th an separately from the net ou tp u t of each division using them . Benham varied his techniques from colony to colony, in accordance w ith the variations in th eir statistics. T hus in the case of Jam aica, w hich was economically m ore advanced and had better statistics than the others, he used m ore refined estim at­ ing techniques. T he details of his proce­ dures w ere as follows: The internally produced unmanufactured foodstuffs, in Benham’s classification, consisted of foodstuffs reaching trade or the final con­ sumer without further processing as well as those consumed by the farmer producing them himself. For Jamaica, he estimated their value by multiplying an estimated per-capita-consumed quantity of each type of foodstuff by its average retail price. As regards the output of root crops, such as yam, in Jamaica for 1942 he found no available data and had to depend on guesses of experts as well as on consump­ tion data. Thus he found that in the principal town of Jamaica, Kingston, the consumption of root products amounted to about 1 lb. per head per day. He was advised that, in the country districts, the consumption of them was about double that in Kingston. He made sample checks, asking families to allow him to weigh their daily ration of these products, con­ sulted the records of caterers supplying workers’ camps, and the records of prisons and other in­ stitutions. He arrived at the conclusion that the average consumption of these products was just over 2 lbs. per head per day; and this conclu­ sion was subsequently borne out by the results of an agricultural census. The value of internally produced manufac­

3 2 . BRITISH W EST INDIES AND MALAYA

tured commodities was likewise estimated by Benham, at least for Jamaica, by multiplying the known or estimated produced or consumed quantity by an average retail price. The value of bread, however, was estimated from the quantity and prices of raw materials entering its production. The annual value of domesti­ cally produced furniture was assumed to be three times the average value of imported fur­ niture in 1930-38. The net value of construction was estimated exclusive of maintenance and repairs, the cost of which was included in the value of the serv­ ices of dwellings or of the services of govern­ ment. It was estimated roughly, except in gov­ ernment construction, for which direct statistics were available. The costs of materials were deducted. The value of the services in marketing im­ ports was estimated as a percentage markup on the import value as a whole, or separately for each product. Services o f dwellings in the case of Jamaica and Granada were evaluated by multiplying an estimated number of dwellings by an average gross rental. No deduction was made for de­ preciation and maintenance costs inasmuch as these were excluded from the value of con­ struction. Transportation included only passenger trans­ portation, inasmuch as the costs of the services of freight transportation were accounted for in the value of the final products. In Jamaica the value of railway and street car transportation was estimated roughly at 50 per cent of the gross receipts of railways. The services of com­ munications were also estimated roughly. The value of imported raw materials and supplies was deducted from the total value of domestic output, as already stated, either in toto or separately from the value of each product. Other duplications were also deducted, such as replacements of existing materials and equipments, certain communication and bank­ ing services and certain excise duties. They were deducted either in a single over-all item from the national total, or separately from the different components of the total. N et income from abroad was estimated fairly accurately on the basis of a substantial body of statistics. In computing it, subsidies on banana and other exports were included. Remittance of emigrants were included; as, too, were pay­ ments to workers of the American Naval Base at Jamaica. B enham analyzed the national incom e of Jam aica by factor shares on the basis of tax statistics and wage statistics; and he broke down personal expenditures by type of prod­ ucts, on the basis of a regrouping of the production estimates and im port statistics. Savings were estimated by him roughly by adding incom e taxes to consum ption ex­

445

penditures and deducting the total from n a­ tional incom e. This was an exceedingly im ­ perfect calculation inasm uch as the total of these items approached m ore closely an ag­ gregate of private incom e than the national income.

II. Estimates for Malaya In his estimates fo r M alaya, Benham fu r­ th er refined his estim ating techniques and covered m uch vaster grounds th an in his W est Indies estimates. In addition to esti­ m ating the national incom e by the n et o ut­ p u t m ethod, he developed also subsidiary estimates, estimates of distributed incomes, and of expenditures fo r consum ption and investm ent, using fo r these purposes in p art independent sources an d em ploying these subsidiary estim ates as checks on the others. The results, u n d er all the three estimates, w ere fairly close.

1. Estimates of the N et Output B enham ’s procedures u nder th e net o ut­ p u t m ethod w ere as follows: Gross output o f agriculture. Benham calcu­ lated the output of various types of vegetal prod­ ucts—rubber, rice, cocoanut, palm oil, fruits and vegetables, etc., largely on the basis of their harvested acreage, which was often different from the planted acreage, and the average an­ nual yield per acre. He allowed for the double cropping and for the differences in the yields in different agricultural districts. Benham esti­ mated the value of the agricultural output at their export or retail prices, which included the costs of distribution, justifying this short cut on the ground that these prices were more readily ascertained and more reliable than prices at the farms. First of all, the latter varied greatly from district to district according to the distances from the market. Secondly, he observed that farmers were inclined to understate their re­ ceipts from the sales because of fear that the information might be used against them in the assessment of taxes. The value of meat production was estimated by him on the basis of the figures of slaughtered cattle reported by the Veterinary Department; the average dressed weight per animal and the average retail price per unit of weight. The officially reported figures of the numbers of poultry in the country were found by him to be gross underestimates. He raised them 50 per cent. Finding that the average age of a bird when killed was 1 year, he accepted the num­ ber of birds in existence at any date as being

446

THE INCO M E OF NATIONS

equal to the number killed during the year. Taking the average weight of the birds (for chickens and ducks separately) and then retail prices per unit of weight, he obtained their total retail values. The value of milk production likewise was computed by him in two steps. First, he took the official figures of production of marketed milk and applied to the portions consumed in the country and in Singapore their respective retail prices, which were $5.50 per gallon in the country and $9.00 per gallon in Singapore. Next, he estimated the value of the milk con­ sumed on the farm, by taking the official fig­ ures of the number of milch cows, estimating their total milk production, and deducting from it the milk fed to the calves as well as the marketed milk, and then applying to the re­ sidual the appropriate retail prices. The value of egg production was separately estimated by him for Singapore and South Johore where a large egg-producing industry existed, on one hand, and for the “up country” on the other. In the first case, the reports that seemed to be accurate indicated that there were one million birds kept for breeding and laying, each laying on the average 100 eggs a year, thus making 100 million eggs a year. At a re­ tail value of 17 cents an egg, this amounted to $17 million. The figures for up country had to be estimated more roughly. Taking the estimate of the number of birds, it was assumed that 30 per cent were layers and that they laid on the average 40 eggs a year. The total output of $25 million in this case was accordingly set at 156 million eggs with a retail value at 16 cents an egg. Gross output of fishing. The gross value of fishing was estimated separately for sea fish and fresh water fish. The first is organized com­ mercially and the weight landed is officially re­ corded. In his Malayan estimate, Benham raised the official figure 25 per cent to allow for un­ derstatements and omissions. He then applied to the quantities of the different types of fish the average retail prices as reported by in­ spectors. As there were 65,000 sea fishermen, he allowed a certain amount of fish per indi­ vidual for family consumption, valuing it at the price applicable to the cheapest grade of fish. Fresh water fish in Malaya is caught mostly for home consumption. The value of it was devel­ oped in consultations with the Fisheries De­ partment. Timber and forest products. The Forestry Department reported fairly good figures of the annual production of sawn timber, hardwood, poles, firewood, charcoal, “free timber” (allowed by the government to be collected by local resi­ dents on payment of a small royalty) and other forest products (bamboo, ratan, etc.). No diffi­ culty was experienced in applying proper values to these products. Mining. The main mining output of Malaya is tin. Nearly half of it is obtained by dredg­

ing operations requiring use of extensive equip­ ment and conducted by large European compa­ nies; about one third of the output comes from small-scale Chinese operators who use the less efficient gravel-pumping method. The figures of output are reported officially and are fairly re­ liable. The British government is the sole pur­ chaser of the output and fixes the price. Hence the export value of it, which includes export duty, transport, and smelting charges, is pub­ lished regularly and is completely accurate. Gross output o f manufacturing. In calculat­ ing the manufacturing output, Benham had to do a considerable amount of personal research, inasmuch as governmental statistical data were exceedingly incomplete. For most manufactur­ ing industries no statistics of output were avail­ able. He found the questionnaire method to be useless, as he had no way of insuring complete and accurate returns. Therefore, he had to rely on information regarding the average gross output per worker, given to him by individual manufacturers whose confidence he was able to gain. He then multiplied this figure by the num­ ber of workers in the industry as ascertained from population and occupational censuses or other sources. The censuses and other sources also gave him the number of establishments in most industries. Where only few firms were en­ gaged, he was able to obtain from one of them an estimate of the total output of the industry. If the products of the industry were subject to an excise duty, as was the case with locally produced beer, stout, and matches, he got the figures of their marketed output from the tax receipts. In some cases where an industry used an important material whose quantity was known because it was imported or was other­ wise controlled by the government, and where the quantitative relation of that material to the product was known, he was able to estimate the output therefrom. Thus having the figures of flour used by bakers, and knowing that 1,000 pounds of it, with other ingredients, produce 1,320 pounds of bread, and that a pound of bread costs at retail 25 cents, he readily ob­ tained the value of bread. As in the case of agriculture, so in the case of manufacturing, Benham evaluated the output of some products (rubber, cocoanut oil, etc.) at export rather than retail prices. His reasons for doing so were given above. The value of the output of rice wine, known as samsu, was estimated for the legally pro­ duced part of it from the records of the 20 licensed distilleries engaged in its production. Its retail price was $40 per gallon in the coun­ try and $39 in Singapore, of which $18.75 was the excise duty. There were numerous illegal “one-man” stills. Benham estimated their out­ put at 25 per cent of the legally produced vol­ ume. Its price in the country was only $24 a gallon, but in Singapore its price was the same as of the legally produced wine.

3 2 . BRITISH W EST INDIES AND MALAYA

Trade. In Malaya there is a large entrepot trade flowing through her free ports. It consists of the re-export to neighboring countries of imports from the west (textiles, etc.) and of imports from neighboring territories of their products (rubber, copra, etc.) for reprocessing and grading and then re-export to the Western countries. The statistics of this trade supply all the necessary data for the computation of the value added in it. But the value added in the trading activities of the local hawkers and other small traders could be only roughly estimated on the basis of their reported number, probable value of goods handled, and probable margin of profits. Private transport and communications. No difficulty was experienced in making the neces­ sary computations for transportation and com­ munications in the hands of large-scale enter­ prise, such as railroads, cable, and wireless, as all the necessary figures on gross receipts and deductible costs were available in their case. But the situation was different in the cases of buses, taxis, trishaws (side cars propelled by cyclists). There it was necessary to base the estimates on the number of reported vehicles and their estimated daily receipts and expenses per vehicle. Service industries. The value added in service industries—hotels, restaurants, cinemas, other entertainment places, laundries, hairdressers, private schools, banks, insurance companies, pawnbrokers, etc.—were estimated by Benham on the basis of the reported numbers of estab­ lishments in each case, an estimate of their probable volume of business, gross receipts, and expenses. Professional services and domestics. The value added in professions was estimated on the basis of the number of persons in each profession, their probable gross earnings, and expenses. On the other hand, the services of domestics were evaluated directly on the basis of their reported numbers, average wages, and average value of the lodgings supplied to them. Food was not included, since the general rule in the country is for the domestics to provide their own food. Services of dwellings. The value of these services was estimated at rents actually paid or at imputed rents, on the basis of sample inves­ tigations and of the reported numbers of dwell­ ings of various types. Free services of government and its enter­ prises. The value of these government services was computed net of the purchases from busi­ ness. The emoluments of civil and military servants include rental subsidies and a vast variety of compensation in kind. Interest on debt incurred for public improvements was in­ cluded. The value added in government enter­ prises was computed on the basis of completely adequate data, in the usual way. Deductible costs. Benham was unable in the case of agriculture and many industrial divi­

447

sions to obtain data on all deductible costs. He was compelled, therefore, in the case of cer­ tain raw materials or other intermediate prod­ ucts, to make block deductions from the gross value of the entire national product. He had to do this for certain imported materials, whose quantity and value was known, such as fuel and containers used in industry, and for certain locally produced materials such as timber used in buildings, manufacture of furniture, etc. In cases where particular materials or services were used partly in industry and partly by con­ sumers, such as gasoline, electricity, and serv­ ices of transport, he had to estimate what pro­ portion was used in the one and what propor­ tion in the other. On the other hand, in the case of a number of industries—particularly those bordering on service industries, such as those of gold and silver smithing, tailoring, dressmaking, em­ broidering and rice-milling—the estimator com­ puted the net output or value added directly, rather than by computing gross value. Treatment o f taxes and subsidies. In comput­ ing the net output, indirect taxes on imports of consumption goods were not deducted by Benham. He took the position that they resem­ bled the personal income tax in that they were paid out of incomes of the local population after these incomes were received. On the other hand, export taxes and excise duties on the local output were deducted by him on the ground that they are paid out of the gross re­ ceipts of the business and not out of the in­ comes of the factors of production. Subsidies paid by government to local output were added to the net value of that output inasmuch as they were used to pay the factor shares in the subsidized industries. Gross investment. Benham sought to include under investment all additions to physical as­ sets, including the increased value of land con­ verted from waste into cultivated land or of land improved by irrigation or drainage. He also sought to include in it all renewals, re­ placements, and repairs other than those enter­ ing into the normal running costs of producing a commodity or service on the ground that the renewals, replacements, and repairs “make good the wastage of assets due to depreciation.” He further exemplified his theoretical position by saying that “if the two items were exactly equal, then in order to estimate correctly the Value of the Net National Product either both should be included (offsetting one another) or both should be omitted.” In estimating total investment, he found a valuable guide in the values of the goods, whether imported or locally produced, which were wholly or mainly investment goods, such as engineering products, machinery and cement. To avoid double counting he had previously included their values in his block deductions from the gross output of all the industrial di­ visions. He also estimated and included in in­

448

TH E IN CO M E O F NATIONS

vestment the total amount of labor employed and wages earned in construction work, land improvements, and other types of capital crea­ tion. For some items of capital investment such as the value of new buildings, he was able to obtain data from local authorities; and for gov­ ernment capital creation he used government accounting records. His estimate of gross investment comprised the following items: new construction and re­ habilitations of properties by public authorities; other building; rehabilitation of industries in­ cluding repair (in rubber planting, tin mining, etc.), making good the destruction and neglect of properties and equipments during the war (under Japanese occupation); new machinery and equipment; major repairs, replacements, and renewals. Depreciation and capital consumption. The final step in Benham’s computations of the net national production was a deduction for de­ preciation and capital consumption of the phys­ ical assets of the country. Benham defined de­ preciation as the loss in the value of assets as a result of their normal deterioration through use over time. A good example is the deteriora­ tion of the rubber trees after a certain period of years as a result of annual tappings. Capital consumption, on the other hand, was defined by him as the losses in capital assets as a re­ sult of fires, floods, and neglect of proper main­ tenance. His estimate of depreciation was ac­ knowledged by him to be “little more than a guess.” He offered no explanation whatsoever of the basis of his “guess.” Deduction for net balance of payments to residents abroad. The foregoing calculations gave the home-produced income of the coun­ try. To obtain from it the figure of the national income of the residents of the country, Benham had to deduct the net balance of payments due to the residents abroad on their investments in the country. He did not deduct, however, the remittances sent by the residents of Malaya to their families or friends in China, India, and elsewhere, inasmuch as they were given volun­ tarily and not in return for any services; nor did he add to the Malayan national income the grants made by the British government or the remittances sent home by Malayan seamen from overseas. Neither did he include in the national product the services rendered by the British forces or the expenditures made by the British government for the maintenance of these forces.

2. Estimates of Final Expenditures T he two m ain categories o f final expendi­ tures in Benham ’s subsidiary estim ates were consum ption and gross investm ent. G overn­ m ent expenditures were not show n sepa­ rately b u t w ere included in these tw o cate­ gories.

a. Consumption Expenditures Consumption included the value of all con­ sumers’ goods and services in Malaya, both domestic and imported, including free govern­ ment services. The goods and services were valued at retail market prices less indirect taxes. Corresponding values were imputed to some nonmarket transactions, such as consumption of home-grown food. Consumption expenditures were classified in rather detailed categories as follows: 1 food; drink; hotels, restaurants, etc.; clothing, shoes, etc.; cigarettes, etc.; cars, bicycles, and repairs, etc.; petrol, etc.; household requisites, etc.; serv­ ices; rental value of all houses; free govern­ ment services. The rental value of all houses and the value of free government services were obtained di­ rectly from the net output estimates.2 The values of domestic agricultural, forest, mineral, and manufactured products were obtained from the net output estimates with some adjustments; e.g., deductions were made for exports, indirect taxes, and the portion of the products used by business. The value of imported consumption goods was computed by subtracting the value of re-exports from official statistics of imports, and then deducting the value of retained im­ ports for industrial purposes from the total of retained imports. b. Investment Expenditure Gross domestic investment consisted of new construction, building, and all other additions and improvements made during the year to physical assets (excluding consumer durable goods, such as private motor cars). It also in­ cluded maintenance and repairs, renewals, re­ placements, and depreciation. The value of the physical change in stocks was not included. The reasons given for this omission were three­ fold: (1) the available data were not complete; (2) the data were dominated by price changes; (3) the value of the physical change in stocks was not large in relation to the gross invest­ ment figure.

3. Comparison of Estimates by the N et Output and Final Expendi­ ture M eth ods3 It was necessary to m ake tw o adjustm ents in th e estimates of consum ption plus in­ vestm ent in order to m ake them com para­ ble to th e estim ate of n et national product m ade by th e n et o u tput m ethod. First, it was necessary to add to consum ption the value o f unrecorded im ports of rice, opium , and gold.4 Second, the deficit in the balance o f paym ents (i.e., the n et value of invest­ m ent financed by foreigners) was deducted. A fter these adjustm ents w ere m ade, the

32.

449

BRITISH WEST INDIES AND MALAYA

two estimates of net national product were fairly close. The final expenditure estimates exceeded the net output estimates in 1947, 1948, and 1949 by $82 million, $43 million, and $82 million respectively. In percentage terms these differences amounted to ap­ proximately 2.5 per cent in 1947, 1.3 per cent in 1948, and 2.6 per cent in 1949. These excesses were due mainly to the fact that all taxes on output were deducted to obtain the net output estimates but only excise duties and entertainment duty were deducted from the final expenditure esti­ mates. Benham states, “It is correct not to deduct Export Duties from the latter, but we should deduct Import Duties (mainly on Petrol other than for private consumption) and ‘Other Taxes on Output’ (licenses on commercial vehicles, etc.) as these enter into the prices of goods and services con­ sumed locally. When allowance is made for these two items, our two sets of estimates correspond fairly closely with one another.” 6

4. Estimates of Private Expenditure and Saving Several adjustments were made to the esti­ mates of consumption in order to estimate the actual sums spent on private consump­ tion, including the imputed value of the consumption of home-grown food, etc. One adjustment, the addition of the value of un­ recorded imports has already been men­ tioned. Secondly, a deduction was made for government subsidies to rice producers.6 A third adjustment was the addition of import duties, excise duties, and entertainment du­ ties on consumers’ goods and services. Lastly, the depletion of stocks of rice for

1947 were added, and the increase in stocks of rice for 1948 were subtracted.7 Table 32-1 shows the amount that was T able 32-1. T he Am ount of Private Savings Available for Investm ent in M alaya, 1947-49 1947

1948

1949

N et national product N e t income paid abroad

3 ,2 1 6 20

3 ,2 6 0 39

3,069 47

N ational income Private expenditures

3 ,1 9 6 3 ,0 4 8

3,221 3,0 1 7

3 ,0 2 2 2,852

D irect personal taxes

148 53

204 81

170 156

A m ount available for private savings

95

123

14

s o u r c e : Benham, The National Income of M alaya, 1947-49, pp. 166-68.

available for savings after deducting total private expenditure and direct personal taxes from national income (estimated by the net output method). Benham estimated that over half the amount of savings was sent abroad as re­ mittances. Thus, the estimates of private savings available for investment were well below the estimates of investment. However, Benham points out that the investment esti­ mates were not too high since (1) some in­ vestment was financed by foreigners, (2) there was a great deal of government invest­ ment, and (3) enterprises probably spent part of their reserves on investing for re­ habilitation purposes, since they expected to get their money back from War Damage Compensation Payments.

Part Four. Developm ents in Sixty-Six Other Countries

EXPLANATORY NOTE TO PART FOUR This part of the book presents a panoramic view of the progress made in the preparation of national income estimates and analyses in recent decades and particularly during the postwar period 1946-57 in some sixty-six countries other than the thirteen selected for special analysis in Part III. While covering a much larger number of countries this part covers each country much less intensely than is done in Part III. The countries are arranged in five groups, partly along ethnological and linguistic lines and partly along geographical lines, and each group is made the subject of a special chap­ ter. The arrangement is as follows:

concepts, methodological approaches, and sources used in their preparation is indi­ cated; the types of income aggregates cov­ ered in the estimates (whether net national income at factor cost or market prices, or home-produced income only, or the aggre­ gate of private incomes, or the gross na­ tional product) are identified together with the years to which these aggregates relate; the nature of the breakdowns of these aggre­ gates (whether by industrial origin or dis­ tributive shares, or by types of final ex­ penditure) is stated; the most important uses made of the estimates are indicated; and the principal bibliographical references to the literature on the subject are given. None of the country sections attempts to describe the concepts, methods, and statis­ tical sources in detail or to reproduce the statistical results of the estimates themselves. The reader interested in obtaining addi­ tional conceptual, methodological, or sta­ tistical information about the estimates for these sixty-six countries is referred to the bibliographies given in each section, as well as to the following general publications:

Num ber of Countries Chapter 33. D evelopm ents in the U nited States, U nited Kingdom , Australia, N ew Zealand, and Union of South Africa Chapter 34. D evelopm ents in E ight­ een European Countries Chapter 35. D evelopm ents in Latin America and the Caribbeans Chapter 36. D evelopm ents in the Near, Middle, and Far E ast Chapter 37. D evelopm ents in Africa T otal

5 18 20 15 8

1. United Nations, Statistical Office, Depart­ ment of Economic Affairs, National Income Statistics of Various Countries, 1938-1948, New York, 1950. ----- , Statistics of National Income and Ex­ penditure, Statistical Papers, Series H, No. 9, New York, 1956, and No. 10, New York, 1957; Yearbook of National Income Statis­ tics, 1957, New York, 1958. ----- , A System of National Accounts and Supporting Tables, Studies in Methods, Sta­

66

The country sections within each chapter are organized on a more or less uniform plan, within the limits of the available in­ formation, as follows: a brief history of the country’s estimates is given; the individuals or agencies responsible for their preparation or publication are named; the nature of the 453

454

THE INCOME OF NATIONS

tistical Series F , N o . 2, N e w Y ork, 1955. U nited N ations, Statistical Office, Depart­ m ent o f Econom ic Affairs, Methods of Na­ tional Income Estimation, Studies in M eth­ ods, Series F , N o . 8, N e w York, 1955.

----- , Concepts and Definitions of Capital Formation, Statistical Papers, Series F, N o. 3, 1953. 2. Organisation for European E conom ic C o­ operation (O E E C ), Statistics of National

Product and Expenditure, 1938, 1947 to 1952, Paris, 1954. ----- , National Accounts Studies, Denm ark, N orw ay, Switzerland, etc., Paris, 1951-53. 3. International A ssociation for Research on Incom e and W ealth, Bibliography on In­ come and Wealth, Bow es and Bow es, Cam ­ bridge, England, vol. I, 1937-48 (b y D aniel C ream er), 1949; vols. II, III, IV, and V , 1949-52 (by Phyllis D e a n e ), 1950-55 (here­ inafter referred to as “International B ibli­ ography” ) .

In examining the development of national income estimates and analyses in these sixtyfour countries, it is important to keep in mind that differences in the degree of the economic development of these countries and in the type of political, social, and eco­ nomic systems maintained by them affect the nature of their estimates. The countries are consequently classified as industrially and agriculturally advanced and as under­ developed; and as countries operating under the free enterprise system and those operat­ ing under the Soviet or Communist system. Only about one third of these sixty-four countries are industrially and agriculturally advanced, while two thirds can be classified as underdeveloped. The first have fairly welldeveloped statistics which were initiated many years ago, and have correspondingly advanced and well-matured national income estimates. The second are only beginning to develop statistics and are still in their swad­ dling clothes so far as national income esti­ mates are concerned. The first are the lead­ ers in the field; the second, the followers. The task of the second is: first, to learn how to adapt to their still primitive conditions the concepts and techniques of the more ad­ vanced countries; and secondly how to or­ ganize as quickly as possible the collection of the most important statistical data on the social and economic facts of the life of their peoples. While they still have a great deal to learn from those countries with more ex­ perience, some of them are becoming leaders in this by no means easy task; and may some day become leaders in national income esti­ mating generally. The developments in both

types of countries are interesting in their own way and are worth watching. More than five sixths of the countries op­ erate under the free enterprise system, less than one sixth under the Communist system. The first base their national income esti­ mates on the comprehensive production con­ cept of national income, use a wide range of income aggregates, and a three-way breakdown of these aggregates (by indus­ trial divisions, by distributive shares, and by types of final expenditure); and some of them even incorporate comprehensive sys­ tems of double-entry social or sector ac­ counts in their estimates. The smaller group of countries bases its national income esti­ mates on the Marxian restricted material production concept; recognizes generally only one type of aggregate, e.g., the net output at market prices; breaks down the total almost solely by industrial divisions (offering but a most cursory analysis, if any, of the distribution and disposition of the national income); and seldom shows any social or sector accounts. Whatever sharing of ideas and experiences takes place among the national income es­ timators and analysts of different countries is generally limited to countries belonging to the same ideological, political, and eco­ nomic group. Estimators and analysts be­ longing to the free enterprise economies share their ideas and experiences freely among themselves, without regard to the degree of economic advancement or under­ development of their countries. Estimators and analysts belonging to the Communist countries do the same among themselves. But there is practically no intercourse be­ tween these two groups, i.e., across the di­ viding ideological line. The main reason for this segregationism in a field of work that is essentially scientific lies in the restrictions imposed by most of the Communist govern­ ments on the publication of materials bear­ ing on the operations of their economies and also on the intercourse of their nationals with those of the non-Communist world gen­ erally (an attitude reciprocrated in some of the countries of the free world as well). We are endeavoring in this work to give equal attention to developments in the free enterprise countries and in the Communist countries. If the treatment of the first is richer than that of the second, it is because of the greater availability of information in their case, rather than because of any lack of desire to treat the second group as fully.

33. DEVELOPMENTS IN THE UNITED STATES, UNITED KINGDOM, AUSTRALIA, NEW ZEALAND, AND SOUTH AFRICA The national income estimates of the five English-speaking countries to be reviewed in this chapter and those of Canada and Ireland—which were reviewed in detail in Chapters 21 and 22 have a great many things in common. They are similar with respect to their underlying concepts, modes of con­ struction, and the uses to which they are put. This is, of course, not surprising. Their com­ mon language has made it easy for any ad­ vances made in one country to be trans­ mitted quickly to the others. Secondly, while obviously differing in their economic, so­ cial, and political development, they share enough history, custom, and outlook to ex­ hibit greater similarity than countries with­ out such a common background. However, within this group there are also subtle simi­ larities resulting more from geographical closeness than from common heritage. For example, the estimates of the United States and Canada resemble each other more closely than either of them resemble the estimates of the United Kingdom or any other member of the English-speaking group.

above, and require, therefore, some elabora­ tion here. In the 1920’s, national income estimates and studies in the United States began to multiply, and in the 1930’s became even more numerous and fruitful, with the re­ sult that by the end of that decade, in less than twenty years, the United States had advanced from a position of relative ob­ scurity in the field to one of unquestioned world leadership. This was accomplished mostly as a result of the work done by private scholars under the auspices of the Brookings Institution, the National Indus­ trial Conference Board, and above all, of the National Conference on Research in In­ come and Wealth. The greatest single con­ tribution to this development was made by Professor Simon Kuznets (Chapter 10, Sec­ tions 2a and b). This pioneering work by private research agencies and individual economists led to active participation and assumption of lead­ ership by the federal government itself. The first official contribution was made in 1926 by the Federal Trade Commission, which published estimates for 1916: National Wealth and Income. In 1934 the Depart­ ment of Commerce established the National Income Unit to prepare annual estimates, and this unit, under the successive leader­ ship of Robert Nathan, Milton Gilbert, and George Jaszi, gradually extended its opera­ tions, eventually becoming the largest na­ tional income estimating unit in the world. A report published by it in 1934 gave es­ timates for 1929-32 and was followed by a similar study for 1929-35. Since then, estimates by the National Income Unit (later National Income Division), published regularly in the Survey of Current Business, have become the backbone of all national income research in this country. As a sup­ plement, a current series of monthly esti­ mates of income payments to individuals (later called “personal income”), going

1. United. States The early history of national income es­ timating in the United States was fully pre­ sented in Chapters 9 (Sections la and b) and 10 (Sections la, 2a, and 3b through 3f) and requires no further treatment here. It was shown there that the first estimate in the United States was prepared by George Tucker in 1843, the second by Charles B. Spahr in 1896, and the third by W. I. King in 1915; and that up to 1920 these three studies constituted practically the entire na­ tional income literature of the United States. Up to that time her position in the world in this field of statistical endeavor was cer­ tainly not distinguished. But the subsequent activities of her scholars and her govern­ ment in this domain, which have been out­ standing, were only slightly touched upon 455

456

THE INCOME OF NATIONS

back to 1929, was inaugurated in 1938. A year later, another important series covering annual payments by states, also going back to 1929, was started. World War II brought about inquiries into gross national product and its distribution by public and private expenditure. This series, a counterpart of the British “White Paper Analysis,” was started in 1942 and produced annual figures for 1929-38 and for subsequent years by quar­ ters. These estimates were supplemented by elaborate studies on the structure of con­ sumer outlays since 1929. In 1944, the National Income Division of the Department of Commerce began a com­ plete overhauling of its national income es­ timates. Basic concepts were revised, in­ dustrial divisions reclassified, improvements in data and techniques introduced, and es­ timates expanded into a system of manifold national accounts. In the course of this revision, account was taken of the agree­ ments on concepts incorporated in a report made by a League of Nations committee (Chapter 10, Sections 3d and 3e). The re­ vised estimates, superseding those for 192946, were published in the Survey of Current Business, July 1947, National Income Sup­ plement. During these years, other federal depart­ ments and agencies made studies of certain aspects of national income. In 1938 the Na­ tional Resources Committee (later the Na­ tional Resources Planning Board), in col­ laboration with the Bureaus of Labor Sta­ tistics and of Home Economics, the Works Progress Administration, and the Central Statistical Board prepared a comprehensive study of consumer incomes in 1935-36 by size and by type of expenditure, based on large samples of family budgets. During World War II such studies were carried through 1941-43 by the Office of Price Ad­ ministration. Bureau of Labor Statistics data on wages were combined with Treasury data on income and data collected by other agencies. In 1947 the Bureau of the Census based income distribution studies for 1944 and 1945 on census income data and sample investigations. The Bureau of Internal Rev­ enue concerned itself with the amount and distribution of individual and corporate tax­ able incomes. The Bureau of Agricultural Economics inaugurated annual studies of gross and net agricultural incomes and their components, which were later used by the National Income Division as a basis for its estimates of the operational component of

the national income. The Securities and Ex­ change Commission prepared estimates of savings and its components. In 1954, the National Income Division of the Department of Commerce published a comprehensive volume describing the con­ cepts and methods used in its estimates and presenting extensive historical and current statistical series on the various income ag­ gregates and components as well as on social accounts. By that time, the estimates for the United States had become among the most distinguished in the world for the extraor­ dinary richness of their underlying statistics and the consistency and refinement of their underlying techniques as well as for their wide scope and detail. The estimates are currently prepared mostly by the income-distributed method, although some elements, such as farm in­ come, are prepared by the net output method. National income is shown in all its three phases. In the analysis of national ex­ penditures, the final products method of estimation is employed. Every important in­ come aggregate and component is covered in the series; e.g., national income at factor cost, both by distributive shares and by in­ dustrial origin; national income at market prices; gross national product or expendi­ ture by types of final consumption and in­ vestment products; personal income; and disposable income. A system of social ac­ counts shows the transactions of four sec­ tors of the economy—households and insti­ tutions; business; government; and the rest of the world— and also a gross saving and investment account (see Chapter 15, Sec­ tion 2).

Bibliography The principal publications on the national income of the United States in the nine­ teenth century have been listed in Chapter 9, Sections la and b; and those of the first half of the twentieth century in Chapter 10, Sections la, 2a, and 3b through 3f. Here we list only those that have appeared since 1940 and only those of primary importance: 1. Sim on K uznets, National Income and Its Composition, 1919-38 (2 v o ls.), N ational Bureau o f E conom ic Research, N e w York, 1941. ------- , National Product Since 1869, N ational Bureau o f Econom ic Research, N ew York, 1945. ------- , “G overnm ent Product and N ational Product,” Income and Wealth Series 1, In­ ternational A ssociation for Research in In-

33.

THE UNITED STATES AND THE BRITISH COMMONWEALTH

com e and W ealth, Cam bridge, England, 1949. 2. N ational Bureau o f E conom ic Research, Studies in Income and Wealth, vols. 1-22, N ew York, 1937-58. 3. U nited States D epartm ent o f C om m erce, N ational Incom e D ivision, National Income, 1954 edition, Supplem ent to the Survey o f Current Business, W ashington, 1954. 4. U .S. Congress, Joint Econom ic C om m ittee, The National Economic Accounts, October, 1957.

For a more complete listing of the United States publications on national income since 1937, see the International Bibliography.

2. United. Kingdom The United Kingdom was the birthplace of the first theories of national income and the first estimates. The contributions of her scholars over the past three hundred years and of her government during the past fif­ teen years or so have been reviewed in fair detail in Chapters 2, 3, and 7, and in Chap­ ter 10, Sections la, 3a, b, d, and e. A few words need to be added here about the na­ tional income estimates prepared in that country since 1920. Bowley and Stamp continued their col­ laboration in the field into the 1920’s and early 1930’s, publishing, in 1927, a joint study of the British national income in 1924, and special studies in the 1930’s. Bowley also published a number of separate studies, including a survey made in 1920 of the dis­ tribution of national income since 1880; one made in 1937 on wages and national income since 1860; and one made in 1942 on na­ tional income during the period 1924-38. All these estimates were based primarily on income tax data, supplemented by census, production, and social security data, as well as on information obtained by means of spe­ cial questionnaires. The method of estima­ tion was the income-distributed method, ex­ cept for farm income estimated by the net output method. A. W. Flux, continuing his earlier studies (1907 and 1911) of national expenditure, published a new estimate in 1929 relating to the year 1924. Where he relied largely on the 1907 production census data in his earlier estimates, in his later estimates he used a variety of data, leaning heavily on a simplified commodity flow method of esti­ mating (see Chapter 20, Section 5e). In some cases, he used the production values of goods, in others the net values at factor

457

cost, and in still others the export or import values. In each case, he made such additions or deductions as seemed to him necessary to avoid double counting, or to obtain fig­ ures expressed in market prices, or to dis­ tinguish as between the domestic supply of goods and the transactions with the rest of the world. Colin Clark produced his first compre­ hensive study of the British national income in 1932. It covered the years 1924-31. Like Bowley and Stamp, he used the incomedistributed method for most of the com­ ponents except agricultural income, which he estimated by the net output method. He relied largely on income tax, wage, employ­ ment, and production data as well as on import and export statistics and dealt with the national income at factor cost. In 1937, he published his second, more important, book in which he estimated the national in­ come at market prices up to and including the year 1935. In these estimates he in­ cluded consumers’ outlay, savings, invest­ ment, and income transfers through taxa­ tion, in addition to national income proper. Finally, after the outbreak of World War II, the newly created Central Statistical Office undertook the annual preparation of estimates of national income and expendi­ tures within the newly conceived framework of national accounts. These official statistics appeared for the first time in “An Analysis of the Sources of War Finance and an Esti­ mate of the National Income and Expendi­ ture in 1938 and 1940,” and were succes­ sively elaborated in later annual White Papers on “National Income and Expendi­ ture” which became known, because of the color of their covers, as the “Blue Book.” The social accounting approach of these estimates has since become a standard not only in the United Kingdom but in many other countries as well. Gross national prod­ uct is shown for a period of years at current as well as constant market prices and at factor cost, and is analyzed by industrial origin, by factor shares (including deprecia­ tion), and by types of expenditure. The ex­ penditures, in turn, are analyzed by the four standard classifications as consumer expend­ itures (by type of product), government current expenditures on goods and services, gross domestic private and government capi­ tal formation, and the overseas surplus or deficit (which is equal to exports and in­ come received from abroad less imports and income paid out). In addition, input-output

458

THE INCOME OF NATIONS

series are presented on an interindustry basis. The official estimates are prepared mostly by the income-distributed method on the basis of tax and census data. The expendi­ ture estimates are based on production sta­ tistics and a variety of other data. The 1954 and other recent estimates no longer showed net national income at factor cost, giving instead the gross national prod­ uct. The latter is shown both at market prices and at a modified factor cost in which profits are given before deducting deprecia­ tion, i.e., gross of depreciation. The attempt to deduct depreciation was abandoned on the ground that the only information avail­ able on depreciation was that represented by the statutory allowances for income tax assessment, which did not cover all capital assets and hence was not sufficiently reliable. The use of the gross rather than the net figure in the historical series was justified in the 1954 White Paper on the ground that “there is good reason to believe that the ‘true’ national income, though smaller than the gross national product, will generally follow fairly close its year to year move­ ments.” At the same time, however, the total figure for depreciation was given, as arrived at on the basis of the data cited. In the social accounts, the following sec­ tors of the economy are distinguished: per­ sons primarily as final consumers; private corporations; public corporations (which control a substantial segment of the British economy since the nationalizations of 194551); central government authorities; local government authorities; and the rest of the world. A distinction is also made for each sector between its production account, its consumption or income and expenditure ac­ count, and its capital or investment account. In 1956, the Central Statistical Office pub­ lished its first comprehensive explanation of the sources used in the preparation of its estimates. The explanations followed closely the tables in the annual “Blue Book.”

Bibliography The most important publications on the national income of the United Kingdom that appeared in that country during the past three hundred years or so have been listed in Chapters 2, 3, and 7; and in Sections la and 3a, b, c, and d of Chapter 10. The more recent ones among them published since the late 1920’s may be recapitulated as follows:

Wages and Income in the United Kingdom since 1860, Cambridge,

1. A . L. B ow ley, 1937.

----- , Studies in the National Income, 19241938, N ational Institute o f Econom ic and Social Research, Econom ic and Social Stud­ ies I, Cambridge, 1948. 2. A . L. B ow ley and Josiah Stamp, The Na­ tional Income, 1924, O xford, 1927.

----- , Three Studies in the National Income, Series o f Reprints o f Scarce W orks, N o . 6, London School o f E conom ics and Political Science, London, 1938. 3. A . W. Flux, “The N ational Incom e,” Jour­ nal of the Royal Statistical Society, 1929, pp. 1-2 5 . 4. a. C olin Clark, The National Income, 19241931, London, 1932. b. , National Income and Outlay, L on­ don, 1937.

c. ——•, Conditions of Economic Progress, d. e.

London, 1940. , “The N ational Incom e in 1932,” Economic Journal, June, 1933, pp. 2 0 5 16. , “N ational Incom e at Its C lim ax,” Economic Journal, June, 1937, pp. 3 0 8 -

20 .

/. ------- , “Further D ata on the N ational Incom e,” Economic Journal, September, 1934. g. , “Investm ent in F ixed Capital in G reat Britain,” London and Cambridge Economic Series, Special M em orandum 38, 1934. h. , “Determ ination o f the M ultiplier from N ational Incom e,” Economic Jour­ nal, September, 1938, pp. 4 3 5 -4 8 . 5. a. British W hite Paper, An Analysis of the

Sources of War Finance and an Estimate of the National Income and Expenditure in 1938 and 1940, H.M.S.O., London, 6.

7.

a.

1941. Central Statistical Office, National In­ come and Expenditure (annual volu m e), known as the “Blue B ook,” London, A u ­ gust, 1952; August, 1953; August, 1954; September, 1955.

b. ----- , National Income Statistics: Sources and Methods, H.M.S.O., London, 1956. a. Richard Stone and D . A . R ow e, The Measurement of Consumer Expenditure and Behaviour in the United Kingdom, 1920-1938 (2 v o ls.), Cambridge, 1954. b. A . R . Prest, Consumer Expenditure in the United Kingdom, 1900-19, Cam ­ bridge, 1954.

3. Australia The preparation of national income esti­ mates in Australia began in the 1880’s with the publication by the government statisti­ cian, Timothy Coghlan, of an estimate of

33.

THE UNITED STATES AND THE BRITISH COMMONWEALTH

the income of New South Wales. He later broadened these estimates to cover the whole of Australia (Chapter 9, Section 4, and Chapter 10, Section Id ). With Coghlan’s retirement in 1904 the preparation of such estimates ended. In 1914-15 it was tempo­ rarily revived when the government included questions on individual incomes in its War Census and published the results. In the 1920’s and 1930’s, several private scholars tried their hands at preparing simi­ lar estimates. Thus, in 1926, J. T. Sutcliffe prepared estimates by the net output method covering the fiscal years 1911-25 and one by the income-distributed method for the fiscal years 1917-22; and Frederick Benham prepared one in 1928 by the income-dis­ tributed method for the fiscal year 1925. J. G. Crawford published some rough cal­ culations in 1936 for the fiscal years 1929 to 1934. Finally, in 1938, Colin Clark and J. G. Crawford jointly published elaborate estimates for the fiscal years 1929 to 1938, linking them to Coghlan’s figures for 19011903, the official estimates for the fiscal year 1905, Sutcliffe’s figures for fiscal years 1917-22, and Benham’s figures for the fiscal year 1925 by interpolating the latter to the fiscal years 1922 and 1928. In 1945, the Commonwealth Census and Statistics Bureau definitely took on the func­ tion of estimating national income, thus re­ establishing the practice introduced by Coghlan half a century before. The Bureau issued its estimate as a White Paper accom­ panying the government’s budget; and has been issuing it annually in that form ever since. The Bureau’s annual estimates cover the national income, national expenditure, and the social accounts of the Australian econ­ omy in both current and constant money values in a continuous series starting with the year 1938-39. National income is esti­ mated mainly by the income-distributed method, and national expenditure by the final products method. The estimates are based on a wide variety of tax, production, sales, and foreign trade statistics. The principal table gives the historical series of the following aggregates: national income at factor cost, personal income of residents, gross national product, gross pri­ vate investment in fixed capital equipment, governmental expenditure on public works, export and other receipts for goods and services furnished to foreigners, imports and

459

other payments for goods and services re­ ceived from foreigners. The social account of national income and expenditure shows on the left-hand side the income payments and other charges making up the net national income (as the sum of wages, company income, surplus of public authority business undertakings, in­ come of unincorporated farm and nonfarm businesses, net rent of dwellings, and other net rent and interest), the gross na­ tional product (as the sum of the net na­ tional income, indirect taxes less subsidies, and allowances for depreciation and mainte­ nance) and total market supplies (as the sum of the gross national product and the import and other payments for goods and services). On its right, or credit, side, the account shows the net expenditure on goods and services as consisting of the gross do­ mestic expenditure (as the sum of the ex­ penditures of public authorities and of finan­ cial enterprises; gross private investment in fixed capital equipment, in nonfarm stocks, and in farm stocks; and personal consump­ tion) and of the total market expenditure (as the sum of the gross domestic expenditure and of the export and other receipts for goods and services). The two sides of the account are shown to be in balance. There are five sector accounts of standard nature covering trading enterprises; financial enterprises; public authorities; individuals as income recipients and consumers; and trans­ actions (of these four sectors) with the rest of the world. In addition, there is a capital account for the economy as a whole, show­ ing the capital funds available on one side and the investments of these funds on the other. The capital funds available are shown as composed of: net increase in indebtedness to rest of world (of public authorities, pri­ vate borrowing from International Monetary Fund, and decrease in international re­ serves) ; public authority capital transfers (to persons and business undertakings, includ­ ing war gratuity, war damage insurance claims and payments to farmers for drought relief); allowances for depreciation; increase in assurance funds; increase in funds of marketing authorities (Australian Wheat Board, Australian Wool Realization Com­ mission, etc.); undistributed profits; and other personal savings. The investments in the Capital Account are shown under the following headings: (a) gross private investment in private

460

THE INCOME OF NATIONS

dwelling construction; other new building and dwelling maintenance; trucks, utilities, etc.; motor cars and cycles; other fixed capi­ tal equipment; and change in the book value of nonfarm stocks, and (6) net increase in indebtedness of public authorities (common­ wealth bonds, including those of states, after deducting redemptions out of loan pro­ ceeds) .

Bibliography References to the works published before 1945 are given in Chapter 9, Section 4, and Chapter 10, Section 1d, above. The principal later works are: 1. Commonwealth of Australia, Bureau of Census and Statistics, National Income and Expenditure, annual publication since 1945; particularly the one for 1954-55, published in 1955. 2. H. P. Brown, The Composition of Personal Income, Reprint from The Economic Rec­ ord, Journal of the Economic Society of Australia and New Zealand, Melbourne, June, 1949. See also the International Bibliography.

4. New Zealand. The earliest estimates of New Zealand’s income were contained in Coghlan’s afore­ mentioned Australian estimates (Chapter 10, Section Id). Official estimating in New Zealand began in 1930, when the Census and Statistics Department of that country published its first income estimate (Chapter 10, Section 2c). Parallel with them, several private estimates were published in the 1930’s and 1940’s. A. G. B. Fisher published an independent estimate in 1930 emphasiz­ ing the distribution of income. F. B. Stephens published one in 1936 which related to homeproduced income in the fiscal years 1926 to 1931 and was based on data of the 1926 Census of Income as well as on income tax, employment, and wage statistics. E. P. Neale published an estimate in 1941 dealing with personal income. At first, the official estimates dealt with the “aggregate private income” of the coun­ try rather than with its national income proper. Only in 1945 did they begin to deal with a wider range of aggregates. These newer aggregates included: national income at factor cost and at market prices, private income, gross national product and national expenditure, and sector accounts. The re­ vised series gave figures from the years 1938-39 on, and showed the distribution of

private income among factor groups and among individuals by size of income. The estimates are prepared on the basis of in­ come tax and census statistics as well as other data.

Bibliography 1. New Zealand, Census and Statistics Depart­ ment, New Zealand Official Year Book, 1932 and later years. 2. F. B. Stephens, “National Income of New Zealand,” The Economic Record, Journal of the Economic Society of Australia and New Zealand, December, 1936. 3. Allan G. B. Fisher, “Distribution of Income in New Zealand,” The Economic Record, op. cit., November, 1930. 4. E. P. Neale, “Recent New Zealand Data Regarding the Incomes of Individuals,” The Economic Record, op. cit., June, 1941. 5. a. New Zealand, Census and Statistics De­ partment, New Zealand Official Estimates of National Income 1938-39 to 1947—48, Supplement to June, 1948 issue of the Monthly Abstract of Statistics, Welling­ ton, 1948. b. , For the same publications for sub­ sequent years, see the supplements to the June-July issues of the same Abstract. c. , Report on the Official Estimates of National Income and Sector Accounts, 1938-39 to 1953-54, Wellington, 1954.

5. Union of South Africa The pioneer in national income estimates in the Union of South Africa was R. A. Lehfeldt who, in 1925, published estimates for 1917 and 1923. In these estimates he distinguished between the incomes of the European population and those of the nonEuropeans, estimating the former to be, on a per capita basis, 13 times as high as the latter. C. G. W. Schuman followed in the late 1930’s with an estimate covering certain years in the period 1910-35. He carried the distinction between the standards of living of the two categories of population beyond Lehfeldt’s point. To facilitate comparisons of the South African per capita national in­ come with the per capita national incomes of other countries, he substituted for the real population of the country an “equiva­ lent population” in which the non-Europeans were counted at an eighth of their real number, on the assumption that their per capita income was one eighth of that of the Europeans. In 1940, S. Herbert Frankel, alone, as well as in association with S. D. Neumark

34.

DEVELOPMENTS IN EIGHTEEN EUROPEAN COUNTRIES

and H. Herzfeld, published estimates for se­ lected fiscal years 1928-39; extending them later, on a revised basis, into a series cover­ ing the fiscal years 1912 and 1918-44. In 1943, he published estimates for the prov­ inces of the Union for fiscal years 1936-39; and in 1945, estimates of consumption and capital investment for fiscal years 1922-41. Frankel’s estimates related to home-produced income and were made by the net output method. No distinction was made in them between the net output of the Europeans and non-Europeans. Such a distinction was declared by him to be “both statistically and theoretically invalid” inasmuch as “national income is the result of the effort of all mem­ bers of the economy” and, therefore, “can­ not be defined or directed on the basis of class, color, or race.” The Social Security Committee later pub­ lished estimates of the per capita disposable incomes of the several ethnic groups in 1941-42, placing them, at 1938 prices, at £ 1 2 5 for the Europeans, £.25 for the col­ ored and the Asiatics, and at £ 1 0 for the natives. After World War II the Central Statis­ tical Office of the Union began publishing annually its own estimates of national in­ come, likewise using the net output method and showing the origin of the national in­ come by industrial divisions. The concepts of national income used in these estimates followed closely the international standards. Steps were being taken in the 1950’s to de­

461

velop statistics necessary for the eventual construction of a complete set of social ac­ counts.

Bibliography 1. R. A . Lehfeldt, National Resources of South Africa, London, 1922; Report of the Eco­ nomic and Wage Commission, Pretoria, 1925. 2. C. G . W. Schum ann, Structural Changes and

Business Cycles in South Africa, 1806-1936, London, 1938. 3. S. Herbert Frankel, alone or with S. D . N eumark or with H . H erzfeld, series o f studies o f the “N ational Incom e o f the U n ion of South A frica,” “C onsumption, Investment, and War Expenditure in Relation to the South A frican N ational Incom e,” “G rowth o f N ational Incom e o f the U nion,” etc., in the South African Journal of Economics, 1 9 40-41, June, 1944, and subsequent issues; U n ion o f South A frica, Provincial Financial Resource Committee, Report, U .G . no. 9, Pretoria, 1944 (paper by S. H . Frankel on the “Estimated Provincial A llocation o f the N ational Incom e o f South A frica” ). 4. D . S. Franzsen, “Som e M ethodological Prob­ lem s Raised by the Calculation o f the U n ­ ion’s N ational Incom e, by Incom e T ype,” South African Journal of Economics, June, 1948. 5. U n ion o f South A frica. Office o f Census and Statistics, National Accounts and National Income, series issued annually and m ore fre­ quently since 1948, as Memorandum 1-9 , etc.; also published in the Office’s Monthly Bulletin of Statistics and published annually in the South African Journal of Economics.

34. DEVELOPMENTS IN EIGHTEEN EUROPEAN COUNTRIES

The national income estimates of seven Eu­ ropean countries, Ireland, Germany, France, the Netherlands, Sweden, Soviet Russia, and Yugoslavia were reviewed in detail in Part III, and in the United Kingdom esti­ mates were discussed in a more limited fashion in Chapter 33. In this chapter we shall survey, likewise in a less detailed way, the recent developments in national income estimation and analysis in some eighteen re­ Notes begin on p. 544.

maining European countries. The only Eu­ ropean country not reviewed here is the little principality of Monaco; and the reason for omitting it is that information on its ac­ tivities in question is lacking. The countries will be taken up in the following order: 1. Italy 2. Switzerland 3. Belgium

4. Luxembourg 5. Denmark 6. Norway

462

THE INCOME OF NATIONS

7. 8. 9. 10. 11. 12.

Finland Iceland Austria Hungary Czechoslovakia Poland

13. 14. 15. 16. 17. 18.

Bulgaria Rumania Albania Spain Portugal Greece

I. General Review 1. Developments in the Free Enter­ prise Economies Among these eighteen countries, twelve operate under the free enterprise system. Their national income estimates in most cases are at least half a century old and are well refined and highly informative. They were perfected particularly during the decade following World War II and have come to be used extensively in the formu­ lation of governmental fiscal and economic policies and, in countries engaged in long­ term national economic planning, in the formulation and execution of such long-term plans. A distinct trend is apparent in this group of estimates toward using the income-dis­ tributed method and even the final expendi­ ture method more extensively than before and to rely more heavily on income tax, social security tax, business accounting, and sales data, in addition to production data. A wider range of breakdowns of the income aggregates is furnished. More and more esti­ mates show the national income in all its three phases—by industrial origin, by dis­ tributive shares, and by types of final ex­ penditure. The number showing sector ac­ counts is also increasing. The standards of construction and presentation of estimates recommended by international organizations are being more closely followed. Among these twelve countries, Austria has had the longest history in national in­ come estimation, but its record has been uneven. A leader in the field in the days of her political, economic, and cultural promi­ nence in the world in the nineteenth and beginning of the twentieth centuries, she has come to play a secondary role in this sphere during the present days of constric­ tion of her political and economic life. In Italy, where for several decades statis­ tics were ill nourished and the national in­ come estimates prepared by private scholars, though more numerous than in any other country, were unreliable, effective steps were taken during the decade 1946-55 to correct

this deficiency. The Central Statistical Insti­ tute placed its statistical collations on a firmer foundation and took over the prepa­ ration of national income estimates as one of its major functions. Its new estimates began to command international respect. In Switzerland, where national income statistics were organized on an official basis even before World War II and were rela­ tively fair, further improvements in them were introduced during the postwar period in line with some of the newer international standards. In Norway, Denmark, Finland, and Bel­ gium, national income estimates were made official during the postwar decade and were raised from positions of relative obscurity to those of considerable prominence. In the first three countries at least they came to be used as tools in the government budgetary and economic development and investment plans. In Norway and Denmark particularly, owing to a successful collabo­ ration of scholars and government officials, they were advanced theoretically and meth­ odologically beyond the points attained in many other European countries. In Iceland, Spain, Portugal, Greece, and the little Grand Duchy of Luxembourg, in which national income estimates were only slightly developed before World War II, earnest efforts were now being made by the governments to establish them on a firmer foundation.

