The World Economy and National Economies in the Interwar Slump

The World Economy and National Economies in the Interwar Slump Theo Balderston

The World Economy and National Economies in the Interwar Slump

Also by Theo Balderston THE ORIGINS AND COURSE OF THE GERMAN ECONOMIC CRISIS, 1923–1932 ECONOMICS AND POLITICS IN THE WEIMAR REPUBLIC

The World Economy and National Economies in the Interwar Slump Edited by

Theo Balderston Senior Lecturer in Economic History University of Manchester

Editorial matter and selection and Chapter 1 © Theo Balderston 2003 Other chapters © Palgrave Macmillan Ltd 2003 All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No paragraph of this publication may be reproduced, copied or transmitted save with written permission or in accordance with the provisions of the Copyright, Designs and Patents Act 1988, or under the terms of any licence permitting limited copying issued by the Copyright Licensing Agency, 90 Tottenham Court Road, London W1T 4LP. Any person who does any unauthorized act in relation to this publication may be liable to criminal prosecution and civil claims for damages. The authors have asserted their rights to be identified as the authors of this work in accordance with the Copyright, Designs and Patents Act 1988. First published 2003 by PALGRAVE MACMILLAN Houndmills, Basingstoke, Hampshire RG21 6XS and 175 Fifth Avenue, New York, N. Y. 10010 Companies and representatives throughout the world PALGRAVE MACMILLAN is the global academic imprint of the Palgrave Macmillan division of St. Martin’s Press, LLC and of Palgrave Macmillan Ltd. Macmillan® is a registered trademark in the United States, United Kingdom and other countries. Palgrave is a registered trademark in the European Union and other countries. ISBN 0–333–73864–0 This book is printed on paper suitable for recycling and made from fully managed and sustained forest sources. A catalogue record for this book is available from the British Library. Library of Congress Cataloging-in-Publication Data The world economy and national economies in the interwar slump/edited by Theo Balderston. p. cm. Includes bibliographical references and index. ISBN 0–333–73864–0 1. Depressions – 1929 – 2. Gold standard – History – 20th century. 3. Economic policy – History – 20th century. 4. International economic relations – History – 20th century. I. Balderston, Theo. HB3717 1929 .W67 2002 330.9’043–dc21 2002026949 10 12

9 11

8 10

7 09

6 08

5 07

4 06

3 05

2 04

Printed and bound in Great Britain by Antony Rowe Ltd, Chippenham, Wiltshire

1 03

Contents List of Figures

vi

List of Tables

viii

Preface

ix

Notes on the Contributors

x

List of Abbreviations

xii

1 Introduction: The ‘Deflationary Bias’ of the Interwar Gold Standard Theo Balderston

1

2 Understanding the Great Depression in the United States versus Canada Pierre L. Siklos

27

3 France in the Depression of the Early 1930s Pierre Villa

58

4 Slump and Recovery: The UK Experience Michael Kitson

88

5 ‘Dancing on a Volcano’: The Economic Recovery and Collapse of Weimar Germany, 1924–33 Albrecht Ritschl

105

6 The Interwar Slump in India G. Balachandran

143

7 New Zealand in the Depression: Devaluation without a Balance-of-Payments Crisis John Singleton

172

8 The Soviet Union during the Great Depression: The Autarky Model Paul R. Gregory and Joel Sailors

191

9 Afterword: Counterfactual Histories of the Great Depression Barry Eichengreen and Peter Temin

211

Index

223

v

List of Figures 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.A1 2.A2 2.A3 3.1 3.2 3.3(a–d) 3.4 4.1 5.1(a–b) 5.2 5.3 5.4(a–d) 5.5 5.6 5.7(a–b) 5.8 5.9 5.10

Short-term interest-rate movements in Canada and the 31 United States, 1930–7 Price-level movements in the US and Canada, 1925–31 31 Business cycles in the US and Canada 33 Selected commodity prices in the US and Canada 1919–37 37 Actual and forecasted inflation: US and Canada, 1930–2 40 Evolution of estimates of measured to permanent income: US and Canada, 1920–36 41 Selected financial indicators in the US and Canada, 1922–37 43 The money supply in Canada and the US: M2, 1919–37 45 The behaviour of spot and futures commodity prices: US, 1920–37 50 The evolution of bank offices in the US and Canada, 51 1919–37 The highs and lows of the Canadian/US dollar 51 exchange rate, 1919–37 Aggregate supply and demand in France, 1926–38 66 Degree of factor utilization in France, 1928–38 66 Interest rates in France, 1928–34 78–79 The structure of money demand (M3) in France, 1928–34. 82 British and world gross domestic product, 1929–37 93 German central-government deficits and the fiscal 109 impulse, 1924–38 The instability of the Keynesian expenditure multiplier in Germany, 1926–38 110 The instability of the Keynesian consumption function in Germany 1926–38 111 Impulse-response relationships between income and consumption 113 New orders to the German machinery industry and 117 ‘Tobin’s Q’ in the Berlin stock market, 1925–35 German industrial labour productivity during the Great 122 Depression Relative PPP prices for Germany 125 Germany’s nominal trade balance 126 German bond issues abroad, 1927–30 130 Actual and simulated GNP with full reparation transfers, 133 1925–34 vi

List of Figures vii

8.1 8.2 8.3 8.4 8.5 8.6 8.7 8.8 8.9

GDP indexes: USSR and other countries, 1920–40 Capital formation indexes: USSR and other countries, 1920–40. GDP indexes: Russia and other countries, 1885–1913 Capital formation indexes: Russia and other countries, 1885–1913 Price indexes: Russia and other countries, 1885–1913 USSR exports and imports, current and constant prices, 1928–37 Composition of USSR exports, 1928–37 Composition of USSR imports, 1928–37 USSR and Europe: various indicators, 1928 and 1937

192 193 194 195 195 204 205 206 208

List of Tables 1.1 2.1 2.2 3.1 3.2 3.3 3.4 3.5 4.1 5.1 5.2 5.3 5.4 5.5 7.1 8.1

Global monetary gold stocks at the ends of 1913, 1925, 1929 Trade, capital and monetary movements: Canada, 1927–37 Modelling the behaviour of inflation The stabilization of the franc The entry of the French economy into the Depression Data on fiscal policy (all levels of government) The assets held as counterparts of the money supply (M2) Apparent average real interest rates UK interwar economic performance: some international comparisons Static Keynesian consumption functions Determinants of investment Money-demand functions Indices of real unit wage cost in the aggregate economy Short-run marginal returns to labour in Germany New Zealand and the Depression, 1926–38 Characteristics of Soviet industrialization, 1928 and 1937

viii

16–17 30 39 62 65 72 75 76 94 112 116 118 121 122 176 200

Preface The idea underlying this book is to offer a series of studies of various countries during the Great Depression in the light, broadly speaking, of the thesis relating it to the revival of the gold standard–the thesis principally associated with the names of Barry Eichengreen and Peter Temin. My grateful thanks are therefore firstly due to the scholars who made space in their busy schedules to write fine chapters for this book, and to Professors Eichengreen and Temin themselves for agreeing to write an ‘Afterword’ and delivering it in very short order. I wish to apologize to all these contributors for the lengthy delay in the completion of the volume, and to the publishers, whose remarkable forbearance deserves acknowledgment here. Of a different order is my debt to my wife, whose ‘worth is far above rubies’ (Proverbs 31:10). THEO BALDERSTON

ix

Notes on the Contributors G. Balachandran teaches at the Delhi School of Economics and the Graduate Institute of International Studies, Geneva. He is the author of John Bullion’s Empire: Britain’s Gold Problem and India between the Wars (1996) and The Reserve Bank of India, 1951–1967 (1998), and is editor of India and the World Economy (2002). Theo Balderston is Senior Lecturer in Economic History at the University of Manchester. His research interests are in the fields of the monetary and macroeconomic history of the interwar world. His publications include The Origins and Course of the German Economic Crisis, 1923–1932 (1993), and Economics and Politics in the Weimar Republic (2002). Barry Eichengreen is George C. Pardee and Helen N. Pardee Professor of Economic and Political Science at the University of California, Berkeley. He has published widely on the history and current operation of the international monetary and financial system. His books include Toward a New International Financial Architecture (1999), Globalizing Capital: A History of the International Monetary System (1997), European Monetary Unification (1997), and Golden Fetters: The Gold Standard and the Great Depression (1992). Paul Gregory is Cullen Distinguished Professor of Economics at the University of Houston, Texas. He is currently a distinguished visiting fellow at the Hoover Institution. His publications include Russian National Income 1885–1913 (1982), Before Command: An Economic History of Russia from Emancipation to the First Five-Year Plan (1994), and (with Robert Stuart) Russian and Soviet Economic Performance and Structure (7th edn, 2001). Michael Kitson is Lecturer in Economics at The Judge Institute of Management in the University of Cambridge. He is a Fellow of the Cambridge-MIT Institute (CMI) and Associate Director of CMI’s National Competitiveness Network. His publications include (with S.Solomou) Protectionism and Economic Revival: The British Interwar Economy (1990), and (with J.Michie) The Political Economy of Competitiveness: Essays on Employment, Public Policy and Corporate Performance (2000). Albrecht Ritschl obtained his degrees and his doctorate in economics from the University of Munich, Germany, in 1983 and 1987. His previous positions include professorships at University Pompeu Fabra, Barcelona/Spain, and at the University of Zurich/Switzerland. Since 2001 he has been Professor of Economic History at Humboldt University, Berlin/Germany. He has published extensively in the fields of interwar and early postwar German economic history. x

Notes on the Contributors xi

Pierre Siklos is Professor of Economics at Wilfrid Laurier University. He was University Research Professor for the year 2000–2001. He has also been a visiting scholar at several institutions including the IMF, Oxford University, and the University of California, San Diego. His publications include War Finance, Hyperinflation and Stabilization in Hungary 1938–48 (1991), the edited volume Great Inflations of the 20th Century: Theories, Policies and Evidence (1995), and The Changing Face of Central Banking. Evolutionary Trends since World War II (forthcoming). Joel Sailors is Professor Emeritus of Economics at the University of Houston. He has published widely in the areas of international economics, interstate US trade, and in economic history. John Singleton is Senior Lecturer in Economic History at Victoria University of Wellington, New Zealand. He is currently working on a history of the Reserve Bank of New Zealand. His publications include (co-edited with R. Millward) The Political Economy of Nationalisation in Britain 1920–1950 (1995), Lancashire on the Scrapheap: The Cotton Industry, 1945–70 (1991); The World Textile Industry (1997), and (with Paul Robertson) Economic Relations between Britain and Australasia, 1945–1970 (2002). Peter Temin is the Elisha Gray II Professor of Economics at the Massachusetts Institute of Technology (MIT). He was President of the Economic History Association, 1995–6, and President of the Eastern Economic Association, 2001–2. His most recent book is Engines of Enterprise: An Economic History of New England (ed.) (2000). Relevant previous publications include Lessons from the Great Depression (1989), and Did Monetary Forces Cause the Great Depression? (1976). Pierre Villa is Associate Professor of Economics at Paris IX Dauphine University, Administrateur at INSEE (the French Institute of Statistics) and scientific advisor at CEPII. His main topics of research concern historical economics, macroeconomic modelling and economic policy. The title of his PHD thesis was ‘A Macroeconomic Analysis of France during the 20th Century’, and the title of his complementary thesis is ‘A Macroeconomic Modelling of Financial Structures’. Previous publications include Une analyse macroéconomique de la France au XXème siècle (1993).

List of Abbreviations $C £NZ £stg bn CPI d. ET FEBD FEBS Fed Ff GDP GNP M2 M3 NBER OLS oz PPP Q QQ RBNZ RM Rs s SRIRL TFP UK US USSR VAR

Canadian dollars New Zealand pounds Pounds sterling Billion (thousand million) Consumer price index (in text) Pence (in the British pre-1971 coinage) (in tables and figures) Rate of change in Eichengreen–Temin Full-employment budget deficit Full-employment budget surplus Federal Reserve System French francs Gross domestic product Gross national product A measure of the money supply including notes, coin and total bank deposits A measure of the money supply including a wider range of liquid assets than M2 National Bureau of Economic Research Ordinary least-squares regression Ounce Purchasing-power parity Question Questions Reserve Bank of New Zealand Reichsmarks Rupees Shillings (in the British pre-1971 coinage) Short-run increasing returns to labour Total factor productivity United Kingdom United States Union of Soviet Socialist Republics Vector autoregression

xii

1 Introduction: The ‘Deflationary Bias’ of the Interwar Gold Standard Theo Balderston

Although the world’s Greatest Depression1 may be justly be regarded as a unique event, the lack of comparable historical events has not prevented scholarly explanation of it from arriving at a unanimity that is almost equally rare in the annals of historiography. This is now the approximate status of the thesis of Peter Temin (1989) and Barry Eichengreen (see especially 1992a) regarding the relationship between the gold standard and the Great Depression. The several chapters of this book were commissioned to review the Depression experiences of various countries in the light of their international monetary relations and broadly in the light of the Eichengreen–Temin thesis. None of them seriously dissents from what (begging their pardons) I will acronymize as the ET thesis. However, their varying emphases do add up to quite an illuminating commentary on it. This introduction firstly summarizes the ET thesis, and secondly seeks to comment on the substantive chapters, in such a way as to examine the roles of (a) diplomatic conflict, especially in relation to Reparations, (b) of the enlarged demand for gold as a consequence of interwar monetary uncertainties, and (c) of ideologies (note the plural) in the history of the gold standard and the Great Depression.

1

The ET thesis

ET attribute the severity of the great depression to international monetary interactions under the gold standard.2 They emphasize that these international interactions served as an intensifying propagation mechanism and explain the peculiar severity of the Depression. This is in contrast to the famous view of Friedman and Schwartz (1963), that the maladroitness of the US Federal Reserve monetary policy suffices on its own to explain this severity (Temin, 1989, pp. 43–54, 83f). As ever, simple statements mislead. Both writers emphasize that the depressive international interactions under the gold standard were made worse by the specific interwar economic and political environment, Temin (1989, pp. 1ff) emphasizing the magnitude of 1

2 Introduction

the economic disturbances resulting from the First World War, Eichengreen (1992a, p. 9) more the effects of the rise of Labour on the credibility of central banks’ commitments to defend exchange-rate parities. Eichengreen (1992a, p. 10), especially, argues that the gold standard worked worse after the First World War because poorer cooperation between central banks after the First World War exacerbated the monetary transmission of deflation. But for both, the counterfactual is that with a system of managed floating exchange rates the catastrophic repercussions of such more immediate and more remote shocks could have been averted (see Chapter 9). One element of the explanation of the catastrophe is epicyclical to their main thesis, namely the analysis of why the international monetary propagation of the slump translated not only into price declines, but also into output and employment declines (cf. Bernanke, 1995, pp. 16ff). Although ET have analysed why monetary deflation produced unemployment (cf. Eichengreen, 1991b, pp. 79ff; Bayoumi and Eichengreeen, 1994, 1996) the analysis stands outside the analytical core of their thesis. Therefore this introduction too, like most of the book, focuses on the monetary aspects assuming the output and employment consequences. It takes the following as its text: The gold standard of the 1920s set the stage for the Depression of the 1930s by heightening the fragility of the international financial system. The gold standard was the mechanism transmitting the destabilizing impulse from the United States to the rest of the world. The gold standard magnified that initial destabilizing shock. It was the principal obstacle to offsetting action. It was the binding constraint preventing policymakers from averting the failure of banks and containing the spread of financial panic. For all these reasons, the international gold standard was a central factor in the worldwide Depression. Recovery proved possible, for these same reasons, only after abandoning the gold standard. (Eichengreen, 1992a, p. xi) Emphasizing that the gold standard system had a deflationary bias – i.e. that it compelled countries losing gold to deflate but did not compel countries gaining gold to reflate – the ET account of the ‘propagatory mechanism’ focuses on the effects of the US and French balances of payments on the balances of payments of other countries in 1929 and succeeding years (Temin, 1989, pp. 22ff). In an immediate sense the story starts with the US Federal Reserve System (hereafter, the ‘Fed’) deploying domestic monetary policy to control the Wall Street boom of 1928–9. US monetary contraction raised US interest rates and reduced US foreign lending; and this particularly affected the capital balances of the countries which until 1928 had been large international borrowers – Latin American and Australasian primary producing countries, and Germany and other countries of Central

Theo Balderston 3

Europe (Eichengreen, 1992b, p. 221).3 From mid-1928 US monetary deflation was also aggravated by French gold accumulation. The causes of French reserve accumulation since 1926 are complex (see Chapter 3 by Villa); its effects were intensified after June 1928 by the increase of gold and decrease of foreign assets because (in Eichengreen’s, 1986, p. 91, view) with the legal re-establishment of the gold standard the Bank of France was prohibited from buying foreign exchange for francs (Eichengreen, 1986, p. 91).4 The global money supply grew much more slowly in 1929 than in the preceding two years (Eichengreen, 1992a, pp. 222f). Some borrowing countries which lost much gold in 1929 moved rapidly to suspend or restrict convertibility: Australia, Argentina, Brazil, Paraguay, Uruguay and Canada in 1929. Reserve losses spread in 1930, and, together with more currency depreciations,5 probably signalled to the New York capital market that the countries concerned were overborrowed and in danger of forms of default beyond redress through the US legal system.6 Maybe lenders interpreted the balance of payments problems as deep-seated on account of the longer-term decline in world primary-product prices – itself perhaps traceable to the general restoration of the gold standard around 1925. For the logic of the ‘gold standard’ thesis the generality of the ‘credit crunch’ in 1929–30 needs more explaining. Before 1914 national ‘overborrowing’ had occurred in different places at different times; and the collapse of creditworthiness in one country had soon been offset by new opportunities in another. The generality of the credit crunch in the later 1920s cannot be disconnected from a general ‘bunching’ of foreign borrowing by many countries in the mid-1920s. This ‘bunching’ partly reflected a postwar backlog of investment opportunities; but also the rhythm of US demand for foreign bonds. This in turn was affected by US monetary policy and the level of US government funding operations in the early 1920s (Abbott, 1937, pp. 176f). To the extent that (i) the postwar slump was a product of the Fed’s monetary restriction in the early 1920s in defence of US gold convertibility in the face of a shrinking gold stock (Eichengreen, 1992a, p. 117), and that this temporarily deterred US foreign bond flotation, and (ii) that the borrowers’ resumptions of the gold standard in the mid-1920s increased their borrowing powers, the bunching of foreign lending in the mid 1920s was also aggravated by gold-standard policies. Monetary expansion was not pursued in the US in early 1930, lest it rekindle the stock-market speculation, and because of commitment to that policy of ‘liquidation’ of prices and speculative positions, which in the early 1920s had apparently laid the basis for prosperity in the remainder of the decade. In 1930 ‘[r]eserves surged towards the principal net foreign creditors’ (Eichengreen, 1992a, pp. 247, 251), denuding the debtor countries. Reserves losses in debtor countries were accompanied by contractions of their monetary bases, though much less so in debtor countries which depreciated (see also p. 10 below).7

4 Introduction

The focus of attention of the ET explanation of the Depression then moves forward to the European financial crisis of summer 1931. Here the accent of their argument moves to another sentence of the above quote: ‘[The gold standard] was the binding constraint preventing policymakers from averting the failure of banks and containing the spread of financial panic.’ Eichengreen unambiguously maintains that failure of international inter-central-bank cooperation was the root cause of the Austro-German financial crisis.8 In his view (e.g. 1996b, pp. 32ff), successful international cooperation was the sine qua non of credible gold-convertibility commitments, which in turn were the sine qua non of a successful gold standard. Using examples such as the Bank of France’s assistance to the Bank of England during the Baring crisis of 1890, he argues that before 1914 the perceived self-interest of each central bank in the survival of the gold standard system as a whole had motivated non-threatened central banks to finance the maintenance of gold payments by endangered central banks – particularly those critical to the gold-standard system (Eichengreen, 1987). Cooperation failed in 1931, especially over the German crisis, and for three or four reasons. The first was the failure of the French and Americans to appreciate the seriousness of the German banking crisis and hence of the conflict between the Reichsbank’s gold-standard commitment to sell gold freely at a fixed price and its lender-of-last-resort duty to save commercial banks from illiquidity during panics. The second reason was that such cooperation as occurred was meagre because guided by a gold-standard mentality (Temin, 1989, pp. 83ff; cf. Eichengreen, 1992b, p. 276), or impeded by conflicting French and Anglo-Saxon diagnoses of the causes of the wider slump (Eichengreen, 1996b, p. 77). The third reason was ulterior and conflicting foreign-political agendas, especially relating to Reparations, which poisoned Franco-German relations (Eichengreen, 1992a, p. 278; cf. Ferguson and Temin, 2001). Given the failure of cooperation, the fourth and ultimate reason for the failure to ‘rescue’ the gold standard was the scale of the assistance required to mount a credible defence of the reichsmark, let alone sterling. Total declared US, French and British gold reserves at the end of June were $7.6 bn, but free reserves (i.e. reserves in excess of statutory minimum note cover) were much less. Credits of this magnitude threatened the US gold cover ratios, hence confidence in the dollar, and hence even US gold payments (Eichengreen, 1992a, p. 286). Kindleberger’s argument (1973, pp. 291ff) that the UK couldn’t lend and the US wouldn’t lend is not true: given the gold standard rules, the US couldn’t lend either. Since diplomatic suspicions prevented a joint British–French–American rescue package for Germany, lack of ‘gold capacity’ was the ultimate obstacle to an adequate policy response to the financial crisis. But what made the task so huge was the collapse of the credibility of parities. Before the war when an exchange rate had depreciated towards its gold export point, the certainty that it would revert towards its central value raised the

Theo Balderston 5

prospect of exchange-rate gains, and stimulated capital inflows. Now such a depreciation triggered the expectation of surrender of the parity and thus generated expectation-fulfilling capital outflows. The postwar rise of class conflictual politics in European countries had politicized monetary policy, especially interest-rate policy, and portrayed a parity-defending exchangerate policy as inimical to the provision of jobs. So the financial markets perceived monetary authorities as having lost their nerve in defending parities, which therefore lost credibility (cf. Simmons, 1994). This loss of credibility itself rigidified the system. In the nineteenth century metallic standards had operated as a ‘contingent rule’. Specified emergencies, if seen as not caused by domestic economic policies, had permitted countries to move temporarily to paper standards without extinguishing the market expectation of the restoration of the unchanged metallic standard once the emergency was surmounted (cf. Bordo and Kydland, 1996). Wars, such as the US Civil War, had been the clearest cases of such emergencies, but other disasters could perhaps have been so treated by the markets. But the weakening credibility of parities circumscribed resort to this ‘escape clause’; and this may perhaps explain the obstinacy and desperation of the German defence of the reichsmark parity in 1931 (Eichengreen, 1996a, p. 380). This catastrophic chain of events in 1931 should be contrasted with the global slump of ten years earlier. As already stated, that slump had commenced with the Fed sharply raising its discount rate in January 1920 in reaction to gold losses – mainly to primary producing countries benefiting from the buoyancy of agricultural demand and prices. It had been prolonged by the maintenance of high Fed discount rates into 1921 (despite the healthy recovery of gold stocks) in conformity to a doctrine of ‘liquidating’ postwar speculative excesses (Eichengreen, 1992a, p. 120).9 However, in the early 1920s the international propagation of these deflationary impulses from the USA had been damped by the number of European currencies whose convertibility was not only suspended but not even yet being targeted by their central bank. Such countries allowed their currencies to depreciate relative to the dollar (cf. Holtfrerich, 1991). In the early 1920s, economic revival in such countries had offset the monetary deflation induced by the US policy in countries such as the UK and its dominions, Sweden, Greece, Netherlands and Portugal, which were targeting a fixed dollar exchange rate. The 1931 European financial crisis ended on 21st September when Britain suspended gold payments and allowed the $ exchange rate of sterling to fall by 30 per cent by the end of the year. Over twenty countries’ exchange rates depreciated against the dollar in the autumn of 1931. Most countries which (a) remained on the gold standard and (b) had escaped serious financial crises gained gold substantially in the immediate wake of the European financial crisis – Belgium, France, Netherlands, Switzerland. These gains, in good part, were the product of official conversions of dollar

6 Introduction

reserves into gold, following the capital losses on their sterling reserves suffered by central banks in September 1931. US gold reserves fell by 12 per cent from the end of August to the end of the year, after having risen by 6 per cent over the preceding four months of the European crisis. Total global gold-plus-foreign reserves contracted, mainly due to this liquidation of sterling and dollar reserve assets; this caused an implosion of the global monetary stock and the global annus horrendus of 1932. This happened although the Fed’s bill discounting and purchase rose in the autumn of 1931 in response to US commercial bank failures. Fed policy financed, in effect, the nonbank sector’s associated shift out of deposits into currency. But the Fed felt unable to counter this shift by even more substantial open market purchases of securities, partly for fear of exacerbating the leakage of gold abroad, and partly because under current legislation the Reserve Banks lacked sufficient legal non-gold reserve assets to counterbalance a large increase in their currency liabilities. Instead, the Fed lent ‘freely but at high interest rates’, raising its discount rate by 2 per cent between August and October (Eichengreen, 1992a, pp. 295–8). The passage of the Glass-Steagal Act of February 1932 relaxed the stipulations governing the non-gold element of the Fed’s required note-issue cover, and it engaged in a $1bn open-market purchase operation in the spring of the year. But the resultant increase in the money supply failed to arrest the decline in production. This failure became a cause celèbre in the later 1930s in the case against the efficacy of monetary policy. It failed, says Eichengreen, because the market correctly divined that the purchasing programme was just a hasty ploy to avert more radical proposals before Congress, and that it would cease (as it did) once Congress adjourned. Since domestic money demand was not stimulated by the operation, the increase in the money supply merely accelerated the gold drain to abroad.10 Some Reserve Banks reached the margins of their gold cover limits; others felt deterred from helping those in trouble for fear of impairing their own lender-of-last-resort capacities in the event of renewed bank failures. These problems came to a head in the banking and convertibility crisis of spring 1933; triggered, in the last analysis, by the effect of the election of Roosevelt on the credibility of the US commitment to the gold standard11 – in fact he suspended the parity in April. Once the US had again stabilized the gold value of the dollar in January 1934 (at a 41 per cent devaluation), it became the recipient of gold inflows on a scale unprecedented except in 1921. But a residual inhibition continued to dominate US monetary policy: the ‘golden fetters’ on the minds of the policy-makers that instinctively chose gold-stock-conserving policies in preference to expansionary ones. The Fed continued to resist extensive monetary expansion (Eichengreen, 1992a, pp. 292f, 343, 387). After 1933 the ‘gold bloc’ countries pursued restrictive monetary policies as long as they tried to defend their parities; and even the countries which had left

Theo Balderston 7

the gold standard were generally slow to exploit their new-found monetary liberty. Without monetary expansion, the beneficial effects of depreciation in switching home demand from imports to home goods, and switching foreign demand to the depreciator’s goods, were more or less offset by the reverse effects on their foreign trading partners; so that, without monetary expansion, depreciation did little for global aggregate demand. Britain’s heavy gold accumulations after 1931 thus negated, in global terms, the domestic expansionary effects of depreciation (Temin, 1989, pp. 32, 74). With monetary expansion depreciation would effectively expand the global money supply, reducing global real interest rates, and stimulating global aggregate demand. But in countries with inflationary memories of the early 1920s, it was feared that abandoning gold convertibility and then shedding the ‘golden fetters’ would knock away a vital ‘crutch’ to domestic political stability. In such countries, commitment to the gold standard ideology had enabled opposed parties to compromise their ideological claims regarding government taxation and spending without alienating their political constituencies, and thus to curb the bitter budgetary and inflation-generating conflicts of the early postwar years (Eichengreen, 1992a, p. 394). Although uncoordinated floating with monetary expansion was a preferable anti-Depression policy to the gold standard, Eichengreen implies, by treating the Tripartite Agreement of 1936 as the goal of his narrative (1992a, pp. 348ff), that jointly managed international floating with substantial exchange-rate stability would have been the ideal. Temin (1989, pp. 91ff) argues that the ‘gold standard ideology’ exerted its depressive influence through the attitudes of business as well as the policymakers. The policy prioritization of the external balance, which the gold standard entailed, conditioned business to expect unemployment and low profits – particularly in the 1920s’ environment of severe global balanceof-payments disequilibria. The abandonment of the gold standard was a ‘regime change’ which, in certain countries, amounted to the promise of policy interventionism in aid of full employment – ‘socialism’, Temin calls it. This induced business optimism and investment (cf. Temin and Wigmore, 1990). Underlying this analysis is an impressive series of lucid and complete empirical studies of elements of the overall thesis (many collected in Eichengreen, 1990). The analysis appeals to economists with its clear policy implications for modern international monetary management (cf. Bernanke, 1995). Its compelling logic has reshaped historians’ understandings of the Great Depression (e.g. Clavin, 2000). Yet it appeals to them also through the complex, historically sensitive story it weaves out of the multiple strands of changing postwar global commodity, capital and labour market conditions, and national and international politics. This introduction will now utilize the several national analyses in the following chapters to comment on the ET thesis.

8 Introduction

2

The chapters of the book

Siklos’s authoritative survey of recent research on the Great Depression in North America (Chapter 2) indicates that the challenges of identifying both the initiating shock, and the influence of monetary contraction in 1930, e.g. on unanticipated deflation, continue to fascinate economists and to elude consensus. The explanation of the highly parallel economic behaviour of the USA and Canada also remains a puzzle, partly because of the simultaneity of the interrelationships between the two economies. Did both suffer a common shock, or was the Depression transmitted from the USA to Canada? In 1928–9 the Canadian gold losses could be held to have transmitted deflation to Canada, unchecked by domestic expansionary monetary policy even after the floating of the Canadian dollar in late 1929. Thereafter Canada didn’t lose much gold except in 1931, nor did the Canadian dollar depreciate much against the US dollar except in the period of the sterling depreciation – also the period of the US banking crises (October 1931–c. April 1933) (cf. Figure 2.A3, p. 51). Perhaps in the earlier part of the Depression the integration of the Canadian with the US financial system was such that incipient losses of Canadian banks’ US-dollar reserves in New York prompted immediate action (e.g. tightening of loan criteria) and rapidly harmonized the Canadian with the US money supply (Canadian monetary contraction proceeded faster than that of the US in 1929/30 – Figure 2.8, p. 45). However, the period of depreciation, which was also a period of lesser monetary contraction in Canada, did not noticeably benefit Canada. This was despite the greater resistance of Canadian than US banking to collapse and hence smaller rupture of the mechanisms of credit intermediation between savers and lenders. Canadian recovery from the slump was slower than that of the US (cf. Figure 2.3, p. 33), as perhaps reflected in the striking trade surpluses of the later 1930s. Maybe this shows in general that the Canadian authorities could not much influence Canada’s real exchange rate vis-à-vis the USA (not even in relation to non-tradables). This would mean that autonomous monetary expansion could not have reduced real domestic interest rates, because it could not have engineered an exchange-rate depreciation that raised the expectation of subsequent appreciation. This would raise the question of whether small open economies possess an incentive to practise a globally beneficial monetary expansion. Villa’s closely argued chapter on France (Chapter 3) clearly blames the gold standard regime for the French gold accumulation, and justifies the ET argument on this subject (if not the details of Eichengreen’s view of the ‘franc Poincaré’). Villa argues that the Bank of France’s foreign-reserve accumulations originated in the trade surplus caused, in Keynesian fashion, by the ‘fiscal surplus’ regime inaugurated by Poincaré, coupled with an interest rate level insufficiently low, given the absence of expected exchange rate

Theo Balderston 9

depreciation, to persuade French investors to invest sufficiently in foreign assets. Instead they wished to hold domestic money and domestic bonds. And this demand for bonds also reduced firms’ demands for bank credit, and hence retarded the expansion of bank deposits. The result was that the trade surplus plus repatriation of French wealth furnished the Bank of France with a net supply of reserve assets, and the resultant expansion of central-bank money met the rising demand for money that the commercial banks, with their sluggishly growing loan portfolios, could not meet. Both fiscal and interest rate policies were thought necessary to rebuild the credibility of the gold-standard franc. Lower French interest rates were in any case inadvisable, as French prices were tending to rise and French real interest rates at the end of the 1920s were close to zero. In a floating exchange-rate regime the balance of payments surplus would have caused the franc to appreciate and increased the attraction of investing the surplus abroad. But with fixed exchange rates and an anti-inflation target, the authorities’ room for manoeuvre to lower interest rates was slight. The only solution would have been to eliminate the balance of payments surplus by an agreed real appreciation of the franc against the pound and the dollar. Thus in the end the choice of exchange rate in 1926, whether accidentally or by design (cf. Sicsic, 1992; Eichengreen, 1996b, p. 64), caused the trouble. Given this, Villa clearly holds the dictates of a gold-standard regime as such responsible for the French gold accumulation; against Eichengreen (1986) he denies that specific regulatory restrictions on the Bank of France’s actions were the reason (though his argument does not entirely explain why the Bank of France accumulated gold rather than foreign reserves especially from 1928). Kitson’s lucid chapter on the UK (Chapter 4) sets the story in a more explicitly international context than the others. Like ET, he is convinced that, given the uneven competitive positions of the various major industrial countries following the First World War, the ‘deflationary bias’ of the gold standard coordinated their economic activities in such a way as to generate the Great Depression. His reading of the ET view is that they think that in the absence of credibility problems and with more public-spirited international cooperation, a gold-standard world could have avoided the Great Depression. By contrast he maintains that under all circumstances the rigidly fixed exchange rates would have forced the deflation of domestic demand in less internationally competitive countries, and that lending from, without domestic reflation in, the surplus countries could only temporarily have staved off the global consequences of this deflationary bias: even the best-functioning gold standard would inevitably have produced a Great Depression sooner or later. On the other hand he does not regard the monetary mechanisms of the gold standard, on which ET lay weight, as having transmitted deflation to the UK in 1930: the UK money supply expanded at that time. He does not say whether he thinks

10 Introduction

that this monetary expansion laid up trouble for the next year, and caused the financial crisis. But even so, Britain’s quick response of abandoning the gold parity spared her the evil consequences of the crisis and ushered in a period of remarkable recovery. Kitson’s argument calls in question the power of the chain reaction of monetary deflations, emphasised by ET, to transmit depression in 1930. The UK shared its modest monetary expansion of 1930 with a group of six creditor countries identified by Eichengreen (1992a, pp. 247ff) as also including Belgium, France, Netherlands, Switzerland and the USA. The aggregated expansion of the monetary bases in 1930 of the five creditor countries excluding the USA may be estimated as $697m – greater than the contraction of the monetary bases of 34 debtor countries ($595m) in the League of Nations sample.12 The aggregated expansion of an approximation of M2 in the creditors ($1,380m) was also somewhat greater than the aggregated contraction ($1,001m) in the debtors.13 Only when the USA is included does the M2 expansion of the creditors drop to $465m. Of course, a growing global economy needs net growth in its money supply; still, excluding the USA, global ‘deflationary bias’ was not excessive in 1930. If eight ‘debtor’ countries had not depreciated their currencies the deflationary bias would have been fiercer, but the significance of this is not clear, since before 1914 smaller countries in payments difficulties had depreciated without thereby breaching the gold-standard system. These calculations suggest that the crux of the ET claim that the goldstandard propagatory mechanism turned a recession into a catastrophe depends on the financial crisis of 1931 having been an endogenous consequence of the prior monetary contraction.14 For in themselves the effects of the 1931 crisis tell us more about the effects of the messy collapse of a particular monetary system and the associated implosion of international ‘key-currency’ reserve assets than they do about the ordinary functioning of the gold standard. Since it was the shockwaves of the German crisis that converted the local Austrian crisis into a global one, Ritschl’s exhaustive analysis in Chapter 5 is central to the question of the ‘endogeneity’ of the 1931 crisis to the operation of the gold-standard system. In his analysis the ‘German slump’ has little to do with the Wall Street boom, with the recession in the USA from 1929, or with US monetary policy in 1930–1 (see, too, Ritschl, 1999). Though he tells a complex story deftly interweaving domestic and international dynamics, its main lines are determined by the vicissitudes of Germany’s Reparations politics. The ‘Dawes Plan’ of 1924, designed under benign US influence as a relief to Germany from the ‘London Plan’ of 1921, stimulated capital inflows into Germany; whereas the revised reparations stipulations of the 1929 ‘Young Plan’ stimulated capital outflows that turned into a stampede in the 1931 financial crisis, having already forced chancellor Brüning’s fiscal retrenchments by making it impossible for the government to borrow.15 True,

Theo Balderston 11

Ritschl agrees that it was the gold standard that tied German monetary conditions to her abruptly changing capital balance, and translated her foreign-borrowing crisis into domestic monetary contraction. His argument implies, however, that even under a floating rate the ‘Young Plan’ would still have produced a binding balance-of-payments constraint and necessitated severe current-account readjustment. The gold standard was itself integral to Germany’s strategy of pursuing agreed Reparations revision and eschewing unilateral action (cf. Krüger, 1985). The USA and France would not, presumably, have been happy to see Germany float the reichsmark in autumn 1931. The payoffs underlying this cooperative strategy were only falsified by the unforeseeable balkanization of world trade, and cessation of international lending in the 1930s.16 Thus, on Ritschl’s argument, the German crisis of 1931 was really the product of Reparations and Reparations politics. Recently Ferguson and Temin (2001) have analysed the German financial crisis as an almost accidental consequence of the effect of the resurgent German Mitteleuropa and Reparations policy on French willingness to lend the credits needed to bail Germany out. Such analyses invite us to weigh the relative contributions of political conflicts and the gold-standard system to the causation of the financial crisis and hence the economic catastrophe. I cannot resist the following flight of fancy: Suppose that in the USA in the middle 1920s the agricultural, other debtor, silver, and taxpayer interests had exerted themselves against the creditor interest in opposing gold sterilization, and in supporting monetary expansion and higher prices. Monetary expansion would not (it could be claimed) have forced the USA, with its ample gold reserves, off the gold standard, and so would not have been inconsistent with a gold-standard ideology. Suppose that higher taxpayer contentment had thus given the USA the latitude to show a more forgiving attitude towards French war debt, and thus enabled, in the Paris Reparations negotiations of 1929, a clearer continuation of the US diplomacy of the middle 1920s – i.e. of minimizing the Reparations burden on Germany in order to strengthen the German market for US goods, and German democracy (Link, 1970; Leffler, 1979).17 Ritschl’s argument implies that US lending to Germany might well have resumed. Suppose that a much milder ‘Young Plan’ had also eviscerated the Hugenberg–Hitler agitation against it, and by denying the Nazis the ‘oxygen of publicity’ had prevented their trail of rising electoral successes from the end of 1929 through to their great gains against the conservative and bourgeois right in the Reichstag elections of September 1930. Suppose that therefore the French had been able to view Curtius’s Austro-German customs union proposal as consistent with the ‘Briand plan’ for a European customs union, and Brüning had not felt forced into his ‘Reparations Declaration’ of June 5th 1931. Suppose with Ferguson and Temin (2001), that the German financial crisis would then have been either

12 Introduction

averted or reparable by means of a Franco-US financial package. Objections could be raised to all of these links – but the sequence is not unthinkable. Britain might have experienced neither the destabilizing collapse of demand for sterling as a reserve asset, nor the exchange rate crisis which enabled her to reverse the ‘mistake’ of 1925. The Fed would not have faced gold losses in autumn 1931, though currency depreciations in agricultural countries might still have put pressure on US farm prices and hence on the US country banks. But Villa’s chapter implies that the French gold accumulations might well have ceased; on any argument they could not have gone on for ever. In these circumstances, could a ‘Great Depression’ have been averted? ET would perhaps reply (see their contribution to this book) that the above adventurous counterfactual misses their point, namely, that however catastrophic the international diplomatic constellation, a floating-rate system would have mitigated, where the gold standard aggravated, its global monetary repercussions. Their argument could be formulated as the claim that, given the continuing destructive fallout of Reparations politics, the gold standard turned the recession into a catastrophe by causing the domino effects of the 1931 financial crisis (cf. Eichengreen, 1992a, p. 278). Their counterfactual is a world of floating exchange rates but the same Reparations politics. In it, presumably the Fed would have practised the same anti-Wall Street deflationary policy in 1928–9, but rising US interest rates would have caused other currencies to depreciate relative to the dollar (the converse happening after the Wall Street crash). However exchange-rate floating would not have prevented the fall in lending to the periphery if this was due to fear of default. The periphery – and Germany – could by depreciation have mitigated the domestic employment consequences of the resultant constraint on their external balances; but in the German case depreciation might not have averted bank insolvency,18 and thus also not have prevented the repercussions on US financial stability alleged by Sumner (1997). Thus the general overborrowing of the 1920s might still have produced a more limited financial crisis in the early 1930s, though floating exchange rates would have prevented a catastrophic unemployment crisis. Either without Reparations but with the gold standard, or with Reparations (plus other causes of abrupt shift in demand for international financial assets) but without fixed exchange rates, the Great Depression would have been less great. ET minimize the impact of the Reparations regime on capital flows to Germany; but they also identify the gold standard as the systemic cause of the slump, partly because this attribution yields present economic-policy recommendations on how to avoid systemic causes of global slump. But since Reparations were the principal obstacle to the successful reconstitution of the concert of great powers which until 1914 had prevented general war in Europe, and hence also to successful inter-central-bank cooperation during gold-standard crises, the

Theo Balderston 13

historian would see an argument for claiming that they were the more fundamental cause of the crisis. This would resuscitate, if on different arguments, a contemporary belief that Reparations caused the slump (Cassel, 1932). The foregoing counterfactuals focus on the gold standard viewed as a fixed exchange-rate system, and on the instability of the demand for ‘key currencies’. But a contemporary commonplace was to blame gold itself, viewed as the reserve asset, and specifically, a long-run excess demand for monetary gold, as the impediment to global economic stability.19 Balachandran’s scintillating chapter (Chapter 6) emphasizes India’s centrality to the gold standard, thus conceived, and hence Britain’s manipulation of Indian monetary policy for its own ends. His account suggests an extraordinary British official acquisitiveness for gold. He traces a consistent motive in the British management of Indian currency policy, remarkable in the fact that the demise of the gold standard made no difference to it. From the futile attempt at a high gold-stabilization of the rupee in 1920; through its lower – but still high – stabilization at the time of the restoration of sterling convertibility; to the pegging of the rupee to sterling in September 1931, the real aim, he says, was to prevent the lower real exchange rate that would tend to raise the real incomes of rural producers of exportables, and stimulate Indian absorption of gold, which was viewed as a menace threatening UK gold reserves. However, it was not until the price deflation of the Great Depression reduced rural incomes in India that the UK-dictated policy succeeded in engineering a gold outflow from India. According to Neville Chamberlain, this ‘astonishing gold mine’ ‘put Britain in clover’, and greatly insulated the sterling float from the pressure of French gold repatriation. Not all agree that greed for Indian gold originally prompted the demand that the rupee be pegged to sterling (e.g. Drummond, 1981, pp. 28–51), but it seems indubitable that this consideration would have strengthened subsequent insistence on the peg. British gold reserves quadrupled between the ends of 1932 and 1938 and increased by one-third more per quantum than the French reserves between 1926 and 1932. The annual rate of US per quantum accumulation of central gold stocks between 1933 and 1938 (16.5 per cent) was about three times the US average of 1914–29 (5.5 per cent).20 In Europe, only the central banks of Germany, Italy, Spain, Hungary, Poland, and Denmark held a smaller quantum of gold in 1935–8 than their peak holding of the 1920s. And of these only Germany, Denmark and Hungary held a significantly smaller dollar values of gold. The quantum of gold held by Australia, New Zealand and Canada (not South Africa) and Latin American countries (where exchange controls were normal) was less in 1935–8 than their 1920s peak.21 But the sterling area as a whole held a much larger quantum of gold in the 1930s than in the 1920s due mainly to the huge British accumulations. Concurrent with this was the already-mentioned collapse in demand

14 Introduction

for foreign-currency reserve assets, which in 1938 formed 12 per cent of the declared reserves of the countries, other than the USA, reported in the League of Nations statistics, as against 34 per cent in 1929.22 The failure to rebuild the foreign asset element of their reserves even after the dollar devaluation suggests that the gold-exchange standard of Genoa intrinsically did not work; perhaps because reserve currencies were always more likely to be devalued against gold than vice versa, but more possibly because, as the Bank of France and the Reichsbank had found out in the 1920s, these official foreign-exchange balances were not fully liquid in terms of gold, despite the convertibility commitments of the gold-centre countries. Significant reductions of these balances had to be negotiated with the gold-centre country whose liability they were. Gold, however, was no one else’s liability. This inference seems to be supported by the fact that the larger reserve-holders reduced their relative foreign-asset holdings the most.23. At any rate, the collapse of the gold standard left the world more dependent on gold as the reserve asset, not less. Under conditions of totally free floating, central banks theoretically need no reserves at all. But countries did have price-stability targets; real exchange rates vary dramatically under floating exchange-rate regimes (Eichengreen, 1991a, pp. 241ff), and floating currencies are always subject to exogenous influences. Evidently central banks of currencies without strict exchange controls still felt the need of large gold reserves for managing their floats. Countries whose foreign reserves after 1933 were less than in 1929 were generally Latin American and East European countries with exchange controls.24 Gold production increased in the 1930s, partly because of the lower commodity price level, partly because of the devaluation of the South African pound; and Indian non-monetary gold was exported. But probably the single main reason for the decline in real interest rates in the 1930s was Roosevelt’s decision to devalue the dollar in 1933–4 (cf. Nurkse, 1944, pp. 131ff). This eased the pressure for gold stocks and allowed the distribution of gold stocks to be maintained by a lowpressure, low interest-rate system, in contrast to the high-pressure, high interest-rate system of the 1920s.25 An excess global demand for monetary gold at interest rates consistent with domestic high-employment macroeconomic objectives is critical to the ET thesis of ‘the deflationary bias’. Without this excess demand, other components of their thesis might suggest an inflationary bias. The argument, that the emergence of anti-deflationary, class-conflictual national politics made the sacrificing of parities to internal high-employment policies more likely, implies such a bias. This is particularly so when conjoined with the emphasis on central-bank cooperation, which suggests that central banks of major countries might be the more emboldened to sacrifice foreign reserves to domestic pressures by the belief that they would be bailed out of consequential crises by other central banks. ET argue that

Theo Balderston 15

this global excess demand for gold was simply a US and French excess demand which denuded the rest of the world.26 But Table 1.1 shows the paradox. Certainly if we consider column 8 in which deflated total monetary gold stocks at the end of 1929 are expressed as indexes of the same at the end of 1913, much of Europe including France, but not Latin America or the leading primary producers and Japan, was relatively short of reserves. However column 6 (in which deflated central gold reserves at the critical turning point of the end of 1929 are expressed as indexes of central gold reserves at the end of 1913) represents ET’s preferred measure, since gold taken out of circulation into central reserves acts as a kind of global ‘super-high-powered’ money controlling the global money supply. And on the basis of column 6, not just France but the world outside Central and Eastern Europe was not short of monetary gold in 1929. Eichengreen does not generally credit the view of contemporary experts that irrational note-cover ratios immobilized gold, leaving countries short of ‘excess’ gold.27 Thus the ET argument seems to require an upward shift (compared with pre-war) in the trade-off between central-bank demand for gold, relative to level of foreign trade, etc., and the satisfaction of internal targets. The persistence of large gold reserves through the 1930s suggests that maybe central bankers welcomed high cover ratios, even though these limited their own freedom of manoeuvre, because these justified the gold accumulations which the Bankers thought necessary on other grounds even in the 1930s. Risk aversion could supply the rationale for this heightened demand (cf. Roberts, 1931, pp. 46ff; Nurkse, 1944, p. 74). If central bankers regarded the level of their gold stocks as less controllable than pre-war, partly because of the politicization of monetary policy at home, and partly because of the greater unpredictability of events abroad and their effects on the competition for gold stocks;28 and if they believed that without internal circulation of gold coin they lost an interest-elastic source of replenishment they had had before 1914, they may well have substantially upped their target gold stocks relative to trade levels, whether under gold-standard or floating-rate regimes. Table 1.1 (panel 3) certainly shows that the interwar era was marked by huge swings in the distribution of reserves – to the primary exporting countries (including the USA) and Japan in the war and early postwar years, denuding Europe; towards France (and not away from the USA) in the Depression era, denuding the primary-good-exporting periphery; and to Britain and the USA in the 1930s, denuding the same countries. Economic historians work backwards from evidence of changes in reserves flows to inferences about how the markets refracted present policy configurations into expectations of future policy (e.g. Eichengreen, 1996b, pp. 55–7). But are these inferences right? And can we wonder if contemporary policy-makers doubted that even the power to depreciate gave them sufficient command over the ‘mind’ of the market to enable them to control the level of their gold reserves?.29

16 Table 1.1

Global monetary gold stocks at the ends of 1913, 1925, 1929 1. Nominal values 1913 1913 Central Total reserves monetary gold~ ($mill) ($mill) (Col. 1) (Col. 2)

UK & European neutrals# France Belgium, Germany, Italy Russia/USSR## Other Europe Latin America Leading primaries* & Japan USA World

423 679 405 787 543 348 475 1290 4945

1060 1700 1135 1041 603 487 731 1924 8773

2. Deflated** 1925 Central reserves (Col. 5) UK & European neutrals# France Belgium, Germany, Italy Russia/USSR## Other Europe Latin America Leading primaries* & Japan USA World

229 62 78 7 23 116 171 184 109

8.6 13.7 8.2 15.9 11.0

1929 Central reserves

($mill) (Col. 3)

($mill) (Col. 4)

1627 711 534 94 212 678 1366 3985 9074

1660 1631 903 198 323 727 1187 3900 10949

indexes: 1913=100 1929 1925~ Central Total reserves monetary gold (Col. 6) (Col. 7) 267 163 152 17 40 142 170 206 151

3. Distribution 1913 1913 Central Total reserves monetary gold (per cent) (per cent) (Col. 9) (Col. 10) UK & European neutrals# France Belgium, Germany, Italy Russia/USSR## Other Europe

1925 Central reserves

12.1 19.4 12.9 11.9 6.9

1929~ Total monetary gold (Col. 8)

95 37 29 5 20 101 125 136 69

110 65 55 10 41 129 123 151 87

1925 Central reserves

1929 Central reserves

1936 Central reserves

(per cent) (Col. 11)

(per cent) (Col. 12)

(per cent) (Col. 13)

17.9 7.8 5.9 1.0 2.3

15.9 15.7 8.7 1.4 3.6

30.0 11.0 3.2 3.3 2.3

Theo Balderston 17 Table 1.1

Global monetary gold stocks at the ends of 1913, 1925, 1929 3. Distribution 1913 1913 Central Total reserves monetary gold (per cent) (per cent) (Col. 9) (Col. 10)

Latin America Leading primaries* & Japan USA World

con’t

1925 Central reserves

1929 Central reserves

1936 Central reserves

(per cent) (Col. 11)

(per cent) (Col. 12)

(per cent) (Col. 13)

7.0

5.6

7.5

7.0

2.6

9.6 26.1 100.0

8.3 21.9 100.0

15.1 43.9 100.0

11.4 37.4 100.0

4.3 41.5 100

~ Total monetary gold adds gold in circulation and in commercial banks to central reserves. This was estimated in the sources as c. $3.8 bn in 1913, $1.2 bn. in 1925 and $0.8 bn. in 1929. * Argentina, Australia, Canada. ** UK foreign trade prices: 1925 = 168.1; 1929 = 147.1 # Denmark, Netherlands, Norway, Spain, Sweden, Switzerland. ## Incl Estonia, Latvia, Lithuania (but not Poland) for 1925 and 1929. Source: Calculated from: League of Nations (1930), pp. 114ff; 1936 distribution from Board of Governors (1943) pp. 544ff. UK foreign-trade prices from Feinstein (1972), p. T139.

These arguments lead to the conclusion that the ‘deflationary bias’ of the gold standard resulted more from its gold commitments than from its fixed-exchange-rate commitments and could have been remedied by a joint decision of all central banks to raise the nominal gold price, without disturbing the parities of exchange rates against each other. But this perhaps brings us to the heart of the ideological problem of the gold standard and the conclusion that the basic problem was the incomplete politicization of the provision of international reserves. Since the late nineteenth century, the dominance of central bank demand for gold made it the world’s first politically managed currency. Within certain large limits set by private demand schedules for gold, the price of gold was largely determined by the unified reserve price maintained on a free global gold market by central banks. By the interwar years central-bank dominance was increased by the virtual cessation of gold-coin circulation outside the USA (and its considerable shrinkage in the USA). Thus the shortage of gold reserves in the 1920s, relative to demand for them, was the artificial product of this reserve price. But for the central bankers to acknowledge this, would have been to deny the ‘objectivity’ of the gold standard as a constraint on their own actions, and to acknowledge the need for a global political mechanism for determining the (artificial) price of gold. The uncoordinated floating of the 1930s, capped by the US devaluation, served the purpose of engineering a world raise of the gold price without confronting the above issue (Mundell, 2000).

18 Introduction

In contrast to India, Whitehall seems to have been indifferent to the New Zealand devaluation of 1930 (see the lively Chapter 7 by Singleton). Liberation from the gold standard facilitated a further, deliberately engineered devaluation in 1933, largely to redistribute income to farmers from their creditors and urban suppliers. Partly, perhaps, because of this redistributional aim, devaluation was also accompanied by the aim of monetary restraint (though in the event there was a jump in the money supply). Another restraint on credit expansion was the arrangement Singleton describes between the government and the banks whereby the government traded 5 per cent treasury bills for the extra sterling balances produced by the import reduction. This was because the banks feared capital losses from the future revaluation they said they expected. This profitable business reduced their incentive to expand domestic credit and hence also limited the New Zealand depreciation. After 1936, however the Labour government initiated a policy of monetary expansion – maybe on behalf of an urban constituency – leading to a foreign-exchange crisis in 1938.30 Singleton’s discussion demonstrates that local monetary expansion that might have added its mite to global recovery did not satisfy local interest politics. It also reminds us that all monetary policy has repercussions on the banking sector, especially in small open economies. These repercussions are seldom modelled in monetary theory.31 Balachandran and Singleton independently argue that Britain forced monetary cooperation on empire and dominions – in relation to the rupee exchange rate and in relation to the degree of monetary expansion in the sterling bloc as a whole. But this cooperation was dysfunctional, because governed by the ‘golden fetters’ that shackled the minds of London policymakers. They imply thus that even a highly cooperative gold-standard system would have damaged the global economy between the wars. Eichengreen and Temin’s ‘Afterword’ (Chapter 9) implies the same, by its sequence of counterfactuals in which the mere avoidance of financial crisis in Germany yields only minor global advantages. The opposite extreme to international economic and monetary cooperation was exhibited by the USSR, whose policies contrasted markedly with those of pre-war Russia. Gregory and Sailors’ instructive chapter (Chapter 8) shows how before 1914 the gold standard tied Russian output and price fluctuations quite closely to those of the world economy. However by 1922 the USSR had almost completely dissipated the still-sizeable share of world monetary gold held in Russia on the eve of the revolution (Nurkse, 1944, p. 233). This is perhaps one measure of the scale of the domestic crisis of civil war etc.,32 but the USSR did not thereafter seek to replenish these stocks. By investing in the gold standard in the 1890s, Russia had gained substantial access to foreign capital at lower interest rates (Gregory, 1979), but the USSR eschewed the dependence on foreign capitalism that such access implied. Nor, Gregory and Sailors explain, was it interested in the

Theo Balderston 19

output mix which the USSR’s comparative advantage would have assigned it under the gold standard. Instead it opted for what they designate as a policy of ‘mercantilism’ which was more radical than, but not wholly divorced from, the policies followed elsewhere in the world in the 1930s. Under the Soviet variant output was maintained in the 1930s by buoyant investment rather than resilient consumption, financed by taxation of the urban sector, since collectivization in the end prevented, rather than enabled, a transfer of savings from agriculture to industry. However, aggregate growth performance was less impressive than thought at the time: over the period 1928–37 Soviet GDP growth was not out of line with that of the UK – something of a vindication for Eichengreen and Sachs (1985). The Soviet contrast also reminds us of the influence of ideologies on global economic systems. The collapse of the gold standard was connected to a collapse of an ideological economic internationalism associated also with free(-ish) trade. It was not only the democratization of electorates that undermined the prioritization of the gold parity in national economic policies. This had to be combined with the weakening belief of that electorate in the long-run benefits of economic internationalism. Increasing pre-1914 critiques of monopoly tendencies in industry were reinforced by the apparent efficiency of wartime ‘economic planning’, by the incompatibility of open economy policies with nationalistic objectives particularly for countries whose comparative advantage did not lie in heavy industry, by the catastrophe of the slump, especially in the USA which sullied the image of capitalist prosperity, and by the image of success radiated by the Soviet Union; all these combained to erode that belief. Contrariwise, the perception of the failure of socialism helped to form a political consensus in favour of ‘globalization’ in the later twentieth century. Interests influence the articulation and evolution of ideologies as Balachandran shows. But interests are defined in relation to expected outcomes of policies, and these expectations are shaped by economic theories, which are the stuff of economic ideologies. The expectations of economic agents then condition the success of policies. Perhaps the history of the Great Depression needs to be written not just in terms of the dominance of one ideology but of a greater diversity of partly overlapping, partly contradicting, economic ideologies than exists today – including ‘the gold standard’, ‘the real bills doctrine’, ‘liquidationism’, ‘the Treasury view’, ‘purchasing power prosperity’33 – and ‘socialism’. This is an agenda for further research into the causes of the Great Depression. Notes 1 For good overviews of the ‘Great Depression’ see Clavin (2000) or Eichengreen (1992a). 2 For a succinct and accessible statement of the ET view of the dynamics of the gold standard, see Eichengreen (1996b), pp. 7–44, and Eichengreen and Temin (1997).