2. Developments in the Six Com­ munist Countries The situation is very different in the six countries that have adopted the Communist system since the end of World War II, namely in Hungary, Czechoslovakia, Po­ land, Bulgaria, Romania, and Albania (the seventh, Yugoslavia, was treated in Chapter 28 above). It is very difficult to ascertain correctly the status of national income esti­ mates and analysis in these countries. First of all, with the possible exception of Poland and, for a time, Czechoslovakia, very little information has been made available about their national income estimates. The pub­ lished figures in most cases lack detail and no explanation of their derivation is given. Secondly, because of the different nature of their economies, the criteria for appraising estimates developed in the capitalist coun­ tries cannot be fully applied to their esti­ mates; and on the other hand criteria that would be fully applicable to them have not

3 4 . DEVELOPMENTS IN EIGHTEEN EUROPEAN COUNTRIES

yet been developed either by these countries themselves or by any outsider. In the four countries first mentioned, Hungary, Czechoslovakia, Poland, and Bul­ garia, the currently prepared national in­ come estimates are largely mere extensions of those prepared under the former political and economic systems. On the other hand, in Rumania and Albania the estimates rep­ resent a wholly new statistical activity, in­ asmuch as no systematic estimates had ever been prepared there before. In all six coun­ tries, national income estimates are pre­ pared by government and are closely linked to their long-range economic plans. Na­ tional income includes only the net value of material production; it is estimated by the net output method, and is broken down by industrial origin. The distribution of the national income is seldom shown, and the disposition or use of the national income by type of expenditure is not developed in de­ tail. The estimates seem to rest on a more solid statistical foundation than the previous estimates did. This is a supposition based on the fact that the collection of economic statistics in these countries has been tre­ mendously expanded since their adoption of the Communist system because of the vast need for them where production is statecontrolled. The information on national income esti­ mates published in Communist countries (Yugoslavia excepted) indicates that these estimates play a much less important role in their economic planning than do other types of statistical collations, and perhaps even a less important role than they do in free enterprise economies. This is not as strange as it may seem at first glance. Gov­ ernments in the Communist countries have direct administrative and financial controls over the operation of state enterprises and other collectives. They do not need, there­ fore, to be as much concerned over the for­ mation and distribution of the national in­ come as do governments and private groups in capitalist economies in which production, consumption, and capital investment are controlled indirectly through the formation, distribution, and expenditure of incomes by private individuals, groups, and governments in a free market, or through free political decisions of the general citizenry. Nor are individuals or private groups as free in Communist countries to inquire into the for­ mation and distribution of national income as they are in capitalist countries. At the same time, those government officials in

463

Communist countries who direct national production or any part of it are not as much concerned with the formation and distribu­ tion of the national income as they are with the physical outputs of the various branches of production and the gross values of these outputs, for it is through the accounts of these outputs that they control the operation of the state enterprises and other collectives and formulate and execute national produc­ tion plans. As E. F. Jackson aptly puts it in a recent paper, “it is unnecessary for a planner [in a Communist state] to have a duplication-free estimate of the national in­ come as long as he knows how many and which enterprises are fulfilling or falling short of their gross output plans.” 1 The planners in the Communist states generally set the goals individually for each branch of material production, specifying in each case: (a) the quantities of goods to be produced by the set date as compared with the currently produced quantities; (b) the number of workers to be employed as compared with the ones currently employed; (c) the percentage increase in productivity per worker to be attained; (d) the per­ centage reduction in the costs of production to be achieved; (e) the gross value of the output at constant prices to be realized as compared with the gross value of the cur­ rent output; and (/) the rise to be achieved in the levels of consumption of the popula­ tion in terms of quantities of meats and other goods supplied per capita as well as in educational, health, and cultural services. The economic planners use the national in­ come aggregate, i.e., the net value of the national output, largely to measure only three things: (a) the planned or expected increase in the total national production; (b) the planned or executed change in the relative importance of the different indus­ trial divisions in the total national produc­ tion; and (c) the planned or executed share of the total national product devoted to capital accumulation. In so using the na­ tional income aggregates, the planners do not generally bother about relating them to other aspects of their economic plan that are expressed in terms of gross outputs. At least, they do not generally publish this relationship. The situation can be illustrated by look­ ing at Hungary. In that country’s five-year economic plan promulgated in 1949, the goals for the output of the manufacturing industry to be reached by December 31, 1954 were set at an increase of 86.4 per

464

THE INCOME OF NATIONS

cent in gross value of the output at con­ stant prices, the increase in productivity at 50 per cent, and the reduction in produc­ tion costs also at 50 per cent. The goals for agriculture and other branches of national production were similarly expressed at vary­ ing percentage increases or reductions under these and other titles. The over-all objec­ tives of the plan were expressed in terms of a rise in the levels of consumption of the population of 35 per cent, an increase in the national income of 63 per cent, and in the share of manufacturing industry in the national income from 49.7 per cent to 58.7 per cent. In 1951, in reporting on the exe­ cution of the first year of operation of the plan, the government raised the goals of the plan to an increase in the gross value of the output in the manufacturing industry to 210 per cent and to a rise in productivity of 92 per cent, but it lowered the goal of reduc­ tion in costs of production to 25 per cent.2 The government reported for 1950 a 20 per cent increase in the national income over 1949, or about one third as much as originally planned; and a rise in the share of the socialized sector in the total national income from 53.8 per cent in 1949 to 70 per cent in 1950. The references to the na­ tional income were disposed of in a single sentence. The published information of these coun­ tries on their national incomes may not give, however, an altogether accurate picture of the true importance of national income esti­ mates in their national accounting and eco­ nomic plans. It is possible that, in unpub­ lished form, as used by the economic plan­ ners confidentially, these estimates are much more elaborate, informative, and better re­ lated to the gross values of the outputs re­ corded in the economic plans and to the flows of incomes and expenditures in the economy; and generally play a much more important role in the actual direction of national production, consumption, and in­ vestment. It is to be hoped, therefore, that the veil of mystery now covering the prepa­ ration and utilization of national income estimates in these Communist countries (as well as in Soviet Russia herself, as already stated in Chapter 25 above) will be lifted, and that the true measure of both the ac­ complishments and the failures of national income estimation in those countries will be revealed. It may well be that E. F. Jack­ son is right when he says that “there can be no doubt that technical progress in the construction of national accounts [in East­

ern European Communist states] has been very fast since the war and little doubt that the Eastern European economists would, if they publish their results, have a lot to teach their western colleagues.” 3 Cer­ tainly the statistical material on file in the statistical offices of those countries must be rich in content and able to provide illumi­ nating information on every aspect of the national income and expenditure of those countries.

II. Reviews of Developments in Individual Countries Having examined the general trends of development in the two groups of the Euro­ pean countries in question, we can now proceed with a more specific examination of the developments in each of those countries in the order indicated above.

1. Italy a. Prewar Estimates In Italy, despite a vigorous development of the science of political economy during the preceding two centuries (see Chapter 1, Section 5), statistical explorations of the size, structure, and growth of national in­ come did not begin until the present cen­ tury, some fifty years after the unification of the country. They had been preceded in the 1880’s and 1890’s by estimates of the coun­ try’s national wealth made by Pantaleoni and other economists. The first national in­ come estimate was made in 1911 by M. San­ toro, who attempted to measure the growth of the country’s national income since 1860. His estimates were based on exceedingly inadequate and questionable statistics, and were, therefore, exceedingly rough. After World War I, interest in the explo­ ration of the size, structure, growth, and distribution of the national income became more active in Italy, spurred on largely by the country’s enhanced role in world af­ fairs, the development of extreme political and economic nationalism among her people, the growth of the tax burden and of the national debt, and the outbreak of extreme unrest among the laboring and peasant classes. In this exploration, Professor Corrado Gini led the way. Among others who made major contributions during the period 1918-39

34.

DEVELOPMENTS IN EIGHTEEN EUROPEAN COUNTRIES

were Tivaroni, Livi, Porri, Sacerdote, Boldrini, Mortara, D ’Addario, Orlandi, Meliado, Barberi, de Vita, Degli Espinosa, and Lasorsa. Each of them prepared a national in­ come estimate of his own. Some, such as Tivaroni, developed long historical series beginning with 1861 or a later year. Others, such as Gini and Livi, started theirs from the immediate prewar period. Still others dealt Strictly with the postwar period or only with a single current year. In no coun­ try in the world did so many national in­ come estimates appear during any period of comparable length as in Italy during that time. Most of the estimates concerned themselves with the measurement of the changes in the size and industrial structure of the national income over time, but a few, such as Gini’s, D’Addario’s, Orlandi’s, and de Vita’s analyzed its distribution as well. Some of the estimates dealt only with the aggregate private income; others with the national income as such. Some, such as de Vita’s, were prepared mainly by the net output method, while others, such as Degli Espinosa’s, mainly by the income-distributed method. Barberi pioneered in estimating national income by the final expenditure method. Several were calculated in both cur­ rent and constant prices. A number, e.g., Gini’s, Meliado’s, and de Vita’s, gave the distribution of national income by divisions of economic activity, and a few, such as Degli Espinosa’s, gave it by factor shares. Barberi gave it by types of consumption. Many of the estimates were well advanced conceptually and some were developed by very ingenious techniques, but all were based on inadequate statistics and hence were of questionable reliability. Most comprehensive among the estimates of the period were those of de Vita, which covered the years 1928, 1936, and 1937; of Degli Espinosa, which related to the years 1936 and 1937; and of Barberi for the years 1928, 1936, and 1938. Each of these was prepared by a different method. Barberi contended that all estimates of the period prepared by the net output or income-distributed method were too low as they omitted certain impor­ tant elements of the income. During World War II, the preparation of national income estimates continued. Vinci, Lenti, de Vita, di Revel, and Cosciani pub­ lished estimates, but they mostly referred to the prewar national income. The Ministry of Finance and the Central Statistical Office prepared some estimates for the current war

465

years (which were later reviewed by Lenti), but none of these estimates was published at the time.

b. Postwar Estimates During the immediate postwar period, Ragazzi published an estimate of the income and consumption of the population for 1944 and 1945, with comparisons for 1938; and Coppola D’Anna, after reviewing the avail­ able estimates for 1945, offered his own esti­ mate for that year. Gini published several essays on national income, criticizing all existing estimates, whether made for Italy or for other countries, on the ground that they do not deduct the costs of maintenance of human labor; but he offered no esti­ mates of his own on the basis of his more restricted concept of national income. In 1947, the official Central Institute of Statistics, under the leadership of Barberi, assumed the function of the annual prepa­ ration and publication of national income estimates. As a first step to that end, the Institute began to organize the systematic and current collection of pertinent economic statistics. In 1950, the Institute published its first major report on the subject, pre­ senting a new series beginning with the year 1947, and providing a comparison with the figures for 1938. It has maintained the an­ nual publication of this series ever since. Thus, the era of reliance for national in­ come estimates primarily or wholly upon the efforts of private scholars came to an end in Italy, just as it had elsewhere. The official series used the concepts of national income at factor cost and at market prices (less the value of government serv­ ices to business) and also the concept of the gross national product or expenditure. It presented breakdowns of these aggregates by industrial origin, by distributive shares, and by types of final product or expendi­ ture. The figures were expressed in both current and constant (1954) prices. By 1955, the Institute completed a historical series of national income at both current and constant prices for the period 18621954 and laid the foundations for the de­ velopment of social accounts.

Bibliography A . Period 1910-31 Michele Santoro, Italy and Her Economic Progress from 1860 to 1910 (in Italian), Rome, 1911; Loria, Nuova Antologia, June 1918; Corrado Gini, “On the Financial Problem of the Post-War” (in Italian), Supplemento Economico

466

THE INCOME OF NATIONS

del Tiempo,

January 3, 1919; “What Is the Size o f Italy’s N ational Incom e” (in Italian ), Tempo Economico, January 5, 1921; A Comparison of

the Wealth and National Income of Several Important Nations Before and After the War (in E n glish ), Provveditorato G enerale dello Stato, R om e, 1925; Index of Concentration and of Dependence (in Italian ), Biblioteca dell’ Econom ista, Serie 6, vol. 20; “The Determ ina­ tion o f the W ealth and Incom e o f N ations A fter the War and Their Com parisons with Pre-War” (in Italian ), Institut International de Statistique, Bulletin, Tom e X X V , 3me liv., 1931, p. 365; and with R. D ’Addario, Distribution of Personal Income in Italy (in Italian ), 1931; Livio Livi, “Inquiry into the Dynam ics o f WarTim e and Post-W ar N ational Incom e” (in Ital­ ia n ), Metron, Vol. I ll, N os. 3 -4 , 1924, p. 556; Basso, “Total Incom e o f Italians” (in Italian), Critica Sociale, D ecem ber 1923; M. Boldrini, “Private Incom e o f Italians Since 1928” (in Italian ), Commercio, August 1929, p. 37; J. Tivaroni, “Tax Contributions and Private Incom es in Italy from 1861 to Present T im es” (in Italian ), Metron, 1926, V ol. IV , N o . 2; Porri, “C om position o f Private Incom e in Italy” (in Italian), Giornale Degli Economisti, 1925; Sacerdote, “W ealth and Private Incom e in Italy” (in Italian ), Pinerolo, 1928; G eorgio Mortara, “N ational W ealth and Incom e and the T ax Burden” (in Italian), Movimento

Economico Italiano, Banca Commerciale Italiana, M ilano, 1929, pp. 387—402; R. D ’Addario, “Inquiry into the D ynam ics o f Incom e Distri­ bution” (in Italian ), Rivista Italiana di Statistica, July-Septem ber, 1930, p. 269.

B. Period 1932-45 Leonardo M eliado, “Private Incom e o f Ital­ ians in 1928” (in Italian), Metron, March, 1932, pp. 25 1 -3 2 1 ; Silvio Orlandi, “Distribu­ tion o f Personal Incom e in Italy in 1929” (in Italian ), La Vita Economica Italiana, 1934, N o . 1; “Distribution o f Personal Incom e A m ong Provinces and Principal Com m unities in 1929” (in Italian ), La Vita Economica, 1935, N o . 4; A gostino de Vita, “N ational Incom e o f Italy of 1928 and Its Regional Distribution” (in Italian ), La Vita Economica Italiana, 1935, N o . 2; “N a ­ tional Incom e o f Italy” (in Italian), La Vita Economica Italiana, 1939, N o . 2, p. 43; “In­ terrelation o f N ational Incom e and Public F i­ nance” (in Italian ), Atti della 2a e 3a Riunione Scient. della Societa Ital. di Statistica, 1940 and 1941; Benedette Barberi, “Com parison o f N ational Incom e o f Italy o f 1936 with 1928” (in Italian ), Economia Italiana, A pril 1938; “M oney Value o f F ood Consumption as R e­ lated to the Incom e o f the Italian P eop le” (in Italian ), Economica, anno X X I, vol. X IX , nos. 5 -6 , M ay-June, 1943; G. Lasorsa, “W ealth and Private Incom e o f Italy and Their Distribution” (in Italian ), L’Economia Italiana, M ay—June, 1937; A gostino D egli Espinosa, “Incom e and W ealth o f Italians in 193 6 -1 9 3 7 ” (in Italian ), Economia, C ollana di M onografie VI, 1939;

C. C osciani, “Recent Inquiries into Italian N a ­ tional In com e” (in Italian ), Economia, F eb ­ ruary, 1940; P. Thaon di R evel, “Indices o f N a ­ tional Incom e from 1928 to 1938” (in Italian),

Atti della 2a Riunione della Societa’ Italiana di Statistica, R om e, 1940; F elice V inci, “Incom e o f Our Country in 1938” (in Italian ), Rivista Italiana di Scienze Economiche, 1943, pp. 1742, and 1931-50; “Socialism o f T oday” (in Italian ), Statistica, January-June, 1944; “F ea­ tures o f Financing a Plan o f Econom ic R e­ construction” (in Italian ), Rivista Bancaria, January-June, 1945; Bruno Rossi Ragazzi, “In­ com e and Consumption o f Italy in 1 9 44-1945” (in Italian ), Congiuntura Economia, April, 1946.

C. Period 1946-55 1. F. C oppola D ’Anna, “N ational Incom e o f 1945” (in Italian ), Congiuntura Economica, N ovem ber, 1946; Population, Income, and

Public Finance of Italy from 1860 to the Present Day (in Italian ), Rom e, 1946. 2. Corrado G ini, “Incom e o f Italy” (in Italian), ’II Globo, June and A ugust, 1946. 3. A ntonnino Giannone, “Public Expenditure in the N ational Incom e o f Italy for 1938 and 1947” (in E nglish ), International Asso­

ciation for Research in Income and Wealth, Conference, Cambridge, 1949. 4. A lessandro M olinari, “N ational Incom e and R egional Incom e” (in F ren ch ), InstitutIntern. de Statistique, 1949 (also in the Bul­ letin de L’Institut, etc., T om e X X X II, 1950); “Incom es from Labor and from Capital in Italy and in Other Countries” (in Italian), Rassegna de Statistica del Lavoro, January and February, 1950; “H istory o f Italian R e­ search in Incom e and W ealth” (in E nglish ),

International Association for Research in Income and Wealth, C onference, September, 1953. 5. Instituto Centrale di Statistica, Studies of National Income (in Italian ), A nnali di Sta­ tistica, 1950, Series V III, vol. I l l (the w hole volum e 447 p p .), R om e, 1950; National Accounts, Lectures by B. Barberi (in Eng­ lish ), R om e, 1954. 6. B. Barberi, “The Econom ic G rowth o f Italy, 1 8 6 2 -1 9 5 4 ” (in E nglish ), International Asso­

ciation for Research in Income and Wealth, Paper presented at the 4th C onference; “The Italian N ational In com e” (in E nglish ), Re­ view of Economic Conditions in Italy, Banco di Rom a, July, 1954. 7. U n ited States, M utual Security A gency, Spe­ cial M ission to Italy, The Structure and Growth of the Italian Economy (including input-output an alysis), R om e, 1953.

2. Switzerland The first estimates of Switzerland’s na­ tional income were made by T. Geering and R. Hotz in their treatise on Swiss eco­

3 4 . DEVELOPMENTS IN EIGHTEEN EUROPEAN COUNTRIES

nomics, published in several successive edi­ tions during the first decade and a half of the present century. They related to the years 1895, 1899, 1905, and 1913, but were crudely made. In 1920, the Federal Tax Department, in a report on the operations of wartime levies, estimated the Swiss na­ tional income for 1915; and in 1921, Salome Schneider, in a study of the fiscal capacities of the cantons, made an estimate of the Swiss national income for 1919. Systematic and comprehensive estimates, however, did not begin until 1926 when the Swiss Association for Statistics and Eco­ nomics placed the subject on the agenda of its meeting. P. Mori and A. Reichlin re­ ported on it, presenting estimates for 1924 and 1925. In 1927, Julius Wyler published the most comprehensive estimate ever pre­ pared for that country, relating to the year 1924. During the succeeding years, he con­ tinued research on the subject under the auspices of the Swiss Federal Bureau of Statistics, publishing a comprehensive series in 1939 covering the national income for the years 1929-38. His estimates dealt with the national income at factor cost and were prepared by the income-distributed method. In 1941, the Federal Bureau began to publish national income estimates annually, extending Wyler’s series to additional years, but generally following the methodology and forms of presentation developed by him. They were based mainly on the statistics of wages, income tax data (for the higher in­ come groups), population and agricultural census statistics, and various financial data, and showed the national income as com­ posed of labor income, pure capital income, and mixed income. The first and the third categories of income were broken down by industrial divisions in which they originated. In 1947-48, the Federal Bureau of Sta­ tistics revised its estimates in accordance with some of the newer approaches recom­ mended by the United Nations. It broad­ ened the estimates to include national in­ come at market prices and personal income before and after taxes, but it did not go as far as to include gross national product and expenditure, or a system of social ac­ counts. The new series covered the years from 1938 on. In 1951, the National Accounts Research Unit of the OEEC (Organisation for Euro­ pean Economic Co-operation), with the aid of Dr. U. Zwingli, former member of the staff of the Swiss Federal Bureau of Statis­ tics, developed, on an experimental basis,

467

an estimate of the Swiss social accounts for the years 1938 and 1946-48; but this experi­ ment led to no immediate modification of the official Swiss estimates. The OEEC report noted that “at first glance . . . [it might seem] surprising that a highly industrialized economy [such as Switzerland] with economic statistics that are in general well developed . . . [should not have expanded] its national income sta­ tistics . . . into a system of national ac­ counts”; and it gave the following four main reasons for this fact: First, “government economic planning plays a less important role [in Switzerland] than in many European countries. . . .” Secondly, “certain basic in­ formation on the economic activity of the country is only partially recorded or not available at all,” e.g., “no census of indus­ trial production has ever been taken in Switzerland and only very few current pro­ duction series are available. . . . No bal­ ance of payments statistics are published.” Thirdly, “the collection of data required for setting up national accounts is in some ways complicated by the [federal] political struc­ ture of Switzerland.” Fourthly, “there ap­ pears to be a markedly cautious attitude among Swiss official statisticians . . . which leads to a reluctance to publish official fig­ ures based predominantly on estimates or on mere assumptions.” Finally, in 1955, the Federal Bureau pub­ lished for the first time the figures of the gross national product or expenditure, broken down into private consumption, public consumption, gross capital formation, and transactions with the rest of the world. In 1955 Erwin Wenk published a private estimate of Swiss consumption expenditure.

Bibliography 1. T. Geering and R. Hotz, Wirtschaftskunde der Schweiz, Zurich, 1902, Fourth Edition, 1910; Fifth Edition, 1914. 2. Statistik der I. Eidgendssischen Kriegssteuer, Eidgenossische Steuerverwaltung, Bern, 1920. 3. Salome Schneider, “The Tax Assets of the Cantons and the Tax Burdens” (in Ger­ man), Zeitschr. schweiz. Statistik und Volkswirtschaft, 1921, pp. 171-79. 4. P. Mori, “The National Income of Switzer­ land” (in German), Zeitschr. schweiz. Sta­ tistik und Volkswirtschaft, 1926, pp. 51237. 5. A. Reichlin, “The National Income of Switzerland, Its Sources and Its Meaning” (in German), Zeitschr. schweiz. Statistik und Volkswirtschaft, 1927, pp. 1-21. 6. Julius Wyler, “The Swiss National Income

468

7.

8.

9.

10. 11.

THE INCOME OF NATIONS

in 1954” (in German), ibid., 1927, pp. 359-412; “National Income” (in German), Handbuch der schweiz. Volkswirtschaft, 1939, vol. II; “Swiss National Income” (in German), Zeitschr. schweiz. Statistik und Volkswirtschaft, 1941. Eidgenossisches statistisches Amt (Federal Statistical Office), Schweiz. Volkseinkom­ men (Swiss National Income), 1924, 1929, bis 1938, Bern, 1941; Volkseinkommen Schweiz, 1938-1942, Bern, 1943; Das Volkseinkommen der Schweiz 1938-1947, Bern, 1948; same for 1948, Bern, 1949; same for later years, see annual publica­ tions Volkseinkommen, etc., and also Sta­ tistisches Jahrbuch der Schweiz; “Das Volks­ einkommen der Schweiz, 1954,” Volks­ wirtschaft, October, 1955. U. Zwingli, “Calculation of Swiss National Income and Its Reliability” (in German), Zeitschr. schweiz. Statistik und Volks­ wirtschaft, April, 1949; “Consumption Ex­ penditures and National Income” (in Ger­ man), ibid., 1949, Heft 4/5. Edward Kellenberger, “National Income, Tax Revenue and Military Expenditures” (in German), Publication in Honour of Dr. Hans Schorer, Bern, 1947. Erwin Wenk, “Consumption” (in German), Handbuch der schweiz. Volkswirtschaft, 1955. Organisation for European Economic Co­ operation (OEEC), National Accounts Re­ search Unit, Switzerland, National Ac­ counts Studies Series, Paris, 1952.

3. Belgium The first estimate of Belgian national in­ come was prepared in 1919 by Charles Clavier, General Director of the Ministry of Finance, and was an approximate average for the years 1912-14. In 1925, Professor Fernand Baudhuin started his income esti­ mates with a similar study for 1913, fol­ lowed by estimates for 1924, 1927, and various years between 1930 and 1954, as part of an annual review of the Belgian economy. Up to the middle of the 1950’s, Baudhuin’s estimates were the only available series covering a substantial period of years in a consistent way. Baudhuin followed the concept of national income at factor cost, showing the aggregate by distributive shares, and giving a partial breakdown by indus­ trial divisions. He used tax statistics, wage and population data, miscellaneous financial reports, and his own personal knowledge, but generally gave little information about his sources and methods. In 1955, he pub­ lished an estimate of national consumption expenditure. In 1948, the official National Institute of

Statistics began to inquire into the subject, without developing, however, a series that would replace Baudhuin’s. In 1949, the gov­ ernment appointed a special commission to inquire into the whole subject of national income accounting, but up to 1955, at least, this Commission made no report on its findings. In 1950 van der Meiren prepared a private estimate of the gross national product and expenditure for the years 1938 and 1946-49, partly from Baudhuin’s fig­ ures and partly from other sources. During this period, a private group of sta­ tisticians, set up by the Institut de Sociologie Solvay under the chairmanship of Professor Krischen, prepared a report dealing with the essential elements for setting up a com­ prehensive system of social accounting for Belgium and actually developing tentative figures of such accounts, including the gross national product and expenditure for the years 1948-51. The group indicated some of the obstacles that stood in the way of the official preparation and publication of such gross national product and social ac­ counting estimates in Belgium. One of the greatest obstacles, it said, was the fear of certain groups lest the publication of such information be used for political purposes and tend to accentuate governmental direc­ tion of the economy. Other difficulties were the lack of consistency and co-ordination among the existing statistics, the poverty of data on consumption and capital formation generally and on production by small busi­ nesses, and the lack of reliability in many of the published statistics. So long as the existing economic statistics were concen­ trated largely on the operation of primary industries and changes in the price level, said the group, the establishment of compre­ hensive national accounting in Belgium was well nigh impossible. In 1955 the Institut brought out a volume with national income figures for the years 1948-54, broken down by industrial divisions, distributive shares, and final expenditures.

Bibliography 1. Charles Clavier, La Fortune Belgique a la Veille de la Guerre (National Wealth of Belgium on the Eve of the War), Brussels, 1919. 2. Fernand Baudhuin, Le Capital de la Belgique et le Rendement de son Industrie avant la Guerre, Louvain, 1924, p. 49; La Structure Economique de la Belgique, Editions de la Societe d’Etudes Morales, Sociales et Juridiques, Louvain, 1926, p. 123; “National

3 4 . DEVELOPMENTS IN EIGHTEEN EUROPEAN COUNTRIES Incom e in 1927” (in F ren ch ), Bulletin d'lnformation et de Documentation de la Banque Nationale de Belgique, M ay, 1927; “N ational Incom e in 1930” (in F ren ch ), Bulletin d’lnformation et de Documentation de la Banque Nationale de Belgique, M ay, 1930, pp. 373— 77; “Incom es” (in F ren ch ), Bulletin de I’lnstitut de Recherches Economiques, Catholic U niversity o f Louvain, February, 1937, pp. 121-33; “Belgian Finances in 1937”; also “in 1939” (in F ren ch ), ibid., February, 1938, 1939; “E m ploym ent” (in F ren ch ),

Principes Permanents d’Economie Privee, Louvain, 1944; “Prices, Consum ption and N ational Incom e” (in F ren ch ), Bulletin de

Vlnstitut de Recherches Economiques et Sociales, July, 1947; and A ugust, 1948; “B el­

3.

4.

5. 6.

gium in 1947” and the same “in 1948” (in F ren ch ), Agence Economique et Financiere, Brussels, M ay, 1948; A pril, 1949; “N ational Incom e and Consum ption in 1954” (in F ren ch ), ibid., A pril, 1955. E. R. van der A a, “C alculation o f the B el­ gian N ational Incom e” (in F ren ch ), Bulletin de Statistique, Institut N ational de Statis­ tique, Brussels, January, 1948. P. van der M eiren, “N ational Incom e, G ross N ational Expenditure and Related T otals in Belgium , 1938 and 1 9 4 6 -1 9 4 9 ” (in E n glish ), Economica Documentatie, A ugust, 1950. Annuaire Statistique de la Belgique, Institut N ational de Statistique, 1949, 1950, 1952. G roupe d’Etudes de la Com ptabilite N ation ­ ale, Premiers Elements d’Une Comptabilite

National de la Belgique 1948-1951 (First Elements of National Accounting of Belgium 1948-1951), Institut de Sociologie Solvay, Brussels, 1953. 7.

Economie Beige et Comptabilite Nationale 1948-1954, Institut de Sociologie Solvay, Brussels, 1955.

4. Luxembourg The first systematic national income esti­ mate for Luxembourg was developed in 1948. It was prepared by a newly established research unit in the Ministry of Economic Affairs. National income was estimated at factor cost by distributive shares for the years 1945-47 and also backward for the years 1913, 1918, 1928, and 1935-37. It was also broken down by industrial origin for the years 1945-47. These studies were continued during the succeeding years, even­ tually incorporating, in 1952-55, the prepa­ ration of estimates of the gross national product or expenditure at market prices for the years beginning 1948.

Bibliography 1. A . Schwinnen, National Income of Luxem­ bourg (in F ren ch ), Etudes Economiques,

469

M inistere des Affaires Econom iques, Luxem ­ bourg, 1948. 2.

National Income of Luxembourg in 1950 and 1951 (in F ren ch ), Service d’Etudes

Econom iques, Cahier N o . 4; M inistere des Affaires Econom iques Luxembourg, 1952; La Comptabilite Nationale du Duche (an­ n u al). 3. U nited N ations, Statistics of National In­ come and Expenditures, Statistical Papers, Series H . N o . 10, N ew Y ork, 1957. 4. Organisation for European Econom ic C o­ operation, Statistics of National Product and Expenditure, 1938, 1947-1952, Paris, 1954.

5. Denmark In 1930, T. Kristensen prepared an esti­ mate of the national income for Denmark by the income-distributed method. In the second half of the 1930’s, the Statistical Department of Denmark prepared an esti­ mate for a single year, 1935. After World War II, the Department took up the prepa­ ration of such estimates on an annual basis, publishing, in 1946, a comprehensive study covering the national product and income for the period 1930-46 and thereafter pub­ lishing the annual figures in its Statistical Yearbook. In the meanwhile two more pri­ vate studies of importance had appeared, one by V. Kampmann in 1942, and the other by H. C. Jorgenson in 1946. In 1948, the Department revised its series both as regards the concept and the coverage of the national product and its components, developing particularly its consumption and investment aspects and presenting the esti­ mates in the form of social accounts. The central part of the new system was the “National Product Account,” from which goods and services were being, so to say, “sold” to households, government, the rest of the world, and to the national capital (resting) account. No separate “Enter­ prises” account was provided. The main emphasis in the official Danish estimates was placed on the gross domestic product, which was computed both at mar­ ket prices, i.e., including indirect taxes, and at factor cost, i.e., excluding them. It was estimated directly from production statis­ tics. Net domestic product was derived from the gross by deducting depreciation and net interest payments to abroad. The Statistical Department also makes an annual forecast of the gross national product as a basis for the formulation by the government of its economic and fiscal policy.

470

THE INCOME OF NATIONS

In 1947, a Ministerial Committee for Economics and Supply was set up to pre­ pare a national economic budget for the uses of the Danish Parliament, the cabinet members, and administrative officers. The Committee consisted of the Prime Minister and the key ministers. The first such budget was published in 1948, the second in 1949. Others followed in the succeeding years. The economic budget, or simply the “na­ tional budget” as it is called, is regarded as a governmental declaration of economic policy. It consists of two sets of parallel columns, one presenting the actual trans­ actions of the completed year, and the other the forecasts for the coming year. The col­ umn and line headings in each set are iden­ tical. They run as follows:

N ational Incom e and N ational Product Account

In 1954, the Department published a his­ torical series of the Danish national income and expenditure from 1870 to 1952, by tenyear averages, with breakdowns by factor incomes, industrial divisions, and by ex­ penditure for consumption and investment.

Bibliography 1. Thorkil Kristensen, Danmarks Driftsregnskab, Copenhagen, 1930. 2. Viggo Kampmann, “Denmark’s National In­ come Estimates” (in Danish), Nationaloekonomisk Tidsskrijt, 1942. 3. H. C. Jorgenson, Hvor stor en Del af Indtaegten unddrager Skattesnyderne det offentlige, Copenhagen, 1946. 4. Denmark, Det Statistiske Department, Na­ tional Product and National Income, 19301945 (in Danish), Statistiske Meddelelser,

Operating A ccount of the Private Sector

Operating Account of the Governm ent Sector

Balance of Paym ents Investm ent

Consumption Private G overnment Gross investm ent Private Government Export Import T otal G N P W ages, Interest, and Profits For private persons Corporation saving Corporation taxes Export duties Profits of governm ent enterprise Interest charged to government buildings Depreciation and m aintenance Private Government Indirect taxes Subsidies T otal G N P Transfers Interest on governm ent debt Social Security, etc. D irect personal taxes, real estate taxes, etc. F ees and Fines Saving Private Governm ent G ifts from abroad Loans from abroad T otal s o u r c e : Economic Survey of Denm ark, National Budget fo r 1949, Prepared by the M inisterial Com­ m ittee for Econom ics and Supply, T able 55.

34.

DEVELOPMENTS IN EIGHTEEN EUROPEAN COUNTRIES

February, 1948; “Income and Wealth, 19491950” (in Danish), Statist. Efterretninger, July 13, 1950; National Accounting 1938 and 1947-1952 (in Danish); Statist. Meddelelser, 1955; Statistisk Arbog, 1948, 1949, 1950, etc.; Statistiske Efterretninger, Janu­ ary, 1950, 1951, and other years; “National Income 1870-1952,” Statistiske Efterret­ ninger, May, 1954. 5. Oekonomiske Secretariat, Economic Survey of Denmark, National Budget for 1949 (in English), Copenhagen, 1950; Economic Sur­ vey of Denmark, 1952 (in English), Copen­ hagen, 1942; same survey for subsequent years in Danish or English. 6. Organisation for European Economic Co­ operation (OEEC), National Accounts Re­ search Unit, Denmark, National Accounts Studies, Paris, 1951.

6. Norway The first recorded estimate of the Nor­ wegian national income was made in 1893 for the year 1891. This was followed by other private estimates made at later dates for the years 1898, 1906, and 1913-23 (see OEEC Studies, Norway, p. 11). Plans for systematic studies were formulated in the 1930’s at the University of Oslo under the guidance of Professor Ragnar Frisch, but they never came to fruition. During World War II and the German occupation, the Central Bureau of Statistics began work on national income estimates for the years 1935-43 based conceptually on national ac­ counting principles, thus paralleling Stone’s and Meade’s work in England. The national product was estimated by the net output method and was accompanied by estimates of expenditures for consumption, domestic capital formation, and exports and imports. By the time the war was over, the estimates were nearly completed, and were published in 1946. These estimates, in the form of social accounts, were continued in the succeeding years in an expanded form and were made an integral part of a national income budget (called simply “national budget”) trans­ mitted by the Finance Ministry to Parlia­ ment annually since 1946. The main concept in these accounts was the “national product” defined as the value of goods and services produced by the factors of production resi­ dent in the economy, valued net of deprecia­ tion but otherwise at market prices. The concepts of the gross national product, in­ clusive of depreciation, and of the “factor income,” corresponding to net national in­

471

come at factor cost, were also used, but only in a minor way. The sector accounts were developed by Professor Aukrust in a unique fashion. Pro­ ductive activity and consumption activity were shown in separate sectors, and within the former, separate sectors were established for the different branches of industry. The accounts for each sector distinguished sharply between product flows and financial flows and between current and capital transactions, and the accounts showed, by means of double entry, both the activity that took place within a sector and the activity that took place between sectors.

Bibliography 1. Ragnar Frisch, National Accounting (in Norwegian), Oslo, 1940; “From National Accounts to Macro-Economic Decision Mod­ els,” International Association for Research in Income and Wealth, Income and Wealth, Series IV, Bowes and Bowes, London, 1955. 2. Hans Luihn, National Accounting in Nor­ way (in Norwegian), Oslo, 1940. 3. Odd Aukrust and Others, Costs of War to Norway (in Norwegian), Oslo, 1945. 4. Odd Aukrust, National Accounting 19301939 and 1946-1951 (in Norwegian with English summaries), Oslo, 1952 (reprint from Norwegian Central Statistical Office, vol. XI, p. 109); ibid., 1948-1953 (in Nor­ wegian), Oslo, 1954; same publication by the Central Statistical Office for succeeding years; Akonmisk Utsyn Over Aret 1954, Oslo, 1955, and the same published annually by the Central Statistical Office; “On the Theory of Social Accounting,” Intern. Assn. for Res. in Income and Wealth, Conference, September, 1949, published in Review of Economic Studies, vol. XVI, no. 41, 19491950; and “Trends and Cycles in Norwegian National Income Shares,” Intern. Assn. for Res. in Income and Wealth, Paper at the Conference, September, 1955. 5. Norway, Central Bureau of Statistics, Na­ tional Income of Norway 1935-1943 (in Norwegian), Oslo, 1946. 6. E. Petersen, National Income 1935-1939 (in Norwegian), Bergen, 1947. 7. Organisation for European Economic Co­ operation, National Accounts Research Unit, Norway, National Accounts Studies, Paris, 1953. 8. Norway, Ministry of Finance, National Ac­ counting and National Budget 1945-1946 (in Norwegian), Oslo, 1946; same publica­ tion for the succeeding years.

7. Finland Professor Valter Lindberg of the Univer­ sity of Helsinki and of the Bank of Finland

472

THE INCOME OF NATIONS

was the pioneer estimator of Finland’s na­ tional income. Starting with a theoretical inquiry into the subject in the middle 1920’s he published a succession of national in­ come estimates during the 1930’s and 1940’s which provided a continuous series for the years 1926-48. In his earlier estimates, he used the concept of national income at fac­ tor cost (adjusted for the net balance of international payments) but, in 1948, he also developed estimates for national income at market prices. He used the net output approach, giving the composition of the na­ tional income by industrial origin, with specifications in each case of the gross value of the output, the pertinent deductions, and the net value. Conceptually and methodo­ logically, Lindberg’s estimates were in line with the progress made in other countries at the time. After World War II, the preparation of national income estimates was taken up by the government, with the Central Statistical Office and the Economics Department of the Ministry of Finance assuming the task. A new series was inaugurated, embracing the year 1938 for the prewar period, and the years beginning 1948 for the postwar era, and covering national income at factor cost, national income at market prices, and gross national product or expenditure. Breakdowns were furnished by distributive shares, industrial origin, and type of con­ sumption and investment expenditure. The larger aggregates were deflated at constant prices. Beginning with 1951, the estimates also included sector accounts. In their newer form, the estimates came to be used ex­ tensively by the government in the prepara­ tion of its annual budget and of its economic development plans. At the same time, other statistical surveys of other aspects of the national income were inaugurated by the government. Thus, in 1948, a national survey was made of the distribution of income by size of individual income and other categories, and in the 1950’s a detailed study was published of the geographic distribution of the levels of in­ come.

Bibliography 1. Valter

Lindberg,

“Incom es

in

Finland,”

Bank of Finland Monthly Bulletin, 1926, N o . 11; Utwecklingen av den Allmanna Inkonstkatlen i Finlands statsskatte-system (in Sw ed ish), H elsingfors, 1934; Suomen Kansantulo (N ational In co m e), H elsinki, 1935; “The N ational Incom e o f Finland,” Bank of

Finland Monthly Bulletin, 1937, N o . 2; Suomen Kansantulo Vuosina 1926-1938 (N ation al Incom e 1 9 2 6 -1 9 3 8 ) (in Finnish with G erm an sum m ary), H elsinki, 1943; Estim ates o f national incom e at market prices 1939-1945, social accounts and de­ velopm ent o f official statistics, articles in Ekonomiska Samfundets Tidskrift, N o . 60, 1943, and N o . 3, 1948, and in Vapaa Yrittajatoiminta, 1948 (in Swedish and F in nish ). 2. Eino H . Laurila, “F inland’s N ational In­ com e A fter the War” (in E n glish ), Bank of Finland Monthly Bulletin, July-A ugust, 1949; articles (in Finnish and Sw edish) on national incom e 1926-1949 and later years, on the volum e o f industrial production 1938-1947, and on social accounts, in Kansantaloudellinen Aikakauskirja N o . 3, 1949, and N o . 2, 1952; Tilastokatsauksia, 1949, N os. 1-2; Statistiska Oversikter, 1950, N os.

11- 12 . 3. G unnar Fougstedt, “Distribution o f N ational Incom e in Finland” (in S w ed ish), Ekono­ miska Samfundets Tidskrift, N o . 4, 1948. 4. Lara W ahlbeck, Orn Inkomstnivans Geografi in Finland ar 1950 (G eographic D is­ tribution o f Incom e in F in la n d ), H elsinki, 1955. 5. Pentti V iita, “Agriculture in the C om puta­ tion o f the N ational Incom e in Finland” (in F in n ish ), Journal of Scientific Agricultural Society of Finland, 1954, N o . 3. 6. Finland, Central Statistical Office, Bulletin of Statistics, N os. 11 -1 2 , 1950; and other issues in later years. 7. Finland, M inistry o f Finance, Econom ics Departm ent, Economic Survey of Finland (in Finnish, Swedish, and E n glish ), pub­ lished annually since 1948, section on na­ tional incom e and social accounts; see par­ ticularly Economic Survey 1954.

8. Iceland. In Iceland, estimates were available in 1953 of the national income and the gross national product broken down by final ex­ penditures for the years 1938 and 1947-52. They have been prepared annually since then by the Iceland Bank of Development.

Bibliography Organisation for European Econom ic C o-opera­ tion, Statistics of National Product and Ex­ penditure 1938, 1947-52, Paris, 1954, pp. 59, 173-74; U.N. Yearbook of National Ac­

counts, 1957.

9. Austria The first estimate of the national income in the old Austrian empire was made in

34.

DEVELOPMENTS IN EIGHTEEN EUROPEAN COUNTRIES

1861 by Karl Czoerning, the eminent Aus­ trian administrator and statistician, and it was followed in 1883 by an estimate by Roschman for 1880 (Chapter 9, Section 2, and Chapter 10, Section lc ). In 1916, Fried­ rich von Fellner published an estimate for the whole of Austria-Hungary for the years 1900 and 1911-13; and in 1918 A. Gurtler published a like estimate covering the war­ time period. After the country’s partition, Ernst von Waizner published an estimate of the na­ tional income of what remained of old Austria, as well of what was separated from her and transferred to other states, for the average 1911-13 year. Next, in 1930 and 1931, F. Savorgnan and W. Winkler pre­ pared national income estimates for Aus­ tria, one for the period 1903-10 and the other for the period of World War I. In 1934, research in the current national income and its growth was undertaken by the Austrian Institute of Business Cycle Re­ search (Konjunktur Institut). In 1937, the results were published covering the period 1929-34. The concept used was that of aggregate private income, rather than na­ tional income proper; and the estimates were prepared by the income-distributed method, largely from income tax and population census data, with the evaluation made both in current and constant prices. During World War II and the immediate postwar years, when Austria was occupied first by Germany and next by other foreign powers, all national income research in the country was suspended, but it was revived in 1949, when Austria’s independence was partially restored and when some degree of stabilization of her money and price level was achieved. In 1950, the Austrian Central Statistical Office (Statistisches Zentralamt) published national income estimates for 1948 and 1949. The Austrian Institute for Economic Research (Institut fur Wirtschaftsforschung), in 1951, published independent estimates for the years 1948-50, using some­ what different concepts and methods and arriving at somewhat different results. In 1952, the two groups joined in the prepara­ tion and publication of a common estimate for the years 1950 and 1951; and their collaboration continued during the succeed­ ing years. The new estimates gave a continuous series from 1948 on, with a comparison in constant prices with the figures for 1937. They gave not only the national income at factor cost but, in line with the general

473

trend, gave the gross national product by industrial divisions and the gross national expenditure by types of consumption and investment. Those published in 1953 also presented a set of social accounts, thus bringing the Austrian estimates once again in line with world developments.

Bibliography 1. Karl Czoerning, Handbiichlein fiir die oesterreichische Monarchie, Vienna, 1861 (figures reproduced in Mulhall’s Dictionary of Statistics, London, 1899, p. 322). 2. Friedrich von Fellner, “National Income of Austria-Hungary” (in German), Statist. Monatschrift, Statist. Zentral Komm. N.F., vol. XXI, 1916. 3. A. Gurtler, “National Income of AustriaHungary” (in German), Weltwirtsch. Archiv. 13 Bd., 1918. 4. Ernst von Waizner, “The National In­ come of Old Austria and Its Distribution Among the Succession States” (in Ger­ man), Metron, vol. VII, No. 4. 5. F. Savorgnan, “Methods for Measuring the Distribution of Austria’s National Income, 1903-1910” (in Italian), Bulletin de Vlnstitut International de Statistique, Tome XXV, 3me livr., 1931. 6. Wilhelm Winkler, “Shifts of Income in Austria during the World War” (in Ger­ man) (Wirtsch. und Sozialgeschichte des Weltkrieges, Oesterr. Ung. Serie), 1930. 7. Oesterr. Institut fiir Konjunkturforschung, “The Development of the National Income in Austria Since 1929” (in German), Monatsberichte, Vienna, August, 1935; also 1937, pp. 130-37. 8. Oesterr. Statist. Zentralamt, Statist. Nachrichten, Vienna, 1950, No. 3; and later num­ bers, particularly Supplements to the De­ cember, 1956, and November, 1957, issues. 9. Oesterr. Institut fiir Wirtschaftsforschung, Estimates of the National Income in Aus­ tria in 1937, 1948 and 1949 (in English), Special Paper No. 7, Vienna, 1951; Gen­ eral Survey of the Austrian Economy, 1954; 1955 (in German), Statistische Ubersichten zu den Monatsberichten, Vi­ enna, February, 1955; February, 1956. 10. Oesterr. Statist. Zentralamt und Oesterr. Institut fiir Weltwirtschaftsforschung, Es­ timate of the National Income of Austria for 1950 and 1951 (in German), Vienna, 1952; Austria’s National Income 1953 (in German), Vienna, 1954; “Gross Value of Austria’s Industrial Output” (in German), Statist. Nachrichten, February, 1955.

10. Hungary The earliest estimates of Hungarian na­ tional income appeared in an over-all esti­

474

THE INCOME OF NATIONS

mate of the national income of AustriaHungary. Most important among them were those made by von Fellner before and dur­ ing World War I. After 1918, when Hungary once more became an independent state, the estimates, prepared again by von Fellner, related to her territory alone. Published in 1923, 1931, and 1937, they covered the years 1911-12 and 1929-36. Based on the Smithian restricted material production con­ cept, they were prepared mainly by the net output method and showed the home-pro­ duced income at factor cost by industrial divisions, plus the net income from abroad. In 1936 and 1937, M. Matolcsy and S. Varga, under the auspices of the Hungarian Institute of Economic Research, prepared estimates that were more comprehensive and detailed than von Fellner’s. They were based on a broader concept which, in addition to material production, included both remu­ nerated and unremunerated private services and even certain government services. They showed net domestic capital formation, and they covered a longer span of years, e.g., 1924/25-1936/37. They were prepared, like von Fellner’s, by the net output method, but for agriculture and mining they also pro­ vided the gross outputs. The revised esti­ mate, prepared in 1938, was published in English, in London, and attracted wide at­ tention among statisticians because of its unique concepts and refined techniques. In 1938, the same institute published an exten­ sion of the Matolcsy-Varga estimates to the year 1937/38. In 1939, Jules Neubauer pub­ lished an independent set of estimates. During World War II and the years im­ mediately succeeding it, the Hungarian In­ stitute of Economic Research continued the publication of national income estimates, extending the Matolcsy-Varga figures suc­ cessively to 1946-47. But eventually, under the new Communist regime, the Central Sta­ tistical Office (Statisztikai Hivatal) took over this activity. It discontinued the old series, substituted the Marxian material production concept of national income (as laid out by Soviet Russia’s economists) for the earlier more comprehensive one, and adjusted the estimates to the needs of the State’s five-year economic investment plans. The estimates as published in 1953 related to the years 1938 and 1947 through 1951, and gave its breakdown by industrial divisions.

Bibliography 1. Friedrich von Fellner, “National Income of Austria-Hungary” (in German), Statist.

2. 3. 4.

5.

6.

7.

8.

9.

10.

Monatschrijt, K.K. Statist. Zentral Kommission N.F., vol. XXI, 1916; “Distribution of Hungary’s National Income Between the Present Hungary and the Succession States” (in German), Metron, vol. Ill, 2, 1923; “Shares of Labor and Capital in the Hun­ garian National Income” (in French), Revue des Travaux de I’Academie des Sci­ ences Morales et Politiques, Paris, 1936— 37; with Varga, Szigeti, and Others, “In­ quiries into the Problems of National In­ come” (in French), Journal de la Societe Hongroise de Statistique, vol. 16, 1938. M. Matolcsy and S. Varga, The National Income of Hungary 1924/25-1936/37 (in English), London, 1938. M. Matolcsy, The Distribution of Income and Taxes in Hungary (in Hungarian), 1938. Jules Neubauer, “The Size of the Hungar­ ian National Income” (in French), Magyar Statisztikai Tarsasag (Journal de la So­ ciete Statistique Hongroise), vol. 77, 1939; also in Revue Econ. Hongroise, vol. 63, 1939. Ungarisches Institut fiir Wirtschaftsforschung, “Hungarian National Income” (in German), Siid Ost Ekonomist, January and July, 1943. Institut Hongrois des Recherches Econo­ miques (same as above), The Hungarian National Income (in French), Monographie No. 15, 1938; Monographie No. 20, 1942; Bulletin, March, 1947. Hungarian Central Statistical Office (Sta­ tisztikai Hivatal). Articles on the Hungar­ ian five-year economic plan and economic growth published in Statisztikai Szemle (Statistical Monthly Review), January, 1951; January, 1952; November, 1952; January, 1953; June-July, 1954; and in the Hungarian Soviet Economic Review (Mag­ yar Szovjet Kozgazdasagi Szemle), Febru­ ary, 1953 (all of them in Hungarian). Istvan Friss, “The Development of the Hungarian National Economy since the Liberation” (in Hungarian), Economic Re­ view of the Academy of Sciences (Kozgaz­ dasagi Szemle, Magyar . . . Akademia Kozgazdasagi), January-July, 1955. U.N. Economic Commission for Europe, The Economic Survey of Europe Since the War, Geneva, 1953; Economic Survey of Europe, 1954, 1955, 1956, and 1957 issues. Plans for the Development of the Econo­ mies of the Peoples’ Democracies, Collec­ tion of Materials (in Russian), Isdatelstvo Inostrannoi Literatury, Moscow, 1952, Part IV

11. Czechoslovakia In 1925, S. N. Prokopovitch, the noted Russian emigre economist, published an esti-

34.

DEVELOPMENTS IN EIGHTEEN EUROPEAN COUNTRIES

mate of Czechoslovakia’s national income. In 1927-29, two Ministers of Finance in Czechoslovakia, Englis and Hotowetz, made crude estimates of the country’s national in­ come by taking certain assumed amounts of per capita consumption expenditures of the people and multiplying them by the number of people. Their figures differed widely from each other, the first being 60 billion Koruns and the second 76 to 83 billions. The first truly systematic estimate was made by Dr. Paul Smutny in 1930 and it related to the period 1927-29. Smutny found both of the foregoing estimates too low, arriving at a figure of 90 billion Koruns for 1928. Smutny used the restricted mate­ rial production concept, apparently under von Fellner’s influence, and made computa­ tions by a combination of the net output and income-distributed methods. He con­ structed an account that he called a “national balance sheet,” which contained some of the features of the social accounts developed years later by Stone and Meade in England; and he urged the International Institute of Statistics, which met in Tokyo that year, to endorse his proposed type of national ac­ counting for wider use. On the left-hand side of his balance sheet or account, he showed the national output by industries, while on the right-hand side he entered the expenditures or uses of the national product. His account for 1928 appeared as follows:

1. 2. 3. 4. S.

National Net Output or Income Agriculture and forestry Mining Manufacturing Trade and transportation Liberal professions Total

National Expenditure 1. Personal consumption 2. National saving Total

Billion KC 30 3 32 18 7 90 Billion KC 81.5 8.5 90.0

During the German occupation, no esti­ mates of the national income seem to have been published in Czechoslovakia, but after the war and the establishment of the Com­ munist regime, the preparation of such esti­ mates was undertaken by the government’s Central Planning Commission, and were re­ lated to the country’s five-year economic plans. They were based on the material pro­

475

duction concept and were made by the net output method. In 1948-50 several studies on the subject were published, but during the years 1951-55 publications of such studies grew sparse. The estimates repro­ duced by the Economic Commission for Europe in its Economic Survey in 1953 show the national income of Czechoslovakia for the years 1937 and 1947-51 broken down by industrial origin.

Bibliography 1. Paul Smutny, “A National Balance Sheet and National Savings” (in Czech), Statisticky Obzor (Statistical Review), Prague, February, 1931, pp. 31-56. 2. Milos Stadnik, National Income and Its Distribution (in Czech), Prague, 1946; National Income and the Method of Esti­ mating It (in English), Czechoslovak Min­ istry of Information, Prague, 1947 (with figures for 1937, 1946, 1947); also articles in Statisticky Obzor, No. 2, 1948; and in Socialni Revue, No. 3, 1948, and No. 3, 1949. 3. I. Kochanowskov, “On Computing the Na­ tional Income in Czechoslovakia,” Socialni Revue, Czechoslovak Ministry of Social Welfare, August-November, 1948. 4. Czechoslovakia, General Secretariat of the Economic Council, “National Income and Expenditure 1946 and 1947” (in Czech), Prubeh Phfineni Hospodarskeho Planu Rok 1947, Prague, 1948. 5. J. Habr, “National Accounts” (in Czech), Nove Hospodarstvi, No. 6, 1948. 6. S. Heretik, “Slovakia’s Share in Czecho­ slovak National Income” (in Czech), Socialni Revue, No. 3, 1948. 7. O. Zabradnicek, “Income and Expenditure of the Population in the Czech Provinces in 1947,” Statisticky Zpravodas, July, 1950. 8. B. Spacil, “The Financial Past and Pres­ ent,” Hospodar, No. 14, 1950 (in Czech). 9. U.N., The Economic Survey of Europe Since the War, Geneva, 1953; Economic Survey of Europe, 1957, Chap. 1. 10. V. Busek and N. Spulber, Czechoslovakia Under the Communists, New York, 1957.

12. Poland The first estimate of Polish national in­ come was prepared by B. Dederko in 1932. It was based on the net output method, but because of the incompleteness of Poland’s production statistics was not altogether sat­ isfactory. In 1935, it was followed by the more comprehensive income studies for 1929 and 1933 made under the auspices of the Polish Institute of Economic Research by Michael Kalecki and Ludwik Landau. These

476

THE INCOME OF NATIONS

estimates were based on an entirely different approach. For 1929, the final expenditure method was used exclusively, and for 1933 it was combined with the income-distributed method. Two other studies in the national income field were published by the Institute at about the same time, one by Landau on labor income and one by Jan Wisniewski on the distribution of income by size. Kalecki and Landau adopted the concept of national income as the net value of the nation’s output of goods and services. How­ ever, they tended to limit the concept of goods and services, in the main, to those passing through the market on the ground that these alone possessed objective value (see Chapter 12, Section 4, above). They computed national income as the sum of the values of: (a) the consumption of goods and services by the urban population, (b) the consumption of industrial products by the rural population, (c) the consump­ tion by the latter of agricultural products, (d ) public and private education, and (e) gross capital formation. Urban consumption was estimated on the basis of available sta­ tistics of the purchases by the city popula­ tion of certain commodities (bread, meat, sugar, etc.), and on the official studies of the family budgets of the various classes of the urban population and the share that the expenditures for these commodities repre­ sented in these budgets. The consumption of the rural population of industrial and agricultural products was estimated partly from production data and partly on the basis of studies of family budgets of landowners, farmers, and agricultural workers. Capital formation was estimated as the sum of in­ ternal gross investment in fixed capital (in­ cluding government investment), increase in inventories, and foreign capital investment. Inasmuch as they estimated consumption at market prices, and calculated capital for­ mation gross of depreciation, the concept of national income used by Kalecki and Landau more closely resembled gross na­ tional product than net national income at factor cost. The aggregates were broken down by divisions of economic activity, by distributive shares, by types of consumption and gross saving, and by the size of indi­ vidual incomes in different social and eco­ nomic groups. During the 1930’s, one more private esti­ mate was published, prepared by PetyniakSanecki. It related to the year 1938 and to the national income at market prices. Dur­

ing World War II and the German occupa­ tion, all national income research was sus­ pended. After the war, with the inauguration of the Communist regime in Poland, national income estimating became an official func­ tion. The Central Planning Administration began publication of annual estimates in 1948, using them in its economic planning work. At the same time, a shift was made to the Marxian material production concept of national income as outlined in the offi­ cial Soviet Russian publications. The first official estimate was for the year 1947. For comparative purposes, the Petyniak-Sanecki figure for 1938 was used, but it was reduced to fit the material production concept by deducting 10 per cent for the value of serv­ ices and 5 per cent for depreciation. From that time on, estimates were made annually by the Planning Administration and the Central Statistical Office. The official estimates were prepared by the net output method, with breakdowns for agriculture and forestry, manufacturing and mining, handicrafts, construction, transport and communications, trade, and the rest of national production. For state industry, the distributive shares were given, such as wages, pensions, subsidies, rents, taxes, state profits, etc. The part of the income originating in the socialized sector (state, local, and co­ operative) was distinguished from the part originating in the private sector. Substantial explanations were published in 1950 regard­ ing the sources and methods of preparation of the figures. The estimates are published annually in the official statistical yearbook.

Bibliography 1. B. Dederko, “National Income,” Statistical Quarterly, 1932, No. 2 (in Polish). 2. Michal Kalecki and Ludwik Landau, Szacunek Dochodu Spolecznego, w.r. 1929 (Estimate of National Income for 1929), Institut Badania Konjunktur Gospodarczych i Cen (Institute of Economic Research), Tom I., Warsaw, 1934; Dochod Spoleczny w.r. 1933 i Podstawy BadaA Perjodycznych nad Zmianami Dochodu (National Income in 1933), Publications of the same Insti­ tute, vol. 4, Warsaw, 1935. 3. Ludwik Landau, Dochody z Pracy Najemnej w.r. 1929 (Labor Income in 1929), Publications of the Institute of Economic Research, vol. 2, Warsaw, 1934. 4. Jan Wisniewski, Rozklad Dochodow Wedlug Wysokosci (Distribution of Income by Size), Publications of the Institute of Eco­ nomic Research, vol. 3, Warsaw, 1934.

34.

DEVELOPMENTS IN EIGHTEEN EUROPEAN COUNTRIES

5. Petyniak-Sanecki, Current Economic Prob­ lems (in Polish), Lwow, 1939, Part II. 6. Central Statistical Office, “Polish National Income 1947” (in Polish), Statystyka Polski, Series D, No. 13, Warsaw, 1949. 7. Central Planning Administration, “Polish National Income, 1838, 1946-1959” (in Polish), Gospodarka Planowa, 1950, No. 7: estimates for 1950-54 appeared in 1951— 55 issues of the same publication; the dis­ tribution of the national income is analyzed in the July-August, 1957, issue. 8. B. Mine, Polish Minister of Finance, Paper on Concept of National Income (in Polish), Nowe Drogi, 1948, No. 8; “Problems of Planning the National Economy,” Gospo­ darka Planowa, March, 1953; “Problems of National Income,” Ksiazka i Wiedza, Warsaw, 1951; also article in Ekonomista, February, 1952 (all of these in Polish). 9. Polish Embassy in the United States, Po­ land’s Six-Year Plan, 1950-1955, Wash­ ington, D. C., 1952. 10. U.N., The Economic Survey of Europe Since the War, Geneva, 1953; Economic Survey of Europe, 1955, 1956, 1957. 11. Plans for the Development of the Econo­ mies of the Peoples’ Republics, Isdat. Inostr. Literatury, Moscow, 1952, Part III (in Russian). 12. J. Nowacki, “Estimates of the National Income for 1956” (in Polish), Wiadomosci Statystyczne, 1957, No. 2. 13. Rocznik Statystyczny (Statistical Year­ book), Warsaw, 1956 and other years.

13. Bulgaria Professor Kiril I. Popoff, director of the Bulgarian Central Statistical Office, prepared the first estimate of Bulgarian national in­ come in 1915. It related to the years 1892 and 1911. In 1929, Dr. A. Tchakaloff, Director of Research of the National Bank, published an income study for 1911, 1926, and 1929, following it in 1937 with one for the years 1924-35. In 1946, he revised and enlarged his study, extending the series to 1945 and providing also an analysis of the national expenditure. Tchakaloff adopted the comprehensive concept of national income. He defined it as the aggregate of all incomes, both mone­ tary and in kind, of individuals and corpo­ rate bodies inhabiting the national territory derived from the production of new goods and services. For the most part, he followed the net output method of estimation, de­ parting from it in favor of the incomedistributed method only in the case of in­

477

comes from professional and other services, petty trades and handicrafts, and small in­ dustry. In addition to estimating national income by divisions of economic activity, he developed subsidiary estimates of its distribution among the various economic groups, and of its disposition for consumer goods, consumer services, direct taxes, and net capital formation or saving. All these analyses were given for the entire period 1924—45. The main aggregates were deflated at constant prices. The estimates were ac­ companied by the estimator’s personal evalu­ ation of the degree of their reliability. Paralleling Tchakaloff’s work, Dr. P. Kiranoff of the Central Statistical Office pre­ pared, in the 1930’s, estimates of the na­ tional income for 1929 and 1932, and, later, for 1935, 1944, and 1945. Kiranoff’s esti­ mates were based on the Marxian restricted material production concept. After the war and the establishment of the Communist regime in Bulgaria, all na­ tional income estimates were prepared by the government and were based on the ma­ terial production concept. The official series related to the years 1939 for the prewar period and the years beginning 1948 for the postwar period and were given at 1939 prices. National income was broken down by industrial divisions and the estimates were used by the government in its eco­ nomic planning. The State Planning Com­ mission reported in 1953 a 53 per cent in­ crease in the national income from 1948 to 1952. Three fourths of the income was said to be devoted to consumption and one fourth to capital accumulation.4

Bibliography 1. Kiril I. Popoff, Stopanska Bolgaria (Eco­ nomic Bulgaria), Sofia, 1915, French trans­ lation under the title La Bulgarie Economique, 1879-1911, Sofia, 1920. 2. P. Kiranoff, National Income of Bulgaria, 1929-1932, Bulletin of the Central Statis­ tical Office of Bulgaria, 1933, No. 1, and article in Stopanski Izvestiva, No. 12, 1934 (in Bulgarian); National Income of Bul­ garia, 1934, 1944, and 1945, Central Sta­ tistical Office of Bulgaria, 1945 (in Bul­ garian). 3. Assen Tchakaloff, Estimate of National Income of Bulgaria, Bulletin of the Bul­ garian Economic Society, 1929, No. 9 (in Bulgarian); The National Income of Bul­ garia, 1924-1935, Statistical Institute for Economic Research, University of Sofia, 1937 (in Bulgarian); National Income and

478

THE INCOME OF NATIONS

Outlay of Bulgaria, 1924-1945, Sofia, 1946 (in Bulgarian). 4. La Bulgarie Nouvelle; or Free Bulgaria, Sofia, January, 1949; February, 1951. 5. Economic Development in S.E. Europe, including Poland, Czechoslovakia, AustriaHungary, Roumania, Yugoslavia, Bulgaria, and Greece, London, Oxford University Press, 1945. 6. Bulgarian Government, Details of 1951 Economic Plan (in English), Sofia, 1950. 7. Khristo Popov, Narodniva Dokhod, etc. (Concept, Methodology, and Calculation of National Income, in Bulgarian), Sofia, 1951. 8. U.N. Economic Commn. for Europe, The Economic Survey of Europe Since the War, Geneva, 1953; also 1954, 1955, and 1956 issues. 9. Plans for the Development of the Econo­ mies of the Peopled Republics, Collection of Materials. Isdat. Inostr. Literatury, Moscow, 1952, Part II (in Russian). 10. V. N. Starodubrovskaia, Building the Eco­ nomic Foundation of Socialism in Bulgaria (in'Russian), Academy of Sciences, Mos­ cow, 1953. 11. E. F. Jackson, “Social Accounting in East­ ern Europe,” International Association for Research in Income and Wealth, Income and Wealth, Series IV, Cambridge, 1955. 12. Statisticki Godisnjak 1957 (Statistical Year­ book, 1957).

14. R um an ia No comprehensive estimate of the na­ tional income was prepared in Rumania prior to 1945, but several private estimates were prepared by non-Rumanian statisti­ cians. In 1930, the Dresdner Bank, in its compilation of economic statistics and na­ tional income data for 39 countries, gave an estimate of Rumania’s national income (at 4 billion Reichsmarks for 1913-14 and 5 billion prewar marks for 1928), with­ out, however, indicating the sources of these figures. Von Waizner, the Austrian expert, estimated Rumania’s national income, exclu­ sive of that of Bessarabia, at 276 million crowns for 1911-13. Colin Clark in his Con­ ditions of Economic Progress gave his own estimate for that country, based largely on Ernst von Waizner’s- calculations. Finally, C. Evelpidi, in his study of the national in­ come of the Balkan countries, published in 1940, reproduced some figures for Rumania. Two British economic studies relating to the Balkans, one published in 1939 and the other in 1945, also gave some figures for Rumania, which likewise were not firmly established.

After the war, with the adoption of the Communist regime in Rumania, national in­ come estimating was taken up by the Cen­ tral Statistical Office and was made a part of the government’s long-range economic planning first inaugurated in 1948; but little was published about it. The Economic Com­ mission for Europe gives the figures of the total national income for 1948 and 1950, but without any breakdowns. The book by V. A. Karra on the Rumanian economy, published in Moscow in 1953, gives one paragraph on the growth of the country’s national income from 1948 to 1952, but gives no breakdowns of it or any other de­ tails. It states, however (page 208), that the national income more than doubled from 1948 to 1952 and that 70 per cent of it is devoted to consumption and 30 per cent to capital accumulation.

Bibliography 1. Dresdner Bank, The Economic Forces of the World (in English), Berlin, 1930. 2. Ernst von Waizner, “National Income of Old Austria and Its Distribution Among the Succession States” (in German), Met­ ron, vol. VII, 1926-1927, No. 4. 3. Colin Clark, Conditions of Economic Progress, London, 1940, pp. 131-32, 175. 4. C. Evelpidi, “National Incomes of the Balkan Countries” (in French), Metron, Rome, 1950. 5. London and Cambridge Economic Series, South Eastern Europe, a Political and Eco­ nomic Survey, Memo. No. 48, London, 1939. 6. Economic Development in S.E. Europe, Oxford University Press, London, 1945. 7. Government of Romania, Planning of the Roumanian Economy (in Russian), Isdat. Inostr. Literatury, Moscow, 1951. 8. V. A. Karra, Building a Socialist Economy in the Romanian Peoples’ Republic (in Russian), Academy of Sciences, Moscow, 1953, p. 208. 9. Plans for the Development of the Econo­ mies of Peoples’ Republics. Isdat. Inostr. Literatury, Moscow, 1952, Part V. 10. U.N. Economic Commn. for Europe, The Economic Survey of Europe Since the War, Geneva, 1953; also 1954, 1955, and 1956 issues.

15. Albania There is no record of any systematic preparation of national income estimates in Albania before her adoption of the Com­ munist system after World War II. Since then, however, national income has been

34.

DEVELOPMENTS IN EIGHTEEN EUROPEAN COUNTRIES

estimated by the Central Statistical Office, but little information about it has been made available. The report issued by the govern­ ment on May 21, 1952 (no. 1469) on the execution of the two-year plan, 1949-50, and of the first year of operation of the fiveyear plan, 1951-55, stated that the income of the population, in prices of 1947, in­ creased from 1948 to 1950 by 16.8 per cent.

Bibliography 1. Plans for the Development of the Economies of the Peoples’ Democracies, Isdat. Inostr. Literatury (in Russian), Moscow, 1953, Part VI, p. 416.

16. Spain The first national income estimate pre­ pared in Spain was by Jose A. Vandellos, published in 1925. His basic estimate was for 1914, for which more statistics were available than for any of the immediate postwar years. This basic estimate was pro­ jected by him to the year 1923 with the aid of special indexes. Vandellos followed the comprehensive concept of national income, defining it as the sum of all factor incomes both in money and in kind. He included interest on unproductive national debt, but excluded profits of public enterprises and deducted taxes. Using mainly the net output method, he relied on census data, agricultural statis­ tics, budget reports, railroad accounts, and his personal knowledge of situations. He estimated the home-produced income by di­ visions of economic activity and adjusted it for the net income from abroad. The ex­ trapolation to 1923 was made with the aid of indexes of the growth of the gross value of agricultural, industrial, and other produc­ tion, on the assumption that the net-gross ratios remained unchanged during the pe­ riod. Only sparse explanations were given of the sources and methods used in the cal­ culations. The reliability of the estimates was low. During the turbulent decade of the 1930’s, no estimates of the Spanish national income were made, but under the Franco regime, in the early 1940’s, the preparation of such estimates was undertaken by the govern­ ment. The National Economic Council, the government’s planning agency, was charged with the task. The first estimate by this agency, published in 1934, was for the years 1929-42. Owing to the continued inade­

479

quacy of the Spanish statistics, accentuated by inflation, the estimates had to be made mainly by indirect methods. Starting with the figures of the gross and net values of agricultural, industrial, mining, and other production for the year 1929, for which the best statistics were available, the estimators developed indexes of changes in the volumes of these outputs and in their prices since 1929, and by applying them to the 1929 base figures, obtained estimates of the gross and net outputs for each division for the succeeding years. In its later reports, the National Eco­ nomic Council carried forward its original series to the years 1943-54, supplementing the indirect estimates with direct data. The estimates were admitted to be only very approximate.

Bibliography 1. Jose A. Vandellos, “Wealth and Income of the Iberian Peninsula” (in French), Metron, vol. V, No. 5, 1925. 2. A. G. Orbaneja, “Evaluation of the Spanish National Income” (in Spanish), Moneda y Credito, Revista de Economia, Madrid, De­ cember, 1943. 3. Spain, Consejo de Economia Nacional, La Renta Nacional de Espaha (Spanish Na­ tional Income), Madrid, 1945; 1952; and volumes under the same title issued in 195355. 4. “Information on the Methods of Estimating National Incomes in Spain” (in Spanish), Revista de Economia Politico, May-August, 1955, Madrid.

17. Portugal Before World War II only private esti­ mates of the Portuguese national income were made. They were reviewed in 1949 by A. D. Guerreiro. The first official estimate was made in 1951 for the year 1949 by the National Institute of Statistics, which fol­ lowed this with others in later years. The estimates became a part of the documents transmitted annually to the Parliament by the Ministry of Finance together with the national budget and the reports on the na­ tional economic development and invest­ ment plans. The estimates deal largely with the national income and gross national product, broken down by industrial origin and by type of final expenditure. The 1952 report transmitted to the Parliament gave figures for the year 1938 and 1947-51; and the later reports carried these figures for­ ward to other years.

480

THE INCOME OF NATIONS

Bibliography 1. A. D. Guerreiro, “National Income Estima­ tion” (in Portuguese), Revista do Centro de Estudos Economicos, May, 1949, Lisbon. 2. A. Correia, Estudos de Economia Applicada o Problema Economico Nacional, Lisbon, 1950; Nota Sobre o Rendimento Nacional, Lisbon, 1951. 3. A. Castro, “Analysis of the First Official National Income Estimate” (in Portuguese), Revista de Economia, December, 1951. 4. Ministry of Finance and National Institute of Statistics, “Provisional Estimate of the National Income” (in Portuguese), Diario das Sessoes, December, 1951, November 29, 1952, and December 10, 1953.

18. Greece According to P. Rediadis, the first rough estimate of the Greek national income was prepared by S. Skiadas in 1891 and was followed, in the 1920’s, by estimates by the Greek Ministry of Finance for the years 1918, 1927, and 1929, by those of A. Angelopoulos and X. Zolotas for 1927, and by one by S. Sakalis for 1929. Rediadis, after publishing several theo­ retical and methodological articles on the subject in the 1920’s, presented his own estimate in 1930 for the year 1929. He fol­ lowed the then standard concept of national income as the aggregate of the net outputs of the different divisions of economic ac­ tivity, including services and net income from abroad, which he calculated as the sum of the net income from capital from abroad and emigrant remittances. Although relying for the greater part of his estimates on available production, tax, and financial data, he also used, where home data were unavailable, known information for other countries whose situation in the particular respects were deemed by him to be similar to Greece’s. Rediadis checked the accuracy of his computations with the aid of an independent rough estimate of the country’s consump­ tion and saving. He estimated first of all the value of domestically produced raw foods, adjusting it for exports; and increased the resulting total by 20 per cent to cover the costs of processing and preparation of these foods so as to obtain the total outlay for food consumption. Assuming the expendi­ ture for food to be 53 per cent of the total consumption expenditure of the people (a ratio obtained apparently from an examina­ tion of the statistics of other countries), he

obtained a figure supposed to represent the total consumption expenditure of the popu­ lation at market prices. To this he added 14 per cent for saving. The resulting figure of national expenditure checked out closely with the previously obtained figure of the national income by the net output method. In 1937 and 1939, C. Evelpidi, using better statistics than those that were avail­ able to Rediadis, published national income estimates for 1927-36. During World War II, when the country was conquered by the German army, the Greek Ministry of Finance prepared a comprehensive estimate of the country’s current national income. It was addressed to the German government in an effort to prove to it that the country, impoverished as it was by the war and by runaway in­ flation, could not possibly pay the costs of occupation with which Germany had as­ sessed it. The estimate was embodied in a lengthy memorandum, written in German, by S. Gotzamanis, Minister of Finance, and addressed to the German Ambassador in Athens for transmission to the authori­ ties in Berlin. The estimate was prepared for the year 1941 by the net output method, with breakdowns of the national income by industrial divisions, and with comparisons, item by item, with the figures for 1939. Dis­ tinction was made between “total income” inclusive of salaries of public employees and of incomes from personal services and pro­ fessions, and “real income” derived from material production only. The figures for 1941 for these two aggregates were shown to be only 38 billion drachma and 23.5 bil­ lion drachma as against 63 billions and 51 billions in 1939, a drop of from 40 to 60 per cent. The minister tried to demonstrate with his figures the advisability of Ger­ many’s treating the country gently, and helping the Greek government to restore industry and trade and to put the thousands of unemployed people back to work. During the war, C. Evelpidi made some further estimates of the Greek national in­ come, publishing one in 1944 for 1939. After the war, continuing his researches, he published, in collaboration with Anastassiades, an estimate of rural income for the years 1927-48. After the war, the government set up a special committee for the investigation of the national income and investment. In 1949, this committee published an estimate of the national income by industrial origin

35.

481

LATIN AMERICA AND THE CARIBBEAN

for 1947 and 1948. In the 1950’s, the offi­ cial estimates were expanded to cover the gross national product and expenditure for the period beginning 1946, with breakdowns by types of consumption and capital invest­ ment. The National Statistical Service of Greece has been publishing its national in­ come estimates annually ever since.

Bibliography 1. S. Skiadas, He Hellas en Europe, Athens, 1891. 2. Perilepsis tes Keimeves Amesou Phorologias, Athens, 1919. 3. P. Rediadis, “New Method for Estimating the National Income” (in Italian), Giornale Degli Economisti, 1921; and “The Greek National Income and Wealth in 1929,” Metron, 1930, No. 4. 4. X. Zolotas, Nomismatike Statheropoiesis, Athens, 1929. 5. Greek Ministry of Finance, S. Gotzamanis, Minister. Memorandum on the Economic Situation of Greece (in German). Mimeo­ graphed Manuscript, September, 1941. 6. C. Evelpidi, The National Income, Athens, 1937; “The National Income of Greece, 1927-1936,” chapter in his Theory and Practice of Agricultural Policy and Eco­ nomics, Athens, 1939; “National Income of Greece, 1939,” in his Agriculture of

7.

8.

9. 10. 11. 12. 13. 14.

Greece, Athens, 1944; “The National In­ come of Greece and Its Composition,” In­ ternational Association for Research in In­ come and Wealth, September, 1949; with Anastassiades, “The Rural Income,” Agri­ cultural Economy, 1945-48 (in Greek). A. Kyskilitsis, “The National Income: Theoretical Problems of Concepts and Methods,” Argyris Papazisis, Athens, 1948 (in Greek). Greece, Committee for Investigation of National Income and Investment, The Na­ tional Income of Greece, 1947, Athens, 1949. Greece, State Secretariat of Public Infor­ mation, “The National Income for 1948,” Bulletin, No. 188, March 9, 1949. Greece, High Board of Reconstruction, A.S.A. National Income and Investment of Greece, 1945-49, April, 1950. Greece, Ministry of Co-ordination, Pri­ vate Consumption in Greece, 1950-1951, Athens, 1955. Greece, National Statistical Service, Sta­ tistical Summary of Greece, 1954. Validis, “Agricultural Income and Recon­ struction,” Bulletin, No. 34, Ministry of Reconstruction, 1948. Demos Politis, Production, Productivity and National Income, Athens, 1954 (in Greek).

35. DEVELOPMENTS IN LATIN AMERICA AND THE CARIBBEAN Despite their close relations with the United States and, in some cases, with Great Britain, the Latin American and Caribbean countries have been slow to take up national income estimation and analysis, just as they have been slow in developing statistics gen­ erally. Before World War II, only three of them—Mexico, Chile, and Peru—had na­ tional income estimates, of which two were privately developed. During the war, two more, Puerto Rico and Venezuela, joined the group, likewise with private estimates. During the postwar decade, however, the number grew rapidly, being actively aided in their development by the United Nations and the International Bank for Reconstruc­ tion and Development, as well as by the United States government. By 1955, at least

twenty-two countries in this part of the world, including Brazil and the British West Indies (reviewed in Chapters 31 and 32), had national income estimates, and every one of them was official in nature. The twenty countries to be treated in this chap­ ter are:

1. Cuba 2. Puerto Rico 3. The Dominican Republic 4. Mexico 5. Guatemala 6. Nicaragua 7. Honduras 8. El Salvador 9. Costa Rica 10. Panam a

11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

British Guiana Colombia Venezuela Ecuador Bolivia Paraguay Uruguay Argentina Peru Chile

482

THE INCOME OF NATIONS

The estimates of these twenty countries show, in varying degrees, the underdevel­ oped state of their economies and also of their statistics. At the same time, their esti­ mates also show the influence of the ad­ vanced concepts and estimating techniques developed in the more advanced countries, inasmuch as the foreign experts who have helped to set up these estimates sought to make them accord with the international standards to the limits permitted by the statistics of these countries. Finally, these estimates also show the influence of eco­ nomic planning as developed during the past decade or two in a number of these countries, as they are linked to these plans wherever such are in operation, and in some countries are even prepared by the planning agencies themselves.

and gross and net capital investment. They are being made more informative from year to year.

1. Cuba

2. Puerto Rico

The first official estimate of the national income for Cuba was prepared by the Na­ tional Bank of Cuba in 1949-50. It showed the national income for the years 1903-48, broken down by distributive shares. The estimate was incomplete, however, as it omitted incomes from professions and do­ mestic services; and was otherwise tentative in nature. In 1950, the Economic and Technical Mission to Cuba, organized by the Interna­ tional Bank for Reconstruction and Devel­ opment in co-operation with the government of Cuba, prepared an estimate for the years 1945-49 which was much more complete. In 1952, the National Economic Council, in a survey of the economic development of Cuba from 1899 to 1950, gave estimates of the growth of the country’s national in­ come from 1937 to 1950. In the meanwhile, the National Bank continued preparation of annual national income estimates and, profiting from the work of the Technical Commission, im­ proved its estimates. The latest of these, at the moment of this writing, is the Bank’s report of 1955. The Bank’s estimates are prepared mainly by the income-distributed method and show the national income by distributive shares in an annual series be­ ginning 1946. The shares are classified as compensation of employees; income from unincorporated enterprises; rent and in­ terest; and income of corporations, includ­ ing direct taxes on corporations. The esti­ mates also show the gross national product,

The first estimate of the income of Puerto Rico was prepared in 1943 for the years 1930-40 by Dudley Smith for the Sugar Producers’ Association of Puerto Rico. The second estimate, or set of estimates, was prepared in 1947-49, under the auspices of the University of Puerto Rico, by Dr. Daniel Creamer, formerly of the National Income Division of the U.S. Department of Commerce, and by his wife, Henrietta Creamer. One estimate, by Dr. Creamer alone, dealt with the net national income at factor cost for 1940-44. The second and third, prepared by the two authors jointly, dealt with the gross national product in 1940-44, and in 1944-46. The estimate of net income aimed at the measurement of the capital and labor income originating in the various branches of Puerto Rican econ­ omy, including its transactions with the rest of the world, and accruing to the residents of the islands. The estimates dealing with the gross national product analyzed the lat­ ter by types of expenditure for consump­ tion and investment. Conceptually, the esti­ mates conformed closely to those of the United States, but owing to less abundant statistics, they were not as detailed. In ad­ dition to providing figures in current money values, the estimates also gave them in 1940 prices. One of the tasks of the investigators was to assist the Puerto Rican governmental agencies in the better organization of their economic statistics and in the setting up of

Bibliography 1. Banco Nacional de Cuba, A. Y. Urosa, Caracteristicas Fundamentales de la Econo­ mia Cubana, Havana, 1950; Memoria (Re­ port) 1949-1950, 1950-1951, and those is­ sued in succeeding years, including one for 1954-1955 issued in 1955; also a paper by C. E. Fargas published by the Bank, de­ scribing the estimating methods. 2. International Bank for Reconstruction and Development, Report on Cuba: Appendix on National Income, Washington, D. C., 1950. 3. G. Guttierez, El Desarollo Econdmico de Cuba, Republica de Cuba, Estudios e Investigaciones Economicos, Junta Nacional de Economia, Havana, 1952.

35.

LATIN AMERICA AND THE CARIBBEAN

an organization that would be able to pre­ pare annual estimates of Puerto Rican in­ come and product. This task they performed successfully. In 1949, the Division of Sta­ tistics of the Puerto Rican Bureau of the Budget published an estimate of Puerto Rican income for the fiscal years 1939-40 to 1945-46, which was the first official ex­ tension of the Creamer study. It presented the income by distributive shares as well as by industrial origin and in both current and constant prices. Eventually, the Bureau of Economics and Statistics of the Puerto Rico Planning Board took over the official preparation of esti­ mates of both the net income and the gross product. With the aid of new statistics it completely revised the earlier series from 1947 on. Its 1955 publication gave, for the fiscal years 1950-54, the net income by industrial origin and by distributive shares, personal income and its disposition, the dis­ position of the gross product and imports by types of expenditure. The classifica­ tions of the industrial divisions, distributive shares, and types of expenditure closely re­ sembled those used in the United States estimates.

Bibliography 1. Dudley Smith, Puerto Rico’s Income, Sugar Producers’ Association of Puerto Rico, Washington, D. C., 1943. 2. Daniel Creamer, The Net Income of Puerto Rican Economy, 1940-44, University of Puerto Rico, Rio Piedras, Puerto Rico, 1947; with Henrietta Creamer, Gross Product of Puerto Rico, 1940-44, University of Puerto Rico, Rio Piedras, 1948; and Gross Product of Puerto Rico, 1944-46, ibid., 1949. 3. Puerto Rico, Bureau of the Budget, The Net Income of Puerto Rican Economy, Fiscal Years 1939—40 to 1945—46, San Juan, 1949. 4. Puerto Rico, Bureau of Economics and Sta­ tistics, Puerto Rico Planning Board, Net Income and Gross Product 1940 and 194755, San Juan, 1956.

3. The Dominican Republic The first estimate of the national income of the Dominican Republic was prepared by the Brookings Institution in 1942 in its report on the Refugee Settlement in that country. The estimate, for the year 1940, was prepared by the net output method, and gave the distribution of the national income by industrial origin. In 1947, the Direccion General de Estadisticas made

483

public an estimate for 1940-46 with a similar breakdown of the national income by industrial origin, but with considerably higher figures than those of the Brookings Institution. In 1952-53, the Central Bank of the Republic published estimates of the gross national product by types of expendi­ tures for the years 1950 and 1951 which were prepared by Dr. Walter Krause.

Bibliography 1. Banco Central de la Republica Dominicana, Producto o Gasto Nacional 1950 (National Product or Income), May, 1952; and Boletin Mensual, January, 1953 (for 1951 figures).

4. Mexico The preparation of national income esti­ mates in Mexico began in the late 1930’s, and from the very start was the concern, if not of the Central Statistical Office, at least of the director in his private capacity as a student. The first estimate was published in 1939, for the year 1929, and was based on the 1930 agricultural and industrial cen­ suses. The second was prepared in 1942-43 by Emilio A. Patino, director of the office at the time, related to the year 1940, and was based on the preliminary and still un­ completed results of the 1940 censuses. The third, by Dr. Josue Saenz, the next director of the statistical office, was for the period 1929-45, and was based on the censuses of 1930, 1935, 1940, and 1945. All three estimates were prepared by the net out­ put method, dealt with the home-produced income only, making no adjustments for the international balance of payments, and showed the national income by divisions of economic activity without any breakdown by distributive shares. After 1945, and particularly following the 1950 census, the scope of the estimates was broadened considerably. They now in­ cluded gross national product by types of product or expenditure as well as net na­ tional income by industrial origin and dis­ tributive shares. The breakdown by indus­ trial origin was given from the year 1929 on, the other two breakdowns from 1939 on. The larger aggregates were given in both current and constant prices. The Bank of Mexico took active part in the preparation of the postwar estimates, and critical and interpretative articles by individual scholars began to appear.

484

THE INCOME OF NATIONS

Bibliography 1. Direction General de Estadistica, Mexico en Cifras, 1938 and subsequent years (in Spanish). 2. F. Bach, “The Distribution of Wealth in Mexico,” Annals of the American Academy of Political and Social Science, vol. 208, March, 1940 (in English). 3. Emilio A. Patino, “The Wealth of Mexico,” Estadistica, Vol. I, No. 1, March, 1943, p. 44. See also supplementary notes appear­ ing in the subsequent issue (June, 1943), pp. 141 ff. (in Spanish). 4. Josue Saenz, “Control of Prices in a War Economy,” Revista de Economia, Febru­ ary, 1943 ( in Spanish); “Net National In­ come of Mexico, 1929-1945,” February, 1945, ibid. (in Spanish). 5. Banco de Mexico, “Estimates of National Income in Mexico,” Memoria de la Primera Reunion de Tecnicos Sobre Problemas de Banca Central, Mexico, 1946; also annual reports of the Bank for 1951, 1952, 1953, etc. (in Spanish). 6. Mexico, Problemas Agricolas e Industrials de Mexico (quarterly publication), articles on national income, July-September, 1952, and January-March, 1953 (in Spanish). 7. Mexico, Informe Anual, Nacional Financiera, Annual reports (in Spanish). 8. M. A. Padilla, “Methodology of National Income,” Revista Hacienda y Finanzas, June, 1947 (in Spanish). 9. G. K. Romero, “Statistics and Calculations of National Income and Product,” Revista de Economia, July 15, 1948 (in Spanish). 10. H. G. Aubrey, “The National Income of Mexico,” Estadistica, Journal of the InterAmerican Statistical Association, lune, 1950. 11. R. M. Ortego, “Adjustment of National Income and of the Population Census 1940,” Revista de Economia, August, 1950 (in Spanish). 12. S. Mujica Montoya, “Effects of the Gen­ eral Characteristics of Industry on National Income,” ibid., November, 1954 (in Span­ ish). 13. Adolph Sturmthal, “Economic Develop­ ment, Income Distribution and Capital Formation in Mexico,” Journal of Political Economy (Chicago), June, 1955.

5. Guatemala The first official estimate of the national income for Guatemala was prepared by the Bank of Guatemala in 1951 and was for the year 1949-50. It followed the United Na­ tions and United States patterns very closely and was rather comprehensive. It showed the gross national product by industrial origin and also by types of final expendi­

ture and was presented in the form of social accounts. Subsequent estimates by the Bank developed a historical series by types of expenditure for the years 1938 and 194654, both in current and constant prices. Special studies have also been made of the distribution of income by size of individual incomes.

Bibliography 1. C. J. A. Palacios, lngreso Nacional de Guatemala, Banco de Guatemala, 1951; “Formas de Redistribucion del Ingreso en Guatemala,” El Funestre Economico, Mexico, July-September, 1952. 2. Banco Nacional de Guatemala, El Producto Nacional Bruto en 1953 (Gross National Product), 1954; and Metodos Usados para el Calculo del Producto Nacional (Methods of Estimating the Gross National Product), 1955.

6. Nicaragua In Nicaragua, the Central Statistical Bu­ reau prepared, in 1952, a national income estimate for 1950 and 1951. During the same year, the Economic Mission of the International Bank for Reconstruction and Development, in its report on a long-range development plan for Nicaragua, elaborated upon this estimate, developing a distribution of the national income and a breakdown of its capital investment share.

Bibliography 1. International Bank for Reconstruction and Development, The Economic Development of Nicaragua. Report of a Mission to the Gov­ ernment of Nicaragua, Baltimore, 1952.

7. Honduras The Central Bank of Honduras published, in 1954, an estimate of the gross national product of the country for the years 192552, with various breakdowns, supplementing it in 1956 with social accounts.

Bibliography 1. Banco Central de Honduras, Estadisticas del Producto e Ingreso Nacional, 1925-52; Cuentas Nacionales, 1925-55.

8. El Salvador Dr. H. C. Wallich and J. H. Adler, econo­ mists of the Federal Reserve Bank of New York, in their study of public finance of El Salvador, prepared, in 1951, with the aid of

35.

LATIN AMERICA AND THE CARIBBEAN

the government statistician of El Salvador, the first estimate of the gross national prod­ uct of the country. The estimate related to the years 1939-46 and showed the gross na­ tional product both by industrial origin and by types of final expenditure in both cur­ rent and constant prices. In 1952, the Cen­ tral Statistical Office prepared the first offi­ cial estimate. It covered the years 1946-50.

Bibliography 1. H. C. Wallich and J. H. Adler, Public Fi­ nance in a Developing Country; El Salvador —A Case Study, 1951, Cambridge, Mass., 1952. 2. W. I. Roller, “Some Considerations Regard­ ing the National Income of El Salvador” (in Spanish), Commentarios Estadisticos, Cen­ tral Statistical Office, El Salvador, Novem­ ber, 1952. 3. United Nations, Technical Assistance Pro­ gramme, Medidas Propuestas para Fomentar el Desarrollo Economico, New York, 1954.

9. Costa Rica The first estimate of the national income of Costa Rica was unofficial and was pre­ pared by members of the staff of the Twen­ tieth Century Fund of New York City in 1952 as a part of its study of the economic development of that country. This estimate related to the year 1949. Work on an official estimate, which was to cover the year 1950, was started by the Central Bank of Costa Rica in 1950. It re­ sulted, as of 1955, in the publication of three studies, each dealing with some one particular segment of the national income and laying a part of the foundation for an eventual complete estimate. The first study was published in 1950, and dealt with the transactions of the economy with the rest of the world in 1950. The second, issued in April, 1955, gave the income originating in the government sector and in the agricul­ tural, fishing, and manufacturing sectors of the economy in 1950. The third, issued in December, 1955, dealt with the income originating in trade, transportation, and communications and also gave provisional estimates of the national income by dis­ tributive shares, by industrial origin, and by type of expenditure for final product, for the year 1950. The estimate was pre­ pared in accordance with the national stand­ ards. It was also intended to present a set of sector accounts.

485

Bibliography 1. Stacy May and Others, Costa Rica—A Study in Economic Development, Twentieth Cen­ tury Fund, New York, 1952 (Appendix 1). 2. Banco Central de Costa Rica, Resumen de las Operaciones en Resto del Mundo, June, 1950; Ingreso Originado en el Sector, etc., April, 1955; Ingreso Originado en el Sector, etc., December, 1955; “Ingreso Nacional,” Revista del Banco Central, December, 1955, No. 24.

10. Panama The first estimate of the national income of the Republic of Panama was prepared in 1953 by Dr. H. R. van Olst of the National Income Unit of the United Nations under the auspices and imprint of the United States Technical Assistance Program. It related to the years 1944-52 and was one of the most comprehensive estimates ever prepared for any Latin American country. It gave the fig­ ures of the gross domestic product at factor cost by industrial origin; the national income at factor cost by distributive shares and by types of income paying organizations; and the gross domestic product by types of ex­ penditure. The estimate also provided a complete set of current sector accounts and a capital account.

Bibliography Republic of Panama, Statistical Office, National Income and National Accounts, 1944-1952 (in Spanish), City of Panama, 1953; also published in English by the United Nations Mutual Assistance Program, New York City, 1953.

11. British Guiana The Development Commissioner’s Office of British Guiana prepared, in 1952, an esti­ mate of that territory’s national income by industrial origin for the years 1948-51. An­ other estimate appeared in a volume issued by the Secretariat of the United Nations in 1952 on the Standards of Living in NonSelf-Governing Territories; and in a special report of the International Bank in 1953.

Bibliography 1. United Nations General Assembly, Secre­ tariat, Standards of Living in Non-SelfGoverning Territories, New York, August, 1952. 2. International Bank for Reconstruction and Development, The Economic Development of British Guiana, 1953.

486

THE INCOME OF NATIONS

12. Colombia The first estimates of Colombia’s national income were prepared simultaneously and independently by two organizations in 1949 —the Bank of the Republic and the Eco­ nomic Mission of the International Bank for Reconstruction and Development. The first was for the years 1945-48; the second cov­ ered a longer period, e.g., 1939-48. Both showed the national income at factor cost by industrial origin and by distributive shares and the gross national product by type of final expenditure. The two estimates differed somewhat in their final results. The Bank of the Republic continued to publish its estimates annually, the latest to date hav­ ing been published in 1955, covering the years 1945-52. The Colombian estimates are among the most developed in Latin America.

Bibliography 1. Banco de la Republica de Colombia (M. Fadul), Ingreso y Producto Nacional 194548, Bogota, 1950; the latest issue to date, for 1945-52, Bogota, 1955. 2. International Bank for Reconstruction and Development, Economic Mission, Report on a Development Program for Colombia, Ap­ pendix (prepared by Jacques Torfs) on Na­ tional Accounts, National Income and Capi­ tal Formation, 1950. 3. J. R. D. Derksen, “New Estimates of Na­ tional Income for Colombia” (in Spanish), Revista del Banco de la Republica, January, 1951.

13. Venezuela Several private estimates of the national in­ come and wealth of Venezuela were pub­ lished in the early 1940’s. One, by Ford, Bacon, and Davis set the national income for 1940 at 900 million bolivares (at the prices of that time); another by Egana y Marquez placed it for 1942 at 869.5 million; and the third, by A. Uslar Pietri, set it for 1943 at 1,244 million. The Central Bank prepared the first official national income estimate in 1949. It related to the year 1936 and was fairly comprehensive. It placed the national income for 1936 at prewar priccs, at 1,500 million bolivares; and it was ac­ companied by a preliminary and less elabo­ rate estimate for 1949. During the succeed­ ing years, the Bank continued to prepare like estimates. In its 1954 report, it presented estimates for 1952 and 1953 that were much more thorough than their predecessors. The

Corporation de Fomento (the official eco­ nomic development agency) issued, in 1952, a collection of studies by different scholars on the national income and wealth of the country for the period 1936-51. The Cen­ tral Bank’s estimate for 1951 was 7,195 mil­ lion bolivares at current prices and 3,340 million at 1936 prices, thus indicating the degree of the rise in prices since prewar times.

Bibliography 1. Ford, Bacon, and Davis, Venezuela en la Encrucijada (Venezuela at the Cross Roads), Caracas, 1940. 2. A. U. Pietri, Sumaria de Economia Venezolana (Summary of the Venezuelan Econ­ omy), University of Venezuela, Caracas, 1945. 3. Central Bank of Venezuela, “Some Data on the National Income of Venezuela in 1936” (in Spanish), Bolet'm del Banco Central, March-April, 1948; Ingreso Nacional de Venezuela (The National Income), Caracas, 1949; Memoria (annual report), 1948, 1949, and subsequent issues. 4. Corporation Venezolana de Fomento, “La Riqueza y el Ingreso Nacional de Vene­ zuela” (National Wealth and Income), Cuadernos de Informacion Economica, JulyAugust, 1952; see also 1954 and 1955 issues.

14. Ecuador In 1952, the Institute of Economic Studies of the University of Ecuador published a study by A. A. Ortiz, a member of the uni­ versity’s faculty, of the concepts and meth­ ods of estimating national income for a country such as Ecuador. This study was based on the existing United Nations and American and British publications and was intended to be used as a textbook in per­ tinent courses at the university. It contained, however, no estimate of the country’s na­ tional income and was purely theoretical in nature. In 1954, however, Dr. H. R. van Olst, an expert with the Technical Assistance Ad­ ministration of the United Nations, prepared a comprehensive estimate of the country’s national income for the Ecuadorian govern­ ment for the years 1950—53. This estimate was prepared in accordance with interna­ tional standards. It showed the gross national product at factor cost by industrial origin, net national income by distributive shares, and gross domestic product by types of final expenditure.

35.

LATIN AMERICA AND THE CARIBBEAN

Bibliography 1. A. A. Ortiz, Estudio de la Renta Nacional, Instituto de Investigaciones Economicas de la Facultad de Ciencias Economicas de la Universidad Central del Ecuador, Quito, 1952. 2. H. R. van Olst, El Ingreso Nacional y Las Cuentas Nacionales de la Republica del Ecuador: ahos 1950-1953. Prepared for the Government of Ecuador by the Technical Assistance Administration of the United Na­ tions, Quito, October, 1954.

15. Bolivia In 1945, L. M. Dominguez, in his study of national income estimates for Latin American countries, conducted under the auspices of the Inter-American Statistical Institute, presented a rough estimate of the national income of Bolivia for 1940, placing it at 7,350 million bolivianos. Labor income was estimated at 60 per cent of the total. In 1948, the Banca Central of Bolivia pub­ lished an estimate of the gross value of the agricultural, mining, manufacturing, and pe­ troleum production (10,602 million bolivi­ anos), excluding trade, transportation, gov­ ernment services, and other branches of pro­ duction. In 1950, the Statistical Office of the United Nations published an estimate of the per capita national income (£ .7 3 ). In 1951 the United Nations Economic Mission to Bolivia, as a part of its inquiry into a longrange development plan for that country, published an estimate of national income and national saving, prepared for it by the National Income Unit of the United Nations. In 1952, a private study of Bolivia’s national income was prepared by Marco D. Pollner, member of the stalf of the Department of Economic Affairs of the United Nations. It related to the years 1948-50 and dealt with the home-produced income at factor cost. The aggregate was broken down both by industrial divisions and by distributive shares. Up to 1955 no official estimate was published in the country.

Bibliography 1. L. M. Dominguez, “National Income Esti­ mates of Latin American Countries,” Studies in Income and Wealth, vol. 10, National Bureau of Economic Research, New York, 1947. 2. United Nations, Statistical Office, National and Per Capita Incomes, Seventy Countries, 1949, Statistical Papers, Series E, No. 7, New York, 1950.

487

3. United Nations Mission of Technical Assist­ ance to Bolivia, National Income and Saving of Bolivia, New York, 1951. 4. M. D. Pollner, “Problems of National In­ come Estimation in Bolivia,” New York Uni­ versity, New York, 1952 (Master’s thesis).

16. Paraguay The Central Bank of Paraguay, in co­ operation with the government statistical de­ partment, prepared, in 1948, an estimate of the country’s national income for the year 1946. The Bank followed this in the suc­ ceeding years with the publication of like estimates. The estimates were prepared by methods similar to those used in the official estimates of the United States. The estimates published in 1955 give a breakdown of the national income and gross national product by industrial origin and by types of final expenditure, for the years 1950-54.

Bibliography 1. Banco Central de Paraguay, La Produccidn Total del Paraguay en 1946 y su Valor Netto, by C. Z. Toler and M. Cirovic, Asuncion, 1948. 2. Banco Central de Paraguay, Renta Nacional del Paraguay en 1951, Asuncion, 1953; also 1950-54, Asuncion, 1956.

17. Uruguay In Uruguay, unofficial estimates of the national income were prepared in 1948 by Raul Ochoa for the years 1931-39 and 1943; and in 1949 and 1950 by Julio Giurio, chief of the Direcion de Industrias. They related to the years 1935, 1940, and 194349. No official estimates were on record as late as 1955.

Bibliography 1. Raul Ochoa, “Mediciones de la Renta Na­ cional,” Revista de Economia, Montevideo, October-November, 1948. 2. Julio Giurio, “Una Estimation de la Renta Nacional del Uruguay,” Revista de Eco­ nomia, Montevideo, June-July, 1949, and September, 1950.

18. Argentina Private estimates of national income ap­ peared in Argentina during World War II, but official estimates were not developed until 1946. The first such estimate was pub­ lished during that year by the Central Bank and was for the year 1945. It showed the

488

THE INCOME OF NATIONS

gross national product by industrial divi­ sions, but was not complete. Other estimates by the Bank followed in the succeeding years, but they also were imperfect. Sub­ stantial improvements in the estimates began to be made only after 1953. In 1955, the Ministry of Finance published an estimate which accorded, in part, with international standards. It showed the gross domestic product at factor cost by industrial origin, year by year, for the period 1935-54 and it gave this series in both current and con­ stant prices. But it did not show the net national income by distributive shares, nor the gross national product by type of final expenditure. The statistics of the country are said to be good enough for the develop­ ment of more comprehensive estimates and analyses than those prepared so far.

Bibliography 1. V. A. Krieger, “A Study of the Estimates of National Income” (in Spanish), Ciencias Economicas, November, 1942. 2. “An Estimate of the National Income of Argentina in 1941” (in Spanish), Revista de Economia Argentina, December, 1944. 3. Banco Central de Argentina, La Renta Na­ cional de Argentina (in Spanish), Buenos Aires, 1946; also, annual reports, 1947, 1948, etc. 4. E. M. Florens, “Correction of the United Nations’ National Income Figures for Ar­ gentina,” Hechos e Ideas, Buenos Aires, July, 1952. 5. Argentina, Secretaria de Asuntos Economicos, Producto e Ingreso de la Repiiblica en el Periodo 1935-54, Buenos Aires, 1955.

19. Peru The first estimates of the Peruvian na­ tional income were prepared by the Depart­ ment of Commerce of the Peruvian Ministry of Foreign Relations in 1938, and were ap­ parently for one year only. Shortly there­ after, Romulo Ferrero of the Agricultural Bank of Peru prepared an estimate. In 1942, the Banco Central de Reserva, in co­ operation with the Department of Com­ merce of the above-mentioned Ministry and with other government agencies, began preparation of its own estimates, publishing the results in 1945, 1949, and every year thereafter. At first, the Bank’s estimates dealt only with the national income at factor cost, and were prepared from production data by the net output method and with breakdowns only by divisions of economic activity. But later, especially for the years beginning with

1947, they were prepared from independent sources also by the income-distributed and final expenditure methods, and provided breakdowns of the net national income by distributive shares and of the gross national product by type of expenditure or product. Conceptually and methodologically the Peruvian estimates are comprehensive and detailed.

Bibliography 1. Peru, Ministry of Foreign Affairs, Depart­ ment of Commerce, “National Income of Peru” (in Spanish), Revista del Departamento, No. 9, 1938. 2. Romulo Ferrero, “National Income of Peru” (in Spanish), Mercurio Peruviano, Lima, 1939. 3. Emilio G. Barreto, “National Income of Peru” (in Spanish), Boletin del Banco Cen­ tral de Reserva del Peru, Lima, August, 1945. 4. Banco Central de Reserva del Peru, La Renta Nacional del Peru 1942-47, Lima, 1949 (with English summary); same pub­ lications for later years, including 1942-52, published in 1954.