20 Introduction 3 On these borrowings see Feinstein and Watson (1995), p. 110. The puzzle why the US equity boom did not induce US bargain-hunting in foreign equity markets is seldom addressed. An argument relating the German recession from 1927 to an external constraint produced by perception of German overborrowing is advanced by Ritschl (1999). 4 Actually Europe excluding France lost only 1 1/2 per cent of its gold stock in 1929, but the rest of the world excluding the USA lost nearly 10 per cent: Board of Governors (1943), pp. 544ff. 5 Including Bolivia, New Zealand, Venezuela. Suspension of convertibility was not the same as serious depreciation against the dollar; though this followed in some important cases such as Argentina and Australia: League of Nations (1937), table I; cf. Board of Governors (1943), pp. 662ff. 6 Cf. Eichengreen, (1996b), p. 72. However the decline of US flotation of foreign bonds in 1930 (excl. the Young Loan), especially to Latin America and Asia/Oceania, was offset by higher foreign lending in France, Switzerland and the Netherlands – but not in the UK. The catastrophic fall in foreign lending came in 1931: cf. League of Nations (1931), pp. 315ff; League of Nations (1938), p. 143; Abbott (1937), pp. 196ff. 7 Excluding the USA and France, the rest of the world’s monetary gold fell from $4,965m in 1928 to $4,370m in 1930 (and to a minimum of $4,244m at the end of 1931: Board of Governors (1943), pp. 544ff). 58 per cent of world central-bank gold reserves was held by the USA and France at the end of 1931 (60 per cent a year later); whereas 53 per cent of total monetary gold had been held by the USA, France and Russia at the end of 1913 (Russia held little in 1930–31): League of Nations (1930), pp. 114f. However at the end of 1930 the US and French share of the total official reserves of gold and foreign exchange of 38 countries, for which data exist, was 46 per cent. See League of Nations (1937), tables III & IV. 8 For a different view: Foreman-Peck, Hallett and Ma (2000), pp. 542f. 9 Eichengreen emphasizes the gold-standard constraint where Friedman and Schwartz (1963, pp. 237ff) blame the inflationary policy pursued by the Federal Reserve System in 1919 so as to facilitate the funding of government debt. Temin (1989, pp. 59ff) has a different explanation for the shortness of the postwar slump. 10 Cf. Friedman and Schwartz (1963), pp. 384–9. Epstein and Ferguson (1984) emphasize rather the effect of the open-market purchases on asset yields, hence on bank earnings and bank stability. Sumner (1997, pp. 67ff) emphasizes the effect of congressional budgetary battles on foreign confidence in the dollar. Harrison emphasized practical limits to the rate at which securities could practicably be bought on the New York market (Wicker, 1966, p. 174). Wheelock (1991, esp. pp. 45–67) maintains that a misreading of the significance of member bank borrowings from the Fed made its monetary policy pro- rather than counter-cyclical. These conceptions flowed from their understanding of their role as one of supporting credit intermediation, rather than of controlling the money supply (cf Wicker, 1966, pp. 58f). 11 Wigmore (1987). However Wicker’s (1966, pp. 172ff) explanation for the inaction of the Fed in the run-up to the 1933 banking crisis centres on their conception of the operation of open-market policy, which they considered efficacious only when the commercial banks had an incentive to increase lending on the basis of the excess reserves thereby acquired. Wicker perhaps also implies that the Fed failed to spot the incipient banking crisis (even although the prevention of such had been one of the main motives for its establishment in 1913) because the crisis started in outlying rural districts and not in financial centres.

Theo Balderston 21 12 The five creditor countries are Belgium, France, Netherlands, Switzerland, UK. The thirty-four debtor countries are (i) Austria, Czechoslovakia, Denmark, Finland, Lithuania, Norway, Sweden (which countries however gained reserves in 1930); (ii) Bulgaria, Chile, Colombia. Ecuador, Estonia, Egypt, Germany, Greece, Hungary Italy, Japan, Latvia, Peru, Poland, Portugal, Rumania, Yugoslavia and Union of South Africa (which lost reserves but maintained their parities); (iii) Australia, Bolivia, Spain and Uruguay (which lost reserves and depreciated); (iv) Argentina, Brazil, Canada, Mexico and New Zealand (for which full data on reserves is not to hand). ‘Monetary base’ is notes and coin issued plus central-bank liabilities. The official foreign reserves of the five creditor countries increased by $636m (or 15 per cent); the reserves of the twenty-nine ‘debtors’ for which there is data reduced by $502m. Official reserves of the USA increased by only 3 per cent. The aggregations for the above calculations proceeded by valuing the relevant national-currency data in League of Nations (1938, tables III and IV) by the December 1929 dollar value of the national currency ascertained from Board of Governors (1943, pp. 662ff), or from comparison of the dollar and national-currency values of 1929 gold reserves in League of Nations (1938, tables IV and VI). If average percentages were substituted for the above aggregates the results would be similar. 13 The approximation to M2 is all notes, coin, central bank liabilities plus all commercial bank deposits less commercial bank cash reserves. No data were available for Egypt; Belgian M2 for 1930 interpolated. For aggregation method and sources, see previous note. It is at an approximation to M1 (as ‘M2’, but including only ‘sight’ as opposed to all commercial bank deposits) that the growth in creditor countries ($675m) is much less than the contraction ($923m) in ‘debtor’ countries. But inspection of the tables in the source shows unexplained abrupt shifts from ‘sight’ to ‘other’ deposits – notably in Switzerland, so this seems an unstable level of definition. 14 Many support the view that the 1931 financial crisis turned a serious recession into a catastrophe: e.g. Foreman-Peck, Hallett and Ma (2000), pp. 539, 542. 15 Cassel (1932) argued that Reparations and War Debts caused the shift of gold to France and the USA, and given the gold standard, this caused the global slump. Ritschl’s argument is that ‘transfer protection’, introduced into the Dawes Plan under US influence to lessen Reparations pressure on the German balance of payments and German demand, temporarily stimulated private lending to Germany by according foreign commercial debts ‘seniority’ over Reparations claims. But the reduction of ‘transfer protection’ in the Young Plan weakened this ‘seniority’ and raised the likelihood of a German default, provoking the capital flights. 16 Ritschl’s analysis competes with other theories of the German banking crisis. For a short discussion see Balderston (2002), ch.5. 17 It is a commonplace that the ‘Young Plan’ schedule of Reparations payments closely mirrored the agreed schedule of French war debt repayments to the USA. The USA was not a recipient of Reparations. 18 Suppose that 30 per cent of the liabilities of a German great bank in 1931 were dollar-denominated. Suppose the reichsmark depreciated by 20 per cent. Then the reichsmark value of its liabilities would rise by 6 per cent. This was about the proportion of the great banks’ own capital to their total liabilities. It is true that the loans which were the counterpart of the foreign-currency liabilities were also foreign-currency denominated. But reichsmark depreciation would reduce the power of the loan debtors to service them.

22 Introduction 19 See League of Nations (1930); Royal Institute for International Affairs (1931). Cf. Johnson (1997). 20 Sayers (1976), vol.III (appendixes), pp. 353ff; Howson (1980), pp. 58f; with Board of Governors (1943), pp. 544ff. Quantum meaning gold valued at $20.67 throughout. At the end of 1937 the US and Britain held 69 per cent of ascertainable global monetary gold (though more than before 1931 may have been held in undisclosed exchange equalization accounts of other countries), at a time when foreign assets were a far smaller share of official reserves than in 1930–1. Cf. Nurkse (1944), p. 90. 21 For a survey of the international implementation of exchange controls, see League of Nations (1938), p. 107. 22 League of Nations (1939), table II, with UK data substituted from Sayers (1976), vol. III (appendixes) pp. 349ff, and Howson (1980). Perhaps the institution of exchange stabilization accounts in the 1930s with undeclared assets operating in foreign-exchange markets (cf. Nurkse, 1944, pp. 143ff) falsifies these proportions; however the evidence for the UK account suggests that they too held meagre amounts of foreign exchange relative to gold. If Belgium, France, Italy, Japan, the Netherlands, Switzerland and the UK are excluded from the sample, the 1937 percentage rises to 34 per cent, as against 37 per cent in 1929. But this is largely because of the inclusion of Germany and Spain, whose gold stocks dropped dramatically. If these countries, too, are excluded the 1937 percentage of the rest is 34 per cent as against 50 per cent in 1929. (See, too, note 32.) All these percentages value both foreign assets and gold in domestic currency, then revalue them in $ at the current exchange rate. They therefore value the gold as the central banks themselves valued it. 23 See previous note. Another possible explanation could be that the foreignreserve proportions fell because of the diminished supply of sterling and dollar assets through the UK and US balances of payments. But in this case one would expect the weaker economies to have smaller foreign-reserve proportions. 24 League of Nations (1938), tables I & IV. 25 On the ‘high pressure’ system, see Eichengreen (1985). 26 Eichengreen (1992a, p. 198ff) and ‘The Gold-Exchange Standard and the Great Depression’ (in Eichengreen, 1990, pp. 238–70) argues that there was no shortage in the supply of gold because wartime centralization of nearly $3 bn of gold coin offset the effects of wartime and early postwar price rises and reductions in gold output. But in 1996b, pp. 61, 68,74, he seems more sympathetic to the possibility that there was also a supply shortage. 27 Eichengreen (1992a), pp. 199f. The main conclusion of League of Nations (1930) (pp. 18–20) was that the prospective gold shortage could be eased by reducing the central-bank note cover ratios. See also Royal Institute for International Affairs (1931). But against the claim that central banks irrationally hoarded gold may be placed the observation that countries suspending convertibility usually did so with considerable gold reserves, and (except for the UK and Brazil) at levels well above their 1913 reserves, perhaps suggesting both that their gold stock was more important to them than their convertibility, and perhaps that their sense of their control of it was weaker than in 1913. Argentina, for example, suspended convertibility with an end-of-year gold stock 70 per cent greater than in 1913 though 30 per cent less than its previous 1920s peak. For general statistics see League of Nations (1937),table II, p. 112f; Board of Governors (1943), pp. 544ff. In 1930 Hawtrey formulated the rule that a country’s desired gold stock was one large enough to allow of ‘the largest

Theo Balderston 23

28

29 30 31 32

33

currency contraction that public policy will allow the country to undergo’ (Royal Institute… 1931, p. 102). In his view no country was unconditionally on the gold standard, but it would not hold gold reserves larger than required to defend its conditional membership. His rule was mis-formulated: most countries suspended convertibility at the point when their gold stocks were still well above exhaustion. Their demands for gold were stronger, relative to their commitment to the gold standard, than his rule implied. Montagu Norman’s description of Bank of England policy between 1925 and 1931 as ‘under the harrow’ seems to encapsulate this recognition: Sayers (1976), vol. 1, pp. 211ff. For example, after January 1934 the US would not sell gold to countries with inconvertible currencies: Howson (1980), p. 36. Drummond (1981), pp. 104–15. The policy led to a proposal for exchange controls which provoked (effective) Whitehall hostility. Cf. also notes 10 and 11 on the relation between US open-market operations and bank earnings.. Similar pressures of domestic crisis would also explain the dissipation of the Spanish stocks after 1936, and also the Third Reich’s dissipation of the less-plentiful gold stocks bequeathed to it by the Weimar Republic – otherwise puzzling in a state bent on war (cf. Mason, 1975). See League of Nations (1937, 1938, 1939), table IV. This theory, adopted by German trade unions from a variety of sources, argued that high wages spelled national economic prosperity, usually on the grounds that both technical progress and monopolistic elements in product markets gave leeway for wage increases without price increases. See, inter alia, Nolan (1994).

References Abbott, C.C. (1937) The New York Bond Market 1920–1930 (Cambridge, Mass.: Harvard University Press. Reprinted New York, 1975). Balderston, T. (2002) Economics and Politics in the Weimar Republic (Cambridge: Cambridge University Press). Bayoumi, T. and B.Eichengreen (1994) ‘Economic Performance under Alternative Exchange-Rate Regimes: Some Historical Evidence’, in P.Kenen, F.Papadia and F.Saccamoni (eds), The International Monetary System (Cambridge: Cambridge University Press): 257–97. ——– (1996) ‘The Stability of the Gold Standard and the Evolution of the International Monetary Fund System’, in T.Bayoumi, B.Eichengreen and M.P.Taylor (eds), Modern Perspectives on the Gold Standard (Cambridge: Cambridge University Press):165–88. Bernanke, B.S. (1995) ‘The Macroeconomics of the Great Depression: A Comparative Approach’, Journal of Money, Credit and Banking, 27: 1–28. Board of Governors of the Federal Reserve System (1943) Banking and Monetary Statistics (Washington, D.C.). Bordo, M.D. and F.E.Kydland (1996) ‘The Gold Standard as a Commitment Mechanism’, in T.Bayoumi, B.Eichengreen and M.P.Taylor (eds), Modern Perspectives on the Gold Standard (Cambridge: Cambridge University Press): 55–100. Cassel, G. (1932) The Crisis in the World’s Monetary System (Oxford: Oxford University Press). Clavin, P. (2000) The Great Depression in Europe, 1929–1939 (Basingstoke: Macmillan).

24 Introduction Drummond, I.M. (1981) The Floating Pound and the Sterling Area (Cambridge: Cambridge University Press). Eichengreen, B.J. (1985) ‘International Policy Coordination in Historical Perspective: The View from the Interwar Years’, in W. Buiter and R. Marston (eds), International Economic Policy Coordination (Cambridge: Cambridge University Press):.139–76; here cited as reprinted in Eichengreen (1990), pp. 113–52. ——– (1986) ‘The Bank of France and the Sterilisation of Gold’, Explorations in Economic History, 23: 56–84, cited here as reprinted in Eichengreen (1990), pp. 83–112. ——– (1987) ‘Conducting the International Orchestra: Bank of England Leadership under the Classical Gold Standard, 1880–1913’, Journal of International Money and Finance, 6: 5–29. ——– (1990) Elusive Stability. Essays in the History of International Finance 1919–1939 (Cambridge: Cambridge University Press). ——– (1991a) ‘The Comparative Performance of Fixed and Flexible Exchange-Rate Regimes: Interwar Evidence’, in N. Thygesen, K.Velupillai and S.Zambelli (eds), Business Cycles: Theories, Evidence and Analysis. Proceedings of a Conference held by the International Economic Association, Copenhagen, Denmark (London: Macmillan): 229–75. ——– (1991b) ‘Relaxing the External Constraint: Europe in the 1930s’, in G. Alogoskoufis, L.Papademos and R.Portes (eds), External Constraints on Macroeconomic Policy: The European Experience (Cambridge: Cambridge University Press): 75–117. ——– (1992a) Golden Fetters. The Gold Standard and the Great Depression 1919–1939 (New York: Oxford University Press). ——– (1992b) ‘The Origins and Nature of the Great Slump Revisited’, Economic History Review, 45: 213–39. ——– (1996a) ‘Déjà vu All Over Again: Lessons from the Gold Standard for European Monetary Unification’, in T.Bayoumi, B.Eichengreen and M.P.Taylor (eds), Modern Perspectives on the Gold Standard (Cambridge: Cambridge University Press): 365–87. ——– (1996b) Globalizing Capital. A History of the International Monetary System (Princeton: Princeton University Press). ——– and J. Sachs (1985) ‘Exchange Rates and Economic Recovery in the 1930s’, Journal of Economic History, 45: 925–46. ——– and P. Temin (1997) ‘The Gold Standard and the Great Depression’, NBER Working Paper series, no. 6060, June. Epstein, G. and T.Ferguson (1984) ‘Monetary Policy, Loan Liquidation and Industrial Conflict: The Federal Reserve Open-Market Operation of 1932, Journal of Economic History, 44: 957–83. Feinstein, C.H (1972) National Income, Expenditure and Output of the United Kingdom 1855–1965 (Cambridge: Cambridge University Press). ——– and K.Watson (1995) ‘Private International Capital Flows in Europe in the Inter-War Period’, in C.H.Feinstein (ed.), Banking, Currency and Finance in Europe between the Wars (Oxford: Clarendon): 94–125. Ferguson, T. and P.Temin (2001) ‘“Made in Germany”.The German Currency Crisis of July 1931’, Working Paper, February, JEL No. N14 E32. Foreman-Peck, J., A.H. Hallett and Y.Ma (2000) ‘A Monthly Econometric Model of the Transmission of the Great Depression between the Principal Industrial Economies’, Economic Modelling,17: 515–44. Friedman, M. and A.J.Schwartz (1963) A Monetary History of the United States 1867–1960 (Princeton: Princeton University Press)

Theo Balderston 25 Gregory, P. (1979) ‘The Russian Balance of Payments, the Gold Standard and Monetary Policy: A Historical Example of Foreign Capital Movements’, Journal of Economic History, 39: 379–99. Holtfrerich, Carl-L. (1991) ‘Germany and the International Economy: The Role of the German Inflation in Overcoming the 1920/1 United States and World Depression’, in W.R. Lee (ed.), German Industry and German Industrialization. Essays in German Economic and Business History in the Nineteenth and Twentieth Centuries (London and New York: Routledge). Howson, S. (1980) Sterling’s Managed Float: The Operations of the Exchange Equalisation Account, 1932–39. Princeton Studies in International Finance no. 46. (Princeton: Princeton University Press). Johnson, H.C. (1997) Gold, France and the Great Depression, 1919–32 (New Haven: Yale University Press). Kindleberger, C.P. (1973) The World in Depression 1929–1939 (London: Allen Lane). Krüger, P. (1985) Die Aussenpolitik der Republik von Weimar (Darmstadt: Wissenschaftliche Buchgesellschaft). League of Nations (1930) Interim Report of the Gold Delegation of the Financial Committee (Geneva). ——– (1931) The Course and Phases of the World Economic Depression (by B.Ohlin) (Geneva). ——– (1937, 1938, 1939): League of Nations Economic Intelligence Service, Monetary Review 1936/7; id. 1937/8; id. 1938/9. Being vol.I of Money and Banking 1936/7; id.1937/8; id. 1938/9 (Geneva). Leffler, M.P (1979) Elusive Quest: America’s Pursuit of European Stability and French Security 1919–1933 (Chapel Hill: University of North Carolina Press). Link, W. (1970) Die amerikanische Stabilisierungspolitik in Deutschland, 1921–32 (Düsseldorf: Droste). Mason, T. (1975) ‘Innere Krise und Angriffskrieg 1938–39’, in F.Forstmeier and H.-E.Volkmann (eds), Wirtschaft und Rûstung am Vorabend des Zweiten Weltkrieges (Dûsseldorf: Droste): 158–88. Mundell, R.A. (2000) ‘A Reconsideration of the Twentieth Century’ American Economic Review, 90(3): 327–40. Nolan, M. (1994) Visions of Modernity. American Business and the Modernization of Germany (New York: Oxford University Press). Nurkse, R.E. (with W.A. Brown, Jr.) (1944) International Currency Experience. Lessons of the Inter-War Period (League of Nations, Economic, Financial and Transit Department) (Geneva). Ritschl, A. (1999) ‘International Capital Movements and the Onset of the Great Depression: Some International Evidence’ (mimeo). Roberts, G.E. (1931) ‘Gold Movements into and out of the United States, 1914 to 1929, and the Effects’, in League of Nations, Selected Documents on the Distribution of Gold Submitted to the Gold Delegation of the Financial Committee (Geneva), pp. 38–63. Royal Institute for International Affairs (1931) The International Gold Problem. A Record of the Discussions of a Study Group of Members of the R.I.I.A 1929–1931 (Oxford: Oxford University Press). Sayers, R.S. (1976) The Bank of England 1891–1944 (Cambridge: Cambridge University Press, 3 vols). Sicsic, P. (1992) ‘Was the Franc Poincaré Deliberately Undervalued?’, Explorations in Economic History, 29: 69–92.

26 Introduction Simmons, B.A. (1994) Who Adjusts? Domestic Sources of Foreign Economic Policy During the Interwar Years (Princeton: Princeton University Press). Sumner, S. (1997) ‘News, Financial Markets and the Collapse of the Gold Standard 1931–32’, Research in Economic History, 17: 39–84. Temin, P. (1989) Lessons from the Great Depression. The Lionel Robbins Lectures for 1989 (Cambridge, Mass: MIT Press). ——– and B.A. Wigmore (1990) ‘The End of One Big Deflation’, Explorations in Economic History, 27: 483–502. Wheelock, D.C. (1991) The Strategy and Consistency of Federal Reserve Monetary Policy 1924–1933 (Cambridge: Cambridge University Press). Wicker, E.R. (1966) Federal Reserve Monetary Policy 1917–1933 (New York). Wigmore, B.A. (1987) ‘Was the Bank Holiday of 1933 Caused by a Run on the Dollar?’, Journal of Economic History, 47, pp. 739–55.

2 Understanding the Great Depression in the United States versus Canada* Pierre L. Siklos

1

Introduction

It is now recognized that the Great Depression should be viewed as a global phenomenon with its roots in the gold standard so arduously restored in Great Britain in 1925. Temin (1989) and Eichengreen (1992) provide arguably the best recent accounts of the Great Depression. Not surprisingly, it has generated a vast literature, and continues to this day to fascinate economists and social scientists more generally. Unlike Friedman and Schwartz (1963, 1982), whose masterful works documented the role of monetary policy in explaining both the depth of the US Depression and its propagation outside the United States, Temin (1989) stresses the influence of a large ‘economic’ shock which was then magnified by an almost slavish adherence to the gold standard. Eichengreen (1992) would not disagree but would emphasize the breakdown in international cooperation among government and central bank officials. Others, such as Kindleberger (1986), might argue instead that the collapse of the post First World War monetary order produced financial chaos which, like the now much discussed ‘Asian flu’ of the late 1990s, caught on across much of the world. The interwar era has fascinated many due to the tremendous severity of the economic collapse, as is documented throughout this volume. Lately, however, interested researchers have been drawn not only to how the Depression was transmitted worldwide but whether the combination of economic ideology and politics conspired to generate such a spectacular slump. Since so much of the debate surrounding the Great Depression centres around financial issues, their real economic consequences, and the global nature of the event, comparisons across countries can provide useful insights about the role played by institutional factors. It is precisely for this reason that a review of the US and Canadian experiences is helpful, not only because of the obvious close ties the two countries share with each other but also because there were significant differences in the financial 27

28 The USA and Canada

sphere which underscore the role played by the gold standard. Yet, despite the differences, there is one unmistakable similarity: both countries experienced a Great Depression at roughly the same time and of comparable economic magnitude. The rest of this chapter is organized as follows. The following section briefly sets the stage for a US–Canadian comparison by outlining what motivated the return to a gold standard following the First World War and why the policy, from a ‘modern’ policy perspective, represents a ‘straitjacket’, otherwise referred to as the ‘cross of gold’. Section 3 explores the origins and transmission of the powerful and large shock that Temin (1989) argues lay behind the Great Depression and how this played out in the US and Canada. Section 4 focuses on the financial aspects of the Great Depression. Arguably, this is a crucial ingredient of the global slump. Section 5 looks at the consequences of the Great Depression for economic reform and recovery. Section 6 offers a few concluding comments by asking how recent economic research sheds light on the overarching factors in the Depression. Needless to say, since the topics of interest are wide-ranging, the present chapter can only selectively survey the extant literature and, hopefully, whet the appetite of the interested student for a more serious study of this remarkable era in economic history.

2

The gold standard as a straitjacket for monetary policy

One cannot evaluate the magnitude of the Great Depression without comprehending the constraints imposed by the gold standard. For much of the nineteenth century commodity money regimes, such as the gold standard, shaped policy-makers’ thinking. Not surprisingly, then, some countries were keen to return to it following recovery from the First World War . But the decision to return to gold was not squarely based on economic considerations. As Winston Churchill who, as Chancellor of the Exchequer, led the United Kingdom back to gold, noted: ‘But this [the return to the gold standard] isn’t entirely an economic matter; it is a political decision…’ (Gilbert, 1977, p. 92). One of the chief attractions of the gold standard among its supporters was that it operated automatically, leaving no room for political interference in the realms of international trade and investments. Combined with a commitment to central bank autonomy, the ‘rules of the game’ ensured that currencies could be converted into gold at the stipulated fixed exchange rate. Indeed, in its purest form, the gold standard comes closest to the fixed exchange rate system discussed in standard macroeconomic textbooks. Students of modern macroeconomics are taught that the demand for money is a function of income and the opportunity cost of holding money, usually measured in terms of some nominal interest rate. Moreover, the

Pierre L. Siklos 29

demand for money is assumed to be proportional to the price level, giving rise to the usual quantity theory relation. But in the purest gold-standard system, i.e., one where the only money is gold coinage, the demand for gold is also a function of the price of gold, relative to those of other goods. This is because gold is a commodity with non-monetary uses too. We can thus write: M/PG = F(Y, P/ PG, q)

(2.1)

where M/PG is the money stock (M) deflated by the price of gold (PG), Y is income, P/PG is the price of consumer goods (P) relative to gold , and q is the ‘cost of holding money’.1 Notice that, unlike in the quantity theory, there is no reason why a change in P/PG should lead to a proportionate change in M/PG, holding Y and q constant. By fixing the price of gold in terms of money, the price of goods in terms of money is also fixed, leaving no room for active monetary policy, making it ‘knave-proof’.2 The actual process which would guarantee this state of affairs was the ‘specie-flow’ mechanism.3 Hence, a country which consumed in excess of production, producing a balance of trade deficit, would have to export gold to cover its debt. Moreover, the resulting fall in gold reserves would normally lead to a contraction in the domestic money supply and, eventually, in the price of goods and services. The reality of the gold standard was, of course, different for a number of reasons. First, countries such as the UK and the US, which had central banks, could manage the monetary system by varying the terms under which they could extend credit to the banking system. Hence, a rise in domestic interest rates could offset the loss of gold reserves from a balance-of-trade deficit by attracting short-term capital flows. The higher interest rates would also serve to depress domestic aggregate demand, thereby helping eliminate these pressures which produced the outflow of gold in the first place. Second, the actual gold standard was a fractional system since only a small portion of the money supply was held in the form of gold. Moreover, a central bank could resort to the practice of ‘sterilization’, that is, conduct open market operations. Hence, for example, a reduction in the money supply from gold outflows could conceivably be offset by buying government bonds held by banks or the public. Third, prices of goods and services do not adjust instantaneously to some imbalance between aggregate demand and supply. Fourth, the adjustment process is also dependent on how mobile capital is. Ostensibly, then, the gold standard could represent a straitjacket under the ‘rules of the game’, as sketched above. However, these rules were contingent on all of the above factors, including the ability of central banks and global financial markets to circumvent them as needed (e.g. see Bordo and Kydland, 1995).

30 The USA and Canada

What then of the US and Canadian experiences? First, it might surprise many that capital markets in industrial economies were highly integrated at the outset of the Great Depression making the current notion of ‘globalization’ seem somewhat dated (Bordo, Eichengreen and Kim, 1998). Dick and Floyd (1992) describe in great detail how Canada operated under the gold standard until the outbreak of the First World War and the degree to which capital markets in Canada and the US were integrated (also see Bordo, Redish and Shearer, 1999). In particular, much of banks’ gold reserves were held in New York where a substantial amount of financial business was carried out. Indeed, these same authors present convincing empirical evidence to suggest that capital market integration is what drove adjustment under the gold standard in Canada rather than adjustment in the price of goods and services.4 Table 2.1 shows the balance of trade, flows of short-term capital, and movements of gold for the period 1926–37. It is clear, for example, that the specie-flow mechanism worked differently in practice than in theory. Figure 2.1 plots the available data on short-term interest rates in Canada and the US. Whereas the traditional view of the specie-flow mechanism predicts that they should move in the opposite direction, to satisfy the condition of equilibrating gold flows from one country to the other, Figure 2.1 instead reveals a considerable amount of parallel movements in these rates before and after the gold standard period. Table 2.1

Trade, capital and monetary movements: Canada, 1927–37

Year

Current account balance

Capital account balance

Net official monetary movements

(Millions of dollars) Gold standard era 1927 1928 1929 1930 1931

–10 –32 –311 –337 –174

+3 +77 +62 –19 +54

+7 –49 –37 +36 –33

Post-gold standard era 1932 1933 1934 1935 1936 1937

–96 –2 +68 +125 +244 +180

+60 +33 +27 +32 –2 –17

–3 –6 +4 +6 +5 +6

Note: A positive sign implies a surplus in the current or capital account balances; a negative sign a deficit. A positive sign in Net Official Monetary Movements means a net inflow of gold; a net outflow occurs when the sign is negative. After 1935, the figure also includes foreign currency holdings held by the Bank of Canada. Source: Urquhart and Buckley (1983), series G83, G109, and G115.

31 6 Gold Standard 5

Per cent

4

US

3

Canada 2

1

0 30

31

33

32

36

35

34

Figure 2.1 Short-term interest rate movements in Canada and the United States, 1930–7 Note: For the US, the yield on short-term government securities. For Canada, the yield on short-term Dominion of Canada bonds. Data on short-term interest rates for Canada not available prior to 1930. The Gold Standard is dated as ending in April 1933. Sources: For the US: Cecchetti (1988); for Canada: Nixon (1937).

US wholesale prices, 1926 = 100

130 100 120 90 110 80

100

90

70 26

27

28

29

30

Canada wholesale prices, 1935 39 = 100

140

110

US Canada

31

Figure 2.2 Price-level movements in the US and Canada, 1925–31 Note: Monthly data on Wholesale Price Index Movements, April 1925–August 1931, inclusive. The Gold Standard was in effect throughout. Source: Brecher and Reisman (1957), table 1 (Canada), table 16 (US), appendix A. The series are not seasonally adjusted.

32 The USA and Canada

Finally, Figure 2.2 plots the behaviour of the wholesale price index of the two countries from April 1925, when the UK returned to gold until the gold standard collapsed in 1931. This figure shows commonality of price movements in the two economies, not the inverse relation predicted by the specie-flow mechanism. (It also disproves the much-touted claim that the gold standard protected against inflation and deflation.) Obstfeld and Taylor (1997) also point out that capital mobility was high not only between the US and Canada, but across all major economies at the time. The onset of the Great Depression led to a further contraction of the order of roughly 25 per cent of cross-border flows of capital even between the US and Canada coming on top of an almost 50 per cent drop in capital flows relative to the 1870–1913 period.5

3 Impulse and propagation of shocks leading to the Great Slump The impulse to a business cycle expansion or contraction refers to an event or a ‘shock’ of some kind which affects economic activity, output in particular, with a lag. This gives rise to changes in economic variables related to each other in time which is then called the propagation mechanism. Under this interpretation, the return of the gold standard sowed the seeds of a deflation, but policy-makers placed themselves in this straitjacket largely under the motivation of the belief that it was necessary to the ‘credibility’ of a regime that they deemed crucial for delivering again the economic prosperity of the pre-First World War era. In Temin’s (1989) view, ‘major deflationary shocks’ were the impulse for the Great Depression and its roots therefore lay in the ‘ideology of the gold standard’. For Eichengreen (1992) the ideology of the gold standard, coupled with a breakdown of international cooperation among key central banks, produced a slump more severe than any previously recorded. That the shock, whatever its origins, was large is clear from inspection of Figure 2.3 which plots industrial production in the US and Canada as well as the dates of reference cycles constructed on the basis of various economic indicators and pioneered by economists at the National Bureau of Economic Research (Burns and Mitchell, 1946). In both Canada and the US, industrial production roughly doubled between 1921 and 1928 or 1929, followed by a contraction of approximately the same magnitude between 1929 and 1933, that is, in about half the time it took industrial production to rise to prior to the collapse.6 The peaks and troughs in the reference cycles – the former dating the beginning of a recession; the latter the end of a recession – tend to mirror the downturns in industrial production. Nevertheless, we do see more frequent and short-lived cycles in economic activity before the Great Depression while longer durations are apparent between recessions and recoveries since the Great Slump.7

Pierre L. Siklos 33 US Business Cycles Industrial production, 1947 49 = 100

70

60

50 Ind. ption 40 30

20 20

22

25

26

28

30

32

34

36

Business Cycles in Canada Industrial production, 1935 39 = 100

140

120

100 Ind. ption 80

60

40 20

22

24

26

28

30

32

34

36

Figure 2.3 Business cycles in the US and Canada, 1919–36 Sources: Industrial production: Brecher and Reisman (1957), table 2 (Canada), table 11 (US), appendix A. Reference cycle dates: www.nber.org (US), and Hay (1967; table 1) for Canada. The shaded areas refer to the peak to trough periods (recession).

Finally, note that, with the exception of the US recession of 1927, Canadian and US recessions are nearly coincident.8 This would suggest a sequence of recessions, including the Great Depression, that originated in the US but were quickly transmitted to Canada. The combination of the gold standard and the close economic ties between the two countries would reinforce the mechanism through which the original shock (i.e., the

34 The USA and Canada

impulse) was transmitted internationally, as emphasized by both Temin (1989; see also 1993) and Eichengreen (1992). There is, however, disagreement on the point. Betts, Bordo and Redish (1996) have explained the parallelisms between the outset, depth and duration of the depression in both countries as a reaction to a common supply shock, namely a technical or resource shock originating in both economies simultaneously, which they did not manage to identify. Cole and Ohanian (1999) use a ‘neoclassical’ model (of the kind popularized by Lucas, 1987), which features a production function with constant returns to scale consisting of labour and capital as inputs. They argue that the recovery from the Depression was much milder than expected, and cannot be explained by the neoclassical approach. They ‘conjecture that government policies toward monopoly and the distribution of income are a good candidate’ for the weaker than expected recovery, a theme also emphasized by Prescott (1999) who also cautions that international evidence is necessary. The strong parallels between the US and Canada detailed in this chapter may weaken the case for the Cole–Ohanian view as we shall see. More importantly, the basis for their contention of a weak recovery from the Depression is an unusual calculation of output performance (Cole and Ohanian, 1999, n. 5) during the 1930s which is not justified by the authors and seems contradicted by Figure 2.3. These two explanations of the relationship between the US and Canadian depressions arise because of problems in the econometrics of distinguishing between them. Even a fairly straightforward model can admit several channels through which the variables of importance are affected. Hence, empirical testing requires that a priori judgments be made about what are, or are not, significant channels of influence, and scholars will differ in such judgments. A simple example (far simpler than the actual econometrics involved) may help make the point clearer. Suppose the economist believes that the correct model (called ‘structural’) linking, say, US and Canadian output can be represented by the following equations: US C C y Ct = b10 – b12 y US t + d11 y t–1 + d12 y t–1 + et

(2.2)

C C US US y US t = b20 – b21 y t + d21 y t–1 + d22 y t–1 + et

(2.3)

where yC is a proxy for output in Canada, and yUS is US output. The model hypothesizes that Canada’s output is determined by its own current and past history and the current and past history of US output; similarly for the US.9 The Greek symbols are ‘parameters’ indicating by how much according to (2.2), say, change in current US output affects current Canadian output. The terms e Ct and e US t respectively, are residuals, that is, output movements not explained by the other right-hand-side variables. Hence, if economic

Pierre L. Siklos 35

agents are assumed to operate with a model such as (2.2) and (2.3), then the residuals capture the influence of unexpected ‘shocks’ such as policy changes on output. Notice that the model presumes a contemporaneous link between US and Canadian output (i.e., y Ct is affected by y US t , and vice versa). These right-hand-side variables are not independent of the lefthand-side dependent variable, and this interdependence makes it difficult to estimate the parameters in such a way that they can be meaningfully interpreted in an economic sense. When economists attempt to empirically estimate a system of equations such as (2.2) and (2.3), they find it convenient to transform them algebraically (i.e. reparameterize) into a ‘reduced US form’ which eliminates the contemporaneous effect of y Ct on y US t and of y t C on y t , The ‘reduced form’ results in equations where the dependent variables, current Canadian (y Ct ) and US (y US t ) output, are expressed as functions of lagged variables only, i.e. only of variables independent of them. Since the details of such transformations are not essential,10 we simply state the result: y Ct = a10 + a11 y Ct–1 + a12 y US t–1 + e1t

(2.4)

US C y US t = a20 + a21 y t–1 + a22 y t–1 + e2t

(2.5)

The economist then would like to work back from the parameters econometrically estimated by applying (2.4) and (2.5) to the actual model, that is, to the ‘structural’ parameters of (2.2) and (2.3). However, this is not possible: only six coefficients are estimated (a10, a11, a12, a20, a21, a22) from (2.4) and (2.5), when eight parameters are of interest (b10, b12, b20, b21, d11, d12, d21, d22). The econometrician’s solution is to impose some economically sensible restrictions. For example, it is sensible to suppose that US output affects Canadian output, but not vice versa. This would mean b21 = 0. Such a priori restrictions enable the remaining structural parameters to be individually identified. But economists may differ about which restrictions to impose. And so in the end competing explanations, e.g. of whether the Canadian slump was imported from the US or the product of a ‘shock’ common to Canada and the US, cannot be decided between by econometric testing alone. Betts, Bordo and Redish (1996) find against the view that Canada ‘imported’ the Great Depression from the US (also see Fremling, 1985). The relative price changes under the gold standard, whose role is highlighted in (2.1), imply that the Great Depression was an event which occurred simultaneously on a global scale. However, it is important to note that the authors are unable to identify short-run supply shocks which may have been common to both countries. And Fackler and Parker (1994) and Burbridge and Harrison (1985) use similar techniques and find in favour of the impulse-propagation explanation of the Great Depression. Cecchetti

36 The USA and Canada

and Karras (1994) identify the sources of the Great Depression in the US, concluding, in line with Temin’s (1989) views, that a collapse in consumption, that is, an aggregate demand shock, was largely responsible for the drop in US output through late 1931. This drop was facilitated by monetary policy actions during this period, as emphasized by Friedman and Schwartz (1963) and Hamilton (1987). This was followed by an aggregate supply shock, via an unexpected decline in prices originating in the banking crises of 1930, 1932 and 1933, which contributed to further declines in output until 1933. It is important to note that these efforts do not formally allow a direct role for the impact arising from the high degree of financial integration between the two economies. As discussed in the next section, it is conceivable that while there was an idiosyncratic shock which led to deflation almost simultaneously in the US and Canada, the integration of financial systems, not explicitly accounted for in the foregoing research, may explain the global transmission of the Great Depression. Is there any single source which can be identified as setting in motion the Great Depression? While some popular histories (e.g., Garraty, 1986) point out the role of the stock market crash of 1929, more recent treatments of this event (e.g., Romer, 1990) suggest otherwise. First, the empirical connection between the stock market crash and real economy activity, namely, lowered expectations for economic activity and hence consumption, appears to be weak at best (e.g., Temin, 1976; Dominguez, Fair and Shapiro, 1988). Moreover, unlike the present-day situation, only a small portion of individuals’ wealth was invested in stocks so that this channel of influence on consumption is also likely not to have been operative. Yet, the combination of loss of confidence in financial institutions’ survival with the fact that banks underwrote share issues and could thereby incur potential losses on their own account – both stemming from the stock market collapse – must also surely be an ingredient in the Great Depression. Romer (1990) argues that uncertainty about the future is what contributed to the drop in output. Although uncertainty can manifest itself in various forms, it is usually captured in the form of some measure of volatility. On this basis it can be shown that real income ‘uncertainty’ appears to have risen before ‘uncertainty’ in stock price movements.11 But even if uncertainty of some form is the proximate cause for the global collapse of the 1930s, was there something in either the structure of the Canadian or US economies which can provide additional clues about the events which transpired during this period? Alternatively, perhaps, was the effect of the original shock amplified by its sheer unexpected size? Although both economies were dependent on the production and trading of commodities, Canada was perhaps relatively more vulnerable. For example, in 1919, the proportion of the labour force in non-agricultural pursuits stood at 73 per cent in the US and 67 per cent in Canada. By 1939, the size of the agricultural labour force shrank considerably in both

Pierre L. Siklos 37

countries (19.8 per cent in the US; 27 per cent in Canada) but the spread between the two actually grew. Hence, any underlying economic uncertainty might spread to uncertainty about future commodity prices. One candidate for the originating source of the Depression could be the pre-depression and depression-period movement of commodity prices. Figure 2.4 plots some available commodity price data for the US and Canada. The general downward trend in commodity prices until about Selected Commodity Prices: Canada, 1919 37 250

100

Oats, cents per bushel

80

200

60

150

40

100

Wheat, cents per bushel Wood, WPI (1935 39 = 100)

Gold Standard ends

Oats Wheat Wood

50

20 20

22

24

26

28

30

32

34

36

Selected US Commodity Prices: US, 1920 37 160

100 Gold Standard ends 80

120 100

60 80 60

40

Oats, cents per bushel

Corn, cents per bushel

140

Corn Oats

40 20

20 20

22

24

26

28

30

32

34

36

Figure 2.4 Selected commodity prices in the US and Canada, 1919–37 Sources: For Canada: Urquhart and Buckley (1983), series K38 (wood products and paper), M228 (wheat), and M233 (oats). For the US: Hamilton (1992). The original data were trimestrial and annual averages were calculated and plotted above.

38 The USA and Canada

1932 is evident in both countries. Despite some stability in commodity prices between roughly 1925 and 1929, commodities fell by over two-thirds of their value in 1919 or 1920. Indeed, commodity prices fell by over 50 per cent between 1919 and 1921 which alone constitutes a substantial shock. But modern macroeconomics would not treat even such large movements in prices as a ‘shock’ if they were anticipated. The reason is that events which are fully anticipated should not upset individuals’ plans even if, as a result, their incomes fell, and if the downturn was expected to be temporary. Many, other than economic historians, would be surprised to know that in the space of 40 years (from 1876 to 1922) the US experienced 12 years of deflation. During the same period Canada recorded a deflation on an annual basis 18 times.12 To what extent, then, were price changes during the 1920s, in particular, anticipated? Cecchetti (1992), Hamilton (1992), and Evans and Wachtel (1993) pose just such a question. Cecchetti (1992) points out that the historical behaviour of inflation in the US was such that the deflation of 1930–2 could have been anticipated. A straightforward explanation involves the notion that inflation and deflation are predictable variables because they are ‘persistent’. ‘Persistence’ means that inflation depends heavily, in the statistical sense, on its own past history. A simple representation of the idea is to write: pt = a + b pt–1 + et

(2.6)

where pt is the inflation rate, a is a constant, b is the degree of persistence in inflation since it measures how last period’s inflation rate affects current inflation. Finally, et measures the error made in specifying that only past inflation affects current inflation. If we estimate (2.6) for the US and Canada we obtain the estimates provided in Table 2.2. We immediately notice that the degree of persistence did not change dramatically before and after 1928 in the two countries.13 Thus, for example, for both the US and Canada a one per cent rise in last year’s inflation rate would raise this year’s inflation rate by roughly a half per cent, and vice versa. To be sure, other variables might be helpful in explaining the current behaviour of inflation but we can go a long way by simply assuming that last year’s inflation rate carries with it considerable momentum. Put differently, asking whether the behaviour of prices in the 1930–2 period was anticipated could be answered by using the estimates for the 1919–28 sample to forecast inflation for the years 1930–2. Figure 2.5 shows the results of this exercise. Clearly, (2.6) implies that the deflation was forecastable with essentially a one period lag. Of course, the term et in (2.6) implies that forecast errors are probable so that some portion at least of the deflation was a surprise. If so, can a more sophisticated model improve on forecast accuracy and, if economic agents are assumed to have, in effect, used such a

Pierre L. Siklos 39

model, would the deflation have been a surprise? Dominguez, Fair and Shapiro (1988) conclude in the affirmative. Burdekin and Siklos (1999) also rely on recent advances in econometrics to estimate whether there may have been a significant change in inflation persistence in several countries using over a century of data. They find significant breaks in US inflation in 1926 and 1933 while for Canada a break is found in 1930. Their findings suggest that if economic agents did not revise their model of inflation, it becomes even more likely that the deflation, to a significant degree, was unanticipated. Hamilton (1992) also presents evidence that, despite the drop in commodity prices in the 1920s discussed earlier, markets in the US did not anticipate a deflation of the magnitude which actually occurred. Quite the contrary. Based on futures-markets price data for key commodities, Hamilton’s analysis suggests that markets were betting at first on price increases. Futures markets are, of course, nowadays widespread in financial assets but their origins can be traced long ago to commodity markets. If these markets operate efficiently then a futures contract represents the market’s estimate of the price of a particular commodity expected to prevail in the future.14 In symbols: fj,t – Et Sj,t+1 = et

(2.7)

where fj,t is the futures price for commodity j at time t, Et Sj,t+1 is the market’s forecast, conditional on information at time t of commodity j’s price in the spot market next period (i.e., time t+1) and et is a ‘residual’ which is defined to have a zero mean and a constant variance. A version of the ‘market efficiency’ hypothesis requires that et fluctuate randomly. This hypothesis did not apparently apply to the markets examined by Hamilton, at least at first.15 Comparable data for Canada are not available but one can speculate that the US results would hold there as well. Evans and Wachtel (1993) point to uncertainty about the stance of monetary and fiscal policies to make the point that the deflation of the early 1930s, which was then Table 2.2

Modelling the behaviour of inflation Dependent variable: annual inflation rate

Independent variables

Constant Lagged inflation rate

US

Canada

1919–28

1919–37

1919–28

1919–37

–3.05(5.26) .64(.07)

–2.33(2.92) .54(.06)

–.17(2.22) .54(.08)

–.96(1.20) .40(.06)

Note: Inflation is the annual inflation rate calculated as (log Pt – log Pt–12) ¥ 100. The Wholesale Price Index was used for the US; the Consumer Price Index for Canada. Lagged inflation is the previous year’s inflation rate. Standard errors are in parentheses.

40 The USA and Canada US

Annual inflation (%)

0 5 Actual inflation Inflation forecasts

10 15 20 30:01

30:07

31:01

31:07

32:01

32:07

Annual Inflation rate (%)

Canada 4 0 4 Actual inflation Inflation forecasts

8 12 16 30:01

30:07

31:01

31:07

32:01

32:07

Figure 2.5 Actual and forecasted inflation: US and Canada, 1930–2 Note: Inflation forecasts based on the application of estimates of (2.6) for the 1919–28 sample for both countries. Estimates are given in Table 2.2. Source: Table 2.2.

followed by inflation, were both unexpected. Hence, if we view events after both economies reached the bottom, around 1933, as signalling a regime change, then the process driving inflation would have changed. While proponents of the rational expectations hypothesis would argue that models such as (2.6) or (2.7) are at fault, because they presume that both the underlying relationship and the coefficients remain constant, Evans and Wachtel (1993) would argue that unexpected forecast errors remain possible so long as economic agents are unsure, in a probabilistic sense, about what drives future inflation. In particular, if individuals expect the deflation to end, and inflation to return, then it is conceivable that the severity of the deflation would be underestimated. While the argument seems persuasive, it ignores that deflation was a ‘normal’ state of affairs since at least 1870. Therefore, there is little reason to think that the deflation of the early 1930s and the subsequent inflation need have been largely unanticipated.16 Nevertheless, the size of the forecast errors in this period may have been large enough to translate into a ‘large shock’ (cf. Figure 2.A1, p. 50).

Pierre L. Siklos 41

Ratio of measured to permanent per capita Income

Friedman’s well-known theory of permanent income17 provides additional perspective on the issue. It predicts that temporary changes in actual income should have little influence on ‘permanent income’. Hence, a shock to aggregate consumption would be signalled by a large and protracted fall in actual income relative to ‘permanent income’. Figure 2.6 shows an estimate of the ratio of per capita actual income to ‘permanent’ income. There are some interesting features in the data. First, actual incomes are persistently below their long-run or permanent levels throughout the 1919–27 era in both countries. Second, the departures from ‘permanent income’ are more prominent for the US than for Canada. Third, measured incomes begin to rise sharply relative to ‘permanent income’ in both countries, beginning in 1933. Given the sharp drop in consumption, especially in the 1928–32 period, it is conceivable that the drop in measured to permanent income, which occurred simultaneously, is symptomatic of an unexpected shock. Since the ratio plotted in Figure 2.6 also shows a sizeable drop during the earlier deflation of the 1920s this still poses the question why a second deflation, temporally close to the first one, was such a surprise. Temin (1989), following Keynes (1931), places relatively more emphasis on the role of investment as the impulse for the chain of events which led to the Great Depression. An obvious candidate indicator of the real cost of

0.95 0.90 0.85 Canada US

0.80 0.75 0.70 Gold Standard period 0.65 20

22

24

26

28

30

32

34

36

Figure 2.6 The evolution of estimates of measured to permanent income: US and Canada, 1920–36 Source: Bordo and Jonung (1987); appendix 1B explains how their measure of permanent income is constructed.

42 The USA and Canada

borrowing money is the real interest rate. Although the empirical connection between interest rates and investment has been found to be multifaceted and complex (e.g., see Romer, 1996, chapter 8), the estimates presented in Figure 2.7, however imprecise, suggest that real interest rates were exceedingly high by historical standards. As shown in Figure 2.7 they averaged 5.37 per cent in the US during the 1922–30 period18 and 10.95 per cent in Canada in the 1929–32 period. An additional indicator of future economic activity, much touted in recent financial economics literature, is the spread between long and short-term interest rates, also plotted in Figure 2.7.19 The potential link between the yield spread and future output operates along two dimensions. If short-term interest rates reflect the stance of monetary policy, then, for given long-term interest rates, a reduction in the spread would, in time, signal a contraction in economic activity as sectors of the economy sensitive to interest rates reduce spending. Alternatively, if firms expect future investment opportunities to improve, they will usually borrow in long-term market for funds. The resulting increase in the supply of bonds reduces bond prices thereby raising current long-term yields. As such, a rise in the spread would be associated with a future upturn in economic activity. As is clear from Figure 2.7, the spread in the US fell sharply in the years leading up to the Great Depression (until about mid-1929). In Canada, the only time the spread might give an indication of a fall in future economic activity is in 1931. Otherwise, the spread presages increases in future output. Either the message in the spread is misleading, or the differences between the US and Canadian experiences have their origins in the structure and response of the respective countries’ financial sectors. We explore this question in the following section. If the impulse for the Great Slump was the deflation in commodity prices (and the price level more generally) the role played by ‘ideology of the gold standard’, as Temin (1989) would perhaps emphasize, is not as clear-cut as might appear to be the case at first glance. After all, Canada’s commitment to the gold standard in the interwar years was not as great as in the US. Moreover, the Act creating the Federal Reserve System prohibited it from inflating the economy by restricting central bank credit to be no larger than 2 1/2 times the gold certificates held by the reserve banks. There was apparently less concern about the risks and consequences of deflation. With the price of gold fixed at $20.67/oz, the Fed felt mandated to use tight monetary policy to offset the potential inflationary effect from the resulting inflow of gold. In the meantime Fed officials looked at the stock market’s performance during the mid-1920s, together with the expansion of credit, and concluded that monetary policy was becoming too easy. 20 At this point the commitment to the gold standard did play a crucial role, especially given the attitudes of the French, who were fiercely committed to the monetary regime. The resulting tightening of monetary policy, evident from Figure 2.7, suggests that a form of international cooperation did work. Canada’s close financial

Pierre L. Siklos 43 Selected Financial Indicators for the US, 1922 37 40

20

2 0 1 20 0 40

1 Great Depression

2 22

24

26

28

30

32

Spread Real Int. rate

Short-term int. rate less inflation

3

Short-term int. rate less inflation

Long-term less short-term yield

4

Spread Real Int. rate

60 34

36

Long-term less short-term yield

Selected Financial Indicators for Canada, 1929–37 2.0

25

1.5

20 15

1.0

10 0.5 5 0.0

0

0.5

5

Great Depression

1.0

10 29

30

31

32

33

34

35

36

37

Figure 2.7 Selected financial indicators in the US and Canada, 1922–37 Note: The spread is the yield on long-term government bonds less the yield on shortterm government bonds. The real interest rate is the yield on short-term government bonds less inflation in the Consumer Price Index (Canada), Wholesale Price Index (US). Inflation is the log difference of the chosen price level evaluated over a three month horizon. The real interest rate calculation assumes investors predicted exactly the inflation rate over the holding period (assumed here to be three months). The period of the Great Depression is dated 1929 October–1933 December. Sources: CPI for Canada, series P10000 from Statistics Canada. WPI for the US from Brecher and Reisman (1957), table 16, appendix A. Yield on short-term government securities from Board of Governors of the Federal Reserve System (1976), 1919–28, and Cecchetti (1988), table A1 (US); Nixon (1937), table VI for Canada.

44 The USA and Canada

and economic links meant that it would inherit the US experience. Nevertheless, if the impulse to the Depression was a combination of real factors built upon faulty institutional foundations, modern macroeconomic analysis also makes it clear that the propagation mechanism was financial in nature.21 Not to be forgotten in all this is the role of the exchange rate. As Eichengreen (1992, p. 240) points out, the Canadian dollar ceased to be convertible after 1929. Floating exchange rates should, at least in theory, have insulated the Canadian economy from a ‘shock’ emanating from the US. Yet the $C/$US exchange rate was, in appearance, relatively stable between 1929 and 1931 (cf. Figure 2.A3, p. 51). Bordo and Redish (1987) argue that the Canadian government was intent on ensuring stability for at least two reasons. First, because of a long historical adherence to gold standard principles; second, because of large $US denominated debt. This state of affairs is somewhat puzzling since some research has demonstrated that, during the Great Depression, countries with flexible exchange rates suffered milder contractions than countries with either fixed exchange rates or ones who adhered to the gold standard (Choudhri and Kochin, 1980). While it is conceivable that credibility and political imperatives dictated exchange rate stability it is less clear why a stable float was deemed preferable over the maintenance of the gold standard orthodoxy. In the absence of credible commitment to exchange rate stability the behaviour of the exchange rate could reflect a ‘peso problem’. This phrase, apparently coined by Milton Friedman, refers to the performance of the Mexican peso during the 1970s which failed to be devalued despite widespread expectations that it would.22 After all, while average exchange rates were stable during the period in question, they were also considerably more volatile than in other years (figure not shown).