20. Chile In the years 1935-40, Raul Simon pre­ pared the first national income estimate in Chile, covering the years 1929-34. He de­ fined the national income as the sum of the incomes available in the country for the requisition of goods and services and he estimated it in five alternative ways as: (1) the net value of goods and services pro­ duced in the various divisions of the econ­ omy, estimated largely on the basis of pro­ duction statistics; (2) the aggregate of labor and capital incomes by the income-distrib­ uted method, with the aid of census statis­ tics and financial reports; (3) the value of goods sold to consumers, on the basis of data on wholesale and retail trade turnover; (4) percentage of national wealth; and (5) the value of energy produced. He obtained somewhat different results in each case. In 1945, the Government Corporation for the Encouragement of Production (Corpora­ tion de Fomento de la Produccion) estab­ lished a permanent unit for national income research. In 1946, it published a compre­ hensive report which included estimates for the years 1942 and 1943. The estimates were prepared mainly by the net output method and showed the national income by its in­ dustrial origin. These estimates were next extended from year to year in the succeed­ ing report of the Corporation, and in 1949

36.

THE NEAR, MIDDLE, AND FAR EAST

they were incorporated into a system of social accounts drawn along the lines of those of the official estimates of the United States. Four sectors were recognized, namely, business, government, individuals, and the rest of the world, and their trans­ actions were combined in a consolidated national income, product, and expenditure account and a consolidated capital account. The Chilean national income and ac­ counts estimates were rated by L. M. Do­ minguez, in 1953, as being among the best in Latin America at the time.

Bibliography 1. Raul Simon, Determination of the National Income of Chile, Santiago, 1935 (in Span­ ish); articles on national income in the Anales del Instituto de Ingenieros de Chile,

489

January, February, and March, 1934, Janu­ ary, 1938, and July-August, 1940 (in Span­ ish). 2. S. Fajardo, “Explanation of a New Method for Computing National Income and Its Ap­ plication to Chile,” Segunda Reunion de Tecnicos de la Banca Central del Continente Americano, Santiago, 1949 (in Span­ ish). 3. Loreto M. Dominguez, “The Status of Na­ tional Income Research in Latin America,” International Association for Research in Income and Wealth Conference, AugustSeptember, 1951 (in mimeographed form). 4. Corporacion de Fomento de la Produccion, National Income 1940—43, Santiago, 1946, in two volumes (in Spanish); Memorandum on National Accounts, May, 1949 (in Span­ ish); National Accounts, 1940-52, Santiago, 1954 (in Spanish).

36. DEVELOPMENTS IN THE NEAR, MIDDLE, AND FAR EAST At least twenty-one countries in the Near, Middle, and Far East have national income estimates today (1957) and their number is rapidly increasing. Before 1946, only five, Egypt, Turkey, Palestine (later Israel), India, and Japan, had them, and in all but one of them, Japan, the estimates were pri­ vately prepared. In twelve years, the number of countries having estimates quadrupled, and in nearly every one of them the esti­ mates have become official. In Part III, we have reviewed in considerable detail the estimates for three countries in this group, Turkey, India, and Malaya (Chapters 29, 30, and 32). In this chapter we will review in a more general way those of the follow­ ing fifteen: 1. 2. 3. 4. 5. 6. 7. 8.

Egypt Israel Lebanon Syria Pakistan Ceylon Burma Thailand (Siam)

9. 10. 11. 12. 13. 14. 15.

Indonesia Hong Kong China Japan Korea (South) Taiwan (Formosa) Philippines

The information on the 1956-57 addi­ tions to this group—Iran, Iraq, and Jordan —came too late for inclusion here. See the Notes begin on p. 544.

U.N., Economic Developments in the Mid­ dle East, 1956-57. With the possible exception of Japan, all of these fifteen countries are underdeveloped and lack adequate statistics. Nearly every one of them has received some technical assistance from abroad in developing its na­ tional income estimates. The assistance came from the United Nations’ Technical Assist­ ance Programme or from the United Na­ tions Economic Commission for Asia and the Far East; or from the experts of the International Bank for Reconstruction and Development, or from British, American, or Western European experts lent to their governments for this work. As a result, the estimates in most of these countries have been developed in accordance with modern international standards within the limitations set by their available statistics; and consid­ erable progress is being made in many of them toward developing better population and economic statistics generally. China is a special case. After an interest­ ing start in national income estimation under the nationalist regime, the results of which were widely published, all work done along this line after the Communist regime was established has been veiled in mystery. That

490

THE INCOME OF NATIONS

substantial developments in the field have taken place there cannot be doubted. But an account of them will have to wait until information on the subject becomes avail­ able. The estimates for these fifteen countries have been developed chiefly by the net out­ put method. In most of them, national in­ come is broken down mainly by industrial divisions; but in a substantial number of them it is also broken down by distributive shares or by type of final expenditure, or by both these categories. This is true of Israel, Ceylon, Thailand, Japan, Korea, Hong Kong, Taiwan, and the Philippines. In at least three cases, Israel, Burma, and Japan, even a system of sector accounts has been set up. Inasmuch as a number of these countries—Israel, Ceylon, Burma, Pakistan, and Japan—have embarked on long-range economic development plans during the past decade, their national income estimates have been related to such plans. Several countries in the group—Israel and Japan particularly (not to mention India)—• have developed native scholars of eminence in the field, presaging a vigorous further development of their national income esti­ mates and analyses in the future.

1. Egypt Interest in national income estimates de­ veloped in Egypt earlier than in any other country of the Near East, the first estimate of her national income having been pre­ pared in 1922, the year of the termination of the British wartime protectorate over Egypt. The initiation of this interest was undoubtedly due to British influence. It must be remembered that although Egypt was nominally a tributary state of the Ottoman Empire, it was, during the period 1882— 1922, largely dominated both politically and economically by Britain. The first estimate was prepared by Dr. J. G. Levi of the Egyptian Ministry of Finance. It covered the year 1921-22 and was prepared primarily for the purpose of revealing useful informa­ tion on ways to improve the tax revenue for the treasury of the newly established inde­ pendent kingdom. Levi’s concept of national income was heavily affected by that purpose. He included in his definition of national in­ come all incomes available to the country’s population from domestic and foreign sources, without deducting the incomes flow­ ing abroad. James I. Baxter, English statis­ tician, criticized Levi’s estimate at the time

as an overestimate. Levi was supported in his defense of his estimate by James T. Craig, another British statistician who quoted his own estimate, made privately before World War I, which, after allowing for the intervening changes in money values and other factors, corresponded with Levi’s results. Levi used the income-distributed ap­ proach. In the face of exceedingly scant sta­ tistical data, he employed whatever informa­ tion he could obtain on the average earnings of certain groups of the population, their average consumption, the net value of agri­ cultural production, and the incomes sub­ ject to taxation; and where he could not find any statistics he made assumptions “in all prudence.” After World War II, Dr. A. M. Anis pre­ pared an estimate of the national income of Egypt for 1937-45, as his doctoral thesis at the London School of Economics, pub­ lishing it subsequently in French, English, and Arabic. His estimate was detailed and was supported with extensive explanations of its sources and methods. In 1952, the Census and Statistics Depart­ ment of the Ministry of Finance published the first official national income estimate for Egypt. It was for the year 1948 and was declared to be provisional in nature. The later publications of the same Department made available definitive estimates for the years beginning 1950. Those issued in 1955 provided figures for 1950-53 and embodied two independent estimates: one prepared in considerable detail by the net output method, with breakdowns by industrial origin; and the other, less detailed, by the income-dis­ tributed method.

Bibliography 1. J. G. Levi, “L’Augmentation des Revenus de L’Etat,” L’Egypte Contemporaine, vol. 13, December, 1922, pp. 596-624. 2. James I. Baxter, “Notes on the Estimates of National Income of Egypt for 1921-1922,” L’Egypte Contemporaine, vol. 14, May, 1923, pp. 405-27. 3. J. G. Levi, “Evaluation du Revenu National de L’Egypte, Reponse a M. J. I. Baxter,” L’Egypte Contemporaine, vol. 14, May, 1923, pp. 428-69. 4. James T. Craig, “Notes on the National In­ come of Egypt,” L’Egypte Contemporaine, vol. 15, January, 1924, pp. 1-9. 5. A. M. Anis, The National Income, Output and Expenditure of Egypt, 1937-1945 (in English and Arabic), Cairo, 1950; also pub­ lished in Cairo in French (originally pre­ pared as a doctor’s dissertation at the Lon­ don School of Economics).

36.

THE NEAR, MIDDLE, AND FAR EAST

6. H. Meunier, “How the National Income of Egypt is Evaluated” (in French), La Revue Economique et Financiere, Nos. 40, 42, 43, December, 1951. 7. J. B. D. Derksen, The National Income of Egypt: Some Comments, Ministry of Na­ tional Economy, Cairo, 1951 (mimeographed paper). 8. Egypt, Ministry of Finance and Economy, Census and Statistics Department, Net Na­ tional Income in Egypt, 1948, Cairo, 1952; National Income of Egypt for 1953, Cairo, 1953; Statistical Pocket Yearbook, 1954, and later years.

2. Israel The first estimate of the national income of Israel (or Palestine as it was called at the time) was prepared in 1941 by Ludwig Gruenbaum, then a member of the staff of the Economic Research Institute of the Jew­ ish Agency for Palestine. It related princi­ pally to the year 1936, but also carried fig­ ures for earlier years. It showed the na­ tional income at factor cost as well as the national product by type of final expenditure and it gave separate figures for the Jewish and Arab segments of the economy. The second estimate was prepared in 1943 by G. E. Wood of the Statistical Office of Palestine and was for the years 1939 and 1942. It showed the national income by in­ dustrial origin and also by distributive shares, but made no distinction between the Jewish and Arab parts of the economy. In 1946, P. J. Loftus, also of Palestine’s Statistical Office, continued Wood’s calcu­ lations to the year 1946, reintroducing, however, Gruenbaum’s distinction between the Jewish and Arab economies. Moreover, he deflated the historical series to constant prices. In 1950, Gruenbaum, who became the leading spirit in the formulation of a longrange economic plan for the country, de­ veloped national income estimates for 194649, while at the same time adjusting his own earlier estimates and those of Wood and Loftus in the light of the changed territory of the country and other factors in order to make a consistent historical series. In the preparation of this new series, on which he built his proposed four-year economic de­ velopment plan, he used both social account­ ing and input-output approaches. In 1953, three more studies of national income appeared. One of them was prepared by the Economics Department of the Min­ istry of Finance and showed the national

491

income for 1951 broken down by industrial divisions, by distributive shares, and by final expenditures. The second estimate was pre­ pared by H. Nachshon, likewise for 1951, but was constructed differently. The third was prepared by A. L. Gaathon of the Sta­ tistical Office, and emphasized the social accounts approach. In 1954-55, Daniel Creamer was brought from the United States to help the Statistical Office set up a more definitive set of esti­ mates. Creamer used the net output method and prepared estimates for 1952-54, break­ ing down the total by industrial origin. The work put into the preparation of na­ tional income estimates in Israel has been very extensive. By 1955, the estimates showed the national income for 1952 and 1953 by industrial origin, by distributive shares, and by types of final expenditure. The estimates were used in the preparation of the national budget and of the govern­ ment’s long-range economic plans.

Bibliography 1. Ludwig Gruenbaum, National Income and Outlay in Palestine, 1936 (in Hebrew), Jewish Agency for Palestine, Economic Re­ search Institute, Jerusalem, 1941; Outlines of a Development Plan for Jewish Palestine, same Institute, 1946; “Palestine’s National Balance,” Palestine Economic Review, Jeru­ salem, July, 1947; “Israel’s National Income 1950,” Shivat Zion (in Hebrew), Jerusalem, 1950; Four-Year Development Plan of Israel 1950-53, Jerusalem, 1950 (both in Hebrew and in English). 2. G. E. Wood, Survey of National Income of Palestine, Government Statistician, Jeru­ salem, 1943. 3. P. J. Loftus, National Income of Palestine, 1944, Government Statistician, Jerusalem, 1946. 4. A. L. Gaathon, “National Income and Product in Israel in 1950” (in Hebrew), Ledioth Statistioth, No. 61, 1951. 5. Israel, Ministry of Finance, Economics De­ partment, Economic Studies No. 2 (in He­ brew), Jerusalem, 1952. 6. H. Nachshon, National Income and Wages in 1951 (in Hebrew), Jerusalem, 1952. 7. Israel, Office of Information, “National In­ come, 1950-51,” Israel Digest, New York, July 19, 1954. 8. Daniel Creamer, Provisional Estimates of Israel’s National Income 1952-53; Harold Lubell, Provisional Estimates of Israel; Na­ tional Expenditure, 1952-53 (Falk Projects for Economic Research), Central Bureau of Statistics, Jerusalem, 1955. 9. Israel Central Bureau of Statistics, Statistical Abstract, 1954-55, Jerusalem, 1955.

492

THE INCOME OF NATIONS

3. Lebanon The first estimate of the national income of Lebanon was made public in 1955 by the Economic Research Institute of the Ameri­ can University at Beirut after more than four years of work. The project was started by the Institute with the support of the United Nations Relief and War Agency in 1950 and was conducted under the direction of Albert Y. Badre, member of the staff of the Institute. National income was to be computed by the net output method from detailed computations of the net output of each of the ten industrial divisions of the economy. Originally designed to cover the years 1948 and 1949, it was eventually ex­ tended to cover, in a consistent series, the years 1948-54. During the first year, esti­ mates of the net output were completed only for agriculture (including fishing) and con­ struction, and these were published in two separate monographs. In 1953 and 1954, the estimates for the remaining divisions were completed and likewise published in separate monographs. Early in 1955, the results were added and the national income totals were made public. Not only were the resulting estimates well developed within the limits of the net output type of approach, but they established a foundation for the development of other types of estimates and analysis of national income as well.

Bibliography 1. A. Y. Badre and Staff of Economic Re­ search Institute of the American University of Beirut, National Income of Lebanon, Monographs Nos. 1-10, Beirut, 1952-55. 2. United Nations Social and Economic Coun­ cil, Implementation of Full Employment and Balance of Payments Policies, New York City, June, 1955.

of requisite data. The United Nations Sup­ plement to the World Economic Report, 1953-54, however, presented some over-all figures of the national income of the country for 1950 and 1953. Considerable statistical work remains to be done before a compre­ hensive estimate for Syria can be developed.

Bibliography 1. “Syria: National Income” (in French), Le Commerce du Levant, Aug., 1952, Beirut. 2. International Bank for Reconstruction and Development, Technical Economic Commis­ sion, The Economic Development of Syria, Baltimore, 1955. 3. United Nations, Economic Developments in the Middle East 1945-1954, Supplement to World Economic Report, 1953-54, New York, 1955.

5. Pakistan The first estimate of the national income of Pakistan was made in 1948, prior to partition, as a part of the estimate of the national income of India, and was an un­ official estimate. Pakistan’s share of India’s national income was calculated largely on the basis of consumption data. The first official estimate of the national income of independent Pakistan was prepared in 1952 by the Pakistan Ministry of Economic Af­ fairs as a part of its report on a six-year development program for the country. The estimate was prepared by the net output method and showed the gross national prod­ uct by industrial origin in 1948-49. The Central Statistical Office continued the preparation of these estimates, and by 1955, a series was available showing the national income by industrial origin for the period 1949-53.

Bibliography

4. Syria In 1952, an unofficial estimate of the na­ tional income of Syria was published in Beirut. It was for the year 1951 and it showed the national income by three major industrial sectors. In 1955, the Economic Mission of the International Bank for Re­ construction and Development, in its report on the economic development of the coun­ try, outlined a procedure for the develop­ ment of a national income estimate for Syria, but abstained from actually develop­ ing such an estimate because of the lack

1. R. C. Desai, “Consumer Expenditure in India 1913-32 to 1940-41,” Journal of the Royal Statistical Society, London, 1948. 2. “National Income of British India and Pak­ istan,” Eastern Economist, December 31, 1948. 3. Pakistan, Ministry of Economic Affairs, SixYear Development Programme of Pakistan, Karachi, 1952. 4. United Nations Economic and Social Coun­ cil, Implementation of Full Employment Policies, New York, June-July, 1952. 5. Pakistan, Central Statistical Office, “National Income Estimation in Pakistan,” Statistical Bulletin, February, 1955, Karachi.

36.

THE NEAR, MIDDLE, AND FAR EAST

6. Ceylon The first comprehensive estimate of the national income of Ceylon was prepared in 1946 by K. Williams, Director of Census and Statistics for the Board of Ministers, in connection with their plan for a longrange economic development for the coun­ try. The second, even more comprehensive estimate was prepared by Dr. B. B. Das Guptas for the Commission on Social Serv­ ices. It dealt with the national income at factor cost and related to the years 1937, 1938, and 1942-44. Thereafter, national in­ come estimates in Ceylon have been pre­ pared annually by the Census and Statistics Department and transmitted to the Parlia­ ment by the Minister of Finance in conjunc­ tion with the budget and with reports on the execution of the economic development plan. The concept used was that of the homeproduced or geographic income of the country rather than that of the national in­ come as the aggregate income of the resi­ dents. In line with the observations made earlier by Phyllis Deane,1 the concept used was considered to be more suitable to a colonial country “where large foreign firms operate . . . or where foreign shareholders have invested significant sums in enterprises of the country.” Although the net output method was used in the computations, no deductions were made for depreciation be­ cause of statistical difficulties. The estimate was presented in two parallel columns. The left-hand column showed the gross national product broken down by types of products, namely domestic exports (tea, rubber, cocoanut products, etc., less im­ ported materials used and less correction for overvaluation); articles produced and consumed locally (rice, cocoanut and cocoanut oil, tea, other food crops, tobacco, live­ stock products, fish, industrial products, both made by factories, which are largely government owned, and cottage-made, etc.); services of trade and transport other than export; professions; personal services, in­ cluding domestics; rent; irrigation work, other land development, and other capital development largely carried on by govern­ ment (less the value of imports of machinery and building materials); services of govern­ ment (other than in capital development); income from abroad; and foreign military expenditure (largely payments to local labor

493

by the British military organization in Cey­ lon). The right-hand column of the account, showing the gross national expenditure at market prices, was broken down into the value of imported consumption goods at landing cost including import duties (in­ cluding rice and sugar imported by the government and distributed at a profit, with the over-all trading deficit being shown as the net subsidy); distribution cost on im­ ports; value of consumption of local prod­ uce; expenditures on personal services in­ cluding domestics; rent; travel (by rail and buses); central and local government ex­ penditures (less cost of public debt, sub­ sidies, pensions, grants to local bodies, rail­ way losses, relief works, and costs of ma­ terials imported by government other than capital goods; plus current expenditures of Government Electrical Department); remit­ tances to foreign countries and expenditures on travel abroad; dividend and interest pay­ ments to abroad (which are subject to local taxation); gross capital formation abroad (governmental, as represented by the loan fund expenditures of the government, and private, represented by building construction and machinery and valued on the basis of the costs of imported building materials and machinery plus costs of domestic labor and other items); net changed assets abroad; less import and excise duties net of subsidies. The totals at the bottom of the two col­ umns balanced with the aid of a slight ad­ justment. The weakest items in the compu­ tation were the value estimates of domesti­ cally produced and consumed articles used in the calculation of the gross national product, and the estimates of distribution costs of imported articles used in the cal­ culation of the gross national expenditures, but a continued improvement in the collec­ tion of statistics was expected to improve the calculation of these items and others and, hence, of the national income totals themselves.

Bibliography 1. Ceylon, Board of Ministers, Post-war Devel­ opment Proposals, Ceylon, 1946. 2. Ceylon, Commission on Social Services, Re­ port, Ceylon, 1947. 3. United Nations, Economic Commission for Asia and the Far East, Economic Survey of Asia and the Far East, 1947, Shanghai, May, 1948. 4. Ceylon, Ministry of Finance, A Six Year Plan for Ceylon, Colombo, 1948; A Full

494

THE INCOME OF NATIONS

Employment Policy, Colombo, July, 1949; Colombo Plan, Colombo, December, 1950. 5. Ceylon, Ministry of Finance, K. Williams, Director of Census and Statistics, National Income of Ceylon, Presented to Parliament, July, 1952; also similar annual reports for previous and subsequent years. 6. K. Williams, “The Estimation of the Na­ tional Income of an Underdeveloped ExportImport Economy,” Bulletin of the Interna­ tional Statistical Institute, vol. XXXIII, Part III, 1951.

7. Burma The first estimate of the national income of Burma was prepared in 1951 by Miss P. Ady, expert for the United Nations Tech­ nical Assistance Program. It was prepared for the Ministry of National Planning and hence was official. It covered the years 1938-39 and 1946-47 to 1949-50 and showed the national product income and expenditure as well as a set of social ac­ counts. These estimates were thereafter car­ ried forward annually by the same ministry and were revised and expanded. By 1955, the ministry’s estimates showed the gross domestic product at market prices by indus­ trial origin for the years 1951-54; and the gross domestic product by type of final ex­ penditure for the years 1938 and 1948-54.

Bibliography 1. Burma, Ministry of National Planning, The National Income of Burma, Rangoon, Sep­ tember, 1951; also annual reports under the same title, August, 1952 and September, 1955 and 1957. 2. United Nations Economic Commission for Asia and the Far East, Analysis of National Income in Selected Asian Countries, Eco­ nomic Bulletin, Bangkok, November, 1952.

8. Thailand The first official estimate of national in­ come of Thailand was prepared by J. S. Gould, Statistical Advisor to the National and Economic Council of that country in 1950. It showed the gross national product by industrial divisions, as well as the net national income at factor cost and at market prices for the years 1938/39 and 1946-48. This estimate was carried forward by Dr. Gould, with some further refinements in the succeeding years. The United Nations Economic Commission for Asia and the Far East, in its several reports and bulletins

issued in 1952-56, also gave data on the national income of Thailand and presented analyses of it. By 1955, the estimates showed the gross domestic product at market prices by industrial origin and by final expenditure for the years 1938 and 1948-53. Only the distribution by factor shares remained un­ developed.

Bibliography 1. Thailand, National Economic Council, Esti­ mates of Gross National Product and Net National Income, 1938/39, 1946-1948; Bangkok, 1950; Preliminary Estimates of the Gross Geographic Product and Domestic National Income, 1938/39 and 1946-1950, Bangkok, 1952. 2. United Nations Economic Commission for Asia and the Far East, Analysis of National Income in Selected Asian Countries, Eco­ nomic Bulletin, Bangkok, 1952; also Eco­ nomic Survey for 1955, Bangkok, February, 1956.

9. Indonesia One of the first estimates of national in­ come for Indonesia was prepared by S. D. Neumark in 1952 for the National Planning Board. It covered the years 1951 and 1952 and provided extensive explanations of the sources and methods used. National income was shown both by industrial origin and by type of expenditure.

Bibliography 1. United Nations Economic Commission for Asia and the Far East, Analysis of National Income in Selected Asian Countries, Eco­ nomic Bulletin, Bangkok, November, 1952. 2. Indonesia, National Planning Board, Na­ tional Income of Indonesia, 1951-52, Report prepared by S. D. Neumark, Economics and Finance in Indonesia, Djakarta, June, 1954.

10. Hong Kong The first estimate of the national income of Hong Kong was prepared in 1954 by R. A. Ma and E. F. Sczepanik of the Uni­ versity of Hong Kong and was for the years 1947-50. It was prepared by the incomedistributed method, and showed the national income by distributive shares. The authors explained that the sparseness of available statistical data did not permit any other estimating approach and breakdown, and that even the estimates by that method and breakdown were none too firmly established.

36.

THE NEAR, MIDDLE, AND FAR EAST

The United Nations Economic Commission for Asia and the Far East in its report for 1955 further elaborated on this estimate, showing not only the national income by distributive shares but also by final expendi­ ture, and giving the figures for the years 1948-52.

Bibliography 1. R. A. Ma and E. F. Sczepanik, National Income of Hong Kong, 1947-50, Hong Kong, March, 1954. 2. United Nations Economic Commission for Asia and the Far East, Economic Survey 1955, Bangkok, February, 1956.

11. China Interest in national income estimates in China developed first under the nationalist regime during World War II, and continued in the immediate postwar years as long as the regime lasted. The resulting estimates were prepared, however, by individual schol­ ars of research institutes and not by the government itself. The first such estimate was prepared in 1941 by Da-Kuen Lieu of the Research Institute of National Economy at Chungking. It related to the period 1931— 36 (preceding the outbreak of the SinoJapanese conflict), for which a reasonable amount of statistics was available. The in­ come-distributed method was followed in its preparation. National income was shown both by distributive shares and by industrial origin. In 1946, the Institute of Social Sci­ ences of Lichung published a study covering the same period, but using the net output approach. In 1947, Mr. Ou and his asso­ ciates at the Academia Sinica at Nanking published a study which, by and large, was more comprehensive than any of the others mentioned. It presented new estimates of national income for the period 1931-36 as well as a preliminary estimate for 1946, and covered not only the national income at factor cost but also the gross national prod­ uct by types of expenditure. It relied on both production statistics, under the net output approach, and family budget data under the final expenditure method. In 1946 and 1947, two young Chinese scholars published in the United States two national income studies, in English. One, by Ta Chung Liu and Shan Kwei Fong, devel­ oped tentative social accounts estimates for the period of the 1930’s; the other, by Mr. Liu alone, embodied an estimate of the gross

495

national product as well as one of the net na­ tional income of China for the same period. It was prepared largely by the net output method. Owing to the dearth of statistics, Liu made bold assumptions in some cases. The reliability of the various parts of his esti­ mates varied considerably. The most interest­ ing part of Liu’s estimate was an attempt, by means of appropriate adjustments, to make the figures for China comparable with those for the United States. To that end he first of all calculated the Chinese products in American prices, which were 47 per cent higher than the Chinese. Secondly, he ad­ justed his figures for the fact that a smaller proportion of agricultural output took the form of cash income in China than in the United States (58 per cent as against 79) and that the proportion of distribution costs in the final sales value was also lower in China (10 per cent as against 36). Accord­ ingly, he raised the gross value of China’s agricultural output proportionately. Thirdly, taking account of the fact that in China a larger proportion of family income was ob­ tained from the unremunerated labor of family members, he raised China’s total in­ come by an additional proportion. All in all, these adjustments doubled the gross na­ tional income figure for China from a per capita of $20.50 to one of $41, comparable with a $540 per capita income for the United States of the same time. With the change in the political regime in 1948, all private research in the coun­ try’s national income came to an end. Esti­ mates of national income must undoubtedly have been prepared by some agencies con­ cerned with economic planning, but if so, they apparently remained unpublished. The five-year economic plan announced in 1953, as published in foreign periodicals, con­ tained no estimate of national income.2 In 1954, Professor Alexander Eckstein of the Massachusetts Institute of Technology published a crude private estimate of China’s national income for 1953 with breakdowns by industrial divisions and types of final expenditure. His calculations were made in American dollars at constant prices. The estimator indicated, however, that these were merely “tentative and crude calcula­ tions . . . intended merely as indications of approximate orders of magnitude rather than as actual estimates.” 3 Dr. Yuan-Li-Wu, a Chinese scholar who had lived in England and the United States,

496

THE INCOME OF NATIONS

prepared, in 1955, a private estimate of China’s national income for 1953.4 He pre­ pared it by projecting Pao-San Ou’s esti­ mate for 1933 with the aid of certain in­ dexes and miscellaneous information. Ou’s estimate was in 1933 prices. Wu also made his original calculations in 1933 prices. He took Ou’s figure of the net value of indus­ trial output in 1933 and projected it to 1953 with the aid of the Communist government’s official index of industrial production. Next, by using the 1933 ratio of the value of in­ dustrial to agricultural production, he ob­ tained a figure for the net value of the combined industrial and agricultural output in 1953. To this he added the values of the services of trade, transportation, govern­ ment, and other sectors of the economy, estimated partly on the basis of current in­ formation and partly on the basis of indexes. The resulting figure, at constant prices, was slightly more than 30 per cent above the 1933 figure. Inasmuch as the population over that twenty-year period had increased 30 per cent (from 450 to 583 millions), Dr. Wu concluded that the government must have kept the 1953 standard of living down to the prewar per capita levels, or even slightly below these levels, through the use of fiscal measures; and that the slight gain in the national income, or curtailment in national consumption, must have been used by the government for national investment through these same fiscal devices. Dr. Wu concluded that the rate of annual net in­ vestment in 1953 was very low. The first official reference to national in­ come seems to have been made in Commu­ nist China in June, 1956. Discussing the operations of the second Five-Year plan (1953-57), the Chairman of the State Planning Commission said that according to preliminary estimates 22 per cent of the na­ tional income was being currently set aside under the plan for capital accumulation. On September 18, in another announcement (re­ ported in the People’s Daily of September 20, 1956), the Chairman gave a corrected lower ratio for the current capital accumu­ lation and also released a set of figures giv­ ing for each year of the 1953-57 period the actual or anticipated percentage of national income devoted to consumption, capital ac­ cumulation, and government revenue. But he did not give the amounts of these aggre­ gates. The authors of the Economic Survey of Asia and the Far East of 1956 (see Ref­

erence No. 9), using the published figures of government revenues of China, attempted to convert the ratios given in the foregoing statement into the sums of the national in­ come. With the aid of population figures, they arrived at the following result:

1953 1954 1955 1956

N ational Income a t M arket Prices (in billion yuan) 74.3 81.7 85.7 94.5

Per Capita Income (in yuan) 127 137 141 154

The second five-year plan proposes a fur­ ther increase of 50 per cent in national in­ come in 1962 over 1957 and a doubling of state investments in capital construction over the previous five-year period (see Ref­ erence 9, Survey, 1957, p. 108).5

Bibliography 1. Da-Kuen Lieu, A Preliminary Estimate of National Income in Pre-War China, Re­ search Institute of National Economy, Chungking, 1941 (in Chinese). 2. Pao-San Ou, “A New Estimate of China’s National Income,” Journal of Political Econ­ omy, vol. LIV, No. 6, 1946; National In­ come of China 1933, 1936, and 1946, Min­ istry of Social Sciences, Academia Sinica, Nanking, 1947; with others, National In­ come of China, 1933, in two volumes, ibid., 1947 (in Chinese). 3. Ta Chung Liu and Shan Kwei Fong, “The Construction of National Income Tables and International Comparisons of National In­ comes,” Studies in Income and Wealth, vol. VII, National Bureau of Economic Research, New York, 1946. 4. Ta Chung Liu, Gross National Product of China, 1931-36, Brookings Institution, Wash­ ington, D. C., 1947. 5. Yu-Ling Pei, “National Income, Consump­ tion, and Investment of China,” Journal of Finance and Industry, Bank of Communica­ tions, Shanghai, May, 1948 (in Chinese). 6. C. H. Chang, “Another Estimate of China’s National Income,” Central Bank Monthly, Shanghai, November, 1948. 7. Alexander Eckstein, “Conditions and Prog­ ress for Economic Growth in Communist China,” World Politics, vol. VII, October, 1954-April, 1955; also Chapter 14 in W. W. Rostow, et al., Prospects for Communist China, New York, 1954. 8. Yuan-Li-Wu, An Economic Survey of Com­ munist China, New York, 1956. 9. United Nations, Economic Survey of Asia and the Far East, 1956; same survey, 1957.

36.

THE NEAR, MIDDLE, AND FAR EAST

12. Japan a. Estimates Before World War II The development of national income es­ timates in Japan can be divided into four periods: (1) the period 1900-20 when the first estimates were prepared by a few pri­ vate scholars, (2) the period 1920-39 when estimates became numerous and some of them were official, (3) the period of World War II when most research in the field was suspended, and (4) the period since the end of the war when the preparation of es­ timates was not only resumed but was estab­ lished on a more solid and broader founda­ tion than ever before. During the first period, at least two na­ tional income estimates are known to have been made: one by K. Nakamura in 1902 for the year 1900, by the net output method; and the other by T. Yamashita in 1906 for 1904 by the final expenditure method. During the second period, interest in the measurement and exploration of the national income was stimulated by the rapid indus­ trialization of the country and growth of Japan’s economic, political, and military power. In 1925, S. Shiomi published esti­ mates by the income-distributed method for the years 1903-19, and in 1931, K. Mori of the Government Bureau of Statistics, us­ ing the same approach but supporting it with better statistics, prepared estimates (which were official) for the period 18871925. Mori’s calculations for 1925 were made from statistics for that year, but those for the other years were made by extrapo­ lating the 1925 data. These earlier esti­ mates were followed during the 1930’s by estimates by Hijikata by the income-distributed method; those by Takahashi by the net output method; estimates by the Mitsu­ bishi Economic Research Bureau which were linked to Hijikata’s; and those by the Japanese Economic Federation which com­ bined Mori’s and Takahashi’s estimates with its own. The estimates for the years sub­ sequent to 1936—that of the outbreak of the Sino-Japanese conflict—were not as good as those for the preceding years, inas­ much as the government substantially cur­ tailed its statistical collections after that year. During World War II, the only new esti­ mates of national income that appeared were those prepared by the Ministry of Finance for the years 1941 and 1942. They were prepared with a view to aiding the govern­

497

ment’s revenue-raising programs and cam­ paigns to stimulate savings, and, therefore, conformed more to the concept of the aggre­ gate private income of the people than to that of the national income proper.

b. Postwar Estimates After the war, during Japan’s occupation by American forces, estimates of the coun­ try’s national income were at first prepared by the American government. In October, 1945, only two months after the cessation of hostilities, the Intelligence Branch of the U.S. State Department prepared an estimate by the income-distributed method, with breakdowns by both divisions of economic activity and distributive shares. In 1946, the U.S. Strategic Bombing Survey made an estimate of Japan’s gross national product for the years 1942-45; and in 1947, a spe­ cial U.S. Statistical Mission surveyed Japan’s statistics and past methods of estimating national income and made extensive recom­ mendations for their improvement. In 1947, the reconstituted government of Japan resumed the work on the preparation of national income estimates, charging its Economic Stabilization Board with the task. The latter proceeded at once with the work of preparation and publication of such esti­ mates with whatever statistics it could as­ semble. At the same time, realizing the in­ adequacies of these statistics, it began to collect new statistics of production, income distribution, and consumption for both cur­ rent and past years which would make pos­ sible an eventual recasting of all its national income series. By the end of 1952, the work of collating the new data was completed. In August of that year, the Economic Sta­ bilization Board gave way to the Economic Counsel Board, and the latter published a comprehensive report in January, 1953, giv­ ing the long-awaited revised estimates. The new series were prepared on a fiscal year rather than calendar year basis, so that they would be closer related to the government budget and would be more useful in the government’s fiscal and economic planning. The new series made possible straight com­ parisons of the postwar national income with that of the 1934-36 period. The new as well as the old Japanese esti­ mates of the postwar period were heavily influenced by the United States and United Nations concepts and techniques. Study trips to the United States and to other countries enabled Japanese statisticians to familiarize

498

THE INCOME OF NATIONS

themselves with the most advanced work being carried on there; and the dispatch to Japan of American and United Nations’ ex­ perts further helped to achieve these results. The new official estimates of Japan em­ phasized the gross national product more than the national income at factor cost. All three phases, output by industrial divisions (plus net balance of income from abroad), distributive shares, and final consumption and investment expenditures, were covered. The industrial divisions were arrayed in three groups, primary production, secondary production, and tertiary production, with a subdivision of each into a number of cate­ gories, and were also regrouped as material production and as services. Distributive shares were classified as labor income, in­ dividual proprietary income, corporate busi­ ness income, and other income. Gross na­ tional expenditure was shown by its four elements of (a) individual spending on con­ sumer goods and services; (b) private capi­ tal accumulation, individual and corporate; (c) net overseas investment; and (d) central and local government procurement of assets and services, with further subdivisions of each. The annual reports on the national income were given the status of a White Paper and were henceforth submitted to the Parliament more or less concurrently with the presenta­ tion of the national budget. In 1954, esti­ mates of sector accounts were under prepa­ ration, thus further extending the scope of Japanese national accounting. Parallel with the extensive government activity in the field, there was substantial activity on the part of private researchers and institutes. Their efforts were mostly in the form of theoretical and methodological investiga­ tions, as well as interpretations of published figures. Thus, in the course of some ten years, 1946-55, national income estimating and research in Japan advanced from a com­ paratively undistinguished position to one of considerable prominence. In 1956, the National Income Section in­ augurated a series of estimates going back to 1930, expressed in 1934-36 prices (see Reference No. 4 ).6

Bibliography Period 1900-1940 Kinzo Nakamura, National Income of Japan, Tokei, Shoshi, No. 224, October, 1902 (in Japanese); Tetsutaru Yamashita, “National In­ come of Japan,” Toyo Keizai Shimpo (Nos.

378-379), June, 1906 (in Japanese); Saburo Shimoi, “Die Entwicklung des Volkseinkommens in Japan in den Jahren 1903-1919,” Archiv fiir Sozialwissenschaft und Sozialpolitik, 53 Bd., 1925, pp. 141-85 (in German); K. Mori, “The Estimates of the National Wealth and Income of Japan Proper,” Bulletin of the Inter­ national Institute of Statistics, Vol. XXV, 2, 1931, pp. 179-204; The Japan Governmental Bureau of Statistics, The Estimated National In­ come of Japan in 1925 (in Japanese); Report of Survey of National Income 1930, Tokyo, 1934 (in Japanese); Shigeyoshi Hijikata, “Component Factors of the National Income,” Tokyo, 1933, Journal of Economists, edited by the Depart­ ment of Economics of Tokyo Imperial Univer­ sity, Vol. 4, No. 2; Vol. 6, No. 8; Vol. 7, No. 11 (in Japanese); K. Takahashi, The War and Japan’s Economic Strength, Tokyo, 1937 (in Japanese); K. Shibata, The Economic Review, edited by the Tokyo Imperial University Eco­ nomic Society, May, 1939 (in Japanese); O. Tanin and E. Yohan, When Japan Goes to War, New York, 1936; Isoki Asahi, The Eco­ nomic Strength of Japan, Tokyo, 1939; Koya, Japan’s Finance and Industry, published by the Foreign Affairs Association of Japan (quoted by Colin Clark, Conditions of Economic Proggress, 1940, p. 124); Mitsubishi Economic Re­ search Bureau, Trade and Industry: Present and Future, London, 1936, Monthly Circular, 19371941; The Japan Economic Federation, An In­ quiry Into the National Income of Japan, 1930-1939, East Asia Economic Intelligence Series, No. 1, 1939; The Oriental Economist, 1939, pp. 381-85; also 1940, p. 138. Period 1945-1955 1. U.S. Department of State, Interim Research and Intelligence Branch, National Income of Japan, October, 1945. 2. United Nations, Economic and Social Council, Economic Commission for Asia and the Far East, Methods of Estimating National Income and Expenditure in Japan, Bangkok, December, 1951. 3. Japan, Economic Stabilization Board, “Analysis of National Income by Distribu­ tive Shares,” Toyo Keizai Tokei Geppo, February, 1950 (in Japanese); Distribution of the National Income in the Post-War World (in Japanese), August, 1950. 4. Japan, Economic Counsel Board (later re­ named “Economic Planning Board”), Divi­ sion of Research, National Income Sec­ tion, Report on National Income (annual), Tokyo, October, 1953; April and Novem­ ber, 1954, 1955 (in Japanese); National Income and National Economic Account­ ing 1930-1955, Tokyo, 1956 (in English). 5. Japan, Cabinet Bureau of Statistics, An Analysis of Family Expenditures as Re­ vealed in the Consumer Price Survey, July, 1950 (in Japanese). 6. The Oriental Economist, notes on National

36.

7.

8.

9. 10. 11.

THE NEAR, MIDDLE, AND FAR EAST

Income, June, 1953; February, 1954; July, 1954; September, 1954; December, 1955. Shigeto Tsuru and Kazushi Ohkawa, “Long Term Changes in the National Product of Japan,” Income and Wealth, Series III, In­ ternational Association for Research in In­ come and Wealth, Bowes and Bowes, Cam­ bridge, 1953; also S. Tsuru and others, “Analysis of Post-War Economy of Japan,” Keizai Kenkyai, October, 1955 (in Japa­ nese). Kazushi Ohkawa, “A Note on Long Term Changes in the National Product of Japan,” The Annals of the Hitotsubashi Academy, Vol. Ill, No. 2, April, 1953; Errata, Vol. IV, No. 1, October, 1953. Yuzo Yamada, “Income Growth and the Rate of Saving,” The Annals of the Hitot­ subashi Academy, April, 1954. M. Shinohara, “Estimate of Capital For­ mation by the Commodity Flow Method,” ibid., April, 1955. S. Tomita, “National Consumption Level in Japan,” Mitsubishi Monthly Circular, February, 1954.

499

the Bureau of Accounts and Statistics of Taiwan and also by the United Nations Economic Commission for Asia and the Far East. The estimates as published in that year showed Taiwan’s net domestic product for the years 1937 and 1948—54 by indus­ trial origin in prices of 1937, and her gross domestic product for 1951-52 by type of expenditure in current Taiwan dollars. In 1955, the Bureau of Accounts and Statistics published an estimate of the na­ tional income by industrial origin for the years 1937 and 1946-56 in 1937 Taiwan dollars.

Bibliography 1. Taiwan, Bureau of Accounts and Statistics, The Structure of National Income of Taiwan, Formosa, 1955. 2. United Nations Economic Commission for Asia and the Far East, Economic Survey 1954, Bangkok, 1955; Economic Survey 1955, Bangkok, 1956.

13. Korea (South)

15. Philippines

The first estimate of the national income of Korea was prepared by R. R. Nathan and Associates, an American group of ex­ perts, engaged by the United Nations Ko­ rean Reconstruction Agency in 1952-54. The estimate showed the gross national product by industrial origin for the years 1949-50, 1952-53, and 1953-54, and was supported by extensive explanations of the sources and methods used. These estimates were continued and further elaborated by the Bank of Korea. The gross national prod­ uct in the latter’s series is broken down by industrial origin, by distributive shares and by types of final expenditure.

The Central Bank of the Philippines pre­ pared an estimate of the national income of the country in 1949 for the years 1938 and 1946-48. It showed both the gross na­ tional product by industrial origin and the net national income, and gave separate cal­ culations of individual and corporate in­ comes. Since then it has published national income estimates in its annual reports. Studies of the national income of the coun­ try have also been made by the Economic Commission for Asia and the Far East in its economic surveys for 1947 and later years; and in 1952, by the experts of the United Nations Technical Assistance Pro­ gramme. The latter developed a distribution of the national income by size of individual incomes. By 1955, the estimates showed the na­ tional income by industrial origin, by dis­ tributive shares, and by final expenditure for the years 1948 to 1954. The estimates are prepared in close accord with international standards.

Bibliography 1. United Nations Korean Reconstruction Agency, An Economic Programme for Ko­ rean Reconstruction, United Nations, New York, April, 1954. 2. Bank of Korea, “Gross National Product of Korea 1952/53, 1953/54, and Forecast for 1954/55,” Monthly Bulletin of the Bank, April, 1954 (English translation by S. H. Lee of the Bank) and the 1955-57 issues.

14. Taiwan The income of Taiwan was originally es­ timated as a part of the national income of Japan, of which it was a province. Esti­ mates of its income as a possession of Na­ tionalist China were prepared in 1955 by

Bibliography 1. Central Bank of the Philippines, Economic Indicators, Manila, 1949; Annual Report 1953, and later issues. 2. United Nations, Economic Commission for Asia and the Far East, Economic Survey 1947, Shanghai, 1948; Methods of National Income Estimation in the Philippines, Bang­ kok, November 29, 1951; Analysis of Na-

500

THE INCOME OF NATIONS

tional Income Estimates for Selected Asian Countries, Economic Bulletin, Bangkok, No­ vember, 1952. 3. United Nations Technical Assistance Pro­ gramme, Report of National Income Ad­

visor, National Income of the Philippines and Its Distribution, New York, 1952. 4. G. S. Watkins, “Labor and the Distribution of National Income,” Central Bank News Digest, December 11, 1952.

37. DEVELOPMENTS IN AFRICA National income estimates for the colonies and protectorates on the African continent began to be developed during the postwar decade of 1946-55. They were prepared either by interested scholars in the countries administering these territories, with the co­ operation or support of the home colonial offices, or by the statisticians of the local colonial administrations themselves. By 1957, such estimates had been prepared for at least eight countries and steps were under way to develop like estimates for some of the others. The greatest contributions towards the development of these estimates have been made by Phyllis Deane and other scholars at Cambridge University and by the East African and the Central African Statistical offices. We have reviewed, in Chapter 33 above, the estimates of the Union of South Africa and in Chapter 36, Section 1, those of Egypt. Here we shall survey the develop­ ments in the same field relating to the fol­ lowing eight colonies or countries: 1. 2. 3. 4.

Nigeria 5. Belgian Congo 6. Kenya 7. Northern Rhodesia8.

Southern Rhodesia Nyasaland Ghana Tanganyika

Most of these estimates deal not with national income proper, defined as income of residents from within and without the country, but as net geographical income or product, or home-produced income unad­ justed for exports and imports and for other international transactions. There are two main reasons for the preference of the esti­ mators for this aggregate: first, because the international transactions in the case of colonies are considerable in size and any Notes begin on p. 544.

error in their calculations might affect the national total substantially; and secondly, because the home-produced income is in many respects more significant in the case of territories of this type than the adjusted national income total. The estimates are prepared mostly by the net output method, except in the case of the two Rhodesias in which they are prepared mainly by the income-distributed method. The income aggregates in most of the esti­ mates are broken down by industrial divi­ sions, distributive shares, and types of final expenditure; and those for the two Rho­ desias even show sector accounts. All of them follow international standards as closely as the available statistics permit. As Phyllis Deane pointed out in her pio­ neering study of the problems involved in preparing national income estimates for colonial territories, the problems are greatest in estimating the incomes and consumption of the native populations. The difficulties are due as much to the incompleteness of the enumerations of these populations as to the inadequacies of the data relating to their per capita outputs and consumption. Among the difficulties encountered in the prepara­ tion of adequate statistics on both subjects, Miss Deane listed “the relative inaccessibility of remote districts; the shortage of skilled personnel [for the collection and tabulation of data]; the hostility in some cases of na­ tives [to questioning] due perhaps to fear of being taxed or to religious taboos” and the frequent inability, owing to language difficulties, of even the willing native to co­ operate effectively. She wrote that “for quantitative information an investigator must rely mainly on administrators’ esti­ mates.” 1

37.

AFRICA

501

1. Nigeria

Bibliography

D. Forde and R. Scott, in their book on Nigeria, published in 1946, estimated the average worker’s income but did not attempt to develop a national income total for that territory. This was done for the first time in 1948 by Margery Pergham and her asso­ ciates in their survey of the Nigerian econ­ omy. Next, A. R. Prest, in his book pub­ lished during the same year, gave data on the income and output of several branches of the Nigerian economy. But the first truly comprehensive estimate of Nigerian national income was prepared in 1952-53 for the Colonial Office by A. R. Prest and I. G. Stewart of the Applied Economics Depart­ ment of Cambridge University. It was made for the year 1950-51 and may be treated as an official estimate. It was fairly detailed and was accompanied by extensive explana­ tions of its methods and sources. A break­ down of the national total was provided. By 1955, estimates were available of the na­ tional income by industrial divisions, dis­ tributive shares, and types of final expendi­ ture for the years beginning 1950.

1. M. Masoin, “The Economic Situation of the Belgian Congo” (in French), Bulletin de I’Institut des Recherches Economiques et Sociales, Brussels, August, 1948. 2. Central Bank of Belgian Congo, Bulletin, October, 1956; and earlier annual issues of the same bulletin.

Bibliography 1. D. Forde and R. Scott, The Native Econo­ mies of Nigeria, London, 1946. 2. Margery Pergham, Mining, Commerce and Finance in Nigeria, London, 1948. 3. A. R. Prest, War Economics of Primary Pro­ ducing Countries, Cambridge, England, 1948. 4. A. R. Prest and I. G. Stewart, The National Income of Nigeria, 1950-51, Colonial Re­ search Studies No. 11, Colonial Office, Lon­ don, 1953. 5. International Bank for Reconstruction and Development, Economic Development of Nigeria, Washington, D. C., 1955.

2. Belgian Congo The first national income estimate for the Belgian Congo was prepared in 1947-48 by Professor M. Masoin of the University of Brussels as a part of a general survey of the economic situation of the colony, and pertained to the year 1947. Soon thereafter, the Central Bank of Belgian Congo took up the preparation of estimates, publishing them annually in its bulletin ever since. Its estimates give a consistent series from 1950 on, broken down by industrial divisions, dis­ tributive shares, and types of final expendi­ ture.

3. Kenya The first national income estimate for the Colony and Protectorate of Kenya was pre­ pared in 1948 by the East African Statistical Department and related to the calendar year 1947. It was followed by other estimates by the same Department in succeeding years, the latest one, at this writing, published in 1955 and giving the figures for the years 1947-53. The concept employed in these estimates is that of the “net geographical income or product.” In addition to money incomes, the estimate includes (a) the value, at local market prices, of foodstuffs produced and consumed by farmers, which is very sub­ stantial, (b) the value of the services of better type of dwellings, and (c) the value of free housing and board given to em­ ployees by private industry. But the esti­ mate does not include the rental value of native huts, of firewood, and of skins used by natives as clothing. “To this extent,” say the estimators, “the income of the territory with a large peasant economy is undervalued vis-a-vis a more developed community.” Nor is any allowance made for hidden emolu­ ments. The estimates are prepared mainly by the net output method on the basis of produc­ tion statistics. However, the value of the subsistence products of African agricul­ turists produced for their own consumption is estimated on the basis of surveys of family budgets and diets of African tribes. The estimates of the net value of manu­ facturing production had to be derived by indirect methods as the sum of estimated wages, salaries, profits and surpluses, inas­ much as no direct production statistics were available. In 1952, however, the Central Legislative Assembly passed a regulation calling for regular collection of industrial statistics, which should work to improve this part of the estimates. The estimate is presented in the form of the following account, one side of which

502

THE INCOME OF NATIONS

shows the factor incomes paid out (in cash and in kind), and the other-—the origins of the net outputs by industrial divisions:

Phyllis Deane in 1946. The estimate was for 1938 and was based on the concept of geo­ graphic or home-produced income (other-

Geographical Income and Net Product of Kenya, 19S1 (million lb) Type of Income 1. Profits and surpluses (a) Profits, interest, and earnings of self-employed persons (b) Public services 2. 3. 4. 5.

Total Rentals Salaries and wages African marketed produce African subsistence agriculture

Total income so urce:

37.8 2.9 40.7 3.0 32.4 4.7 22.1

102.9

Net Product of 6. Agriculture (a) Non-African (b) African marketed produce (c) African subsistence 7. 8. 9. 10. 11. 12. 13. 14.

Total Commerce, finance, insurance Manufacturing Government Transport Construction (private and public) Real estate (services of dwellings) Mining and quarrying Other services Total net product

21.1 4.7 22.1 47.9 16.1 10.0 7.7 6.9 5.7 3.0 1.4 4.2 102.9

Reference No. 1, page 4.

The authors of the estimate classify its various elements according to the degree of their reliability as follows: Geographical Income

wise defined by her as “taxable income”). The study was made as a part of a broader project, initiated in 1941 by the National Geographical Net Product

Category A, Good Estimates Profits and surpluses of public service Net output of non-African agriculture Salaries and wages Services of government Services of transport Category B, Fair Estimates Profits, interest, and earnings of self-employed Net value of African marketed produce Income from African marketed produce Services of commerce, finance, insurance, etc. Net output of manufacturers Construction (private and public) Mining and quarrying Other services Category C, Rough Estimates Rentals Net output of African subsistence agriculture Income from African subsistence agriculture Net value of the services of dwellings

Bibliography 1. Colony and Protectorate of Kenya, East African Statistical Department, Estimates of Geographic Income and Net Produce, 19471952, Nairobi, 1952 (see also the four pre­ vious annual issues of the same publication). 2. , Quarterly Economic and Statistical Bulletin, No. 29, September, 1955.

4. Northern Rhodesia The first estimate of the national income of Northern Rhodesia was prepared by

Institute of Economic and Social Research in London, with the co-operation of the British Colonial Office, and embraced Nyasaland and Jamaica in addition to Northern Rhodesia. It was conducted with the advice of a committee which included Arthur Robinson, Richard Stone, and W. Arthur Lewis. The estimate was based on a variety of data, including the accounts of large companies and government depart­ ments, local surveys of native family budgets and consumption, data on agricultural out­

37. put, and miscellaneous local data on occu­ pations, specific output yields, and prices. The estimates were judged by the estimator to contain a margin or error of from 7 to 11 per cent, with the largest margin apply­ ing to the incomes of the native population. In 1953, the author completed her Central African studies which had been prepared on the basis of a broad social accounting approach, and which gave detailed estimates for Northern Rhodesia and Nyasaland for the years 1945-48, accompanied by de­ tailed explanations of sources and methods used. In the meanwhile, in 1946, the Central African Statistical Office conducted a Cen­ sus of Population that included questions on income. From these answers, the Office built up a table showing the distribution of personal incomes and the geographic and occupational distribution of incomes of the population. These figures were of consider­ able aid in the preparation of its later esti­ mates of the national income of the country. In 1950, the same statistical office pub­ lished an estimate of the national income and social accounts of Northern Rhodesia for the years 1945-49. This estimate showed the components of personal income and of the national expenditure as well as income and outlay sector accounts for the national and local governments, the companies, sav­ ings and investment, and the rest of the world, and incorporated an explanation of the methods of preparation. The Office con­ tinued to prepare and publish estimates dur­ ing the succeeding years, elaborating and expanding them from time to time. By 1955, estimates were available for the years 194554 by industrial origin, by distributive shares, and by type of expenditure; and the totals were deflated in constant 1949 con­ sumer prices.

Bibliography 1. Phyllis Deane, “Measuring National Income in Colonial Territories,” Studies in Income and Wealth, vol. 8, National Bureau of Eco­ nomic Research, N. Y. City, 1946; see also her more elaborate study which covered also Nyasaland and Jamaica, The Measurement of Colonial National Incomes, An Experi­ ment, Institute of Economic and Social Re­ search, Cambridge, 1948; also, Colonial So­ cial Accounting, Cambridge, 1953. 2. Central African Statistical Office, Northern Rhodesia Census of Population, Incomes of the Population, Salisbury, 1948; The Na­ tional Income and Social Accounts of Northern Rhodesia 1945-1949, Salisbury,

AFRICA

503

1950; same for 1945-50, 1945-51, 194552, 1945-53, published in 1952-54.

5. Southern Rhodesia The first comprehensive estimate of the national income of Southern Rhodesia was prepared in 1945 by Professor S. H. Frankel for the Commission of Inquiry into the Mining Industry of Southern Rhodesia. It related to the years 1924-43, and was pre­ pared by the net output method. Since 1947, the Central African Statistical Office has been publishing annual national income es­ timates for that territory. Its first estimate was prepared for the years 1942-44 by the income-distributed method on the basis of income tax returns, workmen’s compensa­ tion statistics, insurance statistics, company reports, and other such data. In 1948, the office expanded the estimates to include fig­ ures on personal income, personal expendi­ ture, government expenditure, and home and foreign investment. In 1949, the esti­ mates were further extended, in accordance with the recommendations of the United Nations publication Measurement of Na­ tional Income and Construction of Social Accounts (1947), to include sector ac­ counts; and in 1952, the name of the annual publication was changed to National In­ come and Social Accounts to express more clearly the broadened nature of the esti­ mates. The preparation of the estimates was greatly aided by the 1946 Census of Popu­ lation, which included questions on income. The estimates are given in the form of social accounts with figures covering the years beginning 1946. The upper half of one account shows the national income, net of depreciation, divided into two parts: (a) income generated in the money economy broken down by distributive shares (wages and salaries of European and non-European workers; income from unincorporated Euro­ pean and non-European enterprises, com­ pany incomes, incomes of statutory boards and commissions, central and local govern­ ment incomes from property, and miscel­ laneous incomes from property), making up the net geographic income in the money economy, plus the net receipts of interest, dividends, and profits from abroad less the net interest, dividends and profits paid to abroad and (b) income generated in the African subsistence economy (value of crops consumed, cattle slaughtered, services of housing and firewood, and increase in

504

THE INCOME OF NATIONS

cattle herds). The lower part shows the national expenditure similarly divided into (a) net expenditure in the money economy (personal and nongovernment current ex­ penditure on goods and services, current expenditure by central and local govern­ ments on goods and services, home invest­ ment, net expenditure by rest of the world on Southern Rhodesia’s goods and services, and the statistical discrepancy, making up the net expenditure in the money economy at market prices) plus subsidies and less in­ direct taxes, and (b) net expenditure in the African subsistence economy (subsistence consumption and investment). The account of personal income and out­ lay is similarly divided into two parts. The upper part shows the personal income broken down into: (a) income from the provision of goods and services (wages and salaries of Europeans and Africans; income from unincorporated European and African enterprises; interest, dividends, and profits paid by companies; and miscellaneous in­ come from property); (b) transfer incomes from central and local governments (debt interest, pensions, grants to missions, other); and (c) transfer incomes from abroad (im­ migrants’ funds, emigrants’ remittances, other transfers from abroad). The lower part of the account of personal income and outlay is broken down into: (a) current expenditure on goods and serv­ ices (food, drink, tobacco, housing, fuel and light, furniture and other household goods, servants’ cash wages, clothing and footwear, new cars, motorcycles, bicycles, etc., run­ ning expenses of cars, etc., railway and other fares in the country, travel abroad, expendi­ ture by missions and lotteries on personal services, and other outlays); (b) transfers of income to government (direct taxes, fees, fines, etc.); (c) transfers of income to abroad (emigrants’ funds, immigrants’ re­ mittances to abroad); and (d ) savings (in­ crease in balances in Post Office Savings and bank deposits, tobacco industry con­ struction, capital portion of life insurance premiums after deduction of claims and paid surrenders, contributions to pension and provident funds, capital repayments of mortgage loans, savings in building societies, and other savings). These general accounts are supported by sector accounts for companies, statutory boards and commissions, central and local governments’ current operations, transac­ tions with the rest of the world, and central government capital account.

The net geographic output is broken down by industrial origin, separately for the part originating in the money economy (Euro­ pean and African) and for the part originat­ ing in the subsistence economy. The Inter­ national Standard Industrial Classification recommended by the United Nations na­ tional income unit in 1948 was fairly closely followed in this breakdown. An additional historical series of the net national income is given for the years 1939-45. The esti­ mates are extraordinarily comprehensive.

Bibliography 1. S. H. Frankel, Report of the Commission of Inquiry into the Mining Industry of South­ ern Rhodesia, Salisbury, 1945. 2. Central African Statistical Office, National Income and Social Accounts of Southern Rhodesia, Economic and Statistical Bulletin of Southern Rhodesia, 1947-55 issues. 3. Government of Southern Rhodesia, National Income and Balance of Payments Financial Statement, 1950; 1951, and succeeding years.

6. Nyasaland The first estimate of the national income of Nyasaland, prepared by Phyllis Deane, was published in 1953 and covered the years 1945-48 (see section on Northern Rhodesia, above). It showed the income of the territory by industrial origin. It was followed by estimates by the Central Afri­ can Statistical Office for certain regions of Nyasaland. Eventually, estimates for Nyasa­ land were incorporated in the over-all esti­ mates for the Federation of Rhodesia and Nyasaland prepared by the Statistical Office.

Bibliography 1. Phyllis Deane, The Measurement of Colonial National Income, Cambridge, 1948; and Co­ lonial Social Accounting, Cambridge, 1953. 2. United Nations, General Assembly, Stand­ ards of Living in Non-Selfgoverning Terri­ tories, Prepared by the Secretariat of the United Nations, New York, August 1, 1952. 3. Central African Statistical Office, The Na­ tional Income and Social Accounts of the Federation of Rhodesia and Nyasaland, Sal­ isbury, 1956.

7. Ghana and Tanganyika After this chapter was set in type, infor­ mation was received about the publication of national income estimates for Ghana for 1950-55, by the Ministry of Finance, and for Tanganyika for 1952-54, by the Colo­ nial Office in London (see U.N., Yearbook of National Accounts Statistics, 1957).

Statistical A ppendix

.

STATISTICAL A P P E N D IX

N A T IO N A L INCOM E AG G R EG A TES O F EIG H TY -SEVEN COUNTRIES A T M ID -C EN TU R Y : 1 9 5 0 -5 5

Explanations T h e table in this A ppendix gives the m ost im p o rtan t national incom e aggregates for eighty-seven countries fo r the year 1955 o r the nearest year to it, as fa r as they m ay be available. The tables in C hapter 16, Section 5, give, for m ost of these countries, break­ dow ns o f these aggregates by com ponents as w ell as their per capita incom es in U.S. dollars at current exchange rates, and the an n u al percentage increases in their national incom e o r gross national product at con­ stan t prices during the period 1950-55. T aken together, the sum m ary table in this A ppendix and the tables in C hapter 16, Sec­ tion 5, give for each country all o r m ost of th e following statistics: (1 ) population; (2 ) foreign exchange rate of the coun­ try ’s currency in U.S. dollars; (3 ) gross or net dom estic o utput in the national currency at current prices, broken down by industrial divisions in w hich the output originates; (4 ) national incom e at factor cost, simi­ larly in the national currency at cur­ ren t prices, broken dow n by factor shares; (5 ) gross national product ( G N P ), broken down by types of final prod­ ucts o r expenditures, sim ilarly ex­ pressed in the national currency at curren t prices. (6 ) th e percentage increase in the coun­ try ’s gross national product deflated at constant prices, during the period 1950-55, and

(7 ) the percentage increase in the coun­ try ’s population during the same period. (8 ) p er capita national income in U.S. dollars at cu rren t foreign exchange rates fo r the period 1952-54, as com puted by D r. J. B. D. D erksen. T he countries are arranged in a geo­ graphic or cultural, rath er th an alphabetic order. In this w ay the data fo r the coun­ tries of the same region o r stage of cultural and econom ic developm ent appear together, m aking easier com parisons of the sim ilari­ ties in the structures of their national in­ com e or product. T he countries are arranged in five groups as follows: (1 ) Principal English-speaking countries (2 ) W estern and E astern E uropean coun­ tries (3 ) L atin A m erican and Caribbean coun­ tries (4 ) N ear, M iddle, and F a r Eastern coun­ tries, and (5 ) A frican countries T he following m ain sources have been used in the preparation of this table and the tables in C hapter 16, Section 5. a. population figures, taken from the U nited N ations, D em ographic Year­ book, 1956, N ew Y ork, 1956. b. foreign exchange rates, taken from P ick’s Currency Yearbook, 1957, N ew Y ork, 1957. c. national incom e and product totals and breakdow ns, taken mostly from U nited

508

Summary Table. Principal National Income Aggregates for 87 Countries in 1955 or Nearest Year *

In Millions of National Currency a t Current Prices Population in Thousands

National Currency, Its Dollar Exchange, Official or Free ( )

Gross (g) or N et (n) Domestic Output a t Factor Cost

National Income at Factor Cost

Gross National Product at M arket Prices

Principal EnglishSpeaking Countries United States Canada United Kingdom Ireland Australia New Zealand Union of South Africa

165,271 15,601 50,968 2,909 9,201 2,136 13,669

U.S. dollar C. dollar = $1.00 (0.96) pound = $2.80 (2.77) I. pound = $2.80 (2.69) A. pound = $2.24 (2.02) N.Z. pound = $2.80 (2.40) S.A. pound = $2.80 (2.68)

355,900* 23,758* 16,639 g 446* 739 g 1952 1,457“

8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24.

France West Germany Austria Switzerland Belgium Netherlands Luxembourg Denmark Sweden Norway Finland Iceland Italy Spain Portugal Greece Cyprus

Communist Countries (:material production) 25. Soviet Russia 26. Poland

43,274 49,995 6,974 4,977 8,868 10,751 309 4,439 7,262 3,425 4,241 158 48,016 28,976 8,765 7,973 520

200,200 27,278

F. franc, 350 (373-90) = $1.00 D. mark, 4.20 (4.24) = $1.00 A. schilling, 26 (30) = $1.00 S. franc, 4.28 = $1.00 B. franc, 50 (50.35) = $1.00 guilder, 380 = $1.00 B. franc, 50 = $1.00 D. krone, 6.91 (6.82-7.11) = $1.00 S. krona, 5.17 (5.20-5.50) = $1.00 N. krone, 7.14 (7.19-7.73) = $1.00 F. markka, 230 (322-61) = $1.00 I. krona, 16.29 (22) = $1.00 I. lira, 625 = $1.00 S. peseta, 38.95 = $1.00 P. escudo, 28.75 = $1.00 G. drachma, 30 (31-32) = $1.00 pound = $2.80

S. ruble, 4.00 (27) = $1.00 zloty, 4.00 (24-84) = $1.00

139,898* 100,400 *

16,378,000 164,000 100,400 26,770 444,0001954 29,140 18,448 28,625 45,121 24,164 907,200 2,8701964 13,529,000

405,000 s 1954 25,770* 17,810* 28,474* 22,088* 812,900* 11,694,000* 236,000“ 1953 47,419s 56,236* 3 6 "1950

315,000 55,022 351S6°

177,000195»* 213,632“

51,094 64,612 40

(

NATIONS

12,440,000 126,157 81,100 23,080 359,0001954 23,680 14,665 23,600 41,358 19,507 767,600

Western and Eastern European Countries

OF

392,800 26,537 19,058 532 5,120 982 1,798

INCOME

322,700 20,535 15,226 448 4,238 846

THE

1. 2. 3. 4. 5. 6. 7.

ostmark, 2.22 (23) = $1.00 C. koruna, 450 1860 = $1.00 H. forint, 11.74 (45) = $1.00 R. leu, 150 (450-500) = $1.00 Y. dinar, 300 (525-710) = $1.00 B. lev, 6.80 (36) = $1.00

Latin American and Caribbean Countries 33. Cuba 34. Jamaica 35. Puerto Rico 36. Mexico 37. Guatemala 38. Nicaragua 39. El Salvador 40. Honduras 41. British Honduras 42. Panama 43. Haiti 44. Dominican Republic 45. British Guiana 46. Surinam 47. Costa Rica 48. Venezuela 49. Colombia 50. Ecuador 51. Brazil 52. Argentina 53. Peru 54. Chile 55. Paraguay 56. Bolivia 57. Uruguay

5,829 1,542 2,263 29,679 3,258 1,245 2,193 1,660 79 910 3,305 2,404 485 220 951 5,774 12,657 3,675 58,456 19,111 9,396 6,761 1,565 3,198 2,615

C. peso = $1.00 pound = $2.80 dollar = $1.00 M. peso, 12.50 = $1.00 G. quetzel = $1.00 cordoba, 5.00 (7.00) = $1.00 colon, 2.50 = $1.00 lempira, 2.00 = $1.00 dollar = $1.00 balboa = $1.00 dollar = $1.00 D. peso = $1.00 dollar = $1.00 guilder, 1.91 = $1.00 colon, 5.60 (9.00) = $1.00 bolivar, 3.35 = $1.00 C. peso, 2.50 (3.00) = $1.0n, sucre, 15 (17.50) = $1.00 cruzeiro, 18.50 (58) = $1.00 A. peso, 18 (31) = $1.00 sol (19.80) = $1.00 C. peso, 200-300 (242-700) = $1.00 guarani, 21.00 (65) = $1.00 boliviano, 42 (105) = $1.00 U. peso, 1.52 (2.50) = $1.00

49,819“ 350,000“ i954a 57,500“ I954* 600,000“ 1S6i» 1,299,100 *i*4 20,000“ i954

f

(1,262,5001964: 21,7001954

1,863 98 e 1952 1,005“ 1964

8468 1950 473 «1952 10.5“ 1948 222 *1954 209“ i**

982 1954 74,760

4201952 10.81946 221 1954

104 !952 1,203 ^ 84,000 609 1,027 1950 588 257 1954 513

151K1951 1,155“ I960 8,771“ 1954 9,923* 551,300“ 136,040* 461,685“ 1®4 11,993* 24,400“ 1949 1,3041949

136 ' “ 1 80 » 1,1201» 11,042 ^ 8,7291964 9,100 549,300 122,311 18,6901®4 456,9381®4 10,920b 25,1391949

1,2981960 13,9741954 10,037 i954 10,775 553,20019" 145,777 25,082 ^ 944,666

APPENDIX

17,944 13,093 9,805 17,490 17,628 7,548

STATISTICAL

East Germany Czechoslovakia Hungary Rumania Yugoslavia Bulgaria

27. 28. 29. 30. 31. 32.

Near, Middle, and Far East Turkey Egypt Israel Lebanon Syria Pakistan India

24,122 22,934 1,735 1,425 3,856 82,439 381,690

T. lira, 2.80 (7.40-10.25) = $1.00 E. pound = $2.87 (2.50) I. pound = $1.00 (.40-.49) L. pound, 2.21 (3.20) = $1.00 S. pound, 2.21 (3.71) = $1.00 P. rupee, 3.31 (6.00) = $1.00 I. rupee, 4.76 (5.00) = $1.00

18,635* 881“ 1954 1,746“ 1,169“ i954 18,156“ i963 99,100“ i964

20,559 867 1964 1,704 1,185 1954 1,6001963 19,086 ^ 99,1001954

1,8301954

509

58. 59. 60. 61. 62. 63. 64.

510

Summary Table. Principal National Income Aggregates for 87 Countries in 1955 or Nearest Year * ( C o n tin u ed )

In Millions of National Currency a t Current Prices

19,434 8,589 20,302 81,900 25,000 6,058 582,603 21,526 8,907 89,100 2,340 21,849

kyat, 4.76 (8.10-9.75) = $1.00 C. rupee, 4.76 (6.00) = $1.00 baht, 12.50 (22) = $1.00 rupiah, 11.40 (19.50) = $1.00 V.N. dollar M. dollar, 3.06 (3.20) = $1.00 yuan, 4.50—5.50 = $1.00 dollar = $1.00 N.T. dollar, 15.65 (30-42) = $1.00 yen, 360 (400) = $1.00 H.K. dollar, 5.71 (5.80) = $1.00 P. peso, 2.00 (3.00) = $1.00

Gross (g) or N et (n) Domestic Output a t Factor Cost 5,010“ 5,232* 31,864*1854 81,369“ 1952 106“ 1964 5,780*1963 85,700“ 22,490“ 6,820,100“ 3,003“ 8,989*

National Income at Factor Cost 4,087 5,172-» 28,1001964b 81,2041952 1021964

13,6491952 6,794,800 2,5911®2 7,804

Gross National Product at Market Prices 4,788 92,987 1962 J 19 19540 5,6901963 1,154 16,559 >®2 8,188,900 2,842 i“2 8,856

Algeria Morocco Tunisia French West Africa Nigeria Belgian Congo Kenya Rhodesia-Nyasaland Mauritius Uganda Gold Coast

9,620 8,495 3,745 4,679 29,730 12,264 6,048 7,069 549 5,508 4,191

F. franc F. franc F. franc F.W.A. franc pound = $2.80 B. franc pound = $2.80 pound = $2.80 rupee = $0.21 pound = $2.80 pound = $2.80

597 g 1950 55,960* 127“ 1954 587*1954 103“ 1964

165,8001953 161,751 1951 5841950 47,340 127 1954b 309 5761954

* Where a figure relates to another year than 1955, such a year is indicated as a superscript. “ At constant prices: Soviet Russia, 1926/27 prices; Czechoslovakia, 1948 prices; Hungary, 1949 prices; Rumania, 1948 prices. b Net domestic product a t factor cost, rather than national income proper. 0 Gross domestic product a t market prices, rather than GNP. d Gross national product a t factor cost. * Net national expenditure: Dr. Vinski's unofficial estimate. * National income (material product) is broken down by type of final expenditure.

233

686,1001963 551,0001983 194,2001963 6091950 57,790 323 649 1964 249

NATIONS

77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87.

OF

V. Africa {for Egypt see Group IV ; for Union of South Africa see Group I)

INCOME

Burma Ceylon Thailand Indonesia Viet-Nam Malaya (Fed.) China Korea (South) Taiwan Japan Hong Kong Philippines

National Currency, Its Dollar Exchange, Official or Free ( )

THE

65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76.

Population in Thousands

STATISTICAL A PPE N D IX

N ations, Statistics o f N ational Incom e and E xpenditure, Statistical Papers, Series H , N o. 8, N o. 9, and N o. 10, N ew Y ork, Septem ber, 1955, M ay, 1956, and January, 1957; th e U nited N ations, E conom ic Survey o f A sia and the Far East, Bangkok, 1955, 1956, and 1957, and E conom ic Survey o f Europe, G eneva, 1955 and 1956. d. per capita national incomes at factor costs in the U.S. dollars, taken from J. B. D . D erksen “H et N ationale Inkom m en— p er hoofd— in 80 L anden — voor 1952-1954,” N etherland C en­ tra l B ureau voor de Statistiek, Statistische en E conom etrische O nderzoekingen, 3eKw. 1956, G ravenhage. e. fo r C om m unist countries and a few

511

underdeveloped free economies, the statistics are derived from special sources given in the chapters and sec­ tions on individual countries in Parts III an d IV. T he estim ates covered by the sum m ary table in this A ppendix and by the tables in C hapter 16, Section 5, are classified ac­ cording to the extent o f their coverage, in C hapter 15, Section 1, and in Table 17-1, C hapter 17; and according to the degree of their reliability, in C hapter 17, Section 5. T he eighty-seven countries reported on in these tables represent about 95 per cent of all the countries know n in 1957 to have had national incom e estimates. F o r the five additional countries— Iran, Iraq, Jordan, G hana, and T anganyika, see pp. 489, 504.

Notes

NOTES TO INTRODUCTION 1 Joseph A. Schumpeter, History of Eco­ nomic Analysis, New York, 1954, p. 278. Frisch’s principal works are the article on “Dy­ namic Economics” in Economic Essays in

Honour o f Gustav Cassell, London, 1933; and the “Okosirk Systemet” in Ekonomisk Tidskrift, 1943, Stockholm.

NO TES TO C H A PTER 1 1 See Chapter 11, Section 4; and Chapter 13, Section le. 2 See reprint of Boisguillebert’s works in Eugene Daire, Economistes Financiers du X V lllm e Siecle, Paris, 1843. In Chapter II (p. 273) of the Factum de la France, Boisguillebert says (translation is ours): “Just as it is possible to estimate the income of a household, a farm, and a village, so it is possible, for one compe­ tent in matters of this sort, to compute the in­ come of a nation. Such a computation has been made for England, whose national income is but one fourth of France’s in whatever way one may measure it; and the claim is made that this income amounts to 700 million (livres) a year.” Manifestly, Boisguillebert could have had Petty’s or King’s estimates in mind, which fixed England’s national income at between 35 and 43 million pounds sterling. 8 William Petty, The Economic Writings of Sir William Petty, edited by C. H. Hull, Cam­ bridge, 1899, vol. I, pp. 108-10. 4 Ibid., p. 306. *lbid., p. 110. «Ibid., p. 313. 7 Gregory King, Two Tracts, edited by G. E. Barnett, Baltimore, 1936, pp. 30-31. 8 Translated from the reprint of Boisguille­ bert’s works in Eugene Daire’s Economistes Financiers du X V IIIm e Siecle, Paris, 1843, pp. 174, 180, 212, 279, 281-82. Even more effec­ tively than Petty three decades or so before and just as effectively as Adam Smith three quarters of a century later, Boisguillebert exposed the

fallacies of the mercantilistic concept of na­ tional wealth as a stock of gold and of national income as the annual accretion to that stock. He defined the true function of money as a mere medium of exchange. Thus he wrote (pp. 162, 209-10, 214): “Wealth does not consist, as you may think, of gold and silver but only of consumable things, including the raw materials furnished to you by agriculture. Gold and silver are wealth only for the countries which pro­ duce them. For all other countries they are but the traditional means of payment for goods, which alone makes the acquisition of these media worthwhile. The richer the country, the more readily it can dispense with precious met­ als, for it has just so much larger proportion of people in whose case the use of metallic money can be supplanted by the use of pieces of paper called ‘bills of exchange.’ ” °lbid., p. 210. 10 Ibid., p. 403. 11 Ibid., p. 405. 12 Ibid., p. 404. 13 Detail de France, op. cit., Chapter V on the “Great Interest Which the King Has in the Improvement of the National Income,” p. 175. The editor of his works, Eugene Daire, a fol­ lower of Adam Smith, 150 years later (1843) in his annotations to the Detail (p. 175) took Boisguillebert to task for committing the “grave error” of including immaterial products in na­ tional income. 14 Reprinted in Eugene Daire, op. cit. See particularly Chapter X, p. 149.

516

NO TES TO CHAPTER

15 For a fuller treatment of the physiocratic doctrine, see J. S. Schumpeter, History of Eco­ nomic Analysis, New York, 1954, pp. 229-30, 238-43; Encyclopaedia of Social Sciences, vol. V, pp. 348-51; and Chapter 5, below. 16 Della Moneta (1751), reproduced in P. Custodi (Italiani Scrittori di Economia Politico), Parte Moderna, Milano, 1803, vol. 3, p. 222. All translations from the Italian writers in this section are the author’s. 17 Dialogues sur le Commerce des BUs (1769), reproduced in G. de Molinari, Necker, Galiani, M onty on, Bentham, Paris, 1848, p. 197. 18 Reproduced in Custodi Collection, op. cit. 19 Ibid., pp. 22-25. 20 Ibid., pp. 25-30. 21 Both of Palmieri’s works are reproduced in the Custodi Collection, op. cit., Parte Mo­ derna, 1805, vols. 37 and 38. 22 Thus Verri wrote: “Money obtained with industry and distributed over a great number of people will speedily remedy any disparity in production. . . . Nature itself (meaning free actions of man in monetary exchanges) when she is allowed to work freely would treat all men as a kindly mother, correcting the ex­ cesses and deficiencies of all factions, distrib­ uting the good things and the bad things in proportion to the activity and wisdom of peo­ ple and leaving among them only those dis­ parities that are sufficient to keep in motion the desire and the industry of man, just as in the ocean, because of the actions of celestial bodies, waters move in tides. Political ob­ stacles thrown into the path of nature because of the quest of politicians for perfection can do no more than retard that equilibrium to which all moral and physical things necessarily tend” (Verri, Custodi Collection, op. cit., p. 31). 23 Adam Smith, Wealth of Nations. All ref­ erences to his work are from the Modern Li­ brary edition, N. Y. City, 1937. 24 Wealth of Nations, op. cit., p. 271. 25 Ibid., p. 236. 26 Ibid., Book II, Chapter III, pp. 314-15. 27 Ibid., Book II, Chapter I, pp. 263-64. 28 Ibid., Book II, Chapter II, p. 271. 29 Schumpeter, op. cit., pp. 307-8. 80 For a fuller description of the wide ac­ ceptance of the Smithian concept of production during the first half of the nineteenth century see: Edwin Cannan, A History of the Theories o f Production and Distribution, P. S. King, London, 1924, Chapter 1; Adolph Wagner, Volkswirtschaftslehre: Grundlegung, Leipzig, 1876, Chapters 1 and 2, and particularly note on p. 17; J. A. Schumpeter, op. cit., pp. 190213, 628-31. 31 A n Inquiry Into the Nature and Origin of Public Wealth, Edinburgh, 1804. 32 Papers in the Publications of the St. Peters­

1

burg Imperial Academy of Sciences, 1806-1827 (in French); his Cours d’Economie Politique, 4 vols., annotated by J. B. Say, Paris, 1823, and his Zur Kritik des Begriffs vom Nazionalreichthum, St. Petersburg, 1827. 33 Traite d’Economie Politique, Paris, 1803, Book I, Chapter III. 34 La Theorie de I’Economie Politique, Paris, 1815. 35 Treatise on Political Economy, translation published in the United States in 1817. 36 Principles o f Political Economy, 1825. 37 Outline o f the Science o f Political Econ­ omy, London, 1836. 38 De la Nature de la Richesse, Paris, 1831, Chapter 2. 30 Cours d’Economie Politique, Paris, 1840. 40 De la Libertd du Travail, Paris, 1845. 41 Staatswirtschaftliche Untersuchungen, Mu­ nich, 1832. 42 Volkswirtschaft, Stuttgart, 1854. 43 National Income— The United Kingdom, London, 1868, Chapter 8, p. 72. 44 Alfred Marshall, Economics o f Industry, 4th ed., London, 1909, pp. 52-53. 45 Alfred Marshall, Principles o f Economics, 8th ed., London, 1936, Book VI, Chapter I, pp. 523-24. 46 Economics o f Industry, London, 1879. 47 History o f the Theories o f Production and Distribution, London, 1893. 48 Volkswirtschaftslehre, Leipzig, 1876. 49 F or a fuller exposition of Marx’s treat­ ment of the Smithian doctrine and correction of it, see his Theories o f Surplus Value, newly translated and published under the title A His­ tory of Economic Theories, Langland Press, New York, 1952, pp. 200-24. Ibid., p. 212. Ibid., pp. 198-200. 52 Ibid., p. 327. Ibid., p. 212. 54 “The total working day of the laborer is divided into two parts. One portion is that in which he performs the amount of labor neces­ sary to reproduce the value of his own means of subsistence. It is the paid portion which is necessary for his own maintenance and repro­ duction. The entire remaining portion of the working day, the entire surplus quantity of labor performed above the value of labor real­ ized in his wages, is surplus labor, unpaid labor, represented by the surplus-value of his entire product in commodities (and thus by a surplus quantity of commodities), surplus-value, which in its turn is divided into differently named parts, into profit [profit of enterprise plus in­ terest] and rent.” Karl Marx, Capital, Kerr & Company, Chicago, 1909, vol. Ill, pp. 970-71. 55 Ibid., vol. Ill, Chapter 49, pp. 978-79. 56 Ibidem. 57 See Chapter 25, Sections 7, 8, and 12; Chapter 28; and Chapter 34, Sections 1-2 and

NO TES TO CH APTER 2

11-10 through 15; also Jan Marczewski, Role o f National Accounts in Planned Economies o f the Soviet Type (in F rench), International Association for Research in Income and Wealth, Income and Wealth Series IV, London, 1955; and E. F. Jackson, Social Accounting in Eastern Europe, ibid.

517

58 J. A. Schumpeter, op. cit., Part V, Chap­ ter 5, “Keynes and Modern Macroeconomics.” Schumpeter points out that Keynes was not ac­ tually original in this approach, as a number of other economists, both in England and else­ where, were thinking along the same lines at the same time.

NOTES TO C H A PTER 2 1 The change in name from “Political Arith­ metic” to “Statistics” had been made under the influence of German fact-collecting scholars, Achenwall, Bielfeld, and others. John Sinclair gave the new name currency in England when he named his great twenty-one-volume work on Scotland, “Statistical Account of Scotland” (1791-99) because, as he said, “a new word might attract more public attention.” 2 William Petty, Economic Writings, edited by C. H. Hull, Cambridge, England, 1899, vol. I, p. 244. 8 Ibid., p. lxiv. 4 In the Preface to this work, Petty wrote: “Sir Francis Bacon, in his Advancement of Learning hath made a judicious Parallel in many particulars, between the Body Natural, and Body Politick, and between the A rts of preserving both in Health and Strength. . . . Now, because Anatomy is not only necessary in Physicians, but laudable in every Philosoph­ ical person whatsoever; I therefore, who pro­ fess no Politicks, have, for m y curiosity, at large attempted the first Essay of Political Anatomy. Furthermore, as Students in Medi­ cine, practice their inquiries upon cheap and common Animals, and such whose actions they are best acquainted with, and where there is the least confusion and perplexure of Parts; I have chosen Ireland as such a Political Ani­ mal, who is scarce Twenty years old; where the Intrigue of State is not very complicate, and with which I have been conversant from an Embrion . . (Ibid., p. 129). 6 Hull, the editor of Petty’s works, cites nu­ merous works published in England in the 1660’s (by Josiah Child, Roger Coke, etc.) claiming that England’s industries were declin­ ing during the reign of Charles II and that her immediate political and economic prospects were dismal (Economic Writings, I, pp. 24142). 6 Op. cit., pp. 241-42. 7 Ibid., pp. 105-8, 112. 8 Ibid., p. 108. °lbid., p. 310. 10 Ibid., pp. 291-94. 11 Ibid., pp. 110-12. 12 Ibid. is Ibid., p. 240. i4 Ibid., pp. 239-40. 16 Ibid., p. lxviii.

i® Two Tracts by Gregory King, edited by G. E. Barnett, Baltimore, 1936, p. 3. 11 Ibid., p. 4. Davenant in his “Memorial Concerning Creditt” dated July 16, 1696, says: “The Returnes Upon the Assessments on Mar­ riages, Birth and Burial have been examined with great care by an able hand,” thus clearly alluding to King and indicating that King must have completed a considerable part if not all of his “Observations” by that time (Ibid., p. 5). ™Ibid„ p. 13. 19 King was not altogether innocent of specu­ lation based on misleading analogies. Thus, after noting that England’s population had in­ creased from 360,000 at the time of her inva­ sion by the Romans fifty-three years before Christ to 2 millions by the time of her con­ quest by the Normans in 1066 and to 5.5 mil­ lions by 1688, he forecast “That in Probability the next doubling of the People of England will be in about 600 years to come, or by the year of our Lord 2300. At which time it will have 11 Mills of People; But that the next doubling after that will not be (in all proba­ bility) in lesse than 12 or 13 hundred Years more, or by the Year of Our Lord 3500 or 3600. At which time the Kingdome will have 22 Mills of Souls, or 4 times it’s present number in case the World should last so long” (Ibid., p. 24). 20 Ibid., p. 56. 21 Discourses on the publick revenues and on the trade o f England, 2 vols., London, 1698; A n Essay on the Probable Methods of M aking the People Gainers in the Balance of Trade, London, 1699; The Political and Com­ mercial Works o f that Celebrated Writer Charles Davenant, collected and revised by Sir Charles Whitworth, London, 1771, 5 vols., hereinafter referred to as “Davenant.” 22 Davenant, op. cit., vol. I, p. 128. 23 Ibid., I, p. 135. 24 Ibid., I, pp. 138-39. 25 Ibid., I, pp. 136-37. 26 Ibid., I, pp. 145-46. 27 Ibid., I, p. 143. 28 Ibid., I, p. 147. 29 Ibid., I, pp. 147—48. 8° Ibid., I, p. 149. 8i Davenant, op. cit., vol. II, pp. 175-220; also A Memorial Concerning Creditt, ad­ dressed “To the Lords Commissioners of his

518

NO TES TO CHAPTER 3

Maj’ties Treasury” dated July, 1696, and pub­ lished for the first time in full by Johns H op­ kins University in 1942 under the title Two Manuscripts by Charles Davenant, with an In­ troduction by A. P. Usher. a2 Ibid., I, pp. 239, 243-44. 88 Ibid., I, p. 248. a* Ibid., I, pp. 249-50. s*lbid„ I, p. 253. so ibid., I, p. 255; see also further elabora­ tions (similarly based on King’s works) of these and other international comparisons of the incomes and expenditures of these three

countries by Davenant in his Collected Works, vol. II, pp. 264-67. 87 Colin Clark, The Conditions o f Economic Progress, London, 1940, pp. ix-x. 88 It should be noted, however, that the sta­ tistical validity of both Petty’s and King’s esti­ mates has been examined recently by Phyllis Deane in her study of the growth of the British national income since the seventeenth century. See her two articles in the Economic History Review, Cambridge, England, April, 1956, and in the Economic Development and Cultural Change, Chicago, November, 1955.

NOTES TO CHAPTER 3 1 Charles King, The British Merchant; or, Commerce Preserved, three volumes, London, 1721, vol. I, pp. 165-67. The author estimated the total house rents of England at £ 3 .5 mil­ lions, and her consumption of foreign goods at another £ 3 .5 millions, thus leaving £ 4 2 mil­ lions for the domestic consumption of her home made goods. See also: Sir Robert Giffen, The Growth of Capital, London, 1889, pp. 83-85. 2 A n Essay on the Causes of the Decline of the Foreign Trade, Consequently of the Value of the Lands of Britain and on the Means to Restore Both, London, 1749. Another tract of his, entitled Serious Considerations on the Sev­ eral High Duties, etc., was first published in 1743, with a third edition dated 1744. 3 Giffen, op. cit., p. 87. 4 Ibid., pp. 89-90. 5 For a review of Young’s works, see Arthur Young on Industry and Economics, arranged by Elizabeth Pinney Hunt, Bryn Mawr, Pa., 1926. 6 Arthur Young, A Six M onths’ Tour Through the North o f England, Containing an account o f the Present State of Agriculture, Manufac­ tures, and Population in Several Counties of this Kingdom, London, 1770, vol. IV, Letter XLII, pp. 543-47. 7 Young on Industry and Economics, op. cit., pp. 132-33. 8 This estimate is discussed by Giffen, op. cit., pp. 92-94. 9 George Rose, A Brief Examination into the Increase of the Revenue, Commerce and Navi­ gation, o f Great Britain Since the Conclusion of the Peace in 1783, London, 1793. 10 Rev. Henry Beeke, Observations on the Produce of the Income Tax and on Its Propor­ tion to the Whole Income of Great Britain, 2d ed., London, 1800. 11 Ibid., pp. 3 and 176. 12 Ibid., p. 128. is Ibid., p. 112. 14 Ibid., pp. 117, 118. is Ibid., pp. 121-24. 16 Ibid., pp. 124-25.

17 Ibid., p. 126. 18 Ibid., pp. 129-30. 19 Ibid., pp. 130-32. 20 Ibid., pp. 139-40. 21 Ibid., p. 141. 22 Robert Giffen held Beeke’s estimates in high esteem, saying that “There is no reason . . . why his estimates should not be made use of as good contemporary estimates, prop­ erly comparable with earlier and later esti­ mates . . . many of his calculations were fully justified by the results of the Income Tax. The total of all the schedules in 1803 for Great Britain was £115,351,000 which compares with Mr. Beeke’s estimate of about £90,000,000 net, excluding certain trade items. . . . I am satisfied that Mr. Beeke’s calculations were so good that we may practically accept his valua­ tion. . . . As regards Schedule A in particular the gross assessments on the Income Tax were in 1803 for G reat Britain £38,691,000 which would include lands, houses, tithes, mines and other items, while the total of Schedule B was £24,279,000, not far short of Mr. Beeke’s figure for lands only. The total of £35,000,000 for trades and professions agrees fairly well with Mr. Beeke’s estimate of income from home and foreign trade, allowing for differ­ ences in arrangement and classification between his method and that of the Income Tax Sched­ ules.” Giffen, op. cit., pp. 97, 100. Phyllis Deane also considers Beeke’s estimates credit­ able, except for an underestimate of agricul­ tural incomes (see her article in Economic Development and Cultural Change, Chicago, vol. IV, no. 1, November, 1955. 23 The full title of Bell’s book was: “Three Essays on Taxation o f Income, with remarks on the late Act of Parliament on that subject; on the National Debt; the Public Funds; on the Probable Consequences of the Law for the Sale of the Land Tax; and on the Present State of Agriculture in Great Britain with a scheme for the Improvement of Every Branch of it, and Remarks on the Difference between National Produce and Consumption,” London, 1799. The title of the second book was Essays on Agricul­

NOTES TO CHAPTER 4

ture with a Plan for the Speedy and General Improvement of Land in Great Britain, Edin­ burgh, 1802. 24 Three Essays on Taxation of Income, op. cit., p. 61. 25 See her article in Economic Development and Cultural Change, op. cit., p. 29. 26 Robert Giffen, The Growth of Capital,

519

London, 1889; A. L. Bowley, Studies in the National Income, Cambridge, England, 1942. 27 Friedrich Gentz, Essai sur I’Etat Actuel de VAdministration des Finances et de la Richesse Nationale de la Grande Bretagne, London and Berlin, 1800. 28 Ibid., p. 85. 29 Ibid., pp. 88-89.

NOTES TO CHAPTER 4 1 Economistes Financiers du X V lllm e Siecle, E. Daire (ed.), Paris, 1843, p. 272. (Transla­ tion is ours.) 2 Daire, ibid., pp. 222 ff. 3 Daire, ibid., pp. 264-66. 4 Daire, ibid., pp. 403-4. Boisguillebert elabo­ rated further upon this subject: “There is not a workman who would not wish to sell his product at a price three times higher than its costs and who would not wish to have the product of his neighbor at one third its costs. It is only at the point of the sword that justice is insured in these encounters; and it is Nature or Providence that has charged itself with this task. Just as it has provided the weakest animal with retreats and other protections against fall­ ing prey to the stronger ones . . . so in human commerce Nature has provided such an order which, if only permitted to operate, would prevent the more powerful individuals from forcing upon the feebler ones a sale which would deprive the latter of a power of sub­ sistence— an order designed to assure each par­ ticipant in commerce a share of the general prosperity proportional to his station. It is said, therefore, that it is of utmost importance that the State Let Nature Take Its Course, allowing every individual full liberty of action and inter­ fering with commerce only to the extent neces­ sary to assure equal protection to all partici­ pants in it and to prevent violence.” Ibid., p. 280. 5 See his Traiti des Grains (Treatise on G rain), Daire, op. cit., pp. 352-93. 6 Excerpt from Saint-Simon’s Mtmoires, re­ produced in Daire, op. cit., pp. 167-68. 7 Le Supplement au D itail de la France; Traiti des Grains; Causes de la Rarete de I’Argent; and Dissertation sur la Nature des Richesses. 8 Testament Politique de Monsieur de Vauban, published in 1707, without identification of publisher and place of publication and in­ cluding Boisguillebert’s Detail and some other works without designation of his authorship as well as Vauban’s Dime Royale itself; followed in 1712 by other foreign editions of complete works of Boisguillebert similarly without men­ tion of his name. 9 Daire, op. cit., p. 175. i° Ibid., pp. 174, 253. i i Ibid., p. 273.

12 Voltaire, Siecle de Louis X IV , Chapter XXVIII on “Finances.” 13 Daire, op. cit., p. 28. 14 Daire, op. cit., pp. 34-35, Preface. The Projet d’une Dime Royale was translated into English under the title A n Essay for a General Tax; or, a Project for a Royal Tythe, 2d ed., London, 1710. 15 Daire, op. cit., Part I, “Second Fonds,” p. 77. 16 Vauban disclaimed any credit for being original in proposing a “Royal Tithe,” pointing out that it is mentioned in the Bible many times and was at least three thousand years old. 17 Daire, op. cit., “Premier Fonds,” pp. 62-63. 18 After completing this calculation based on arable land only, Vauban added, “and much more were the woods, meadows and pasture ground tithed.” 19 Daire, op. cit., Part I, “Second Fonds,” p. 72. 29 Ibid., p. 73. 21 Ibid., p. 121. 22 Ibid., pp. 75-76. For Paris alone he esti­ mated the rental income from dwellings at 20 million livres. 23 According to Eugene Daire, this was a considerable underestimate. Ibid., p. 82. 24 Ibid., pp. 92, 102. 25 Vauban, A n Essay for a General Tax, etc., op. cit., p. 62; Daire, op. cit., p. 93. 26 Daire, op. cit., p. 101. 27 Reflexions Politiques sur le Commerce et les Finances, La Haye, 1st ed., 1738. Reprinted in Daire, op. cit. 28 Says Dutot: “Mr. Marshal Vauban in his ‘Dime Royale’ shows us that the revenues of the Kingdom amounted to 2,336,450,000 livres” (Daire, op. cit., p. 877). He also says that the “louis d’or” was fixed in 1699—the year cov­ ered by Vauban’s figures— at 30 livres 12 sous. 29 Ibid., p. 877. 30 Ibid., p. 71. 31 Ibid., p. 443. 32 Law wrote some of his works on banking and finance in England, before his migration to France and organization there of his great financial “bubbles”; but his most important pamphlets outlining his complete “Financial System” were published in France, or, after his exile, in Italy. 33 Daire, op. cit., p. 877.

520

N O TES TO CHAPTER 5

NOTES TO CHAPTER 5 1 See Quesnay’s article “Grains (Econ. Polit.),” published in the Encyclopedic, vol. 7, in 1757 (Economistes Financiers du X V lllm e Siecle, E. Daire (ed.), Paris, 1843, pp. 193— 249). 2 Most of the quotations in this section are taken from the Oeuvres Economiques et Philosophiques de F. Quesnay, edited by Auguste Oncken, Paris, 1888. The translations are ours. 8 Ibid., p. 309. 4 Ibid., pp. 314-16. 6 At this point Quesnay inserted a footnote explaining that each sum received by the productive class and by the sterile class “pre­ supposes a double value, inasmuch as there is both a sale and a purchase in each case”; but that there is “real consumption only for the value of 5 billions which form the total receipts of the productive class.” He also explained some of the workings of money and other fac­ tors responsible for a rise in prices. 6 Ibid., pp. 316-17. 7 Ibid., p. 311. 8 Ibid., pp. 494-515, especially p. 500. 9 Ibid., p. 63. For a more recent scholarly analysis of Quesnay’s Tableau, see Henri Woog, “The Tableau Economique of Francois Quesnay, An Essay in the Explanation of Its Mechanism and a Critical Review of the Interpretation of Marx, Bilimovic, and Oncken,” Staatswissenschaftliche Studien, Vol. 7, Bern, 1950 (in Eng­ lish). 10 Oeuvres de Lavoisier, Paris, 1893, vol. VI, pp. 413-14. The Philosophie Rurale, published in three volumes in Amsterdam in 1764 as the work of Victor Mirabeau, is sometimes con­ sidered to be the work of Quesnay. However, in the copy of the work available to us, we were unable to find the figure mentioned by Lavoisier. 11 F or a sketch of his life and reprint of his Principes Economiques, see Daire and de Molinari, Melanges d’Economie Politique, Paris, 1847. 12 Daire and de Molinari, op. cit., pp. 26878, 360. 18 A. R. J. Turgot, Reflexions sur la Forma­ tion et la Distribution des Richesses, written by him for the benefit of two Chinese students in 1766 and first published in Du Pont’s Ephemerides in 1769-70. 14 Schumpeter, History of Economic Analysis, New York, 1954, pp. 248-49. 15 Schumpeter, op. cit., pp. 246-47. 16 Voltaire, The Man of Forty Crowns, trans­ lation from French, London, 1768, pp. 10-12; Oeuvres de Lavoisier, op. cit., p. 413. 17 See Note 21, below, re d’lvemois. 18 Discours prononce a l’Assemblee Natio­ nale, by M. Du Pont, Sur I’Etat et les Ressources des Finances, Imprime par ordre de l’Assemblee,

Versailles, 1789, pp. 9-12. Du Pont was a phy­ sician, but he was more interested in political economy and, later, in the publishing business than in medicine. He managed to escape the guillotine during the Revolution, and emigrated to America with his two sons in 1800, first only temporarily and later in 1815 permanently, founding there with them what eventually be­ came one of the greatest industrial and com­ mercial enterprises of the world, but retaining to the end of his life an active interest in social reform. 19 Memoires sur le Commerce de la France et de ses Colonies, Paris, 1789. 20 Moreau de Jonnes, Statistique de I’lndustrie de la France, Paris, 1856, pp. viii, ix. 21 Francis d’lvernois, Historical and Political Survey o f the Losses Sustained by the French Nation, etc., London, 1799, p. 256. 22 A rthur Young, Travels During the Years 1787, 1788, 1789, etc., London, 1792. 28 Ibid., p. 454. 24 Ibid., p. 455. 25 Ibid., p. 462. 26 “If, by net produce, we are to understand rent, and if it does not mean that, I know not what it can mean . . .” Ibid., p. 463. ™lbid., p. 463. 28 Biographical material based on Daire and de Molinari, op. cit., vol. I, pp. 577-80; and on Douglas McKie, Antoine Lavoisier, New York, 1952. 29 Translated from Daire and de Molinari, op. cit., p. 580. 80 In the preface to an accompanying work on The State o f the Finances o f France, pub­ lished by him in January, 1792, and addressed to the National Assembly, Lavoisier wrote: “At a time when people are prone to exaggerate everything, the good as well as the bad; when everyone views things with an instrument which magnifies or reduces them, lengthens their dis­ tance or brings them closer; when no one sees them in their true dimensions or at their actual location, I thought that it would be useful if some one undertook to discuss dispassionately the state of affairs, and to subject the finances of the State to a rigorous arithmetical calcula­ tion. . . . Let the reader not expect to find here anything but what originates from the im­ portance of the subject: this work will be as cold as reason” (Oeuvres de Lavoisier, op. cit., VI, p. 464). 81 Oeuvres de Lavoisier, op. cit., VI, pp. 404-5. 82 Ibid., p. 405. 88 Ibid., pp. 415-16. 84 Ibid., p. 414. 85 Ibid., p. 406. ™Ibid., pp. 418-19. v Ibid., pp. 411, 420-21, 442.

NO TES TO CHAPTER 6

88 Ibid., pp. 426, 452. H e estimated the total production of oats at 400 million measures valued at 200 million livres, but allowed for four fifths of it to be used by animals and only one fifth (40 million livres) by human beings. 88 Ibid., p. 424. 40 Ibid., pp. 458-59. 41 Ibid., pp. 462-63; and 425-26. 42 Ibid., p. 426. 43 Ibid., p. 427. The “produit net” before the imposition of the tax, according to Lavoisier, was composed in money values as follows: 728 million livres from wheat, 80 from wine, 169 from meat and animal products, 120 from forest products, 50 from wool products, 32 from oats, 17 from hay and straw, and 2 mil­ lion livres from silk. 44Ibid., pp. 416, 467-68. 46 Ibid., p. 428.

521

46 Daire and de Molinari, op. cit., pp. 608-14. 47 See the critical observations on these two estimates made some two decades later by Charles Ganilh in his La Theorie de I’Economie Politique, Paris, 1815, vol. I, pp. 74-77. 48 Colin Clark, The Conditions o f Economic Progress, London, 1940, p. 2. 49 Corps Legislatif, Conseil des Anciens, Opinion d’Arnould, Seance du Messidor an 6 (1798), pp. 6-10. 50 Sir Francis d’lvernois, Historical and Po­ litical Survey o f the Losses Sustained by the French Nation, in Population, Agriculture, Colonies, Manufactures and Commerce in Con­ sequence o f the Revolution and the Present War (English edition), London, 1799. 51 Ibid., pp. 256-57. 52 Ibid., pp. 249-55.