4 The Great Depression as a financial crisis: Canada versus the US As noted in the preceding section, deflation meant that real interest rates rose even in the face of relatively low nominal interest rates.23 But lower prices also raise the real cost to borrowers of repaying outstanding principal. Prospective borrowers, if they feel that prices are expected to fall further will also find credit to be expensive. In addition, default rates will be exacerbated if the deflation is unanticipated, leading potential lenders to be scared away from making loans, thereby producing a severe ‘credit crunch’, to use the modern term. The foregoing ingredients are harbingers of a financial crisis and place the banking system at the centre of the story of the Depression. While Kindleberger (1986) was one who emphasized the role of financial crises in the Great Depression, Bernanke (1983) is credited with working out the analytical details of how the Great Depression led to

Pierre L. Siklos 45

2600

45000

2400

40000

2200

35000

Millions of $US

Millions of $CDN

a massive failure of the credit allocation process (also see Bernanke, 1995). Not to be forgotten is monetary policy. After all, few economists would nowadays disagree with Friedman and Schwartz’s (1963) view that the sharp contraction in the money supply was the main causal factor in the chain of events that led to the ‘great contraction’. In fact, Hamilton (1987) makes the persuasive case that the monetary contraction led to an unanticipated deflation, and hence the financial crisis which ensured, together with the attendant high real interest rates, the collapse of intermediation services. Figure 2.8 shows just how massive the monetary contraction was in both countries. If monetary policy and the banking system are important ingredients in the story, then do differences in institutional structure matter? There were two notable differences between the US and Canadian financial systems. First, until 1935, Canada had no central bank. Second, US banking laws aimed at preventing banks from forming oligopolies. The Canadian banking system permitted branch banking while the US system placed severe restrictions on this type of activity.24 As a result, Bordo, Rockoff and Redish (1994) report the relatively greater stability of the Canadian over the US banking system, in the sense of a reputation for soundness (also see

Canada US

30000

2000 Great Depression

25000

1800 20

22

24

26

28

30

32

34

36

Figure 2.8 The money supply in Canada and the US: M2, 1919–37 Note: For the US, M2 is the sum of currency held by the public and total deposits (demand and time). For Canada, M2 is currency and total deposits. The Great Depression is dated 1929–33. Sources: Bordo and Jonung (1987), Courchene (1969), and Metcalf, Redish and Shearer (1998).

46 The USA and Canada

Grossman, 1994, in this connection). This translates into a lower probability of failure. Stability is then synonymous with the absence of bank failures. The importance of this feature lies in the concept of asymmetric information and the particular role it plays in the financial system. If insiders in the banking system are thought to possess a relatively better understanding of the state of the financial system than the outsiders (e.g., depositors), a shock to a particular bank or segment of the financial system may lead outsiders to paint all participants with the same brush, thereby prompting a run on the banks. This would lead to a further contraction of the money supply, through the well-known multiplier process,25 leading to a further diminution of the intermediation process. The consequences of asymmetric information for the severity of the Great Depression were noted by Mishkin (1978). While the number of bank offices or branches fell off substantially in both countries, and bank runs (whose social costs continue to be debated: e.g., Calomiris and Mason, 1997), figured prominently, the Canadian experience was marked by a contraction in the size and reach of existing banks (from eighteen in 1920 to ten in 1929), as only one bank, the Home Bank, failed in 1923 (cf. Table 2.A2, p. 51). Thus, adjustment in Canada occurred mainly via mergers and acquisitions and not widespread bank failures as in the US.26 It is, among other factors considered, an important element in the story of the Great Depression because Haubrich (1990) argues that the contraction in intermediation services did not have a separate impact on Canadian output as did the US crisis in banking (also see Calomiris, 1993). Hence, whereas the costs of providing intermediation services should not have risen in Canada they did so in the US and so further exacerbated the output decline. Yet banks in Canada (and the US) did engage in portfolio reallocation as loans declined by over 50 per cent from 1929 until 1936 while holdings of government securities rose about fourfold over the same period.27 A more satisfactory explanation may be that Canadian depositors had more confidence than their American counterparts in the efficiency and productivity of the financial system. As a result, the US banking system was perhaps more ‘fragile’ than the Canadian one. While there is no evidence to support this hypothesis for Canada, Cooper and Corbae (1997) develop a model to suggest that the US banking system was fragile at the outset of the Great Depression. So we are still left with the perhaps startling conclusion that the much-touted stability of Canadian banking did not make the country immune to the severity of the ‘great contraction’. Therefore, other factors must be at play. How important was the absence of a central bank (i.e. lender of last resort)? Under a gold standard rule, the automatic nature of the system would suggest the lender of last resort function is rather unimportant or unnecessary. As we have seen, however, the ‘rules of the game’ were not strictly followed. In the Canadian context, monetary policy actions were

Pierre L. Siklos 47

influenced by government via the Finance Act of 1914, originally meant to assist the agricultural sector through crisis, but subsequently extending loans more generally. The government of Canada essentially acted as a lender of last resort, though it was a modest role at best.28 An additional source of funds for the Canadian banking system were reserves held in US dollars. Hence, while other countries (e.g., France) resisted treating the US dollar as a reserve currency, Canada did not.29 Moreover, Bordo and Redish (1987) argue that political considerations, not the desire to have a lender of last resort after the gold standard ended, explain the emergence of a Canadian central bank in 1935. But its creation may be viewed as one of the events responsible for recovery from the Great Depression. Indeed the formation of a central bank is seen as essential to permit the use of instruments of monetary policy in a world that no longer operates with the automatic constraints provided by the gold standard (Shearer and Clark, 1984). Nevertheless, it is worth noting that in spite of institutional differences between the US and Canada in the realm of monetary policy, economic performance in the two countries is similar, so it is hard to believe that the transmission of economic shocks from the US to Canada was insignificant (Siklos, 1999).

5

The consequences: reform and recovery

If economists are interested in the impulse and propagation of the Great Depression, they are equally interested in the timing of the recovery from that same event. As shown in Figures 2.3 and 2.4 both economic activity in general, and commodity prices in particular, rebounded quickly and sharply from Depression era lows. About explanations for this outcome there is relatively less debate. Friedman and Schwartz (1963) argue that the deposit insurance in 1934 in the US marked the turning point. Wigmore (1987) suggests instead that the end of the gold standard in April 1933 marked the beginning of recovery. Since these two events are temporally close to each other it seems difficult to empirically identify the relative importance of one impulse over the other.30 Vernon (1994) points a finger at the enhanced role of fiscal policy in the aftermath of the slump. Eichengreen and Sachs (1985) and Choudhri and Kochin (1980), place a great deal of emphasis on the exchange rate regime, though neither article considers the Canadian experience. Beyond the initial impetus for change, and following the debacle in monetary policy, lay almost a decade’s worth of reforms that, as Timberlake (1993, p. 274) puts it: ‘is almost too much for mortal mind to assimilate, let alone explain’. The abandonment of the gold standard and the centralization of Federal Reserve power at the Board of Governors, embodied in the passage of the Banking Act of 1935, were the highlights of US monetary reforms of the 1930s. By contrast, the recovery in Canada was rather more

48 The USA and Canada

placid with the introduction of the Bank of Canada being the major milestone. Unlike the US experience, deposit insurance was not introduced in Canada.31 It is interesting to note that studies referred to earlier, which purport to show that the impulse for Canada’s Depression was not the US experience, fail to come to grips with why Canada’s recovery so closely paralleled that of the US. Was it part of a natural secular renewal reflecting the depths of the downturn? Was international cooperation, in the form of the World Economic Conference in June 1933, a factor in the return to growth? In Eichengreen’s (1992) view, the Conference was an ‘utter failure’. Rather, the Tripartite agreement between the US, the UK and France, combined with a large US devaluation undertaken by President Roosevelt, interpreted as a regime change in Temin’s (1989) view, were catalysts in freeing major economies from their ‘golden fetters’. Yet there are aspects of the recovery that have not been fully investigated. For example, the apparent asymmetry in business cycles has not been fully explained nor the role played by institutional considerations or government involvement in the process. It seems clear, however, that recovery was not as complete in Canada as in the US (e.g., see Figure 2.3). The incompleteness of the Canadian recovery might suggest that an important element in its economic performance originates from domestic sources. Yet both capital flows (Obstfeld and Taylor, 1997, table 2.1), and the terms of trade (Safarian, 1959, chapter 5), recovered slowly so the influence of the international transmission of business cycles in the Canadian recovery should not be underestimated . All studies of the recovery must contend with the ‘shock’ of the Second World War which, in historical terms, came shortly after the bottom of the Depression had been reached. In part for this reason an alternative approach asks: What if policy makers had not been blinded by the ideology of the gold standard? In other words, what if the Great Depression had never happened? Speculations of this kind require constructing a ‘counterfactual’ experiment. A simple example is Eichengreen’s (1992, chapter 10) argument that an open market operation (i.e., buying of government bonds) would have offset the fall in the money supply in 1931–2 in the US. A more elaborate approach (e.g., Bordo and Eichengreen, 1997) supposes that ‘stable’ monetary policies in a (fractional) gold standard would have averted disaster including the unexpected deflation of the early 1930s. Unfortunately, the authors are unable to nail down the impact of the ‘No Great Depression’ hypothesis on overall economic performance. Instead, they speculate that the gold standard in some form would have lasted beyond the Second World War. Bordo, Choudhri and Schwartz (1998) also use a counterfactual experiment to show that an expansionary monetary policy in the 1931–3 period would have not only averted the banking panics but also the consequent loss of gold reserves would not have been large enough to reduce the gold ratio below the statutory minimum

Pierre L. Siklos 49

requirement. Interestingly, none of the counterfactuals consider how an earlier introduction of a large fiscal stimulus may also have avoided the onset of the Great Slump. While interesting, counterfactuals assume that correct monetary decisions would have been possible. But as the masterful works of Eichengreen (1992), Friedman and Schwartz (1963) and Temin (1989) show, one ignores the impact of the intellectual milieu of the era at some risk. Yohe (1990), Timberlake (1993), and Calomiris and Wheelock (1997) provide fascinating accounts of how ideology is a notion not easily embedded in any model. Notwithstanding the internal consistency of the gold standard system, simulations also presume that, if only a better policy option has been presented to decision-makers, the Great Depression could have been avoided. This type of optimism is not shared by those, such as adherents to the modern political-economy literature, who argue that only clearly spelled out, accountable, and binding institutional constraints will lead to desirable and credible economic outcomes. There is, of course, another way out of the dilemma facing policy-makers in how much institutional constraint should be placed on them. Thus, for example, if an autonomous monetary authority – deemed desirable – implements ‘bad’ policy, the fiscal authorities may simply decide not to ‘play the game’ and introduce policies to offset the negative outcomes from the prevailing monetary policy. But then there still remains the potential for economic activity to become more volatile than through institutional devices which ensure that ‘good’ monetary policy is practised.

6

Concluding remarks

Almost seven decades after the end of the Great Depression, economists have learned much from the Canadian and US experiences about how economic shocks are transmitted across countries and the role and importance of personalities who make and implement economic policies. Is there a consensus then amongst economists about why the Great Depression took place, why it was so severe, and the ingredients of the recovery? Eichengreen and Temin (1997) would have us believe this is so. Contrary to historians, the authors suggest that there ‘now prevails a remarkable degree of consensus among specialists about the causes of the Great Depression’. They return to the themes listed in the opening section of this chapter, namely the ideology of the gold standard and central bankers’ attitudes about the proper policies in the face of economic shocks. While this much is probably true there is still room for academic disagreement. Whaples (1995) reports what is, in effect, a lack of consensus on the causes of the Great Depression. Based on a survey of 178 US members of the Economic History Association (EHA), only 31 per cent of historians in the EHA agreed or partially agreed with the notion that monetary forces

50 The USA and Canada

caused the Depression, while 52 per cent of economists in the EHA agreed with this principle. By contrast, 61 per cent and 51 per cent of economists and historians, respectively, traced the impulse to the Great Slump to a real shock. Consensus, if any, appears greater about Temin’s great shock hypothesis than Friedman and Schwartz’s view. Interestingly, far more consensus was found concerning the proposition that the right mix of policies could have prevented the Depression (75 per cent of economists and 78 per cent of historians agreed with this idea). Indeed, this survey has also suggested that there is disagreement about the extent to which shocks are transmitted across countries. The foregoing survey of views suggests that it is clearly unsatisfactory to suggest that a single cause is responsible for such a cataclysmic global event. Nevertheless, what is striking about the US–Canadian experience is that, despite differences between the two countries’ financial structures, both countries suffered a tremendous slump. Either institutional structures are not sufficient to insulate a country from large foreign shocks or the role of institutional factors, or how economic policy can short-circuit their role in the Depression and its aftermath is not yet sufficiently understood. One thing is certain, however. Regardless of how the Great Depression began and whatever the significance of US shocks on the Canadian economy, the principal lesson of the interwar years is that, in the area of monetary policy, ideology is no substitute for good judgment.

Futures price less spot price, one period ahead (4 months)

Appendix

80 60 40 20 Fut. less spot(t+4) OATS Fut. less spot(t+4) CORN

0 20 40 60 20

22

24

26

28

30

32

34

36

Figure 2.A1 The behaviour of spot and futures commodity prices: US, 1920–37 Note: Data from Hamilton (1992). Also, see notes to Figure 2.4.

Pierre L. Siklos 51 4800

35000

4200

25000

4000 20000

3800 3600

15000

3400

no. bank branches

US no. bank offices

4400

Canada

4600 30000

No. bank off. No. bank off.

US Canada

3200

10000 20

22

24

26

28

30

32

34

36

Figure 2.A2 The evolution of bank offices in the US and Canada, 1919–37 Source: Bordo and Jonung (1987).

Canadian dollars per 1 U.S. dollar

1.30 Convertibility ends

1.25 1.20 1.15 1.10 1.05 1.00 0.95 20

22

24

26

Average

28

30 High

32

34

36

Low

Figure 2.A3 The Highs and Lows of the Canadian/US Dollar Exchange Rate 1919–37 Source: Urquhart and Buckley (1983), series J560, J561, J562.

Notes * Troy Elyea provided valuable research assistance. I am grateful to the Social Sciences and Humanities Research Council of Canada for financial support. Theo Balderston, Richard Burdekin, Lars Jonung and LeRoy Laney provided helpful comments on earlier drafts. Previous versions of this paper were presented at the Western Economics Association Conference, San Diego, July 1999, and the Stockholm School of Economics. 1 In its purest form, the cost of holding money is part opportunity cost, part cost of storing gold reserves, part cost of transacting in gold coins.

52 The USA and Canada 2 As argued by one of the supporters of the gold standard, Sir John Bradbury, at the time Churchill was debating whether or not the UK should return to gold . 3 David Hume, the famous eighteenth-century Scottish philosopher, first described the process. ‘Specie’ is gold , silver, or other precious metal used as money. 4 Trade and investment ties between the US and Canada were also strong, of course. During the period in question roughly two-thirds of imports were from the US while approximately 60 per cent of foreign capital invested in Canada came from US sources. But ties with the UK also remained strong since over one-third of investment in Canada came from the UK while over one-third of Canadian exports went to the UK. Imports from the UK, by contrast, accounted for less than 20 per cent of Canada’s total imports. Unfortunately, there are limitations in the available data which make it difficult to determine precisely how trading patterns were affected by the Great Depression. 5 Thus, financial dislocation, in the form of reduced, though still large, capital flows suggests one avenue through which the Great Depression may have been transmitted from the US to Canada. Yet, there is a suspicion that there is more to it than that, since Figures 2.1 and 2.2 suggest that other forces may also have been at work. 6 While it is hazardous under the circumstance to compare the relative sizes of the two economies, the sheer size of the US economy is surely a factor by itself. US nominal GDP over the 1926–37 period was approximately 18 times Canada’s nominal GDP. 7 Indeed, Diebold and Rudesbusch (1990) find substantial differences in the duration of pre-First World War versus post-Second World War business cycles. I do not discuss the comparative behaviour of unemployment rates in the two countries. Zagorsky (1998) shows that existing data, which purport to show significantly lower unemployment rates in Canada during the Great Depression, are misleading. Each country handled depression relief workers differently. The US treated these individuals as unemployed, Canada did not. 8 Peaks in the US business cycle are usually dated 1 to 5 months before those in Canada, while the dating of troughs is usually 1 month ahead in the US. 9 To keep the analysis simple I only allow a one period lag. Extensions allowing more lagged terms would not fundamentally change the arguments to be outlined below. 10 See, for example, Enders (1995, chapter 5) for a description. 11 Sumner (1992) suggests that news contradicting the public’s expectations about the likely degree of central bank cooperation may have contributed to heighten uncertainty about the future monetary policy. A proxy for ‘uncertainty’ would be the standard deviation in either a measure of real income (e.g., real GDP) or in a stock price index. Modern empirical macroeconomics tends to rely on more sophisticated measures of volatility derived from the variance of the residuals from some regression estimate describing the relationship between economic series of interest. A fuller analysis of this procedure is, however, beyond the scope of this paper. 12 Based on data in column (2) of table B.2 in Dick and Floyd (1992). 13 The choice of 1928 is arbitrary in a sense but it has the advantage that it omits the period most economists would describe as the Great Depression period. 14 That is, at the time the contract expires at which time the relevant commodities are delivered at the agreed upon price.

Pierre L. Siklos 53 15 For interested readers, Appendix Figure 2.A3 plots et for the case of corn and oats, using Hamilton’s data. 16 Indeed, an equation such as (2.6), estimated during the period 1919–28, continues to perform quite well when forecasts for the 1935–7 period are made. 17 According to this theory, consumers look beyond their current level of income in determining consumption and savings levels. See Siklos (1997, chapter 9). 18 The average real interest rate in the US rises to 10.21 per cent if we begin calculations in 1920. 19 Bonser-Neal and Morley (1997) provide a thorough and introductory exposition on the subject. 20 For example, inflows of gold totalled $1,432 million between 1921 and 1924. Meanwhile, loans by investment brokers to their clients rose 400 per cent between 1919 and 1928 while instalment credit rose almost 300 per cent between 1919 and 1928. Based on figures in Historical Statistics of the United States: Bicentennial Edition (1997 CD-ROM), series X547 and X551 and Board of Governors of the Federal Reserve System (1976), table 2.2.158. 21 An aspect of the debate not discussed here is the role played in the propagation mechanism by the flexibility of nominal wages. Unlike the apparent downward rigidity in nominal wages prevalent today, there was little such rigidity in the interwar era. Bernanke and Carey (1995) point out that Eichengreen (1992), for example, makes only a passing reference to the development of real wages during this era. See also note 4 above for an explanation of data limitations in the Canadian context. 22 A similar occurrence took place after Canada devalued the dollar in 1949. The Canadian dollar floated until 1961 over objections by the International Monetary Fund about violating the Bretton Woods monetary standard. At the time of devaluation the Canadian dollar was set at 90.9¢ US. In 1961, when the $C was pegged anew the rate was 92.5¢. 23 Obviously, nominal interest rates cannot go below 0 per cent but real interest rates can rise with the severity of the deflation. 24 Through various regulatory measures meant to significantly raise the cost of entry into banking. See Siklos (1997, chapter 17). Wheelock (1992, 1995) also provides a thorough review of the US financial system during this period. 25 See Siklos (1997, chapter 16) for a basic introduction to the concept of a money multiplier. 26 If the adjustment was somehow assisted by a government bail-out then the stability of the Canadian banking system was a myth. This is the argument of Kryzanowski and Roberts (1993) but is persuasively refuted by Carr, Mathewson and Quigley (1995). Kryzanowski and Roberts (1999) reply that when one corrects for the difference between market value and book value on a bank’s balance sheet, some Canadian banks during the Depression were technically insolvent. 27 Based on data in Urquhart and Buckley (1983), series J191, J146 and J138. 28 Advances under the Finance Act never exceeded $82 million CDN (in 1929), less than 2 per cent of bank assets at the time. Based on data in Urquhart and Buckley (1983), series J182 and J198. 29 For this reason, authors such as Rich (1988), and Dick and Floyd (1992), define the Canadian monetary base to include bank reserves of US dollars. 30 Coe (1998, chapter 3) uses some sophisticated econometric techniques and concludes that the introduction of deposit insurance is the event which launched recovery in the US.

54 The USA and Canada 31 Siklos (1997, chapter 19) discusses the successful lobbying by Canadian chartered banks to prevent the introduction of deposit insurance in Canada (until 1967).

References Bernanke, B.S. (1983) ‘Non-Monetary Effects of the Financial Crisis in the Propagation of the Great Depression’, American Economic Review, 73 (March): 257–76. ——– (1995) ‘The Macroeconomics of the Great Depression: A Comparative Approach’, Journal of Money, Credit, and Banking, 27 (February): 1–28. ——– and K. Carey (1996) ‘Nominal Wage Stickiness and Aggregate Supply in the Great Depression’, Quarterly Journal of Economics, 111 (August): 853–83. Betts, C.M., M.D. Bordo and A. Redish (1996) ‘A Small Open Economy in Depression: Lessons from Canada in the 1930s’, Canadian Journal of Economics, XXIX (February): 1–36. Board of Governors of the Federal Reserve System (1976) Banking and Monetary Statistics, 1914–1941 (Washington: Federal Reserve). Bonser-Neal, C. and T.R. Morley (1997) ‘Does the Yield Spread Predict Real Economic Activity? A Multicountry Analysis’, Economic Review of the Federal Reserve Bank of Kansas City, 3: 37–53. Bordo, M.D. and B. Eichengreen (1997) ‘Implications of the Great Depression for the Development of the International Monetary System’, NBER Working Paper 5883 (January ). Bordo, M.D. and L. Jonung (1987) The Long-Run Behavior of the Velocity of Circulation (Cambridge: Cambridge University Press). Bordo, M.D. and F.E. Kydland (1995) ‘The Gold Standard as a Rule: An Essay in Exploration’, Explorations in Economic History, 32 (October): 423–64. Bordo, M.D. and A. Redish (1987) ‘Why Did the Bank of Canada Emerge in 1935?’, Journal of Economic History, XLVII (June): 405–17. Bordo, M.D., E.U. Choudhri and A.J. Schwartz (1998) ‘Was Expansionary Monetary Policy Feasible during the Great Contraction? An Examination of the Gold Standard Constraint’, Working Paper, Carleton University (July). Bordo, M.D., B. Eichengreen and J. Kim (1998) ‘Was There Really an Earlier Period of International Financial Integration Comparable to Today?’, NBER Working Paper 6738 (September). Bordo, M.D., A. Redish and R. Shearer (1999) ‘Canada’s Monetary System in Historical Perspective: Two Faces of the Exchange Rate Regime’, Working Paper, University of British Columbia (April). Bordo, M.D., H. Rockoff, and A. Redish (1994) ‘The U.S. Banking System from a Northern Exposure: Stability versus Efficiency’, Journal of Economic History , 54 (June): 325–41. Brecher, I. and S.S. Reisman (1957) Canada–United States Economic Relations (Ottawa, Ont.: Royal Commission on Canada’s Economic Prospects). Burbridge, J. and A. Harrison (1985) ‘A Historical Decomposition of the Great Depression to Determine the Role of Money’, Journal of Monetary Economics, 16: 45–54. Burdekin, R.C.K. and P.L. Siklos (1999) ‘Exchange Rate Regimes and Shifts in Inflation Persistence: Does Nothing Else Matter?’, Journal of Money, Credit and Banking, 31 (May): 235–47. Burns, A.F. and W.C. Mitchell (1946) Measuring Business Cycles (New York: National Bureau of Economic Research).

Pierre L. Siklos 55 Calomiris, C.W. (1993) ‘Financial Factors in the Great Depression’, Journal of Economic Perspectives, 7 (Spring): 61–85. ——– and J.R. Mason (1997) ‘Contagion and Bank Failures during the Great Depression: The June 1932 Chicago Banking Panic’, American Economic Review, 87 (December): 863–83. ——– and D.C. Wheelock (1997) ‘Was the Great Depression a Watershed for American Monetary Policy?’, NBER Working Paper 5963 (March). Carr, J., F. Mathewson and N.C. Quigley (1995) ‘Stability in the Absence of Deposit Insurance: The Canadian Banking System’, Journal of Money, Credit and Banking, 27: 1137–38. Cecchetti, S.G.(1988) ‘The Case of the Negative Nominal Interest Rates: New Estimates of the Term Structure of Interest Rates During the Great Depression’, Journal of Political Economy, 96 (6): 1111–41. ——– (1992) ‘Prices During the Great Depression: Was the Deflation of 1930–32 Really Unanticipated?’ American Economic Review, 82 (March): 141–56. ——– and G. Karras (1994) ‘Sources of Output Fluctuations During the Interwar Period: Further Evidence on the Causes of the Great Depression’, The Review of Economics and Statistics, 76: 80–102. Choudhri, E.U. and L.A. Kochin (1980) ‘The Exchange Rate and the International Transmission of Business Cycle Disturbances: Some Evidence from the Great Depression’, Journal of Money, Credit and Banking, 12: 565–74. Coe, P. (1988) ‘Money, Output and the United States Interwar Financial Crisis: An Empirical Analysis’, Ph.D. dissertation, University of British Columbia. Cole, H.L. and L.E. Ohanian (1999) ‘The Great Depression in the United States from a Neoclassical Perspective’, Quarterly Review, Federal Reserve Bank of Minnesota, (Winter): 2–24. Cooper, R. and D. Corbae (1997) ‘Financial Fragility and the Great Depression’, NBER Working Paper 6094 (July). Courchene, T.J. (1969) ‘An Analysis of the Canadian Money Supply: 1925–1934’, Journal of Political Economy, 77: 363–91. Dick, T.J.O. and J.E. Floyd (1992) Canada and the Gold Standard (Cambridge: Cambridge University Press). Diebold, F.X and G.R. Rudebusch (1990) ‘A Non-Parametric Investigation of the Duration Dependence in the American Business Cycle’, Journal of Political Economy, 98 (June): 596–616. Dominguez, K.M., R.C. Fair and M.D. Shapiro (1988) ‘Forecasting the Depression: Harvard versus Yale’, American Economic Review, 78 (September): 595–612. Eichengreen, B. (1992) Golden Fetters. The Gold Standard and the Great Depression 1919–1939 (New York: Oxford University Press). ——– and J. Sachs (1985) ‘Exchange Rates and Recovery in the 1930s’, Journal of Economic History, 45: 925–46. ——– and P. Temin (1997) ‘The Gold Standard and the Great Depression’, NBER Working Paper 6060 (June). Enders, W. (1995) Applied Econometric Time Series (New York: John Wiley & Sons). Evans, M. and P. Wachtel (1993) ‘Were Price Changes during the Great Depression Anticipated? Evidence from Nominal Interest Rates’, Journal of Monetary Economics, 32: 3–34. Fackler, J.S. and R.E. Parker (1994) ‘Accounting for the Great Depression: A Historical Decomposition’, Journal of Macroeconomics, 16 (Spring): 193–220. Fremling, G.M. (1985) ‘Did the United States Transmit the Great Depression to the Rest of the World?’, American Economic Review, 75 (December): 1181–5.

56 The USA and Canada Friedman, M. and A.J. Schwartz (1963) A Monetary History of the United States 1867–1960 (Princeton: Princeton University Press). ——– (1982) Monetary Trends in the United States and the United Kingdom (Chicago and London: University of Chicago Press). Garraty, J. (1986) The Great Depression (New York: Harcourt Brace Jovanovich). Gilbert, M. (1977) Winston S. Churchill, Volume 5: 1922–1939 (Boston: Houghton Mifflin). Grossman, A.S. (1994) ‘The Shoes that Didn’t Drop: Explaining Banking Stability during the Great Depression’, Journal of Economic History, 54 (September): 654–82. Hamilton, J.D. (1987) ‘Monetary Factors in the Great Depression’, Journal of Monetary Economics, 19: 145–69. ——– (1988) ‘Role of the International Gold Standard in Propagating the Great Depression’, Contemporary Policy Issues, VI (April): 67–89. ——– (1992) ‘Was the Deflation during the Great Depression Anticipated? Evidence from the Commodity Futures Market’, American Economic Review , 82: 157–78. Haubrich, J.G. (1990) ‘Nonmonetary Effects of Financial Crisis: Lessons from the Great Depression for Canada’, Journal of Monetary Economics, 25: 223–52. Hay, K.A.J. (1967) ‘Money and Cycles in Post-Confederation Canada’, Journal of Political Economy, 75 (June): 262–73. Keynes, J.M. (1931) ‘An Economic Analysis of Unemployment (Harris Lectures)’, as reprinted in The Collected Writings of John Maynard Keynes, ed. D.E. Moggridge, vol. 13 (London: Macmillan, 1973): 343–67. Kindleberger, C.P. (1986) The World in Depression, 1929–1939, 2nd edn (Berkeley: University of California Press). Kryzanowski, L. and G.S. Roberts (1993) ‘Canadian Banking Solvency, 1922–40’, Journal of Money, Credit and Banking, 25: 361–76. ——– (1999) ‘Perspectives on Canadian Bank Insolvency During the 1930s’, Journal of Money, Credit and Banking, 31: 130–34. Lucas, R.E., Jr. (1987) Models of Business Cycles (Oxford: Basil Blackwell). Metcalf, C., A. Redish and R. Shearer (1998) ‘New Estimates of the Canadian Money Stock, 1871–1967’, Canadian Journal of Economics, 31 (February): 104–24. Mishkin, F.S. (1978) ‘The Household Balance Sheet and the Great Depression’, Journal of Economic History, 38: 918–37. Nichans, J. (1978) The Theory of Money (Baltimore: Johns Hopkins University Press). Nixon, S.E. (1937) ‘Interest Rates in Canada: The Course of Interest Rates, 1929–1937’, Canadian Journal of Economics and Political Science , 3: 421–34. Obstfeld, M. and A.M. Taylor (1997) ‘The Great Depression as a Watershed: International Capital Mobility over the Long-Run’, NBER Working Paper 5960 (March). Prescott, E.C. (1999) ‘Some Observations on the Great Depression’, Quarterly Review, Federal Reserve Bank of Minnesota, (Winter): 25–31. Rich, G. (1988) The Cross of Gold : Money and the Canadian Business Cycle, 1867–1913 (Ottawa, Ont.: Carleton University Press). Romer, C.D. (1990) ‘The Great Crash and the Onset of the Great Depression’, Quarterly Journal of Economics, 105 (August): 597–624. Romer, D.(1996) Advanced Macroeconomics (New York: McGraw-Hill). Safarian, A.E. (1959) The Canadian Economy in the Great Depression (Toronto: University of Toronto Press).

Pierre L. Siklos 57 Shearer, R.A. and C. Clark (1984) ‘Canada and the Interwar Gold Standard, 1920–35: Monetary Policy without a Central Bank’, in M.D. Bordo and A.J. Schwartz (eds), A Retrospective on the Classical Gold Standard (Chicago and London: University of Chicago Press): 277–310. Siklos, P.L. (1997) Money, Banking and Financial Institutions 2nd edn (Toronto: McGraw-Hill Ryerson). ——– (1999) ‘US and Canadian Central Banking: The Triumph of Personalities Over Politics?’, in C.-L. Holtfrerich, G. Toniolo and J. Reis (eds), The Emergence of Modern Central Banking (Leicester, UK: Scolar Press for the European Association of Banking History): 231–78. Sumner, S. (1992) ‘The Role of the International Gold Standard in Commodity Price Deflation: Evidence from the 1929 Stock Market Crash’, Explorations in Economic History, 29: 290–317. Temin, P. (1976) Did Monetary Forces Cause the Great Depression? (New York: Norton). ——– (1989) Lessons from the Great Depression (Cambridge, MA: The MIT Press). ——– (1993) ‘Transmission of the Great Depression’, Journal of Economic Perspectives, 7: 87–102. Timberlake, R.H. (1993) Monetary Policy in the United States (Chicago and London: University of Chicago Press). Urquhart, M.C. and K.A.H. Buckley (1983) Historical Statistics of Canada (Ottawa: Statistics Canada). Vernon, J.R. (1994) ‘World War II Fiscal Policies and the End of the Great Depression’, Journal of Economic History, 54: 850–68. Whaples, R. (1995) ‘Where is there Consensus among American Historians? The Results of a Survey on Forty Propositions’, Journal of Economic History, 55 (March): 139–54. Wheelock, D.C. (1992) ‘Monetary Policy in the Great Depression: What the Fed Did, and Why’, Review of the Federal Reserve Bank of St. Louis (March/April): 3–28. ——– (1995) ‘Regulation, Market Structure, and the Bank Failures of the Great Depression’, Review of the Federal Reserve Bank of St. Louis (March/April): 27–38. Wigmore, B.A. (1987) ‘Was the Bank Holiday of 1933 Caused by a Run on the Dollar?’ Journal of Economic History, 47: 739–55. Yohe, W.P. (1990) ‘The Intellectual Milieu at the Federal Reserve Board in the 1920s’, History of Political Economy, 22 (3): 465–88. Zagorsky, J.L. (1998) ‘Was Depression Era Unemployment Really Less in Canada than the US?’, Economics Letters, 61: 125–31.

3 France in the Depression of the Early 1930s Pierre Villa

1

Introduction

Neither the degree of responsibility of the French economy for the Great Depression, nor its influence upon it, are easy to appraise, because France entered the Depression late and many of its adjustments seem at first glance to be idiosyncratic. One is thus tempted to regard the French case as sui generis. However, some contemporaries, and economists such as A. Sauvy, do blame the lengthening and deepening of the Depression in France on the misunderstanding of economics by contemporary policymakers, and especially their fixation on certain monetary and financial doctrines. Such criticisms have been reinforced by Anglo-Saxon, especially British, accusations, firstly that France implemented a restrictive monetary policy as of 1928 in order to accumulate gold reserves and thereby exert political pressure in relation to the Reparations question, and secondly that France did not employ expansionist, open-market monetary operations to accommodate the depression after 1930. They claim that in a certain sense economic policy was overloaded by the political objectives it was also made to serve. Eichengreen considers that the gold standard had a deflationary bias. Countries experiencing relatively rapid growth implemented restrictive monetary policies which induced gold inflows from countries whose recessions had loosened the gold-standard policy constraints on them. Capital flight was quicker and greater than current account adjustment, so gold shifted into growth countries because of high interest rates, even although they had current account deficits. In ‘follow the leader’ fashion other countries with current account deficits had to implement restrictive monetary policies in order to maintain the credibility of their gold parity. In this way the lack of intelligent coordination of monetary policies lengthened the Depression. In my opinion such analyses overestimate the impact of monetary policy on the money stock; in fact much of the rise in the French gold stock was, 58

Pierre Villa 59

given the demand for money, simply a response to the decline of other components of the assets held by the central bank and the commercial banks against their monetary liabilities, i.e. of Bank of France credit to the state and of commercial bank credit to the private sector. Furthermore the relative efficacy of monetary policy, compared with fiscal and real exchange-rate policy, needs to be re-evaluated. In essence, monetary policy is coarse and should basically peg the long-term interest rate to the growth rate. But fiscal policy is more precise, and the greater variety of fiscal than monetary policy instruments makes it possible to employ them for the simultaneous achievement of a variety of specific targets. However, in France during this period economic policy was never thought of as a means of managing activity . Rather than being deliberately adjusted to achieve certain internal targets, policy was simply made to conform to two doctrines: the ‘balanced budget’ doctrine, and the ‘stabilization of credit conditions’ doctrine. Budgetary policy was, therefore, restrictive both in intention and in result until the end of 1930. This allowed the stabilization of the franc, the promotion of growth and the implementation of a counter-cyclical policy at a time when the economy was at full factor utilization. From 1931 onwards, when France was in a Keynesian unemployment regime, both monetary and fiscal policies remained restrictive in intention in order to defend the gold parity of the franc and to obey the ‘balanced budget’ doctrine. But whilst monetary policy was also restrictive ex post, fiscal policy was for a large part expansionary and countercyclical in its out-turn, because of the stickiness of expenditure, of automatic stabilizers and of transfers to the colonial empire. However, ideological resistance to counter-cyclical fiscal policy was strong in the private sector. This resistance induced an increase in real long-term interest rates from 1931 onwards, and hence (even though there was no financial crisis) a dramatic fall in private investment. On the other hand, consumption played a counter-cyclical role, because the rise (due to deflation) in the real value of private-sector wealth held as public debt and money reduced the propensity to save. Lastly, real exchange-rate appreciation and foreign trade-quota policies caused a fall in export profitability, which exerted a large negative influence on firms’ supply. To manage the depression would have meant stabilizing competitiveness and export profitability by lowering the parity of the franc to gold, lowering short-term interest rates and boosting domestic demand by public expenditure increases of a magnitude that was unthinkable at the time. The chapter is divided into four parts. The first explains why France entered the Depression late, contrary to the claims of some economists that recession commenced in 1928 due to an alleged fall in profitability or an excess supply of goods, or because mass-consumption had not yet developed. But the deferral of the French recession was not, as has also been alleged, due to some undervaluation of the franc, because in fact the trade

60 France

balance, competitiveness and export profitability all deteriorated during the 1926–30 period. In this chapter a dynamic portfolio model is used to show that the French growth in this period resulted from an appreciation of the real exchange rate, and not from its depreciation, as erroneously suggested by Eichengreen and Wyplosz (1988). This model will show that, near full employment, real-exchange-rate appreciation increased long-run supply by increasing the purchasing power of salaries whilst decreasing real unit labour costs. This explains the fall of nominal interest rates during the 1926–30 period, that accompanied the high growth and deficitary trade balance. In such a framework French monetary policy was not restrictive (or not more restrictive than elsewhere) in respect of the long-term interest rate. In this period a policy of international cooperation would have entailed an appreciation of the franc and a lowering of interest rates in the US and the UK. It is then shown that a change in the underlying economic regime occurred in France during 1930 with the appearance of Keynesian unemployment and ex post budget deficits. France was hit by a negative demand shock. The last two parts of the chapter show that this shock was not accommodated by monetary and fiscal policy, even if the latter was not as restrictive ex post as intended. Gold inflows were basically initiated by the return of previous French asset outflows. Nevertheless, an expansionary policy could have been implemented by lowering interest rates, because the decreasing money demand from 1933 onwards was not due to banking panics and a financial crisis, but to the fall in investment as a consequence of the lack of demand and of the high level of interest rates (i.e. of the user cost of capital).

2

France’s late entry into the Great Depression

Although the prevalent thesis maintains that France entered the Great Depression late and that recession was to a great extent imported, some authors date it from 1928, and claim that it had internal causes. In this section it is shown that such theses are not consistent with the facts. Two well-known theses predicated upon the late entry of France are then discussed, and shown also to be incompatible with the facts. Lastly an explanation of the macroeconomic chain of events from 1926 until 1931 is put forward, according to which Poincaré’s restrictive fiscal policy, in a full employment regime, induced a late expansion of supply in a way that was prejudicial to the growth of foreign trade. Theses about the precocity of the crisis in France According to Marseille (1980) the French Depression began as early as 1928. Though concealed by the undervaluation of the franc, it appeared both in traditional export-orientated but inadequately protected industries

Pierre Villa 61

(such as textiles), and in modern ones suffering excess capacity (such as cars and metallurgy). According to Marseille, France was in a generalized overproduction regime, because the growth of labour productivity increased supply more rapidly than the growth of domestic and foreign demand. Internal demand grew slowly because the purchasing power of salaries was lagging and because traditional patterns of consumption still characterized the peasantry and the general population of ‘domestic’ and ‘self-employed’ workers. In the same vein, Boyer (1979) considers that the purchasing power of wages was lagging behind productivity gains and preventing France from developing forms of urban mass consumption. In his words, ‘Fordism’ came late in France. Contrary to these analyses, however, the French economy was at full utilization of productive capacity and at full employment until the end of 1930. The length of the working week did not decrease below the official 48-hour threshold until 1930 and fell only in 1931.1 Moreover, unemployment, computed from census and employment-bureau statistics, rose little above its low 1926 level (except during the 1927 recession) until 1931, after which it rose.2 Furthermore, there was no overproduction crisis. From 1926 to 1930 aggregate demand for French goods increased faster (3.45 per cent per year) than output (GDP: 2.5 per cent per year), inducing an increase in prices. Domestic absorption grew even more rapidly (4.3 per cent per year). These figures imply that inventories were being cut and that the tradebalance deficit was increasing over the period. Lastly, the rate of profit (including dividends) hovered at a high level until the end of 1930 (see Table 3.1). This was the net effect of three processes. Firstly, wages were geared to prices, but with lags which tended to enlarge profits as prices rose. However, secondly, export profitability was falling as international competition compelled exporters to reduce their prices on foreign markets. But, thirdly, the decrease in raw material prices was lowering the cost of imported inputs. From an analytical point of view, the theses of Marseille and Boyer generalize incorrectly from partial examples. The recession in textiles was not peculiar to France, and the troubles in iron and steel arose from the capacities inherited after the annexation of Alsace-Lorraine. Rather, the increase in the trade deficit seems to have been the logical consequence of general excess demand in an economy at full employment, and of the spread of quantity rationing and quotas in international trade.3 Explanations of the late entry of France into the Depression The traditional explanation of France’s late entry into the Depression is derived from A. Sauvy.4 In his view, undervaluation of the franc caused strong GDP growth from 1926 onwards; and it avoided excessive inflation because of the very high supply elasticity with respect to prices. But this thesis, too, is not compatible with the data. It cannot explain the following

62 Table 3.1

The stabilization of the franc 1925

1926

1927

1928

1929

1930

1931

Public deficit/GDP (in per cent)

–2.37

2.09

2.04

–0.17

1.92

1.66

–1.00

Taxes/GDP (in per cent)

12.6

14.6

15.6

16.0

15.7

15.2

16.1

Trade balance/GDP (in per cent)

1.12

0.65

1.10

–0.07

–1.58

–2.13

–2.90

Foreign interest income/GDP (in per cent)

0.61

0.85

0.67

0.84

1.30

1.25

1.04

Nominal effective exchange rate1

0.59

0.76

0.61

0.62

0.62

0.62

0.61

Real effective exchange rate1

1.00

1.15

0.88

0.90

0.87

0.80

0.73

Purchasing power of wages (base 1 in 1928)2

1.04

1

0.93

0.98

1.05

1.08

1.09

Real unit wage costs (base 1 in 1928)3

1.02

1

0.92

0.96

0.92

0.92

0.86

Growth differential (in per cent)4

–2.0

–2.0

–10.4

3.8

6.4

6.3

5.8

Long term interest rate difference (in per cent)5

4.7

4.2

2.0

0.9

0.3

-0.6

–0.8

15.0

14.3

13.5

14.3

13.7

12.7

10.7

5.4

6.0

7.9

14.8

16.9

10.6

5.1

Private bonds 2.0 issues (billion of Ff.)

2.5

5.7

5.4

10.2

14.5

9.1

Profit rate including dividends (in per cent) Stock issues (billion of Ff.)

1

Weighted exchange rates of the 6 main competitors of France. Related to the consumer price index. 3 Related to the GDP price index. 4 Growth difference between France and its six main competitors (The UK, Germany, Italy, Belgium, Switzerland and the US). 5 France minus the UK (public debt and consols). Source: Villa (1994), base 1938 except when it is specified. 2

Pierre Villa 63

facts: the trade balance became negative in 1928; and the real exchange rate began to appreciate and the profitability of exports to decrease, both as of 1926, i.e. of the very beginning of the Poincaré stabilization. In order to take account of the buoyancy of demand and the trade deficit, Eichengreen and Wyplosz (1988) suggest that the strong French growth was the consequence of the stabilization of the franc in the following way. The budget surplus resulting from a restrictive fiscal policy was applied to the reimbursement of the public debt. This policy therefore induced an increase in the holding of financial assets denominated in foreign currency. But to make the private sector accept these assets in place of the retired government debt, either the return on foreign assets must be increased (i.e. the expectation of a depreciation of the real exchange rate of the franc against foreign currencies must be created), or the yield of French assets must decrease. Where the marginal productivity of physical capital remains constant, the yield on French assets can fall only if the expected yield on shares falls due to an expected decline in their capital value. This implies that the price of shares will have increased in the short run by ‘overshooting’. This short-run increase in share prices will increase the demand for investment goods and the supply of exportable goods, and the excess supply of exports induces a real depreciation of the franc. If the French interest rate decreases enough, there must be an expectation of a real appreciation of the franc and so short-run ‘overshooting’ of the degree of depreciation required in the longer term to raise the rate of return on foreign, relative to domestic, assets. That is, the depreciation of the real exchange rate in the short run will be greater than in the long run.5 The model has two kinds of drawbacks: It gives simulation results which do not fit the facts and it is questionable in its description of the economic system. First as to the simulation results. The direct impact of the stabilization policy was in reality an appreciation of the real exchange rate in both the short and long term, a decrease in competitiveness and also in the profitability of exports. Thus the trade balance deteriorated throughout the period, 1927 excepted (Table 3.1). The Eichengreen–Wyplosz model gives exactly the opposite results. Moreover, while a decrease in the excess of the French over the British interest rate is to be observed, the price of French shares kept on increasing throughout the stabilization period until 1930 without any ‘overshooting’. Finally, fiscal policy produced a recession in 1927 big enough to increase unemployment and to boost foreign trade by reducing absorption. In our view, the discordance between the simulation and the actual effects of the policy springs from three misrepresentations of the economic system. • Firstly, there was no perfect substitution between shares and bonds. The nominal long-term interest rate decreased continuously from 1926 until

64 France

1931, whereas the yield on shares (price change plus dividends) remained stable until 1930 (as measured by the rate of profits), or increased until 1929 (as measured by the price of shares: see Tables 3.1 and 3.2). So the risk premium on shares was large and variable. Moreover, statistics of the financial market show a big increase in bond financing during 1929 and 1930; and these were years of high investment. These observations show that the share market remained limited to a few agents (large banks did not buy them), and that their price did not matter nearly as much in investment decisions as the interest rate.6 • Secondly, the interest rate is not the variable determining the equilibrium between supply and demand on the goods market as it is in the Eichengreen–Wyplosz model. In the short run, the interest rate was determined by monetary policy, and there was a link between the interest rate on bonds and the discount rate. So in the short run a decrease in demand induced a decrease in the nominal interest rate, as in all Keynesian models, but had an ambiguous effect on the rate of profits: the marginal rate of productivity of capital increased, but the scale of output diminished. That is why nominal interest rates fell after the restrictive policy of 1926, while the rate of profits was almost constant. • Thirdly, the model should explain the short-run recession in 1927 due to a budget surplus, the short-run expansion following the budget increase of 1928, and the full employment regime which took place afterwards should form a long-run equilibrium. An explanation with a portfolio model We suggest a unified explanation of the stabilization of the franc and of France’s late entry into the Depression, using a dynamic model which distinguishes short and long-term effects (Villa, 1993a, ch. 5). The economy is described by an open-economy Mundell–Fleming model. Its specific features are confined to a price–wage loop and the modelling of export behaviour. Econometric results show that prices were pegged by firms, with some adjustment lags, in relation to unit labour costs and the price of imported raw materials. Wages were geared to consumption prices, with adjustment lags, but consumption prices were dependent only on the GDP price index, because there were few imported consumption goods other than from the colonial empire. Moreover, exports were a function of profitability (i.e. export price/GDP price) and not of competitiveness (export price/foreign price). This was the consequence of quantitative rationing in foreign trade and of a high level of specialization. Exports were oriented in sectors with high profitability and not designed to gain market shares. Agents’ portfolio choice is introduced into this neo-Keynesian model. Like Eichengreen and Wyplosz, we assume that there is no perfect substitutability between national and foreign assets. Agents want to hold only a part of their wealth in foreign-currency assets, the share depending on

Pierre Villa 65

differences between the expected yields (i.e. the difference between the French interest rate, and the foreign interest rate plus the expected rate of depreciation of the franc exchange rate). The balance of payments determined the required change in French holdings of foreign assets and gold; this change is equal to the sum of interest earned on foreign assets, the actual appreciation of the exchange rate (taking into account the effect of exchange-rate variation on the franc value of foreign assets) and the level of the trade balance. A surplus on the trade balance requires an increase in Table 3.2 The entry of the French economy into the depression (1938 base if not specified)

Rate of growth (per cent) Capacities GDP Demand excluding inventories Domestic expenditures excluding inventories

1928

1929

3.1 6.1 4.9 5.5

Levels Trade balance (billion 1.4 of Ff.) Degree of capacity 93.3 utilization (per cent) Unemployment rate 1.3 (per cent) Inflation rate-CPI 0.2 (per cent) Nominal long-term 5.33 interest rate (per cent) User cost of capital 10.5 (per cent)1 Profitability 12 1.0 Profitability 23 1.0 Competitiveness4 1.39 Export profitability5 1.26 1

1930

1931

1932

1933

4.2 8.9 4.5

5.1 –2.6 1.5

2.3 –3.9 –6.8

–0.4 –8.8 –5.7

0.1 3.0 2.0

6.4

3.7

–5.3

–5.7

2.3

–0.7

–10.6

–14.6

–22.6

1.8

97.5

90.3

84.8

77.7

80.0

1.2

1.2

2.4

3.7

3.7

4.2

3.5

–2.9

–6.7

–3.3

4.89

3.82

8.6

8.3

1.14 1.47 1.37 1.20

1.08 1.32 1.41 1.13

3.70

10.4 0.80 0.73 1.34 0.98

4.73

13.6 0.49 0.49 1.36 0.87

5.74

13.2 0.60 0.51 1.35 0.83

Putty-clay model, with depreclation and a finite lifetime of investment. Rate of profits/user cost of capital, base 1 in 1928. 3 Real price of shares (deflated by investment prices)/user cost of capital, base 1 in 1928. 4 Exports price/price of the six main competitors of France (the US, the UK, Germany, Belgium, Italy and Switzerland). 5 Exports price/GDP price. Source: Villa (1994) 2

66 530 510 490 470 450 430 410 390 370 350 330

Capacities GDP Aggregate demand excl. inventories

310 1926 Figure 3.1

1928

1930

1932

1934

1936

1938

Aggregate supply and demand in France, 1926–38 (at 1938 prices)

100

Unemployment rate Capacities

90 80 70 60 50 40 30 20 10 1926 Figure 3.2

1928

1930

1932

1934

1936

Degree of factor utilization in France, 1928–38 (per cent)

1938

Pierre Villa 67

the part of wealth held in foreign-currency assets. This increase will not occur unless it corresponds to the portfolio choices of private agents. The domestic interest rate must decline, to induce French agents to hold more wealth in such assets. This fall in the home interest rate increases absorption, and hence reduces the trade surplus. This stabilizes the balance of payments. However this is not the mechanism which ensures the instantaneous equilibrium between the demand for foreign assets as determined by portfolio choice, and the supply of foreign assets, as determined by the balance of payments equilibrium. This equilibrium is achieved through variation of the real exchange rate. If there is a balance of payments surplus – i.e. if wealth held in foreign assets and gold has to increase – the real exchange rate appreciates to increase the expectation of a later depreciation, and thus increases the yield of foreign-currency-denominated assets. So the demand for foreign currency increases, whilst the exchange-rate losses as the franc appreciates reduce the supply of foreign assets. In 1926 Poincaré launched a permanent budget surplus policy which, by depressing internal demand, induced a trade surplus. And as the surplus accumulated, it led to an increase in the foreign-currency wealth of France. The long-term equilibrium effect of this increase was to generate interest flows which had to be offset by a trade deficit. Such a deficit could only be brought about by a higher rate of growth of production in France than abroad, and by an appreciation of the real exchange rate, lowering competitiveness and export profitability. Moreover, this real appreciation of the franc meant a decrease in import prices, hence an increase in firms’ profitability. This in turn induced an extension of the supply of goods. At the same time, because wages were pegged on lagged consumption prices, and because inflation was cut down, the restrictive policy could increase the purchasing power of wages to the employee (without raising the ‘own product real wage’ to the employer), and so induce an extension of the labour supply. This double extension of supply had an expansionary effect on production which is – in the long run – supply-determined. In addition, in the long run, the interest rate varies to adjust demand to supply. The consequence of this was both to stimulate the demand for goods and to make the French willing to hold a larger part of their wealth in foreign currency and gold. The long-run effects are thus classical: a restrictive fiscal policy has a positive impact on growth. This explains why a restrictive fiscal policy, maintained for five years, and starting from a situation of full employment in 1926, induced sustainable growth in 1929 and 1930. In the short run the consequences of the policy are different, and Keynesian in nature. The decrease in public spending lowered demand and production, and caused a trade surplus. In order to induce French investors to hold this additional wealth in foreign currency assets, the French interest rate had to decrease and the nominal and real exchange rate to appreciate, as outlined above.

68 France

The permanent fiscal-surplus policy should thus have resulted in a trade surplus in the short term and a trade deficit in the long term; lower growth than abroad in the short term but faster growth than abroad in the long term; an increase in assets in gold and foreign currency held by the French in the short and long run; an increase in the income from these assets; a permanent appreciation of the nominal and real exchange rate of the franc; and a permanent decrease in the interest-rate differential against foreign countries. This is what a simulation of the model shows and what can be observed in the data during the period of budget surpluses between 1926 and 1930, if the temporarily expansionist fiscal policy of 1928 is excluded (see Table 3.1). Thus, under circumstances of adequate global aggregate demand, the same policy as accounts for the stabilization of the franc, and the inflow of gold, also explains why France avoided the Depression for so long, by accounting for the real appreciation of the franc, the trade deficit and the exceptional growth at the end of the 1920s.

3 The unfolding of the Depression in France from the end of 1930 The commencement of the Depression in France represented a change in the underlying economic regime: industrial production declined in June 1930, de-seasonalized consumer prices in December 1930, and the deseasonalized unemployment rate rose as of December 1930 onwards.7 Until this time, France was in an aggregate supply regime as has been shown: the increase in the purchasing power of salaries increased the labour supply and the decrease in labour costs boosted the supply of goods. Aggregate demand was growing (4.5 per cent in 1929 and 1.5 per cent in 1930), but its structure was distorted. The trade balance worsened with competitiveness stagnating, export profitability decreasing and world demand receding (Table 3.2). The first two factors are a consequence of domestic inflation, and the third of the prior recession among France’s principal trade competitors. In contrast, domestic demand (excluding inventories) grew rapidly because of housing investment and productive investment by firms. The latter evolution resulted not only from the general equilibrium relationship, but also from special factors. As to the former, full employment and the restrictive fiscal policy reduced the real interest rate and increased profitability, whether measured both by the ratio of profits to user cost of capital, or by the ratio of shares to user cost of capital (see Table 3.2).8 So from 1927 onwards and especially in 1929 and 1930, it is observable that a huge amount of share and bonds were issued by quoted firms. Nevertheless, it is not possible to provide a complete econometric explanation of this surge in investment in 1930. We must invoke contingent

Pierre Villa 69

historical circumstances as special factors. Indivisible investment programmes in capital equipment industries (mechanical and electrical engineering), launched in 1928 and 1929, ended in 1930 and 1931: these had stemmed from the building of car and hydro-electric plants, the duration of whose construction had lasted about three years. Even these factors remain statistically insufficient. The remainder of the boom in equipment investment in 1930 must thus be read either as a measurement problem or as an error by firms in their expectation of demand. This last interpretation gains plausibility from the degree of capacity utilization, and the unemployment and inflation rates, which were the only economic variables observable by contemporaries. From 1931 onwards France was in a Keynesian unemployment regime. This assertion can be verified in two ways: on the one hand by observation of demand, supply and unemployment statistics, and on the other by indirect evidence such as the nominal long-term interest rate and the price level. As suggested by Temin (1976) the decrease in aggregate demand should have induced a decrease in prices and nominal interest rates, but the fall in the latter can be counteracted by a restrictive monetary policy.9 The data (see Table 3.2 and figures) corroborate the diagnosis of a Keynesian recession. This fall stemmed basically from investment. An econometric study (see Villa, 1993b) shows that housing purchases fell suddenly because the government stopped distributing subsidies based on war damage. This fall was later aggravated by the high real interest rates. Likewise, the fall in firms’ investment can be econometrically explained by the decrease in aggregate demand and the fall in the profit rate. The profit rate fell even though employment and the weekly working hours adjusted very rapidly. The rise of salaries was more than offset by redundancies and the decrease in working hours.10 Thus real unit labour costs still decreased in 1931, but the rate of profits did so too, because the scale of production diminished. The depreciation of the pound reinforced the downturn of absorption. This situation continued in 1932, and the recovery in 1933 had three origins: good competitiveness until the dollar devaluation; a rising budget deficit; and a surge in household consumption fuelled by wealth effects. Then investment increased via an accelerator effect. According to some authors (Romer, 1990), consumer spending would have behaved procyclically during the Depression, because of the fall and uncertainty of stock prices. Estimates on French data show the contrary to have been true: the propensity to consume rose during the Depression years (1931, 1932), and fell in 1933 during the recovery. There are three reasons for the countercyclical role of consumption. The increase of the share of salaries in total household income increased the propensity to consume; deflation and the fall in consumer prices enlarged real financial wealth and especially real monetary wealth; and finally, households cut their savings (by which they

70 France

had previously sought to rebuild their financial wealth), because prices were decreasing (see Villa, 1996). After 1933 and the dollar devaluation, and until 1936, the economic evolution in France can be explained by the overvaluation of the franc, the Gold Bloc, and the deflationary policy. These points raise neither factual questions nor controversial interpretations among economists. In any case, they will not be investigated, because they are beyond the aim of this chapter. Given this chronological limit, the next two sections examine to what extent French economic policy and the gold standard might have triggered or magnified the Depression in France and abroad.