NOTES TO CHAPTER 6 1 M. V. Ptucha, Otcherki po Istorii Statistiki v SSSR, second printing, Moscow, 1955, vol. 1. 2 Professor G. Vernadsky in his History of Russia (revised edition, Philadelphia, 1944, p. 114) says: “In the sixteenth and seventeenth centuries the Russian population numbered ap­ proximately fifteen million. This figure varied from time to time in view of wars and revolu­ tions, but the general total remained approxi­ mately the same. There was no accurate census of the population of Russia prior to the eight­ eenth century. During the first quarter of that century the population did not increase; in fact, it probably decreased in view of the hardships of Peter’s reign and its increasing wars. At the time of Peter’s death in 1725, Russia had a population of about thirteen million. In the beginning of the nineteenth century, the total rose to forty million, while by the middle of the nineteenth century it had reached almost seventy millions.” 3 F or a description of the development of the Russian bureaucracy see M. N. Pokrovsky, Russkaia Istoria, Moscow, 1933, vol. II, pp. 213-33, and 313-19. 4 Akademia Nauk, SSSR, Institut Ekonomiki, Otcherki po Istorii Statistiki, Moscow, 1955, pp. 36-55. 5 Ptucha, op. cit., pp. 130-36. 6 Ptucha, op. cit., pp. 152-54. 7 Ibid., pp. 154-62. 8 Akademia, Otcherki, op. cit., pp. 57-64. 9 Akademia, Otcherki, op. cit., pp. 57-69. 10 Ptucha, op. cit., pp. 296-97. n Ibid., pp. 299-305. i 2 Ibid., pp. 297-99. i® Ibid., pp. 182-90. 14 Ibid., pp. 163-82. 15 Paul Miliukoff, in his article on Pososhkov (Encyclopaedia of Social Sciences, New York City, 1934, vol. 12, pp. 266-67), says: “Some

of Pososhkov’s proposals are similar to those made by Walter Raleigh in England, Peter de la Court in Holland, and Marshal de Vauban in France; while others anticipate programs ad­ vanced by non-Russian writers at a later date; in either case direct foreign influence is im­ probable.” On the other hand M. N. Pokrovsky (op. cit., vol. II, pp. 200-1) takes a some­ what different view of Pososhkov’s book. He ridicules Pososhkov’s philosophy of “naive na­ tional economic self-contentment . . . and colbertianism.” He refers specifically to that au­ thor’s insistence that everything Russia needs should be produced in Russia, including even toys for children and eye glasses so that “noth­ ing should be bought from foreigners even when offered at half-price,” and also mentions Pososhkov’s belief that all that is needed to accomplish this result is the will of the people; and that when this result is attained Russia will have “the money that she needs.” Pokrovsky believes that Pososhkov “borrowed these mercantilistic notions from the stories he heard from Russians who had been abroad, about the nature of the policies by which foreign govern­ ments fostered the development of manufac­ tures.” Pokrovsky takes the view that “the his­ torical importance of Pososhkov’s book is very problematical,” thus taking issue with the ap­ praisal of it by other Russian economists as the work of a great and original pioneer. i® Akademia, Otcherki, op. cit., pp. 69-72. 17 Pokrovsky, op. cit., vol. Ill, p. 20; Bolshaia Sovietskaia Encyclopaedia, Moscow, 1947, pp. 451-54. 18 P. I. Liashchenko, Istoria Narodnogo Khosiaistva SSSR, Moscow, 1956, 4th ed., vol. I, pp. 403-09. 19 Ptucha, op. cit., p. 342 (the translation is ours); for other information about Bernoulli

522

NO TES TO CHAPTER 6

see ibid., pp. 306-18; and Encyclopaedia of So­ cial Sciences, op. cit., vol. 2, p. 525. 20 For a fuller account of the works of Shcherbatov, Tchulkov, Golikov, and other Russian economists and statisticians of this period, see Ptucha, op. cit., pp. 226-34 and 248-67; and P. I. Liashchenko, op. cit., vol. 1, pp. 410-16. Ibid., pp. 215-22, 318-29. 22 Ptucha, op. cit., pp. 436-50. 28 Ptucha, op. cit., pp. 72-81, 115-19, 190215. Although claiming for Lomonosov inde­ pendence from foreign influence in his politicoarithmetical approach, Ptucha concedes that the work of Graunt, Petty, and Edmund Hailey “having been printed in the Proceedings of the British Royal Society . . . must have been known to him” (ibid., p. 200). 24 R. D. Grehov, “Economic Surveys of the Eighteenth Century,” Letopis Archeographitcheskoi Kommissii, Leningrad, 1929, p. 39. 25 ptucha, op. cit., p. 354. 26 Gemalde von St. Petersburg, Theil 1, Riga, 1794, p. 112, summarized in Ptucha, op. cit., p. 364 (translation is ours). 27 H. Storch, Historisch-Statistisches Gemalde des Russischen Reichs, etc., 8 vols. and supple­ ment, Riga-Leipzig, 1797-1803. 28 A. I. Hodnev, Istoria Imperatorskogo Volno Ekonomicheskogo Obshchestva, St. Pe­ tersburg, 1865, p. 1. 29 Ptucha, op. cit., pp. 92-93. 80 M. N. Pokrovsky, Istoria Rossii, Moscow, 1933, vol. 3, p. 72. S1 Ibid., vol. 3, p. 95; also Liashchenko, op. cit., vol. 1, p. 422. 82 Ptucha, op. cit., pp. 380-94. 88 Ibid., pp. 92-95, 392-94. 84 Ptucha, op. cit., p. 105. 85 Bolshaia Sovietskaia Encyclopaedia, op. cit., pp. 465-66 (the translation is ours). 86 Liashchenko, op. cit., vol. 1, p. 422, quoted from V. A. Bilbasov, Diderot in St. Petersbourg, 1884, pp. 101-03 (the translation is ours). 87 M. V. Ptucha explains the first appearance of the national income estimates in Russia to­ wards the end of the eighteenth century as fol­ lows: “Towards the end of that century the problem of the calculation of the national in­ come came strongly to the fore and assumed reality. The economy was becoming capital­ istic. . . . It became possible to evaluate eco­ nomic activity in monetary terms. . . . Statis­ tical thought was ready for an attempt to meas­ ure the national income of the country in money terms on a per capita basis so that com­ parisons could be made with the economic sit­ uations of other countries. . . . A second cir­ cumstance that prompted the Russian scholars to attempt such calculations during that period was their intensive patriotic feelings and a de­ sire to controvert the wholly unfounded notions about the extreme poverty of the Russian peo­ ple that were being spread abroad at that time

by foreign critics.” (Ptucha, op. cit., p. 47; translation is ours.) 38 The book appeared in German under the title Abriss der Physikalischen Beschaffenheit der Osterreichischen Staaten, etc., by B. F. Hermann, St. Petersburg-Leipzig, 1782, 374 pp. (Description of the Physical Characteristics of the Austrian States and Their Development as Regards Agriculture, Trades, Manufactures, Factories, and Commerce.) 39 The first of these two was entitled Beitrdge zur Physik, Okonomie, Mineralogie, Chimie, Technologie und zur Statistik besonders der Russischen und angrenzenden Lander, 2nd vol­ ume, Berlin-Stettin, 1787; the second book was called Statistische Schilderung von Russland in Riicksicht auf Bevdlkerung, etc. (Statistical Picture of Russia as Regards Her Population, Physical Characteristics, Natural Resources, Agriculture, Mining, Manufactures, and Com­ merce), by B. F. J. Hermann, St. Petersburg, 1790. For a fuller statement of Hermann’s life and writings see Russki Bibliographitcheski Slovar, Moscow, 1916; and Ptucha, op. cit., pp. 234-46 and 359-63. 40 Ptucha, op. cit., pp. 234-38; Akademia Nauk, Otcherki Istorii Leningrada, Moscow, 1955, vol. I, p. 295. 41 Ibid., p. 362. 42 Ptucha, op. cit., p. 362. 43 Ibid., p. 363. 44 A. W. Hupei, Versuch die Staatsverfassung des russischen Reichs darzustellen (Attempt to Present the Political Structure of the Russian Empire), Theil 1, Riga, 1791; Theil 2, Riga, 1793. 45Hiipel, op. cit., part 1, pp. 265 and 584. (Hiipel’s criticisms of Hermann’s national in­ come estimates are summarized by Ptucha, op. cit., pp. 240-41 and 270.) 46 For a sketch of Radishchev’s life and for the texts of his principal works see his Com­ plete Works, St. Petersburg, 1907, 2 vols.; and his Complete Works, Leningrad and Moscow, 3 volumes, editions 1941, 1947, and 1952; also a summary and a characterization of them by M. V. Ptucha, op. cit., pp. 394-414. 47 Ptucha, op. cit., pp. 395-96 (translation is ours). 48 Ptucha, op. cit., p. 242 (translation is ours). 49 This work is reproduced in the Collected Works, op. cit., 1941, vol. 2, etc. A summary of it appears in Ptucha, op. cit., pp. 402-04. Ptucha concedes that Radishchev used Petty’s methods, but he makes an important qualifica­ tion. He writes: “Like the English economist and statistician, William Petty, did before him, Radishchev, by ingeniously combining the quantitative with the qualitative methods, was unveiling the laws of the social phenomena. But there is one great difference between them: Petty conducted his analysis in the interests of the ruling classes of his country, whereas Radi­ shchev wrote his works in order to expose the

N OTES TO CHAPTER 7

miserable situation of the working classes and to protect their interests.” (Ptucha, op. cit., p. 404). This comment exaggerates the differ­ ence between the two writers. Petty was also a rebel, and a populist, and did not enjoy the King’s favor; and most of his works similarly could not be published while he was alive (see Chapter 2, sections lb and d). But it is un­ questionably true that Radishchev was more of an idealist than was Petty. It must not be forgotten, too, that the system under which he lived was far more oppressive than that which existed in England. 60 Akademia, Otcherki, op. cit., pp. 112-16. 51 For a complete text of Radishchev’s esti­ mate see his Collected Works, op. cit., 1907, vol. 2, p. 217; or the corresponding volume in the Soviet editions of the same works, 1941-52. It is reproduced also in Ptucha, op. cit., pp. 401-02. 52 The Preface to the Geography read as fol-

523

lows: “Several years before, while publishing a monthly journal The Discoursing Citizen, in which we placed mostly articles of moral phi­ losophy, history, and fine literature in order to learn the reader’s interest, we learned to our utmost satisfaction that . . . the Russian people were most interested to learn all about the land and history of their fatherland and of the foreign countries as well, as the most im­ portant road to the acquisition of a most useful knowledge to humanity. . . . Therefore, since that time we have been laboring on the prepara­ tion of this Universal Description of the Lands of the World, with special emphasis on our own Fatherland” (quoted from Ptucha, op. cit., p. 289). 68 A summary of this estimate is given in the book of Ptucha (op. cit., pp. 107-08), who had access to the Geography in the Moscow and St. Petersburg Archives. ™Ibid., p. 107.

NOTES TO CHAPTER 7 1 P. Colquhoun, LL.D., A Treatise on the Wealth, Power, and Resources of the British Empire, etc., 2d edition, London, 1815. 2 Encyclopaedia of the Social Sciences, E. R. A. Seligman (ed.), 1930, New York, vol. 3, p. 663. 8 The History of Statistics, John Koren (ed.), New York, 1918, p. 367. 4 Colquhoun, op. cit., pp. 8-9. 5 Robert Giffen, The Growth of Capital, London, 1889, pp. 50, 101. 6 Colquhoun, op. cit., p. 62. 7 Ibid., p. 91. 8 Ibid., p. 96. 9 Encyclopaedia of the Social Sciences, op. cit., vol. 3, pp. 663-64. 10 Colquhoun, op. cit., pp. 63-64. 11 Giffen, op. cit., p. 104. 12 John Gray, A Lecture on Human Happi­ ness, a Lecture on the Existing Evils of Society and Development of Means by which they May be Permanently and Effectively Removed, Lon­ don, 1825; Philadelphia, 1826, pp. 12-17. See also Max Beer’s article on John Gray in the Encyclopaedia of Social Sciences, New York, 1932, vol. 7, p. 159. 18 Simon Gray, The Happiness of States— An Inquiry Concerning population, the modes of subsisting and employing it and the effects of all on human happiness: in which is devel­ oped the new or productive system of statis­ tics. Republished with an additional book and Appendix, London, 1819 (first edition, Lon­ don, 1815). Gray wrote in his letter to J. B. Say (p. 636 of that book): “In a visit which I made to your country last year, I confess I did not find such striking or brilliant results . . . Considering the extensions of buildings and an improvement in their style, which show

the increase of population combined with the concomitant increase of wealth, as the surest symptoms of a thriving country, I paid par­ ticular attention to your towns and villages in these points, and am sorry to say, I saw no progress whatever. I have no recollection of any strictly additional buildings . . . In truth, though we also are suffering from an unusual stagnation, I found at my return, more new houses going on in the petty suburb of London, Camden Town, and its neighbourhood, than I had seen in the whole route of my travel through France. Every town and village seemed sta­ tionary. I own, however, I found much of what I expected, on my principles, from a state so long well peopled. There was an appearance of wealth . . . but of little capital. Your soil is almost universally under cultivation, but, with some exceptions, in a very inferior style. Your people are generally employed and busy, yet not very effectively. Though the popula­ tions of France be to that of England only as about 150 to 230 squire per square mile, France seems to be at a still more considerable rate behind our island in capital, and the results of active capital. In several statistical points we have got the start of a full century before you.” See also S. Gray’s Remarks on the Production of Wealth and the Influence the Various Classes Have on Carrying on That Process: in a letter to the Reverend T. R. Malthus, London, 1820 (The Pamphleteer, London, 1820, vol. 17, pp. 385-416). 14 Joseph Lowe, The Present State of Eng­ land in Regard to Agriculture, Trade, and Fi­ nance, with a Comparison o f Prospects of Eng­ land and France, London, 1822; second edition, 1823; German translation, Leipzig, 1823; New York, 1824. Previous to the writing of this,

524

N O TES TO CH APTER 8

Lowe published A n Inquiry into the State of British West Indies (London, 1807), which was a plea on behalf of the colonial planter for a more liberal colonial policy on the part of the British government, and particularly for a re­ duction of the British duties on the imports of sugar from that colony. Palgrave’s Dictionary of Political Economy carries a short note about Lowe’s books, but gives no dates of his life or any personal information about him. 15 The Present State of England, op. cit., p. 269. 16 Phyllis Deane, “Contemporary Estimates of National Income in the First Half of the Nineteenth Century,” Economic History Re­ view, Cambridge, England, April 1956, pp. 344, 346. 17 Lowe, op. cit., p. 256. ™lbid., p. 384. 19 Ibid., pp. 332-33, Appendix, pp. 94-100; also J. Schumpeter, History of Economic Anal­ ysis, New York, 1954, p. 526. 20 Lowe, op. cit., pp. 404-5. 21 Ibid., Appendix, p. 77. 22 Ibid., p. 75. 23 Pablo Pebrer, Taxation, Revenue, Expen­ diture, . . . and Debt of the British Empire, etc., London, 1833. A French edition was pub­ lished in Paris in 1839 under the title Histoire Financiere de I’Empire Britannique. 24 Ibid., pp. 350, 477. The table is not dated but seems to be based on figures relating to the years 1828-31. 25 Deane, op. cit., Econ. Hist. Rev., pp. 34849. 26 Pebrer, op. cit., pp. 506-11. 27 G. R. McCulloch, A Statistical and De­ scriptive Account o f the British Empire, first edition, 1837; fourth edition, two volumes, 1854. 28 The full title of Porter’s work was: The Progress of the Nation in Its Various Social and Economic Relations from the Beginning of the Nineteenth Century. Published first seriatim in 1838-42, issued next in a single volume in 1843. The second and third editions appeared in 1846 and 1851. It has been termed “an in­ valuable record of the first half of the nine­ teenth century . . . remarkable for the accu­

racy and variety of its information.” (Palgrave’s Dictionary o f Political Economy, London, 1926, vol. 3, p. 170.) 20 W. F. Spackman, A n Analysis o f the Oc­ cupations o f the People, showing the Relative Importance of the Agriculture, Mining, Ship­ ping, Colonial, Commercial, and Manufacturing Pursuits of the United Kingdom and Its De­ pendencies in Numbers, Capital, and Annual Productions. Compiled from the Census of 1841 and other Official Returns. Second enlarged edition, London, 1847. so Ibid., pp. 11-12. 31 The full title of Smee’s pamphlet on the income tax (a copy of which is in the rare book collection of the Harvard University Graduate School of Business) is The Income Tax—Its Extension at the Present Rate Proposed to A ll Classes, Abolishing the M alt Tax, Window Tax, Duty on Hops, on Licenses to Sell and Make Beer, the Tax on Railways, the Excise on Bricks; and Reducing the Duty on French Wines, London, 1846. Smee’s other two bro­ chures (available in the New York Public Li­ brary) are: A paper read before the Society of Antiquarians of London, March, 1851, on the Gold Discoveries and Effect Thereby Pro­ duced on the Relative Value o f Silver and Gold, London, 1851; and A Proposal to Increase the Smaller Salaries Under Government, Cornhill, 1860. 32 Economic History Review, op. cit., p. 351. 83 R. Dudley Baxter, National Income of the United Kingdom, London, 1868, pp. 1-2. 34 Ibid., pp. 2-3. 35 Ibid., p. 3. so Ibid., p. 69. M lbid., pp. 15-16. 38 Ibid., p. 22. 39 Ibid., pp. 52, 56, 60, 61. 40 Ibid., p. 64. 41 Ibid., p. 66. 42 Ibid., pp. 73-74. 43 Giffen’s most important works related to national income were: Essays in Finance, First Series, London, 1880; Second Series, London, 1886; The Growth o f Capital, London, 1889; Economic Inquiries and Studies, 2 vols., Lon­ don, 1904.

NOTES TO CHAPTER 8 1 The History of Statistics, John Koren (ed.), American Statistical Association, New York, 1918, p. 287. 2 Ganilh wrote Essai Politique sur le Revenu Public, etc., Paris, 1806; Des Systemes d’Eco­ nomie Politique, Paris, 1809 (translated into English in 1812 as A n Inquiry into the Various Systems of Political Economy); and La Theorie de I’Economie Politique, Paris, 1815. 3 The English translator of his book, D. Boileau, himself an author of A n Introduction

to Political Economy, wrote in his Preface to the 1812 edition: “No study can be more at­ tractive to a benevolent mind than that which investigates the means of providing a plentiful national income.” He commended Ganilh for being “an able advocate” in France of the truth long recognized in England “that commerce is one of the most powerful of . . . means” for increasing the country’s national income, “at a time when her ruler is bent upon destroying commerce.” A n Inquiry into the Various Sys­

NOTES TO CHAPTER 9

tems o f Political Economy, New York, 1812. 4 Ganilh, La Theorie, etc., op. cit., vol. I, p. 36. 5 Ibid., vol. I, p. 38. 6 Ibid., p. 92. 7 Ibid., pp. 96-99. 8 Ganilh interpreted Beeke’s estimates as showing in England a per capita income for labor of 216 livres, and for the propertied classes of 1,400 livres; and that the laboring class in England, embracing % of the popula­ tion, received %o of the total national income, whereas the propertied classes representing % of the population received of the national income. Ibid., vol. II, pp. 430-31. 9 Ibid., vol. I, pp. 99-100. i° Ibid., vol. II, pp. 458-59. 11 Montalivet, Expose de la Situation de I’Empire, 1813; also Ancienne et Nouvelle France, Paris, 1813. 12 Quoted from Koren, The History o f Sta­ tistics, op. cit., p. 287. 13 Comte Chaptal, De I’Industrie Frangaise, 2 vols., Paris, 1819. 14 Ibid., vol. I, pp. 238-47. 16 Ibid., vol. II, pp. 203-4. 16 Moreau de Jonnes, Statistique de I’Indus­ trie de la France, Paris, 1856, p. 316. 17 J. B. Say, Traiti d’Economie Politique, etc., 6th ed., Paris, 1841, Book II, p. 354; see also Cours Complet d’Economie Politique Pratique, 2d ed., Horace Say (ed.), Paris, 1840, vol. II, pp. 10-11. 18 Say, Cours, op. cit., vol. II, p. 18. 19 Alfred de Foville, La France Economique, Paris, 1887, pp. 437-38; also A. Chabert, Essai sur les Mouvements des Revenus et de VActivite Economique en France, 1798-1820, Paris, p. 428.

N O TES TO 1 History of Statistics, John Koren (ed.), American Statistical Association, New York, 1918, p. 672. 2 George Tucker, Progress o f the United States in Population and Wealth, with an Ap­ pendix, New York, 1855, pp. 169—71. 3 Ibid., pp. 195-96. 4 Ibid., Appendix, pp. 6 3 -6 6 . 5 History of Statistics, op. cit., p. 573. 6 Charles B. Spahr, A n Essay on the Present Distribution of Wealth in the United States, New York, 1896. 7 Ibid., p. 95. 8 Ibid., Chapter V. 9 Ibid., pp. 104-5. Ibid., pp. 128-29. 11 Ibid., pp. 7 9 -8 1 . ™lbid., p. 160. 13 Karl von Czoernig, Statistisches Handbiichlein fur die Oesterreich. Monarchie, Vienna, 1861, pp. 64, 71; the total figure is reproduced

525

20 Moreau de Jonnes, op. cit., pp. xiv, xvixvii. 21 Charles Dupin, Forces Productives et Commerciales de la France, 2 vols., Paris, 1827; and Bien-etre et Concorde des Classes du Peuple Frangais, Paris, 1848 (quoted in Le Revenu National, La Croissance depuis 1780, Institut de Science Economique Appliquee, Paris, 1952, p. 31). 22 J. M. Dutens, Essai Comparatif sur la For­ mation et la Distribution du Revenu de France 1815-1835, Paris, 1842, p. 28; and Analyse Raisonnee des Principes d’Economie Politique. 23 J. H. Schnitzler, Statistique Generate, Methodique et Complete de la France, 4 vols., Paris, 1846. 24 Ibid., pp. 344-47. 25 Maurice Block, Dictionnaire de la Poli­ tique, Paris, 1874, vol. 2, pp. 844—45. 26 Paul Leroy-Beaulieu, Essai sur la Repar­ tition des Richesses, etc., Paris, 1883. 27 Ibid., pp. 7, 492-93. 28 Revue des Deux Mondes, 1883, pp. 512— 51. 29 De Foville, op. cit., pp. 428-33. 80 Ibid., pp. 432-34. 31 Chamber of Deputies, Report No. 1130 on the proposal of an income tax, Oct. 14, 1886. 32 De Foville published his estimated income distribution in the Economiste Frangais, July 10, 1880. 33 Coste, “Etude Statistique sur les Salaires des Travailleurs et le Revenu de la France,” Journal de la Societe de Statistique de Paris, August, 1890, p. 237. 34 Reproduced in A. de Lavergne and L. Paul Henry, La Richesse de la France, 1908, p. 161.

CHAPTER 9 in Mulhall’s Dictionary o f Statistics, London, 1899, p. 322. 14 G. Hufeland, Neue Grundlegung, 1807, 1813; Heinrich Storch, Handbuch der Politischen Oekonomie, 1823; Kaufmann, Politische Oekonomie, 1830; F. B. W. Hermann, Staatswissenschaftliche Untersuchungen, 1832, 1870, p. 309; A. E. F. Schaeffle, National Oekonomie, 1860; R. H. Rau, Lehrbuch, 1868; E. Menger, Volkswirtschaftslehre, 1871; William Roscher, Grundlagen der Nationaloekonomie, 1875, pp. 322-33; Adolph Wagner, Allgemeine oder Theoretische Volkswirtschaftslehre, Grundlegung, Leipzig, 1876, p. 100; and his Grundlagen der Volkswirtschaft, vol. I, 1892, p. 419. 15 Heinrich Storch, “Zur Kritik des Begriffs vom National Reichthum,” Akademia Nauk, St. Petersburg, 1827 (reprint); Gustav Schmoller, “Die Lehre vom Einkommen,” Zeitschrift fiir die Gesamte Wissenschaft, 1863, vol. XIX; F. B. Hermann, Lehre vom Einkommen, 1878;

526

NO TES TO CHAPTER

A. E. F. Schaeffle, National Vermoegen und Einkommen, 1884; R. Meyer, Das Wesen des Einkommens, 1887, and his article in Handwoerterbuch der Staatswissenschaft, 2d ed., vol. 3, p. 348; Fr. Kleinwaechter, Das Einkommen und Seine Vertheilung, Leipzig, 1896. 16 Adolph Wagner, Grundlegung, 1876, op. cit., p. 100. 17 Kleinwaechter, op. cit. 18 Prussia established a “Royal Statistical State Office” in 1805, Bavaria in 1808, Wiirttemberg in 1820, Saxony in 1851, Baden in 1852, etc. (History of Statistics, op. cit., pp. 338-39.) 10 The principal estimates of the period 18601900 and the publications in which they ap­ peared were as follows: G. Rumelin, Beitraege zur Ermittlung des Volkseinkommen Konigr. Wuerttemberg, 1863. Ernst Engel, “Die Ergebnisse der Klassensteuer, etc. im Preuss. Staat,” Zeitschrift des Preussischen Statistischen Bu­ reaus, 1868; “Die Klassensteuer, etc. und die Einkommensvertheilung im Preussischen Staat, 1852-1875,” Zeitschrift des Kgl. Preussischen Statistischen Bureaus, 1875. A. Soetbeer, “Das Gesamte Einkommen und dessen Verteilung im Preussischen Staat,” Arbeiterfreund, 1875; Umfang und Verteilung des Volkseinkommen im Preussischen Staat, 1872-1878, Leipzig, 1879; “Zur Einkommenstatistik Preussens, Sachsens und Grossbritanniens,” Vierteljahrsschr. fuer Volkswirtsch., Politik und Kulturgeschichte, 1887-1888; “Volkseinkommen im Preussischen Staat 1876 und 1888,” Conrad’s Jahrbuch fuer National Oekonomie und Statistik, 1889. A. Samter, “Das Einkommen der Bevoelkerung in Preussen,” Kgl. Preussischen Oekonomische Gesellschaft, Koenigsberg, 1873; H. Losch, Volksvermoegen, Volkseinkommen und Ihre Verteilung, 1887. L. Hermann, Volksvermoegen und Volkseinkommen und Deren Verteilung, 1887. Wilhelm Boehmert, Die Verteilung des Einkommens in Preussen und Sachsen, 1898. R. E. May, “Das Verhaeltniss des Verbrauchs etc., und die Marxistische Theorie,” Schmoller’s Jahrbuch fiir Gesetzgebung, Verwaltung und Volkswirtschaft im Deutschen Reich, 1899, pp. 271 ff. 20 H. W. Arndt, “A Pioneer of National In­ come Estimates,” Economic Journal, December, 1949, p. 616. 21 Coghlan’s estimates were published in the official Yearbook of New South Wales, Wealth

10

and Progress o f New South Wales, 1887-88, p. 552; 1890-91, pp. 639, 644; 1892, p. 849; 1893, p. 781; 1897-98, p. 967; 1900-01, pp. 860-61. 22 Arndt, op. cit., p. 617. 23 Ibid., p. 618. 24 Ibid., p. 625. 25 S. Skiadas, He Hellas en Europe, Athens, 1891. 26 H. Storch, Historisch-Statistisches Gemalde des russischen Reichs, Riga, Leipzig, 1797-1803, 3 vols., vol. 2, p. 574. 27 H. Storch, Cours d’Economie Politique, St. Petersburg, 1815, 6 vols. 28 P. I. Liashchenko, Istoria Narodnago Khosiaistva, Moscow, 1956, vol. II, chapter 6, section 7. 29 V. I. Pokrovsky, Concerning the Stability o f Russia’s Favorable Balance o f Trade (In Russian), St. Petersburg, 1897, chapter 8. In 1896 Mulhall published in London his Indus­ tries and Wealth o f Nations, in which he gave a new estimate for Russia which related to the year 1894. Pokrovsky’s per capita figure corre­ sponded closely to Mulhall’s new figure. 30 Leone Levi, “On the Distribution and Pro­ ductiveness of Taxes with Reference to The Prospective Ameliorations in the Public Rev­ enue of the United Kingdom,” Journal o f the Royal Statistical Society, vol. XXIII, 1860, p. 42. 31M. G. Mulhall, The Dictionary o f Sta­ tistics, London, 1884. In the Preface to this work, Mulhall wrote: “Hitherto every science has had its Dictionary, except Statistics, one of the reasons, perhaps, why so many philosophers deny this branch of study to be a science. This is the first Statistical Dictionary ever published in any language.” In the fourth edition (1899), Mulhall wrote that “since the first edition was published, in 1884, no fewer than 6,000 copies have been disposed of, chiefly among the libra­ ries of Europe, United States, and the British colonies.” 32 M. G. Mulhall, Dictionary o f Statistics, 4th ed., London, 1899, p. 320. 33 Alfred de Foville, La France Economique, Paris, 1887, pp. 444-45. 34 OEEC, National Accounts Studies, Nor­ way, Paris, 1953, p. 11: “The first estimate of Norwegian national income was made in 1893 and concerned the year 1891.”

NOTES TO CHAPTER 10 1 See Chapters 21 through 37. 2 A. L. Bowley, Statistical Studies Relating to National Progress in Wealth and Trade Since 1882, London, 1904. 3 Ibid., pp. 19-20. 4 See Bowley’s article in the Statistical Jour­ nal, September, 1908; and his article “The

Census of Production and the National Divi­ dend,” Economic Journal, vol. XXIII, 1913. 5 L. G. Chiozza Money, The Nation’s Wealth, Will It Endure? London, 1914. 6 Josiah C. Stamp, British Incomes and Prop­ erty, London, 1916. 7 A. de Lavergne and L. Paul Henry, La

N O TES TO CHAPTER 1 0

Richesse de la France, Fortune et Revenus Prives, Paris, 1908, C. Colson, Cours d’Econo­ mie Politique, vol. Ill, edition definitive, Paris, 1927. Emile Levasseur, Questions Ouvrieres et lndustrielles en France sous la Troisieme Re­ pub lique, Paris, 1907, pp. 618 ff. 8 W. I. King, The Wealth and Income o f the People of the United States, New York, 1915, pp. 129, 160, 249-55. Like Spahr’s book, King’s appeared in a series edited by Richard T. Ely. 9 K. Nitschke, Einkommen und Vermoegen in Preussen, 1902. Adolph Wagner, “Weitere Statistische Untersuchungen Ueber die Ver­ teilung des Volkseinkommens in Preussen,” and “Zur Methodik und Statistik des Volkseinkom­ mens und Volksvermoegens,” Zeitschr. des Kgl. Preuss. Statist. Bureaus, 1904. E. Wuerzburger, “Die Saechsische Einkommensstatistik als Masstab fuer die Beurteilung der Einkommensverhaeltnisse,” Zeitschr. des Kgl. Saechs. Statist. Bureaus, 1904. K. Peris, Die Einkommenentwicklung in Preussin seit 1896 nebst Kritik an Material und Methoden, Berlin, 1911. F. Kuehnert, “Einkommensgliederung der Preus­ sischen Bevoelkerung im Zeitraum, 1902-1914,” Zeitschr. des Kgl. Preuss. Statist. Landesamtes, 1916. Biedermann, “Einkommens und Vermoegensverhaeltnisse in Preussen,” Zeitschr. des Kgl. Preuss. Statist. Landesamtes, 1918. A. Friedman, “Die Wohlstandsentwicklung in Preussen von 1891-1911,” Jahrbiicher fiir Nationaldkonomie und Statistik, 48 Bd., 1914. N. Tabacovici, “Das Koenigreich Sachsen,” Volks- und Staatswirtschaftliche Abhandlungen, 3 Folge, Heft 4, 1913. E. Fuhrmann, Das Volksvermoegen und Volkseinkommen des Koenigreichs Sachsen, 1914. 10 R. E. May, “Das Deutsche Volkseinkom­ men im Jahre 1900,” op. cit., 1903, pp. 196 If.; “Das Volkseinkommen und der Zuwachs des Deutschen Volksvermoegens in 1907,” op. cit., 1909, pp. 85 ff. 11 Karl Helfferich, Deutschlands Volkswohlstand, 1888-1913, Berlin, 1913, translated as Germany’s Economic Progress and National Wealth, 1888-1913, Germanistic Society of America, New York, 1914. Helfferich referred to a previous estimate by Schmoller of 25 bil­ lion marks for 1895 and one by SteinmannBucher of 35 billion marks for 1908, as con­ firming his figure (Ibid., p. 100). 12 Ibid., p. 100. 18 Ibid., p. 103. i* Ibid., p. 105. 16 Friedrich von Fellner, “Die Schatzung des Volkseinkommens,” Bulletin de I’Institut Inter­ national de Statistique, vol. XIV, 3, 1905, pp. 109 ff.; “Das Volkseinkommen Oesterreichs und Ungarns,” Statistische Monatschrift, Statis­ tische Zentral Kommission, vol. XXI, 1916, pp. 485-609 (published as a reprint in Vienna, 1917, and separately in Hungarian, in Buda­ pest, 1916). 16 A Statistical Account of the Seven Colo­ nies o f Australasia, 1901-2. Published by the

527

Authority of the Government of the State of New South Wales and of the Commonwealth of Australia, 9th issue, Sydney, 1902, pp. 76369; A Statistical Account o f Australia and New Zealand, 1902-3; Ibid., 1903-4, pp. 530-34; 1904, pp. 321-26. 17 OEEC, National Accounts Studies, Nor­ way, 1952, p. 11. 18 T. Geering and R. Hotz, Wirtschaftskunde der Schweiz, Zurich, 1902; 4th ed., 1910; 5th ed., 1914. 19 Michele Santoro, L ’ltalia nei suoi Progressi Economici dal 1860 al 1910, Rome, 1911. 20 Cyril Popoff’s Economia Bulgaria, Bul­ garian Academy of Sciences, Sofia, 1915 (re­ published in French translation as Bulgarie Economique, Sofia, 1920). 21 Andre Barthe, “Essai d’Evaluation de la Richesse de 1’Espagne,” Journal de la Societe de Statistique de Paris, vol. 58, 1917, pp. 157-61. 22 S. N. Prokopovitch, “Opyt Ischislenia Narodnago Dokhoda,” Trudy Volno Ekonomickeskago Obshchestva, St. Petersburg, 1906, vol. II, pp. 20-35. 23 Speeches of A. A. Bublikov and A. I. Konovalov, Stenographic Debates, 4th Duma, February 18 and 19 and March 22, 1916; an article by M. I. Friedman, Vestnik Finansov, St. Petersburg, 1916, No. 8 (in Russian). See also P. I. Liashchenko, Istoria Narodnago Khosiaistva, Moscow, 1956, vol. II, p. 641 (in Russian). 24 The Economist, London, December 18, 1915, Special Supplement; Vestnik Finansov, St. Petersburg, 1916, No. 1, and No. 8. 25 S. N. Prokopovitch, Voina i Narodnoie Khosiaistvo (The War and the National Econ­ omy), Moscow, 1917; 2nd expanded edition of the same, Moscow, 1918. 26 Colin Clark in 1939 in the Introduction to his Conditions o f Economic Progress, a book in which he assembled data on national income estimates the world over, bitterly complained of the lack of financial and technical assistance in his colossal undertaking. He wrote: “This subject of national income, of central impor­ tance in what is supposed to be one of the most advanced countries of the world, has been left entirely to the unaided efforts of a single indi­ vidual with a number of other duties to per­ form, the only clerical assistance available being that which he cared to pay himself.” Conditions of Economic Progress, London, 1940, p. ix. 27 Among the publications of the Brookings Institution, the following were particularly nota­ ble: I. G. Nourse and Associates, America’s Ca­ pacity to Produce, Washington, 1934; M. Leven, H. G. Moulton, and Clark Warburton, Am er­ ica’s Capacity to Consume, Washington, 1934. Among those of the National Bureau of Eco­ nomic Research the following deserve special mention: the volumes on national income by W. C. Mitchell, W. I. King, O. W. Knauth, and M. Leven, published in 1921-30; Simon Kuz-

528

NO TES TO CH APTER 1 0

nets’ Commodity Flow and Capital Formation, 1938; and the annual series entitled Studies in Income and Wealth (Conference on Research in Income and Wealth) first begun in 1937. 28 The references to all these private works are as follows: J. T. Sutcliffe, National Divi­ dend, Economic Series, 1, Melbourne, 1926; Frederick C. Benham, Prosperity of Australia, London, 1928, Ch. 2; J. G. Crawford, paper published by the Bank of New South Wales, May 4, 1936; and Colin Clark and J. G. Craw­ ford, National Income of Australia, Sydney and London, 1938. 29 F. B. Stephens, “National Income of New Zealand,” The Economic Record, Journal of the Economic Society of Australia and New Zealand, December, 1936, 231-56; and New Zealand Official Year Book, 1932 and later issues. 30 Ragnar Frisch, Nationalregns Kapet (N a­ tional Accounting), Oslo, 1940. 31 The first White Paper was published im­ mediately after the Chancellor submitted his budget in April, 1941. It was called Analysis of the Sources of War Finance and Estimate of the National Income and Expenditure in 1938 and 1940 (Md. 6261, H.M. Stationery Office, 1941). The part on the sources of war finance was prepared by the Treasury; the other part, by the Central Statistical Office. 32 R. F. Harrod, Life of John Maynard Keynes, New York, 1951, pp. 497-503. 33 See U.S. Budget Message, Jan., 1945. See also Gerhard Colm, “Experiences in the use of Social Accounting in Public Policy in the United States,” Income and Wealth, Series I, International Association for Research in In­ come and Wealth, Cambridge, 1951. 34 See the articles of L. Duge de Bernonville, “Les Revenues Prives,” Revue d’Economie Poli­ tique, May-June, 1938; and A. Sauvy and R. Rivet, “Fortune et Revenue National,” ibid., January-February, 1939, pp. 356-92. 35 John Kaldor in reviewing the White Paper wrote: “. . . the publication of this document marks the official recognition of a very impor­ tant principle: that in order to pass judgment on the financial system of the nation, it is neces­ sary to review the national accounts, and not merely the government accounts, which are prescribed in the Budget. It is impossible to judge intelligently the system of taxation, or the scale of public expenditures, without a quantitative record of the total economic ac­ tivity of the nation, which forms its back­ ground. This is perhaps even more important in wartime . . . but it is vital in peacetime as well. If a statement of this kind had been presented year by year, simultaneously with the Budget, many financial mistakes of past Governments might have been avoided.” Eco­ nomic Journal, June-September, 1941, p. 181. 80 The idea of the nation’s economic budget seems to have originated in the United States Bureau of the Budget in 1941 and 1942. See

article by Grover W. Ensley, “A Budget for the Nation,” Social Research, September, 1943; also Chapter 15, Section 3. 37 In the United States official estimates, the shift from net to gross national income or prod­ uct was made first in March, 1942, in Milton Gilbert’s article (U.S. Department of Com­ merce, Survey o f Current Business, March, 1942). 88 The simultaneity of the birth of these two ideas is indicated by the fact that the same issue of the Economic Journal that contained Kaldor’s aforementioned review of the new White Paper with its “national account” also contained the following article by Meade and Stone suggesting the construction of sector ac­ counts: J. E. Meade and Richard Stone, “The Construction of Tables of National Income, Expenditure, Savings and Investment,” Eco­ nomic Journal, June-September, 1941, pp. 21633. 89 Articles in the Economic Journal, June, 1946, and September, 1947; and Appendix on “Definition and Measurement of the National Income and Related Totals” published by the League of Nations, Committee of Statistical Experts, Subcommittee on National Income Statistics, Measurement o f National Income and the Construction o f Social Accounts, Studies and Reports on Statistical Methods, No. 7, Geneva, United Nations, 1947. The greater part of the report was taken up by Mr. Stone’s Appendix outlining the theoretical and tech­ nical problems and advantages of the social accounts. 40 United Kingdom, Central Statistical Office, National Income and Expenditure, 1938 to 1946, Cmd. 7099, London, 1946. 41 U.S. Department of Commerce, National Income and Product Statistics o f the United States, 1929-46, National Income Supplement to Survey of Current Business, Washington, D. C., July 1947. 42 See OEEC, National Accounts Studies, Paris, 1951-53: Denmark; Switzerland; France; The Netherlands; Norway; Sweden. 43 F or a critical discussion of the shift to the social accounting approach and to gross national product see article by S. Kuznets, “National Income: a New Version,” Review o f Economics and Statistics, Cambridge, Mass., August, 1948, and the reply to it by Milton Gilbert and his associates in the U.S. Depart­ ment of Commerce National Income Division, appearing in the same issue of the same journal. See also Chapter 15, Section 2b. 44 United Nations, International Standard Classification o f A ll Economic Activities, Sta­ tistical Papers, Series M, No. 4, New York, 1948; and by the same organization, A System o f National Accounts and Supporting Tables, Studies in Methods, Series F, No. 2, New York, 1952; OEEC, A Standardised System o f Na­ tional Accounts, Paris, 1952; and Edward F. Denison, “Report on Tri-Partite Discussions of

NO TES TO CHAPTERS 1 0

National Income Measurement,” Studies in In­ come and Wealth, Conference on Research in Income and Wealth, National Bureau of Eco­ nomic Research, vol. X, New York, 1947. 45 Wassily Leontieff, The Structure o f Ameri­

TO 1 2

529

can Economy, New York, 1941; 2d ed., 1919— 39, New York, 1951; and Studies in the Struc­ ture o f the American Economy, edited by the same author, New York, 1953. 46 See Chapter 25, Sections 8-12.

NOTES TO CHAPTER 11 1 Irving B. Kravis mentions these criteria for distinguishing economic from noneconomic ac­ tivity: (1) the material criterion that identifies as economic all activity resulting in a material product satisfying a human want; this criterion plays a particularly important role in a primi­ tive society; (2) the exchange criterion that treats as economic all activity resulting in an exchangeable good or service, which is used in economies with highly developed markets; and (3) the psychological criterion that looks to the sensitivity of the activity to rewards and accordingly classifies as economic an activity carried on within the household if it diminishes in response to higher rewards for remunerated activities outside the household. He believes that for the purposes of international compari­ sons of national incomes all three criteria should be used (“The Scope of Economic Ac­ tivity in International Income Comparisons,” Paper presented before the Conference on Re­ search in Income and Wealth, October 29-30, 1954, National Bureau of Economic Research, New Y ork). 2 Simon Kuznets and Raymond Goldsmith, Income and Wealth of the United States, Trends and Structure, Income and Wealth, Series II, International Association for Research in Income and Wealth, Cambridge, England, 1952, pp. 63-69. s J. R. Hicks, The Social Framework, Lon­ don, 1942, Chapter XI. 4 Carl S. Shoup, Principles o f National In­ come Analysis, Boston, 1947, pp. 2, 3; Chap­ ters 5, 6. 5 U.S. Department of Commerce, National Income 1951, Supplement to the Survey of Current Business, Washington, 1951, p. 31. 6 J. Schumpeter, “Das Socialprodukt und die Rechenpfennige,” Archiv. fiir Sozialwissensch. und Sozialpolitik, vol. 44, pp. 627-715, 191718. 7 Statist. Reichsamt, Das deutsche Volksein­ kommen vor und nach dem Kriege, Berlin, 1932, p. 10.

8 “Current Problems in Measurement of Na­ tional Income,” Bulletin de Vlnstitut Interna­ tional de Statistique, vol. 28, No. 2, 1934, p. 283. 9 J. R. Meade and R. Stone, “The Construc­ tion of Tables on National Income, Savings, and Investment,” Economic Journal, vol. 51, 1941, pp. 216-33. 10 Francois Perroux, Le Revenu National, Inst, de Science Econ. Appl., Pragma V. 3, Paris, 1947, p. 26. 11 See for example its issue, National Income Statistics, 1938-1948, New York, 1950, p. 8. 12 For a more complete discussion of the problems involved in the classification of the branches of production under the net output method of estimating national income, see Chapter 18, Section 2. 18 The problems of classification of factor incomes are discussed more fully in the chap­ ter on the income-distributed method (Chapter 19, Section 1-1). 14 The classification of expenditures for final products recommended by international or­ ganizations is discussed in Chapter 20, Section 2.

15 For a further discussion of this subject see the articles by Simon Kuznets and Milton Gilbert in The Review o f Economics and Sta­ tistics, Cambridge, Mass., vol. 30, 1948, pp. 157, 158, and 190. 16 For a fuller discussion see Chapter 14, Section 2b. 17 Organisation for European Economic Co­ operation (OEEC), A Standardised System of National Accounts, Paris, 1952, pp. 28 ff. and 76, and United Nations, A System o f National Accounts and Supporting Tables, Studies in Methods, Series F, No. 2, New York, 1953, p. 17. 18 U.S. Department of Commerce, National Income 1954, Supplement to Survey of Current Business, 1954, p. 32.

NOTES TO CHAPTER 12 1 This percentage is for goods valued at agri­ cultural prices. When these goods are valued at retail prices, farmer’s own consumption rises to 34 per cent of total farm income (Central Statistics Office, Irish Statistical Survey, 1953, Dublin, 1954, pp. 50, 57).

2 Dominion Bureau of Statistics, Reference Paper No. 25, Handbook o f Agricultural Sta­ tistics, Part II, Farm Income, Ottawa, Febru­ ary, 1952, pp. 46, 77. 8 “The Content and Use of Estimates of the National Income,” Banca Nazionale del Lavoro

530

NO TES TO CHAPTER 1 2

Quarterly Review, No. 5, April, 1948, pp. 271 ff. 4 The exclusion of the services of owneroccupied dwellings was thus rationalized by the U.S. Department of Commerce in its first offi­ cial estimates: “The services yielded by such items are not ordinarily paid for in the usual course of business, nor are they commonly evaluated as a result of a bargain in the market place. Usually the motive behind the acquisition of such owned durable goods for personal use is not concerned with a pecuniary return on the investment, as is the purchase of securities” (U.S. Department of Commerce, National In­ come in the United States 1929-35, Washing­ ton, D. C., 1936, p. 4). 5 See Edward F. Denison, “Report on Tri­ partite Discussions of National Income Meas­ urement,” Conference on Research in Income and Wealth, Studies in Income and Wealth, vol. 10, National Bureau of Economic Research, 1947. 6 International Association for Research in Income and Wealth, Income and Wealth Series II, Bowes & Bowes, Cambridge, England, 1952, p. 164. 7Ibid., pp. 164 ff. 8 OEEC, National Accounts and Research Division, A Standardised System o f National Accounts, Paris, 1952, p. 64. 9 One of Adam Smith’s followers among the national income estimators, Colquhoun, got himself involved in this problem of derivative income. He first calculated national income, or the “new property created every year” as the aggregate of the products of agriculture, mining, manufacturing, trade, shipping, fisher­ ies, and fine arts. Next he proceeded to calcu­ late the derivative incomes of unproductive laborers, whose exertions do not create any new property, namely of the royalty, nobility, gentry, government officials, army and navy personnel, pensioners, the clergy, legal and medical practitioners, school and university employees, and paupers. He concluded that “more than one fifth part of the whole com­ munity are unproductive labourers, and that these labourers receive from the aggregate la­ bour of the productive classes about one third part of the new property created annually.” Having thus presented the two categories of in­ come, Colquhoun was unable, however, to reconcile them in some single total and limited himself to the observation that “it does not follow . . . that a very great proportion of these unproductive labourers are not highly useful in their different stations in society. On the contrary, with a few exceptions, in addition to the benefits derived from personal exertions, they eminently tend to promote, invigorate, and render more productive the labour of the creating classes” (P. Colquhoun, Treatise on the Wealth, Power and Resources of the British Empire, London, 1815, p. 109). 10 Josiah Stamp recognized this as he wrote

in 1934: “As civilization advances, it may well do so by the very fact that merely material production in wealth tends to render a less and less proportion of total human enjoyment” (“Methods Used in Different Countries for Estimating National Income,” Journal o f the Royal Statistical Society, vol. XCVII, 1934, p. 426). Colin Clark in his Conditions o f Eco­ nomic Progress postulated as a general propo­ sition that with the advance of civilization, the proportion of services, i.e., of tertiary produc­ tion, in the total national income is increasing. Thus he wrote: “From Sir William Petty’s day to the present time the transfer of working population from primary production to second­ ary and tertiary has been continuing, and per­ haps will continue for as many centuries more” (London, 1940, p. 341). Matolcsy and Varga in their Hungarian estimates confirmed this proposition, saying: “The importance of the latter [immaterial production] is increasing with the advances of civilization to such an extent that the proportion of the value of the produc­ tion of material goods, although still the bulk of the national income, is constantly diminish­ ing” ( The National Income o f Hungary, 1924/ 25-1936/37, London, 1938, p. 7 ). 11 Said M arx: “In the capitalist mode of pro­ duction, the independent peasant or artisan is . . . divided into two persons. In the role of possessor of the means of production he is a capitalist; in the role of worker, he is his own hireling. As capitalist he pays himself his wages and himself gains the profit of his capital, ex­ ploits himself as a wage worker and pays him­ self, out of the surplus value, the tribute which labor owes to capital. Perhaps he even pays himself a third part, the rent . . .” ( Theorien iiber den Mehrwert, edited by Karl Kautsky, Stuttgart, 1905, pp. 422-23; also its English translation, A History o f Economic Theories, Langland Press, New York, 1952, pp. 324-25). 12 In Zur Kritik der Sozial Demokratischen Programme von Gotha (Berlin, 1920, pp. 14-15) Marx defined the “social product” as consisting of: (a) the costs of the means of production; ( b ) additions to capital, reserves, and insurance; and (c) the rest, which is for consumption and from which must be deducted the costs of government, schools, health care, poor relief, etc. 13 Marx does not use the term “net national income” in reference to the sum of wages, profits (including interest), and rent. He pre­ fers to call this aggregate “gross income.” On the other hand, he uses the term “gross prod­ uct” not in the modern sense of factor incomes plus depreciation, but in the sense of total re­ ceipts from sales including the value of raw materials and unfinished products and services of other enterprises or branches of production consumed in the production of a good. This concept of “gross product” was similar, there­ fore, to the modern concept of business turn­ over (what the French call “chiffres d’affaires”),

N OTES TO CHAPTER

inclusive of the value of sales at all stages of production, from the raw material stage, through manufacturing, wholesaling, and re­ tailing. All these terms were first used by the physiocrats and next by Adam Smith, but Marx gave them a different meaning. 14 Marx, Das Kapital, C. H. Kerr & Co., Chicago, 1909, vol. Ill, p. 979; also Ibid., p. 971, where Marx says: “The value of the an­ nual product, in which the new labor added during the year is incorporated, is equal to the wages, or the value of the variable capital, plus the surplus value, which in its turn is di­ vided into profit and rent.” Ibid., p. 328. 16 Ibid., pp. 330-56. 17 A History of Economic Theories, op. cit., pp. 328-29. Marx says: “In addition to min­ ing, agriculture, and heavy industry, there exists a fourth sphere of material production. . . . This industry is transportation, whether of men or goods. The relationship of the productive or wage workers to the capitalists is absolutely the same in this as in other spheres of material production.” But in the succeeding passages he says that “the transportation or mankind, one might say, is simply a service rendered by an entrepreneur,” thus indicating that it belongs in immaterial production. 18 D. I. Chernomordik et al., Narodny Dokhod, U.S.S.R., Akademia Nauk, Institut Economiki, Moscow, 1939; M. B. Kolganoff et al., Narodny Dokhod, U.S.S.R., Akademia Nauk, Institut Economiki, Moscow, 1940. 19 Finansovy Institut, Ministerstvo Finansov, Finansy i Kredit SSSR, Gosisdat, Moscow, 1953, p. 36. 20 The Kolganoff study reduces the 1929 estimate for the United States from $81.0 to $55.9 billion and the 1928 estimate for Ger­ many from 75.4 to 55.5 billion Reichsmarks to adjust them to the U.S.S.R. concept of na­ tional income (Op. cit., pp. 67-71 and 80-81). On the other hand, a later work by Prof. A. I.

13

531

Petrov, Nazionalny Dokhod, Moscow, 1949, p. 13, places the so-called “overstatement” of the national income estimates in non-Communist countries due to their inclusion of services at approximately 15 per cent. 21 Strumilin maintained that “it is incorrect in estimating national income to consider only the net value of the output created by the eco­ nomic enterprises of the nation. The compu­ tation should embrace the incomes of the en­ tire population. The latter is comprised, how­ ever, not only of the net value of the material production, but also of the services furnished without charge to the citizens by the socialist State and the social organizations. Doctors, teachers, and other workers not participating directly in the material production receive a part of the social product created thereby and furnish in exchange for it the results of their own labor in the form of services. The repro­ duction of the material product implies also the reproduction of services. In our socialist practice, the teaching trades are already being treated in all the accounting of the economic enterprises on an equal footing with the work of laboratories, repair shops, etc.” (quoted by Chernomordik, op. cit., pp. 75-76, Problemy Planirovania, Moscow, 1932, p. 397). In June, 1957, at a national conference of statisticians held in Moscow, Strumilin again raised some of these issues, but with no greater success than before (Voprosy Ekonomiki, Moscow, Sep­ tember, 1957, pp. 99-119). 22 Op. cit., pp. 198 and 203. 23 Michael Kalecki and Ludwig Landau, Szacunek Dochodu Spolecznego w.r. 1929, Instytut Badania Konjunktur Gospodarczych, Warsaw, 1931; and Matolcsy and Varga, op. cit., pp. 4-7 (see also below Chapter 14, Sec­ tion le and Chapter 34, Section 11-12). 24 Otto Kraus, “Sozialprodukt und Volksein­ kommen,” Volkswirtschaftliche Schriften, Heft 3, Berlin, 1952, pp. 39 ff.

NOTES TO CHAPTER 13 1 Article by Milton Gilbert et al., The Re­ view o f Economics and Statistics, Cambridge, Mass., vol. 30, August, 1948, p. 183. 2 Simon Kuznets, “Government Product and National Income,” International Association for Research in Income and Wealth, Income and Wealth, Series I, Cambridge, 1951, p. 182. 3 Irving Fisher, The Nature of Capital and Income, New York, 1912; Constructive Income Taxation, New York, 1942; and article, “The Concept of Income: A Rebuttal,” Econometrica, vol. 7, October, 1939. 4 Wrote Heinrich Storch: “L’usage d’une chose, fut-ce meme celui d’une chose materielle, est toujours une chose immaterielle.” (“The use

of a thing, even be it a material thing, is al­ ways an immaterial thing.”) Le Revenu Na­ tional, Akademia Nauk, Memoires, 1817-18, Series 5, p. 420. 5 John Stuart Mill, Principles, Bk. V, Chap. I, § 4: “The proper mode of assessing an in­ come tax would be to tax only the part of in­ come devoted to expenditure.” 6 In this and in other ways saving is as much a part of the currently enjoyed income as con­ sumption. The fact that different values may be placed upon it from those placed on con­ sumption does not justify its exclusion from the concept of income. This was well pointed out by Antonio de Viti de Marco in his First

532

N O TES TO CHAPTER 1 4

Principles of Public Finance, New York, 1936 (translation from Italian), Chapter 10, §7, pp. 225-27. 7 “The Content and Use of Estimates of the National Income,” Banca Nazionale del Lavoro Quarterly Review, No. 5, Rome, April, 1948, p. 301. 8 Corrado Gini, “The Content and Use of Estimates, etc.,” op. cit., pp. 284-85. 9 Carl S. Shoup, Principles o f National In­ come Analysis, Cambridge, 1947, p. 149. 10 Alfred Lowe, “A Structural Model of Pro­ duction,” Social Research, vol. 19, June, 1952. 11 Simon Kuznets, “National Income and In­ dustrial Structure,” Proceedings of the Inter­ national Statistical Conference, Washington, 1947, vol. V, Calcutta, 1951. Reprinted in Economic Change, New York, 1953. 12 Said Marshall: “The terms National In­ come and National Dividend are convertible; only the latter is the more significant when we

14 The problem is treated fully in: Dwight R. Yntema, “National Income Originating in Financial Intermediaries,” Studies in Income and Wealth, vol. 10, National Bureau of Eco­ nomic Research, New York, 1947; United Na­ tions, Measurement o f National Income and the Construction o f Social Accounts, Studies and Reports on Statistical Methods, No. 7, Geneva, 1947; United Nations, National Income Statis­ tics 1938-1948, Statistical Office, New York, 1950, pp. 22-23; U.S. Department of Com­ merce, National Income 1954, Supplement to Survey of Current Business, Washington, D. C., 1954, pp. 46-48. 16 The U.S. Department of Commerce pub­ lication, National Income, 1954, op. cit., p. 46, illustrates the manner in which “the ordinary methods of measuring value added to total output in terms of income and product flows break down in the case of commercial banking” by the following sample account:

Income and Product Account of a Commercial Bank, Monetary Transactions Only (thousands of dollars) Wages paid Net interest paid Interest paid on deposits Less: interest received Profit

50 -9 5 5

Income originating

-1 5

Service charge receipts Less: current account purchases from other firms

10 25

100

30

are looking at the national income in the char­ acter of the sum of the new sources of en­ joyments that are available for distribution” (Principles of Economics, London, 8th ed., 1920, Book VI, Chapter I, pp. 523-24). 13 J. A. Hobson, Taxation in the New State, New York, 1920, Chapter 2.

Product originating

-1 5

16 United Nations, Measurement o f National Income, etc., op. cit., pp. 87—90. 17 U.S. Department of Commerce, National Income 1954, Washington, 1954, p. 47. 18 National Income 1954, op. cit., p. 47.

NOTES TO CHAPTER 14 1 We say “sometimes” because interest on the public debt is often considered a transfer payment that is not included in national in­ come. 2 This assumption may be correct in societies that are well governed but would be incorrect in those that are not. 8 Simon Kuznets, National Income and Capi­ tal Formation, 1919-1935, New York, 1937; see also Studies in Income and Wealth, vol. I, National Bureau of Economic Research, New York, 1937; Dominion Bureau of Statistics, National Income of Canada 1919-1938, Ot­ tawa, 1941. 4 Simon Kuznets, “Government Product and National Income,” International Association for Research in Income and Wealth, Income and Wealth, Series I, Bowes & Bowes, Cambridge, England, 1951, p. 189.

5 Matthias Matolcsy and Stephen Varga, The National Income o f Hungary 1924/25-1936/ 37, London, 1938, p. 6. 6 Ibid., p. 6. 7 “Public Finance in the National Income,” The Review o f Economic Studies, vol. VI, No. 2, February, 1939, p. 150. 8 Economica, vol. XV, New Series, August, 1948, p. 164. 9 Studies in Income and Wealth, op. cit., vol. I, New York, 1937; see Section lg. 10 Gerhard Colm, “The Government Sector in National Income Accounts: A Re-examina­ tion of a Few Controversial Issues.” Paper De­ livered before the Conference on Research in Income and Wealth, New York, October 29, 1954. 11 Income and Wealth, Series I, op. cit., p. 219.

NOTES TO CHAPTER

12 Ibid., pp. 193-94. 13 Studies in Income and Wealth, op. cit., vol. I, pp. 234-35. 14 Income and Wealth, Series I, op. cit., pp. 180-81, 184. 16 Ibid., pp. 192-96. 13 Ibid., p. 194. *7 Ibid., pp. 197-98. 18 Das deutsche Volkseinkommen vor und nach dem Kriege, Einzelschriften zur Statistik des deutschen Reiches, Berlin, 1932, pp. 14-16, 134-41. i° National Income of Sweden, by Eric Lin­ dahl, Einar Dahlgren, and Karin Kock, Lon­ don, 1937, vol. I, pp. 226-31. 20 Studies in Income and Wealth, vol. I, op. cit., pp. 173-227. 21 Ibid., pp. 210-11. 22 “Allocation of Benefits from Government Expenditures,” Studies in Income and Wealth, vol. II, op. cit., 1938, p. 329. 28 See Chapter 10, Section 3e, for other fea­ tures of this agreement. For a fuller account of it see E. F. Dennison, “Report on Tripartite Discussions of National Income Measurement,” Studies in Income and Wealth, vol. X, Confer­ ence on Research in Income and Wealth, N a­

15

533

tional Bureau of Economic Research, New York, 1947, pp. 3-22. 24 S. Kuznets, National Income and Capital Formation, 1937, op. cit.; also Studies in In­ come and Wealth, op. cit., vol. I, 1937, p. 237. 25 Statistisches Reichsamt, Das deutsche Volks­ einkommen vor und nach dem Kriege, Berlin, 1932. See also Chapter 26, Section l-3f. 26 Gerhard Colm, “Public Revenue and Pub­ lic Expenditure in National Income,” Studies in Income and Wealth, op. cit., vol. I, 1937, p. 186. 27 Colin Clark, The National Income, 19241931, London, 1932, p. 1. 28 M. Matolcsy and S. Varga, The National Income o f Hungary, London, 1938, p. 31. 29 See Chapter 25, Section 11. Abram Berg­ son and Hans Heymann, Jr., in their review of Soviet National Income and Product 194048 (New York, 1954, p. 167), say: “One kind of loss which is not deducted in Soviet figures on total profits comprises the losses of economic organizations covered by subsidies from the government budget”; and these authors also believe that the Russians also fail to deduct from profits the consolidated losses of enter­ prises financed by bank credit (op. cit., p. 170).

NOTES TO CHAPTER 15 1 Hicks, “The Valuation of Social Income,” Economica, 1940. 2 F or a treatment of these reservations, see J. L. Nicholson, “National Income at Factor Cost or Market Prices?” The Economic Jour­ nal, June, 1955. 3 National Bureau of Economic Research, National Product in Wartime, New York, 1945, pp. 20-31; see also Chapter 14, Sections lb, lg for a later reversal of his position. 4 OEEC, National Accounts Research Unit, National Accounts Studies. The publications include separate monographs on France, Swit­ zerland, Sweden, Denmark, the Netherlands, and Norway; and A Standardised System of National Accounts, all issued during the years 1951-55. 5 Simon Kuznets, “National Income: A New Version,” The Review o f Economics and Sta­ tistics, Cambridge, Mass., vol. XXX, 1948, pp. 152-53. 6 Milton Gilbert, E. F. Dennison, George Jaszi, C. E. Schwartz, “Objectives of National Income Measurement: A Reply to Professor Kuznets,” The Review of Economics and Sta­ tistics, Cambridge, Mass., vol. XXX, 1948, p. 180: “We agree entirely on the first point [made by Professor Kuznets]. The mechanism of double entry bookkeeping is a convenient tool for recording economic transactions but it cannot, of course, tell you what kind of transaction you want to record.”

7 Kuznets, “National Income: A New Ver­ sion,” op. cit., p. 154. 8 Milton Gilbert, op. cit., pp. 181-82. 9 Ingvar Ohlsson, On National Accounting, Konjunkturinstitutet, Stockholm, 1953 (Eng­ lish translation), p. 9. i° Ragnar Frisch, Okosirk-systemet, Oslo, 1942 (mimeographed; quotations taken from Ohlsson, op. cit., p. 8). 11R. Ruggles, A n Introduction to National Income and Income Analysis, New York, 1949; also his National Income Accounting and Its Relation to Economic Policy, ECA, Paris, 1949. U.N., A System o f National Accounts and Sup­ porting Tables, 1953. For a fuller bibliography on social accounts see Ohlsson, op. cit., pp. 336-43; and International Association for Re­ search in Income and Wealth, Bibliography (annual series). 12 See his Les Comptes de la Nation, Paris, 1949. 13 For a description of the “Nation’s Eco­ nomic Budget” and its operation in Sweden, see Ohlsson, op. cit., Chapter VIII. 14 OEEC National Accounts Studies: Den­ mark, Paris, 1951, pp. 9-10, 66-67; Norway, Paris, 1953, pp. 13-14, 83-85. 15 Wassily Leontieff et al., Studies in the Structure o f the American Economy, New York, 1953, p. 8. i® Ibid., pp. 8-10.

534

NOTES TO CHAPTER

16

NOTES TO CHAPTER 16 1 See J. B. D. Derksen’s paper on “Inter­ temporal Comparisons of Real National In­ come” and S. Kuznet’s “Long-Term Changes in the National Income of U.S. Since 1870,” in In­ come and Wealth, Series I and II, International Assn. for Res. in Inc. and Wealth, 1951-52. 2 OEEC, National Accounts Studies, Den­ mark, Paris, 1951, pp. 95-98. 8 Matolcsy and Varga, The National Income of Hungary 1924/25-1936/37, London, 1938, pp. 39-41. 4 W. S. Woytinsky (coauthor with Emma Woytinsky of World Population and Produc­ tion, and of World Commerce and Govern­ ments, New York, 1955) in a debate with Colin Clark held over the latter’s predictions made in November, 1953, of a forthcoming serious slump in the United States, referred to the many unrealized predictions of disaster made by some economists in the United States dur­ ing the postwar period. The “danger signs of a slump,” he said, “were first discovered by some American economists . . . in 1943-44. . . . When nothing happened, the prophets of doom postponed the fulfillment of their proph­ ecy first to 1947, then to 1948 . . . [then] for 1950 . . . [then] for 1952 . . . [then for 1953].” He criticized forecasting via the con­ struction of “models” of the operations of the national economy, saying: “The fallacy of mathematical models of the economic system as a means of economic forecasting was con­ clusively demonstrated in the 1940’s. To pull a white rabbit from his hat, the magician must first put the animal there. If you know the trick, you need only look at the systems of equations to know what kind of white rabbits they contain and what ‘assumptions’ must be used in order to let the rabbits loose.” Woytin­ sky explained his own optimistic prognostica­ tion of developments in the United States for the postwar period, which turned out to be more accurate, by his better sense of general trends, saying with respect to the prospects for 1954: “My prognostication of a prompt eco­ nomic upturn rests rather on the faith in the dynamism and flexibility of our economy, the genius of the American people and their ability to take proper decisions at the proper time” (Commercial and Financial Chronicle, New York, January 28, 1954, pp. 4, 54-55). 6 G. Colm, The American Economy in 1960, National Planning Association, Washington, D. C., December, 1952. 6 Potential Economic Growth o f the United States During the N ext Decade (Materials prepared by the committee staff, Grover W. Ensley, staff director), U.S. 83d Congress, 2d session, Joint Economic Committee, Washing­ ton, D. C„ 1954. 7 Ibid., pp. 1-2.

8 Milton Gilbert and Irving B. Kravis, A n International Comparison o f National Products and the Purchasing Power o f Currencies, OEEC, Paris, 1954, pp. 17, 21-28; see also paper by Irving Kravis, “The Scope of Eco­ nomic Activity in International Income Com­ parisons,” Conference on Research in Income and Wealth, New York, October 29-30, 1954, National Bureau of Economic Research. For a criticism of the Gilbert-Kravis approach, see J. L. Nicholson’s book review in the Economic Journal, June, 1955, pp. 352-59. The reviewer denies the possibility of making valid welfare and productivity comparisons between econo­ mies with different relative prices and prod­ ucts; and maintains that such comparisons can be made only in terms of goods that have the same relative prices in the countries or engage the same relative quantities of factors of pro­ duction. 9 Nicholson, op. cit., p. 356. 10 Simon Kuznets, “Comparability of N a­ tional Product Measures Among Different So­ cieties,” Paper presented before the Conference on Research in Income and Wealth, October 29-30, 1954, New York, National Bureau of Economic Research (p. 8 of preliminary re­ port). 11 For a discussion of some phases of this subject see Paul Studenski, Measurement o f Variations in the Fiscal Capacity o f States, United States Social Security Board, Bureau of Research and Statistics, Memo. No. 50, March, 1943, pp. 48-53. 12 Netherlands Central Bureau of Statistics, Statistische en Ekonometrische Onderzoekingen, 1956, no. 3, pp. 125-26. 13 Poland, Central Statistical Office, Statistical Yearbook 1956 (in Polish). 14Pick’s Currency Yearbook 1957, New York, 1957. 15 Planovoie Khosiaistvo, Moscow, August, 1957, p. 76. 16 Voprosy Ekonomiki, Moscow, July, 1957, p. 141. 17 Countries o f Socialism and Capitalism in Figures, Gospolitisdat, Moscow, 1957, p. 107 (in Russian). 18 After this chapter was set in print, an article recently published by S. Kuznets has come to the author’s attention in which a some­ what similar international comparison is made (“Quantitative Aspects of the Economic Growth of Nations,” Economic Development and Cul­ tural Change, October 1956 and July 1957). Many of Kuznets’ conclusions coincide with those presented here. Inasmuch as the two com­ parisons have been made independently, the similarity of their conclusions makes them so much more likely to be true.

N O TES TO CHAPTERS 1 7

TO 1 9

535

NOTES TO CHAPTER 17 1 K. Williams, The National Income of Cey­ lon, Colombo, August, 1952, pp. 15, 16. 2 For example, the British West Indies, India, Malaya. 3 United States, United Kingdom, Canada, Australia, New Zealand, Germany, and Sweden. 4 National Income and Expenditure, 19461952, Central Statistical Office, London, Au­ gust, 1953, p. 74. 5 Ibid., p. 3. 6 Estimates o f Geographical Income and Net Product, 1947-1951, East African Statistical Department, Nairobi, November, 1952, p. 3. 7 Ministry of Finance, National Income and

Expenditure, 1938-1944, Report presented to the Parliament, Dublin, 1946, p. 23. 8 Eidgenoessisches Statistisches Amt, Schweizerisches Volkseinkommen, 1924, 1929, bis 1938, pp. 72, 73, and 89. 9 Kuznets, National Income and Its Composi­ tion, 1919-1938, New York, 1941, vol. II, pp. 502-3; also Table 98, pp. 510-13. 10 Kuznets attributed the lowest margin of error, 7.50 per cent, to net income originating in railroads, telephone, and telegraph; and the highest margin of error, 36.78 per cent, to in­ come originating in real estate (if we omit the miscellaneous category where the error margin was 54.56 per cent).

NOTES TO CHAPTER 18 1 Economic and Social Council Official Rec­ ords, Supplement No. 5B, Statistical Commis­ sion, International Standard Industrial Classi­ fication of all Economic Activities, United N a­ tions, 1948, pp. 3-5. 2 Organisation for European Economic Co­ operation, A Standardised System of National Accounts, Paris, 1952, p. 17. 3 OEEC, National Accounts Studies, Norway, Paris, 1953, pp. 87-99. 4 UN, Statistical Office, Methods o f National Income Estimation, Studies in Methods, Series F, No. 8, New York, January, 1955, pp. 53-54. 6 UN, Methods of National Income Estima­ tion, op. cit., p. 54; OEEC, A Standardised System of National Accounts, Paris, 1952, p. 53.

6 For a discussion of depreciation and its re­ lation to the definition of income, see George F. Break, “Capital Maintenance and the Con­ cept of Income,” Journal o f Political Economy, vol. 42, February, 1954, pp. 48-62. 7 See the discussion on the provisions for the consumption of fixed capital in UN, Methods of National Income Estimation, op. cit., pp. 39-42. 8 See Simon Kuznets, National Income and Its Composition, 1919-1938, National Bureau of Economic Research, New York, 1941, pp. 271-74 and 412; S. Fabricant, Capital Con­ sumption and Adjustment, National Bureau of Economic Research, New York, 1938.

NOTES TO CHAPTER 19 1 Colin Clark and J. G. Crawford, National Income of Australia, Sydney and London, 1938. For a description of the Netherlands estimates see Chapter 23. 2 The degree of breakdown of capital incomes of the various country estimates is shown in Table 4 in the United Nations, Statistical Of­ fice, Statistics of National Income and Expendi­ ture, Statistical Papers, Series H, No. 8, New York, September, 1955. 8 The United States practice is, perhaps, the best example of this method of estimating wages and salaries, since social security systems in 1950 covered 95 per cent of privately paid wages and salaries and about 80 per cent of total wages and salaries. See the discussion in United States Department of Commerce, Na­ tional Income 1954 edition, A Supplement to the Survey of Current Business, Washington, 1954, pp. 68-75.

4 United Nations, Methods o f National In­ come Estimation, Statistical Office, Studies in Methods, Series F, No. 8, New York, January, 1955, p. 12. 6 Organisation for European Economic Co­ operation, A Standardised System of National Accounts, Paris, 1952, p. 76. 6 National Resources Planning Board, Family Expenditures in the United States, 1935-36. 7 For a discussion of farm income and sources of data used see Chapter 18, Section 3a. 8 UN, Methods o f National Income Estima­ tion, op. cit., p. 19. 9 Simon Kuznets, National Income and Its Composition, 1919-1938, New York, 1941, vol. H, pp. 413-15. 10 UN, Methods o f National Income Estima­ tion, op. cit., p. 19. 11 United States, National Income 1954, op. cit., pp. 44-45.

536

NO TES TO CHAPTERS 1 9

12 Ibid., p. 45. 13 Ibid., p. 45. 14 A Standardised System of National Ac­ counts, op. cit., p. 78. 15 UN, Methods of National Income Estima­ tion, op. cit., pp. 16 and 48. 16 This discussion is based on United States, National Income 1954, op. cit., pp. 97-103. See also the summary of U.S. interest estimates in UN, op. cit., pp. 17-18.

TO 2 1

17 Statutory tax exemptions, reported for 1933-34, had to be projected to other years; all other exemptions had to be estimated for all years. 18 Salaries and wages subject to withholding tax based on tax proceeds for certain years. 10 United States, National Income 1954, op. cit., p. 92.

NOTES TO C H A PTER 20 1 Colin Clark, A Critique of Russian Statis­ tics, Macmillan, London, 1939, and Julius Wyler, “National Income Estimates of Soviet Russia,” Social Research, December, 1946. 2 Raul Simon, Determination de la Entrada Nacional [National Income] de Chile, Santiago, 1935. 3 Simon Kuznets, Commodity Flow and Capital Formation, vol. I, National Bureau of Economic Research, New York, 1938. 4 Wassily Leontieff, The Structure of Am eri­ can Economy, 1919-1929, Cambridge, H ar­ vard University Press, 1941. 5 J. E. Meade and R. Stone, “The Construc­ tion of Tables of National Income, Expendi­ ture, Savings and Investment,” The Economic Journal, June-September, 1941. 6 Central Statistical Office, National Income and Expenditure 1946-1953, London, 1954, Table 1, pp. 2-3; U.S. Dept, of Commerce, National Income 1954, Table 4, pp. 164-65. 7 These classifications do not apply, however, to estimates based on the restricted material or market production concepts. See, for example, the prewar Polish estimates, Chapter 34, Sec­ tion 11-12). 8 In the Swedish estimates up to 1934, the value of the services of consumer durables was included with consumer expenditures, and the purchase of new consumer durables was deemed part of capital formation.

9 The prewar Polish estimates were not ad­ justed for international payments of this kind, but the payments were relatively small (see Chapter 34, Section 11-12). 10 OEEC, National Accounts Studies, Den­ mark, 1951. 11 OEEC, A Standardised System o f National Accounts, Paris, 1952, pp. 23, 58-62. 12 Until 1946, the estimates for the United Kingdom gave net capital formation directly. 18 Commonwealth of Australia, National In­ come and Expenditure 1953-54, Canberra, 1954, p. 10. 14 UN, Concepts and Definitions o f Capital Formation, Series F, No. 3. 15 OEEC, Standardised System, op. cit., pp. 25, 69-70. 16 UN, Methods o f National Income Estima­ tion, Series F, No. 8, 1955, p. 29. 17 OEEC, Standardised System, op. cit., pp. 21, 63-66. 18 UN, Methods of National Income Estima­ tion, Series F, No. 8, p. 24. 19 U.S. Department of Commerce, National Income 1954, op. cit., p. 115. 20 Clark, op. cit. 21 UN, op. cit., Series F, No. 8, p. 36. 22 OEEC, National Account Studies—■Nor­ way, Paris, 1953, pp. 107-9.

NOTES TO CHAPTER 21 1 Patrick Colquhoun, Treatise on the Wealth, Power, and Resources of the British Empire, London, 1814, p. 313; Pablo Pebrer, Taxation, Revenue, Expenditure, Power, Statistics and Debt of the Whole British Empire, London, 1833, table facing p. 386; Sir Robert Giffen, Economic Inquiries and Studies, London, 1904, vol. 2, p. 366. 2 R. H. Coats, National Wealth and Income of Canada, Monetary Times Annual, Ian. 3, 1919; Canada Year Book, Dominion Bureau of Statistics, Ottawa, 1925, p. 816, and subsequent volumes, chapter on National Wealth and In­ come; National Income of Canada, Ottawa,

1934; National Income of Canada, 1919-1938, Part I, Ottawa, 1941; and since 1943, Economic Conditions in Canada and M onthly Review of Business Statistics. These publications contain articles on national income with methodological comments. See also Major A. R. Lawrence, The Canadian Income, Its Sources, Distribution and Expenditure by Counties, Census Divisions or Districts, Toronto, 1933. 3 Bank of Nova Scotia, Monthly Review, No­ vember and December, 1935; May, 1937; July, 1938; September, 1940; and June, 1941. 4 Royal Commission on Dominion-Provincial Relations (the Rowell-Sirois Commission); Na-

NO TES TO CHAPTERS 2 1

tional Income, by D. C. MacGregor, J. B. Ruth­ erford, G. E. Britnell, and J. J. Deutsch, Ot­ tawa, 1939; and Dominion-Provincial Confer­ ence, National Income, 1937-1940, January, 1941. 6 Dominion Bureau of Statistics, National In­ come of Canada, 1919-1938, Parts I and II, Ottawa, 1941. 6 Dominion Bureau of Statistics, National Ac­ counts, Income and Expenditure. Reference Book for Dominion-Provincial Conference, Ot­ tawa, 1945. For an account of the circumstances surrounding these innovations see Simon A. Goldberg, “Development of National Accounts in Canada,” Canadian Journal of Economics and Political Science, XV, 1, 1949, pp. 34-52.

TO 2 3

537

7 See Reference No. la, pp. 96-99. 8 See Reference No. la, pp. 87-91. 9 See Reference No. la, pp. 91-96, 115-16. 10 For details of the estimating procedure see Reference No. la, pp. 92-93. 11 See Reference No. la, pp. 97-98. 12 See Reference No. la, pp. 121-22. 13 See Reference No. la, pp. 99-115, 122-24. 14 See Reference No. la, pp. 103-8. is Ibid., pp. 108-9. i® Ibid., pp. 109-14. 17 O. J. Firestone (University of Toronto Press, 1951). i g See Reference No. la., p. 114. 19 Ibid., pp. 124-27.

NOTES TO CHAPTER 22 1 Since 1955 various other inquiries have been used in place of the former annual No­ vember sample. 2 Since subsistence farming plays an impor­ tant part in the Irish economy, the Central Sta­ tistics Office publishes figures for farm house­ hold consumption at full retail prices for pur­ poses of comparisons of per capita national in­ come with other countries. (See table in righthand column.) 3 See note 1, above. 4 This paragraph was based on the commu­ nication from R. C. Geary.

Farm Produce and Fuel Consumed on Farms W ithout Process of Sale, for Selected Years, 1938, 1949, 1953, and 1954 (million £) 1938 1949 1953 1954

Agricultural Prices 13.2 30.0 34.9 33.3

Retail Prices 21.5 45.6 52.5 49.2

s o u r c e : Irish Statistical Survey, 1953, p . 57 (for years 1938 and 1949); figures for 1953, 1954 from communication from R. C. Geary.

NOTES TO CHAPTER 23 1 a. W. A. Bonger, “Vermongen en Inkomen in Nederland,” De nieuwe Tijd, vol. XV, March, 1910; ibid., vol. XX, April and May, 1915. b. , Vermongen en Inkomen in Neder­ land gedurende den oorlogstijd, 1913-1920 (Wealth and National Income of the Nether­ lands During the War Period), Amsterdam, 1923. c. , “Het nationale inkomen van Neder­ land,” D e Socialistische Gids, vol. 16, March, 1931, p. 227. Bonger was not the first income estimator, since an official publication refers to older literature on national income and wealth estimates listed in C. A. Verrijn Stuart, Inleiding tot de Statistiek, vol. Ill, Part 2, p. 310. (Enkele Berekeningen over het Nationale Inkomen van Nederland, De Nederlandsche Conjunctuur, Speciale Onderzoekingen, No. 2, p. 7.) 2 a. J. J. Bekaar, “Op Zoek N aar ons Nationaal Inkomen,” Algemeen Handelsblad, De­ cember 12 and 14, 1924. b. , “Nationale Inkomen,” De groene Amsterdammer, October 15, 1927. c. C. J. P. Zaalberg, “De Opbrengst van den arbeid in Nederland” (The Product of Labor

in the Netherlands), Economisch-Statistische Berichten, 1931. d. M. J. H. Smeets, Inkomen en vermogen en hun beteekenis voor de welvaart, Amster­ dam, 1931. 3 Sub-Commissien uit de comissie voor de economische politiek (Sub-Commission of the Commission for Economic Policy), Inkomen en vermogens in Nederland, 1929. 4 See References la, b, c, and d. 5 The estimates for the period before World War II and during the period of the war are described in a series of articles by Dr. Derksen and Professor Tinbergen in various issues of the monthly review of the Netherlands Central Bu­ reau of Statistics, e.g., M ay-June, 1940, August, 1940, February, 1941, March-April, 1941, and February-M arch, 1942; and also in several books published in Holland during the period such as the ones by C. P. de Groot in 1936, J. D. Kruisheer in 1937, and J. Van der Wijk in 1939. 6 Special studies on the net output of con­ struction were available, namely M. J. de Bosch Kemper, “Bouwvolume en werkloosheid

538

NOTES TO CHAPTER 2 4

in het bouwbedrijf,” De Nederlandsche Con­ junctuur, May, 1933; and C. Th. J. Delfgaauw, “De beteekenis van het bouwbedrijf voor het economisch leven van Nederland,” EconomischStatistisch Maandbericht, March, 1939. 7 About the usefulness of the various tax statistics in the Netherlands, see C. P. de Groot, Methoden voor de berekening van het nationale inkomen, Purmerend, 1936, pp. 101-28. 8 The results of this computation agreed closely with the number of children reported in the income tax statistics. 9 This group includes mainly agricultural la­ borers, farmers, youthful workers in industry and trades, fishing and river shipping, small shopkeepers, domestic servants, and unem­ ployed workers. 10 Reports of the Pensioenraad, Statistical Yearbook of the Netherlands East Indies (Re­ tired Civil Servants of the East Indies Govern­ m ent), and death duty statistics. 11 Ph. J. Idenburg and J. B. D. Derksen, Reference No. lc, p. 22. 12 Profits included capital gains and losses, but for lack of available statistical material the latter could not be eliminated. 13 This method was first applied by M. J. de Bosch Kemper, “Bouwvolume en werkloosheid in het bouwbedrijf,” De Nederlandsche Conjunctuur, May, 1933. 14 Reference No. lb, p. 33. 15 Business cost taxes were also deducted. 16 The estimate of entrepreneurial income was said to be fair for the years after 1931, but too low for the earlier years. 17 This treatment of banking in national in­ come differs from the method adopted by the United States Department of Commerce, which imputes all services rendered free by banks to depositors (not to borrowers). This may partly

be explained by the different nature of banking in the Netherlands. Private persons do not have accounts with banks in general, and checking facilities are not provided to the same degree. 18 This statement in the official publication, however, is not confirmed by the fact that the wages and salaries paid in ecclesiastical services equal the total income, namely 25 million guilders for each year (see Enkele Berekeningen over het Nationale Inkomen van Neder­ land, op. cit., pp. 56 and 63). The interest on debts of churches is estimated elsewhere at 5 million guilders. 19 The wage and salary figures for 1903-20, however, were incomplete, since they excluded wage incomes in agriculture, shipping, and pro­ fessional and domestic services. 20 J. B. D. Derksen and J. Tinbergen, Berekeningen over de economische beteekenis van Nederlandsch Indie voor Nederland, Maandschrift, 1945. (An Evaluation of the Economic Value of the Dutch East Indies for the Nether­ lands.) 21 Ed. van Cleeff, “Nationale boekhounding: proeve van een jaaroverzicht Nederland 1938,” De Economist, luly, 1941, pp. 415-24. (Na­ tional Bookkeeping: an Attempt at an Annual Survey for the Netherlands, 1938); “Beteekenis en inrichting eener nationale boekhounding,” De Economist, November, 1951, pp. 608-23. (On the significance and the Organization of a National Book-keeping System.) 22 For a summary in English see J. B. D. Derksen, A System o f National Book-keeping illustrated by the experience o f the Netherlands economy, Occasional Papers X, National In­ stitute of Economic and Social Research, Cam­ bridge University Press, 1946. 23 Reference No. 2c.

NOTES TO CHAPTER 24 1 The estimates for the United States pre­ pared by the National Industrial Conference Board covered a longer period, going back to 1799, but they were organized by decades for the period 1799-1899 and were rather crude. Kuznets’ historical series for the United States, prepared in 1946 and extended in 1952, covered the period 1869-1948 and were also organized by decade averages and were similarly only approximate in character (see note 1, Chap. 16). 2 Reference No. 5, pp. 245-46 and 273-74. 3 The balance of exports and imports was derived from statistics of foreign trade. Since exports are valued f.o.b. Swedish ports and the corresponding produced goods are valued at the place of production, the value of the ex­ ports was reduced by the transportation and merchandising charges to the ports, which gen­ erally amounted to 10 per cent. The values of

the imports included import duties. Transpor­ tation and commercial services rendered for exporting goods and in overseas shipping was considered as export of services. 4 Reference No. 1, Part I, p. 173; Part II, p. 442. 5 Ibid., Part II, p. 553. 6 Ibid., Part I, p. 55. 7 Reference No. 1, Part II, pp. 73-79. 8 For instance, one ton of roots was assumed equivalent to 100 food units, one ton of dredge was assumed equivalent to 900 food units, etc. (Reference No. 1, Part II, p. 121.) 9 It is worth mentioning that these averages rose from 830 kg. in 1861 to 2,180 kg. in 1930. 10 Sveriges Tradgardsodling, Arsbok, 1923. 11 This summation was incomplete as logs used in building and construction work or worked in small sawmills and wood fuel used in industry were excluded.

N O TES TO CHAPTER 2 5

12 Agricultural statistics, information given by the National Defense Committee and the Tariff and Treaty Committee, liquor control statistics, industrial statistics. 13 The export prices were adjusted by the ratio between export prices and domestic prices available in the years after 1896. 14 The production of cheese and butter on farms was included in the estimates for milk. 15 Handicraft was excluded from this com­ putation, as the allocation of the value of its gross product was estimated separately accord­ ing to assumed percentages. 16 They are: (1) mining and metal indus­ tries, (2) stone, clay, and glass industries, (3) wood products industries, (4) paper and print­ ing industries, (5) food products industries, (6) textile and clothing industries, (7) leather, hair, and rubber industries, (8) chemical in­ dustries, (9) power stations, gas works, water works. 17 Reference No. 1, Part II, pp. 244-45. 18 The relation between gross and net value of construction is shown in Table 24-1, Part C. 19 It should be noted that trade excludes wholesale trade in products sold to domestic

539

industries. However, an amount was added for the income from export trade. 20 The state’s share in the profits of alcoholic liquor trade, certain turnover taxes, excise duty in spirits, and tobacco tax. 21 Reference No. 2a, p. 216. 22 For Stockholm, Gothenburg, other towns, agrarian and other rural districts (Reference No. 1, Part II, Appendix). 23 Ibid., Part I, p. 162. 24 Ibid., Part II, pp. 547-48. 2' Ibid., Part II, p. 549. 2« Ibid., Part II, p. 550. 27 In the prewar estimates for Germany the value of services from government were simi­ larly split into services rendered to producers and to consumers (see Chapter 26, Section 3f). 28 Reference No. 1, Part I, p. 253. 29 Reference No. 2b, p. 4. 30 Reference No. 5, p. 246. 31 Reference No. 4, p. 15. 32 This section and those following are based mainly on the OEEC publication for Sweden, Reference No. 4. 33 Reference No. 5, p. 246. 34 See the discussion in Reference No. 5, pp. 247—49.

NOTES TO CHAPTER 25 1 Storch’s biography and list of statistical works are given in Chapter 6, Section 4c. His Cours d’Economie Politique was published orig­ inally in six volumes in St. Petersburg in 1815. It was followed by another edition in four vol­ umes, annotated by J. B. Say, published in Paris in 1823; and by a German translation in three volumes annotated by K. H. Rau, published in Hamburg in 1819-20. A series of papers in French on the nature of income, theory of value, and nature of national income were published by the Akademia Nauk, St. Petersburg, in its Memoires, ser. V, vols. 2, 3, 5, and 8; 1807-12; 1817-18. A pamphlet in French, “Considerations sur la Nature du Revenu National,” was published in Paris in 1824. 2 One of the important Russian works in political economy of that period was the brief treatise by Professor N. Bunge (later Minister of Finance and reformer of Russia’s tax sys­ tem) published in St. Petersburg in 1870. It contained 2 pages on the theory of national income, reflecting the ideas of Storch, Say, and Rau, but no reference to the methods of esti­ mating national income. The works of Gorloff and Danielson did not go even that far. 8 Leone Levi, “On the Distribution and Pro­ ductiveness of Taxes with Reference to the Prospective Amelioration in the Public Rev­ enue,” Journal of the Royal Statistical Society, vol. XXIII, 1860, p. 42. Levi estimated Russia’s national income to be £ 4 0 0 millions, or £ 6

(40 rubles) per capita. He did not explain how he arrived at this figure. For M. G. Mulhall’s figures see Chapter 9, Section 7 (as taken from his Dictionary o f Statistics, London, 1884; 4th edition, London, 1899; and his Industries and Wealth o f Nations, London, 1896). See also V. I. Pokrovsky, Concerning the Stability of Russia’s Favorable Balance o f Trade (in Rus­ sian), St. Petersburg, 1897, Chapter 8. 4 “Opyt Istchislenia Narodnogo Dokhoda” (“Attempt at an Estimate of the National In­ come”), Trudy Volno Ekonomitcheskogo Obtchestva, 1906, II, pp. 20-35 (in Russian). 5 S. N. Prokopovitch, Voina i Narodnoie Khosiastvo, Moscow, 1917, 2d ed., 1918 (in Russian). 6 The official estimates appeared in various publications of the Gosplan (State Planning Commission) and the TZUNHU (Central Ac­ counting Administration of the National Econ­ om y), such as Kontrolnyie Tzifry and Socialistitcheskoye Stroyitelstvo. The 1928-29 vol­ ume of Kontrolnyie Tzifry contained national income estimates in detail. Of the special works on the subject, all in Russian, and using offi­ cial figures, the most important are: Litoshenko, Narodny Dokhod (National Income), U.S.S.R., Narkomfin, Moscow, 1925; S. G. Strumilin, Narodny Dokhod (National Income), Moscow, 1929. See also reference items 1-6 in the Ref­ erences at the head of this chapter. 7 S. Turetsky, “Dynamics of Prices,” Planovoie Khosiaistvo, 1939, No. 1, pp. 122-26; V.

540

NO TES TO CHAPTER 2 5

Katz, Planovoie Khosiaistvo, 1934, No. 7, pp. 124-25. 8 See Reference No. 7. 9 See Reference No. 8. 10 “Soviet National Income and Product in 1937,” Quarterly Journal of Economics, 1950, pp. 208-41 and 408-11. 11 “Intricacies of Russian National Income Indexes,” Journal of Political Economy (Chi­ cago), vol. 55, 1947, pp. 299-322; “The Econ­ omy of the Soviet Union,” American Economic Review, Proceedings, May, 1951, pp. 484-89. 12 See Reference No. 11. 13 “National Income and Product of the U.S.S.R. in 1940,” Review o f Economic Statis­ tics, Cambridge, Mass., vol. 29, 1947, pp. 22634. 14 “Volkseinkommen und Sozial Produkt der Sovietischen Volkswirtschaft,” Zeitschrift fiir Schweiz. Volkswirtschaft und Statistik, vol. 46, 1950, pp. 49-64. 15 Reference No. 2, pp. 3, 11. 16 A. B. Batcharin and Associates, Finances and Credit in U.S.S.R., Gosfinansisdat, Mos­ cow, 1955, p. 36. It is further claimed that “in the U.S.S.R. the entire national income be­ longs to the laboring class . . . [whereas] in capitalist countries more than one half of the national income is appropriated by the exploit­ ing classes and the share of the laboring class in the national income constantly diminishes.” The latter part of this statement is completely contradicted by statistics in western countries (e.g., see Chapter 16, Section 5c). 17 Reference No. 3, pp. 67-71 and 80-81. 18 Reference No. 5, p. 13. 19 For a fuller discussion of the Soviet con­ cept of national income, see Reference No. 9. 20 V. Katz, Planovoie Khosiaistvo, op. cit., pp. 126-27. 21 For a statement of the arguments on both sides of the issue, see Reference No. 2, pp. 55-58 and 108-20. 22 Ibid., p. 129. 23 The annual depreciation on plows and other cultivating machines as well as on wheat threshers was computed at 10 per cent of their original cost, that on harvesting machinery at 12 per cent, and that on combines at 14 per cent. The Chernomordik study pointed out (Reference No. 2, p. 125) that these depre­ ciation charges no longer corresponded to the actualities, inasmuch as substantial changes had occurred in the machines and their uses. Lower depreciation charges were urged. 24 Ibid., p. 126. 25 Reference No. 2, p. 56. Up to 1937 it was customary to include in gross value all unfinished production. This was said to have tended to undermine the effectiveness of the state’s production programs, for the industries were inclined to accumulate large inventories of idle raw materials and to leave much of their production in an unfinished state. As the supply of goods to other productive enterprises

and consumers was thereby being retarded, the government issued a regulation that all pro­ duction programs should be made in terms of finished output. 26 Quoted in Reference No. 2, from Problemy Planirovania, Moscow, 1932, pp. 397, 398. 27 Reference No. 2, p. 75. 28 Ibid., p. 186. 29 Ibid., p. 221. 30 According to the Chernomordik study (Reference No. 2, pp. 87-88), “From an eco­ nomic point of view the turnover tax and the socialist profit are identical in nature. Both are produced by collective socialized labor. Both constitute the portion of the national income that is accumulated for investment and other social needs. . . . Profits increase not only with output but also with labor efficiency and lower costs. The purpose of allowing enter­ prises to retain a portion is to afford an addi­ tional stimulus to efficiency and cost reduction. Though only a portion of profit is paid into the national budget, the entire proceeds of the turnover tax are made available to it. The with­ drawal of a portion of the ‘accumulatable’ in­ come in the form of the turnover tax ensures the fulfillment of the revenue estimates of the budget, and, hence, also of its expenditure, and of the entire economic program of the nation. The tax is also an instrument of supplementary financial control over the entire system of pro­ duction and distribution.” 31 F or a fuller description of this tax see Alexander Baykov, Development o f the Soviet Economic System (Cambridge University Press, 1945), Chapter XIX. 32 Reference No. 3, p. 103. ™lbid., p. 105. 34 For a more complete analysis of these methods and their results, see Reference No. 10; also (in Russian) V. Katz, Planovoie Khosiaistvo, 1934, No. 7; A. Pervuhin, Plano­ voie Khosiaistvo, 1934, No. 4; M. Kolganoff, Problemy Ekonomiki, 1937, Nos. 3-4; S. N. Prokopovitch, Quarterly Bulletin o f Soviet Rus­ sian Economics, March, 1941; and his Russland’s Volkswirtschaft unter den Soviets, Zur­ ich, 1944. 35 See Reference No. 10. 36 S. N. Prokopovitch suggested the use of a 1926-27 cost of construction per square foot of floor space or cubic foot content as a usable constant price, but it obviously was a highly inadequate and imperfect measure. (See Ref­ erence No. Id , Chapter VI.) 37 Kolganoff, Problemy Ekonomiki, op. cit., p. 108, and Turetsky, Planovoie Khosiaistvo, op. cit., p. 126. 38 Reference No. 11 b. 39 Reference No. 8a. 40 See Maurice Dobb, Soviet Economy and the War (London, 1941), pp. 24, 31-32. 41 Reference No. 11a. 42 Reference No. 116, pp. 30-32.

NO TES TO CHAPTERS 2 5

43 Abram Bergson and Hans Heymann, Jr., Soviet National Income and Product 1940-48, New York, 1954; Oleg Hoeffding, Soviet Na­ tional Income and Product, 1928, New York, 1954.

TO 2 7

541

44 1928 figures from Hoeffding, op. cit.; 1937-1948 figures taken from Bergson-Heymann, op. cit., pp. 70-71. 45 Ibid., pp. 70-71.

NOTES TO CHAPTER 26 1 Karl Diehl, V erein fiir Sozialpolitik, Schriften, 1926, vol. 173. 2 Treaty of Versailles, Part III, Annex II, 12 (b ). 3 Reference No. 1. 4 Ferdinand Gruenig, Der Wirtschaftskreislauf, Berlin, 1933. 5 Including insufficiently reported rent on farmers’ dwellings and value of the farmers’ consumption of own products. 6 For technical reasons the incomes from land reserved to retired farmers (altenteile) and social security benefits were estimated to­ gether with the tax-exempt incomes. 7 The percentage for corporations was de­ termined by an examination of the ratio be­ tween dividends and net profits reported in the statistics of corporations and a comparison of these profits with the profits reported for tax purposes. The same percentage was applied to co-operatives and similar enterprises, while the profits of other institutions subject to the cor­ poration tax (autonomous estates, societies) were fully included in undistributed profits. 8 The ratio increased steadily from one third to one half during the period 1925-31. 9 To this interest a small amount was added for interest derived from savings in foreign banks. 10 Mortgages held by individuals included also individual loans to business. 11 The official rates for preinflation mort­ gages were 1.8 per cent in 1925, 3 per cent in 1926 and 1927, and 5 per cent from 1928 to 1931. For new mortgages, 7.5 per cent was considered to be the average interest. 12 This adjustment excludes the interest on

German bonds issued abroad and on foreign savings in German banks and German savings in foreign banks, which were deducted in the beginning. 13 These estimates were reported in: Statis­ tisches Jahrbuch fiir das Deutsche Reich, 1938, p. 560. Statistisches Reichsamt, Wirtschaft und Statistik, 1933, p. 686; 1934, p. 610; 1935, p. 622; 1936, p. 690; 1937, p. 738; 1938, pp. 2, 802; 1939, p. 705. “(jber die Moglichkeiten Einer Kurzfristigen Einkommens Beobachtung,” Vierteljahreshefte zur Statistik des Deutschen Reichs, 1933, Heft 4. “Das Deutsche Arbeitseinkommen in der Wirtschaftlichen Erholung,” 1934, Heft 3. Wandlungen in der Schichtung der Arbeiter- und Angestellteneinkommen von 1929-1936, 1937, Heft 3. Reference No. 3, Jostock, 1940, pp. 766-74. 14 Wirtschaft und Statistik, op. cit., 1938, p. 850. Wilhelm Bauer and Peter Dehen, “Landwirtschaft und Volkseinkommen,” In­ stitut fiir Konjunkturforschung, Vierteljahres­ hefte, 1938-39, pp. 411 ff. 15Iostock, op. cit., p. 770; see also Wirt­ schaft und Statistik, 1939, p. 707. 16 Maxine Y. Sweezy, The Structure o f the Nazi Economy, Cambridge, Mass., 1941, p. 199. 17 Donner, “Die Grenzen der Staatsverschuldung,” Weltwirtschaftliches Archiv, September, 1942. H. W. Singer, “The Sources of War Fi­ nances in the German War Economy,” Review of Economic Studies, 1943, also in London and Cambridge Economic Service Bulletin, 1943, p. 106. 18 Wirtschaft und Statistik, 1939, p. 565.

NOTES TO CHAPTER 27 1 This section is based mainly on Reference No. 4b, pp. 1-9. 2 This section is based mainly on Reference No. 2 (OEEC), pp. 121-28; and Reference No. 1. 3 Reference No. 2 (OEEC), p. 102. 4 This section is based mainly on the OEEC report (Reference No. 2), pp. 134-37. 6 OEEC report, op. cit., p. 136. 6 This section is based on Reference No. 2, pp. 113-20, 129-34; and Reference No. 1, pp. 17-41. 7 Reference No. 2, pp. 116-17.

8 Ibid., p. 118. 9 Reference No. 2, p. 130. Ibid., p. 131. 11 Ibid., p. 131. 12 This section is based mainly on Reference No. 4a, pp. 20-23 and 102-14; Reference No. 4b, pp. 9—11 and 17-19. 13 Reference No. 4a, p. 17. 14 Reference No. 4a, pp. 102-4. 16 Reference No. 4a, pp. 107-8. 16 Reference No. 4a, pp. 109-10. 17 Reference No. 4a, pp. 113-14. 18 This explanation was furnished by Mr.

542

NOTES TO CHAPTERS 2 7

Claude Gruson, Chief of the Research Unit of the Ministry of Finance, in a letter to the au­ thor dated September 20, 1955. 19 This section is based mainly on Reference No. 4a, pp. 23-24 and 114-19. 20 This section is based on Reference No. 4a and Reference No. 5.

TO 2 9

21 Reference No. 4a, p. 20. 22 This section is based mainly on Reference No. 4b and Reference No. 4c. 23 Reference No. 4a, p. 82, Reference No. 4b, pp. 18 and 19, and Reference No. 4c, pp. 114, 118-21. 24 Reference No. 4d.

NOTES TO CHAPTER 28 1 Reference No. 4a. 2 Reference No. 7. 3 For instance, Novi Sad for grains, Spalato for olive oil, etc. 4 Reference No. 4a, p. 286. e Ibid., p. 287. eibid., p. 288. 7 Ibid., p. 290.

8 Ibid., p. 291. The value of products handled by trade was fixed at the following percentages of total gross value of the output or imports: plant production 30 per cent, animal husbandry 50, forestry 75, mining 100, rural industry about 30, artisans 33, factory industry 100, and imports 80 per cent. 9 Reference No. 8.

NOTES TO CHAPTER 29 1 Reference No. 3, p. III. 2 Reference No. 1. 3 The first as well as the second estimate was said by the estimators to relate to the agri­ cultural year running from July to June. How­ ever, only the agricultural output was estimated for this period, while budgetary data (indirect taxes, salaries of government employees) in fact referred to the fiscal year (June to M ay), and the rest of national income referred to the calendar year. 4 Reference No. 2. 5 Reference No. 3. 6 Reference No. 4. 7 Reference No. 5a and 5b. 8 Although it was repeatedly stated by the estimators that their estimates related to the net national income, the value added in in­ dustry and the output in trade had been com­ puted as gross items including indirect taxes, interbusiness services, expenses for certain ma­ terials, and depreciation allowances. 9 Reference No. 2, p. 48; in the estimates for 1942-44 different per cent allowances for the costs were made, namely: 20 per cent for field crops and vegetables and 10 per cent for fruits (Reference No. 3, pp. 32-34). 10 It should be noted that the calculated net changes in stocks were not included in the na­ tional income estimates. However, in the esti­ mates for 1942-44 the production of meat was estimated on the basis of data on the slaughter­ ing of animals at slaughterhouses and outside of them. The figures thus obtained were cor­ rected for any increase in the livestock during the period covered. No adjustments were made for the decreases in the number of livestock on the theory that these would be reflected in a corresponding decrease in the next year’s in­ come (Reference No. 3, p. 25).

11 Reference No. 3, p. 63. 12 Reference No. 3, pp. 31, 63. 13 Reference No. 4a, pp. 29, 31. 14 The definition of gross value is given in note 8, above. 16 Reference No. 3, pp. 67-85. 16 These profits of government monopolies and indirect taxes include the following items: monopoly profits on tobacco, salt, alcohol, matches and cartridges, customs duties, trans­ action, consumption, and franchise taxes, and commerce’s share in the income tax, the wheat protection tax and all provincial, municipal, and chamber of commerce charges. It will be noted that the deduction of these taxes and monopoly profits does not leave net output in trade, be­ cause no deduction was made for interbusiness services and depreciation. Neither does it lead to gross national income at factor cost, since certain indirect taxes are still left in the gross value added in industry. 17 Reference No. 2, p. 126. The percentages applied were as follows: 1933-34 Agriculture (including forestry) 75.0 Industry 100.0

1934-35

1935-36

73.3 97.7

71.3 95.0

They were determined by extrapolating assumed similar percentages for 1927 on the basis of the changes in wholesale and retail prices. The procedure was rather rough since the number of commodities whose prices were actually studied at different levels of processing was very small. 18 Reference No. 3, pp. 97-122. 19 Reference No. 3, p. 166. 20 The income of professions was roughly

NO TES TO CHAPTER 3 0

estimated on the basis of the statistics of the income tax. 21 Reference No. 5b and c. 22 Other sources were: new income tax statis­ tics (available since 1950); improved agricul­ tural statistics on quantities produced and es-

543

pecially on prices received by farmers; labor statistics; several agricultural surveys taken dur­ ing recent years; and data provided by some newly created official agencies such as the Soil Products Office, Meat and Fish Organization, FAO National Committee, etc.

NOTES TO CHAPTER 30 1 V. K. R. V. Rao listed the earlier estimates and examined some of these carefully. The names of the estimators and the years for which estimates were made were given as fol­ lows: for British India, Dadabhai Naoroji, 1868; Baring and Barbour, 1882; F. G. At­ kinson, 1875, 1895; Lord Curzon, 1897-98; William Digby, 1899; Sir B. N. Sarma, 1911; G. Findlay Shirras, 1921, 1922; and for all of India, Shah and Khambhata, 1921; Wadia and Joshi, 1913-14; Vakil and Muranjan, 1925. Rao, 1939, Reference No. 1, pp. 1-45, 164-70. 2 G. Findlay Shirras, Poverty and Kindred Economic Problems in India, London, 1931, 2d ed., Delhi, 1936; Science of Public Finance, London, 1936, 3d ed., vol. I; “India’s National Income,” Revue de I’Institut International de Statistique, 4ieme Annee, 1936; Federal Fi­ nance in Peace and War, London, 1944. V. K. R. V. Rao, References No. 1 and No. 2. The year covered was not exactly defined by the estimators. It mainly refers to the agricultural year, may vary with the nature of the crop. 3 The following figures show the population involved.

1921 1931 1941

British India Burma (in millions) 250.2 13.2 274.8 14.7 295.8 16.8

Tot al 263.4 289.5 312.6

The Statesman’s Year-Book, 1943, p. 133; Statistical Year Book of the League of Nations, 1942-44, pp. 16 and 18. The figures refer to the British Provinces only, excluding the inde­ pendent Indian States. 4 Reference No. 4, p. 5. 8 “When a country has a favourable balance of trade,” wrote Rao (Reference No. 2, p. 25), “it means that a part of its production ceases to be available for the country; and the value of this part should be excluded from a compu­ tation of the country’s net income. Of course, this favourable balance may really be due to an export of capital and, therefore, offset by an import of securities. In such a case, the favourable balance does not really represent the loss of income for the country. In addition, there may either be an increase of the country’s indebtedness abroad or a diminution of its credit balances abroad, which really should be

counted as an item for deduction in arriving at net income.” 6 Rice, wheat, sugar cane, jowar, grain, bar­ ley, corn sugar, tea, coffee, tobacco, indigo, and rubber. 7 Other oil seeds, other fibers, fruit, vege­ tables, root crops, spices, and miscellaneous crops. 8 V. R. Gaitonde, Village Tanning. 9 The average price per head was determined in a discussion with Professor Gangulee, mem­ ber of the Royal Commission on Agriculture. The results were checked against data on con­ sumption available in family budget studies. 10 Reference No. 2, p. 106. 11 Reference No. 2, p. 85 (quotation from a report of the Royal Commission on Agri­ culture). 12 This is illustrated by the following figures:

Milk produced Milk consumed

Millions of Pounds 1925-29 1931-32 36.0 36.6 38.3 38.5

13 The NSS was established in 1950 as a continuing sampling organization with the prime objective of quick collection and tabu­ lation of data needed by the government. 14 Reference No. 4, p. 46. 15 Reference No. 3, p. 21. 16 Reference No. 2, p. 150. 17 This average was 6 rupees against 109 in the cities. 18 Reference No. 4, p. 100. 19 Article by G. D. Birla, “Notes on the General Situation of India,” Economic Journal, September, 1932, p. 485. 20 Figures for 1931-32 from League of Na­ tions, Balance o f Payments, 1933, pp. 99-100. 21 Mukherjee and Ghosh, Reference No. 1, p. 60. 22 In January, 1957, for the first time, a Con­ ference on National Income and Wealth was held in India, arranged by India’s Central Sta­ tistical Office. Papers on many aspects of the subject were presented by representatives of the central and provincial statistical offices, univer­ sities, and trade organizations. That such a con­ ference was held is one indication of the grow­ ing interest in the subject in that country, as well as of the great gains being made there in the organization of economic statistics.

544

NOTES TO CHAPTERS 3 2

TO 3 7

NOTES TO CHAPTER 32 1 Benham, pp. 163-65. 2 Entertainment duty was subtracted. 3 The value of these imports did not include their costs of distribution, since these costs were also omitted in the net output estimates. 4 The two estimates were not independent (and, therefore, one is not a reliable check on the other) since some of the expenditure esti-

mates were taken from net output estimates. 5 Benham, p. 161. 6 The retail value of rice had to be increased by the amount of the subsidy, since the actual retail value and the subsidy were available for distribution as incomes to rice producers. 7 This adjustment could not be made for other consumer goods because of lack of data.