4

Fiscal policy in France from 1929 until 1933

Broadly speaking, the French government did not adopt any deliberate policy over this period of managing aggregate demand by using autonomous public expenditures. The countercylical characteristics of the public balance are merely the consequence of ex post rigidities in spending, and of the automatic stabilizers. Policy was not Keynesian in this period, despite the presence of Keynesian demand-deficiency in the economy since 1930. Until the end of 1930 fiscal policy remained restrictive: Poincaré and his successors tried – if the (involuntary11) fiscal stimulus in 1928 is disregarded – to maintain a financially orthodox budget surplus of 2 per cent of GDP, and to reduce the public debt (as well as long-term interest rates). As explained above, this policy owed its success to the fact that France was at full employment. The only autonomous expenditures contemplated by policy-makers were the so-called ‘big public works programmes’. The ‘Tardieu’ programme, proposed in 1929, aimed to increase supply (by infrastructural expenditures). It was rejected by Parliament. If it had been carried out, it would have had an inflationary impact and would have increased the trade-balance deficit, because the economy was at full capacity utilization both of labour (unemployment stood at 1.2 per cent) and of capital (capacity utilization stood at 97 per cent).12 In fact, contemporary policy-makers rejected the plan for other reasons: they were afraid that by suppressing the budget surplus it would call into doubt the credibility of Poincaré’s stabilization policy and cause capital flight. This is an example of the deflationary bias of the gold standard. Eichengreen (1992, p. 10) invokes this effect in the radically different post-1918 context, when big fiscal deficits had become the arena of internal conflicts over who would ‘pay for’ the needed (real) debt reduction. The deflationary bias inherent in the decision to link money to gold, and associated restrictive policy, eliminated any expectation of inflation or of taxation levels based upon the size and possible monetization of public debt; without this ‘bias’ capital flights

Pierre Villa 71

and associated losses of gold reserves would render the gold standard regime impossible. An expansionary fiscal policy can thus run up against the gold standard rules – in the shape of the required ratio between gold and the note issue. Ironically, this regime worked best at the end of the 1920s, since it forced France to adopt a restrictive policy at full employment.13 Other public works programmes were planned and actually implemented when France was in depression.14 The macroeconomic consequences of these programmes, directed as they were towards sectors with a high labour/capital ratio and facing little competition through international trade (education, agriculture, public works), have been disputed (Saly, 1980). It is difficult to assess them since sometimes they represented monetary expenditures not applied to the purchase of goods, and at other times subsidies substituted for credits already granted by the government. It is thus hard to distinguish between consumption, investment, and operating expenditures. In spite of these ambiguities, it may be considered that this policy did contribute to the increase in public expenditure and have a countercyclical effect (see Table 3.3). But these works were not the reason for the abrupt appearance of a budget deficit in 1931. Public consumption and investment had been growing from 1928. But the change of fiscal regime in 1931, which occurred in spite of an increase in the average tax rate, had in fact other, if undesigned, origins. These were, rather, the increase in social allowances (unemployment), the stickiness of civil servants’ salaries (when the level of consumer prices was decreasing), and international income transfers to the colonial empire. Together these outweighed the decline in government interest payments since 1931, due to debt repayment and changes in borrowing techniques.15 The balanced-budget doctrine carried full weight in 1933, when the government tried to reduce the number of civil servants, to increase the income tax, to tax fuel, etc. After the fall of several governments, a restrictive fiscal policy was implemented: an increase of 10 per cent in income tax, a special additional tax on the income of civil servants, and the ‘contrivance’ of a new tax levied on the excess of the 1933 nominal income over the average of the 1931 and 1932 nominal incomes.16 This designedly restrictive fiscal policy did not balance the budget ex post as expected, but it did succeed in interrupting growth, contrary to what was happening in other countries.17 In conclusion, fiscal policy was inspired by the balanced-budget doctrine and was weakly related to the gold standard. Only Poincaré and his immediate successors can be supposed to have acknowledged a relationship between public debt and capital flows. The tendency of orthodox fiscal policy was countercyclical before the onset of the Depression and procyclical thereafter; however in the Depression the results differed from their intentions due to the action of automatic stabilizers.

72 France Table 3.3 per cent)

Data on fiscal policy (all levels of government) (as share of GDP, in

1928

1929

1930

1931

1932

1933

12.5

10.2

11.2

15.3

18.1

19.4

2.9

3.0

3.6

4.7

5.5

5.1

of which: social and 5.2 unemployment benefits

3.7

4.8

5.7

6.9

7.7

of which: net foreign –0.4 transfers (e.g. colonial empire)

–0.6

–1.9

–0.6

–0.4

0.8

Expenditures (excluding interest on debt) of which: consumption and investment

Interest on debt

4.0

3.8

3.2

3.3

3.9

3.8

Budget surplus

–0.2

1.9

1.7

–1.0

–2.4

–4.7

Taxes and social contributions

16.3

16.0

16.1

17.6

19.6

18.6

Public debt in francs

81

70

72

78

87

97

Apparent average interest rate on debt (per cent)

4.8

5.3

4.4

4.3

4.3

4.3

Source: Villa (1994). Central government, social security and local authorities are included.

5

Monetary policy and the gold standard

According to Eichengreen, the gold standard has a deflationary bias. Countries in recession cannot carry out expansionary monetary policies because these bring them up against capital outflows and thus against vanishing gold reserves. Countries which implement restrictive monetary policies benefit from capital inflows, which can even reverse their current account deficit and cause them to accumulate gold. Their competitors are thus constrained to implement restrictive policies of a ‘follow the leader’ type, even when these policies do not fit their own activity levels. The system promotes restrictive policies and systematically prevents any country from escaping from depression through policy measures that are not coordinated with those of their trading partners – unless it first abandons gold convertibility or devalues when the gold cover ratios are breached.18 By accumulating gold, France was thus partly responsible for triggering the Depression, deepening it, and breaking the gold standard. From this point of view, Foreman-Peck, Hughes Hallet and Ma (1992) try to

Pierre Villa 73

show, by estimating a monthly model for France, Germany, the UK and the US over the 1929–33 period, that a cooperative monetary policy between the four countries could have limited the Depression and avoided uncoordinated beggar-thy-neighbour devaluations. These analyses, which to some extent also emphasize the influence of the Depression on French policy, are questionable from several points of view: In the first place, they assume that in a gold-standard regime, French governments had the power to control the money supply quantitatively, by sterilizing foreign-currency and gold reserves through open-market policies. However the gold-standard regime was functioning in France as a fixed exchange-rate regime with imperfect mobility and substitutability of private assets. For a given interest rate, the sum of the private and public gold inflows and of the public foreign-currency reserves was thus related to the private foreign assets’ movements and the level of the current account. So at a constant interest rate, a pure open-market policy (i.e. a repurchase of public assets) could not reduce the gold and foreign currency counterpart of the money stock. It could only increase the money stock or decrease the refinancing of banks by the central bank. The only three ways to fight gold inflows are the following: (1) reducing the interest rate to cause the private sector to hold foreign-currency assets as substitutes for domestic franc assets or for money; (2) buying foreign assets directly on the foreign exchange market for gold; (3) appreciating the exchange rate of the currency in gold. The Bank of France refused these policies because they threw into question the external value of the franc, and it feared above all a return to instability on the foreign-exchange market. So it favoured pegging and stabilizing the discount rate (see Mouré, 1991). Secondly, the consequences of a monetary policy depend on the macroeconomic regime currently obtaining. In 1929 and 1930 France was at full employment. So any decrease in the interest rate would have increased price levels at a time when short- and long-term real interest rates were negative. From 1931 onwards the French economy was in a Keynesian unemployment regime with budget deficits. But so great was the mistrust of financial markets – which believed in the balanced-budget doctrine – that a great part of the budget deficits was financed by money or short-term bonds, and long-term real interest rates rose again, strengthening the restrictiveness of monetary policy. The following sections aim to describe French monetary conditions over the period, and to evaluate the degree of restrictiveness of monetary policy, and its responsibility for the Depression. Monetary policy from 1928 until the end of 1930 Gold and foreign-currency reserves accumulated from the end of 1926 to May 1928 as a consequence of current-account surpluses (Ff 22.4 billion)

74 France

induced by the restrictive fiscal policy, and of the capital inflows (Ff 13.6 billion). The total (Ff 36 billion) was used to buy French private assets or was held in French money (see Table 3.4). In both cases this induced an increase in the proportion of gold and foreign reserves in the assets-side counterpart of the money stock. However, where the inflowing capital was invested in French assets (other than money) it also reduced the credit demand of firms, by facilitating security financing. The substitution of marketable securities for bank credit increased the gold/money ratio mechanically, without having to invoke any credit-multiplier theory.19 In 1929 gold and foreign reserves were at a standstill because the demand for foreign assets increased again when foreign interest rates rose relative to French and because the current balance deteriorated after the real appreciation of the franc. However, in 1930 and until the sterling depreciation in September 1931, gold again flowed in through the repurchase by the Bank of France of foreign assets held by French speculators. The Hoover moratorium, and the fear of sterling and dollar devaluations, induced a dramatic inflow of previous capital outflows (Sicsic and Villeneuve, 1993). Banking credits increased moderately during the last two years of high activity (1929 and 1930). Firms preferred to finance investment from their high profits than from bank loans (see Tables 3.3 and 3.4). Bond issues were stimulated by low real interest rates and by the fact that budget surpluses restricted government bond issuing. This theory of the substitution of gold and foreign assets for public and private credit claims in the structure of the assets-side counterparts of the money demand demonstrates how monetary policy has to manage both the cost of finance, and the reconciliation of the discrepancy between the structure of the wealth desired by private agents and the structure of private and public indebtedness. In France the policy followed to perform these tasks was basically one of pegging the discount rate. This policy was not restrictive when viewed in relation to short- and long-term real interest rates – which were always lower than the American and British except for 1928 (see Table 3.5 and Figure 3.3). The apparent restrictiveness of monetary policy in 1928 reflects the reluctance of the Bank of France to decrease the discount rate in line with the rate of inflation. However it was not as restrictive as it seems, since the banks’ refinancing needs were limited (Table 4.4), because the substitution of French assets for foreign in private wealthholding reduced the credit demands on the banks, and because the budget surplus reduced the issue of new public debt. The smallness of the Bank’s purchases of discountable paper was not due to its refusal to refinance commercial banks. On the contrary, in 1928 the interbank offer rate was lower than the discount rate. A discount-rate constraint, if any, was not binding. When it began to bind at the end of 1929 and beginning of 1930, the Bank lowered its discount rate, so that the real short-term interest rate became negative (see Figure 3.3 and Table 3.5).

Table 3.4

1

The assets held as counterparts of the money supply (M2) (Ff billion)

M3 M21 of which: central bank money (including postal deposits)1 Gold and foreign reserves counterpart of M22 Government financing counterpart of M23 Banks loans counterpart of M21 of which: refinancing of banks by the central bank4 Variation of private reserves (gold and foreign currencies)5 Variation of public reserves (gold and foreign currencies)5 Current account6 French private bonds and shares issues7 Public debt in French currency1 Nominal GDP1

1926

1927

1928

1929

1930

1931

1932

1933

122 107

141 120

165 138

177 145

192 154

213 162

216 159

210 151

56

60

65

70

78

88

87

87

10

33

64

67

79

91

88

78

59 38

45 42

12 62

11 67

9 66

16 55

26 45

25 48

6.5

3.9

4.6

8.4

8.4

6.9

3.2

4.1

13.1

–8.6

–5.0

4.8

–6.8

–27.2

–2.4

0.7

–1.5 10.4

15.2 5.8

13.2 8.7

3.4 8.9

7.1 2.0

12.1 –4.0

–0.8 –1.6

–2.7 –3.0

8.6 290 331

13.6 299 343

20.2 289 356

27.1 282 400

25.1 282 392

14.2 284 366

16.2 276 316

10.2 304 313

75

Sources: 1 Villa (1994), money stock held by residents, excluding Treasury deposits at the post offices. 2 Bulletin de la SGF (1901–1931) and (1929–1939), including appreciation of gold and foreign currencies. According to the law of 1928, the Gold/Central Money ratio should not be lower than 35 per cent. It is obvious from the computations reported in this table that, from 1927 onwards, there was room for an intervention on the foreign exchange market, because reserves exceeded the ratio. But this impression stems from the view of the Gold Exchange Standard, because foreign-currency reserves are included in the counterparts. 3 Patal–Lutfalla (1990), appendix, but excluding Treasury deposits at the post offices. 4 Bank of France balance sheets. 5 Rist–Schwob (1939) and Sicsic–Villeneuve (1993). 6 Rist–Schwob. 7 Marnata.

76 France Table 3.5

Apparent average real interest rates (per cent)

Short term (discount rates) France UK US Long term (bond and consol rates) France UK US

1928

1929

1930

1931

1932

1933

3.34 5.62 2.94

–0.65 5.84 5.31

–0.73 3.77 5.52

5.05 6.57 11.30

9.25 6.41 13.00

5.77 3.40 4.66

5.14 5.59 2.89

0.74 4.94 5.09

0.36 4.86 7.09

6.64 6.93 13.80

11.48 7.34 16.14

9.00 4.78 7.86

Sources: Annual statistics of the different countries and Villa (1994) for French prices.

Thus monetary policy consisted in fixing the nominal discount rate whilst inflation fluctuated considerably, and indeed reflected, with a lag of about one year, the fluctuations of output (see Table 3.2). Even if there had been a stabilization policy pursuing an internal target, (which was not the case) it would have been difficult to formulate the corresponding interestrate policy. Two criticisms have been made of this policy. First, that the Bank of France did not accommodate the increase in money demand which followed the Poincaré stabilization. If this were true, an increase in the nominal long-term interest rate should have been observed, since this rate governs the choice between money and financial assets. But the opposite occurred. Therefore, gold accumulation and its rising importance among the assets held against the money stock resulted only from capital inflows and the consequential decline in the demand for bank credit. According to the second criticism the Bank should have used an active open-market policy.20 Eichengreen thinks that this was precluded both by the regulations governing the Bank’s operations and by its failure to create a market for government bonds by selling them in 1928 at a moment of increasing gold reserves.21 But in fact open-market operations were possible, since the Bank used to buy second-hand treasury bills and many government securities through the discount market. However a repurchase of these securities could have had three different effects. The first case would be where, with the discount rate held constant, these bills and securities were sold by the banks. In order to balance their accounts the banks would then have had to decrease their refinancing at the Bank of France; the share of discountable papers in their asset structure would have risen and of government bonds fallen, but there would have been no impact on the money stock at the M2 level. The second case would be where the securities were sold by households, and here in turn two configurations must be considered. If households had increased their money demand by the same

Pierre Villa 77

amount as the repurchase of securities, their money stock would have increased without changing the gold stock, but the government-credit element in the Bank’s reserves against the note issue would have increased by the same amount as the money stock. But if households had not wanted to increase their money demand, the interest rate would have had to fall, in order partly to make them accept the extra money and partly buy French and foreign assets.22 In this case, the increase in foreign assets corresponds to the capital flight induced by the interest-rate discrepancy with the foreign interest rate and/or by the expectation of exchange-rate depreciation.23 An open-market operation under the gold standard is also equivalent to buying and selling foreign bonds directly on the foreignexchange market. The unwillingness of the Bank of France to implement an expansionary open-market policy was thus the counterpart of its unwillingness to lower interest rates.24 The gold inflows at the end of the 1920s did not originate in some special functioning of the gold standard in France. The Bank could not unilaterally implement an interest-rate-reducing policy, firstly because the economy was at full employment and, with inflation fluctuating but persisting, the real long-term interest rate was near zero; and secondly because it was afraid of triggering a confidence crisis and an unstoppable capital outflow. Only a cooperative monetary policy combining an appreciation of the franc in relation to gold with an expansionary policy in the US and UK could have solved the problem. But such a policy could only have been contemplated if it had been recognized that France was not in the same economic regime as the two other countries. Moreover, for the success of such a policy, the positive spillovers through international trade would need to have been large. But this condition was not satisfied because, at the same time, countries were increasing their tariff barriers and quotas.25 Monetary policy during the Depression (1931–3) From 1931 onwards France was in a Keynesian unemployment regime, with falling domestic and foreign demand, and falling investment. Restrictive monetary policy no doubt played a great part in deepening and protracting the Depression. This can be measured by the increase in the short-term real interest rate and the unwillingness of the Bank of France to reduce the nominal discount rate. To this may be added the increase in the spread between the long-term and short-term real interest rates (see Table 3.5). Why were long-term real interest rates so high? Several explanations are generally put forward. The first invokes an external reason. After the sterling devaluation, private agents were anticipating a depreciation of the franc between 1932 and 1933. This idea is supported by the measures of exchange-rate expectations computed by Haucoeur (1993), who compares the price of the 4 per cent

78 9

France The US The UK

8.5 8 7.5 7 6.5 6 5.5 5 4.5 4 3.5 3

1926

Figure 3.3(a)

1927

1928

1929

1930

1931

1932

1933

Nominal long-term interest rates in France, 1928–34

20 15 10 5 0 –5 –10 France The US The UK

–15 –20 1926 Figure 3.3(b)

1927

1928

1929

1930

1931

1932

Real long-term interest rates in France, 1928–34

1933

79 7

France The US The UK

6.5 6 5.5 5 4.5 4 3.5 3 2.5 2

1926

Figure 3.3(c)

1927

1928

1929

1930

1931

1932

1933

Nominal short-term interest rates in France, 1928–34

20 15 10 5 0 –5 –10 France The US The UK

–15 –20 1926 Figure 3.3(d)

1927

1928

1929

1930

1931

1932

Real short-term interest rates in France, 1928–34

1933

80 France

consols indexed on the exchange rate of the pound sterling (emprunt Caillaux), and the price of the non-indexed 4 per cent consols. However a measure of expectations based on the forward exchange rate does not confirm this result.26 Three other ‘internal’ explanations have been put forward to explain the spread between long-term and short-term interest rates: 1 Gold hoarding: by withdrawing resources from savings, this would have produced a lack of liquidity, which could have been resolved only by an increase in the interest rate. However this explanation, which is a species of the Keynesian liquidity trap, seems to be exaggerated, since gold hoarding by the private sector was small in relation to all moneyholding (M2 or M3).27 2 The credibility crisis of the government in the eyes of the financial markets and the private sector. Since 1931 public budgets were in deficit despite governments’ intentions of restoring balance. The Treasury had difficulties in selling public debt in this period, as especially demonstrated in two episodes: • the bond conversion of September 1931. The government was only able to convert the old bonds into 4.5 per cent consols instead of 4 per cent as expected. This credibility crisis centred upon the financial markets, for private banks and the Caisse des Dépôts28 held 55 per cent of the bonds.29 • the treasury bill issue of May 1932, which, in the view of the financial markets, confirmed that the government was abandoning its balanced-budget target. A risk premium was instantly added to the long-term interest rate30 (see Haucoeur, 1993). This crisis of confidence in the state signified that small investors and banks would not buy long-term government bonds except at high interest rates; otherwise they preferred savings deposits (M3 minus M2 in Table 3.2) or central money (see Figure 3.4). At the same time they turned away from shares, the yields of which were smaller than those of fixed-interest assets (see Table 3.2 and Sauvy 1984, vol.I, p. 127). This shift was accentuated by the fact that the government raised the ceiling on savings deposits, and that its rate of interest on these deposits was higher than the discount rate.31 Thus in a sense the Bank of France implemented a more restrictive monetary policy than the discount rate indicates. However as the Caisse des Dépôts was in charge of managing the savings bank, it used to invest a large part of its resources in bonds. It could thus, by buying or selling longterm and treasury bonds, have reduced the long-term interest rate. Since this reduction is not observed. the increase in the long-term interest rate must be explained by a third hypothesis. 3 Conflicts between the monetary authorities and the state. These developed in three special circumstances:

Pierre Villa 81

• In September 1932, during the bond and consol conversion, the Bank of France refused the government a monetary advance to fund the operation. • In 1933 the Bank of France refused to discount treasury bills unless they corresponded to real operations.32 • By permanently fixing a ceiling on the discount (repurchase) of treasury bills, the Bank of France constricted the short-run financing of government and forced it to ask the Caisse des Dépots to take part in these operations. It did so from 1932 onwards. In 1933 the Treasury had to borrow from the English banks. This made the long-term interest rate climb again.33 Would it have been possible to implement an expansionary policy? Certainly, during a period of mistrust of government bonds (for fear of taxation and inflation), central-bank purchases of treasury bills or of bons de la défense nationale (a sort of treasury bill) could have allowed the government to obtain finance at a lower interest rate. Moreover, this could have allowed the banks to offer a lower interest rate to firms, which would have favoured private investment. However, the most efficient measure would have consisted of lowering short-run interest rates34 in order to reduce both the banks’ interest rate and the long-term interest rate. But such a policy would have been frustrated by two doctrines: the balanced-budget doctrine and the ‘real bills’ doctrine. Thus in 1932 the Bank of France refused to reduce the discount rate on the pretext that this would have allowed firms to maintain production levels at a time when the outlets for trade were limited. It is impossible to be more anti-Keynesian at a time of demand weakness. However, the consequence of this chilly monetary management were not wholly bad, for no banking crisis occurred in France. There was incipient private-sector mistrust toward banks in 1930, with deposit transfers within the banking system, from small banks to larger. This was followed by substitution of high-powered money (banknotes) and of savings deposits for bank deposits, as of 1931 (see Figure 3.4). Nevertheless, no banking crisis took place. Bank bankruptcies were few and were redeemed without loss to the depositors. More especially, the banking system as a whole was always very liquid in France. Deposits always exceed credits (see Table 3.4), and the discount of commercial bills by the Bank of France (discount window) was small. Finally, banks invested their excessive liquidity in public longterm and treasury bonds and gold, rather than in shares and bonds of the private sector, thus avoiding large losses from the fall in stock-exchange prices (see Tables 3.2 and 3.4). In fact their liquidity, measured by the loans/deposits ratio, improved from 1930 to 1933.35 This improvement did not reflect a tightening in the supply conditions for credit due to a banking crisis having broken the information structures that allow lenders to select

82 France 90 85 80

Central bank money Bank deposits Savings deposits

75 70

FF Billion

65 60 55 50 45 40 35 30 25 20 1928 Figure 3.4

1929

1930

1931 Years

1932

1933

Structure of money demand (M3) in France, 1928–34

borrowers (cf. Bernanke, 1983) – because bank failures were the exception. The improvement of the loans/deposits ratio came basically from the decrease in the demand for bank credits, which was related to the fall in investment and the excessively high real interest rates. Thus only an interest-rate-reducing policy could have had an impact on investment.36 Moreover, the safety of deposits and of their income allowed consumption to play a countercyclical part, as opposed to what was observed in the US (Romer, 1990). Money was the main form of savings, and the rise in the value of the real money stock increased the propensity to consume,37 because it reduced the rate of saving previously devoted to rebuilding the money stock that the inflation had eroded. Thus monetary collapse did not magnify the Depression as in the case of the US; nevertheless monetary policy was procyclical because it increased real short-term interest rates. Moreover, the spread between long- and short-term interest rates increased from 1931 onwards because the financial sector mistrusted the government for not succeeding in balancing the budget. This was procyclical. Yet, by succeeding in balancing the budget, the government would have deepened the depression. In effect, the government had to manage a credibility-deflation trade-off. The high real interest rates and avoidance of a banking crisis explain why the gold

Pierre Villa 83

standard did not quickly collapse in France (i.e. after 1932), despite an overvalued exchange rate – an explanation on which all economists agree.

6

Conclusion

During the interwar period, and in particular between 1926 and 1933, economic policy in France was not managed in order to achieve certain internal targets, but rather made to conform to two doctrines: the ‘balanced-budget’ doctrine, and the ‘stabilization-of-credit-conditions’ doctrine. The aim of the latter was basically to maintain a constant nominal cost of credit and to provide liquidity to the banking sector and to firms. The Bank said that it had to fight against ‘managed money’. This assertion can be understood as a combination of the ‘real bills doctrine’, which advocates discounting only such bills as correspond to real transactions, and the Wicksellian theory of ‘the credit economy’, according to which the ‘natural rate of interest’ – i.e. the long-term interest rate – is determined by the short-term rate plus a risk premium for inflation, exchange-rate depreciation and illiquidity. After the exchange rate crisis of 1926, the economic authorities did not want to diverge from these two principles. Until the (late) entry of France into the Depression in 1931, the two doctrines expressed themselves in a restrictive fiscal policy and a neutral monetary policy – which was actually rather expansionist if one looks at real interest rates. So, ex post and de facto, the chosen policy-mix proved effective in fighting inflation and boosting growth, because it induced an appreciation of the real exchange rate (which had been at a low level in 1926), and because it allowed labour costs to decrease and the purchasing power of wages to increase. But this policy was unconscious, so that the doctrines did not change after the descent into the Depression in 1931. The uncertainty about the gold value of the franc and the public debt (even though the gold/central money ratios were far higher than the 35 per cent officially required) expressed itself in the risk premium on the long-term interest rate. At the same time, the budget deficits being unintentional, the government constantly tried to reduce them, and this sent the wrong signals to the private sector about the depressed aggregate demand for goods. Thus it was not some special malfunction of the gold standard in France which induced the gold inflow, exporting depression abroad. Rather it was a misunderstanding of the efficiency of fiscal compared with monetary policy. Two lessons can be drawn from study of this period. The first is that coordination might have improved the situation. But, for this, it would need to have been recognized that until the end of 1930 France was at full employment, unlike the US and the UK; and that exchange-rate and fiscal policy would have been more efficient than monetary policy in rectifying the imbalances between them; whereas monetary policy should be targeted

84 France

on the real interest rate (GDP growth) as adjusted for the public debt tradeoff. The second lesson is that coordination of economic policies requires some freedom. Before 1931 France would have to have implemented a restrictive fiscal policy and a rather neutral monetary policy, while the US and UK would have to have launched expansionist fiscal and monetary policies. The opposite was true during the 1931–3 period, when France should have adopted an expansionary monetary policy. But these coordinated policies are difficult to implement when they are constrained by prudential ratios like the gold standard, and gold-to-money ratios – or, equally, by the fiscal budget-deficit and the debt-to-GDP ratios of the Maastricht treaty. Prudential ratios and their consequences for credibility are inconsistent with coordination. Notes 1 See the degree of capacity utilization in Table 3.2. The capacity of production is measured by means of a Cobb-Douglas production function estimated from gross fixed capital in equipment and total working population including the unemployed and the military. Working population was nearly constant in France between the wars. The average working week in private firms was 48.45 hours in 1929, 47.33 in 1930 and 45.48 in 1931 (Villa, 1994). The official working week was 48 hours. 2 In 1930 unemployment amounted to 260,000, compared with 248,000 in 1926 – much lower than the 438,000 during the 1927 recession following the restrictive policy of Poincaré. 3 A measure of quantitative protectionism has been computed for France’s four main competitors: Belgium, Germany, the UK and the US. The resultant index is 87 in 1926, 100 in 1928, and 124 in 1930. 4 Sauvy (1984), vol.I, pp. 81–3. 5 Eichengreen and Wyplosz (1988), tables 3 and 4. 6 An example of the fact that share prices had small effects on investment decisions is the case of electrical firms. They distributed large dividends to shareholders, who reinvested them in the firms in order not to disseminate shares among a wider public. In this case, new shares are equivalent to invested profits and there is no interaction between investment and market prices. 7 We measure unemployment from the census and statistics of the employment bureaux, excluding individual entrepreneurs and the self-employed. Even on this restricted definition, unemployment rose 226,000 during the first quarter of 1931. A discussion of the low level of unemployment in France can be found in Salais (1988). We do not think that a correction for the self-employed would change anything qualitatively in the evolution of the matter: unemployment appeared suddenly. 8 The decrease of the real interest rate explains the boost in housing investment, and the increase in profitability the surge of physical investment. 9 The long-term interest rate upturn came at the end of 1931, partly from the devaluation of the pound (expectation of a depreciation of the franc), and partly from the mistrust of French public debt on the part of French financial markets. During October 1931 the rate on bonds rose rapidly from 3.48 per cent to 3.88 per cent. 10 Wage-earner employment in the private sector fell from 10.3m in 1930 to 9.85m in 1931, and unemployment increased from 260,000 to 487,000.

Pierre Villa 85 11 In 1928 the budget surplus vanished as a consequence of social security expenditures and large unemployment benefits. The restrictive fiscal policy had induced a recession in 1927 and increased unemployment. With the return to full employment in 1928, the budget surplus reappeared in 1929. 12 Contrary to Eichengreen’s statement (1992, p. 255), fiscal policy remained restrictive in 1929 and 1930, even though public investment increased. 13 Budgetary discipline, whether prescribed by the gold standard or by Maastricht’s ‘prudential ratios’, is extremely efficient at full employment. Yet is this the case in a Keynesian unemployment regime? 14 In 1931, the ‘Steeg’ (Ff 0.7 bn) and ‘Laval’ programmes (Ff 2.7 bn), and subsequently, ‘Marquet’ programmes in 1934 (Ff 2.5 bn) and 1935 (Ff 1.3 bn). 15 The government had difficulty in issuing a new long-term debt at low interest rates because of mistrust on the part of the financial markets. So it issued money and short-term bonds, which lowered the apparent interest rate (see next paragraph and Table 3.3). 16 This curious tax was motivated by the idea that only the increase of nominal income had to be taxed in a period of falling prices. 17 Sauvy (1984 ,vol I, ch.10) does not understand why French growth failed in 1933, but he never mentions the fiscal burden and the overvaluation of the franc. 18 See Eichengreen (1992, pp. 274–86), and Temin (1993). 19 Other things being equal, the mechanism is the following: the French exchanged their foreign assets for gold, then gold for money to buy domestic assets denominated in francs. The firms issued shares and bonds because the decrease in the interest rates made them a more attractive means of financing. Firms then reduced their demand for bank credit. The demand for money did not change and, as regards supply, the gold and foreign reserves counterpart was substituted for commercial credits. The increase in the gold and reserves counterpart was exactly equal to the decrease in the private-sector holdings in foreign currencies. 20 Eichengreen (1992, p. 197). 21 Eichengreen (1992, p. 223). 22 If they bought only French assets, this would have lowered firms’ demands for bank credit, and have had no impact on gold. 23 There were no exchange controls, and few capital controls: the French could not lend in foreign currencies to foreigners, but could buy foreign assets. 24 The Bank of France cited the trade-balance deficit to explain that it did not violate gold-standard rules (Eichengreen, 1992, pp. 254–5). 25 The index of trade quotas (see note 3) moves from 100 in 1928 to 126 for France and from 100 to 124 for its four main competitors. 26 According to this last method, there had been an expectation of a franc appreciation of 1.7 per cent on average in 1931, of 0.3 per cent on average in 1932, and of a depreciation of 1.1 per cent in 1933. 27 The variation of gold holdings by the French after 1928 was not more than 2 per cent of the 1933 M2 money stock, using either Sicsic and Villeneuve (1993) or Villa (1993a) as sources. Explaining the increase in the interest rate by gold hoarding was more an intuition of the time than a statistical result. 28 The Caisse des Dépôts et Consignation still exists as an institution which manages savings (deposits). It operates like a commercial (savings) bank, issuing loans and purchasing bonds, discounting bills and government securities. 29 On this subject it is worth consulting Sauvy (1984, vol.I, p. 122). In statements reported at the time, the aim was ‘not to rob small investors’.

86 France 30 Average interest rates on long-term government stocks increased from 4.67 per cent in April 1932 to 4.97 per cent in May 1932 and to 5.12 per cent in June 1932 (Villa, 1994). 31 Ceilings on savings deposits increased from Ff 2,000 to 20,000 for individuals, and were set at 100,000 for firms in April 1931. The interest rates fluctuated between 3.25 per cent and 3.5 per cent, with a discount rate of 2.5 per cent. 32 Should this behaviour be seen as a re-emergence of the real bills doctrine, or perhaps rather as a desire not to finance public deficits, because they could be ‘bad money’? 33 See Haucoeur (1993, p. 110f). 34 The discount rate and the rate on savings deposits. 35 The credits/deposits ratio of the four main commercial banks goes from 0.90 in 1930 to 0.78 in 1933. In 1931, following the failure of the Adam bank (the Loustric case), the consolidated balance sheet of these four banks is as follows (Rist, 1937): Liabilities: deposits, Ff 38 bn; refinancing less reserves, Ff 3 bn. Assets: gold, Ff 13 bn; private and public discounted bills, Ff 18 bn; loans to firms, Ff 10 bn. It would be difficult to go bankrupt with such a balance sheet! 36 Econometric estimates (Villa, 1993b) show that the decrease in aggregate demand almost completely explains the fall of firms’ investment. The remainder is due to the decrease in profitability, that is, the ratio profits/user cost of capital. 37 See Villa (1996) for estimates of the consumption function over this period.

References Bernanke, B.S. (1983) ‘Non-Monetary Effects of the Financial Crisis in the Propagation of the Great Depression’, American Economic Review, 73 (3): 257–76. Boyer, R. (1979) ‘La crise actuelle; une mise en perspective historique’, Critiques de l’économie politique (April–September). Eichengreen, B. (1992) Golden Fetters. The Gold Standard and the Great Depression 1919–1939 (New York: Oxford University Press). ——– and C. Wyplosz (1988) ‘The Economic Consequences of the Franc Poincare’, in E. Helpman, E. Razin and A. Sadka (eds) Economic Effects of the Government Budget (Cambridge, MA: MIT Press): 257–86. Foreman-Peck J., A. Hughes Hallett and Y. Ma (1992) ‘The Transmission of the Great Depression in the United States, Britain, France and Germany’, European Economic Review, 36: 685–94. Haucoeur, P.C. (1993) ‘Surévaluation ou crise de confiance, Hausse des taux d’intérêt et durée de la grande dépression en France’, in M. Levy-Leboyer, A. Plessis and M. Aglietta (eds), Du franc Poincaré à l’écu (Paris: Comité pour l’histoire économique et financière de la France): 97–123. Marnata, F. (1973) La bourse et le financement des investissements (Paris: Armand Colin). Marseille, J. (1980) ‘Les origines inopportunes de la crise de 1929 en France’, Revue économique, 31 (4): 648–84. Mouré, K. (1991) Managing the Franc Poincaré. Economic Understanding and Political Constraint in French Monetary Policy 1928–1936 (Cambridge: Cambridge University Press). Patat, J.P and M. Lutfalla (1990) A Monetary History of France in the Twentieth Century (New York: St Martins Press). Rist, C. (1937) L’évolution de l’économie française (1910–1937) (Paris: Sirey).

Pierre Villa 87 Rist, C. and P. Scwob (1939) ‘Balance des paiements’, Revue d’economic politique 53 (1), special issue: 528–50. Romer, C. (1990) ‘The Great Crash and the Great Depression’, Quarterly Journal of Economics, CV: 597–624. Salais, R. (1988) ‘Why Was Unemployment So Low in France during the 1930s?’, in B. Eichengreen, and T.J. Hatton (eds), Interwar Unemployment in International Perspective (Boston: Kluwer Academic Publishers): 247–88. Saly, P. (1980) ‘La politique française des grands travaix 1929–1939: fut-elle keynésienne?’, Revue économique, 31 (4): 706–42. Sauvy, A. (1984) Histoire économique de la France entre des deux guerres, vols 1 & 3 (Paris: Economica). Sicsic, P. and B. Villeneuve (1993) ‘L’afflux d’or en France de 1928 à 1934’, in M. Levy-Leboyer, A. Plessis and M. Aglietta (eds), Du franc Poincaré à l’écu (Paris: Comité pour l’histoire économique et financière de la France): 21–55. Temin, P. (1976) Did Monetary Forces Cause the Great Depression? (New York: Norton). ——– (1993) ‘Transmission of the Great Depression’, Journal of Economic Perspectives, 7: 87–102. Villa, P. (1993a) Une analyse macroéconomique de la France au XXème siècle (Paris: Presses du CRNS). ——– (1993b) ‘Productivité et accumulation du capital en France depuis 1896’, Observation et Diagnostics Economiques, no.47 (October): 161–200. ——– (1994) Une siècle de données macroéconomiques (Paris: Insee, Economie générale, No.86–87). ——– (1996) ‘La fonction de consommation sur longue période en France’, Revue économique, 47 (1): 111–42.

4 Slump and Recovery: The UK Experience Michael Kitson

1

Introduction

The world economy was in considerable disarray following the end of the First World War. Initially the international trading and payments system was dominated by flexible exchange rates but from the mid-1920s the cornerstone of international economic management was a reconstructed form of the gold standard, which the UK joined in April 1925. This chapter argues that the interwar gold standard encouraged beggar-my-neighbour deflation and after 1929 it became a vehicle for transmitting recession. In the UK the Depression was severe in absolute terms causing a massive increase in unemployment and poverty. But compared to other countries the UK Depression was relatively mild, reflecting the importance of national specific factors in both transmitting and ameliorating external shocks. The policy response to Depression was particularly important: those countries, such as the UK, that withdrew from the strictures of the gold standard early had the potential to implement expansionist policies that encouraged recovery and growth in the 1930s.

2

The Depression: international mechanisms

The cornerstone of international economic management from the mid1920s was the gold standard. The interwar variant was founded on the questionable success of its predecessor, the classical gold standard, which was in operation during the quarter of a century before the First World War. The reconstructed gold standard of the 1920s was intended to bring stability into international trading relations and increase world prosperity. It failed to achieve these objectives. The success of the system depended on monetary adjustment, so that real exchange rates converged, but its actual effect was to depress real variables such as output and employment and undermine the capacity of individual governments to deal with domestic economic problems. 88

Michael Kitson 89

To be effective the gold standard required a process of automatic adjustment, the classical price-specie-flow mechanism, to correct payments imbalances. Under this system the price level would adjust in response to deficits and surpluses on the balance of payments. A deficit would lead to a loss of gold and a contraction in the money supply, leading to fall in prices and the eradication of the deficit. Similarly, a surplus would lead to an accumulation of gold, a rise in the money supply and prices and balance of payments equilibrium. Unfortunately, this approach is based on a number of assumptions that are not grounded in economic reality. First, it assumes the ‘law of one price’ – that the process of global competition will ensure that a basket of goods will cost the same in all countries. In reality, imperfectly competitive markets, imperfect information and so on ensured significant and protracted deviations from the law of one price. Second, it assumes that the demand for money is stable – this is a source of much controversy between monetarists and Keynesians – but most empirical evidence suggests that the demand for money is not stable (see Goldfeld, 1973). Third, it assumes that the monetary authorities do not intervene (‘sterilize’) to prevent increased (decreased) gold reserves adding to (reducing) the money supply. Fourth, it assumes that the burden of adjustment would be borne by prices and not by quantities. This is the pre-Keynesian assumption that the economy tends towards full employment – an assumption that is so obviously inappropriate to the interwar period. The inappropriateness of the key assumptions which underpinned the gold standard led to a number of systemic problems in the system. First, it combined together countries with different initial conditions and different economic structures in a fixed exchange rate system. Second, the system operated asymmetrically – balance-of-payments imbalances had to be accommodated or rectified by weaker countries that tended to generate deficits (the accumulation of surpluses is not a problem that has to be rectified). Third, the system could not accommodate shocks; on the contrary it operated to amplify the slump. The initial conditions The adjustment process integral to the gold standard created a severe deflationary bias for the world economy. To capture this bias the main trading countries can be broadly classified (after Cripps, 1978) into those constrained, and those unconstrained, by their trade performance. Those countries that could maintain a sufficient level of exports, relative to imports, at a high level of economic activity were not balance-of-payments constrained. Such countries could pursue full employment strategies by regulating domestic demand or could accumulate increased reserves. Conversely, those countries that could not maintain balance-of-payments equilibrium at a high level of economic activity had to reduce domestic demand in order to import only those goods and services which they could afford to finance.

90 United Kingdom

The two key unconstrained countries in the 1920s were France and the United States. The two major constrained countries in the 1920s were Britain and Germany, which both emerged from the aftermath of war with severe economic problems. The reconstructed gold standard, therefore, created a fixed-exchange-rate regime with members at different stages of economic development, with different economic structures and different economic problems. The deflationary bias As the main trading nations entered the exchange-rate system with different initial conditions it was apparent that the strength of the adjustment process would be central to the regime’s effectiveness. The option of adjusting the nominal exchange rate was effectively precluded. The adjustment of the real exchange was slow and erratic. For the UK, most studies indicate a significant overvaluation of the sterling effective exchange rate in the 1920s: with the overvaluation being greatest in the early 1920s, following the announcement of the intention to return to gold, and with only a slow movement towards purchasing power parity throughout the rest of the decade (Redmond, 1984). In fact Keynes’s (1925 [1972]) contemporary estimate of a 10 per cent overvaluation has proved, although based on limited data, to be a reasonable approximation of more recent empirical estimates. As the downward pressure on wages and prices was protracted due to established price- and wage-setting behaviour the result was slow growth and higher unemployment. Conventionally, it has been assumed that the unravelling of the price–quantity adjustment process would eventually return the economy to its previous position with only a temporary loss of output and jobs. However, the legacy of slow growth lowered the long-run capacity of the economy due to its impact on physical and human capital accumulation and the permanent loss of export and importsensitive markets. The slow adjustment of the real exchange rate – the failure of the ‘law of one price’ – left two alternative means of adjustment: first, changes in the level of demand – deflation in constrained countries and reflation in surplus countries; second, the financing of the deficits of constrained countries by capital flows from the unconstrained countries. In fact the ultimate burden of adjustment was borne by domestic deflation as the surplus countries were reluctant to reflate (and there was no effective disciplining device requiring them to do so). As noted above, the classical adjustment mechanism assumes that gold flows will provide the means of changing the level of demand with the impact falling on prices. But price adjustment was slow and the reflationary impact of gold flows into France and the United States was negated by domestic monetary policy. Both countries, which by the late 1920s had accumulated 60 per cent of total gold reserves, prevented these reserves from boosting their domestic money supplies (the issue of

Michael Kitson 91

sterilization raised earlier). American policy-makers were increasingly concerned with curbing stock market speculation whereas the French were wary of inflation. The prioritization of domestic economic issues transmitted deflation abroad. Low import demand, particularly in America, led to widening balance-of-payments deficits in many of the key European economies. The growth of world trade and, through the trade multiplier, world income was therefore limited by the domestic policies of the unconstrained countries. Whereas these nations could choose whether to reflate or pursue domestic policy concerns the constrained countries had no such option. The entire burden of adjustment fell on them – they could either deflate to eradicate balance-of-payments deficits or borrow, at least in the short term, to fund them. Deflation could be achieved either through allowing reserves to flow out depressing the money supply and domestic expenditure – the classical mechanism – or by policies that directly affected the components of demand. In Britain it was interest rates that acted as the key deflationary tool. From 1923 there was a trend rise in the Bank of England’s discount rate as the authorities adopted policies consistent with the return and maintenance of the exchange rate at the pre-war parity. At the same time the general trend of other central banks’ discount rates was downward (Eichengreen, 1994). The deflationary impact of such policies helped to keep the balance of payments in surplus and prevented the loss of gold.1 Additionally, the Bank of England also deployed gold-market and foreignexchange operations to maintain its stock of international reserves (Moggridge, 1972). The impact on the real economy was to slow growth, with the economy failing to reap its growth potential (Kitson and Solomou, 1990). Despite the level of GDP in 1924 being significantly below that of 1913, the growth rate of the British economy was significantly below the world average.2 Similarly unemployment remained persistently high, averaging just under 8 per cent for the period 1924–9 according to Feinstein’s (1972) figures. Furthermore, although the UK managed to maintain a balance-of-payments surplus, its export performance was poor – exhibiting slow growth and a declining share of world markets. The UK share of world exports in 1929 was 3.2 percentage points below its 1913 level, an average annual decline of 1.6 per cent (Lewis, 1949). Thus the deflationary bias of the gold standard did not merely fail to deal with the structural problems of constrained countries, such as the UK, it accentuated them. It not only lowered growth and raised unemployment but hampered long-run competitiveness. The dampening of domestic demand reduced the benefits of mass production and the exploitation of scale economies. Deflation to maintain external equilibrium raised unit costs and generated a further loss of competitiveness and declining shares of world markets. Such a process of cumulative causation led the UK to suffer a vicious cycle of stagnation. Locked into a fixed-exchange-rate system there were few policy options to reverse the process.

92 United Kingdom

The gold standard and the Great Depression If the gold standard failed to maximize world growth in the 1920s its shortcomings were also evident with the onset of the Great Depression. The causes of the Great Depression are subject to continual debate. Many studies focus on domestic developments in the American economy which were transmitted to the world economy. Friedman and Schwartz (1963) have emphasized tight monetary policy;3 Kindleberger (1973) stresses the fall in consumption due to redistribution of income from the agricultural sector where prices were falling; Romer (1990) considers the decline in the consumption of durables due to increased uncertainty created by the Wall Street Crash; Lewis (1949) considers the collapse of American capital exports; other policy factors include the alleged failure of fiscal policy to provide automatic stabilizers and the argument that the Smoot-Hawley tariff of 1930 initiated a mutually destructive trade war (Friedman, 1978; Capie, 1992). Others studies have focused on the Great Depression as being initiated through changes and structural problems in the international economy (Fearon, 1979). It is apparent that there is no single satisfactory explanation of the Great Depression. In fact an explanation which embraces the cumulative impact of structural problems, adverse demand shocks and policy mistakes, such as adherence to the gold standard, would seem to be a more realistic approach than a monocausal view. The extent of the Great Depression can be attributed to the operation of the gold standard (Temin, 1989; Eichengreen, 1992; Kitson and Michie, 1994). The impact of adverse shocks, such as the recession in the US and the collapse in capital exports, was transmitted to the rest of the world through the exchange-rate regime. As foreign loans were called in due to developments in the domestic economy, the gold flows to the United States increased. The draining of reserves in the debtor countries accelerated and monetary policy was tightened to ensure gold convertibility. Thus the deflationary bias of the gold standard system resulted in a perverse reaction to adverse demand shocks. Rather than facilitating an expansion of demand to ameliorate the Depression the system magnified the problem, leading to a collapse in world trade. The main impact of the collapse of world trade was to push the UK economy deeper into recession in 1929 through a collapse in exports (Corner, 1956). There is significant evidence, however, that the UK economy was already in recession in 1928 (Solomou and Weale, 1993). This suggests that the Great Depression may be best described as two-phase recession with troughs in 1928 and 1932. The recessionary forces in 1928 can be attributed to the direct and the indirect impact of the gold standard. First, there was a collapse in trade in services which was a consequence of the impact of the collapse of US overseas investment on world purchasing power (Solomou, 1996). Second, there was a significant fall in construction activity, which was, in part, a consequence of the high real rates of interest required to keep sterling on gold.

Michael Kitson 93

Between 1929 and 1931, the main recessionary impulse was the real demand shock caused by the collapse of exports of both goods and services, which fell by 38 per cent during these two years (Thomas, 1994). Monetary shocks were relatively unimportant during the early stages of the depression – the money supply was increasing during 1929–30 and only fell slightly during 1930–1 – but were more important during the later stages of the recession. Monetary effects became apparent in 1931 as the impact of the financial crises and rising fiscal and trade deficits led to rising real rates of interest as the government strove to maintain the value of sterling.

3

The Depression: the role of national specific factors

The impact of the Great Depression varied across countries, and, to a large extent, these variations reflected national-specific factors. As shown in Figure 4.1, which plots the path of GDP for the UK and the ‘world’ economy (here, the ‘world’ is Maddison’s (1982) leading 12 countries ), the UK depression, although severe in real terms, was relatively mild compared to the experience of the world economy; an annual output decline of less than 2 per cent compared to a world average of over 6 per cent. An illustration of the range of experience is shown in Table 4.1, which provides performance data for the four leading countries: it shows that the depth of the Depression was more than six times greater in the US compared to the

Figure 4.1 British and world gross domestic product, 1929–37 (1929 = 100) Sources: Feinstein (1972) and Maddison (1982).

94 United Kingdom Table 4.1

UK interwar economic performance: some international comparisons Depression 1929–32

Recovery 1932–7

Inter-period (1929–37)–(1924–9)

GDP (average annual per cent change) UK USA Germany France

–1.7 –10.4 –5.6 –5.1

4.3 6.5 8.5 2.6

–0.6 –3.6 –2.5 –3.3

Industrial production (average annual per cent change) UK USA Germany France

–3.7 –18.6 –15.3 –9.6

7.8 13.8 15.8 –2.6

0.4 –6.0 0.0 –6.8

Source: Author’s calculations from Maddison (1991).

UK. There are a number of factors that can help to explain why, although deep in historical terms, the UK depression was relatively mild compared to the experience of the other industrialized countries. The financial system The stability of the domestic financial system also helped to moderate the extent of the Great Depression in the UK. The fragmented banking structures in Europe and the United States led to a series of bank failures which caused chaos in capital and currency markets. These divergent experiences emphasize the importance of stability in the financial system for the rest of the economy. This reflects the issue of ‘systemic risk’ – that the social costs of the failure of a financial system exceed the private costs: a bank failure can lead to a ‘contagion’ effect which can cause solvent banks to become insolvent, leading to a collapse in the financial system with reverberations throughout the real economy. Greenspan (1996 [1998]) has recently commented, ‘There will always exist a remote possibility of a chain reaction, a cascading sequence of defaults that will culminate in financial implosions if it is allowed to proceed unchecked.’ In the early 1930s, such a ‘remote possibility’ was a reality (‘manias, panics and crashes’ to use Kindleberger’s (1996) phraseology) in many countries – but not in all. There were major banking crises in the US during which over 2,000 banks collapsed (Thomas, 1994); similarly, in 1931, there were widespread bank failures in Central Europe, (Bernanke and James, 1991). In the UK, however, there was no major financial crisis despite some monetary tightening during 1930–1. This relative stability , in the face of such a large shock, reflected the maturity of the UK financial

Michael Kitson 95

system and its concentrated structure: mergers before the First World War had led to the five largest clearing banks holding the vast majority of current account deposits, and no UK commercial bank reported a loss during 1929–32 (Capie and Wood, 1994). The agricultural sector The extent of the depression was also moderated by the UK’s low dependence on agriculture. Free-falling primary product prices severely affected the incomes of those countries that depended on agricultural output, whereas countries such as the UK with a relatively small agricultural sector suffered less direct adverse effects. The UK also benefited from the terms-oftrade gain of lower import prices – during 1929–32, the collapse of world primary prices generated a 20.8 per cent improvement in the UK’s terms of trade. This helped consumer real wages to increase during the period, which provided a positive demand boost to the economy. An early policy regime change An important factor which moderated the extent and duration of the slump in Britain, and initiated the start of recovery, was a shift in policy regime – and, it should be emphasized, a relatively early shift compared to many other countries. The series of fiscal and balance of payments crises led to a number of macroeconomic policy changes, starting with the suspension of the gold standard and the devaluation of sterling in September 1931. The devaluation of the exchange rate also allowed the government to pursue a more expansionist (‘cheap money’) monetary policy. Additionally, the protection of manufactures was extended by the emergency Abnormal Importations Act in November 1931 and the Import Duties Act in February 1932. The combination of these changes in trade and monetary policies increased aggregate demand for British products, thus helping to promote recovery from the Great Depression. The suspension of the gold standard and the accompanying devaluation had a number of beneficial impacts on the domestic economy. First, it improved trade performance and alleviated the balance of payments constraint on growth. The competitive gain of the devaluation was particularly large in 1932. Taking Redmond’s (1980) figure of a 13 per cent depreciation during 1931–2, Broadberry (1986) undertook an elasticities’ analysis of the policy change, measuring the impact of the change in relative prices on the volume of exports and imports. The estimated improvement in the balance of trade, resulting from this competitive advantage, amounted to £80 million. Assuming a multiplier of 1.75, a 3 per cent increase in GDP can be attributed to the relative price effect, a large part of the turning point in 1932. The second benefit of devaluation was that it removed the exchange rate constraint on monetary policy so that interest rates could be determined by

96 United Kingdom

domestic economic conditions rather than having to maintain the exchange rate or prevent excessive loss of reserves. The ‘cheap money’ policy has been identified as a permissive policy for economic revival (see Richardson, 1967), especially important in stimulating a housing boom (Worswick, 1984). Although the fall in building society rates had a major effect on lowering the cost of mortgage repayments, there is, however, significant evidence that this was not the major source of economic revival (Kitson and Solomou, 1990). First, much of the stimulus to housing was a result of the growth and development of building societies: building society deposits grew faster in the 1920s than in the 1930s due to an increase in the proportion of savings going to building societies and also to a peak in repayments. Thus, the increased availability of funds in the 1930s was mainly due to institutional developments in financial markets in the 1920s and not the monetary policy shift of 1932 (Humphries, 1987). Second, even if a link between monetary policy and the housing boom can be established, the contribution of housing to the recovery was limited, as it was a relatively small sector in the economy (Kitson and Solomou, 1990). Devaluation and the accompanying introduction of other expansionist policies led to a third, less mechanistic, benefit. Under the prevailing world conditions of uncertainty and monetary and financial turbulence, the reorientation of policy towards the domestic economy significantly improved business confidence. The prospect of a stable and growing economy encouraged home producers to increase, or bring forward, investment and expand production. The extension of protectionism also led to increased demand for domestic products through a number of mechanisms (Kitson and Solomou, 1990). First, it improved competitiveness of domestic manufactures, which reduced Britain’s dependence on imported manufactures and encouraged the production of domestic substitutes. Second, the resulting increase in domestic incomes generated a demand stimulus for the whole economy. Third, the more favourable conditions for manufactures and the expanded domestic market allowed the exploitation of economies of scale and increased productivity.

4

Policy and the recovery

Many of the policy initiatives not only contributed to the turning point from depression to recovery in 1932 but also contributed to the strength of the recovery throughout the 1930s. As noted previously, there was a sustained increase in GDP from 1932 and, as shown in Figure 4.1, the rate of increase was similar to that of the world economy. This was a notable achievement, as the British recovery followed a relatively mild depression whereas much of the world recovery was a cyclical bounceback in response to a deep depression. Thus the level of GDP in Britain, based on the 1929

Michael Kitson 97

benchmark, was significantly higher than that of the world economy. Many of the mechanisms through which policy sustained recovery were similar to those discussed above that initiated the turning point from recession. There were, however, a number of differences, the most important of which was the dissipation of the relative price advantage of the devaluation of sterling due to the global collapse of the gold standard and a series of devaluations abroad – there was a large nominal devaluation of the sterling effective exchange rate between 1931 and 1932 but this advantage was subsequently eroded.4 Although the competitive effect of the suspension of gold was moderated, the gains of the early 1930s may have provided longterm advantages as the short-term benefits of improved trade performance could be sustained through the establishment of distribution networks and customer loyalty. Throughout the 1930s Britain maintained a reduced propensity to import which increased the share of domestic demand for domestic products – although much of this improvement can be attributed to the benefits of protectionism which remained throughout most of the decade. Additionally, Britain’s share of world export markets stabilized, whereas it had been falling for the previous fifty years; this, however, did not lead to export-led recovery due to the low volume of world trade during the 1930s. Furthermore, and despite competitive devaluations abroad, breaking the exchange-rate constraint on monetary policy allowed lower interest rates throughout the period. The policy-regime change initiated in 1931/32 was central to Britain’s improved economic performance in the 1930s. There are, however, a number of additional factors which need to be considered. First, policy acted in the context of favourable supply conditions, such as the existence of a technological gap with the US which provided the potential for ‘catching-up’ growth and the development of new industries. Second, there were favourable demand shifts that were, at least in part, independent of macroeconomic policy – such as the favourable terms of trade shift in the early 1930s. Third, firms and employees did not offset all the output effects of increased demand through increased prices and wages. Although economic recovery was stimulated by a policy-induced demand shift, this was not of the traditional Keynesian sort – that is, via expansionary fiscal policy. The ‘Treasury View’ and balanced budgets were, and remained, the order of the day. Although the actual budget balance went into a small deficit during the Great Depression this was due to the operation of automatic fiscal stabilizers – such as falling tax revenue and increased expenditure on such items as unemployment benefits. In fact the government attempted to limit the operation of the stabilizers through public expenditure cuts including cutting the standard rate of unemployment benefit. The discretionary component of fiscal policy, that which excludes the automatic component of stabilizers, was in surplus during the Depression, suggesting, from a Keynesian perspective, a deflationary fiscal

98 United Kingdom

stance during the Depression (Middleton, 1981).5 This tight fiscal policy was continued during the early 1930s and was only relaxed towards the end of the decade with the advent of rearmament. Although the policy shift of the 1930s improved economic performance it has been argued that, as it was a response to economic crisis, the shift lacked coherence, and was not discretionary but was forced on the authorities (Beenstock, Capie and Griffiths, 1984). This argument, however, overstates the case. Booth (1987) has argued that policy-makers had a coherent strategy which was to increase profitability through raising prices. Although, as we have argued, the impacts of policy changes were primarily through other mechanisms, namely output effects rather than price effects, this does not negate the fact that the policy shift had internal coherence. The increased management of the economy can be illustrated by looking at exchange-rate policy. First, the timing of policy was important for moderating the recession and promoting recovery. Those countries than untied themselves from gold early were more likely to experience faster economic growth (Newell and Symons, 1988; Kitson and Michie, 1994; Solomou, 1996). Furthermore, some countries, such as France, Belgium and the Netherlands, despite prolonged recession were reluctant to leave gold. This suggests that leaving gold was not an automatic act, although those countries that remained locked into the system did so because of their fear of inflation (resulting from their experience of the 1920s), whereas British policy was more concerned with raising the price level. Secondly, following the suspension of gold the exchange rate was not allowed to float freely but was managed through the newly created Exchange Equalisation Account. This allowed the monetary authorities to buy or sell sterling to offset movements in the exchange rate caused by private sector trading.

5

The gold standard and UK economic performance

In his evaluation of the interwar gold standard Eichengreen (1992) argues that the lack of credibility and cooperation were the main weaknesses of the system. Credibility required acceptance of the (temporary) adverse impacts on output and employment, cooperation was required to ensure adjustment was symmetric rather than solely dependent on deflation by the constrained countries. Despite many similarities the argument here has differences. Simply, the system was structurally flawed.6 First, although cooperation would have limited the deflationary bias, it may have only extended the life of the system rather than prevented its ultimate demise. Even if the stronger countries had been encouraged to reflate this would not have completely eradicated the deflationary burden on constrained countries, a burden which may have had persistent effects on productive capacity. Indeed the system may have resulted in divergent growth paths. Reflation in the stronger countries could have led to faster growth of

Michael Kitson 99

output and productivity leading to a virtuous cycle of growth while deflation limited the growth potential of the relatively weaker countries. Second, the system combined together countries with different economic conditions and problems. These problems were not eradicated by the regime, rather they were accentuated. Discretion over the use of monetary, fiscal and exchange rate policy was removed. Third, the regime was not able to accommodate adverse shocks. As Eichengreen notes, the operation of the international monetary system magnified the impact of recessionary forces. As discussed above, the regime was incompatible with members possessing different economic structures and with a recession phased differently between countries. The UK returned to the gold standard at an overvalued rate – and the announcement of the intention to do so led to the problems of overvaluation being greater in the early 1920s than later in the decade. This induced a severe demand constraint for the UK economy, transmitted though poor trade performance and tight monetary policy. Furthermore, the gold standard regime had removed any policy flexibility to alleviate the demand constraint. The overall impact was to slow growth and increase unemployment; the problem was particularly severe in those areas that were highly dependent on tradable sectors, and so the regional disparities in prosperity increased. Growth and improved economic performance during the 1930s were significantly dependent on countries untying themselves from the strictures of the gold standard and adopting independent expansionist policies. Leaving the gold standard was not, however, the simple solution for economic recovery. First, timing was important: leaving gold early, as in the UK case, generated greater benefits as it not only provided the potential to stimulate demand but it mitigated the adverse long-run impacts of the Depression. Second, the exchange rate regime was not the only factor in recovery: some countries also reaped the advantages of increased protectionism (as in the case of the UK) and fiscal expansion (as in the case of Germany). Third, national-specific factors were important: the characteristics of labour, product and capital markets were different – shocks and policy changes had different impacts in different countries. Fourth, although a cooperative growth framework in the 1930s might have improved overall global economic performance, the uncoordinated policies of the 1930s were a vast improvement on the deflationary coordination of the 1920s.

6

Conclusions and lessons

The reconstruction of the gold standard in the 1920s was misconceived and damaging. Misconceived, because it was founded on the questionable notion that the classical gold standard had been a source of stability and

100 United Kingdom

growth. And damaging, because it reduced world growth and prosperity. The lessons for the construction of international monetary systems are twofold. First, systems must allow individual countries some form of adjustment to deal with domestic problems and shocks; and, second, systems must evolve as the world economy evolves. The first lesson was learnt relatively quickly – as the desire not to repeat the turmoil of the interwar period led to the construction of a new global financial architecture at Bretton Woods in 1944 – but the second one was not. The Bretton Woods system was based on an adjustable peg exchange rate system, which was a source of global stability, but individual countries retained some control over domestic macroeconomic policy as they could impose capital controls. Bretton Woods was flawed – as with the gold standard, the burden of adjustment was put on those countries with balance-of-payments deficits and not those with surpluses – but it was successful and it helped to create the conditions for the rapid growth of world output (Kitson and Michie, 2000). The demise of Bretton Woods and, ultimately, its collapse in 1973 reflected the failure of the system to evolve to changing global conditions, in particular the relative decline of the United States and the increasing inconsistency between US domestic policies and the needs of the world economy. The deflationary bias of the gold standard had a direct impact on the UK economy through reducing world growth and thus UK exports. The additional problem for the UK was that of entering a fixed exchange rate system at an overvalued rate. This increased the deflationary impact of the system as well as removing the potential to introduce policies that would stimulate the domestic economy. The operation of the gold standard was a major contributing factor to the Great Depression and its global reach. In the UK the depression was relatively mild and this illustrates the importance of national-specific factors in transmitting and amplifying shocks. National economies, then as now, were dependent on different markets; they had different institutional, and often legal, frameworks; there were differences in preferences, uncertainty and social norms; and there were different policy responses. With such contrasts, it is not surprising that the timing, extent and duration of the Great Depression varied so significantly between countries. Economists from different schools of thought frequently quibble over transmission mechanisms (such as how interest rates affect investment, the demand for money, and so on). What the Great Depression showed was that the variation in transmission mechanisms across time and space are more important than theoretical disputes. In the UK the stability of the financial sector was critical in reducing the domestic impact of global deflation. This illustrates the importance of the institutional response to shocks – countries with more fragile and dispersed financial systems suffered deeper recessions. The policy response,

Michael Kitson 101

and its timing, were also important in reducing the Depression. In particular, those countries such as the UK that withdrew from gold early and introduced other measures to boost demand had milder depressions. They also tended to have better recovery paths, which showed that deep depressions can have a lasting impact on growth by reducing productive potential and increasing uncertainty. Temin (1989) has argued that there was not a significant policy regime change in the UK in 1931, and that the recovery came from two factors: first, the revival in world output and world trade; and second, the housing boom. Here we have posited a different perspective: that the policy regime change was both significant and central to the UK’s improved economic performance in the 1930s. There were three major elements to the regime change. First, the suspension of the gold standard and the introduction of a managed exchange rate. Second, lower interest rates, with ‘cheap money’ replacing the ‘dear money’ of the 1920s. Third, the introduction of protectionism represented a major change as the UK had previously pursued a unilateral adherence to free trade. The combination of these changes in trade and monetary policies increased demand for British products and so helped to dampen the Depression and promote growth. Furthermore, with such high levels of uncertainty in the global economy, the reorientation of policy towards the domestic economy improved business confidence and encouraged producers to increase investment and expand production (Kitson, 1999). Where does this leave Temin’s contention that policy was not significantly expansionary? Simply, it is not tenable. First, the contention that the UK benefited from a significant external stimulus from expansionary regimes in the US and Germany is too simplistic. Although a growth in exports helped to start the recovery, exports were not a main driving force in the overall revival and, of course, they recovered, at least in part, because of a more competitive exchange rate. The UK’s share of world trade stabilized in the 1930s but the level of world trade, although increasing, remained below the pre-Depression level; thus the volume of UK exports in 1937 was only 80 per cent of the level in 1929. Second, as discussed above, although housing grew fast, it was a relatively small sector; the growth in manufacturing output made a much larger contribution to the overall growth in the economy. Notes 1 The balance of payments on current account was in surplus from 1924 to 1929 apart from 1926 when the impact of the General Strike resulted in a small deficit (see Feinstein, 1972). The adverse impact of the overvaluation on competitiveness, however, led to smaller surpluses than those achieved in the immediate pre-war period. 2 During the period 1913–29, Britain’s growth rate of 0.7 per cent per annum was approximately one-third the world average (Kitson and Solomou, 1990).

102 United Kingdom 3 Cooper (1992, p.2125) observed that Friedman and Schwartz ‘having never met a central bank they liked, of course attributed the severity of the depression to the perverse behaviour of the Federal Reserve Board’. 4 As well as the appreciation of the nominal exchange rate from the mid-1930s due to competitive devaluations abroad the real exchange rate also appreciated due to higher domestic inflation. 5 It is questionable whether the actual budget balance or the discretionary budget balance (sometimes referred to as the constant employment budget balance) is the best measure of fiscal stance. Additionally, there is the issue of how budget deficits affect private sector expectations and confidence. In the context of the early 1930s, prior to the establishment of Keynesian demand-management policies, budget deficits were associated with economic instability and inflation. Thus, their psychological impact on private sector expectations was likely to lead to reduced investment and, therefore, the fiscal orthodoxy of the 1930s may have contributed to recovery, albeit not through the mechanisms (‘crowding-out’) of orthodox economics. 6 Cooper (1992) has also argued that the system was structurally flawed focusing on the prospective failure of gold supplies to maintain the system without requiring continual deflation.

References Beenstock, M., F. Capie and B. Griffiths (1984) ‘Economic Recovery in the United Kingdom in the 1930s’, in Bank of England Panel of Academic Consultants, The UK Economic Recovery in the 1930s: Panel Paper No.23 (London: Bank of England): 57–85. Bernanke, B and H. James (1991) ‘The Gold Standard, Deflation and Financial Crisis in the Great Depression: An International Comparison’, in R.G. Hubbard (ed.), Financial Markets and Financial Crises (Chicago: Chicago University Press). Booth, A. (1987) ‘Britain in the 1930s: A Managed Economy?’, Economic History Review, 40: 499–522. Broadberry, S.N. (1986) The British Economy Between the Wars: A Macroeconomic Survey (Oxford: Basil Blackwell). Capie, F. (1992) Trade Wars: A Repetition of the Interwar Years? (London: IEA Current Controversies, No.2). ——– and G. Wood (1994) ‘Money in the Economy, 1870–1939’, in R. Floud and D.N. McCloskey (eds), The Economic History of Britain Since 1700, Vol. 2: 1860–1939 (Cambridge: Cambridge University Press. 3 vols, 2nd edn), pp. 217–46. Cooper, R.N. (1992) ‘Fettered to Gold? Economic Policy in the Interwar Period’, Journal of Economic Literature, 30: 2120–8. Corner, D.C. (1956) ‘British Exports and the British Trade Cycle’, The Manchester School, 24: 124–60. Cripps, F. (1978) ‘Causes of Growth and Recession in World Trade’, in Cambridge Economic Policy Group, Economic Policy Review, No.4 (Department of Applied Economics, Cambridge). Eichengreen, B. (1992) Golden Fetters; The Gold Standard and the Great Depression, 1919–1939 (New York: Oxford University Press). ——– (1994) ‘The Interwar Economy in a European Mirror’, in R. Floud and D.N. McCloskey (eds), The Economic History of Britain Since 1700, Vol. 2: 1860–1939 (Cambridge: Cambridge University Press. 3 vols, 2nd edn), pp. 291–319. Fearon, P. (1979) The Origins and Nature of the Great Slump (London: Macmillan).

Michael Kitson 103 Feinstein, C.H. (1972) Statistical Tables of National Income and Expenditure and Output of the UK, 1855–1965 (Cambridge: Department of Applied Economics and London: Royal Economic Society). Friedman, M. and A.J. Schwartz (1963) A Monetary History of the United States, 1867–1960 (Princeton: Princeton University Press). Friedman, P. (1978) ‘An Econometric Model of National Income, Commercial Policy and the Level of International Trade: The Open Economies of Europe, 1924–1938’, Journal of Economic History, 38: 148–80. Goldfeld, S.M. (1973) ‘The Demand for Money Revisited’, Brookings Papers on Economic Activity: 577–638. Greenspan, A. (1996 [1998]) ‘Remarks at the VIIIth Frankfurt International Banking Evening’, quoted in C. Goodhart, P. Hartmann, D. Llewellyn, L. Rojas-Suarez and S. Weisbrod (eds) (1998) Financial Regulation: Why, How and Where? (London: Routledge): 9. Humphries, J. (1987) ‘Inter-War Housebuilding, Cheap Money and Building Societies: The Housing Boom Revisited’, Business History, 29: 325–45. Keynes, J.M. (1925 [1972]) ‘The Economic Consequences of Mr Churchill’, reprinted in The Collected Writings of John Maynard Keynes, vol. IX. Essays in Persuasion (London: Macmillan for the Royal Economic Society): 207–30. Kindleberger, C.P. (1973) The World in Depression, 1929–1939 (London: Allen Lane). Kindleberger, C. P. (1996) Manias, Panics and Crashes (New York: John Wiley & Sons). [First edition, New York, 1978] Kitson, M. (1999) ‘Recession and Economic Revival: The Role of Policy in the 1930s and 1980s’, Contemporary European History, 8 (1): 1–27. ——– and J. Michie (1994) ‘Depression and Recovery: Lessons from the Interwar Period’, in J. Michie and J. Grieve Smith (eds), Unemployment in Europe (London: Academic Press). ——– and J. Michie (2000) The Political Economy of Competitiveness: Essays on Employment, Public Policy and Corporate Performance (London: Routledge). ——– and S. Solomou (1990) Protectionism and Economic Revival: The British Interwar Economy (Cambridge: Cambridge University Press). Lewis, W.A. (1949) Economic Survey, 1919–1939 (London: George Allen & Unwin). Maddison, A. (1982) Phases of Capitalist Development (Oxford: Oxford University Press). ——– (1991) Dynamic Forces in Capitalist Development (Oxford: Oxford University Press). Middleton, R. (1981) ‘The Constant-Employment Budget Balance and British Budgetary Policy, 1929–39’, Economic History Review, 34: 266–86. Moggridge, D.E. (1972) British Monetary Policy 1924–1931: The Norman Conquest of $4.86 (Cambridge: Cambridge University Press). Newell, A. and J.S.V. Symons (1988) ‘The Macroeconomics of the Interwar Years: International Comparisions’, in B. Eichengreen and T.J. Hatton (eds), Interwar Unemployment in International Perspective (Dordrecht: Kluwer). Nurkse, R. (1944) International Currency Experience (Geneva: League of Nations). Redmond, J. (1980) ‘An Indicator of the Effective Exchange Rate of the Pound in the 1930s’, Economic History Review, 33: 83–91. ——– (1984) ‘The Sterling Overvaluation in 1925: A Multilateral Approach’, Economic History Review, 37: 520–32. Richardson, H.W. (1967) Economic Recovery in Britain, 1932–9 (London: Weidenfeld & Nicolson). Romer, C. (1990) ‘The Great Crash and the Onset of the Great Depression’, Quarterly Journal of Economics, 105: 597–624.

104 United Kingdom Solomou, S (1996) Themes in Macroeconomic History: The UK Economy, 1919–1939 (Cambridge: Cambridge University Press). ——– and M. Weale (1993) ‘Balanced Estimates of National Accounts where Measurement Errors are Autocorrelated: The UK 1920–38’, Journal of the Royal Statistical Society (Series A), 156 (1): 89–105. Temin, P. (1989) Lessons from the Great Depression (Cambridge, MA: MIT Press). Thomas, M. (1994) ‘The Macroeconomics of the Interwar Years’, in R. Floud and D.N. McCloskey (eds), The Economic History of Britain Since 1700, Vol. 2: 1860–1939 (Cambridge: Cambridge University Press. 3 vols. second edition), pp. 320–58. Worswick, G.D.N. (1984) ‘The Recovery in Britain in the 1930s’, in Bank of England Panel of Academic Consultants, The UK Economic Recovery in the 1930s, Panel Paper no.23, pp. 5–28.

5 ‘Dancing on a Volcano’: The Economic Recovery and Collapse of Weimar Germany, 1924–33 Albrecht Ritschl

1

The theme

In late 1928, the Agent-General for Reparations, Parker Gilbert, had a conversation with Gustav Stresemann on the reasons for the country’s regained prosperity and its further outlook. Germany had witnessed a spectacular recovery from its postwar muddles: the hyperinflation was over and the gold standard restored, democracy had stabilized, reparation payments proceeded in an orderly manner, and the French were retreating from one occupation zone after another. The international confrontation of the early 1920s seemed a thing of the past, and a certain degree of political cooperation and détente had established itself. Germany enjoyed the fruits of restored international confidence, and more and more international capital flowed into the country. German bonds sold well in foreign markets, and US direct investors flocked into the country to spot profitable investment opportunities. It was in those pre-Depression years when companies like General Motors, Ford, and General Electric arrived in Germany. In a matter of only a few years, the German way of life would become a perfect imitation of the American one: motorization and the electrification of private households were just around the corner. So at least it seemed. However, Stresemann’s reply was gloomy: ‘We are dancing on a volcano.’ No doubt, this was the answer of an old and sick man, to be taken with a grain of salt. But only half a year later, the volcano had started to emit thick clouds of smoke. In the spring of 1929, a major German insurer failed, exhibiting the fragility of the country’s financial system. In March, a reparations committee clashed over a revision of the Dawes Plan. A foreign exchange crisis followed which could be averted only at the last moment. And for the first time since the stabilization of 1924, a public loan flotation failed. It failed so badly that from this time onwards, central government depended on the central bank even for short-term credits to bridge payment peaks at the end of a quarter. A loud controversy over the need 105

106 Germany

for budget stabilization arose between the Reichsbank’s president and the finance minister. By the end of the year, the finance minister and his budget director resigned in protest. The president of the Reichsbank, the notorious Dr Schacht, followed two months later. In March 1930, the cabinet fell, Germany was put under emergency rule, and the Great Depression had begun.

2

Motivation

Interpretations of the collapse of the Weimar Republic have long been controversial. Most authors, writing in the Keynesian tradition of the 1950s, identified the problem to have been rooted in a misguided deflationary reaction to a demand crisis (see Grotkopp, 1954; Kroll, 1958), if not in a deliberate, sinister attempt to deflate the economy in order to get rid of reparations (influential among the many: e.g. Sanmann, 1965; Mommsen, 1978). According to the first view, German politicians just made the same mistake as their American counterparts, not recognizing that the source of the problem was not in relative prices but rather in lack of confidence on the part of consumers and investors. According to the second view, German politicians knew exactly what they were doing but sacrificed social welfare to the political struggle against reparations. In either case, scholars concluded that fiscal and monetary restrictions only aggravated the slump, and that adherence to the gold standard served as those artificial ‘golden fetters’ (Eichengreen, 1992) which helped transmit the bad shock from one country to the other. A dissident view was presented by Borchardt (1979) who challenged the orthodoxy on the German slump on two grounds. First, he claimed that all was not well with the recovery of the late 1920s. He noted that unit wage cost had gone up steeply and that private investment had remained unimpressive. Second, he conjectured that some constraint must have operated on monetary and fiscal policy in Germany between 1929 and 1932. Reviewing the evidence on political decision-making at the time, he argued that German policy would have been less restrictive had government enjoyed easier access to credit. These hypotheses have created a fairly large literature. Almost every detail of the conjectures has been questioned, refuted, and restated in various ways (see the essays in von Kruedener, 1990, for an overview of the first decade of this debate). This chapter will review some of these debates and also present a new perspective on the problem. Ritschl (1995) argued that a fresh look at Germany’s reparation problem was needed to explain why Germany’s depression was so deep and came so early. Diplomatic historians have pointed out long ago (Link, 1970; Schuker, 1976, 1988) that the Dawes Plan of 1924 opened access for Germany to foreign credit in spite of the unfulfilled reparation demands. The financial design of the Dawes Plan

Albrecht Ritschl 107

distorted the incentives for German policy-makers and for international lenders. Consequently, Germany engineered a foreign credit rush, temporarily paying her reparations on credit, consuming more than the disposable income she had, and living beyond her means by all standards. In short, the Germans danced on a volcano, and they knew it. This chapter is arranged as follows. The next section examines the major components of aggregate demand and reviews the case for a Keynesian interpretation of the German interwar business cycle. Section 4 turns to the supply side and traces the wage cost issues that puzzled Borchardt and his followers. Section 5 looks into reparations and the incentive problems they caused, arguing that in this way, the views of Borchardt and his Keynesian opponents may not be so difficult to reconcile. Section 6 concludes with remarks on the German debt default and the beginnings of Nazi autarky policy.

3

The demand side

The most firmly established hypothesis holds that the German interwar business cycle was driven by demand forces and deepened by policies of deflation and balanced budgets. Failure of the state and the central bank to stimulate aggregate demand and expand credit during the slump had already been criticized by contemporary scholarship in Germany. Garvy (1975) documents that in those days prior to Keynes’s (1936) General Theory, German academics had already independently developed very similar doctrines, including a theory of the multiplier and a fairly robust idea of speculative money demand. Much of this scholarship derived its lessons from the early stages of the German hyperinflation of the early 1920s, when successive injections of money appeared to have increased real activity before raising the price level (Holtfrerich, 1986). Moreover, German central banking had a solid tradition of ‘productive credit creation’ under an extreme version of the banking doctrine (James, 1998). Thus there was ample experience with credit expansion in Germany, albeit overshadowed by the hyperinflation of 1923. A first test of Germany’s continued propensity to stabilize the economy in the presence of a slump had been passed in 1925/26, when the first recession after the 1924 reconstruction of the gold standard hit the German economy. The government undertook expansionary budget policies to help create employment and buffer the decline in output (Blaich, 1977; HertzEichenrode, 1982). A complicated scheme of sinking funds and equalization budgets was operated to conceal these budget deficits from the reparation creditors and their Agent General in Berlin. Germany’s official statements of public revenues and expenditures continued to show surpluses at a time when the central government budget was actually running deficits (Netzband and Widmaier, 1964).

108 Germany

After 1929, many scholars argued that similar policies should be followed again (for an account of this scholarship, see e.g. Hagemann, 1984). Plans and schemes for credit expansion were designed and published in large numbers, and the government was routinely criticized for clinging stubbornly to its balanced-budget policies (see the survey in Borchardt, 1990). During early 1932, Werner Sombart, then a grand old man in German historical economics and a professor emeritus in Berlin, even gathered a study group to hold seminars on credit expansion. This series provided a forum for economic activists of all kinds, ranging from the trade unions to the left wing of the Nazi party (Barkai, 1990). Many of its members would later be the rank and file of the bureaucracy that engineered the Nazi recovery. Writing in retrospect after the Second World War, these former junior technocrats credited themselves with having applied Keynesian policies in the Nazi recovery, even before Keynes had fully spelled them out conceptually (Grotkopp, 1954, is a typical example). It was largely their criticism of fiscal and monetary policy during the slump from 1929 to 1932 that shaped later scholarly debates. Deflation and balanced budgets came to be regarded as the principal reason why the German slump had been so severe, and work creation plus deficit spending received credit for having brought about the recovery. However, what exactly the effects of German fiscal and monetary policies were during the interwar years has been notoriously difficult to quantify. This is largely due to the opaqueness of Germany’s budgeting procedures at the time and the secrecy with which budget data were later treated in Nazi Germany. The seminal work in the field is still Erbe (1958), who produced estimates of public budget deficits to conclude that while budget policies during the slump were clearly recessionary, the Nazi recovery was far less Keynesian than one might think. On the other hand, in a study of FullEmployment Budget Surpluses (FEBS) for Germany, Cohn (1992) assigned very strong macroeconomic effects to central government spending during both the Great Depression and the subsequent recovery. If this were so, it would clearly be an international exception. Research on FEBS by Brown (1956) for the United States and by Peppers (1973) for the United Kingdom suggests that once proper adjustment for the endogenous fluctuations in tax revenues were made, the remaining possible effects of discretionary fiscal policy on the aggregate economy were minimal. Although the particular method employed by Peppers met with criticism (Broadberry, 1984), the broad picture still appears to be that in either country, the public sector was too small for deficits or surpluses to have sizeable aggregate effects. For interwar Germany, the use of archival data on central government budgets has greatly improved the accuracy of public deficit figures over earlier estimates. Ritschl (1998a) employed reconstructed budget data for the German central government to repeat the full-employment budget exercise for Germany. Results are given in Figures 5.1a and b.

Albrecht Ritschl 109 14 12 10 8 6 4 2 0 2

Deficit Figure 5.1(a)

1938

1936

1934

1932

1930

1928

1926

1924

4

FEBD Cohn

FEBD

German central government deficits, 1924–38 (bn RM)

5 4 3 2 1 0 1

d Deficit Figure 5.1(b)

Fiscal Impulse

The fiscal impulse, Germany, 1925–37 (bn RM 1913)

1937

1935

1933

1931

1929

1927

1925

2

FI Cohn

110 Germany

At a first glance, the information conveyed by these figures does seem to be in line with conventional wisdom: during the Depression, the fullemployment budget surplus indeed rises and declines sharply thereafter (in Figure 5.1, this is represented by a decrease in the full-employment deficit and its subsequent increase). That is, according to this Keynesian measure, fiscal policy was recessionary during the recession of 1929–32 and expansionary during the subsequent expansion. Looking at the fiscal impulse (i.e. the year-to-year change in the FEBS) in Figure 5.1(b), the effect becomes even more pronounced: while fiscal impulses remain adverse through 1931, a full Keynesian policy cycle becomes visible for the subsequent upswing that ranges from 1932 to 1935. (Note also the renewed surge of deficit growth during the Four-Year Plan from 1937 on.) Policy clearly became more expansionary towards the end of the Depression, and the strongest impulse was generated in the early phases of the upswing around 1934. However, these deficits were too small to provide a satisfactory Keynesian account of the German business cycle. In order to explain the contraction and the subsequent expansion of aggregate output from fiscal deficits, the two should be related by a Keynesian income–expenditure mechanism. The value of the expenditure multiplier would then have to be somewhere between unity (which underlies the FEBS concept) and a magnitude of two to three (if the income elasticity of tax revenues is markedly less than one). One simple way of confronting this with the evidence is to relate national income growth between any two years to the fiscal impulse (Figure 5.2). As Figure 5.2 bears out, the movement of GNP in most years was clearly too high to be explained by fiscal policy. In 1930/31, for example, real GNP fell by 5 times as much as the revenue-adjusted government deficit. In the two following years, the ratio was 6 and 12, respectively, while the FEBS concept assumes a multiplier of one. 40 30 20 10 0 10 20 30

31.39

5.04 6.46

1.22

15.70

12.74 2.89 1.75

5.89

2.20 1.48

2.11 18.98 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 d Y/d FEBD

Figure 5.2 1926–38

Theoretical Value of FEBS Multiplier

The instability of the Keynesian expenditure multiplier in Germany,

Albrecht Ritschl 111

Government budget policy in Germany during the Great Depression thus did look deflationary: however, the fiscal impulse it generated was by orders of magnitude too small to explain the contraction of national output in a Keynesian fashion. We must therefore conclude that much of the popular criticism of Brüning has been exaggerated: German national product apparently declined for reasons other than misguided fiscal policy. If all is not well with the old-time Keynesian explanation of the German slump, it may be rewarding to look into the deeper parameters of the German economy at the time. As a first check we look into the most traditional of all Keynesian consumption functions, relating the year-to-year change in private consumption to the annual change in disposable income (Figure 5.3). German consumers in the interwar period were not exactly Keynesian: the marginal ratio of consumption to income fluctuates wildly and repeatedly exceeds the range permitted by Keynesian doctrine. In short, the ‘consumption function’ is highly unstable in interwar Germany, or to put it more drastically: there is probably no Keynesian consumption function at all. In the estimates shown in Panel a of Table 5.1, there is no evidence of a well-behaved, static consumption–income relation, be it for subperiods or the whole span from 1925 to 1938 for which we have halfway reliable quarterly data. During the sub-periods, the coefficient on the income variable is insignificant and has the wrong sign, while for the period as a whole, it is significant but pathologically low.1 Of course we could force our econometrics to yield the desired Keynesian results by just omitting trends and seasonal components (see the righthand side of Table 5.1). Then, the income variable picks up the trend and seasonals and reproduces something close to what a traditional textbook would have promised. However, this correlation is a spurious one, as we have suppressed all deterministic trend components. Consumer behaviour in interwar Germany was probably shaped by a combination of credit 2.5 1.84 2 1.47 1.5 1.13 0.99 0.72 1 0.5 0.32 0.01 0 0.5 1

0.71 0.18

0.50 0.46 0.34

0.60

1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 dC (priv) / dY (disp) Figure 5.3 1926–38

The instability of the Keynesian consumption function in Germany,

Static Keynesian consumption functions; interwar Germany (Quarterly data, 1925:1 – 1938:4; 2-stage least squares)

Dependent variable = consumption (a) Including seasonals (not shown) and trend 1925:1 1925:1 1930:1 1933:1 to 1938:4 to 1929:4 to 1932:4 to 1938:4

(b) Excluding seasonals and trend 1930:1 1933:1 1925:1 to 1932:4 to 1938:4 to 1938:4

Constant

1.706

6.659

4.071

13.159

0.816

0.936

1.481

Y_DISP

0.229

–1.920

–0.459

–7.137

0.639

0.557

0.345

TREND

0.005

0.027

–0.026

0.153

Adj. R2 0.8536 –1.116 0.8373 DW 0.8988 1.0797 1.3899 Variables: Cons = constant; Y_Disp = disposable income

0.5707 2.0411

0.9319 1.1232

0.9446 1.9326

0.6351 0.3667

TREND*CRISIS

–0.013

(Y_DISP)*NAZI

–0.081

(Y_DISP)*CRISIS

0.110

(1) t values in brackets. (2) CONS and Y_DISP in logs. (3) NAZI = 1, 1933:1 to 1938:4. (4) CRISIS = 1, 1930:1 to 1932:4.

112

Table 5.1

113 0.02 0.01 0 0.01 1 Figure 5.4(a)

2

3

4

5

6

7

8

Impulse response of income to shock in consumption

0.02 0.01 0 0.01 1 Figure 5.4(b)

2

3

4

5

6

7

8

Impulse response of consumption to shock in consumption

0.02 0.01 0 0.01 1 Figure 5.4(c)

2

3

4

5

6

7

8

Impulse response of income to shock in income

0.02 0.01 0 0.01 1 Figure 5.4(d)

2

3

4

5

6

Impulse response of consumption to shock in income

7

8

114 Germany

constraints, wildly oscillating incomes, and vague expectations about the future. But it cannot be adequately captured by a Keynesian income–expenditure mechanism. In order to track the income–expenditure dynamics of the German economy more closely, Figure 5.4 shows the impulse-response functions of a bivariate vector autoregression (VAR) in income and consumption, employing quarterly data and a trend. The impulse-responses are obtained in the usual way: first, a vector autoregression is estimated and the variance–covariance matrix is obtained. Performing a Choleski decomposition on the latter,2 the so-called orthogonal errors are obtained. Feeding these back into the VAR, the dynamic multipliers or responses to a one-standard-error shock to either of the error terms can be obtained and traced over time. In the present case, we followed the propagation of any shock over a period of eight quarters. According to the Keynesian income–expenditure paradigm, a positive one-off shock to income should cause a slightly smaller shock to consumption in the same period. Lagged responses would occur only to the extent to which consumption exhibits habit persistence. Nor would the income process itself react persistently to one-time shocks, except in the case of habit persistence or accelerator effects on investment. In contrast, an exogenous one-time shock to consumption would tend to propagate itself through time (by its multiplier effects on income) and peter out in geometric decay. The same should be observable for the response of income to one-off consumption shocks. Consumption–income dynamics in inter-war Germany hardly followed any such pattern. Income shocks decayed rapidly and petered out after 3–4 quarters, while their effects on consumption were somewhat more persistent. Consumption shocks practically did not propagate themselves at all: the direct response of consumption to its own past shocks falls to zero after two quarters and is even slightly negative at longer horizons. In the same vein, income shows no initial response to autonomous consumption shocks at all but exhibits a slight negative reaction at 3–5 quarters after the consumption shock. Drawing the different pieces of evidence together, results are quite discouraging from a Keynesian point of view. In the period in question, there does not seem to be any such thing as a stable Keynesian consumption function, neither in static nor in dynamic perspective. Although income shocks do have effects on consumption, the converse evidently does not hold. That means, there were no visible multiplier effects at the time. Hence, we must doubt that a Keynesian income–expenditure mechanism was operative that could have translated the small fiscal shocks of Brüning’s deflation policy into big swings in national income. This does evidently not mean that effective demand puzzles as such played no role in the German slump. Voth (1993), Tilly and Huck

Albrecht Ritschl 115

(1994), and Ritschl (1994) presented various different estimates of fixed investment, in which the role of aggregate demand was evaluated against distributional variables. One central element of Borchardt’s (1979) revision of the received wisdom on Weimar Germany had been his claim that investment was adversely affected by redistribution towards labour. The result of the subsequent debate has remained inconclusive: while Broadberry and Ritschl (1995) found evidence in favour of Borchardt’s claim, Voth (1995) presented evidence to the contrary, arguing that interest rates rather than wage pressure had explanatory power for investment in Weimar Germany. Clearly, from a neoclassical point of view, both wages and interest rates should have a negative impact on investment. Empirically, however, econometric estimates of investment functions have been notoriously plagued by problems of specification (Chirinko, 1993). This may also have been a problem of the debate about the determinants of investment in Weimar Germany. As the neoclassical theory of investment (e.g. by Hayashi, 1982) points out, all information pertaining to investment should be included in aggregate measures of Tobin’s q, namely, the stock market index deflated by a price index of equipment, provided only that capital markets are weakly efficient. As Chirinko (1993) points out, empirical estimates of this specification usually obtain highly significant but numerically very low coefficients for the stock market q, and other information such as lagged values of investment and other investment determinants continue to have explanatory power for investment activity. The same holds true of interwar Germany when we estimate such an investment function along with cost elements like wages or interest rates. In one such estimate, we obtained a strongly negative and significant coefficient for wages, but an insignificant and positive coefficient for interest rates. Changing the specification slightly, either coefficient was negative and significant (Table 5.2). The conclusion from that would be that not only were capital markets inefficient in Germany (variables other than the stock market index are also significant) but also that either side in the debate was right (both wages and interest terms play a role – which would not be surprising from a theoretical point of view). Real wage cost as well as interest rates increased in Germany during 1927 to 1929, which may lend credibility both to a cost-push and a monetary interpretation of investment decline. However, as Broadberry and Ritschl (1995) have noted, very similar increases in wage costs and interest rates were observable also in Britain at the same time, albeit without having similarly devastating effects on investment after 1929. Hence, a closer look is needed. Temin (1971) had argued from inventory investment that there were signs of a beginning recession in Germany already in 1927. Although the evidence met with criticism (Falkus, 1975; Balderston, 1977), Balderston

116 Germany Table 5.2 Determinants of investment in Germany (Quarterly data, 1925:1 to 1932:4; ordinary least squares) Dependent

Invest

Domestic M-orders

Constant

4.88984 0.14193 0.66970 –1.13290 0.02155 0.00293 0.86763 2.27053

0.52924 0.45475 0.60224 –0.47870 –0.00888 –0.00654 0.96061 2.15855

Lagged dependent Tobin’s Q Wages Interest Trend Adj. R2 Dw

Domestic M-orders = Domestic machinery orders. (1) All variables in logs except for INTEREST. (2) t values in brackets.

(1983, 1993) has argued convincingly for tightening capital market conditions in Germany before 1929. Voth (1999) found strong indications of a stock market bubble and bust in Germany in 1927, and Ritschl (2002) showed that there exists strong co-movement between a properly deflated stock market index and domestic orders of machinery in Germany. As Figure 5.5 bears out, domestic machine orders peaked in mid-1927, two years before foreign orders to German machine builders reached their maximum. Given that machine orders are a leading indicator of equipment investment, this suggests that the investment climate worsened in Germany since 1927, long before it did so in the economies of Germany’s main trading partners. This early decline of German investment is no doubt a key issue, as is its close connection with the beginning stock market decline of 1927. One possible candidate is monetary policy. After the stabilization of the mark in 1924, Germany was linked to the gold standard, and the Reichsbank adhered to a rather strict policy of defending its reserves. Bordo and Eschweiler (1993) employed a Taylor-type rule approach to examine Reichsbank policies and found it to be roughly consistent with defending a target reserve ratio. According to Keynesian orthodoxy, these ‘golden fetters’ (a catch-phrase coined by Eichengreen, 1992) caused German monetary policy to be too restrictive. Hardach (1976) and others have claimed that Germany’s monetary stance could probably have been less restrictive even within the gold

Index 1928 = 100

Albrecht Ritschl 117 160 140 120 100 80 60 40 20 0

01

5:

2 19

7:

2 19

2 19

8:

Domestic Orders

01

01

01

01

01

6:

2 19

9:

2 19

0:

3 19

01

1:

3 19

01

2:

3 19

Foreign Orders

01

01

01

3:

3 19

4:

3 19

5:

3 19

Berlin Stock Market Index

Figure 5.5 New orders to German machinery industry and ‘Tobin’s Q’ in the Berlin stock market, 1925–35

standard. The argument obviously rests on limited capital mobility: if the gold standard did not work without frictions, there was still room left for domestic monetary policy. One way to examine this proposition is to check out the stability of the money demand function. If capital mobility was imperfect, a stable relation should exist between domestic money demand and the Reichsbank’s money instruments. Table 5.3 shows a number of estimates of such static money demand functions. As can be seen from the left-hand panel, the results vary quite strongly when the estimation is repeated for the several sub-periods. In the estimate for the whole period of 1925 to 1938, the coefficient on the interest variable remains highly insignificant, the income coefficient is too low, and prices enter with the wrong sign. As the dummy variables indicate, there is evidence of structural breaks interfering with the regression. In order to eliminate the effects of the monetary regime switch which Nazi foreign exchange control introduced from 1933 on, the regression was run again for the period before and during the slump as well as for the whole span from 1925 to 1932. As the results bear out, the money demand function exhibits pathological behaviour already during the Depression. Only for the years of 1925 to 1929 do we find halfway plausible parameter values. Even here, the income and price elasticities of money demand seem to be too low. However, this is a well-known defect of the specification we chose here and does not necessarily imply pathological behaviour of the underlying structure. To check out the possible reasons for the instability of the parameters, two different alleys may be pursued. First we may ask if the relation is stable within any period, such that the breaks can indeed be associated with the breakpoints we chose here. In a co-integration framework, the above equations could be conceived of as equilibrium co-integration

118 Germany Table 5.3

Money demand functions (quarterly data)

1925:2 to 1938:4

OLS 1925:2 1930:1 to 1929:4 to 1932:4

1925:2 to 1932:4

2SLS 1925:2 1925:2 to 1929:4 to 1932:4

Const.

17.963 (5.3338)

7.3083 (2.5761)

31.654 (5.9327)

11.0719 (2.6416)

–12.2945 (31.9172)

11.2726 (6.2303)

Y

0.2394 (0.5096)

0.7614 (0.2235)

–0.4146 (0.2634)

0.5819 (0.2444)

1.5735 (1.4591)

0.5781 (0.2661)

PRICES

–1.224 (0.7091)

0.0802 (0.3250)

–2.3940 (0.7642)

–0.3902 (0.3471)

2.5692 (4.0509)

–0.4172 (0.8354)

INTEREST

0.0051 (0.0172)

–0.0405 (0.0120)

0.0076 (0.0055)

–0.0134 (0.0083)

–0.1641 (0.1984)

–0.0122 (0.0338)

TREND

0.0246 (0.0092)

0.0158 (0.0039)

–0.1308 (0.0243)

0.0181 (0.0044)

0.0024 (0.0245)

0.0181 (0.0047)

CRISIS

1.5280 (0.8934)

NAZI

–1.9871 (0.7351)

TREND* CRISIS

–0.7919 (–1.8341)

TREND* NAZI

0.1022 (1.0721)

0.4922 (0.4427)

0.5175 (0.8383)

–0.0297 (0.0213)

–0.0308 (0.0382)

R2

0.8677

0.9533

0.8924

0.8935

0.5398

0.8934

DW

0.5333

1.8277

1.3864

1.4059

2.0767

1.3979

(1) All data in logs except for INTEREST. (2) NAZI = 1. 1933:1–1938: 4. (3) CRISIS = 1. 1930:1–1932: 4. (4) Standard errors in parentheses. (5) INTEREST in 2SLS estimates instrumented with US commercial paper rate.

relationships among integrated processes. However, this is necessarily a long-run concept, and any test of ‘stationarity’ would be meaningless for lack of power if performed for a time span this short.3 A second approach to the same problem is to look for a more structural explanation. An obvious candidate is the endogeneity of the domestic interest rate. If under the interwar gold standard the Reichsbank targeted a certain reserve ratio and if capital mobility was imperfect, the Reichsbank’s reaction to capital movements displacing the domestic money demand functions would be to adjust the interest rate. Thus, the interest rate and

Albrecht Ritschl 119

the residuals of the money demand function should be correlated. To overcome this ordinary least-squares (OLS) bias we need an instrument which is correlated with the domestic interest rate but not with the residuals. For this we can employ a short-term US interest rate, observing that the German economy was small relative to the US. The results of this exercise are shown in the right-hand-side panel of Table 5.3. As can be seen, all domestic variables become insignificant: German monetary demand appears to have been driven mostly by US interest rates, and given an apparently high degree of international capital mobility, the policy of the Reichsbank was entirely endogenous. Apparently there was no stable relation between domestic monetary aggregates and monetary instruments that monetary policy could have employed. If adherence to the gold standard paralyzed the Reichsbank’s manoeuvring capability so strongly, then obviously Germany should have left the gold standard in time as the Depression worsened. Why did it not just follow Britain in abandoning gold in 1931? Conventional wisdom as in Holtfrerich (1982) has it that failure to escape from the gold standard was Germany’s single most severe policy mistake during the Great Depression. Borchardt (1980, 1984) attempted a defence of Brüning, arguing that fear of renewed hyperinflation would have made any escape from the gold standard self-defeating. According to this view, maintaining a nominal anchor for the currency was essential to keep the German public calm and prevent interest rates from rising in anticipation of future inflation. The German inflation trauma, if it existed, was put to test in the German banking and balance-of-payments crisis of mid-1931. Although Keynesian critics typically argued that it was grossly misleading to think of fear of inflation playing any role during a deflationary slump, Borchardt (1985) argued that fear of creating inflation hysteria had been one of the major motivations preventing German policy-makers from abandoning gold. Following the approach of Mishkin (1981), Voth (1998) found evidence that anticipated inflation may indeed have gone up sharply in Germany during 1931. Indeed, short-term interest rates increased sharply during the summer of 1931, when the Austrian banking crisis spilled over to Germany. If this was not just a reaction to Germany’s balance-of-payments problems at the time (James, 1985) but indeed reflected inflationary expectations, sticking to the gold standard was probably the best response on the part of monetary policy.4 Summing up, the perspective on German monetary policy has shifted toward stronger emphasis on the constraints imposed by the gold standard. Adherence to gold was a credible commitment against inflation and budget deficits (Bordo and Kydland, 1995). Obtaining this ‘good housekeeping seal of approval’ (Bordo, Edelstein and Rockoff, 1998) of monetary policy was obviously much needed in Germany after the

120 Germany

hyperinflation. Tying the hands of the Reichsbank appears to have been successful, as the econometric evidence on money demand suggests that the gold standard indeed left little manoeuvring space for monetary policy. Leaving the gold standard was apparently not an easily available option, as the market reacted to fears of abandoning the gold standard with interest rate hikes. Given the record of the hyperinflation, Germany was probably in stronger need of a nominal anchor for its currency than any other major economy of the interwar period, irrespective of the damage to the real economy.

4

Supply-side puzzles

In 1979, a paper by Benjamin and Kochin on British unemployment in the Great Depression aroused widespread fury in the academic community. Benjamin and Kochin argued that interwar unemployment was largely a supply-side phenomenon. In the same year, Borchardt (1979) presented his revision of Weimar’s economic history, which in large parts arrived at similar conclusions, although in distinctly less radical fashion. In Borchardt’s view, high wage growth combined with the eight-hour day to create a sick and weak recovery from the hyperinflation, one in which supply conditions remained precarious. While, in the US, prominent writers in the Keynesian tradition had fully embraced the role of wage growth in aggravating the slide into the Depression (Bernanke, 1983), Borchardt’s proposition on German labour market conditions met with deep scepticism. In an influential contribution to the debate, Holtfrerich (1984) presented aggregate and sectoral evidence to show that unit labour cost in Weimar Germany was no higher in relation to labour productivity than in 1913. Ritschl (1990) went through the evidence again and argued from a more plausible time series that the profit squeeze identified by Borchardt could indeed have happened. Voth (1994) and Spoerer (1994) took the debate on unit wage cost further, while Broadberry and Ritschl (1995) examined comparative evidence for Germany and Britain to conclude that wage pressure in both countries must have grown in similar fashion at the time. Spoerer (1996) examined tax audit records of major joint-stock companies for the 1920s and found evidence of depressed profit rates already prior to the Great Depression. As of today, a consensus seems to have established itself that there was indeed some shift in the distributional position of labour in the late 1920s, although its precise consequences for the German economy still remain unclear. Table 5.4, adapted from Broadberry and Ritschl (1995), summarizes the available evidence on unit wage cost in Germany and provides a comparison with Britain and the US. As the data bear out, real unit wage cost in Germany rose very much in line with Britain during the years preceding the Depression. Very much the

Albrecht Ritschl 121 Table 5.4

1925 1926 1927 1928 1929

Indices of real unit wage cost in the aggregate economy (1913 = 100) Germany

UK

113.4 111.4 114.9 117.2 115.5

110.2 115.2 116.6 114.1 114.5

Source: Broadberry and Ritschl (1995).

same upward movement is visible for the unionized sectors of the American economy at the same time. Note that there is no such upward tendency in the non-unionized sectors. Prima facie this would suggest that union activity indeed drove the rise in unit labour cost before the Depression. However, the issue is not so easily resolved. If firms reacted to a demand shock with labour hoarding, observed wage shares would go up without there being wage pressure on employment. Holtfrerich (1984, 1990) pointed out that, as demand problems in the German Depression were all too obvious, rising wage shares as such are insufficient evidence for Borchardt’s (1979) wage-push hypothesis. One way of identifying the nature of the shock is to look into the behaviour of average and marginal labour productivity. If labour hoarding occurred, there should be evidence of declining average labour productivity during the slump, which should combine with spuriously high marginal products of labour. Bernanke and Parkinson (1991) examined US industry data for the Great Depression and found that short-run returns to labour were even increasing. This is consistent with a demand shock that hits output proportionately more strongly than employment. In such a case, the observed marginal product of labour measured at high frequencies over the business cycle is higher than unity. For Germany, contemporary data on output and total hours in industry were collected from 1929, but only a summary publication of the results at a monthly frequency seems to have survived. Figure 5.6 provides the result of calculating these data on hours into the index of industrial production. Surprisingly enough, hourly productivity in German manufacturing seems to have continued to increase during the Depression. There is no apparent evidence of an adverse productivity shock (which would be consistent with real-business cycle theory), nor do we find evidence of labour hoarding (which would be consistent with a Keynesian interpretation of increasing unit wage cost). To explore further whether short-term labour hoarding pushed German manufacturing below its efficient production frontier, we apply the rule-

122 Germany

Index 1928 = 100

130 120 110

Y/worker

100

Y/hour

90

Figure 5.6

1934:01

1933:01

1932:01

1931:01

1930:01

1929:01

1928:01

1927:01

80

German industrial labour productivity during the Great Depression

of-thumb approach of Bernanke and Parkinson (1991) described above, estimating the equation: OUTPUTt = a + b LABOURt + ci DETERMINISTICit + ut

(5.1)

where OUTPUT and LABOUR represent the logs of output and hours worked, respectively, and where DETERMINISTIC is a set of seasonal indicator variables and a time trend. For short time periods, we can assume the capital stock to be constant, such that eq. (5.1) can be regarded as a shortterm production function. If labour hoarding persists, output should fluctuate more strongly than employment. Then, we should obtain an abnormally high coefficient estimate on the labour input term. Bernanke and Parkinson (1991) reported coefficients above unity, a phenomenon known as short-term increasing returns to labour (SRIRL). Table 5.5 provides the results from our own estimates. Table 5.5

Short-run marginal returns to labour in Germany (quarterly data, OLS) 1927:1 to 1934:4

CONSTANT

1930:1 to 1932:4

0.9835

2.1558

HOURS

0.8391

0.7143

TREND

–0.00004

–0.0121

Adj. R2

0.9850

0.9915

DW

1.0319

1.6348

(1) All data in logs. (2) t values in brackets.

Albrecht Ritschl 123

As the table bears out, the elasticity of labour in production is below one and arguably close to the wage share in value added in manufacturing. Contrary to the evidence for the US, gathered by Bernanke and Parkinson, SRIRL is thus absent from the German data. Producers apparently stayed near the production function despite the slump in output. This phenomenon clearly needs further exploration. Bernanke and Parkinson had employed their measure in order to identify the source of an adverse shock to average productivity. Under the assumption of constant returns to scale in production, an adverse total factor productivity shock in the sense of real-business cycle theory would be consistent with declining average labour productivity but not with SRIRL. Despite an inward-shrinking production possibility frontier, the marginal relation between labour input and output would remain unaffected. On the contrary, a Keynesian demand shock would be consistent with average labour productivity decreasing and marginal labour productivity increasing, which is what Bernanke and Parkinson find for the US economy. For Germany, we find neither of the two: Average labour productivity is evidently not falling and SRIRL is rejected by the data as well. As a conclusion, we would have to accept that neither Keynesian labour hoarding nor adverse productivity shocks describe the German supply side adequately. Surprisingly enough, the German supply-side evidence through 1932 is still more consistent with Borchardt’s (1979) wage-push interpretation of the slump than with the rivalling Keynesian or real-business cycle explanations.

5

Reparations and foreign debt issues

Political historians (notably, Schuker, 1976, 1988) have long noticed a puzzle: given that Germany emerged from the Treaty of Versailles hopelessly indebted with reparations, how could it be that the country attracted massive foreign capital inflows in the late 1920s? Between 1924 and 1929, Germany borrowed abroad an amount equivalent to roughly one-third of her 1929 GNP and apparently had a marvellous time spending the proceeds: sport stadiums were built and urban transport connections expanded, public housing flourished through generous public subsidies, a mayor named Konrad Adenauer had Germany’s first four-lane motorway built (from his office in Cologne to nearby Bonn where he used to go for the weekend), and the national railway system even started electrifying its major long-distance routes, a luxury that the winners of the First World War could not dream of affording for themselves (James, 1985, 1986). Lest it be forgotten, there was a reparations bill, but that one was being paid entirely on credit, too. This ‘Ponzi’ scheme collapsed some time in 1928/29. From then on, Germany found it increasingly hard to attract new foreign funds and soon had to be glad if existing debts, most of them short-term, could be rolled

124 Germany

over (see Balderston, 1983, 1993). A run on the central bank’s reserves in April was fended off only with some effort, a major domestic loan flotation in May 1929 failed, and from the early summer of 1929 on, Germany switched to an increasingly strict austerity policy. Commercial lenders at home and abroad conditioned rolling over of existing debt to ever more deflationary policies. In the summer of 1929, the government turned to the notorious Dr Schacht of the Reichsbank as a lender of last resort. In the event, the central bank did extend some credit, but only after a law had been passed committing the treasury to a rigorous repayment scheme for its floating short-term debt. The adoption of this law, which came to be known as lex schacht, in December 1929 marks the beginning of deflationary policy in Germany. The political consequences of these regime changes were dramatic: the following months saw the resignations, in turn, of the finance minister and his budget director, of Dr Schacht as president of the Reichsbank (he would be reappointed by Hitler in 1933), and, finally, of the centre-to-left cabinet under chancellor Müller. Some time in between, a crash occurred on Wall Street, but the Germans were so busy with their own recession that they hardly noticed what happened (see James, 1985. A detailed account of the political decision process in German public finance in 1929 is Bachmann, 1995). In March 1930, a minority government of financial experts under Bruening was formed that did everything to meet a mounting foreign debt crisis by classical austerity measures. Short of majority support for its policies, it relied increasingly on presidential emergency decrees and on a tacit agreement with the moderate left to keep parliamentary sessions suspended for most of the time. A possible answer to the question why Germany accumulated high external deficits in the late 1920s could lie in an overvaluation of the currency. Figure 5.7 looks into this issue by comparing German indices of wholesale and consumer prices with their British and US counterparts. To provide a link with the pre-war period, we index all data to 1913. As the data show, Germany emerged from the condition of an undervalued currency around 1926. A large deflationary shock, combined with a return to the gold standard at the pre-war parity, had helped to restore the credibility of German central banking after 1924 (Sargent, 1986). A major slump had followed in 1925/26, and politicians indeed became worried that deflation had gone too far. A widespread feeling at the time was that some credit expansion should be permitted and that in international comparison, German price levels were, if anything, too low (Netzband and Widmaier, 1964; Hertz-Eichenrode, 1982). During 1927 and 1928, when capital imports were the highest, this readjustment came to an apparent standstill; Germany had ceased to be an emerging market, as Voth (1999) has termed it. Notice, however, that throughout the late 1920s, Germany’s consumer prices remained lower in international comparison than before the First World War.

Albrecht Ritschl 125 120 100 1913 = 100

80 60 40 20

1936

1938

1936

1938

1934

vs. GB

vs. US Figure 5.7(a)

1932

1930

1928

1926

1924

1913

0

Relative PPP indices for Germany (wholesale prices)

120

1913 = 100

100 80 60 40 20

vs. US Figure 5.7(b)

1934

1932

1930

1928

1926

1924

1913

0

vs. GB

Relative PPP indices for Germany (consumer prices)

High capital imports during the late 1920s were partly mirrored in German trade deficits. However, as Balderston (1993) has argued, German export performance was actually quite satisfactory at the time. Exports maintained themselves close to their 1929 level through most of the summer of 1930, and the nominal trade balance improved steadily (Figure 5.8).