NOTES TO CHAPTER 34 1 E. F. Jackson, “Social Accounting in East­ ern Europe,” International Association for Re­ search in Income and Wealth, Income and Wealth Series IV, Bowes & Bowes, London, 1955, p. 256. 2 Plan for the Development of the Economies

o f the Peoples’ Democracies (in Russian). Col­ lation of Materials, Isdatelstvo Inostrannoi Litteratury, Moscow, 1952, Part IV. 3 E. F. Jackson, op. cit., p. 259. 4 Reference No. 10, p. 206.

NOTES TO CHAPTER 36 1 Phyllis Deane, The Measurement o f Colo­ nial National Incomes: A n Experiment, Cam­ bridge, England, 1948. 2 Pravda, September, 1953; People’s China, November 16, 1954; World Politics, vol. VII, October, 1954-April, 1955. 3 Reference No. 7, p. 279. 4 Reference No. 8, pp. 246-49. 6 According to the U.N. Economic Survey o f Asia and the Far East, 1957, Communist China’s national income at constant (1952) prices increased under the first five-year plan (1952-56) at a decelerating rate, averaging 9.1 per cent per year: Population 1952-56, millions: 569; 581; 595; 608; 621. National income, billion yuan: 63.62; 75.59; 81.71; 85.53; 90.00; consumption, billion yuan: 53.63; 61.75; 64.06; 68.00; 71.00; ac-

cumulation, billion yuan: 9.99; 13.83; 17.65; 17.53; 18.90. Per capita national income: 112; 130; 137; 141; 145. Per cent annual increase in national income, 1953-56: 18,8; 8.1; 4.7; 5.2. The per capita consumption was estimated to have risen over the period 13 per cent for the peasants and 19 per cent for workers and employees. 6 Real national income of Japan (1952 prices), according to the U.N. Economic Sur­ vey o f Asia and the Far East, 1957, is expected to increase under the new economic plan in 1957-61, 5.8 per cent per annum, i.e., at a much lower rate that it did in the decade 1947-57 (11.4 per cent). The goal of the new plan is to achieve economic stability and bal­ anced growth as between different sectors of the economy.

NOTES TO CHAPTER 37 1 Phyllis Deane, “Measuring National In­ come in Colonial Territories,” Studies in In-

come and Wealth, vol. 8, National Bureau of Economic Research, New York, 1946, p. 167.

Index

T he Index is divided into tw o p arts: one relating to the authors m entioned in the text, the others to the countries and regions re­ ferred to in it. In both Indexes the n atu re of the subjects treated is indicated by th e fol­ lowing sym bols: H -history of national in­ com e concepts, m easurem ents, and analyses; T -theory, current; M -m ethodology and esti­ m ates, current; S-statistics. W here m ore th an one o f these subjects is treated, a com bina­ tion o f these symbols is used. T he num bers refer to the text pages; the m ost im portant references are in bold-face type. F o r guidance as to the location in the text o f subjects other than countries and re­ gions, see the detailed T able o f Contents.

INDEX OF AUTHORS This Index consists alm ost wholly of nam es o f persons m entioned in the text— princi­ pally estim ators of national incom e and authors o f statistical and theoretical w orks o n the subject. O f the m any governm ental an d private organizations carrying on activi­ Aa, E. R. van der, MS-469 Achenwall, Gottfried, H-85, 517 Adler, J. H„ H-157; MS-484 Ady, P., MS-494 Aktan, Resat, MS-414 Albon, Count d’, H-65 Alembert, I. L. D \ H-75 Angelopoulos, A., MS-480 Anis, A. M., MS-490 Antonovski, M. I., H-89, 98 Arndt, H. W., H-136, 137, 526 Arnould, H-76, 77, 123, 521 Atkinson, F. G„ H-137 Aubrey, H. G., MS-484 Aukrust, Odd, T-213; MS-471 Bach, F., MS-484 Bacon, MS-486 Badre, Albert Y„ H-157; MS492 Bagehot, Walter, H-117 Baibour, H-137 Bakker, O., MS-326 Ballue, H-128 Baran, Paul, MS-352 Barberi, Benedetto, TM-3; H150; MS-465, 466 Baring, H-137 Barnett, G. E., H-30 Barreto, Emilio, G., MS-488 Bartels, Hildegarde, MS-374 Barthe, Andre, H-145, 156, 527 Batcharin, A. B., T-540 Baudhuin, Fernand, H-150; MS468 Baudrillart, Henri J. L., H-20 Bauer, Wilhelm, M-541 Baxter, James I., MS-490 Baxter, Robert Dudley, H-20,

ties in the field and referred to in the text, only those internationally know n are in­ cluded here. F o r the nam es of other organi­ zations, see the bibliographies of the indi­ vidual countries in Chapters 21-37.

102, 111, 112, 114-17, 118, 119, 132 Baykov, Alexander, T-540 Bearde de LAbbe, H-88 Beccaria, Cesare, H -18-19 Beeke, Henry, H-43-47, 50, 51, 104, 106, 114, 115, 118, 121, 122, 158, 518 Beer, Max, H-106, 523 Bekaar, J. J., M-537 Bell, Benjamin, H-47-50, 51, 114, 118, 518 Benham, Frederick C., H-15051, 156, 157; M-262, 267, 268, 274, 289; MS-303, 44349, 459, 528 Bergson, Abram, MS-352, 372; M-533, 541 Bernoulli, Daniel, H-84 Biedermann, H-527 Bielfeld, J. F., H-86, 517 Bilkur, Selflk, MS-414, 415, 417 Birla, G. D„ MS-543 Block, Maurice, H-20, 119, 127 Bodin, Jean, H-13 Boemert, Wilhelm, H-526 Boileau, D„ H-524-25 Boisguillebert, Pierre, H -ll-1 6 , 52-56, 60, 62, 66, 77, 82, 120, 158, 515, 519; T-168; M-255 Boldrini, M., MS-465, 466 Bolotov, A. T., H-88 Bonger, W. A., H-145, 156; MS325, 328, 537 Bonnet, V., H-128 Bowley, A. L., TM-3; H-50, 114, 117, 142-43; MS-457, 526 Break, George F., M-535 547

Britnell, G. E., M-537 Brookings Institution, The, TS527—28 Bunge, N. C., H-138, 539 Burke, Edmund, H-103 Biisching, A. F., H-85 Caillaux, loseph, H-143 Campion, Harry, H-152 Cannan, Edwin, H-21, 118; T516 Carli, Gian, H-18 Castro, A., MS-480 Catherine the Great, H-78, 80, 82-84, 87, 89, 90, 91, 94, 95, 96, 98, 138 Ceickners, A., H-156 Chailley, H-128 Chalmers, George, H-20, 30 Chamillart, H-53 Chang, C. H„ MS-496 Chaptal, lean, H-20, 120, 123, 128; M-255 Chernomordik, D. I., TM-3; T186; MS-349, 354, 355, 356, 358, 360, 361; T-531, 540 Chernyshev, Z. G., H-87 Child, Josiah, H-13, 517 Cirovic, M., H-156 Clark, Colin, TM-3; H-40, 76 114, 150-51; T-192, 203, 225-26; TS-233; M-259, 261, 277, 286, 288, 293, 297; MS349, 352, 371, 457, 459, 478, 498, 518, 521, 527, 528, 530, 533, 534, 535, 536 Clavier, Charles, MS-156, 468 Claviere, fitienne, H-67, 77 Clay, Henry, H-152 Cleef, Ed. van, MS-538

548

Coats, R. H., H-149, 150; MS303; M-536 Cochut, Andre, H-125, 127, 128 Coghlan, Timothy, H-135-37, 141, 145, 150; MS-458, 459, 526 Coke, Roger, H-517 Colbert, J. B., H-53 Colm, Gerhard, TM-3; H-153; T-197-98, 200, 203, 214, 532, 533, 534; TS-223 Colquhoun, Patrick, H-20, 1027, 108, 109, 111, 114, 115, 118, 131, 143, 159; T-168, 530, 536 Colson, Clement, H-21, 128, 142-43, 527 Condillac, Etienne Bonnet, H65, 66 Condorcet, M. I. A., H-75, 95 Conring, Hermann, H-85 Copeland, Morris, TM-3; T-214 Coppola D A nna, F., MS-465, 466 Correia, A., MS-480 Cosciani, C., MS-465, 466 Coste, Alfred, H-125, 126-27, 128 Cournot, Antoin Augustin, H138 Craig, James T., MS-490 Crawford, J. G„ H -150-51; T-192; M-277, 286; MS-459, 528, 535 Creamer, Daniel, MS-454, 482, 483, 491 Creamer, Henrietta, MS-482, 483 Cromwell, Oliver, H-27 Cummings, John, H-129 Curzon, Lord, H-137 Czoernig, Karl von, H-134, 141, 159; MS-472, 473 D Addario, R„ MS-465 Dahlgren, Einar, TM-3; MS334, 341, 533 Daire, Eugene, H-20, 59, 515, 519, 520 Davenant, Charles, H-13, 30, 37-40, 42, 51, 54, 65, 69, 86, 90, 91, 92, 97, 100, 106, 158, 517, 518 Davis, MS-486 Deane, Phyllis, H-50, 108, 109, 114, 156; MS-454, 493, 500, 503, 504, 518, 544 Decker, Matthew, H-41, 51 Dederko, B„ H-156; MS-475 Degli Espinosa, Agostino, MS465, 466 Dehen, Peter, M-541 Denison, Edward F., M-528, 530, 533 Deparcieux, Antoine, H-85 Derksen, J. B. D., H-150; TS228, 231; MS-325, 331, 332, 424, 486, 491, 534, 537, 538; S-507, 509, 511 Descartes, Rene, H-66 Desai, R. C., MS-490 Destutt de Tracy, A., H-20

INDEX OF AUTHORS

Deutsch, J. J., M-537 Diehl, Karl, MS-375, 541 Dieterici, C. F. W., MS-374 Digby, William, H-137 Djordjevic, M. R., MS-408 DjuriciC, V. M., MS-408 Dobb, Maurice, MS-540 Dominguez, Loreto M., MS487, 489 Donner, MS-541 Doumer, Paul, H-125, 128 Drilier, H „ H-157 Dugdale, William, H-30 Duge de Bernonville, Leo, H150, 153; T-204; M-255; MS391, 528 Duncan, G. A., T-203; M-261; MS-318 Dunoyer, Charles, H-20 Dupin, Charles, H-124-25, 128 Dupont de Nemours, Pierre, H -15, 61, 67, 70, 76, 77, 99, 520 Dutens, J. M., H-125, 128 Dutot, Charles de F., H-59, 60, 519; M-255 Eckstein, Alexander, MS-495 Egana y Marquez, MS-486 Eldem, MS-414, 415 Elsass, MS-375 Engel, Ernst, H-135; MS-374, 526 Englis, MS-475 Ensley, Grover W., T-214, 528 Eppenstein, Franz, MS-414 Evelpidi, C., MS-480, 481 Evelyn, Sir George, H-108 Fabricant, Solomon, M-276, 535 Fajardo, S., MS-489 Faure, H-120 Fellner, Friedrich von, TM-3, 21; H-144, 527; T-203; M267, 287; MS-409, 473, 474, 475 Ferrero, Romulo, H-156; MS488 Firestone, O. J., M-537 Fisher, A. G. B„ MS-460 Fisher, Irving, T-190-91, 531 Flaix, Fournier, H-128 Florens, E. M„ MS-488 Flux, A. W., M-288; MS-457 Forbonnais, F. V. D. de, H-65, 66

Ford, MS-486 Forde, D„ MS-501 Forg, Shan Kwei, MS-495 Fougstedt, Gunnar, MS-472 Foville, Alfred de, H-21, 119, 125, 126, 127, 128, 139, 143, 526 Franger, Otto von, M-408 Frankel, S. Herbert, H-150, 156; MS-460, 461, 503, 504 Franzsen, D. S., M-461 Friedman, A., H-527 Friedman, M. I., T-527 Frisch, Ragnar, TM-3, 4, 26; H-150, 151; T-213; MS-471, 515, 528, 533

Friss, Istvan, MS-474 Fuhermann, E., H-527 Gaathon, A. L., MS-491 Gainsbrugh, Martin R., T-214 Gaitonde, V. R., MS-428, 543 Galiani, Ferdinando, H-16, 17 Ganilh, Charles, H-20, 61, 12023, 521, 525 Garnier, Germain, H-20 Geary, R. C., TS-226; MS-318, 322, 537 Geering, T„ H-145, 156; MS466, 467, 527 Gentz, Friedrich, H-50-51, 519 Georgi, T. G„ H-95 Ghosh, A. K., M-293; MS-424, 425, 434, 438, 543 Giffen, Sir Robert, H-41, 50, 51, 102, 103, 106, 111, 114, 117, 118, 132, 518, 536 Gilbert, Milton, TM-3; H-150, 154; T-209, 212-13, 528, 529, 531, 533, 534; TS-22526; MS-420, 455 Gini, Corrado, TM-3; H-150; T-178, 190-91, 532; MS-464, 465, 466 Giurio, Julio, MS-487 Glazunov, I., H-97 Goldberg, Simon A., M-537 Goldsmith, Raymond W., TM-3; T-214, 529 Golikov, I. I., H-85, 522 Gould, J. S., MS-494 Graunt, John, H-26, 84, 86, 522 Gray, John, H-102, 106-7, 118, 523 Gray, Simon, H-102, 106-8, 523 Greenwood, Arthur, H-152 Grehov, R. D., H-522 Groot, C. P. de, M-537, 538 Gruenbaum, Ludwig, H-156; MS-491 Gruening, Ferdinand, MS-541 Gruson, C., TM-3; MS-391, 542 Guerreiro, A. D., MS-479, 480 Guptas, B. B., Das, MS-493 Gurtler, A., MS-473 Guttierez, G., MS-482 Habr, J., MS-475 Hales, John, H-13 Hailey, Edmund, H-522 Harrod, R. F., H-152, 528 Helfferich, Karl, H-144, 527; T-168; MS-374 Hemming, Francis, H-152 Henderson, Sir Hubert, H-152 Henry, Paul, H-143, 158, 526 Heretik, S., MS-475 Hermann, B. F. G., H-78, 87, 90-93, 95, 96, 97, 99, 100, 146, 522; MS-375 Hermann, F. B. W., H-20; T-525 Herodotus, H-52 Herzfeld, H „ MS-461 Heymann, Hans, MS-372; T-533, 541

INDEX OF AUTHORS

Hicks, J. R., TM-3; T-168, 197, 206, 529, 533 Hijikata, Shigeyoshi, MS-497 Hobbes, Thomas, H-66 Hobson, John, T-191, 532 Hodnev, A. I., H-87, 522 Hoeffding, Oleg, MS-372, 541 Hogben, L. T., H-40 Hopkins, Richard, H-152 Hotowetz, MS-475 Hotz, R., H-145, 156; MS-466, 467, 527 Hufeland, Gottlieb, H-20, 525 Hull, C. H., H-30, 517 Hupei, A. W., H-85, 93, 138, 522, 523 Idenburg, Philip J., MS-325, 538 Inan, MS-414 Inonu, MS-414 Institut International de Statis­ tique, M-529 International Association for Research in Income and Wealth, TMS-434, 456; M334; MS-489, 517, 529, 530, 531 International Bank for Recon­ struction and Development, MS-157, 443, 482, 484, 485, 486, 492, 501, 547 International Statistical Insti­ tute, MS-424, 443, 543 Ivernois, Francois d’, H-67, 76, 77, 520, 524

King, W. I., H-143-44; MS455, 527 Kiranoff, P., MS-477 Kirillov, I. K„ H-82 Kleinwaechter, Fr., H-134, 526 Knauth, O. W., MS-527 Knopp, MS-375 Kochanovskov, I., MS-475 Kock, Karin, MS-334, 532 Kolganoff, M. B„ TM-3; MS349, 366; T-531, 540 Koller, W. I., MS-485 Kotchubey, V. P., H-96 Kotuchihin, I. K , H-80 Kraft, W. L., H-86, 91 Krasnolobov, I. M., M-349 Kraus, Otto, H-20; T-187, 531 Krause, Walter, MS-483 Kravis, Irving, TS-225-26; T529; MS-534 Krieger, V. A., MS-488 Krischen, MS-468 Kristensen, Thorkil, H-156; MS-469 Krizamc, G., H-80 Krug, Leopold, MS-374 Kruisheer, J. D., MS-537 Kuczinski, Robert R., MS-375 Kuehnert, F., H-527 Kuznets, Simon, TM-3; H-150, 155; T-165, 169, 177, 179, 180, 189, 191, 195, 197, 198200, 202, 206, 211, 212, 528, 529, 531, 532, 533, 534; TS225, 226; M-261, 275, 276, 281, 288, 535, 536; MS-424, 455, 538 Kyskilitsis, A., MS-481

Jackson, Donald, T-200 Jackson, E. F., MS-463, 464, 478, 517, 544 Jacquart, Camille, MS-414 Laffemas, Barthelemy de, H-13 Jakob, L. H. von, H-20 Lagrange, Joseph Louis, H-61, Jasny, Naum, MS-352 75-76, 121, 123 Jaszi, George, MS-455; T-533 Landau, Ludwik, H-12, 150; Jewkes, John, H-152 T-187, 203; M-288; MS-475, Jones, N. S. Carey, H-157 476, 531 Jonnes, Moreau de, H-20, 67, 123, 124, 125-26, 128, 159, LaPlace, Pierre Simon, H-120, 138 520; M-255 Lasorsa, G., MS-465, 466 Jorgenson, H. C., MS-469 Jostock, P., TM-3; H-150; MS- Lassalle, Ferdinand, H-135 Lauderdale, Lord, H-20 374, 541 Laurilla, Eino H., MS-472 Kaldor, John, H-154, 528 Lavergne, A. de, H-143, 158, Kalecki, Michael, H-12, 150; 517, 526, 527 T-187, 203; M-288; MS-475, Lavoisier, Antoine Laurent, H476, 571 16, 61, 65, 68-75, 77, 91, 99, Kampmann, Viggo, MS-469 104, 106, 121, 123, 127, 158, Karra, V. A., MS-478 520, 521 Katz, V., MS-349, 373, 540 Law, John, H-60, 61, 519 Kautsky, Karl, T-530 Lawrence, A. R., M-536 Kellenberger, Edward, MS-468 League of Nations, M-154, 193 Kemper, M. J. de Booch, M- Lehfeldt, R. A., H-156; MS-460, 537, 538 461 Keynes, John Maynard, TM-3; Leibnitz, G. W. von, H-66 H-12, 25-26, 108, 151, 153, Lenti, MS-465 154, 158; TS-221, 517 Leontieff, Wassily, H-155; TKiernan, T. J., H-156; MS-318 214-16; M-288, 529, 533, King, Charles, H-518 536 King, Gregory, H-13, 15, 26, Leroy-Beaulieu, Pierre Paul, H30-36, 37, 38, 39, 40, 41, 42, 21, 119, 125, 126, 127, 128, 47, 51, 62, 66, 69, 92, 100, 132, 134 102, 108, 114, 121, 147, 160, Le Tresor de la Roque, H-128 515, 517, 518 LeTrosne, G. F., H-61, 67, 77

549

Levasseur, Emile, H-143, 527 Leven, M., MS-527 Levi, J. G., MS-490 Levi, Leone, H-116, 118, 139, 156; MS-350, 526, 539 Lewis, Arthur, MS-502 Liashchenko, P. I., H-79, 89, 521, 522, 526, 527 Lieu, Da Kuen, H-156; MS-495 Lindahl, Erik, H-150, 156; MS334, 533 Lindberg, Valter, H-156; T-177; MS-471, 472 Litoshenko, T-539 Liu, Ta-Chung, MS-495 Liverpool, Lord, H-114 Livi, Livio, MS-465, 466 Locke, John, H-66 Loftus, P. S., MS-491 Lomonosov, M. V., H-86, 522 Losch, H., H-526 Lowe, Adolph, T-191, 532 Lowe, Joseph, H-20, 102, 1079, 118, 131, 158, 523-24; TS-221 Lundberg, Erik, TM-3; T-213 Ma, R. A., MS-494 McCulloch, John Ramsey, H20, 103, 110, 111, 114 MacGregor, D. C., M-537 Maitland, James, H-20 Malthus, Thomas R., H-20, 107, 108 Marchewski, Jan, T-517 Marco, Antonio de Viti de, T531 Margolin, N., MS-373 Marshall, Alfred, H-20-21, 118, 136, 158; T-191, 516, 532 Marx, Karl, TM-6, 12, 19, 20, 21-25; H-89, 135, 138, 14546, 158; T-181, 183-87, 352, 357, 411, 474, 476, 477, 516, 530, 531 Mas, H. A., H-157 Masoin, M., H-156; MS-501 Matolcsy, M„ TM-3; H-12, 150; T-187, 196-97, 203, 204; TS-219; M-260, 265, 267, 268, 290; MS-474, 530, 531, 532, 534 May, R. E„ H-135, 141, 144, 374, 526, 527 May, Stacy, et al., MS-485 Meade, James, H-152, 154; T169, 209, 213; M-288; MS471, 475, 528, 529, 536 Meiren, P. van der, MS-468, 469 Meliado, Leonardo, MS-465, 466 Melon, Jean F., H-60 Menger, E., H-525 Meunier, H., MS-490 Meyer, Robert, MS-374, 526 Michandiere, H-72 Miliukoff, Paul, H-521 Mill, James, H-20 Mill, John Stuart, H-20; T-190, 531 Minz, B., MS-477

550

Mirabeau, Victor, H-15, 61, 75, 520 Mitchell, Wesley C., T-169, 527 Moheau, H-72 Moliere, Jean B. P., H-16 Molinari, Alessandro, MS-466 Molinari, Gustave de, H-520 Money, Sir Chiozza, H-143, 144, 526 Montalivet, H-120, 122-23, 128 Montchretien, Antoine, H-13 Montesquieu, H-48 Montoya, S. M., MS-484 Mori, K„ MS-497, 498 Mori, P., H-150, 467 Mortara, Georgio, MS-465 Moulton, H. G„ M-527 Mukherjee, M., M-293; MS424, 425, 434, 438, 543 Mulhall, G. M„ H-21, 139, 140-41, 145, 146, 526, 539; MS-350 Mun, Thomas, H-13, 65, 66 Nachshon, H., MS-491 Nakamura, Kinzo, H-145, 156; MS-497 Naoroji, Dadabhai, H-137 Nathan, Robert R., H-157; MS455, 499; M-535 National Bureau of Economic Research (U.S.), M-456-57, 529, 530, 532, 533; MS-149, 534, 543 National Industrial Conference Board (U.S.), MS-149, 204, 205, 538 Neale, E. P., MS-460 Nelson, R. W„ T-200 Neubauer, Jules, MS-474 Neufchateau, H-120 Neuman, John von, H-20 Neumark, S. D., MS-460, 461, 494 Newton, Sir Isaac, H-61 Nicholson, J. L„ TS-226, 533, 534 Nitshke, K„ H-527 Nourse, I. G., M-527 Nowacki, J., MS-477 Ochoa, Raul, H-156; MS-487 Ohkawa, Kazushi, MS-499 Ohlsson, Ingvar, TM-3; T-213; MS-334, 533 Orbaneja, A. G., MS-479 Organisation for European Eco­ nomic Cooperation (OEEC), Division of Statistics and N a­ tional Accounting, M-154, 160, 209, 271, 279, 282, 28991, 292; MS-225, 262, 26566, 325, 391, 467, 468, 469, 471, 472, 529, 530, 533, 534, 536, 538, 541; S-454 Orlandi, Silvio, MS-465 Ortego, R. M., MS-484 Ortiz, A. A., MS-486 Ou, Pao-San, MS-495, 496 Padilla, M. A., M-484 Palacios, C. J. A., MS-484

IN DEX OF AUTHORS

Palgrave, Sir R. H. I., H - l l l , Reclus, Elisee, H-128 524 Rediadis, P., M-268; MS-480, Palmieri, Giuseppe, H-16-17, 481 516 Reichel, I. G., H-85 Pantaleoni, Matteo, H-137; MS- Reichlin, A., MS-467 464 Renoult, Rene, H-143 Revel, P. Thaon di, MS-465, Passy, Hippolyte, H-128 466 Patino, Emilio A., MS-483, 484 Pebrer, Pablo, H-20, 102, 109- Ricardo, David, H-20, 110 10, 118, 131, 158, 536 Rickman, Joan, H-103 Peel, Robert, H - l l l , 112 Rivet, R., H-153, 528; M-255 Pei, Yu-Ling, MS-496 Riviere, Mercier de la, H-15, 90 Pergham, Margery, MS-501 Robbins, Lionel, H-152 Peris, K„ H-527 Robinson, Arthur, MS-502 Perroux, Francois, TM-3; T- Robinson, Austin, H-152 169, 214, 529 Rogowski, MS-375 Pervuhin, A., T-540 Romero, G. K , MS-484 Petrof, A. I., MS-349, 353, 373, Roscher, Wilhelm, H-20, 375, 571 525 Petty, Sir William, TM-5; H- Roschman, MS-473 11-16, 26-30, 32, 34, 37, 38, Rose, George, H-43, 51, 518 39, 40, 41, 42, 51, 62, 65, Rosen, J., MS-352 66, 69, 84, 86, 90, 92, 97, Rossi, Pellegrino, H-20 99, 100, 102, 104, 108, 114, Rousseau, H-48, 94 121, 158, 160; 517, 522, 523; Rouvier, H-128 T-168; M-255 Rubenstein, N. L., H-89 Petyniak-Sanecki, MS-476 Rudcenko, P., MS-408 Peytral, H-128 Ruggles, R., T-213, 533 Pietri, A. Uslar, MS-486 Rumelin, G., H-135, 141, 526; Pitt, William, H-41, 43-45, 46, MS-374 48, 49, 50, 51, 113, 158 Rutherford, J. B., M-537 Plestcheiev, H-91 Rytchkov, P. N., H-88 Poincare, Raymond, H-143 Pokrovsky, M. N., H-521, 522 Sacerdote, MS-465 Pokrovsky, V. I., H-21, 88, Saenz, Josue, M-259, 261, 269; 139-40, 145, 146, 526, 539; MS-483, 484 MS-350 Saint-Simon, Louis, H-53, 519 Politis, Demos, TMS-481 Sakalis, S., MS-480 Pollner, Marco D., MS-487 Popoff, Kiril I., H-145, 156, Samter, Adolph, H-135, 526 Santoro, Michele, H-145, 156; 527; MS-477 MS-464, 465, 527 Popov, Khristo, MS-478 Sauvy, Alfred, H-153, 528; MPorri, MS-465 255 Porter, George R„ H-110-11, Savorgnan, F., MS-473 524 Say, Jean Baptiste, H-20, 107, Pososhkov, I. T., H-82, 521 123-24, 126, 128, 138, 539 Poussielque, H-128 Sayer, Server, MS-414 Prest, A. R„ MS-501 Prokopovitch, S. N., H-145-49, Schaeffle, A. E. F„ H-525, 526 156; MS-349, 351, 352, 371, Schall, MS-374 Schlotzer, A. L., H-85 474, 527, 539, 540 Schmoller, Gustav, H-375, 525 Ptucha, M. V., H-78, 84, 87, Schneider, Salome, MS-467 93, 100, 521, 522, 523 Schnitzler, J. H., H-125, 128 Pulteney, H-43, 51 Schorry, Otto, MS-374 Quesnay, Franfois, H-15, 16, Schuman, C. G. W., MS-460, 461 26, 53, 60, 61-65, 66, 67, 89, Schumpeter, Joseph A., TM-4, 138, 520; TS-221 25; H-66, 108, 169, 515, 516, Quetelet, Adolphe, H-138 517, 529 Schwartz, C. E., M-533 Raabe, H. H „ MS-374 Schwinnen, A., MS-469 Rademont, Goevin, H-60 Radishchev, A. N., H-78, 86, Scott, R„ MS-501 89, 90, 93-97, 98, 99, 100, Sczepanik, E. F., H-157; MS494 101, 146, 522, 523 Seligman, E. R. A., H-127 Ragazzi, Bruno Rossi, MS-465, Senior, N. W., H-20 466 Rao, V. K. R. V., H-137, 150, Shcherbatov, M. M., H-85, 522 156; M-267, 268, 288; MS- Shelborne, Lord, H-29 Shinohara, M., MS-499 424, 425-33, 543 Shiomi, Saburo, MS-497 Rau, K. H., H-525, 539

INDEX OF AUTHORS

Sutcliffe, J. T., H-149, 150-51; Shirras, G. Findlay, MS-424, MS-459, 528 543 Sweezy, M. Y., MS-541 Shoup, Carl, T-168, 191, 529, 532 Tabacovici, N., H-527 Simon, Raul, H-156; M-288; Takahashi, K , MS-497 MS-488, 489, 536 Sinclair, John, H-41, 103, 517 Talandier, H-128 Tanin, O., MS-498 Singer, H. W., MS-541 Tatishchev, V. N„ H-82 Sismondi, H-20 Taubert, I. L, H-87 Skiadas, S., H-137; MS-480, Tchackaloff, Assen, M-268; MS481, 526 477 Smart, William, H-21, 118, 119 Tchulkov, M., H-85, 89, 522 Smee, William Kay, H-102, 111-14, 115, 117, 118, 524 Tchuprov, A. A., H-138, 139 Teisserenc de Bort, H-128 Smeets, J. H„ M-537 Smith, Adam, H - ll, 12, 14, 16, Thiers, Louis Adolphe, H-124 17, 18-20, 21-23, 25, 47, 48, Thorne, A. P., MS-443 50, 66, 76, 85, 89, 90, 92, 93, Tinbergen, J., M-537, 538 95, 97, 99, 100, 102, 103, Tivaroni, J., MS-465, 466 104, 107, 115, 119, 121, 123, Tolosan, H-67, 77, 123, 125, 126 136, 138, 158; T-181-83, 184, Tomita, S., MS-499 187, 350, 409, 516 Smith, Dudley, H-156; MS-482 Tosic, M. D„ MS-408 Tretiakov, N. A., H-89 Smith, Harold, H-153 Tsuru, Shigeto, MS-499 Smutny, Paul, MS-475 Tucker, George, H-20, 66, 129Sobol, V., MS-373 32, 134, 141, 159; T-168; Soetbeer, Adolph, H-135, 141; MS-455 MS-374, 526 Tugan-Baranovsky, M. I., HSoler, C. A., H-156 146 Sonesson, N., MS-340 Turetsky, S., T-539 Spacil, B„ MS-475 Spackman, W. F., H-20, 102, Turgot, A. R. J., H-19, 62, 65, 66, 89, 520 110-11, 118, 524 Spahr, Charles, H-132-34, 143, United Nations, Statistical Of­ 158; MS-455 fice; also Economic Commis­ Spallart, Neumann, H-134 sions for Europe, for Near Speransky, M. M., H-96 and Middle East, and for Stadnik, Milos, MS-475 Asia and the Far East; and Stamp, Josiah, TM-3; H-114, Missions, M-155, 160, 271, 142, 143, 152; MS-457; T333, 532; MS-154, 155, 157, 526, 530 160, 228, 232, 265, 453-54, Starodubrovskaia, V. N., 478 469, 474, 475, 478, 485, 487, Stephens, F. B., H-150; MS-460, 492, 493, 494, 495, 496, 499, 528 500, 504, 511, 528-29, 532, Sternheim, MS-326 535, 536, 544; S-250 Stewart, I. G., MS-501 Stone, Richard, TM-3; H-152, Vacher, L„ H-125, 127, 128 154, 528; T-169, 209, 213; MS-481 M-288, 529, 536; MS-424, Validis, Van Cleeff, Ed., MS-332 475, 502 Jose A., H-156; MStorch, Heinrich, H-20, 87, 138, Vandelos, 267; MS-479 516, 522, 525, 526, 531, 539; Van Olst, H. R., H-157; MST-190; MS-350 485, 486 Strachey, John St. Loe, H-119 Van Yzern, J., TS-226 Strukof, A., MS-350, 373 Varga, S., TM-3; H-12, 150; Strumilin, S. G., T-186, 531, T-177, 187, 196-97, 203, 539; MS-373 204; TS-219; M-260, 265, Studenski, Paul, M-349, 534 267, 290; MS-474, 530, 531, Sturmthal, A., MS-484 532, 534 Siissmilch, J. P., H-86 Vargentine, P., H-86

551

Vauban, Sebastien Le Prestre, H-13, 14, 15, 53, 54-60, 68, 77, 82, 120, 158, 160, 519 Vegner, A., H-156; MS-408 Vernadsky, G„ H-521 Verri, Pietro, H-16-18, 516 Vignes, H-128 Viita, Pentti, MS-472 Vinci, Felice, T-177, 179; MS465, 466 Vinski, Ivo, T-186; MS-408, 409, 411, 413 Vita, Agostino de, MS-465, 466 Voltaire, F. M. A. de, H-54, 61, 67, 74, 77, 91, 94, 519 Vorontzov, A. R., H-93, 94, 95, 96 Wagner, Adolph, H-21, 134, 144; T-168, 525, 526, 527 Wahlbeck, Lara, MS-472 Waizner, Ernst von, MS-473, 478 Walker, Francis A., H-131 Wallich, H. C„ H-157; MS-484 Walras, Antoine Auguste, H-20 Walras, Leon, H-138 Warburton, Clark, M-527 Watkins, G. S., MS-500 Wenk, Erwin, MS-467, 468 Wijk, J. Van der, M-537 Williams, K„ MS-493, 494, 535 Winkler, W., MS-473 Wisniewski, Jan, MS-476 Wolff, Christian, H-86 Wolowski, Louis, H-21 Wood, G. E., MS-491 Woog, Henri, H-520 Woytinsky, W. S. and Emma, TS-534 Wu, Yuan-Li, MS-495, 496 Wyler, Julius, H-151; TS-226; M-288, 297; MS-349, 352, 371, 372, 467 Yamada, Yuzo, MS-499 Yamashita, Tetsutaru, MS-497 Yanson, Y. A., H-139 Yntema, Dwight R., T-532 Yohan, E„ MS-498 Young, Arthur, H-41-43, 44, 47, 51, 67-68, 69, 77, 90, 100, 103, 104, 121, 123, 518, '520; T-165 Zaalberg, C. J. P., M-537 Zabradnicek, O., MS-475 Zhulkowski, N. D., H-89, 98 Zolotas, X., MS-480, 481 Zwingli, U., MS-467, 468

IN D E X O F CO UNTRIES A N D REGIONS

Africa, TM-7; H-98; TS-231, 244; M-264; MS-453; S-510, 500-504 Albania, H-156; S-230, 232, 233, 249; MS-462, 463, 47879 Algeria, H-157; TS-254; S-510 America, North, M-264 America, South, M-264 Argentina, H-141, 156; TS-229, 236, 238, 241, 248, 254; MS481, 487-88; S-509 Asia, H-98, 148; TS-231, 232, 244; M-264; MS-461, 499 Australia, H-129, 135-37, 141, 145, 149, 150, 151, 153, 526, 527, 528; T-192, 204, 207; TS-230, 238, 241, 254; M263, 264, 272, 274, 275, 277, 278, 279, 285, 286, 290; MS352, 453, 455, 458-60; S-508 Austria, H-17, 79, 80, 90, 134, 139, 141, 150, 159; T-204; TS-229, 232, 236, 238, 241, 245, 248, 254; M-263, 274; MS-462, 472-73; S-508 Austria-Hungary (see also Austria, Hungary), H-129, 134, 144, 147, 525, 526, 532 Belgian Congo, H-156; TS-230, 234, 236, 239, 242, 243, 249, 254; MS-500, 501; S-510 Belgium, H-141, 147, 149, 150, 156; T-214; TS-229, 231, 235, 238, 239, 241, 248, 254; M260, 263, 278; MS-461, 462, 468-69; S-508 Bolivia, H-157; TS-230, 235, 239, 254; MS-481, 487; S-509 Brazil, TM-7; H-157; TS-229, 234, 235, 238, 242, 249, 254; M-263, 268, 272, 274, 277, 279, 285; MS-301, 303, 44043, 481; S-509 British Guiana, H-157; TS-254; MS-481, 485; S-509 British Honduras, H-157; TS254; S-509

British India (see also India), Czechoslovakia, H-21, 149, 156; TS-230, 232, 233, 237, 240, M-270; MS-424 242, 249, 254; MS-462, 463, British West Indies (see also 474-75; S-509 Jam aica), H-150, 156; M257, 263, 267, 268, 303; Denmark, H-80, 141, 149, 156; MS-443-49, 481 T-181, 192, 215, 218; TSBulgaria, H-145, 146, 147, 150, 229, 234, 235, 238, 241, 248, 156, 527; T-204; TS-230, 232, 254; M-258, 263, 269, 273, 233, 237, 240, 242, 249, 254; 289, 294; MS-461, 462, 469M-263, 268, 271; MS-462, 71; S-508 463, 477-78; S-509 Dominican Republic, H-156; Burma, H-157; TS-230, 232, TS-242, 254; MS-481, 483; 235, 242, 254; MS-489, 490, S-509 494; S-510 Canada, TM-6; H-141, 149, 150, 151, 153, 154, 156; T177, 178, 181, 192, 195, 201, 204, 207, 209, 219; TS-229, 231, 235, 238, 241, 248, 254; M-257, 258, 259, 260, 263, 264, 265, 270, 271, 272, 273, 275, 277, 278, 280, 281, 283, 284, 285, 290, 292, 294, 529, 536, 537; MS-302, 303-17, 455; S-508 Caribbean, TM-6; TS-231; MS453,481; S-509 Ceylon, H-156; TS-242, 253, 254; M-270, 274, 289; MS489, 490, 493-94; S-510 Chile, H-149, 156; TS-229, 235, 238, 241, 248, 254; M-288; MS-481, 488-89; S-509 China, H-94, 95, 156; TS-225, 227, 231, 232, 233, 247, 249, 250, 254; M-263; MS-446, 489, 495-96, 544; S-510 Colombia, H-157; TS-229, 235, 238, 241, 254; M-274; MS481, 486; S-509 Costa Rica, H-157; TS-229, 235, 242, 254; MS-481, 485; S-509 Cuba, H-156; TS-238, 254; MS481, 482; S-509 Cyprus, H-157; TS-229, 235, 238, 241, 254; S-508 552

East Europe, MS-462-64, 544; S-508 East Germany (see also Ger­ many), TS-230, 232, 233, 237, 240, 249, 250, 254, 46364; S-509 Ecuador, H-157; TS-229, 235, 238, 239, 241, 254; MS-481, 486-87; S-509 Egypt, TM-7; H-149, 156; TS230, 235, 236, 254; M-263; MS-489, 490-91, 500; S-509 El Salvador, H-157; TS-254; MS-481, 484-85; S-509 England. See United Kingdom Estonia, H-90 Europe (also European), H-66, 98, 103, 139, 141, 145; TS225, 226, 231, 250; M-264; MS-446, 453, 461, 478, 503, 504 F ar East, TM-7; TS-231, 232; MS-453, 489, 494, 499; S509 Finland, H-79, 91, 139, 146, 149, 156; T-177; TS-229, 235, 238, 241, 248, 254; M263; MS-462, 471-72; S-508 Formosa. See Taiwan France, TM-3; H -ll, 13, 14, 15, 16, 19, 20, 21, 26, 28, 29, 32, 33, 34, 36, 39, 40-41, 48, 51, 52-60, 61-77, 78, 80, 82,

IN DEX OF COUNTRIES AND REGIONS

84, 86, 89, 90, 91, 92, 94, 95, 98, 100, 102, 103, 107, 108, 119-28, 129, 132, 134, 135, 137, 140, 141, 142, 143, 144, 145, 146, 147, 150, 153, 158, 159, 160, 515, 516, 519— 21, 524-25, 526, 527, 528; T-168, 169, 178, 205, 207, 214; TS-221, 225, 226, 229, 235, 238, 239, 241, 244, 248, 254; M-255, 258, 259, 262, 263, 267, 269, 270, 271, 272, 275, 276, 278, 286, 292; MS302, 317, 350, 374, 384, 385, 388, 391-407, 461, 490, 54142; S-508 French West Africa, H-157; TS-254; S-510

258, 263, 268, 271, 274, 288, 293; MS-424-39, 489, 490, 492, 543; S-509 Indonesia, H-156; TS-230, 235, 242, 243, 254; MS-489, 494; S-510 Iran, MS-489 Iraq, MS-489 Ireland, TM-6; H-27, 28, 103, 104, 105, 106, 107, 116, 149, 156; T-177, 203, 209; TS-229, 235, 236, 238, 241, 248, 254; M-261, 263, 267, 269, 270, 271, 272, 273, 275, 277, 278, 279, 297, 529, 537; MS-31824, 455, 461; S-508 Israel, H-156; TS-229, 235, 236, 238, 241, 243, 245, 248, 254; M-263, 264; MS-489, 490, 491; S-509 Italy, TM-3; H-16, 19, 80, 86, 90, 137, 141, 145, 147, 150, 156, 516, 527; T-173, 177, 179, 204; TS-225, 226, 229, 232, 235, 241, 244, 248, 254; M-263; MS-408, 411, 461, 462, 464-66; S-508

553

Mongolia, TS-232, 233 Morocco, TS-254; S-510

Near East, TM-7; TS-231; MS453, 489-92; S-509 Netherlands, H-26, 28, 29, 32, 33, 34, 36, 39, 80, 86, 100, 139, 141, 145, 150, 151, 154, 156; T-181, 204, 207, 214; TS-228, 229, 231, 235, 241, 248, 254; M-257, 258, 263, 264, 265, 268, 269, 273, 277, 278, 285, 286, 287, 537-38; MS-302, 328-34, 461; S-508 Netherlands East Indies (see also Indonesia), MS-332 New South Wales, H-135, 136, 145, 150 New Zealand, H-145, 149, 150, Gambia, H-157 151, 156, 527, 528; T-178, Germany (see also East Ger­ 204; TS-229, 234, 235, 238, many, West G erm any), TM241, 254; M-263, 264, 273, 3, 20, 21; H-50, 78, 80, 82, 278, 282; MS-453, 455, 46084, 85, 86, 87, 90, 93, 102, 61; S-508 129, 132, 134-35, 137, 138, Nicaragua, H-157; TS-254; MS139, 140, 141, 144, 145, 146, 481, 484; S-509 147, 150-51, 159, 525-26, Nigeria, H-157; TS-230, 236, 527; T-168, 169, 180, 181, 239, 242, 254; MS-500, 501; 187, 192, 200, 202, 203, 204; Jamaica (see also British West S-510 TS-225, 254; M-253, 259, Indies), TM-7; TS-229, 235, North America, H-98; M-264 263, 265, 267, 271, 272, 285, 236, 241, 248, 254; MS-443- North Korea, TS-232, 233 286, 287; MS-302, 317, 331, 49, 502; S-509 Northern Rhodesia (see also 350, 371, 374-90, 393, 408, lapan, H-145, 150, 156; T-178; Rhodesia-Nyasaland Federa­ 461, 471, 473, 475, 476, 480, TS-229, 235, 238, 241, 248, tion), H-156; MS-500, 502254; M-258, 260, 263, 272, 541 3, 504; S-510 Ghana, MS-504 277, 289, 293; MS-446, 489, Norway, TM-3; H-141, 145, Gold Coast, H-157; TS-239, 490, 495, 497-99, 544; S-510 150, 153, 526, 527; T-213, 242, 254; S-510 Jordan, MS-489 215; TS-229, 235, 238, 241, Great Britain. See United King­ 243, 248, 254; M-258, 263, dom Kenya, H-156; TS-230, 236, 265, 268, 269, 270-71, 272, Greece, H-129, 137, 149, 156, 239, 254; M-257; MS-500, 273, 274, 275, 294, 297; MS501-2; S-510 526; T-204; TS-229, 232, 235, 342, 461, 462, 471; S-508 241, 248, 254; M-263, 268; Korea, Republic of (South Nova Scotia, H-150; MS-303, MS-462, 480-81; S-508 K orea), H-157; TS-230, 232, 306 Guatemala, H-157; TS-229, 242, 243, 254; M-293; MS- Nyasaland (see also Rhodesia489, 490, 499; S-510 241, 248, 254; M-293; MSNyasaland Federation), H481, 484; S-509 156; MS-500, 502, 503, 504; Latin America, TM-6; TS-231, S-510 244; M-264; MS-453, 481-89; Haiti, H-157; TS-230, 254; MS-509 293; S-509 Oceania, M-264 Latvia, H-149, 156 Hawaii, M-272 Holland. See Netherlands Lebanon, H-157; TS-229, 235, Pakistan, H-157; TS-230, 232, Honduras, H-157; TS-229, 235, 254; M-273; MS-489, 492; 235, 249, 254; MS-489, 490, S-509 238, 249, 254; MS-481, 484; 492; S-509 Luxembourg, H-156; TS-229, Palestine. See Israel S-509 Hong Kong, H-157; TS-229, 235, 238, 241, 248, 254; MS- Panama, H-157; TS-229, 235, 235, 236, 238, 239, 254; MS461, 462, 469; S-508 236, 238, 241, 254; MS-481, 489, 490, 494-95; S-510 485; S-509 Hungary (see also Austria-Hun­ Malaya, TM-7; H-150, 157; TS- Paraguay, H-156; TS-230, 235, 241, 254; M-257, 262, 263, gary), TM-3; H-12, 21, 90, 254; MS-481, 487; S-509 267, 268, 270, 274, 275, 289; 150; T-177, 187, 196, 204, MS-303, 443-49, 489, 544; Peru, H-149, 156; TS-230, 239, 219; TS-230, 232, 233, 240, 242, 254; MS-481, 488; S-509 S-510 242, 249, 254; M-260, 263, 265, 267, 268, 287, 290; MS- Mauritius, H-156; TS-229, 235, Philippines, H-157; TS-230, 239, 242, 254; M-268, 275; 241, 254; S-510 462, 463, 473-74; S-509 MS-489, 490, 499-500; S-510 Mexico, H-149, 156; TS-229, 235, 238, 241, 254; M-259, Poland, H-12, 79, 139, 146, Iceland, H-157; MS-462, 472; 149, 150, 156; T-187; TS-230, 261, 263, 269, 271, 278; MSS-508 231, 232, 233, 237, 240, 242, 481, 483-84; S-509 India (see also British India), 249, 254, 260; M-263, 268, TM-7; H-129, 137, 149, 150, Middle East, TM-7; TS-231; 274, 288, 293, 536; MS-462, 156; TS-221, 225, 227, 230, MS-453, 492-94; S-509 463, 475-77; S-508 232, 235, 244, 249, 254; M- Monaco, MS-461

554

INDEX OF COUNTRIES AND REGIONS

Portugal, H-141, 157; TS-229, 235, 241, 248, 254; MS-462, 479-80; S-508 Prussia (see also Germany), H79, 135; MS-374, 376 Puerto Rico, H-156; TS-229, 236, 238, 241, 243, 254; M263, 264, 275; MS-481, 48283; S-509

MS-352, 461, 462, 466-68; S-508 Syria, H-157; TS-254; MS-489, 492; S-509

Scandinavia, H-154; T-213 Scotland, H-28, 41, 43, 45, 46, 52, 103, 116 Serbia, H-147 Siam. See Thailand Siberia, H-79, 80, 91, 94, 96 Silesia, H-144 South America (see also Latin America), H-98; TS-231; M264; MS-303 South East Asia, MS-443 South Korea. See Korea, Re­ public of Southern Rhodesia (see also Rhodesia-Nyasaland Federa­ tion), H-156; MS-500, 503-4 Soviet Russia. See Union of Soviet Socialist Republics Spain, H-109, 141, 145, 149, 156, 527; TS-229, 231, 235, 254; M-263, 267; MS-462, 479; S-508 Surinam, H-157; TS-254; S-509 Sweden, TM-3; H-78, 80, 82, 86, 141, 149, 150, 153, 156; T-177, 179, 180, 192, 200, 202, 203, 213, 214; TS-230, 238, 241, 248, 254; M-257, 263, 264, 265, 268, 270, 271, 272, 275, 290, 292, 536, 53839; MS-302, 335-49, 461; S508 Switzerland, H-80, 84, 137,141, 145, 150, 151, 156, 527; T180, 181, 204, 209; TS-230, 238, 241, 248, 254; M-260, 261, 263, 267, 272, 281, 287;

Uganda, H-157; TS-239, 254; S-510 Ukraine (see also Union of Soviet Socialist Republics), H-79, 84, 91 Union of South Africa, TM-7; H-149, 150, 156; TS-229, 241, 254; M-264, 277, 278; MS-453, 455, 460-61; S-508 Union of Soviet Socialist Re­ publics, TM-3; H-12, 21, 78101, 122, 129, 137-40, 141, 145-49, 151, 155, 159, 52123, 525, 526, 527; T-178, 185, 186, 203, 204, 207; TS219, 220, 230, 231, 232, 233, 237, 240, 242, 249, 250, 254; M-255, 263, 265, 268, 270, 271, 288, 293, 297; MS-301, 302, 303, 349-73, 454, 461, 464, 474, 476, 531, 539-41; S-508 United Kingdom, TM-3; H -ll, 13, 14, 16, 19, 20, 21, 2640, 41-51, 52, 54, 65, 66, 69, 70, 76, 77, 78, 84, 85, 86, 90, 91, 93, 99, 100, 10119, 120, 121, 122, 125, 129, 131, 132, 134, 135, 137, 139, 141, 142-43, 144, 145, 146, 147, 152, 153, 154, 158, 159, 160, 515-18, 523-24, 526, 528; T-168, 169, 178, 192, 197, 198, 201, 203, 204, 207, 209; TS-208, 221, 224, 225, 226, 227, 229, 235, 238, 239, 241, 243, 244, 245, 248, 254;

Taiwan, H-157; TS-230, 234, 239, 242, 243, 254; MS-489, 490, 499; S-510 Tanganyika, MS-504 Thailand, H-157; TS-230, 249, 254; MS-489, 490, 494; SRhodesia-Nyasaland Federation, 510 TS-239, 242, 243, 249, 254; Tunisia, TS-254; S-510 M-289; S-510 Turkey, H-78, 79, 147, 149, Rumania, H-156; TS-230, 232, 151, 156; T-173; TS-229, 233, 240, 242, 249, 254; M235, 241, 248, 254; M-260, 263; MS-462, 463, 478; S-509 263, 274; MS-301, 302, 414Russia. See Union of Soviet 23, 489, 542-43; S-509 Socialist Republics

M-253, 255, 257, 259, 262, 263, 273, 278, 282, 286, 288, 289, 290, 294, 297; MS-301, 302, 303, 304, 317, 323, 332, 350, 352, 371, 374, 384, 385, 388, 424, 443, 453, 455, 45758, 461, 471, 474, 475, 481, 486, 489, 490, 493, 495; S508 United States, TM-3, 20; H-28, 30, 40, 43, 45, 76, 94, 100, 101, 102, 103, 129-34, 139, 141, 142, 143, 149, 150, 151, 153, 154, 155, 158, 159, 525, 527, 528, 529; T-165, 168, 169, 173, 174, 177, 178, 179, 189, 192, 193, 195, 198, 200, 201, 203, 204, 205, 207, 209, 210, 211, 212, 213, 214, 215, 216; TS-208, 220, 221, 222, 223, 224, 225, 226, 227, 229, 230, 231, 232, 235, 236, 238, 241, 244, 245, 248, 254; M253, 255, 258, 261, 263, 264, 265, 267, 269, 272, 273, 275, 277, 278, 279, 280, 281, 283, 284, 285, 287, 288, 289, 290, 291, 292, 293, 294, 297; MS301, 302, 303, 304, 316, 352, 353, 371, 384, 385, 388, 424, 453, 455-57, 481, 482, 485, 486, 489, 495, 497, 498, 531, 534, 535, 538; S-508 Uruguay, H-156; TS-229, 236, 238, 254; MS-481, 487; S509 Venezuela, H-157; TS-228, 241, 243, 254; MS-481, 486; S-509 Vietnam, TS-232, 233, 254; S510 Wales, H-43, 45, 46, 50, 116 West Germany (see also Ger­ many), TS-228, 229, 232, 235, 238, 241, 244, 245, 248, 250, 254; M-275; MS-302, 382-90; S-508 Western Europe, H-27, 149; M-264; MS-350, 489; S-508 Yugoslavia, H-21, 149, 151, 156; T-178, 186, 203; TS230, 232, 233, 237, 242, 249, 254; M-263, 267, 268, 270, 271; MS-301, 302, 303, 353, 408-13, 461, 462, 542; S-509

Date Due

D e m c o 293-5

30498

m is e r s

3 =1030 0 l 2 S 5b l 0 5

C O L L E G EOF SO U TH JE R S E Y

HB601 . S74

Stu d en sk i, P. in c o m e o f nation s

30498 DATE

NAME

C LA S S

•C O L L E G EOF

SOUTH JERSEY

30498

H B