126 Germany 600 mill reichmarks

400 200 0 200 400 600

Figure 5.8

1938

1937

1936

1935

1934

1933

1932

1931

1930

1929

1928

1927

1926

1925

800

Germany’s nominal trade balance

Had an initial overvaluation of the currency been the reason for the trade deficits of the late 1920s, the purchasing-power parity indices of Figure 5.7 should have fallen over time while the trade balance recovered. However, both series appear to grow over time, which means that during the period of high external deficits, Germany’s prices rose instead of falling relative to international price levels. Hence, initial overvaluation can hardly be seen as the driving force; the high deficits of the late 1920s appear to have had different reasons. One such reason may lie in the sovereign-debt aspects of the German reparation problem. Large portions of Germany’s capital imports of the 1920s served to pay out reparations on credit and thus were not reflected in the trade balance. The modern theory of country debt has made important contributions to explaining these ‘autonomous capital movements’, as a more traditional approach to international finance used to call such phenomena. Often, such capital flows and their seemingly abrupt reversal are motivated not so much by interest-rate differentials, domestic absorption, and similar standard determinants of external equilibrium, but rather by the highly incomplete nature of international capital markets. In the absence of an international authority that could fully enforce credit repayments, the only incentive for debtor nations to honour their debts is fear of sanctions and embargoes (see, e.g., Bulow and Rogoff, 1988, 1989). As such sanctions are seldom perfect, countries have the alternative of retreating into autarky and repudiating their external debt at a limited welfare loss, which would be given by the welfare costs of trade diversion and of building up import-substitution industries. Anticipating the risk of default but often ignorant of the exact economic conditions in the debtor country, international lenders would attempt to limit their exposure to country risk, an attempt which often results in an abrupt end to a lending boom, at the risk of producing self-fulfilling expectations of a debt crisis. To resolve a debt crisis, the parties involved would often avoid going to sanctions and

Albrecht Ritschl 127

embargoes by rescheduling and downsizing the debt payments, thus sharing the risk between debtors and creditors. Political historians have long pointed to the inherently political nature of Germany’s foreign-debt problem of the interwar years (Link, 1970; McNeil, 1986; Schuker, 1976, 1988). All of these writers argue for a causal connection between Germany’s capital imports between 1924 and 1929 and the later defaults on her reparations in 1931–2 and commercial debts, beginning in 1933. In the Dawes Plan of 1924, currency stabilization and a rescheduling of reparations had been combined with each other. Under the influence of Keynes’s (1919, 1922) warnings against exceeding Germany’s capacity to pay, a clause was inserted into the plan that obliged the Allied Reparation Agent in Berlin to defer reparation transfers abroad whenever a foreignexchange crisis threatened. In practice, this meant that commercial claims on Germany would be redeemed first, while the Reparation Agent would have to go back to the end of the queue waiting at the Reichsbank’s foreign exchange counter. In terms of the theory of incomplete credit markets, the seniority of commercial credits and reparations had thus been reversed (on the concept, see e.g. Bulow and Rogoff, 1989). More plainly, Germany was now allowed to issue new senior debt on top of her reparations burden. The reparation debt itself was already all too significant. Counting its more relevant portions, reparations about equalled Germany’s national product of 1913 (Ritschl, 1996).5 By modern standards, Germany was overindebted: not capacity to pay but rather her willingness to pay was the issue, as Germany’s reparation debt was higher than what could be exacted through sanctions and other coercive measures. The Allied occupation of the Ruhr valley in 1923 had demonstrated the limited power of sanctions quite clearly. In the end, damage had been done not only to the German side but also to France, which had emerged from the conflict weakened both politically and economically (Schuker, 1976). The Dawes Plan, of American design, aimed to solve the reparation problem by postponing it. Germany would stabilize financially, return to the gold standard, and thus accept playing by the international rules again. Besides embracing political détente, this also included resuming reparation payments. A payment scheme was established which offered the Germans quite substantial initial reductions over previous plans. To make this offer attractive, the US agreed to postpone a settlement of the inter-allied war debt, and there appeared to be good chances that these debts would altogether be forgiven (Link, 1970). Moreover, a stabilization loan was granted to the Germans in order to help restock the economy. As already mentioned, transfer protection against reparations promised to save the currency and downgraded reparation payments relative to commercial claims on Germany.

128 Germany

The Germans soon realized that this principle worked to their advantage. Although it was against the spirit of the Dawes Plan, the government did little to prevent reparations from being paid on credit. The sharp recession of 1925/26, partly provoked by the Reichsbank’s desperate attempts to create a primary trade surplus through monetary means, was soon overcome by public work creation and credit expansion through the public banking system. A controversy developed between the Reichsbank and the central government on how to stem the fast growth of public bond issues abroad. Under the Weimar tax constitution of 1920, central government could set most major tax rates and determine the level of tax revenue transfers to states and municipalities almost unilaterally (Pagenkopf, 1981). During the hyperinflation, it had abused this system as a weapon against reparations, keeping tax collection artificially low (Feldman, 1993). After the currency stabilization in 1924, it turned to the other extreme of minimizing transfers to lower-level governments. This helped to create the favourable impression of balanced central government accounts but shifted deficits to lower-level governments (see Netzband and Widmaier, 1964; Witt, 1982). As a consequence, states and municipalities increasingly turned to the international market to finance their budget gaps. Funds were raised partly through public bond issues, and partly through public utility and urban transport companies, which were often formally private but usually under state and municipal ownership. Tax legislation even supported foreign borrowing by granting tax privileges to foreign bonds (Hertz-Eichenrode, 1982). Upon the pressure of Schacht of the Reichsbank, this tax privilege was removed in the summer of 1927 (McNeil, 1986). As a consequence, foreign bond issues came to a sudden standstill, inflows of foreign exchange decreased rapidly, and the government became worried that reparations could no longer be paid on credit. Faced with a choice between stabilizing the budget or permitting more capital inflows, government opted for the latter. The tax break was reintroduced, and in late 1927, capital imports resumed at their previous levels. Political historians tell us that this decision was taken consciously in order to avoid having to pay out reparations from primary balance-of-payment surpluses, which is what Schacht would have preferred. As a senior official in Stresemann’s foreign ministry put it in 1927: ‘the more commercial debts we take in, the less we will have to pay out in reparations’ (McNeil, 1986). Having understood the logic of imperfect capital markets, the German government played a reparation gamble. Fresh money obtained in international markets would hopefully drive out reparations at the margin: in the case of a payment crisis, a conflict of interests between reparation creditors and commercial creditors would arise in Allied countries. In the end, no creditor country would risk a financial crisis for the sake of reparations (see Schuker, 1988, p. 39). In other words, the Germans took their

Albrecht Ritschl 129

commercial creditors hostage to the reparation conflict (Lüke, 1958; Link, 1970). As creditors to Germany grew increasingly nervous about their investments, it became obvious that further lending would have to stop at some point and that the credit pyramid might even collapse. At the central bankers’ meeting on Long Island of spring, 1927, Benjamin Strong of the Federal Reserve Bank of New York had already warned that within two years’ time, the worst depression in history would break out, the only question being whether it would begin in the US or in Germany (Link, 1970). Pressure on the Dawes Plan also mounted from the French side. In the Berengér-Mellon treaty of 1926, France had already committed herself to paying back her portion of the inter-allied war debt in 62 annual instalments. Although payments had been made tacitly since 1926, France ratified this treaty only in 1928, when the French economy had recovered sufficiently well to quell fears in the French parliament about France’s capacity to pay (Schrecker, 1978). From that time on, the former link with reparations was re-established: there would be no forgiveness of reparations unless the inter-allied debt was cancelled as well. If inter-allied debt was not forgiven, the rationale of the Dawes Plan had gone. During the Dawes Plan period, credit markets had been able to absorb an amount of reparations that was roughly equivalent to the indemnité nette from the initial reparations bill of 1921 (Ritschl, 1996). However, they would not absorb the inter-allied war debt, just as bankers had predicted in 1921: this debt was apparently too large to be ‘commercialized’ by Germany without creditors running into serious country risk. Consequently, reparations as well as the foreign debt accumulated during the Dawes Plan would now have to be paid from trade surpluses. This was what Benjamin Strong had predicted: the credit machinery for recycling German reparations would break down soon, as probably would the foreign credit pyramid that Germany had built up since 1924. To ensure payment of the inter-allied war debts through German reparations, the Young Plan introduced two new elements. First, transfer protection was essentially abolished. Payments of half the annuity could still be deferred, but only for a maximum of two years and without implying a general standstill. Second, reparations were set to match the payment schedule for inter-allied credits, which resulted in a payment plan of 59 years. Stated in present value, the Young Plan was actually more beneficial for Germany than the Dawes Plan, whose annuities were to reach a steady-state level in 1928–9. The Young Plan also included certain provisions for debt forgiveness. However, the terms of payment were far stricter than in the Dawes Plan, as suspending transfers was possible only for two consecutive years at the maximum and only for a portion of the annuity.

130 Germany

This ensured that paying reparations on credit became almost impossible now: reparations would be an additional block of debt ranging ahead of commercial debt, and both elements together amounted to a good 70 per cent of 1929 GNP. This presents an obvious puzzle: why did Germany accept the Young Plan if its conditions were so adverse for the German economy? Was Schacht right when he opposed the plan? In fact, the hostage doctrine of reparations and commercial debt held by the German Foreign Office offered hopes that the seniority scheme of the Young Plan would not prove robust if it came to a test. Under the threat of a serious payment crisis, creditors might prefer abandoning reparations to losing their loans to Germany, and reparations would go without affecting German creditworthiness in international markets. All this implies that the Young Plan had a credibility problem. Consequently, foreign lending to Germany did not cease immediately when the Young Plan was implemented. A major loan granted in connection with the Young Plan made the transition easier and postponed the onset of a foreign debt crisis by another year. Otherwise, however, German bond issues abroad soon came to a virtual standstill (Figure 5.9). As Figure 5.9 shows, there is a major revival of foreign borrowing in 1930, which is however almost entirely due to the Young Loan (which is included in ‘Public Administration’ and ‘Public Enterprises’). With increasing difficulties in raising money abroad, the German side was soon forced to resort to fiscal austerity and deflation. In the summer of 1930, Brüning’s emergency government tried to pass the new deflationary budget by presidential emergency decree in order to avoid defeat in parliament. However, upon French diplomatic pressure, this plan was given up and the bill was 800 700 mill. RM

600

PC

500 400

MB PA

300

PE

200

30 19

29 19

28 19

19

27

100 0

Figure 5.9 German bond issues abroad, 1927–30 Key: PC = Private Companies, MB = Mortgage Bonds, PA = Public Administration, PE = Public Enterprises. Sources: Balderston (1993); Statistisches Jahrbuch fuer das Deutsche Reich, various issues.

Albrecht Ritschl 131

presented to the Reichstag, where it failed.6 As a consequence of this defeat, general elections were called for September. Their result was a further weakening of the political centre and a dramatic rise in the votes for the Nazi party. This episode shows how vulnerable Germany had become to foreign diplomatic interference. In the summer of 1930, the last portion of the Young loan had yet to be floated, and France used this fact as a political weapon to make sure that the commitment to austerity policy was shared by all major political forces in Germany – with doubtful success, as we have seen. The fact that the Young loan could be used to exert political pressure also reveals that Germany’s access to foreign credit was now apparently very limited. Trying to avoid outright unilateral default, the government attempted instead to tighten the budget and bring down the general level of prices and wages. Salaries in the public sector were cut several times, as were fiscal transfers to states and municipalities. Amidst sharp deflation, rapidly growing unemployment and a mounting financial crisis, the Brüning government came under intense domestic pressure to declare international default and switch to credit expansion (Borchardt, 1990). In June of 1931, the Austrian banking crisis spilled over to Germany and turned into an all-out banking crisis there (on Austria, see Schubert, 1991; on Germany, see James, 1986). One of the country’s largest banks, the Darmstädter and Nationalbank, failed and a bank holiday was declared. Faced with the banking crash and a run on the Reichsbank’s gold reserves, the German government issued a memorandum asking for a change in reparation policies so that default on commercial debt could be avoided. The solution that was found in multilateral negotiations was a rescheduling scheme, announced in the Hoover moratorium on reparations and inter-allied credit (Heyde, 1998). Service on either debt element would be suspended for half a year, and extension would be considered if Germany met certain requirements, including continued adherence to the gold standard. For the first time, the US now recognised the de facto existence of a link between the two issues, albeit not yet formally. As a second element, a standstill agreement was concluded, which rolled over Germany’s short-term debt for half a year. As a result, Germany remained current on her commercial interest payments but was temporarily relieved from reparations. It almost seemed as if the German strategy from the 1920s had paid off: under the threat of a potentially contagious financial crisis, protection of commercial creditors had proved to matter more than reparations. However, this was not quite the outcome that German policy-makers had hoped for: reparations had been suspended only temporarily, and this did nothing to restore German creditworthiness in international financial markets.

132 Germany

6

Digression: was the Brüning deflation necessary?

We are now in a position to ask the counterfactual question whether alternative scenarios for economic policy existed after the banking crisis of 1931 that could have avoided the slide into the Nazi dictatorship. This issue has two parts to be considered. First, we ask whether Nazi rule was inevitable given Brüning’s deflationary policy, or whether Germany could have successfully recovered from the Brüning deflation without debt default, autarky, and the rise of Nazism. To put it differently, we want to know under what conditions the austerity policies of Brüning could have been more successful. A key issue here is that the next parliamentary elections were due only in 1934. However, Brüning’s cabinet fell victim to a political intrigue in May 1932, more than two years before its term was exhausted, and without having lost its support in parliament. The tactics of the emergency cabinet under Brüning had consisted in riding out the Depression in the hope of getting rid of reparations and restoring Germany’s external creditworthiness. Then, the fruits of recovery could be reaped in the next elections and the extremist vote would subside (see van Riel and Schram, 1993 on the interconnections between unemployment and the Nazi vote). Germany reached the spring of 1932 with a ramshackle economy but without having defaulted on her commercial debt. Reparations were indeed forgiven in the summer of 1932. Klug (1993) found that German bond prices in New York recovered immediately after reparations were cancelled. This would indicate that Germany’s policies towards her commercial debt indeed still had credibility. For our counterfactual history of the Brüning deflation and what might have been, this is actually good news, for it indicates that Brüning’s policies of financial conservatism indeed achieved their main economic goal. Had Brüning’s cabinet survived the summer of 1932, the fruits of abandoning reparations could probably have ripened in a climate of slowly returning confidence. Instead, his successor, Von Papen, complied with the orders from the president’s office, removed the remainders of the democratic coalition in the state of Prussia in a cold coup d’état, and thus provoked the dissolution of parliament on the national level (see e.g. Bracher, 1971). Counter to his expectations, two consecutive national elections in July and November of 1932 established a joint majority of the Nazis and the communists, shaking the credibility of Germany’s international commitments and finally dashing the hopes of her international creditors. A second, even more important counterfactual we want to study is the question of whether the Brüning deflation could have been avoided altogether. Conventional wisdom as in Jochmann (1978) has it that Brüning was obsessed with reparations and that his attempt to get rid of the repara-

Albrecht Ritschl 133

tion burden motivated his deflationary policy. In his criticism of this orthodoxy, Borchardt (1979) maintained that Brüning’s manoeuvring space was constrained by credit restrictions, whatever his intentions regarding the reparation issue. We have argued above that reparations under the Young Plan were apparently not irrelevant for Brüning’s manoeuvring space; instead they directly conditioned austerity and deflation. That is, Brüning’s deflation was the only feasible response to the Young Plan as long as outright default was to be avoided. However, we may ask how these constraints could have been relaxed if German policies during the Dawes Plan period had been different. Had Germany taken in less foreign credit after 1924, would a less restrictive policy have been possible during the slump? Let us suppose a counterfactual in which no credit rush had taken place in the 1920s such that Germany would not have been cut off from foreign credit after 1929. Ritschl (1998b) studied this in a Keynesian framework, assuming that the balance-of-payments constraint was tighter in the 1920s and looser in the 1930s. More precisely, we assume that reparations during the Dawes Plan had been fully transferred out of trade surpluses (Figure 5.10). The simulation shown in Figure 5.10 assumes an income-dependent Keynesian import function.7 If reparations had been fully transferred in the 1920s, in each year a primary trade balance surplus of the same amount would have had to be generated. Under Keynesian assumptions, this would ceteris paribus have depressed national income in the 1920s. On the other hand, a policy of full reparation transfers would have avoided accumulating foreign debt, which would have diminished the burden on the trade balance after 1929, when the Young Plan effectively drove out further foreign lending. As Figure 5.10 shows, the effect of this simulation is to 70 bn reichsmarks

60 50 GNP m = 0.15 m = 0.20

40 30 20 10 0 1925 1926

1927 1928 1929 1930 1931 1932

1933 1934

Figure 5.10 Actual and simulated GNP with full reparation tranfers, 1925–34 Key: GNP Actual GNP series m = 0.15 GNP simulation with marginal ratio of imports to income equal to 0.15 m = 0.20 GNP simulation with marginal ratio of imports to income equal to 0.20 Source: Ritschl (1998b). Reproduced with the permission of Cambridge University Press.

134 Germany

shift the Depression backward in time. The most drastic adjustment would have occurred in the mid-1920s, not in the 1930s. Germany would almost have followed the British trajectory, with unimpressive performance during the 1920s and a rather mild depression in the early 1930s.8 This confirms the hypothesis of Borchardt (1979) who claimed that in Germany, overly expansionary macroeconomic policies in the 1920s were a cause of the rather stern conditions prevailing in the Depression. The reasoning presented here suggests that with the Young Plan in place, there was little else for German policy to do than to wait until the scheme had proved to be unsuccessful. The sharp austerity policies pursued during the crisis were the price that Germany paid for the gamble it had played during the Dawes Plan, when it had attempted to drive out reparations through foreign credit. This strategy ultimately paid off, but at the high social cost of aggravating the Depression for Germany. In sum, during the Great Depression we find Germany caught in a condition of a mounting foreign debt crisis. The Brüning deflation from 1930 to 1932 was not an application of misguided doctrines, nor was it the sinister attempt to run down the economy in order to get rid of reparations. Instead, we see the German deflation as the rather passive austerity reaction to the foreign credit constraint imposed by the Young Plan. As long as German policy-makers had a desire to reintegrate Germany into the world economy, it made sense to comply with the Young Plan, hoping to get rid of reparations while not having to default on the commercial debt.

7 Default after default: reparations and the transition to Nazi autarky policies The debt arrangements of 1931 had pursued the objective of maintaining and later restoring Germany’s access to international credit markets. To this end, the short-term debt was frozen (while interest was still being paid), whereas the long-term debt remained entirely unaffected. Thus, Germany was still fully current on her interest obligations; the crisis was handled as a transitory contingency, not as a fundamental change in German debt policies. This arrangement made sense if all parties anticipated a future return to a regime of free capital movement and international economic integration. As the lion’s share of German foreign debt was either directly or indirectly American in origin, Germany had a strong incentive to follow America’s policies of adhering to the gold standard and resisted the temptation to engineer a joint escape from the gold standard with Britain. Much of Germany’s short-term debt was actually owed to Britain, as London was the financial agent for much of Germany’s foreign trade. This implies that freezing Germany’s short-term debt hit Britain disproportionately, working

Albrecht Ritschl 135

more to the benefit of predominantly American long-term credit. Joining Britain in the departure from gold would probably have made sense for Germany had its foreign credit status been different. As things stood, a few American hints as to the country structure of Germany’s foreign debt appear to have sufficed to convince the Germans that it was a good idea for them to employ an American, not a British recipe for curing the crisis (Lüke, 1958; Link, 1970). Thus, Germany formally remained on the gold standard, suspended short-term convertibility but retained her long-term commitments, such that in the long run, full return to the gold standard without debt default would be possible. How come this incentive changed over time? The slide into Nazi autarky policies provides us with a surprising element. From 1933 on, German foreign debt policies changed their objective and gradually started to unblock British credits (Wendt, 1971), while reducing debt service on longterm loans systematically (this is still a most superficial account of Nazi foreign debt policies, but it suffices here to make the point). That is, while during the slump the stability of long-term credit was favoured at the expense of short-term debts, policies were reversed now, and long-term debts were discriminated against. One main reason for this seems to be the disintegration of the gold standard and of the free trade order, which sharply reduced the incentive for Germany to keep current on her long-term obligations. The new Roosevelt administration in the US accepted the logic that in order for the Germans to repay their debt, they would have to be allowed to run export surpluses vis-à-vis the US, which the Americans preferred to avoid (Schuker, 1988). Given the international slide into protectionism and notably the change in US trade policies, the risks from defaulting on long-term debt were much lower now than before. In contrast, not defaulting on short-term trade credit made even more sense now than before. Germany had channelled much of her European trade into bilateral trade agreements but still needed credit facilities for her overseas trade. By giving preference to debt service on previous trade credit, Germany managed to regain some degree of creditworthiness in London in a general mood of British policies that have aptly been described as ‘economic appeasement’ (Wendt, 1971).

8

Conclusions and implications

This chapter has dealt with the economic recovery and collapse of Weimar Germany between 1924 and 1933. After 1924, Germany underwent a very volatile business cycle, with a large upswing first and a deep depression after. We reviewed the rivalling Keynesian and more supply-side oriented explanations that have traditionally been offered. The traditional Keynesian account viewed the German Depression after 1929 as an imported recession

136 Germany

which was aggravated through domestic deflationary policies. According to this view, dwindling capital exports from the US in the wake of the New York stock market boom had ended the German recovery from the hyperinflation, and instead of engineering a soft landing, German policy turned the recession into a deep slump. However, taking a closer look at the evidence, either element of this account seems doubtful today. As emphasized by Temin (1971), the beginning of the Depression in Germany sets in too early to be explained by US capital exports. During the later slump, data on full employment budget surpluses suggest that although policy was in a Keynesian sense procyclical, the impulses generated by the public budget were too weak to affect the economy to a significant degree. As to the supply side, evidence would suggest a business cycle similar to Britain at the same time: in either country, unit labour cost increased throughout the late 1920s and gradually fell thereafter. The same was true for long-term interest rates. Thus, had the German business cycle been driven mostly by supply factors, Germany would have had to follow largely the British pattern, where an anaemic, unimpressive upswing before 1929 was followed by stagnation and eventually a mild decline in output. Germany experienced nothing of this sort; instead her business cycle was far more pronounced than the British one. Searching for additional elements explaining the German slump, several authors, notably Balderston (1993), have paid renewed attention to international factors and especially foreign credits. One interpretation that we examined in some detail asserts that the foreign credit cycle of the German economy was largely caused by the adverse incentive effects of the Dawes Plan. To protect the central bank from foreign exchange outflows due to reparations, an escape clause had been written into the plan. This transfer protection clause, adopted under the influence of Keynes, permitted Germany to issue fresh senior debt in international markets, which she did in high amounts, effectively paying her reparations on credit. Once this credit pyramid collapsed, Germany was burdened with the double load of reparations plus debt service, and it was a matter of time until one of the two would have to go. The Young Plan of 1929/30 introduced tighter terms of payment to save both debt elements but soon proved to be unsustainable. Germany was caught in a mounting foreign debt crisis, and deflation was the austerity reaction attempting to prevent unilateral default. The crisis broke out in full after the banking panic of June 1931. The Hoover moratorium allowed Germany to suspend reparations while staying current on her commercial interest payments. Given the double burden of reparations and commercial debt, deflating the German economy in order to create balance-of-payment surpluses was logical under the Young Plan. Avoiding deflation would either have implied default or a different set of starting conditions in 1929. We exam-

Albrecht Ritschl 137

ined a counterfactual in which Germany would not have enjoyed transfer protection under the Dawes Plan. Without the artificial credit boom of the 1920s, the German economy would probably have passed through a slump in the mid-1920s, only to recover later on, and the slide into the Depression would have been notably less severe than it actually was. Our conclusions must still be tentative at this point. It seems that the German slump of the early 1930s does have to do with the reparation problem, albeit in an indirect manner. Focusing on the incentive effects of the reparation arrangements, an explanation of the German foreign debt cycle between 1924 and 1933 emerges. Germany manipulated the Dawes Plan to drive out reparations and then suffered the consequences under the Young Plan. In this way, Germany inflicted a transfer problem on herself, which backfired during the Depression, severely aggravating the slump and undermining the foundations of the Weimar Republic.

Notes 1 We are not the first to note this effect. Already the very first international econometric study of the consumption function by Stone and Stone (1938–9) obtained pathologically low parameter values for Germany. Similarly low estimates were obtained for Germany by Erbe (1958), Borchardt and Ritschl (1992), and Tilly and Huck (1994). 2 This involves imposing a triangular structure on the variance–covariance matrix, ordering the variables according to their hypothetical exogeneity. Here, the ordering was income–consumption. We also experimented with reversing the order, but without much change in the results. 3 Powerful or not, the Dickey-Fuller tests we performed on the residuals from the money demand functions were unable to reject the unit-root hypothesis, i.e. of lack of stability of the money demand functions. 4 Borchardt and Schötz (1991) provide a detailed account of the internal policy debates in the government and the central bank at the time. Indeed, restoring confidence by not leaving the gold standard appears to have played a major role in political decision-making during 1931. 5 This includes the A and B bonds, i.e. the indemnité nette and the redemption of inter-allied war debt, but leaves out the controversial and mostly propagandistic C bonds, which would account for another 152 per cent of 1913 GNP. For comparison, France’s reparations to Prussia after the war of 1871 accounted for about 20 per cent of her 1869 GNP (White, 1999). 6 On the background, see the diaries of the German budget director at the time, Hans Schäffer, held in the Institut für Zeitgeschichte Munich/Germany (folder IFZ Da 03.03). 7 As is implicit in such counterfactual exercises, this analysis disregards the Lucas (1976) critique. This may indeed be a problem, as we tacitly assume here that the government had a technology for preventing financial markets from counteracting a more restrictive balance-of-payments policy through countervailing capital movements during the 1920s. 8 This side result is interesting insofar as Britain was indeed suffering continuous balance-of-payment problems in the 1920s, which were the main obstacle to less

138 Germany restrictive monetary policies. See e.g. Cairncross and Eichengreen (1983). On a comparison of the German and British business cycles during the 1920s, see Broadberry and Ritschl (1995).

References Bachmann, U. (1995) Reichskasse und öffentlicher Kredit in der Weimarer Republik 1924–1932 (Frankfurt/Main: Lang). Balderston, T. (1977) ‘The German Business Cycle in the 1920s: A Comment’, Economic History Review, 30: 159–61. ——– (1983) ‘The Beginning of the Depression in Germany, 1927–1930. Investment and the Capital Market’, Economic History Review, 36: 395–414. ——– (1993) The Origins and Course of the German Economic Crisis (Berlin: Haude & Spener). Barkai, A. (1990) Nazi Economics (New Haven: Yale University Press). Benjamin, D. and L. Kochin (1979) ‘Searching for an Explanation of Unemployment in Interwar Britain’, Journal of Political Economy, 87: 441–78. Bernanke, B. (1983) ‘Nonmonetary Effects of the Financial Crisis in the Propagation of the Great Depression’, American Economic Review, 73: 257–76. ——– and H. Parkinson (1991) ‘Procyclical Labor Productivity and Competing Theories of the Business Cycle: Some Evidence from Interwar US Manufacturing Industries’, Journal of Political Economy, 99: 439–59. Blaich, F. (1977) Die Wirtschaftskrise 1925/6 und die Reichsregierung. Von der Erwerbslosenfürsorge zur Konjunkturpolitik (Kallmünz: Lassleben) Borchardt, K. (1979) ‘Zwangslagen und Handlungsspielräume in der grossen Wirtschaftskrise der frühen dreissiger Jahre. Zur Revision des überlieferten Geschichtsbildes’, in Jahrbuch der Bayerischen Akademie der Wissenschaften (1979): 1–46; translated in K. Borchardt, Perspectives on Modern German Economic History and Policy (Cambridge: Cambridge University Press, 1991): 143–60. ——– (1980) ‘Wirtschaftliche Ursachen des Scheiterns der Weimarer Republik’, in K.-D. Erdmann and H. Schulze (eds), Weimar, Selbstpreisgabe einer Demokratie: Eine Bilanz heute (Düsseldorf: Droste): 211—49. ——– (1984) ‘Could and Should Germany Have Followed Britain in Leaving the Gold Standard?’, Journal of European Economic History, 13: 471–98. ——– (1985) ‘Das Gewicht der Inflationsangst in den wirtschaftspolitischen Entscheidungsprozessen waehrend der Weltwirtschaftskrise’, in G.D. Feldman (ed.), Die Nachwirkungen der Inflation auf die deutsche Geschichte 1924–1933 (Munich: Oldenbourg): 233–60. ——– (1990) ‘A Decade of Debate About Bruening’s Economic Policy’, in J.v. Kruedener (ed.), Economic Crisis and Political Collapse. The Weimar Republic 1924–1933 (Oxford: Berg): 99–151. ——– and A. Ritschl (1992) ‘Could Brüning Have Done It? A Keynesian Model of Interwar Germany, 1925–1938’, European Economic Review, 36: 695–701. ——– and H.-O. Schötz (eds) (1991) Wirtschaftspolitik in der Krise. Die (Geheim-) Konferenz der Friedrich-List-Gesellschaft in September 1931 über Möglichkeiten und Folgen einer Kreditausweitung (Baden-Baden: Nomos). Bordo, M. and B. Eschweiler (1993) ‘Rules, Discretion, and Central Bank Independence: The German Experience, 1880–1989’, in P. Siklos (ed.), Varieties of Monetary Reform: Lessons and Experience on the Road to Monetary Union (Boston: Kluwer).

Albrecht Ritschl 139 Bordo, M. and F. Kydland (1995) ‘The Gold Standard As a Rule: An Essay in Exploration’, Explorations in Economic History, 32: 423–64. Bordo, M. , M. Edelstein and H. Rockoff (1998) ‘Was Adherence to the Gold Standard a “Good Housekeeping Seal of Approval” during the Interwar Period?’, mimeo. Bracher, K.-D. (1971) ‘Brünings unpolitische Politik und die Auflösung der Weimarer Republik’, Vierteljahrshefte für Zeitgeschichte, 19: 113–23. Broadberry, S. (1984) ‘Fiscal Policy in Britain During the 1930s’, Economic History Review, 37: 95–102. ——– and A. Ritschl (1995) ‘Real Wages, Productivity, and Unemployment in Britain and Germany During the 1920s’, Explorations in Economic History, 32: 327–49. Brown, E.C. (1956) ‘Fiscal Policy in the Thirties: A Reappraisal’, American Economic Review, 46: 857–79. Bulow, J. and K. Rogoff (1988) ‘Multilateral Negotiations for Rescheduling Developing-Country Debt’, Journal of Political Economy, 97: 155–78. ——– (1989) ‘A Constant Recontracting Model of Sovereign Debt’, Journal of Political Economy, 97: 155–78. Cairncross, A.K. and B. Eichengreen (1983) Sterling in Decline (Oxford: Blackwell). Chirinko, R. (1993) ‘Business Fixed Investment Spending: A Critical Survey of Modeling Strategies, Empirical Results, and Policy Implications’, Journal of Economic Literature, 31: 1875–911. Cohn, R. (1992) ‘Fiscal Policy in Germany during the Great Depression’, Explorations in Economic History, 29: 318–42. Eichengreen, B. (1992) Golden Fetters. The Gold Standard and the Great Depression 1919–1939 (New York: Oxford University Press). Erbe, R. (1958) Die nationalsozialistische Wirtschaftspolitik 1933–1939 im Lichte der modernen Theorie (Zürich: Polygraphischer Verlag). Falkus, M.E. (1975) ‘The German Business Cycle in the 1920s’, Economic History Review, 28: 451–65. Feldman, G.D. (1993) The Great Disorder: Politics, Economics and Society in the German Inflation 1914–1924 (New York/Oxford: Oxford University Press). Garvy, G. (1975) ‘Keynes and the Economic Activists of pre-Hitler Germany’, Journal of Political Economy, 83: 391–405. Grotkopp, W. (1954) Die Grosse Krise. Lehren aus der Überwindung der Weltwirtschaftskrise (Düsseldorf). Hagemann, H. (1984) ‘Lohnsenkungen als Mittel der Krisenbekämpfung? Überlegungen zum Beitrag der “Kieler Schule” in der beschäftigungspolitischen Diskussion am Ende der Weimarer Republik’, in H. Hagemann and H.D. Kurz (eds), Beschäftigung, Verteilung und Konjunktur. Festschrift fuer Adolphe Lowe (Bremen): 97–129. Hardach, G. (1976) Weltmarktorientierung und relative Stagnation. Währungspolitik in Deutschland 1924–1931 (Berlin: Duncker & Humblot). Hayashi, F. (1982) ‘Tobin’s Marginal q and Average q: A Neoclassical Interpretation’, Econometrica, 50: 213–24. Hertz-Eichenrode, D. (1982) Wirtschaftskrise und Arbeitsbeschaffung. Konjunkturpolitik 1925/26 und die Grundlagen der Krisenpolitik Brünings (Frankfurt: Campus). Heyde, P. (1998) Das Ende der Reparationen. Deutschland, Frankreich und der Youngplan 1929–1932 (Paderborn: Schoeningh). Holtfrerich, C.-L. (1982) ‘Alternativen zu Bruenings Politik in der Weltwirtschaftskrise?’, Historische Zeitschrift, 235: 605–31.

140 Germany ——– (1984) ‘Zu hohe Loehne in der Weimarer Republik? Bemerkungen zur Borchardt-These’, Geschichte und Gesellschaft, 10: 122–41. ——– (1986) The German Inflation 1914–1923: Causes and Effects in International Perspective (Berlin/New York: de Gruyter). ——– (1990) ‘Was the Policy of Deflation in Germany Unavoidable?’, in J.v. Kruedener (ed.), Economic Crisis and Political Collapse. The Weimar Republic 1924–1933 (Oxford: Berg): 63–80. James, H. (1985) The Reichsbank and Public Finance in Germany, 1924-1933: A Study of the Politics of Economics during the Great Depression (Frankfurt: Knapp). ——– (1986) The German Slump: Politics and Economics 1924–36 (Oxford: Clarendon). ——– (1998) ‘The Reichsbank, 1876–1945’, in Deutsche Bundesbank (ed.), Fifty Years of the Deutsche Mark (Oxford: Oxford University Press): 3–53. Jochmann, W. (1978) ‘Brünings Deflationspolitik und der Untergang der Weimarer Republik’, in D. Stegmann et al. (eds), Industrielle Gesellschaft und politisches System (Düsseldorf: Droste): 97–112. Keynes, J.M. (1919) The Economic Consequences of the Peace (London: Macmillan). ——– (1922) A Revision of the Treaty (London: Macmillan). ——– (1936) The General Theory of Employment, Interest and Money (London: Macmillan). Klug, A. (1993) The German Buybacks, 1932–1939: A Cure for Overhang? (Princeton: Princeton University Press. Princeton Studies in International Finance 75). Kroll, G. (1958) Von der Weltwirtschaftskrise zur Staatskonjunktur (Berlin: Duncker & Humblot). Kruedener, J.von (1990) ‘Could Brüning’s Policy of Deflation have been Successful?’, in J. von Kruedener (ed.), Economic Crisis and Political Collapse. The Weimar Republic 1924–1933 (Oxford: Berg): 81–98. Link, W. (1970) Die amerikanische Stabilisierungspolitik in Deutschland 1921–32 (Düsseldorf: Droste). Lucas, R. (1976) ‘Macroeconomic Policy Evaluation: A Critique’, Journal of Monetary Economics, 1: 19–46. Lüke, R. (1958) Von der Stabilisierung zur Krise (Zürich: Polygraphischer Verlag). McNeil, W.C. (1986) American Money and the Weimar Republic (New York: Columbia University Press). Mishkin, F. (1981) ‘The Real Interest Rate: An Empirical Investigation’, CarnegieRochester Conference Series on Public Policy, 15: 151–200. Mommsen, H. (1978) ‘Heinrich Brünings Politik als Reichskanzler: Das Scheitern eines politischen Alleinganges’, in K. Holl (ed.), Wirtschaftskrise und liberale Demokratie. Das Ende der Weimarer Republik und die gegenwärtige Situation (Göttingen: Vandenhoeck & Ruprecht): 16–45. Netzband, K. and H. Widmaier (1964) Währungs- und Finanzpolitik der Ära Luther 1923-1925 (Basel/Tübingen: Kyklos-Verlag/J.C.B. Mohr). Pagenkopf, H. (1981) Der Finanzausgleich im Bundesstaat (Stuttgart: Kohlhammer). Peppers, L. (1973) ‘Full Employment Surplus Analysis and Structural Change: The 1930’s’, Explorations in Economic History, 10: 197–210. Riel, A. van and A. Schram (1993) ‘Weimar Economic Decline, Nazi Economic Recovery, and the Stabilization of Political Dictatorship’, Journal of Economic History, 53: 71–105. Ritschl, A. (1990) ‘Zu hohe Löhne in der Weimarer Republik? Eine Auseinandersetzung mit Holtfrerichs Berechnungen zur Lohnposition der Arbeiterschaft 1925—1932’, Geschichte und Gesellschaft, 16: 375–402.

Albrecht Ritschl 141 ——– (1994) ‘Goldene Jahre? Zu den Investitionen in der Weimarer Republik’, Zeitschrift für Wirtschafts- und Sozialwissenschaften, 114: 99–111. ——– (1995) ‘Was Schacht Right? Reparations, the Young Plan, and the Great Depression in Germany’, mimeo, Universitat Pompeu Fabra. ——– (1996) ‘Sustainability of High Public Debt: What the Historical Record Shows’, Swedish Economic Policy Review, 3: 175–98. ——– (1998a) ‘Deficit Spending in the Nazi Recovery, 1933–1938: A Critical Reassessment’, mimeo, Universitat Pompeu Fabra. ——– (1998b) ‘Reparation Transfers, the Borchardt Hypothesis, and the Great Depression in Germany, 1929–1932: A Guided Tour for Hard-Headed Keynesians’, European Review of Economic History, 2: 49–72. ——– (2002) ‘International Capital Movements and the Onset of the Great Depression: Some International Evidence’, in H. James (ed.), The Interwar Depression in an International Context (Munich: Oldenbourg). Sanmann, H. (1965) ‘Daten und Alternativen der deutschen Wirtschafts- und Finanzpolitik in der Ära Bruening’, Hamburger Jahrbuch für Wirtschafts- und Gesellschaftspolitik, 10: 109–40. Sargent, T.J. (1986) Rational Expectations and Inflation (New York: Harper & Row) Schrecker, E. (1978) The Hired Money: The French Debt to the United States, 1917–1929 (New York: Arno Press). Schubert, A. (1991) The Credit-Anstalt Crisis of 1931 (New York: Cambridge University Press). Schuker, S. (1976) The End of French Predominance in Europe. The Financial Crisis of 1924 and the Adoption of the Dawes Plan (Chapel Hill: University of North Carolina Press). ——– (1988) American Reparations to Germany, 1924–1933 (Princeton: Princeton University Press. Princeton Studies in International Finance 61). Spoerer, M. (1994) ‘German Net Investment and the Cumulative Real Wage Position, 1925–1929: On a Premature Burial of the Borchardt Debate’, Historical Social Research, 19: 26–41. Spoerer, M. (1996) Von Scheingewinnen zum Ruestungsboom. Die Eigenkapitalrentabilität der deutschen Industrieaktiengesellschaften 1925–1941 (Stuttgart: Steiner). Stone. R. and W.M. Stone (1938–9) ‘The Marginal Propensity to Consume and the Multiplier. A Statistical Investigation’, Review of Economic Studies, 6: 1–24. Temin, P. (1971) ‘The Beginning of the Great Depression in Germany’, Economic History Review, 24: 240–48. Tilly, R. and W. Huck (1994) ‘Die deutsche Wirtschaft in der Krise 1924 bis 1934. Ein makrooekonomischer Ansatz’, in C. Buchheim, M. Hutter and H. James (eds), Zerrissene Zwischenkriegszeit, Knut Borchardt zum 65. Geburtstag (Baden-Baden: Nomos): 45–95. Voth, H.-J. (1993) ‘Investitionen in den “Goldenen Jahren” der Weimarer Republik’, Zeitschrift für Wirtschafts- und Sozialwissenschaften, 113: 629–33. ——– (1994) ‘Much Ado about Nothing? A Note on Investment and Wage Pressure in Weimar Germany, 1925–29’, Historical Social Research, 19: 124–39. ——– (1995) ‘Did High Wages or High Interest Rates Bring Down the Weimar Republic?’, Journal of Economic History, 55: 801–21. ——– (1998) ‘The True Cost of Inflation: Expectations and Policy Options During Germany’s Great Slump’, mimeo, Universitat Pompeu Fabra. ——– (1999) ‘With a Bang, Not a Wimper: Pricking Germany’s “Stockmarket Bubble” in 1927 and the Slide into Depression’, mimeo, Universitat Pompeu Fabra.

142 Germany Wendt, B.J. (1971) Economic Appeasement, Handel u. Finanz in der britischen Deutschland-Politik 1933-1939 (Düsseldorf: Bertels). White, E. (1999) ‘The Cost and Consequences of Napoleonic Reparations’, NBER Working Paper No. 7438. Witt, P.-C. (1982) ‘Finanzpolitik als Verfassungs- und Gesellschaftspolitik’, Geschichte und Gesellschaft, 8, 386–414.

6 The Interwar Slump in India: The Periphery in a Crisis of Empire* G. Balachandran

1

India in the Great Depression

India’s encounter with the Great Depression has unfortunately been of greater interest to students of the country’s history than to students of the interwar world economy. Historians of India have conventionally regarded the Depression as a turning point in the country’s political and economic history. Nationalist opposition to colonial rule intensified in the 1930s partly because of growing support from a peasantry burdened by falling prices and fixed rent and revenue liabilities. Thus, according to one view, the Depression ‘precipitated the end of empire’ (Rothermund, 1992). Nearly two decades later, the Depression also seemed to justify India’s resort to a strategy of import-substituting industrialization sustained by sales in the domestic market. In contrast, studies of the worldwide impact of the interwar Depression do not carry many references to India. Until recently, accounts of the slump revolved around the propagation of an ‘initial shock’ and individual countries’ resistance or vulnerability to that shock. As might be expected of a small economy, India did not set off any worldwide shocks. Nor did it actively resist the Depression because government policies, which were formulated mainly in conformity with London’s priorities and preferences, were determinedly pro-cyclical throughout these years. But even in the 1920s India was less dependent than other developing countries on foreign trade while exporting a more diversified basket of commodities. According to some estimates, its exports made up about 10 per cent of estimated national income and imports about 8 per cent (Chaudhuri, 1983, p. 804; Goswami, 1986, p. 165). The proportions for commodity trade excluding treasure were even lower. Thanks to the relatively large size of its domestic economy, endogenous stimuli to reallocation and expansion were also not entirely absent. Consequently, in the aggregate, the Indian economy did not passively succumb to the Depression. Given this record, it is perhaps not surprising that India should 143

144 India

have eluded attention in most ‘shock-response’ accounts of the interwar Depression. The Depression undoubtedly caused a steep decline in Indian trade. According to figures the Indian government submitted to the League of Nations, commodity exports at current prices bottomed out in 1932–3 at about 43 per cent of the 1929–30 level, while imports fell to slightly below half the 1929–30 level in 1933–4 before beginning to recover.1 Export prices (1927–8 = 100) plunged to 53.5 in 1933–4. The drop in import prices (to 62 in 1935–6) was less sharp though more prolonged. But the adverse movement in the terms of trade was relatively short-lived. The bulk of the fall in Indian export prices (from 90 to 59) took place between April 1929 and March 1931. In contrast import prices fell more steadily through the early 1930s and India’s terms of trade (1927–8 = 100) hit the trough at 82.6 in 1931–2, though doubtless recovery remained slow thereafter (Chaudhuri, 1983, p. 840). Broadly similar trends are discernible in the movement of agricultural and non-agricultural prices. Aggregate price movements confirm the intensity of the shock between 1929 and 1931, the weighted index of all commodity prices (1928 = 100) collapsing to 60 in 1931 and sliding to 52 in 1934 before beginning to recover (McAlpin, 1983, pp. 903f.). The Depression is also widely regarded as a grim event for the large majority in rural India. Commercialized segments of agriculture such as jute bore the brunt, but elsewhere, too, the slump took its toll. The burden of rural indebtedness, in particular, became more crushing. According to Goldsmith, the nominal value of agricultural debt increased from Rs 11,500 million in 1929 to 18,000 million in 1939. The rural debt/income ratio also doubled. A quarter of the farmer’s income went towards debt servicing in 1939, as against an eighth in 1929.2 Estimates of Indian national income vary widely.3 Yet there is widespread agreement that taken as a whole the Indian economy was not amongst the worst hit: to a considerable extent, unlike in the industrial economies, the Depression affected prices more than aggregate output.4 Although 1929–30 levels of real national income were not attained once again until 1934–5 (Sivasubramonian, 1965) or 1936–7 (Heston, 1983), the peak to trough (in both cases 1929–30 to 1931–2) variation works out to 1.3 per cent and 1.9 per cent respectively. (In contrast, Sivasubramonian and Heston estimate a real national income decline of about 6 and 10 per cent respectively between 1919–20 and 1920–1.) The rural credit system all but collapsed in the Depression. But contrary to trends elsewhere, the largely urban commercial banking system saw significant expansion during these years. There was even a minor stock-market boom after the early 1930s reflecting and sustaining an accelerated tempo of industrial growth and diversification which had its origins in a combination of higher tariffs and rising urban incomes, and which was assisted by low interest rates and a flexible reallo-

G. Balachandran 145

cation of resources from agriculture to industry. However, neither the nature and objectives of public intervention nor the growth and diversification of industry warrant comparison with countries following a more active policy.5 After over two decades of stagnation, India’s population (enumerated at regular decennial intervals since 1871) began growing from the 1920s. Population increased nearly 15 per cent between 1931 and 1941 and outstripped the growth in national income. According to Maddison’s (1971) estimates, real per capita incomes fell by about 3 per cent between 1931 and 1935 and about 5 per cent between 1926 and 1935. Maddison also estimates the decline to have been steeper over 1936–40 (about 5.5 per cent) and 1941–5 (about 7 per cent). Unlike Maddison, both Sivasubramonian and Heston see only a small growth in real per capita incomes in the 1920s which lasted till 1930. Yet even they estimate Indian income per capita to have fallen by less than 3 per cent between 1931 and 1935.6 Though not insignificant for a country with low incomes and widespread inequalities, this decline was mild by contemporary Latin American standards. Having said this, however, we must also note Maddison’s estimates of growth in other Asian economies of the order of about 4 per cent between 1929 and 1933.7 Middling outcomes do not make riveting headlines. They might make footnotes, however. Overlords in London allowed Indian policy-makers little autonomy. For example, they overruled Delhi to peg the rupee to the sterling when the latter left gold in September 1931. Besides, policy and performance in the colony shadowed Britain’s in the 1930s and are sometimes discussed in the imperial context (e.g. Drummond, 1981). Contemporaries, however, took a different and rather wider view. In the 1920s, British policy-makers closely oversaw economic management in India to ensure that the colony’s policies did not conflict with the needs of European stabilization, including that of the pound sterling. W.A. Brown Jr, a perceptive contemporary commentator on the gold standard, pointed out on the eve of the Depression that India was one of the three countries – South Africa and the US being the other two – which were of ‘chief interest and significance in a study of the restoration of the gold standard after the war’ (Brown, 1929, pp. 12, 42f.). As the Depression intensified and capital outflows resumed in 1931, officials in London believed the pound would not survive a fall in the rupee and pressed strong deflationary measures on the colony. Finally, India began exporting gold in considerable quantities after September 1931. Until the mid-1930s British observers regarded these exports as the principal expansionary influence on the world economy and sterling policy was partly motivated towards sustaining them. In recent years we have grown familiar with a wider canvas where the Depression is relocated within its interwar context of economic dislocation and restructuring in Europe, economic and political friction, flawed

146 India

management by governments challenged by unemployment and payments imbalances, and the absence or failure of policy coordination. Regimes and ideologies have also received their fair share of attention. There has been renewed stress, in particular, on the role of the gold standard, with both Temin (1989) and Eichengreen (1992) attributing the intensity of the slump, if not its origins, to policy-makers’ faith in the prevailing theology on gold. A great virtue of the new focus is that it evokes contemporary perceptions and concerns (though naturally not the diagnosis). It also provides a suitable framework within which to reassess India’s significance to the interwar world economy and its encounter with the Depression. Framing India’s Depression experience in a wider multilateral setting also helps shed interesting new light on British external economic priorities and strategies during these years. It is sometimes suggested that Britain sought the resolution of its formidable external economic and political problems between the wars either in a policy of ‘internationalism’ or ‘imperialism’ (‘empire integration’).8 Thus, according to a common variant of this argument, while Britain was resolutely ‘internationalist’ in the 1920s, it began to forge closer economic ties with the empire as the world economy moved into Depression. The Indian experience however reveals a more sure-footed (and less schizoid) interwar British effort to meld the two approaches into a coherent policy bent at all times on leveraging the resources of the empire to restore or boost Britain’s international economic position. Such a policy was dictated in this instance by India’s importance to the smooth functioning of the world economy, and enabled by Britain’s control over macroeconomic management in the colony. Following Keynes, British officials believed India played a counter-cyclical role in the world economy. A central aim of this chapter is to identify and elucidate this role and relate Britain’s interwar efforts to regulate it against the background (assumed for the most part) of European – especially British – economic and financial instability. The resulting effects on the Indian economy will be addressed briefly, and finally, the wider resonances of the Indian story, in particular for the other parts of the analysis advanced by Temin and Eichengreen.

2

Modelling the Indian demand for gold

Keynes was perhaps the first commentator to draw attention to India’s counter-cyclical role in the contemporary world economy. India had long been regarded as a ‘sink’ for precious metals. After the crash in silver prices and the closure of local mints to silver in 1893, India’s taste for precious metals turned increasingly towards gold. The metal was both a commodity with social and ritual uses (though Benjamin Strong on a visit to India in 1920 was stretching a metaphor when he reported that Indian women could

G. Balachandran 147

not be buried without their gold jewellery) and a store of value. As incomes rose in India on the back of a worldwide increase in primary product prices, Indian imports of gold grew rapidly after the turn of the century. Exchange intervention – i.e. selling rupees (or ‘council bills’ in contemporary parlance, and other instruments) to intending remitters more cheaply than they could set down gold in India – reduced but did not eliminate gold flows to the colony.9 According to estimates presented by the London gold dealer Joseph Kitchin to the Macmillan Committee, in a good year before the war India might import a quarter of the world’s gold output.10 A few winters before the First World War Indian gold imports began to cause great anxiety in London as the major factor behind the recent hardening of interest rates in the City. Writing his first book – Indian Currency and Finance – in the midst of a controversy over Whitehall’s management of Indian currency affairs, Keynes (1911 [1971], pp. 70f) argued that while Indian gold imports might cause short-term strains in London, particularly when they occurred during an ‘inconvenient week’, on the whole a ‘creditor nation’ like Britain should welcome their contribution to dampening inflation. India, he declared, was a ‘true friend to the City’ and an ‘enemy’ of inflation. Keynes’s implicit model of production, consumption, and saving by Indian households was quite simple. Although foreign trade made a limited contribution to its national income, India was a small open economy in 1913: it was a global price-taker in most commodities, and with no restrictions on the external account, prices of Indian tradeables broadly reflected trends in world prices. More than half the national income and nearly three-quarters of employment were accounted for by a largely peasant agriculture. In this setting, a rise in international prices and levels of business activity helped raise prices in India. Provided harvests did not fail, this also translated into higher disposable incomes for a mainly peasant-based economy.11 These incomes were either spent on tradeables (say cloth) or saved. In the absence of a developed banking system, savings were principally held in gold and silver. While poorer households could perhaps only afford silver, those who could do so bought gold because its prices were now more stable. Setting silver aside for the present, households’ decisions to buy gold depended in the first instance on their incomes, i.e. the prices fetched by their produce. Policy-makers who relied on this insight to frame policy in interwar India embellished it from wartime experience to take relative prices and expectations into account. Indian households, they noted, preferred gold to cloth if the latter’s prices were high and were expected to fall. Whether Indian householders based their decisions on incomes or price expectations, a global boom was believed to deliver larger quantities of gold to India. These gold flows, Keynes argued, helped arrest the boom before it turned inflationary.

148 India

Not a great votary of the yellow metal, Keynes believed changes in local tastes and banking habits would soon make India disgorge its hoards. But his analysis also implied that gold flows to India would be checked or reversed in a slump. Keynes’s analysis was put to the test not long afterwards during and immediately after the First World War. Though currency crises and controversies in India and a temporary return to bimetallism complicated the picture, the response of Indian households to wartime inflationary expectations (thanks to good harvests and trading controls, retail prices of Indian foodgrains and exports did not increase greatly until 1917) and postwar inflation largely confirmed Keynes’s view. By 1920, as Britain grappled with a falling pound in the midst of rising unemployment, the economist’s insight began to provide the discourse and overt basis for policy.

3

Managing the rupee 1914–20

The wartime growth of the Indian economy was far from spectacular (Heston, 1983, pp. 397–9). Exports and trade surpluses were also both lower in 1914/15–1918/19 than in the preceding five years.12 But as prices the world over increased as a result of the war, India’s demand for precious metals became more pressing. With Britain needing all the gold it could get to finance imports from the US, gold exports to India were restricted, at first by the Bank of England exerting moral pressure on traders and banks wishing to remit metal to the colony. When this failed to have the desired impact, the Governor of the Bank, Walter Cunliffe, used the threat of higher ‘money rates’ to prevail upon Whitehall to impose physical controls on gold movements to India.13 Thanks to these restrictions, gold traded on the private account liquidated a mere tenth of India’s wartime trade surplus (as against over a third before the war). Consequently, the burden of financing Indian trade and meeting its economy’s need for precious metals was thrown on to silver. Rising demand for the metal and difficulties of increasing supply in the short term had already led to a surge in world silver prices. The Indian demand led to their further hardening.14 Until the early 1920s, India followed a de facto currency board system in which remittances from abroad – whether sterling bought and accumulated in London or gold landed in India – determined currency expansion. This system endured the war, even if expansion took place almost entirely against sterling assets during these years. But the worldwide shortage of precious metals threatened to undo the Indian currency system in other ways. Until the mid-1930s when bank deposits rose markedly, currency made up the overwhelming (and a largely stable) proportion of money supply in India. Currency comprised notes and token silver rupees. A concession to

G. Balachandran 149

the presumed Indian preference for metallic coinage, silver rupees were available against currency notes from the government’s currency offices. To ensure the conversion of notes into silver, the government maintained a paper currency reserve in India which until the war comprised mainly silver. But difficulties of procuring the metal meant that silver holdings in this reserve failed to keep pace with currency requirements during the war. Since moderating currency expansion in line with silver supplies would have meant an unacceptable monetary squeeze on wartime trade, the note issue in India now expanded against UK Treasury Bills. Soon these securities came to account for the larger part of the paper currency reserve, giving rise to fears of a run on the government’s stocks of silver. There were good reasons for such fears. As gold grew scarce, silver became the preferred store of value in India. Demand for the metal was also reinforced by inflationary expectations, while wartime restrictions on trading in the metal and rising silver prices (which also increased the metallic worth of silver rupees and undermined their token character) directed the demand for silver towards the government’s currency offices. The obligation to ‘convert’ currency notes into silver coins bearing the head of the king-emperor was widely regarded as the linchpin of the colony’s currency system and the basis of public faith in it. By the end of 1916 it seemed only a matter of time before the linchpin broke and the wheels came off the Indian currency system. Domestic ‘inconvertibility’ in India was feared to be little short of a disaster. It would put an immediate end to the supply of soldiers and war materials from India. The government’s borrowing programme might never recover from the resulting blow to public confidence. By vindicating popular faith in precious metals as the most reliable long-term store of value it would indefinitely retard India’s monetary and financial development and stimulate larger gold flows to the colony. Finally, the collapse might provoke peasant and industrial unrest and boost political agitation. For these reasons, some officials felt, rupee inconvertibility would deal a greater blow to British prestige in India than a military defeat, than even a German landing at Norfolk!15 It was possible to relieve the danger by revaluing the rupee since doing so would lower the rupee price of imported silver and the metallic value of the token. Indeed, as world silver prices rose, the rupee was revalued in two stages from the pre-war 16d. (or old pence) to 18d. in April 1917.16 But war having made the demand for Indian exports inelastic, further revaluations merely threatened to embarrass Britain and its allies without promising lasting relief in India. As Indian purchases of silver in the world market only pushed up its price and worsened the colony’s problems, an international agreement to stabilize the metal was crucial to the prospects for currency stability in wartime India. The US held the key to such an agreement: it was a major

150 India

silver producer, it was capable of augmenting supply immediately as the Treasury Department held large stocks of the metal, and as Britain’s wartime ally and supplier, its government maintained an interest in the smooth flow of raw materials from India. After some negotiation Britain and the US reached an agreement in April 1917 under which the latter agreed to melt a large part of its Treasury’s stock of silver dollars for export to India. Silver prices were stabilized at one dollar per ounce under the deal which was to last till the end of the war. After the war, as trade and prices came under the grip of the postwar boom, India continued to run large surpluses and accumulate sterling reserves. Nor did the colony’s demand for silver – or its domestic ‘convertibility’ crisis – abate. However, with the cap on its price removed following the cessation of hostilities, silver resumed its dizzy climb. The rupee, stabilized at 18d. under the Anglo-American agreement, was moved up in stages to protect the token character of silver coinage in India and it soon exceeded 22d. Successive revaluations of the rupee may have helped preserve silver coinage in India but they aggravated the colony’s gold problem. Thanks to import curbs and unrequited demand, the metal now sold in the colony at a considerable premium over world prices. Each successive revaluation of the rupee lowered the nominal rupee price of imported gold, widened this premium, and postponed the restoration of free trade in the metal and the gold standard in India. Against this background, an expert committee was constituted in Britain in 1919 to help stabilize the Indian currency system. Initially the committee’s explicit concern was to preserve silver coinage in India, and its preferred means of doing so was a steep revaluation of the rupee. More than halfway through its deliberations, in the course of hearing Keynes’s testimony, the committee (and the India Office) stumbled upon an alternative justification for a steep rupee revaluation, namely, inflationary pressures in the colony. Though only a cloak for a policy intended to get India out of Britain’s path towards postwar stabilization and recovery, the new justification had the advantage of helping to frame the colony’s currency problem in more general terms. Besides, as Keynes pointed out to the committee, appearing to fight inflation made shrewder politics. Thus in the end, the committee’s chosen instrument and Keynes’s justification for it came together, and citing inflationary pressures in India the committee recommended revaluing the rupee to two shillings (gold).17 When implemented early in February 1920, this rate translated into nearly three shillings sterling because of the fall meanwhile in the pound–dollar (or gold) parity. Despite enormous reserve losses sustained in intervention this rate carried little conviction in the markets: the rupee’s gold peg could not be sustained and it went into virtual free fall. The new rate also led to commodity exports collapsing from Rs 3,300 million to 2,600 million. With imports ballooning from Rs 2,100 million to 3,360 million (a significant proportion

G. Balachandran 151

of these imports accumulated in warehouses as domestic demand plunged in the aftermath of the shock) a surplus of Rs 1,200 million in 1919/20 became a deficit of over Rs 750 million in 1920/1.18 Prices (weighted basket of all commodities) dropped nearly 10 per cent between 1920 and 1921. Particularly affected were prices of non-agricultural commodities which fell almost 30 per cent (McAlpin, 1983, pp. 203f). There was also a steep decline in real national income of between 6 (Sivasubramonian) to 10 (Heston) per cent (Heston, 1983, pp. 397–9). As anticipated, the Indian demand for gold lost its edge. Moreover, as households liquidated their metallic balances to cope with the income effects of the shock, the earlier condition of excess demand for gold and silver was abruptly reversed. Not only was the ‘convertibility’ problem overcome, India briefly emerged alongside some other countries as a net exporter of gold. It is not possible, given the nature of the available estimates and data, to separate the effects of the policy shock from those of the collapse of the postwar boom. But it is worth noting that the run on the rupee and the fall in world silver prices began in February 1920, i.e. some months before the onset of the slump.

4

The high rupee and Britain’s agenda 1919–20

The 1920 stabilization package is noteworthy for two reasons. First, it marked the explication of a policy discourse which thereafter made curbing inflation the major objective of economic policy in interwar India. This, as pointed out above, was Keynes’s singular contribution to the 1920 committee. More significantly for the argument presented in this chapter, the stabilization package signified the first effort to use short-term policy to disengage India from global expansionary pressures. A discourse focused on inflation helped justify this disengagement. It is widely recognized that while some halting measures were taken, especially in Britain which faced a steep slide in the sterling, to rein in inflation by November 1919, the postwar boom went practically unchecked until late in the spring or nearly the summer of 1920. It is also widely recognized that with its precarious balance of payments position having rendered Britain’s ability to influence world trade and employment weaker than before the war, its leaders sought an expansionary global environment to enable sterling to appreciate (and after 1925 stay on the gold standard) with minimum loss of employment. Only the US had the ability now to follow expansionary policies without triggering a balance of payments crisis, and British officials spared no effort to urge this course upon their Atlantic neighbour. Urging inflation on the US and deflation on India might seem at first glance a conflicting stance for Britain to adopt. But the key to understand-

152 India

ing the apparent contradiction, reconciled in practical terms in 1920 by the intended severity of the Indian shock, is that both outcomes were felt to be necessary to ease monetary conditions in Europe and particularly in Britain. As Basil Blackett, Controller of Finance at the British Treasury, explained in a lengthy memorandum he wrote on the Indian currency committee’s recommendations, higher commodity prices represented ‘the one element of hope in the present confusion’ since they lowered the value of war debts and reduced ‘the distance to be travelled from a paper to a gold standard’: The policy of the British Government … is to try and get back to the gold standard as soon as … possible without a violent break in prices the social consequences of which were disastrous. So long as the purchasing power of gold is only about 9s. in the sterling as compared with pre-war prices … [i.e. commodity prices were more than double their prewar levels – GB], the ideal of a restoration of the gold standard in this country is reasonably within reach. If India is allowed to set up an effective demand for gold …, all hope of restoring sterling to parity must be abandoned. Blackett underlined that India should therefore agree not to import gold freely ‘for an indefinite period’. Then, in a markedly bullionist vein, he argued that, if the colony continued to press its claim to gold, the rest of the world would be bound in self-defence … to take steps to restrict their purchases from India. British rule in India and the consequent internal prosperity have built up Indian commerce, till India has become a creditor country on an important scale. In the world as it is at present organized, a creditor country … becomes a menace to the world unless it adopts the recognized method of foregoing present enjoyment of a part of its income for the sake of investing it in new capital developments at home and abroad. If India demands payment in precious metals to an extent that is unreasonable, the British Government will be doing a doubtful service to the world in protecting India against the marauding invaders whose raids in olden times served the purpose of relieving her of superfluous hoards of gold.19 Unfortunately for Britain, while it succeeded in preventing India from taking gold in 1920, it was less successful in persuading the United States to resume expansionary policies until 1923.

5

Keeping the rupee up 1920–5

The global economic expansion which Britain sought and anticipated framed London’s approach towards Indian economic policy until the mid-

G. Balachandran 153

1920s. Already in 1920, Ralph Hawtrey, Director of Financial Enquiries at the British Treasury, had noted that though welcome, higher commodity prices would also tend to increase the worldwide demand for gold. ‘This is particularly conspicuous in India … where absorption is expected to be very large’, he warned.20 As in 1919–20 so through the first half of the following decade, officials in London viewed Indian gold imports as a menace to economic revival. But the menace might be controlled if India were disengaged from an expanding world economy. Securing this disengagement thus became an important objective of British policy in India in the 1920s. The exchange rate was the principal instrument for the purpose. The rupee having collapsed to about 11d. gold by the spring of 1921, policy became geared towards achieving its appreciation. To this end the government – whose intervention represented the biggest influence upon it – kept off the exchange market entirely until 1923, preferring to finance its sterling obligations through recourse to borrowing in London rather than through sales of council bills. By 1923 the rupee had risen to 16d. sterling. The limits of contraction and foreign borrowing were also becoming apparent and domestic opinion favoured pegging the rupee at this rate which would equal the pre-war rate once Britain returned to the gold standard. Though the government resumed buying sterling, with American expansion barely under way its intervention was modest and commitment to further appreciation unshaken. As Blackett who was now the Finance Member of the Government of India declared to a Calcutta business audience, ‘premature stabilization’ of the rupee at 16d. would make gold ‘look cheap’ and encourage bullion imports.21 By dint of its deflationary monetary policies (currency was not expanded sufficiently or even contracted and the quasicentral Imperial Bank of India’s busy-season – i.e. principal harvest and export season – lending rate regularly touched the exceptional pre-war peak of 8 per cent and climbed to 9 per cent in February) the rupee was raised to 18d. by the winter of 1924.22 As the Indian government acknowledged to the 1926 currency commission, this rise was possible only because the ‘supply’ of currency was kept below the ‘demand’ for it.23 With the pound also rising in the months preceding its return to gold, the rupee was effectively stabilized at 18d. gold from May 1925. Despite pressure from London, including from Montagu Norman who uncharacteristically urged India to attend to stabilizing prices rather than the exchange rate until European currencies had returned securely to the gold standard, further appreciation proved impossible because of growing public outrage in the colony over the government’s policies.24 The 18d. rate was endorsed by the Hilton-Young Commission in 1926. India, as it happened, was the only country in the world to revalue its currency after the war.

154 India

The majority on the commission was persuaded about the stability of the recommended rate by the decline in prices which followed the rupee reaching 18d. (sterling) at the end of 1924 and the absence of active supporting intervention at that rate. But even in March 1926 the 18d. rate was threatening to come unstuck. Publicly the government blamed this on speculation but, in private, officials conceded that a 18d. rupee appeared stable when harvests were plentiful and exports were rising. But one indifferent harvest and export season had proved enough to depress the rate.25 It was too late, however, to backtrack from recent policy. Though the rupee strengthened in March 1926 after the authorities warned against speculation and threatened to intervene, unstable exchange market conditions persisted through the next twelve months and necessitated substantial currency contraction. Unable to buy sterling even to meet its immediate remittance needs during much of the 1926/7 busy season despite high interest rates and a tight rein over currency expansion because of fears of weakening the rupee, the government liquidated reserves (and contracted currency) to pay its bills in London. Similar conditions obtained during the following busy season, with the government being forced to sell sterling in December 1927 at the height of the export season to steady the rupee. These measures proved temporary. Not only did the rupee never reach the gold-import point during the years that followed, it remained below par throughout the month during eight and six months respectively in 1927/8 and 1928/9; only in five of the remaining ten months did it stay at or above 18d. throughout the month. In 1929/30 and 1930/1 it never reached 18 pence.26 Not surprisingly, the government was forced to dip into reserves or borrow abroad to pay its way in London. No long-term loans were raised between 1924 and 1927, but the next three years saw the Indian government borrowing about £35 million in London to finance itself, besides raising £11 million through short-term bills. Reserves were also depleted by nearly £20 million.27

6

Indian gold imports 1923–8

As Keynes recognized in A Tract on Monetary Reform, India’s exchange rate policy largely succeeded in insulating it from inflationary pressures in the early 1920s.28 Indian prices fell steadily from 1921 to 1924 before rising by nearly 3 per cent in 1925. Then followed a downward trend which was generally uninterrupted until 1933.29 Real incomes, though rallying smartly for two years from the 1920–1 trough, largely stagnated thereafter for the rest of the decade, significant growth being confined to two years – 1924–5 and 1929–30. Expectedly, the growth rate of national income between April 1921 and March 1930 averaged less than 2 per cent per annum.30 Thus, as

G. Balachandran 155

Aldcroft (1977, p. 216) noted, owing to its deflationary exchange rate and monetary policies the booms of the 1920s largely bypassed India. Despite the deflation, India imported large quantities of gold between 1923 and 1926. These imports undoubtedly reinforced deflationary impulses in the economy: the official gold price being below the market price, gold imports, besides diverting expenditures from other forms of consumption and investment, did not trigger currency expansion. Several explanations may be invoked for the failure of policy to stem gold inflows. With the appreciation of the rupee and the pound, there was a decline through these years in the nominal rupee price of gold. Thanks to an expanding world economy and a sequence of four good crop years, Indian exports also grew rapidly after 1921. In 1923 the value of Indian exports exceeded the earlier peak reached at the height of the postwar boom in 1919 and stayed above this level through the next two years. Agricultural prices recovered after 1923, and an appreciating rupee may have briefly helped improve India’s terms of trade during these years of rapid worldwide revival. Apart from coinciding with a period of expanding world trade and incomes, the rate of currency appreciation during the 1920s was perhaps too slow to depress incomes sufficiently to check the encouragement to gold imports resulting from the decline in the metal’s nominal rupee price. Besides, even if they experienced some income losses, Indian households whose annual gold purchases during 1915–22 averaged about a third of their purchases before the war may have reduced or postponed consumption to replenish their stocks of the metal. The high price of manufactures during the mid-1920s and the presence of deflationary expectations may have also together strengthened the move towards gold.31 Finally, falling silver prices would have caused portfolio shifts towards the superior metal. Attributing the Indian demand for gold during these years to the low real price of the metal, a recent study argues that a global revaluation of gold would, among other things, have served to arrest the Indian demand (Johnson, 1997, esp. ch. 3). Had it been possible, a revaluation of gold (or a coordinated devaluation of the major currencies) would undoubtedly have augmented international liquidity and helped ease post-First World War instability. But as the above account suggests, Indian gold imports were prey to many influences, and the effect of a gold revaluation on such imports would have depended on more factors than merely its impact on relative prices and the latter’s impact on the demand for gold.32 India’s trade surpluses and gold imports began to decline from the mid1920s. Though the world economy was largely stable or growing until 1928, the Indian trade surplus halved between 1925/6 and the following two years. Indian gold imports fell from a peak of £52 million in 1924 to £28 million in 1925, and to an average of about £15 million per year over

156 India

the next four years. And by January 1928 officials in New York could confidently predict that future Indian gold imports would be ‘spasmodic and in small quantities’.33 In addition to the above factors, in 1928–9 India faced another disturbing medium-term portent. This was the bunching of maturities in the next three years of sterling loans floated during 1921–3 to help the government pay its way in London.34 Thus the Indian economy was already in a state of stress when the Depression arrived.

7

Gold exports and the exchange rate in the Depression

As pointed out above, the slump was particularly intense in India during the first two years. While pro-cyclical policies reinforced the effect of declining trade and capital flight, the slump was cushioned in India after 1931 by Britain leaving the gold standard, sterling depreciation, and the large exports of gold from the colony. Enough has been said above about the effects of the Depression on Indian trade. Let us note, in addition, that India ran a trade deficit only once during these years, in 1932–3. Thanks to falling gold imports and service outflows, India’s current account deficit also fell sharply from Rs 488 million (or 0.88 per cent of estimated national income) in 1926/7 to Rs 100 million (0.17 per cent) in 1929/30. But a shrinking trade surplus led to the current account gap increasing to Rs 250 million in 1930/1. This amounted, however, to less than half of one per cent of national income. Thanks to its gold exports, India ran current account surpluses for the next six years.35 Capital flows posed the more difficult challenge. Between 1927 and 1929, short-term inflows covered the larger part of India’s current account deficit. From 1929–30, however, India experienced a net outflow of short-term capital which rose from about Rs 30 million that year to Rs 180 million the following year and to nearly Rs 400 million in 1931–2. These outflows were, of course, not peculiar to India. Apart from the factors which made for uncertain conditions in currency markets everywhere, the rupee was also affected by the run on the pound to which it was anchored. As the latter weakened against the dollar and the franc, a flight from the rupee may have reflected the ongoing flight from the sterling area. As capital outflows grew, the Indian authorities had little choice than to meet them in the best possible way. Some officials proposed controls but Whitehall and the Bank of England would have nothing to do with such suggestions until September 1931. Maintaining India’s uninterrupted repatriation of short-term balances and the servicing of its sterling obligations were both equally vital in London’s scheme of things, and other objectives were subordinated to these.36 It is significant, for instance, that India’s

G. Balachandran 157

ability to service its capital and current account obligations were the only criteria Whitehall officials ever used to assess the colony’s ‘recovery’ in the Depression. Monetary and fiscal policies were thus geared towards financing these outflows. London’s financial priorities in India were dominated by its own problems. A weak current account, for example, greatly enhanced the importance of transfers from the rest of the world including India.37 With its ability to attract capital from the other major centres in jeopardy, London now relied increasingly on free capital inflows from the empire. Finally, an Indian default would have had catastrophic consequences for sterling. As Whitehall officials admitted to their counterparts in Delhi after swearing them to the ‘utmost secrecy’ only days before Britain left the gold standard, the ‘sterling and rupee are intimately connected and … collapse of one might bring down the other’.38 As the rupee came under pressure, the exchange rate controversy, which had lurked near the surface since the mid-1920s, reared its head again. Indian businessmen and politicians, officials in Delhi, and even some Whitehall advisers, favoured a devaluation of the rupee. But not surprisingly in the light of our discussion of policy in the 1920s, the Bank of England and the India Office steadfastly refused to allow a parity change. Official arguments against devaluation were based on the assumption of a fully monetized (Indian) economy characterized by universal and nearly instantaneous adjustments to exchange rate changes.39 In 1933 the India Office presented a model to the League of Nations according to which less than a third of transactions in India were mediated through the market and two-thirds of marketed farm output went to settle rent, debt, and revenue obligations which were fixed in nominal terms. Other payments including wages were made in kind. In October 1931 Indian businessmen invoked similar premises to argue that rupee devaluation would benefit the large majority of rural Indian households. But only for their India Office interlocutors to reject the argument by embellishing its counter-model of a fully monetized economy with an agricultural wage-bargaining process characterized by downward real wage rigidity!40 The Bank of England used similar arguments, though interestingly, a memorandum Norman commissioned on the rupee also suggested that devaluation, presumably by promoting recovery, might well increase gold flows to India.41 With devaluation rejected, the authorities could only borrow abroad to replenish reserves and enforce more contraction in India. Even in 1928 India’s dependence on London for loans had irked Norman who saw it as a symptom of the government’s failure to contract credit. Despite Norman’s reservations, the Indian government borrowed £31 million in three issues between May 1930 and February 1931. But in May 1931 nearly two-thirds of another loan issue for £10 million devolved to the underwriters. Hence

158 India

the pressure grew to intensify contraction, Norman seeking ‘money famine’ in India ‘to frighten bears of exchange’.42 Trade and policy contractions caused gross currency circulation to fall from Rs 1,867 million in September 1929 to Rs 1,487 million in September 1931.43 For all but four months during this period (when it was 5 per cent), the Imperial Bank rate remained at or above 6 per cent (Reserve Bank of India, 1954, p. 693). The Depression, particularly the decline in customs revenues, also affected the government’s budgetary position. Actual revenues fell short of estimates by about Rs 110 million in 1930/1, and with actual expenditure exceeding estimates, the year’s accounts yielded a deficit of Rs 115 million (or 0.2 per cent of national income). Although duties on many goods were increased and expenditures slashed in the budget of February 1931 and in the emergency budget of September 1931, this year too ended with a deficit of similar magnitude.44 There were two further problems. The Imperial Bank’s efforts to tighten the market were neutralized by banks discounting government paper. But the Imperial Bank could not discourage these discount practices without further weakening the market for government securities already damaged by high interest rates, or affecting profitability and dividend payouts to shareholders. Secondly, the Depression and government policy adversely affected incomes and forced Indian households to sell their gold holdings to settle dues and debts and finance consumption. The selling tendency, which set in about July 1929, grew quite pronounced by the end of the year. By the end of 1930 upcountry centres were reportedly sending 1,600 ounces of gold to Bombay alone every day. The government bought the bulk because the bazaar price was now lower than the official price, and by early 1931 monthly gold flows into the mint amounted to nearly £1 million.45 These receipts augmented the government’s capacity for intervention. But in the meantime they threatened to undo the monetary contraction achieved so far. The reserve losses of 1929–31 and the seemingly spontaneous offsetting gold sales by Indian households invite comparison with the 1920 episode. In 1919–20 reserves accumulated in London while the large demand for gold and silver by Indian households went unmet. Rather than satisfying this demand, Britain preferred to administer a strong deflationary shock which reduced incomes and prices in India and smothered the local demand for precious metals. A converse process was at work in the Depression as gold accumulations in India offset its government’s reserve losses in London. A repetition of the counter-cyclical precedent of 1920 (and the succeeding years) would have involved some devaluation to restore trade and incomes if not reassure the markets. As the devaluation ran its course, improved

G. Balachandran 159

exports and capital inflows may have helped boost reserves in London, while domestic expansion would have seen to the release of gold in the government’s vaults in India.46 But the authorities ceased to be sensitive to price stability as prices began to fall so that policy – which had been counter-cyclical in inflations – now turned pro-cyclical. As a consequence Indian economic policies remained persistently deflationary throughout the interwar years.

8

London policy and Indian gold exports

Though policy in India in the 1930s might appear to be inconsistent with that in the early 1920s, the outcomes were similar. As in 1920, so in the 1930s, Indian households liquidated their savings to finance consumption and settle debts. In so doing, they once again financed the transfer of liquid resources to Britain. When the pound came off the gold standard London overruled Delhi’s preference for an independently floating rupee and had its way over retaining the currency’s sterling peg (Schuster, 1979, p. 115; Rothermund, 1992, p. 43). The depreciation of the sterling, and alongside it of the rupee, led to a rise in the nominal price of gold in these two currencies. In addition, gold prices being higher in London than at home, India began exporting large quantities of the metal to London from September 1931.47 In all, Indian gold exports during 1931–8 exceeded £250 million, or about a third of South Africa’s exports during the same period.48 Lower Indian gold prices relative to those prevailing in London reflected the influence of private reserve liquidation in easing local supplies of the metal. Though the post-September 1931 increase in gold prices would undoubtedly have motivated some sales, for most sellers the higher prices were only ‘a mitigation of … [their] misfortune’ (Indian Currency and Exchange, 1935, p. 30). Note here that Indian gold sales commenced in 1929 and exports from the end of 1930. Indian commodity prices continued to fall until 1933–4 and did not return to their 1931 levels until nearly the end of the decade. Though industry showed signs of growth and one cannot rule out investors moving resources from gold to more rewarding uses, there is no evidence of any respite in agriculture and so no reason to suppose that distress sales ceased after September 1931. Indian gold exports came as a source of relief and rejoicing in London. While the Secretary of State for India was congratulated by the cabinet for his ‘accomplish(ment)’, others in the City were quick to note the greater freedom of policy now afforded Britain.49 As Neville Chamberlain, Britain’s Chancellor of the Exchequer, wrote to his sister: [the] astonishing gold mine we have discovered in India has put us in clover. The French can take their balances away without our flinching.

160 India

We can accumulate credits for the repayment of our £80m loan and we can safely lower the bank rate. So there is great rejoicing in the City.50 Authorities in London did everything they could to encourage these outflows. Even an export tax on gold was disallowed in India despite the government’s budgetary problems, not to mention suggestions to promote expansion through official intervention in the local gold market. The Bank of England was quick to activate banking channels to arrange for intervention on its behalf in the Bombay gold market.51 The British preference for using the exchange equalization account (EEA) to undertake intervention and hold down the pound is also widely recognized (Howson, 1980; Eichengreen, 1992, p. 303). But apart from their other obvious and well-advertised advantages, an important justification for persisting with the managed float regime and a low pound at least in 1932 was that they promised to get more gold out of India. As Frederick Phillips, the chief treasury adviser who opposed raising or stabilizing the pound observed in May 1932, the EEA helped to ‘draw out gold from hoards here and in India as has been seen in recent months. This results in an addition to gold stocks available for monetary purposes in the longer run.’52 Earlier in February 1932 Phillips argued for a low pound, citing among other factors the encouragement it gave to Indian gold exports: the most desired objective is a general rise in world … prices and … the most single powerful force to that end at the moment is the flow of gold from India. … that flow depends … on the depreciation of the rupee, that is the depreciation of the sterling. At $3.60 or $3.70, gold exports from India would not be checked, but at $3.90 they could cease. ‘A rise in the sterling would have a deadening effect [on Indian gold exports] which would be most unfortunate.’53 The availability of Indian gold, in turn, enabled smoother management of the pound sterling in a climate marked by impending or actual competitive depreciation of the major currencies.54 But within India itself, neither the low sterling nor the liquidation and export of gold helped raise prices. Officials in London were relieved to see India reduce foreign debt and replenish exchange reserves, but remained oblivious to other indicators of the continuing slump.55 In January 1932, officials at the Bank of England were worried about Indian prices rising on the back of gold exports and low interest rates.56 Shortly afterwards the Indian government urged Whitehall to manage the pound with an eye to the interests of primary producers and sought a package to stimulate the economies of the sterling area and relieve agrarian distress in India. But these pleas fell on deaf ears since India’s ability to meet its external liabilities was assured (‘430 million of gold absorbed in last 70 years, 55 million

G. Balachandran 161

so far exported’, a treasury official noted on the margin of an Indian memorandum). In any case, the treasury declared, ‘the tail can’t expect to wag the Bulldog’.57

9 Ideologies, interests and cooperations under the gold standard In turning their attention towards regimes, institutions, and ideologies, modern economic historians of the slump have come closer than those of an earlier generation to the central preoccupations of social and political historians of interwar Europe. As the latter have long been aware, no account of interwar Europe can ignore the gold standard and the conservative social and political agendas of which it formed an important part. This agenda was reinforced by the European middle classes’ experience or fear of wartime and postwar inflation because of which the restoration of the gold standard came to be widely regarded as the obvious route to economic revival and stability in societies contending with upheavals released by the First World War. Even those who agonized over the costs of restoring or preserving the gold standard believed the alternative was chaos and dissolution. For Polanyi (1957, p. 25), writing towards the end of the Second World War, ‘belief in the gold standard was the faith of the age’. The United States has generally received the lion’s share of scholarly attention for this period. This is partly a legacy of the earlier ‘shockresponse’ approach to the 1930s Depression. But Britain is much more indispensable to analyses of interwar economic instability, particularly as it affected the empire; and that country’s importance for a rounded understanding of even the Depression grows once the latter phenomenon is restored to its proper interwar context and our attention directed towards the principal structural and institutional features of the interwar world economy. It is therefore not surprising that several major themes in recent scholarship – such as fears about the social and political consequences of deflation, asymmetric sharing of adjustment burdens by surplus and deficit countries, distribution of gold reserves, the apparent wilfulness of American overseas lending, questions about French motives and actions, and the immense difficulty of coordinating short-term objectives and policy in the face of domestic constraints – should echo contemporary perceptions of interwar ills, particularly in Britain. As this chapter has shown, Britain’s unrelenting control over macroeconomic management in interwar India ensured that these perceptions determined economic policy-making in the colony – so much so that there exists perhaps no better example of ‘international cooperation’ between two countries in this troubled period. But as Temin (1989, p. 87) and Cooper (1992, p. 2127) have both pointed out whilst contesting Eichengreen’s diagnosis of the ills of the interwar economic system,

162 India

cooperation may be dysfunctional when its object is fundamentally flawed.58 This is perhaps nowhere more true than in the case of interwar India. As Britain attempted to return durably to the gold standard in the midst of a worldwide ‘scramble for gold’ – to quote Norman – deflationary counter-cyclical measures ensured that India would not experience the booms of the 1920s. Though growth was poor, these policies failed to stop the flow of gold to India. In contrast, London pressed pro-cyclical policies on India during the Depression which intensified its effects and turned the colony into an exporter of gold and a source of relief to Britain. It is apparent from the protests which the colonial government’s economic policies provoked in India that relational asymmetries and a regime collapse intensified the involuntary nature of the ‘cooperation’ which obtained between the two countries. But as London recognized especially in the context of its dominions and undoubtedly in relation to India as well, without the assured cooperation of these countries, the credibility of London’s policies to restore sterling stably in investors’ preferences, and so faith in the currency, and perhaps the credibility of the gold standard itself, would have been even weaker than it was during these uncertain years. Nor was the possibility of intensifying such cooperation hidden from view when Britain ‘bargained’ with other players, notably the United States.59 When such bargaining proved difficult (the First World War), or did not yield agreements acceptable to Britain (early 1920s), or failed (1930s), it pondered or threatened closer ‘cooperation’ with the empire and turning its side, if not its back, on globally multilateral trade.60 Furthermore, although Britain sought an inflationary global outcome in the 1920s, its approach towards its colonies remained resolutely deflationist.61 Besides, as some studies have shown, even dependencies like Argentina (which agreed to service sterling debts promptly in return for improved access to British markets for its beef) came off worse in the 1930s from cooperating with Britain than countries which kept their distance from the declining imperial power (O’Connell, 1984; Jorgensen and Sachs, 1989). Finally, Britain was not above deploying empire resources to encourage a more forthcoming US response towards its own concerns while often ignoring those of the affected colony. London, for instance, pressed India not to commit to discontinuing silver sales in a falling market in 1930–1 unless US proposals for an international silver conference were widened to address the gold problem. Two years later Britain agreed over Indian reservations to cap silver sales in order to persuade Roosevelt to stabilize the dollar (Balachandran, 1996a, pp. 586–9). Cooperation between Britain and its dependencies and colonies, in particular India, thus yielded the latter few benefits; and by contributing to the credibility of a flawed and deflationary regime, prolonged its demise. The Indian story told in these pages also raises interesting questions about the relationship between economic doctrines and policies during the

G. Balachandran 163

interwar period upon which so much of the recent literature on the Depression has turned. The history of interwar economic policy-making rests on two interrelated premises. The first is the powerful grip of an unchanging ‘orthodoxy’ over the minds of its practitioners who took an idealized view of the working of the gold standard and were prone to overlook the domestic costs of adjusting to external disequilibria. In the other premise, a contest of ideas is suggested which pitted these orthodox practitioners against younger economists, among whom Keynes was the most notable, whose policy prescriptions were based on a more sensitive understanding of the incidence of these costs. Doctrines and ideas are no doubt important influences over policy. But they do not represent the whole story for several reasons. Doctrines are often retrospective constructs which promote unintended coherences and ignore the disparate social and political contexts and processes producing their raw material. As Fetter warned us many years ago, older ideas are also sometimes articulated into doctrines and ‘orthodoxies’ in the very process of contest with a new set of ideas.62 Interwar orthodoxy was not entirely a postwar construct – in fact this orthodoxy was being constructed, mediated and deployed, challenged, reinforced, all at the same time in the troubled 1920s, until it was grievously wounded in the Depression. But economic historians have either tended to take an ‘essentialist’ (and consequently ahistorical) view of ‘orthodox’ economic doctrines during this period, or attention has focused too closely on the contest between its practitioners and their (Keynesian) challengers. What is ignored as a result are the numerous adjustments interwar policy-makers made to their orthodox ideas and the willingness, more generally, of the state and the dominant social groups in 1920s Europe to use new means to achieve their objectives, including that of restoring the conservative ancien régime. In Maier’s account of interwar social stabilization efforts in Germany, for example, the state was forced by domestic conflicts to tread a path where monetary policies were deflationary to reassure orthodox international creditors, while wage, industrial, and tariff policies were used ‘to reinforce a corporatist and bourgeois social settlement’ (Maier, 1975, p. 590; Maier, 1987, pp. 165f., 182f.). When the world went into a slump and Germany’s creditors failed to produce a package to rescue both external finances and a banking system overexposed in domestic industrial equities, the authorities resorted to effective currency depreciation through stringent exchange controls (Eichengreen, 1992, pp. 270–8). The manner in which Britain’s interwar financial-policy establishment reacted to heterodox ideas was also far from geriatric. Often it was not averse to experimenting piecemeal with new policy options including those advanced by its opponents when these suited its objectives (Booth and Pack, 1985, pp. 2f., 191–5). The interwar British government might

164 India

generally appear orthodox in its employment and public expenditure policies. But its policy on the bank rate was notably interventionist, whereas before the war a bank rate change was of equal unconcern to the Treasury as the colour on the Bank’s ‘front door’.63 The odour of bullionism, not altogether absent from the Bank of England’s management of the pre-war gold standard, also became more noticeable between the wars.64 British efforts to promote international monetary cooperation, the gold exchange standard, and central banking between the wars, too, did not accord with perceptions of the functioning of the pre-war gold standard which officials, desiring a ‘return to status quo ante bellum’, wished to encourage, or of the role of the state and quasi-state institutions such as central banks in it.65 In the 1920s, Britain, while restricting its own overseas lending, led the way in pressing countries running surpluses (principally the US) to expand their economies.66 At the same time, it did not hesitate to press deflationary policies upon other deficit countries. By 1929–30 Britain was even willing to try some soft blackmail with its creditors while advising debtors such as Argentina to default on their dollar obligations (O’Connell, 1984, p. 208; Diaz-Alejandro, 1984, p. 27); and it was left to the Bank of France, which had earlier supported ‘unorthodox’ exchange-rate policies at home, to press upon British officials the orthodox ideology of paying one’s bills punctually and without fuss (Mouré, 1991, pp. 61–3). One of the more striking aspects of Britain’s interwar policy ideas about the reconstruction of the international monetary system is their innovativeness, and the criticism they implied – whether in relation to gold backing for national money supplies, the monetary role of gold, or to central banking – of some interwar dogmas about the pre-war gold standard such as those embodied in the Cunliffe Report.67 This innovativeness was underlined by a policy discourse whose supple, and often transparent, opportunism was unburdened by much new theorizing and whose principal object was to restore the diminishing influence of a sclerotic Britain over an uncertain and unstable world economy. Whatever their historical and internal inconsistencies, Britain’s interwar policy ideas were consistent with its external and domestic objectives. Equally, even as some orthodox rules were held up as inviolable, some others (such as a high gold cover for money supply or in the Indian exchange reserve) were declared ‘out of date’;68 and some new formulae such as central banks owned by shareholders or the gold-exchange standard became ‘the latest guiding principles’ in their respective spheres.69 In this way, rather than a seamless ‘orthodoxy’ commanding blind adherence and absolute faith, new ideas and older ones – a modern policy discourse and more time-worn objectives – jostled in the practical repertoire of economic policy-makers in interwar Britain. This mixing and matching of objectives, strategies, policy instruments, and the discourse deployed to justify them is a notably conspicuous feature of London’s monetary policy-making for interwar India.70

G. Balachandran 165

The course of interwar economic policy-making in India thus illustrates the difficulty of privileging doctrines or doctrinal contests, or of accepting the former as given. The colony’s subordination of exchange-rate stability to price stability during the 1920s attracted praise from Keynes (1923 [1971], pp. 127f). But his prescriptions were applied to India – if not always at their behest, certainly with the knowledge and approval of policy-makers in London who refused to countenance similar policies for Britain. There is thus more than mere irony in the fact that Montagu Norman, who was the major protagonist of deflationary policies in Britain to restore the pound to the gold standard, favoured at the same time a policy of stabilizing prices in India through a floating exchange rate. Nor is it without significance that Norman’s conversion to Keynes’s advocacy in the Tract failed to outlive the period of rising world prices and Indian policies returned to the pro-cyclical mode in the Depression. Notes * I am grateful to Theo Balderston for helpful comments on a earlier draft of the chapter. 1 The estimates are summarized in Balachandran (1996b), p. 171. 2 Goldsmith (1983), pp. 126–8; Goswami (1984); Baker (1984), p. 577f.; more generally Rothermund (1992). 3 Contemporary national income accounts were compiled in India only from the 1950s. Figures for earlier periods are researchers’ estimates. 4 Sources for national income estimates cited below are: Sivasubramonian (1965), and Heston (1983, pp. 397–9), which also gives a summary of Sivasubramonian’s estimates. 5 Tariffs, for example, were imposed mainly for political reasons – to placate domestic industrialists and arrest their growing support for the nationalist Congress party, and to curb the influence of left-wing movements amongst industrial workers – and not as part of a conscious anti-cyclical policy. See Chatterji (1992). 6 Maddison’s estimates in (1971), appendix B, p. 167, converted to 1946–7 prices and index numbers in Heston (1983), p. 402f.; also see Goswami (1986), p. 175. 7 Maddison (1984) cited in Goswami (1986), p. 175. 8 Boyce (1987), pp. 6–34, 45–52, 62–71, 101–25; also see ch. 11. 9 The council bills mechanism was instituted originally to finance the Indian government’s expenditures in London. But it quickly became an instrument of wider exchange intervention and remained so until 1923 when council bill sales in London were supplemented, and shortly replaced, by the sale of rupees in Bombay against sterling credits to the Indian government’s account in London. In this chapter, the terms ‘council bills’ and ‘council bill mechanism’ are used generically to represent all forms of rupee exchange intervention involving the pound sterling. Keynes (1971) still offers the best introduction to the pre-war Indian currency system; the intervention system is discussed in ch. 5. 10 Committee on Finance and Industry (1931), J. Kitchin’s statement, para. 7; Reserve Bank of India (1954), p. 970. 11 Despite the intense debate surrounding India’s bewilderingly diverse and complex organization of agriculture, we may safely regard prices as a proxy for (disposable) incomes of agricultural households which produced their essential

166 India

12 13 14

15 16 17

18 19 20 21

22

23 24 25 26 27

consumption requirements and faced rent, revenue, and debt obligations fixed in nominal terms. Hilton-Young Commission (1926) vol. 2, McWatters’s Memorandum, app. II to app. 3, p. 25. The context and nature of wartime British efforts to restrict gold flows to India are elaborated in Balachandran (1996b), pp. 48–55. Wartime silver prices rose from less than 68 cents per ounce before Indian needs became known to nearly 90 cents by early 1917. The Anglo-American agreement, discussed below, fixed the price of silver at one dollar. See India Office Records, L/F/6/169, ‘Silver Supply, Demand, and Prices’, Memorandum by J. Kitchin, 25 Feb. 1931. The wartime crisis of the Indian currency system is discussed in greater detail in Balachandran (1996b), ch. 3. Before British currency and coinage went metric, a pound equalled 20 shillings. Twelve old pence made up a shilling. For a more detailed discussion of this stabilization episode, see Balachandran (1993). A rupee parity of two shillings (gold) meant the rupee equalling two shillings when the sterling was at its pre-March 1919 parity with gold (or dollar). As soon happened, a depreciation of the pound from this parity meant a rise in the sterling price of rupees as well. Hilton-Young Commission (1926), Memoranda, app. 69, p. 453. Treasury Papers, T160 18/F571, ‘Indian Exchange and Currency Committee’, c. 18 Dec. 1919. Hawtrey Papers, 1/13, ‘Return to Gold Standard’, July 1920; also Hawtrey (1939), p. 56. Hawtrey Papers, 1/3/1, text of speech, 4 Dec. 1923; there is an apparent paradox in Blackett’s argument that rupee stabilization at a lower rate might make gold ‘look cheap’, since the higher rupee which the Finance Member sought would actually reduce the nominal rupee price of gold. But this paradox is easily resolved once we recognize that Blackett had in mind not the nominal price of gold, but the metal’s price in terms of commodities. By depressing prices in India (including of commodities competing with gold such as cloth) a higher rupee was expected to increase the ‘real’ price of gold (i.e. its price in terms of commodities). A rupee which floated upwards as world prices rose on the back of the widely anticipated US-led expansion of the world economy would also help check the fall in the ‘real’ price of gold in India. Hence Blackett’s warning against ‘premature stabilization’. With sterling purchases restricted to secure rupee appreciation and gold not being tendered to the government because its official price (of two shillings or Rs 10 fixed in 1920) was below the market price, whatever currency expansion took place in the mid-1920s was through the discretionary seasonal issue of ‘emergency currency’ by the government. This currency was typically withdrawn at the end of the peak season. Hilton-Young Commission (1926), Minutes of Evidence, Q. 10479, 11830. For Norman’s views, see Hilton-Young Commission (1926), Minutes of Evidence, QQ. 13724–34. India Office Records, Mss. Eur. E397/32, Blackett’s letter to Benjamin Strong, 21 Dec. 1927. Government of India (various years). India Office Records, L/F/7/897, colln 107, central legislative assembly proceedings, 23 Feb. 1931; L/F/5/100, Finance Department statistics.

G. Balachandran 167 28 Keynes (1923 [1971]), p. 127f. 29 McAlpin (1983), pp. 903f. – ‘weighted all commodities’ series. 30 Heston (1983), pp. 397–9; population grew about one per cent per annum during this decade; see Visaria and Visaria (1983), p. 488. 31 On the possible impact of anticipated currency appreciation on prices, see Faini and Toniolo (1992), pp. 132–6. Indian government officials blamed high cloth prices for boosting the demand for gold. On the other hand, expectations of currency appreciation would have reduced the demand for gold. But this effect may have been offset by uncertainty about the future of the pound sterling and trading access to the yellow metal. 32 While arguing (p. 46) that ‘Indian hoarding demand was largely an inverse function of gold’s real price’, Johnson (1997, pp. 46–8) also invokes other variables, such as the size of harvests, export prices, incomes, and silver prices to explain movements in Indian gold demand in the 1920s and 1930s. 33 Federal Reserve Bank of New York Archive, C252, Crane to Case, 4 Jan. 1928. 34 India Office Records, L/F/5/100, Finance Department statistics, pp. 117–19; L/F/5/101, p. 119. 35 Trade and current account figures from the Indian government’s estimates presented to the League of Nations; national income figures from Heston (1983), pp. 397–9. 36 Bank of England Archive, OV9/17, Denning to Kisch, 5 March 1930; Kisch to Niemeyer, 22 March, and reply, 24 March 1930; Niemeyer to Schuster, 22 April 1930. 37 Treasury Papers, T172/1756, Treasury memorandum ‘The Balance of Payments’, n.d. but summer of 1931; also Eichengreen (1992), p. 280. 38 India Office Records, L/F/5/189, Secretary of State to Viceroy, 10 Sept. 1931. 39 For a more detailed account of the discussion, see Balachandran (1996b), pp. 175–7. 40 India Office Records, L/F/6/1183 f. 6619, meeting on financial questions, 16 Oct. 1931, p. 34. 41 Bank of England Archives, G14/96, Committee of Treasury, 26 Nov. 1930; Gubbay to Shaw, 1 Dec. 1930. 42 India Office Records, L/F/6/1177, f. 3693, Norman to Kisch, 5 May 1931. 43 Government of India (various years). 44 HMSO (various years); national income figures from Heston (1983), pp. 397–9. 45 India Office Records, L/F/6/1172 f. 37, P&O Bank to Kisch, 1 Jan. 1931; Viceroy to Secretary of State, 1 April 1931 in L/F/6/1175 f. 2222; 20 Jan. 1932 in L/F/6/1188; Schuster (1979), p. 115; also letters of the Controller of Currency in Finance Department papers, f. 1(8)–F–1931–1. 46 On the effects of a devaluation in the depression, see Eichengreen and Sachs (1985). 47 For gold price differentials between India and London, see Rothermund (1992), pp. 48–9 and Drummond (1981), p. 48. 48 South African figures from Katzen (1964), pp. 18–19, 40–1, 60–1. 49 Treasury Papers, T160/400 f. 12471, annex 7, ‘The Repayment of an Indian Loan: An Incident in the History of Whitehall Control’, CP 23(32), 15 Jan. 1932 and cabinet resolution, 20 Jan. 1932. 50 Quoted in Bridge (1986), pp. 73f. 51 Bank of England Archives, C43/279, copies of cables exchanged between the London and Bombay offices of the Hongkong and Shanghai Banking Corporation, Jan.–May 1932; US National Archives, GB 120, American Consul’s despatch from Bombay, 23 Jan. 1933.

168 India 52 Bank of England Archives, C43/22, Phillips to Hambro, 23 May 1932 and Hambro’s undated reply; also see Federal Reserve Bank of New York Archives, C261, Osborne to Crane, 13 Jan. 1932. 53 Treasury Papers, T175/57, part 2, Phillips’s memorandum, 27–29 Feb.; letter to Henderson, 26 Feb. 1932; see Howson and Winch (1977), pp. 104–5 on the significance of this memorandum; for contemporary perceptions of the significance of Indian gold exports, see Keynes (1982), pp. 70–2; Keynes (1972), p. 353; Bank of England Archives, OV 48/9, ‘Factors Influencing the Movement in Gold Prices’, by Per Jacobsson, 12 March 1932; OV 48/10, ‘The Future of the Gold Standard’, also by Per Jacobsson, April 1934, p. 24. 54 See Howson (1980), p. 62 for the changing composition of assets in Britain’s exchange equalization account after about the middle of 1932; also more generally, Balachandran (1996b), pp. 183f. 55 Reserve Bank of India (1954), pp. 881 and 668 respectively. 56 Bank of England Archives, OV 56/51, note by Catterns, 14 Jan. 1932; the concern about inflation at the bottom of the slump was, however, not confined to India. 57 Treasury Papers, T177/8, ‘Memorandum on Effects of the Fall in Prices’, June 1932 and Phillips’s pencilled comment; T160/474 f. 12471/06/2, Leith-Ross’s remarks on Waley’s note, 5 March 1932; Drummond (1981), pp. 23–7; also John Singleton’s chapter on New Zealand in this volume. 58 See, too, Pierre Siklos’s chapter in this volume for a similar point about Canada. 59 Costigliola (1977). He also emphasizes the importance Britain attached in the 1920s to retaining Australia and South Africa within the sterling area. Also see in this connection, Tsokhas (1994). 60 For the First World War and the depression, see Cain and Hopkins (1993), pp. 49–61 and ch. 5 respectively; for the early 1920s, Costigliola (1977); for the 1930s, also see Drummond (1981). It is not surprising, in the circumstances, that many in post-Second World War USA and Europe believed the sterling area and imperial preferences posed a threat to the revival of multilateral trade. 61 For a survey, see Cain and Hopkins (1993), chs 6 and 9. 62 See the analysis of mercantilism in Coleman (1980); on the so-called gold standard orthodoxy, see Fetter (1965), pp. 143, 216, and 237–9. 63 This exaggerated view is attributed to Otto Niemeyer who was Controller of Finance at the Treasury until he left to join the Bank of England in 1927: Clarke (1988), p. 35; on interwar Treasury interventionism, see Boyce (1987), pp. 134f. 64 See Bloomfield (1959), p. 25; de Cecco (1985), p. 46; Eichengreen (1985), p. 24. 65 Flanders (1990), p. 83; more generally see pp. 67–83. 66 Moggridge (1971); Boyce (1987), pp. 95–8, 175–6; Committee on Finance and Industry (1931), Report , para. 314. 67 Eichengreen (1992), pp. 36–7 argues that Cunliffe dogmas did not entirely prevent contemporaries from holding other views on the working of the pre-war gold standard. 68 India Office Records, colln. 43, f. 16, L/F/7/468, Kisch’s minute, 22 May 1923 dismissing a recommendation of the pre-war Chamberlain Commission (1914), para. 96, proposing at least a 50 per cent gold cover in the exchange reserve; this recommendation was more valid in the troubled 1920s than when it was first made. 69 See Hawtrey’s advocacy of the gold-exchange standard in Hawtrey Papers, 1/3/2, ‘Indian Currency’, 28 Aug. 1925; on central banking, see Balachandran (1996b), ch. 8. 70 For a more elaborate discussion, see Balachandran (1996b), pp. 8–18, 223–6.

G. Balachandran 169

References Unpublished references Bank of England Archive, London. Federal Reserve Bank of New York Archives, Federal Reserve Bank, New York. Finance Department Papers, Government of India, National Archives of India, Delhi. India Office Records,British Library, London. Hawtrey papers, Churchill College Archives, Cambridge. Treasury Papers, Public Record Office, Kew. US National Archives, Washington, DC.

Published references Aldcroft, D.H. (1977) From Versailles to Wall Street, 1919–1929 (London: Allen Lane). Baker, C. (1984) Indian Rural Economy, 1880–1955: The Tamilnad Countryside (Delhi: Oxford University Press). Balachandran, G. (1993) ‘Britain’s Liquidity Crisis and India, 1919–20’, Economic History Review ,46 (August): 575–91. ——– (1996a) ‘Gold, Silver, and India in Anglo-American Monetary Relations, 1925–1933’, International History Review ,18 (August): 573–90. ——– (1996b) John Bullion’s Empire: Britain’s Gold Problems and India between the Wars (London: Curzon). Bloomfield, A.I. (1959) Monetary Policy under the International Gold Standard: 1880–1914 (New York: Federal Reserve Bank of New York). Booth, A. and M. Pack (1985) Employment, Capital and Economic Policy: Great Britain, 1918–1939 (Oxford: Basil Blackwell). Boyce, R.W.D (1987) British Capitalism at the Crossroads, 1919–1932: A Study in Politics, Economics, and International Relations (Cambridge: Cambridge University Press). Bridge, C. (1986) Holding India to the Empire: The British Conservative Party and the 1935 Constitution (Delhi: Sterling). Brown, W.A. Jr (1929) England and the New Gold Standard, 1919–26 (London: P.S. King). Cain, P.J. and A.G. Hopkins (1993) British Imperialism. Vol.2: Crisis and Reconstruction, 1914–1990 (London: Longman). Chamberlain Commission (1914) [Royal Commission on Indian Finance and Currency], Report (London: HMSO). Chatterji, B. (1992) Tariffs, and Empire: Lancashire and British Policy in India, 1919–1939 (Delhi: Oxford University Press). Chaudhuri, K.N. (1983) ‘Foreign Trade and Balance of Payments’, in The Cambridge Economic History of India, vol. 2: c. 1757–1970, ed. Dharma Kumar (Cambridge: Cambridge University Press): pp. 804–77. Clarke, P. (1988) The Keynesian Revolution in the Making 1924–1936 (Oxford: Clarendon). Clay, H. (1957) Lord Norman (London: Macmillan). Coleman, D.H. (1980) ‘Mercantilism Revisited’, Historical Journal , 23 (4): 773–91. Committee on Finance and Industry [i.e. the Macmillan Committee] (1931) Report (London: HMSO). Committee on Finance and Industry [i.e. the Macmillan Committee] (1931) Minutes of Evidence, 2 vols (London: HMSO). Cooper, R.N. (1992) ‘Fettered to Gold? Economic Policy in the Interwar Period’, Journal of Economic Literature , 30 (December): 2120–28.

170 India Costigliola, F.C. (1977) ‘Anglo-American Financial Rivalry in the 1920s’, Journal of Economic History, 37 (December): 911–34. de Cecco, M. (1985) ‘The International Debt Problem in the Interwar Period’, Banca Nazionale del Lavoro Quarterly Review, no. 152: 45–64. Diaz-Alejandro, C. (1984) ‘Latin America in the 1930s’, in R. Thorp (ed.) Latin America in the 1930s: The Role of the Periphery in the World Crisis (London: Macmillan and St Martin’s Press): 17–49. Drummond, I.M. (1981) The Floating Pound and the Sterling Area, 1931–39 (Cambridge: Cambridge University Press). Eichengreen, B.J. (1985) ‘Introduction’, in B. Eichengreen (ed.) The Gold Standard in Theory and History, 1st edn (New York: Methuen). ——– (1992) Golden Fetters: The Gold Standard and the Great Depression 1919–1939 (New York: Oxford University Press). ——– and J. Sachs (1985) ‘Exchange Rates and Economic Recovery in the 1930s’, Journal of Economic History, 45 (December): 925–46. Faini, R. and G. Toniolo (1992) ‘Reconsidering Japanese Deflation during the 1920s’, Explorations in Economic History, 29 (April): 121–43. Fetter, F.W. (1965) Development of British Monetary Orthodoxy (Cambridge, MA: Harvard University Press). Flanders, M.J. (1990) International Monetary Economics 1870–1960: Between the Classical and New Classical (Cambridge: Cambridge University Press). Goldsmith, R.W. (1983) The Financial Development of India, 1860–1977 (Delhi: Oxford University Press). Goswami, O. (1984) ‘Agriculture in the Slump: The Peasant Economy of East and North Bengal in the 1930s’, Indian Economic and Social History Review, 23 (3): 347–58. ——– (1986) ‘The Depression, 1930–1935: Its Effects on India and Indonesia’, Itinerario,10 (1: special issue ‘India and Indonesia from the 1920s to the 1950s: The Origins of Planning’). Government of India (various years) Controller of Currency, Report. Hawtrey, R.G. (1939) The Gold Standard: In Theory and Practice, 4th edn (London: Longmans Green). Heston, A. (1983) ‘National Income’, in The Cambridge Economic History of India, vol. 2: c. 1757–1970, ed. Dharma Kumar (Cambridge: Cambridge University Press) pp. 376–462. Hilton-Young Commission (1926) [Royal Commission on Indian Currency and Finance] (London: HMSO). HMSO (various years) East India Finance and Accounts (London). Howson, S.K. (1980) Sterling’s Managed Float: The Operations of the Exchange Equalization Account, 1932–39 (Princeton Studies in International Finance no. 46, Princeton, International Finance Section, Department of Economics). ——– and D. Winch (1977) The Economic Advisory Council, 1930–39 (Cambridge: Cambridge University Press). Indian Currency and Exchange (1935) Congress Golden Jubilee Brochure no. 11 (Allahabad: All-India Congress Committee). Johnson, H. Clark (1997) Gold, France, and the Great Depression, 1919–1932 (New Haven: Yale University Press). Jorgensen, E. and J.S. Sachs (1989) ‘Default and Renegotiation of Latin American Foreign Bonds in the Interwar Period’, in B. Eichengreen and P.H. Lindert (eds) The International Debt Crisis in Historical Perspective (Cambridge, MA: MIT Press): 57–68. Katzen, L. (1964) Gold and the South African Economy (Amsterdam: A.A. Balkema).

G. Balachandran 171 Keynes, J.M. (1911 [1971]) Indian Currency and Finance, reprinted as vol.I of The Collected Writings of John Maynard Keynes, eds D.E. Moggridge and E. Johnson (London: Macmillan and St Martin’s Press for the Royal Economic Society). ——– (1923 [1971]) A Tract on Monetary Reform, reprinted as vol.IV of The Collected Writings of John Maynard Keynes, eds D.E. Moggridge and E. Johnson (London: Macmillan and St Martin’s Press for the Royal Economic Society). ——– (1972) The Collected Writings of John Maynard Keynes, vol. IX: Essays in Persuasion (Macmillan and St Martin’s Press for the Royal Economic Society). ——– (1982) The Collected Writings of John Maynard Keynes, vol. XXI: World Crises and Policies in Britain and America, eds D.E. Moggridge and E. Johnson (Macmillan and Cambridge University Press for the Royal Economic Society). McAlpin, M. (1983) ‘Price Movements and Fluctuations in Economic Activity’, in The Cambridge Economic History of India, vol. 2: c. 1757–1970, ed. Dharma Kumar (Cambridge: Cambridge University Press) pp. 878–904. Maddison, A. (1971) Class Structure and Economic Growth: India and Pakistan since the Moghuls (London: George Allen & Unwin). ——– (1984) ‘Growth, Crisis, and Interdependence, 1929–38 and 1973–83’ (unpublished mimeo). Maier, C.S. (1975) Recasting Bourgeois Europe: Stabilization in France, Germany, and Italy in the Decade after World War I (Princeton: Princeton University Press). ——– (1987) In Search of Stability: Explorations in Historical Political Economy (Cambridge: Cambridge University Press). Moggridge, D.E. (1971) ‘British Controls on Long-Term Capital Movements, 1924–31’, in D.N. McCloskey (ed.) Essays on a Mature Economy: Britain since 1840 (Princeton: Princeton University Press). Mouré, K. (1991) Managing the Franc Poincaré: Economic Understanding and Political Constraint in French Monetary Policy (Cambridge: Cambridge University Press). O’Connell, A. (1984) ‘Argentina into the Depression: Problems of an Open Economy’, in R. Thorp (ed.) Latin America in the 1930s: The Role of the Periphery in the World Crisis (London: Macmillan and St Martin’s Press): 188–221. Polanyi, K. (1957) The Great Transformation: The Political and Economic Origin of Our Times (Boston: Beacon Press). Reserve Bank of India (1954) Banking and Monetary Statistics of India (Bombay: Reserve Bank of India). Rothermund, D. (1992) India in the Great Depression, 1929–1939 (Delhi: Manohar). Schuster, G. (1979) Private Work and Public Causes: A Personal Record (Cowbridge: D. Brown). Sivasubramonian, S. (1965) ‘National Income of India: 1900–10 to 1946–47’, (unpublished Ph.D. thesis, Delhi School of Economics). Temin, P. (1989) Lessons from the Great Depression (Cambridge, MA:. MIT Press). Tsokhas, K. (1994) ‘The Australian Role in Britain’s Return to the Gold Standard’, Economic History Review, 47 (1): 129–146. Visaria, L. and P. Visaria (1983) ‘Population (1757–1947)’, in The Cambridge Economic History of India, vol. 2: c. 1757–1970, ed. Dharma Kumar (Cambridge: Cambridge University Press) pp. 463–532.

7 New Zealand in the Depression: Devaluation without a Balance-ofPayments Crisis* John Singleton

1

Introduction

New Zealand, with a population of less than two million souls, was the smallest, except for Newfoundland, of the politically independent British dominions in the 1920s and 1930s. New Zealand enjoyed one of the highest levels of income per capita in the world; the efficiency of its livestock industry was unsurpassed; and its state was a pioneer of social reform. Economically, the ties between New Zealand and Britain were extremely strong. Eighty per cent or more of the country’s exports, which consisted of wool, meat, and dairy produce, were regularly consigned to the mother country. It is clear that the Depression was transmitted to New Zealand by falling commodity prices in the British market. The story of the New Zealand government’s response to the slump is broadly consistent with the observations of Eichengreen (1992), and especially Temin (1989), on the mesmerizing effect of orthodox economic theory on politicians and their advisers. As the crisis deepened, however, the grip of orthodoxy began to loosen, and the government started to modify its policies. After the Labour Party came to power in 1935 there was an unmistakable drift towards socialism, although by this time the economic recovery was virtually complete. The New Zealand pound depreciated against sterling during 1930–1, and was devalued against sterling by an additional amount in 1933. The 1933 devaluation was unusual because the balance of payments was healthy; the purpose of this measure was to raise the incomes of farmers. Changes in economic policy only partially explain New Zealand’s recovery from the Depression. Credit is also due to persistent growth in agricultural productivity, the Ottawa conference, and the eventual stabilization and strengthening of commodity prices in Britain. New Zealand’s economic institutions were unsophisticated but robust in the 1920s and 1930s. In economic and financial matters, governments 172

John Singleton 173

attempted to follow what they understood to be the orthodox British line. Academic expertise in the discipline of economics was very scarce, except in the rather limited area of agricultural economics. New Zealand economists deferred to the views of their eminent British colleagues and teachers in high theory and monetary economics. Although they were aware that some British economists, such as Keynes, had transcended the gold standard mentality, they were reluctant to take the initiative in proposing radical policies for New Zealand. In the 1920s, the New Zealand pound was on a sterling exchange standard, so that the preservation of the gold standard was never an issue. But this does not mean that New Zealand was free of the gold standard mentality. On the contrary, the link between the New Zealand pound and sterling assumed an importance in New Zealand equivalent to that of the gold standard elsewhere. New Zealand accepted the other elements of financial orthodoxy, including the sanctity of balanced budgets, without hesitation. The next section outlines the structure of the New Zealand monetary system. Following sections examine the transmission of the slump to New Zealand, the government’s resort to deflation, the devaluation campaign, the causes of economic recovery, and the fate of proposals for an imperial economic strategy to fight Depression.

2

The New Zealand monetary system

The monetary system of New Zealand, like that of Australia, was on a sterling exchange standard, although it was indirectly linked to the gold standard through sterling’s convertibility into gold. Gold was held in New Zealand in the vaults of the commercial banks but it rarely left the country. According to Tocker (1924), the main determinant of the money supply was the quantity of sterling held in Britain by the New Zealand commercial banks, in accounts known as London or sterling balances. These banks had branches in London because so much of New Zealand’s trade was with Britain. Tocker stated that a fall in British commodity prices would result in a reduction in New Zealand’s sterling export earnings and a decline in the London balances. Banks would interpret the fall in London balances as a signal to reduce advances to their customers in New Zealand. A rise in British commodity prices would set in motion the opposite sequence of events. London balances would rise and the New Zealand banks would increase their lending to customers at home. It was assumed that domestic cost and price adjustments would follow upon a change in advances. A fall in advances would lead to reductions in costs and prices in New Zealand; exports would rise and imports would fall, and the London balances would return to their original level. A rise in advances would lead to increases in costs and prices; exports would fall and imports would rise, and the London balances would return to their original level.1 Tacit agreement

174 New Zealand

among the commercial banks precluded the variation of interest rates in response to short-term fluctuations in the London balances. The commercial banks also fixed the exchange rate between the New Zealand pound and sterling; governments did not attempt to determine the exchange rate before 1933. The banks tried to hold the exchange rate at parity, so that one New Zealand pound was worth one pound sterling (and one Australian pound), in deference to the almost universal belief that a pound should be worth the same whether it was British, New Zealand, or Australian. Defence of the exchange rate was also left in the hands of private sector financial institutions. New Zealand did not have a central bank until 1934, and government financial policy was largely confined to the perennial struggle to balance the budget (Hawke, 1971; Endres and Fleming, 1995; Fleming, 1997a, pp. 1–6). As noted earlier, the banks did not always succeed in their efforts to preserve the parity, and the local currency depreciated by 10 per cent against sterling in 1930–1.2 When Britain was forced off the gold standard in September 1931, the New Zealand commercial banks chose to maintain the prevailing sterling exchange rate, and in consequence the New Zealand pound fell against gold (Drummond, 1981, p. 6). Any other policy would have been disastrous for New Zealand borrowers, who were already smarting from the decline in commodity prices. If the banks had chosen to defend the exchange rate against gold, the New Zealand pound would have risen against sterling, precipitating a further sharp reduction in farmers’ export receipts. Since New Zealanders were price takers who sold their produce in London for sterling, revaluation against sterling would have slashed the domestic-currency value of farmers’ export receipts. Revaluation would also have reduced the domestic-currency value of the London balances. The banks’ decision to follow sterling was not regarded as a departure from orthodoxy, nor was it seen a matter for official comment. Ministers were more concerned about the British departure from gold, which it was initially feared would damage the City of London to such an extent that the New Zealand government would be unable to renew maturing sterling loans. A secret emergency credit was obtained from the Bank of England, but, in the event, the loans were renewed without undue difficulty (Hawke, 1988, pp. 122–3). It is worth noting that New Zealand gained little in third markets from its joint float with sterling. The vast bulk of its exports were sold in the Sterling Bloc, and the remainder encountered import restrictions in many countries.

3

Transmission of the slump to New Zealand

The Depression was transmitted to New Zealand by the collapse in London commodity prices (Hawke, 1985, pp. 127–9; Hawke, 1988, p. 113). Several factors contributed to the fall in prices in the British market. First, the

John Singleton 175

slump caused the demand curve for primary commodities to shift to the left. Second, the supply curve for commodities shifted to the right. Depression did not halt the gradual increase in world agricultural productivity and output. Moreover, countries such as Holland and Denmark diverted food exports to Britain in response to growing protectionism on the European continent. Farmers bore the brunt of the Depression. In the worst year, 1931–2, net farm income in New Zealand was actually negative (Fleming, 1997b, p. 66). Rural distress was intensified by the fact that many farmers had taken on large mortgages when land was very expensive at the end of the First World War. Declining commodity prices increased the already onerous burden of servicing and repaying these loans. The fall in rural incomes exerted a powerful ripple effect on the rest of the economy, which was heavily dependent on orders placed by farmers. Meat freezing works, woollen mills, and dairy factories were unavoidably caught in the spiral of decline. Farmers had less to spend on locally produced and locally finished consumer goods. Service providers of all sorts, including railways, pubs, and dentists, suffered as a result of the agricultural depression. Only debt collectors could look forward to an increase in trade. Table 7.1 presents some key economic statistics from the slump era. Real incomes fell by a lesser amount in New Zealand than in certain other countries, including the United States and Germany. It was easier for a nation of farms and small businesses to adjust costs and prices after a shock than for countries dominated by large corporations, cartels, and mass unions.3 Nevertheless, the experience of living through the Depression profoundly affected many New Zealanders, and shaped their view of the international economy for decades to come. As a result of the decline in export prices, New Zealand’s balance of payments was under severe pressure in 1930–1, the London balances fell, and the exchange rate slipped below parity. Even so, New Zealand’s balance-ofpayments crisis was far milder than Australia’s, not least because the Australians had been the heavier overseas borrowers during the 1920s. Moreover, it did not take long for the decline in New Zealand’s export receipts to bring about a large fall in demand for imports. Farmers were forced to reduce their purchases of imported machinery and consumer goods. By 1932, the New Zealand banks were accumulating London balances again (Cain and Hopkins, 1993, p. 135). Thus the devaluation of 1933 cannot have been due to weakness in the balance of payments, nor can it have been rendered necessary by the Australian devaluation of 1931.

4

Orthodox reaction to the slump

Most New Zealanders endured the Depression in typically stoical fashion, which sometimes verged on the fatalistic. It is hardly surprising that they

176

Table 7.1

New Zealand and the Depression, 1926–38 (1) (2) (3) Nominal Per capita Per capita GNP (£m) real GNP economic 1910/11 growth prices (£) (per cent)

(4) Export values at current prices (1929 =100)

(5) Export volumes (1929 =100)

(6) Retail price index (1929 =100)

(7) Commodity terms of trade (1929=100)

(8) Unem ployment (per cent)

(9)* Nominal interest rates (per cent)

(10)* Real interest rates (per cent)

(11) Money supply (M3) (£ m)

(12) Nominal govern ment budget surplus (£ m)

Year to: March

March

March

Dec

Dec

Dec

Dec

Dec

March

March

Dec

March

1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938

71.9 68.7 75.5 77.3 73.0 65.9 63.7 67.5 70.4 73.3 86.4 90.4 95.8

–7.2 –4.4 9.8 2.4 –5.6 –9.8 –3.3 6.0 4.3 4.1 18 4.6 6.0

82 88 101 100 82 63 65 75 86 85 103 121 106

87 94 97 100 104 105 118 136 127 128 137 137 130

100 100 100 100 98 90 84 79 80 83 86 92 95

83 86 102 100 79 66 62 61 76 75 86 96 88

3.4 5.2 6.5 5.8 6.5 8.8 9.4 11.1 11.6 9.3 9.5 7.9 6.3

6.2 6.5 6.5 6.5 6.4 6.3 6.3 5.9 5.6 5.1 4.7 4.6 4.7

10.3 9.9 5.1 6.4 2.4 18.0 12.3 7.0 2.8 –2.7 2.1 –2.7 2.5

48.3 49.3 55.5 57.3 54.3 52.1 53.1 62.0 62.2 67.2 69.1 74.2 74.2

–5.2 –5.1 –5.3 –6.0 –4.7 –6.8 –5.1 –1.3 –2.1 2.6 –0.7 –1.7 –2.6

157.5 147.8 166.5 172.3 158.4 127.9 117.2 123.7 133.5 150.5 183.1 207.0 226.3

* Nominal interest rates refer to mortgages registered under the land deeds system. These are adjusted by an official wholesale price index to give real interest rates (Fleming, 1997b, p. 67). Note: Figures are in New Zealand pounds unless stated otherwise. Sources: Cols 1–3, Rankin (1992, p. 59); Cols 4–8, Hawke (1988, p. 11); Cols 9–10, Fleming (1997b, p. 66); Col. 11, adapted from Sheppard, Guerin and Lee (1990, p. 55); Col. 12, Abbott (1995, p. 66).

John Singleton 177

were bemused by the unfolding chaos in the world economy. The fact that the centre-right United Party was re-elected to power, in 1931, indicated that the electorate, which was organized on the basis of universal suffrage, was either content with, or else saw no alternative to, orthodox economic policies. As an individualistic society, New Zealand was infertile ground for movements peddling collectivist solutions to the nation’s woes (Mabbett, 1995). Agitation for a socialist response to the Depression was not converted into a Labour majority until 1935, by which time the economy had already climbed out of the trough. In 1929, the elderly, but vastly experienced, Sir Joseph Ward was both Prime Minister and Minister of Finance in the United government. The main opposition party, Reform, was led by Gordon Coates, who had been Prime Minister until 1928. Both United and Reform supported orthodox economic policies, and there was little to choose between them. Labour, the third party, held the balance of power, but it was unwilling to bring down the government because it had no money for a snap election campaign. A visiting Oxford academic remarked that Ward’s ‘reddish nose and waxed moustaches’ made him look like a ‘prosperous bartender’ (Bassett, 1993, p. 275). The ailing Ward retired in 1930 and died soon afterwards. His successor as leader of United and Prime Minister was the bland but ‘cunning’ George Forbes, who did not allow public business to interfere with his regular visits to the movies (Bassett, 1995, pp. 174–5). Forbes’s intention was to ride out what was expected to be a brief recession with a balanced budget. Observing that tax and customs revenues were falling due to the contraction in incomes and the decline in imports, Forbes promised to balance the budget by slashing government expenditure. Business leaders were delighted with this uncompromising stance; the general manager of the National Bank purred that the new premier was ‘a good solid sensible man’ (Bassett, 1995, p. 160). From the perspective of 1930, Forbes’s conviction that the recession would be ephemeral was not unreasonable. New Zealanders were accustomed to sharp fluctuations in commodity prices (Fleming, 1999, p. 336). The severity of the economic crisis was not recognized until 1931, and no significant change in policy took place until 1933. Before discussing the pros and cons of alternative policies, it is necessary to explain the beliefs and actions of the government. United, Reform, Treasury, the farmers, the bankers, and most economists were initially convinced that deflation was the correct response to the fall in incomes and the decline in London balances. It was taken for granted that market forces would eventually drive domestic costs and prices down to levels that were appropriate in the light of external conditions. As New Zealand’s commodities became cheaper, export volumes would grow, and in due course the London balances and personal incomes would be restored to normal levels. Adjustment could be hastened by judicious government intervention; in particular deflation

178 New Zealand

would bring forward reductions in costs and prices (Endres, 1990a; Endres and Fleming, 1995). Modern scholarship, however, suggests that the elasticity of supply of New Zealand’s exports was low. Export volumes were determined by capacity. Farmers did not hold produce back from the market when prices were low, and could not easily increase supply when prices rose (Fleming, 1999, p. 336). Under these circumstances, the prospects for an export-led recovery, induced by cost reductions, were at best moderate. Public works expenditure was reduced from £9.0 million in 1930 to £2.4 million in 1933. Old age, invalid, and widows’ pensions were cut, as were family allowances. In April 1931, and once more in March 1932, public sector wages were reduced by 10 per cent. Interest rates were cut on new and existing government debt. But government expenditure did not show any appreciable decline until the year ending March 1933. Despite his brave promise, Forbes had not been able to cut spending overnight (Abbott, 1995, pp. 66–7). Measures were also taken to reduce costs in the private sector. Interest rates were reduced, in 1931, under an informal agreement between the government and the banks. A further cut in interest rates was secured by legislation in 1932. The government was particularly worried about the difficulties experienced by farmers in servicing their mortgages (MacDonald and Thomson, 1987; Fleming, 1997b). In 1931, the Arbitration Court, which was responsible for ensuring that wage settlements were fair, ordered a 10 per cent cut in wages in all industries. Unemployed workers were strongly encouraged not to hold out for high wages before returning to work. No outright dole payments were made until 1934. The only recourse for men without work was to sign on for hard labour, such as road building, on a government job creation scheme in return for food and accommodation in a work camp (Robertson, 1982; Endres, 1990b; Endres and Jackson, 1993). It should be noted that the Labour Party consistently dissented from the orthodox interpretation of the crisis. Labour initially attributed the rise in unemployment to excessive British immigration, the mechanization of agriculture, and low consumer demand arising from the unequal distribution of income. As the crisis deepened, Labour concluded that the ultimate cause of the slump was the inherent volatility of the international monetary system. Eichengreen and Temin explain international instability in terms of the clash between the gold-standard mentality and the circumstances of the interwar era. But, for Labour, the instability of the international monetary system was an essential feature of banking in a capitalist world. Many socialists in New Zealand, Australia, and Canada were attracted, during the 1930s, to the unorthodox monetary theories of the Englishman, Major Douglas, founder of the social credit movement (Bennett, 1997). Douglas argued that private-sector financial institutions were incapable of creating sufficient spending power to maintain full

John Singleton 179

employment (King, 1988, pp. 146–60). The Douglas thesis was an intoxicating mixture of algebra – the famous A + B theorem – and fantasy worthy of the X-Files in which it was alleged that the world was controlled by a conspiracy of international bankers. Although the New Zealand Labour Party never officially endorsed the doctrine of social credit, which was too extreme for its more sober supporters, it absorbed the underlying message that the world monetary system was rotten to the core. It followed from this assessment that, instead of supinely waiting for international conditions to improve, the government should actively stimulate recovery. During the early 1930s, Labour began to campaign for fiscal and monetary expansion as an antidote to the deflationary impact of the international banking system (Sinclair, 1976, pp. 90–8; Gustafson, 1986, pp. 132–52). Such policies presupposed tight controls over international transactions in order to prevent the complete collapse of the external value of the currency. Multi-party talks on the economic crisis were held in 1931. Forbes was reluctant to take all of the blame for retrenchment, and was delighted when Reform agreed to enter into a coalition with his party. Labour, however, refused to compromise and opted for isolation. Forbes retained the premiership, while Coates was given responsibility for tackling unemployment. Coates was at first inclined to follow the orthodox economic line, but, as we shall see in the next section, in view of the non-appearance of the promised recovery, in 1932 he became a leading advocate of devaluation (Bassett, 1995, pp. 180, 191).

5

Devaluation

The sterling value of the New Zealand currency was altered twice during the Depression. During 1930 and early 1931, the commercial banks succumbed to pressure from the international currency markets, and allowed the value of the New Zealand pound to depreciate from parity to a rate of £NZ110 to £stg100.4 The banks hoped that the parity would soon be restored, and the use of devaluation to promote domestic economic recovery was not even considered at this juncture. Only the second devaluation, in 1933, was a deliberate act of public policy, and followed a period of intensive debate. The slide in the exchange rate in 1930–1 was viewed with disquiet in New Zealand because it appeared to signal a loss of confidence in both the currency and the overall stability of the economy. Treasury and the ‘Canterbury School’ of economists, based in Christchurch, predicted that a lower exchange rate would actually hinder recovery. They argued that a lower exchange rate would raise the price of imported farm inputs, and thus increase the costs of New Zealand producers, making their plight even more difficult in overseas markets. Since reductions in costs and prices were

180 New Zealand

deemed necessary for recovery, the Canterbury economists concluded that the exchange rate should be returned to par as quickly as possible (Endres, 1990a, pp. 65–6). Critics attributed the depreciation of 1930–1 to high official overseas borrowing in the 1920s, and Australian control over the commercial banks. Four of the six main banks in New Zealand were in Australian ownership, and it was alleged that they used the London balances earned by New Zealanders to finance transactions by wicked Australians. This was pure paranoia, and it should not have been necessary to look beyond New Zealand’s temporary balance of payments difficulties to account for the depreciation (Sinclair and Mandle, 1961, pp. 190–4; Hawke, 1973, pp. 19–22). Support for a further devaluation emerged in 1931, and gathered strength in 1932. Devaluation was recommended by Professor D.B. Copland, a New Zealander at Melbourne University, who had already advised the Australian government on this issue (Groenewegen and McFarlane, 1990, pp. 136–42), and by A.C. Davidson of the Bank of New South Wales, an Australian bank with branches in New Zealand. Copland and Davidson pointed out that devaluation would increase farmers’ revenue, expressed in New Zealand pounds, per unit of exports. Since New Zealand was a price taker, devaluation would have no effect on sterling prices. As farm incomes picked up, spending by farmers would increase, and incomes would rise throughout the economy. In February 1932, the Economic Committee, a panel of economists whose advice had been sought by the government, threw its weight behind the theory that devaluation would assist recovery. In arriving at this conclusion, the Economic Committee took into account the persistence of the recession, the demonstrable failure of existing policies, and the growing respectability of devaluation in British academic circles (Copland, 1931; Endres, 1990a; Fleming, 1997a). In so far as devaluation could be expected to encourage an increase in exports relative to imports it would raise aggregate demand and GDP. However, in view of the inelasticity of export supply, and the fact that much industrial activity in New Zealand was based upon the assembly and finishing of imported semi-manufactures, the direct effect of devaluation would be constrained. Moreover, the losses of other groups would have to be set against the gains of farmers. Consumers would have to pay more for imports, real wages would fall, and sterling debtors, including the government, would face higher interest costs. While manufacturers would benefit from the protection afforded by devaluation, they would be forced to pay higher prices for imported machinery and semi-finished goods. Thus, the potential of devaluation to increase aggregate activity in New Zealand would depend upon whether it was accompanied by a loosening of monetary conditions. In deference to orthodoxy, the advocates of devaluation were forced to play down, or even deny, the potential for monetary expan-

John Singleton 181

sion in the wake of a reduction in the exchange rate. Devaluation was portrayed as a means of securing a fairer distribution of the misery produced by the slump, which had so far fallen most heavily on farmers. While farmers welcomed the prospect of devaluation, considerable opposition was aroused in other quarters. With the exception of the Bank of New South Wales, the commercial banks maintained that devaluation would undermine New Zealand’s reputation for financial stability. The bankers added, quite rightly, that there was no need for devaluation on balance of payments grounds since the London balances were rising. The Labour Party, notwithstanding its monetary radicalism, concluded that devaluation was unacceptable because it would increase the cost of living of the working class and the elderly (Sinclair, 1976, p. 101). Even the manufacturers opposed devaluation. The government was split over devaluation. Downie Stewart, the dour Minister of Finance, and A.D. Park, the Secretary of the Treasury, set their jaws firmly against a further change in the exchange rate. Downie Stewart warned that devaluation would lead to a fall in imports and customs revenue, and thus intensify the government’s budgetary problem. Moreover, a further fall in the external value of the currency would increase the burden of servicing the government’s sterling debt. Forbes could not make his mind up, and limply suggested that the decision should be left to the discretion of the banks. This was a recipe for further procrastination. Coates, the main supporter of devaluation in the cabinet, belatedly forced the issue in January 1933, with the result that the government asked the banks to devalue the currency to the level of £NZ125 to £stg100. With great reluctance the banks complied, in the knowledge that if they were defiant Coates would change the exchange rate by legislation. Stewart could not accept devaluation and promptly resigned, and Coates became Minister of Finance and the most powerful member of the government (Bassett, 1995, pp. 190–4). There was little support in New Zealand for a monetary or fiscal stimulus as an accompaniment to devaluation. Labour regarded devaluation and reflation as alternatives, and rashly promised to return the currency to parity with sterling. Some younger Treasury officials, including Bernard Ashwin, who had studied Keynes’s writings on monetary theory, believed that looser monetary conditions would speed economic recovery, with or without devaluation, but senior officials refused to listen to such heresy (Hawke, 1973, p. 118). Professor Allan Fisher of Otago University College was even more heretical, recommending floating the exchange rate, increasing the money supply, and encouraging the transfer of factors of production from farming into manufacturing and services. Fisher felt that the time was ripe for New Zealand to move beyond agriculture – but he was ploughing a lonely furrow (Endres, 1988). But most economists, including the members of the Economic Committee, argued that devaluation should be accompanied by further retrenchment aimed at limiting the increase in

182 New Zealand

domestic prices. They also feared that the banks might be tempted by the anticipated rise in London balances to initiate a credit boom. Ministers were very sensitive to warnings about inflation, and the devaluation package included the introduction of a new sales tax, further cuts in government expenditure, and the scaling down of mortgages. It was, however, accepted that devaluation would make it possible to retrench at a more leisurely pace (Copland, 1931; Bassett, 1995, p. 195). Only the commercial banks were capable of administering a monetary stimulus in the immediate aftermath of devaluation. New Zealand did not possess either a central bank or an official monetary policy until August 1934. The exchange rate debate in 1932 actually distracted ministers from the proposal for a central bank. In any case, the reason for establishing the Reserve Bank of New Zealand (RBNZ) was not to facilitate monetary expansion. The central bank originated in the determination of Treasury to achieve a clear separation between the monetary systems of New Zealand and Australia, so as to avoid repetition of the confusion which it was believed had led to the depreciation of 1930–1. Except in left-wing Labour circles, there was no suggestion that the central bank should be used to boost the money supply (Hawke, 1973, pp. 19–22). Parliament eventually decided that the RBNZ should be a private sector institution under close government control. Conservatives argued that its mission should be to defend the exchange rate and fight inflation. The left was split between pessimists who opposed the RBNZ on the grounds that it would be a creature of the international bankers, and optimists who pointed out that its powers of credit creation could be used to eliminate unemployment. Under the United–Reform coalition there was little prospect that the RBNZ would pursue radical policies, although there would be plenty of scope for a socialist government to point the central bank in a different direction (Hawke, 1973, pp. 31–6; Bassett, 1995, p. 204).

6

Assessment of devaluation

The 1933 devaluation gave a major boost to the farming community. As farmers emerged from the doldrums, they began to spend more on locally produced goods and services. Devaluation also helped those local manufacturers who did not depend upon imported materials to compete more effectively against imports. But not all segments of the farming community derived equal benefits from devaluation. Sheep farmers gained more from the change in the exchange rate than dairy farmers, since the latter had to contend with Danish retaliation. The Danes, who exported large quantities of dairy produce to Britain, mistakenly believed that New Zealand intended to reduce the sterling prices of its butter and cheese after devaluation. Denmark acted to protect its market share by devaluing the kroner and cutting the sterling prices of its own dairy goods. New Zealand exporters

John Singleton 183

were now forced to cut their prices in order to maintain sales. The only beneficiaries of this episode were British consumers (Kindleberger, 1934). Devaluation, without an accompanying monetary or fiscal stimulus, would have achieved little, other than to redistribute poverty from farmers to consumers, wage earners, and sterling debtors. Table 7.1 shows that in 1933 there was a significant increase in the money supply and a gradual fall in nominal interest rates, suggesting that the commercial banks made some adjustment to their strategies following devaluation. The question is whether this change of tack was sufficient to have made an appreciable impact on the economy. Hawke argues that the monetary stimulus accompanying devaluation was severely constrained by the Banks’ Indemnity Act, a measure that was designed to appease the commercial bankers. New Zealand bankers felt that the decision to devalue was mistaken, and predicted that within a couple of years it would be reversed. But in the interim there would be a large increase in London balances. The banks were uncomfortable about this prospect, and pointed out that the domestic currency value of the sterling balances would fall in the event of revaluation. They wanted the government to hold some of the additional London balances. Under the Banks’ Indemnity Act, the government agreed to sell 5 per cent Treasury bills to the banks in exchange for any surplus sterling in their possession. According to Hawke, this arrangement meant that the banks had no incentive to cut interest rates below 5 per cent (Hawke, 1985, pp. 119–20). Put another way, the banks believed that the situation had become more risky as a result of devaluation, and caution was their watchword. This attitude would have prevailed irrespective of the Banks’ Indemnity Act. Nominal and real interest rates increased in New Zealand relative to Britain in the early 1930s, indicating that the risk premium of investing in the dominions was rising, probably due to concern about the possibility of debt default. The fall in British interest rates after the abandonment of gold in 1931 was much more pronounced than the decline in New Zealand rates after devaluation in 1933.5 The action of New Zealand banks after devaluation may have been further constrained by the increase in country risk. On the other hand, there is no evidence that the threat of private capital flight was an issue in New Zealand until Labour was in office in the late 1930s (Hawke, 1985, pp. 164–5). As regards fiscal policy, coalition politicians did not jettison their commitment to a balanced budget. Although both the nominal and real budget deficits rose in the year to March 1934, a surplus was at last achieved in the following year, mainly due to a sharp recovery in government revenue.6 Keynes (1983, p. 438), writing in the mid-1930s, doubted whether the United–Reform coalition could have done any more to accelerate New Zealand’s recovery. He pointed out that a combination of reflation and low sterling prices would have led to a serious balance-of-payments crisis,

184 New Zealand

which is exactly what transpired in New Zealand under Labour in 1938–9. While the risks associated with reflation were genuine, the strength of the balance of payments in 1933 gave New Zealand more leeway for expansion than had been available to Australia when it devalued under duress in 1931. There is no doubt that the economy improved after devaluation: per capita incomes rose, unemployment fell, and industrial production increased. Although devaluation assisted recovery, its effects were muted by the lack of a central bank, the caution of the commercial bankers, the physical constraints on farm exports, and the persistent attractions of orthodoxy in government circles. The change in the exchange rate was only one of several factors contributing to recovery. By 1933, some commodity prices were already beginning to pick up. The sterling prices of mutton and wool reached their lowest points in 1932, and thereafter started a slow recovery. The sterling butter price reached its nadir in 1934 (Board of Trade, 1940, pp. 258–61). If London prices had continued to decline, the 1933 devaluation might not have led to higher incomes for farmers, although it might have prevented them from tumbling even further. It should be borne in mind that farm incomes failed to rise after the 1930–1 depreciation because sterling prices continued to fall. The improved outlook in 1933, at least for mutton and wool prices, was greatly to New Zealand’s advantage. Dominion farmers should also have been thankful for the comparative shallowness of the Depression in Britain (Eichengreen, 1988, p. 50). They were beneficiaries of Britain’s departure from gold to the extent that this contributed to the revival of the British economy and the recovery of commodity prices. Rising export volumes provided New Zealand farmers with some compensation for declining prices during the Depression. Output growth was a result of improvements in agricultural productivity and was unrelated to the exchange rate (Hawke, 1988, pp. 121–2). However, in so far as competitors, including Australian, Danish, Argentinian, and British farmers, also achieved productivity gains, further downward pressure must have been exerted on prices. Commercial-policy developments also made a contribution to the stabilization and recovery of the New Zealand economy. In 1932, Britain finally abandoned free trade and imposed tariffs on products, including some agricultural items, imported from foreign countries. It was not made clear whether agricultural imports from the dominions would also be subject to tariffs. This crucial question would be discussed later in the year during the imperial economic conference at Ottawa. At the Ottawa conference, the British granted a number of concessions to the dominions. They agreed to exempt dominion produce from the new butter tariff (Drummond, 1974, pp. 269–71), and, although they refused to impose tariffs on foreign meat, a system of meat quotas, which discriminated in favour of Australia and

John Singleton 185

New Zealand at the expense of Argentina, was introduced (Capie, 1978). No tariff or quota restrictions were applied to imports of wool. New Zealand undertook to increase tariffs on imports of manufactures from foreign countries in order to give British goods a wider margin of preference. Although New Zealand played a minor role at Ottawa, and the most important negotiations were between Britain, Australia, and Canada, it reaped the rewards of British concessions to the larger dominions. Cain and Hopkins (1993, pp. 83–7) state that Britain’s key objective at Ottawa was to avert a sterling debt crisis by arranging for an increase in the dominions’ export revenues. Ottawa gave a boost to New Zealand farmers’ confidence and helped them to combat foreign competition in the markets for butter and, in particular, meat. Drummond (1974, pp. 280–3), however, concludes that the Ottawa settlement was a secondary factor in the economic recovery of the dominions. He argues that the imperial conference focused on the wrong issues, and missed a great opportunity to devalue the empire currencies against sterling. In 1935, the coalition parties were punished for having been in power earlier in the decade, and the first Labour government was elected. The economy was well on the way to recovery when Labour entered office, although this is rarely acknowledged in popular mythology. One of Labour’s first moves was to nationalize the RBNZ. This was a symbolic act because the government already possessed extensive powers over the central bank. More significantly, Labour instructed the RBNZ to manufacture credit to fund its expenditure plans, including a large public works programme and a scheme to guarantee the incomes of dairy farmers. The 1933 devaluation gave Labour a certain amount of freedom in the areas of fiscal and monetary policy. Nevertheless, ministers promised to restore the currency to parity with sterling in order to reduce the cost of living. They did not achieve their goal until 1948. Higher government spending helped to maintain the recovery in 1936 and 1937, but aggravated a severe balance of payments crisis, caused by a sharp dip in world commodity prices, in 1938–9. This external crisis was managed by the introduction of exchange and import controls, an ominous departure that set the stage for the postwar policy of insulationism (Hawke, 1973, pp. 65–71, 82–100; Gustafson, 1986, pp. 165–6, 171, 185, 225; Sinclair, 1976, pp. 123–4, 155).

7

New Zealand and imperial monetary cooperation

While this chapter is principally concerned with New Zealand, it should not be forgotten that New Zealand was part of a wider imperial financial network called the sterling bloc. This section considers whether there was any scope for a coordinated response to the Depression by Britain and the sterling dominions. The lack of international monetary cooperation between the wars is a theme of Eichengreen and Temin. Much of their

186 New Zealand

work deals with relationships between the great powers, such as the United States, Britain, France, and Germany, but it is also worth examining the extent of financial cooperation between other groups of countries including the members of the British empire. Although monetary cooperation in the empire was rudimentary between the wars, the foundations were laid during the 1930s for the much closer coordination of imperial economic policy that emerged in the 1940s. Aside from Canada, the fiscally autonomous members of the empire fixed their currencies against sterling during the 1930s (Drummond, 1981). Refusal to accompany Britain off gold would have entailed a painful revaluation against sterling, as South Africa discovered when it initially declined to join sterling’s float. Eichengreen and Irwin (1995) suggest that the sterling bloc gained from its collective depreciation against the gold standard countries in the 1930s. Sterling bloc members, including Britain, were able to relax their tight monetary policies once gold had been abandoned. Notwithstanding the general desire for monetary cooperation within the empire, there was bitter disagreement over the nature of such cooperation. The British thought in terms of establishing a mechanism to ensure that the ‘colonial savages’ in the dominions, to use the terminology of Sir Otto Niemeyer of the Bank of England, followed an orthodox monetary line (Cain, 1996, p. 342). London advocated the creation of a chain of imperial central banks. It was hoped that these banks would be placed under the supervision of the Old Lady of Threadneedle Street (Cain and Hopkins, 1993, p. 111). However, while dominion governments were not averse to the formation of central banks, they were not prepared to hand over their control to the Bank of England. Hawke (1973, pp. 40, 43) argues that, by 1934, when the RBNZ was established, the Bank of England was reconciled to the fact that it could not manipulate the dominion central banks, and was willing to settle for a genuine partnership. But, in view of the Bank of England’s attempt to coerce Canada into joining the sterling bloc, it seems unlikely that the British had a real change of heart (Cain, 1996). In New Zealand, Coates, and his Labour successor as Minister of Finance, Walter Nash, were adamant that the RBNZ should implement the policies of the government in Wellington and not take orders from London. The first governor of the RBNZ was a Bank of England man, Leslie Lefeaux, but he was not permitted to influence policy, and his repeated protests over Labour’s plans for spending and credit creation were overruled (Hawke, 1973, pp. 66–9). Dominion governments did not share the opinion of the Bank of England that the purpose of imperial economic cooperation was to achieve passive conformity to British monetary strategy. On the contrary, the dominions desired a role in the formulation of British monetary policy. They maintained, with some justification, that a more expansionary British monetary policy would greatly benefit imperial primary producers. In 1931, Professor J.B. Condliffe, a New Zealander working for the League of Nations, suggested that

John Singleton 187

a network of imperial central banks could be used to implement a collective monetary stimulus (Endres, 1991, p. 189). New Zealand supported a Canadian move to put empire monetary policy on the agenda of the Ottawa conference (Drummond, 1974, p. 215). At Ottawa, Coates urged the British to use monetary policy to raise the sterling price level (Bassett, 1995, p. 185; Hancock, 1940, p. 195). This proposal met with an icy British response to the effect that the dominions had no right to interfere in the management of sterling (Drummond, 1981, pp. 22–6). The British maintained that the only safe way to increase commodity prices was for dominion producers to restrict output. British officials regarded their dominion counterparts as amateurs whose advice on monetary policy was at best worthless. In 1933 Australia and New Zealand called for the establishment of an imperial committee to discuss measures to raise sterling commodity prices, but this proposal was branded as another attempt to foist inflationary policies on the British government (Drummond, 1981, p. 160). After 1935, the New Zealand Labour government continued to advocate expansionary policies at the imperial level. At the 1937 imperial conference, New Zealand opposed a South African proposal that preparations be made to restore sterling to the gold standard (Drummond, 1981, p. 232). It was not irrational for the Forbes–Coates coalition to uphold caution in domestic monetary policy, while calling for monetary expansion in the empire. New Zealand farmers sought higher sterling commodity prices. Although sterling prices would not have been affected by reflation in New Zealand alone, they would have been increased by reflation in Britain and the rest of the sterling bloc. Unfortunately, this strategy was not feasible in the 1930s. As Temin (1989, p. 104) points out, the British never really abandoned the gold standard mentality.

8

Conclusion

New Zealand’s Depression experience was broadly similar to that of other temperate primary producers such as Australia and Canada. Political leaders, civil servants, and economists were unprepared for the severity and duration of the slump. Understandably, New Zealand had no homegrown solution to the crisis, and chose to follow the orthodox policy of deflation, at least until 1933. The devaluation of 1933 was not forced on New Zealand by external financial pressure, and was justified as a measure to boost farm incomes. Even after devaluation, the government remained committed to retrenchment and balancing the budget. Although the redirection of economic policy assisted New Zealand to recover from the Depression, devaluation was not the only, or necessarily the most important, stimulant of growth. A firmer British market, rising agricultural productivity, and the protection afforded by the Ottawa Agreement all contributed to the revival of incomes and employment in New Zealand in the mid-1930s. Some loos-

188 New Zealand

ening in monetary conditions occurred between 1933 and 1935, but a deliberate policy of reflation was not attempted until Labour came into office in 1935. In any case, reflation in Britain would have been of far greater benefit to New Zealand during the slump than reflation at home. New Zealand, in alliance with Australia and Canada, called for an expansionary imperial monetary policy, but Britain steadfastly refused to accept responsibility for stimulating imperial economic recovery, and portrayed the advocates of imperial reflation as cranks. New Zealand’s revival was smoother and speedier than the recoveries of the United States and the remaining gold standard economies. Even more satisfactory results would have been achieved if Britain had made constructive use of its financial leadership of the empire. Notes * 1

2 3

4

5 6

I would like to thank Gary Hawke and Theo Balderston for commenting on an earlier draft. Tocker’s understanding of the adjustment mechanism has been widely accepted by New Zealand economists. But it is possible to conceive of alternative mechanisms. For instance, it could be argued that a drop in London commodity prices would result in a fall in the expectation of profits in New Zealand, a decline in capital inflows, and a reduction in the demand for money in the dominion. I owe this point to Theo Balderston, who makes reference to work on Canada under the gold standard by Dick and Floyd (1991). This occurred before Britain left gold. There is some dispute over the course of aggregate income and the extent of unemployment during the Depression. For an introduction to this debate, see Chapple (1994) and Rankin (1994a, 1994b). Throughout this chapter, a fall in the exchange rate is understood in its modern sense, as a reduction in the value of domestic currency in terms of foreign currency. But many interwar authorities describe devaluation as an increase in the exchange rate, by which they mean that foreign currency becomes more expensive in terms of domestic currency. When reading the economic literature of the 1920s and 1930s, the reader must be on guard in order to avoid confusion. For data on British interest rates, see Mitchell (1988, pp. 682–3), and British retail prices, see Feinstein (1972, p. T140). The constant employment budget balance, which corrects the nominal balance for cyclical changes in expenditure and revenue, was already in surplus in the year ending March 1933 and remained in surplus until the year ending March 1937 (Abbott, 1995).

References Abbott, M. J. (1995) ‘The Real Structural Imbalance and Fiscal Stance in New Zealand During the Interwar Years’, New Zealand Economic Papers, 29: 63–76. Bassett, M. (1993) Sir Joseph Ward (Auckland: Auckland University Press). ——– (1995) Coates of Kaipara (Auckland: Auckland University Press). Bennett, J. (1997) ‘Social Security, the “Money Power” and the Great Depression: The International Dimension to Australian and New Zealand Labour in Office’, Australian Journal of Politics and History, 43: 312–30.

John Singleton 189 Board of Trade (1940) Statistical Abstract for the United Kingdom, 1924 to 1938 (London: HMSO). Cain, P.J. (1996) ‘Gentlemanly Imperialism at Work: The Bank of England, Canada, and the Sterling Area, 1932–1936’, Economic History Review, XLIV: 336–57. ——– and A.G. Hopkins, (1993) British Imperialism Vol.2: Crisis and Deconstruction 1914–1990 (London: Longman). Capie, F. (1978) ‘Australian and New Zealand Competition in the British Market 1920–39’, Australian Economic History Review, 18: 46–63. Chapple, S. (1994) ‘How Great Was the Depression in New Zealand? Neglected Estimates of Inter-war Aggregate Income’, New Zealand Economic Papers, 28: 195–203. Copland, D.B. (1931) New Zealand Exchange and the Economic Crisis (Christchurch: Whitcombe & Tombs). Dick, T.J.O. and J.E. Floyd (1991) ‘Balance of Payments Adjustment under the International Gold Standard: Canada 1871–1914’, Explorations in Economic History, 28: 209–38. Drummond, I.M. (1974) Imperial Economic Policy 1917–1939 (London: George Allen & Unwin). ——– (1981) The Floating Pound and the Sterling Area (Cambridge: Cambridge University Press). Eichengreen, B. (1988) ‘The Australian Recovery of the 1930s in International Comparative Perspective’, in R.G. Gregory and N.G. Butlin (eds), Recovery from the Depression: Australia and the World Economy in the 1930s (Cambridge: Cambridge University Press). ——– (1992) Golden Fetters. The Gold Standard and the Great Depression 1919–1939 (New York: Oxford University Press). ——– and D.I. Irwin (1995) ‘Trade Blocs, Currency Blocs and the Reorientation of World Trade in the 1930s’, Journal of International Economics, 38: 1–24. Endres, A.M. (1988) ‘“Structural” Economic Thought in New Zealand: The Interwar Contribution of A.G.B. Fisher’, New Zealand Economic Papers, 22: 35–49. ——– (1990a) ‘The Development of Economists’ Policy Advice in New Zealand, 1930–4: With Particular Reference to Belshaw’s Contribution’, Australian Economic History Review, 30: 64–78. ——– (1990b) ‘The Economics of Wages and Wage Policy in the Depression and Recovery Period: Distinctive Elements in the New Zealand Debate’, New Zealand Journal of Industrial Relations, 15: 1–18. ——– (1991) ‘J.B. Condliffe and the Early Canterbury Tradition in Economics’, New Zealand Economic Papers, 25: 171–97. ——– and G.A. Fleming (1995) ‘Monetary Thought and the Analysis of Price Stability in Early Twentieth Century New Zealand’, New Zealand Economic Papers, 29: 173–94. ——– and K.E. Jackson (1993) ‘Policy Responses to the Crisis: Australasia in the 1930s’, in W.R. Garside (ed.), Capitalism in Crisis: International Responses to the Great Depression (London: Pinter). Feinstein, C.H. (1972) National Income, Expenditure and Output of the United Kingdom, 1855–1965 (Cambridge: Cambridge University Press). Fleming, G. (1997a), ‘Keynes, Purchasing Power Parity and Exchange Rate Policy in New Zealand during the 1930s Depression’, New Zealand Economic Papers, 31: 1–14. ——– (1997b) ‘Economists and Mortgage Relief in New Zealand in the 1930s’, Australian Economic History Review, 37: 54–68. ——– (1999) ‘Agricultural Support Policies in a Small Open Economy: New Zealand in the 1920s’, Economic History Review, LII: 334–54.

190 New Zealand Groenewegen, P. and B. McFarlane (1990) A History of Australian Economic Thought (London: Routledge). Gustafson, B. (1986) From the Cradle to the Grave: A Biography of Michael Joseph Savage (Auckland: Reed Methuen). Hancock, W.K. (1940) Survey of British Commonwealth Affairs, Vol. II, Part 1 (London: Oxford University Press). Hawke, G.R. (1971), ‘New Zealand and the Return to Gold in 1925’, Australian Economic History Review, 11: 48–58. ——– (1973), Between Governments and Banks: A History of the Reserve Bank of New Zealand (Wellington: A. R. Shearer). ——– (1985) The Making of New Zealand (Cambridge: Cambridge University Press). ——– (1988) ‘Depression and Recovery in New Zealand’, in R.G. Gregory and N.G. Butlin (eds), Recovery from the Depression: Australia and the World Economy in the 1930s (Cambridge: Cambridge University Press). Keynes, J.M. (1983) ‘Review of Report of Monetary Committee, 1934, New Zealand’, in The Collected Writings of John Maynard Keynes, ed. D. Moggridge, vol. XI (London: Macmillan): 435–9. Kindleberger, C.P. (1934) ‘Competitive Currency Depreciation Between Denmark and New Zealand’, Harvard Business Review, 12: 416–26. King, J.E. (1988) Economic Exiles (Basingstoke: Macmillan). Mabbett, D. (1995) Trade, Employment and Welfare: A Comparative Study of Trade and Labour Market Policies in Sweden and New Zealand, 1880–1980 (Oxford: Clarendon Press). MacDonald, B. and D. Thomson (1987) ‘Mortgage Relief, Farm Finance, and Rural Depression in New Zealand in the 1930s’, New Zealand Journal of History, 21: 228–50. Mitchell, B.R. (1988) British Historical Statistics (Cambridge: Cambridge University Press). Rankin, K. (1992) ‘New Zealand’s Gross National Product: 1859–1939’, Review of Income and Wealth, 38: 49–69. ——– (1994a) ‘Comment: How Great Was the Great Depression’, New Zealand Economic Papers, 28: 205–9. ——– (1994b) ‘Estimating Unemployment: New Zealand, 1933’, Paper Presented at Conference of the Economic History Society of Australia and New Zealand, University of New England, NSW, July. Robertson, R.T. (1982) ‘Government Responses to Unemployment in New Zealand, 1929–35’, New Zealand Journal of History, 16: 21–38. Sheppard, D.K., K. Guerin and S. Lee (1990) ‘N.Z. Monetary Aggregates and the Total Assets of Leading Groups of Financial Institutions, 1862–1982’, Victoria University of Wellington, Money and Finance Group Discussion Paper No. 11. Sinclair, K. (1976) Walter Nash (Auckland: Auckland University Press). ——– and W.F. Mandle (1961) Open Account: A History of the Bank of New South Wales in New Zealand 1861–1961 (Wellington: Whitcombe & Tombs). Temin, P. (1989) Lessons from the Great Depression (Cambridge, MA: MIT Press). Tocker, A.H. (1924) ‘The Monetary Standards of New Zealand and Australia’, Economic Journal, 34: 556–75.

8 The Soviet Union during the Great Depression: The Autarky Model Paul R. Gregory and Joel Sailors

1

The Soviet economy as an outlier

The Great Depression affected the industrialized powers at different times and in different ways. Some suffered steep, others small, production declines; some recovered slowly, others more quickly. Despite these differences, no major industrialized market economy escaped significant economic losses from the Great Depression/Slump of the 1920s and 1930s. Other studies in this collection examine the experiences during the Great Depression of economies that belonged to the global trading and monetary systems, trying to isolate its causes – monetary, exchange rate, trade restrictions, technology shocks, and so on. The Soviet Union represents the sole example of a planned socialist economy, bent no less on autarky. Having created an administrative-command system designed to insulate the domestic economy from demand and external shocks and independent of international capital flows, the Soviet economy experienced rapid economic growth and industrial transformation during the very period when other economies were stagnating. Moreover, the Soviet Union was the sole case of a country that followed a deliberate development strategy, executed by administrative actions rather than markets, to oversee its rapid industrialization. It was this exceptional Soviet experience that lent it much of its appeal in the 1950s and 1960s in Western Europe and more so in Africa, Asia, and Latin America. The leadership of the Soviet Union approved the first Five-Year Plan in 1928 and began forced collectivization in 1929, at the same time as the United States was entering its Great Depression. The Soviet decisions on industrialization and collectivization followed a decade of debate and political infighting between the two major factions of the Communist Party over the proper course of development and the appropriate economic system for the world’s first major socialist power. A short ten years later, on the eve of the Second World War, the Soviet economy was transformed from private ownership in agriculture and small-scale industry to almost total state ownership, from producing agricultural products and industrial 191

192 The Soviet Union

raw materials to producing heavy industrial products, and was transformed from a predominantly rural society to an urban society. The Soviet Union’s economic relations with other countries changed radically as well: while the tsarist economy had been strongly integrated into world capital and product markets, the Soviet planned economy had virtually cut itself off from the outside world. In gauging the speed of Soviet transformation, Simon Kuznets (1963, pp. 345–7) wrote the following: the rapidity of this shift was far greater in the USSR than in the other developed countries … the shift of labor out of agriculture of the magnitude that occurred in the USSR in the 12 years from 1928 to 1940 took from 30 to 50 years in other countries. Comparative Soviet growth: the Depression years Figures 8.1 and 8.2 compare Soviet real national income and real investment spending (using the most widely accepted Western recalculations of Soviet GDP) with the major industrialized European countries and the United States during the 1920s and 1930s.1 Unlike the market-economy series, which commence in 1920, the Soviet figures begin in 1928 – the first year of the Five Year Plan era. The Soviet economy had recovered by 1928 from the ravages of the First World War and the Russian civil war, and 1928 is hence a good starting point for the Russian series. Figure 8.1 and especially Figure 8.2 distinctly show the uniqueness of the Soviet experi-

250

200

Germany gdp Italy gdp UK gdp USSR gdp France gdp US gdp

150

100

50

0 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40

Figure 8.1

GDP indexes: USSR and other countries, 1920–40 (1929 = 100)

Paul R. Gregory and Joel Sailors 193 500

400

300 Germany cf Italy cf UK cf USSR cf

200

100

0 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 –100

Figure 8.2

Capital formation indexes: USSR and other countries, 1920–40 (1928 = 100)

ence: the USSR experienced rapid expansion of output and investment between 1928 and 1937 while the major industrialized countries recorded stagnant or declining real output and investment. While, in the major industrialized economies, consumption smoothing made the decline in investment spending greater than in GDP, investment spending outgrew GDP by a significant margin in the Soviet Union. In 1937, the Soviet investment rate was double what it had been in 1928. If we had used official Soviet statistics for this period, the Soviet growth performance of the 1920s and 1930s would have been even more astonishing, but official Soviet statistics have long been discredited, particularly for the 1930s. With the collapse of the Soviet Union in December of 1991, Russian economists have increasingly disputed measures of Soviet growth during the late 1920s and 1930s. A growing number of Russian and émigré economists now argue that even the widely used Western recalculations of Soviet economic growth for the period 1928 to 1940 are seriously overstated. Although recent criticisms of Soviet historical statistics provide some intriguing evidence of overstatement of 1930s growth, we still lack solid evidence to dispute the fact that Soviet growth was substantial during the period 1928 to 1940 (Khanin, 1993; Erickson, 1990). This chapter seeks to explain why the Soviet Union’s experience was so different from the rest of Europe’s during the period of the Great Depression. It re-examines the basic features of the Soviet administrativecommand economy and how this type of economic system shielded the domestic economy from external disturbances and generated capital solely through domestic saving. The chapter contrasts the Soviet experience with

194 The Soviet Union

that of the Russian Empire during its final years, and it concludes with some simple cost/benefit considerations.

2

The late Tsarist period

The 30 years leading up to the 1917 Bolshevik revolution saw substantial progress and economic growth of the Russian economy. This growth was made possible by the gradual emancipation of the serfs (beginning with juridical emancipation in 1861), the building of a railway network, and more than a decade of fiscal austerity that made it possible to become a gold standard country in the late 1890s. The result of this activity was an annual growth rate in excess of 3 per cent, that was shared equally by industry and agriculture. By 1913, the Russian Empire was the world’s fifth largest economy, but still relatively poor on a per capita basis.2 In 1913, Russia was indeed an open economy. However, as a very large and resource-rich country, its trade ratios were a relatively low 15 per cent of national income (Gregory, 1982, pp. 185–6). Moreover, Russian economic development during the late tsarist period was strongly dependent upon Western capital. In 1914, Russia was the world’s largest debtor nation accounting for some 15 per cent of world external indebtedness. Some 15 per cent of Russia’s net capital formation had been supplied by net foreign investment (Gregory, 1982, table 1; Gregory, 1996, pp. 466–9). As an open economy, the Russian economy was subject to external shocks emanating from abroad. Figures 8.3, 8.4 and 8.5 show that Russian 120

100 Russia gdp UK gdp Germany gdp France gdp US gdp Denmark gdp Sweden gdp Norway gdp Italy gdp

80

60

40

20

85 86 87 88 89 90 91 92 93 94 95 96 97 98 9 19 9 00 1 2 3 4 5 6 7 8 9 10 11 12 13

0

Figure 8.3

GDP indexes: Russia and other countries, 1885–1913 (1913 = 100)

Paul R. Gregory and Joel Sailors 195 140 120 100

Russia cf UK cf Germany cf France cf US cf Denmark cf Sweden cf Norway cf Italy gdp

80 60 40 20

85 86 87 88 89 90 91 92 93 94 95 96 97 98 9 19 9 00 1 2 3 4 5 6 7 8 9 10 11 12 13

0

Figure 8.4 Capital formation indexes: Russia and other countries, 1885–1913 (1913 = 100) 120

100

Russia p UK p Germany p France p US p Denmark p Sweden p Norway p Italy p

80

60

40

20

0 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 900 1 2 3 4 5 6 7 8 9 10 11 12 13 1

Figure 8.5

Price indexes: Russia and other countries, 1885–1913 (1913 = 100)

real output and investment followed European cycles via a complex lag structure. Price movements appeared even more closely linked. As an economy largely dependent on agriculture, some of Russia’s output and investment fluctuations were caused by agricultural shocks, such as the

196 The Soviet Union

famine of 1891–2. The major independent shocks to the Russian economy, however, emanated from political events: namely, the revolution of 1905 – which Figures 8.3 and 8.4 show had a major negative impact on real output and investment – and then the First World War, Bolshevik Revolution, and Russian Civil War. That tsarist Russia was caught up in the Western business cycle comes as no surprise. Gold standard economies are supposed to be interdependent through financial- and price-transmission mechanisms. A price shock in one country (say, a deflation in the rest of the world) will cause money to flow from the domestic economy to the rest of the world and the domestic economy will automatically deflate. There have been no special studies of Russia’s terms of trade during the period 1885 to 1913, but given the similarity between Russia’s and the United States’ export and import composition we would expect these two countries to have experienced similar trends in the terms of trade. Studies of the United States for this period show fairly constant terms of trade (see Taussig, 1927, pp. 418f.). We suspect therefore that Russian terms of trade were also relatively unchanged during the period 1885 to 1913. The Soviet Union entered the interwar period with an economy that produced less than half of its pre-war peak, about one-third of peak industrial production, and an economy largely cut off from international product and capital markets. The Soviet default on tsarist debt of 1914 cut off the USSR from world credit markets, despite some efforts by the Soviet leadership to regain access. The period from 1914 to 1928 witnessed the severe losses in output associated with the First World War and the Russian Civil War and the recovery of output to pre-war levels during the New Economic Policy of 1921 to 1928. Similarities: USSR and other countries: the interwar years There were many parallels in the unfolding events experienced by Russia and Germany between the First and Second World Wars. Both had very large foreign debts after the First World War. Russia deliberately accumulated hers in the three decades leading up to the war, while German debt was imposed by the Allies. Russia repudiated its foreign debt, while Germany had no choice but attempt to pay, which ultimately proved impossible. Both were devastated by the First World War. Both had greater difficulties in regaining their former level of output than other European countries. Both suffered severe wrenching economic shocks – the Soviet Union in remaking the institutional basis of its economy from capitalism to command, while Germany experienced both hyperinflation and deep depression (see Ritschl in this volume) in trying to service its external debt and turned toward ‘state capitalism’ in its frustration. Both Germany and the USSR went through severe political shocks. In the Soviet Union the whole social and political fabric of country was rent asunder and replaced by a dictatorial political system. Germany entered a somewhat similar experience with the rise of the Nazi era.

Paul R. Gregory and Joel Sailors 197

Both the USSR and Germany were attracted to mercantilist policies. Russia continued mercantilist autarkic economic policies that were made easier by the vast geographic size of the country. Germany increasingly adopted autarkic policies but with policies and trade aimed toward enlarging and absorbing political and economic areas that would enable fully autarkic conditions to be met. Russia emphasized military preparedness as facilitated by the mercantilist policy, as did the Germans. Germany asymptotically approached a command economic by using some economic tools such as confiscation, bureaucratic control of exports and imports, licences, quotas, embargoes, barter arrangements, and exchange restrictions. These measures isolated Germany from the rest of the world except some of its near neighbours in their currency area. As other countries entered the Great Depression their policies came to resemble those of mercantilism and Soviet practices. In international trade, international traders of Soviet trading companies filled out their inputquantity requirements for their national ‘input-output table’ while the many individuals and enterprises of capitalist countries filled out their individual niches in their national and international prices ‘table of comparative advantage and disadvantage’. The above differences in the two types of economy also applied to strictly domestic economy decisions. For example, in one domestic policy area alone, to influence economic conditions Russian changes in savings were ‘forced’ by the policy of imposing the ‘non-equivalent exchange’ in command transactions involving agricultural products (see below), while other industrial countries would alter the background of saving decisions, i.e., the interest rate, in order to entice individuals and enterprises to change their savings rates. However, during the Great Depression other industrial countries increasingly relied upon more mercantilist-type intrusive measures in economic life. Germany, as noted above, served as the more extreme example of industrial country response to the Great Depression. One example may suffice to illustrate a more intrusive government in economic life. The US during the Great Depression increased the content of mercantilism in its panoply of domestic and international policies. Domestically, the US thought for quite some time that the problem was one of low prices – especially in agriculture. Thus, the US embarked on attempts to raise agricultural prices through the provisions of the Agricultural Adjustment Act by reducing the supply of many specific farm products. But to reduce specific crop acreage of private farms a majority of the farmers had to join the government-arranged ‘cartel’ and agree to the reduction. This brought more government involvement in private affairs since bureaucrats has to be assured that individual farmers complied on the limitation of acreage of their crops or numbers of farm animals and/or their product. Throughout the nation other government agencies hired people, for example, to measure the land that each individual farmer had in each

198 The Soviet Union

of the targeted crops. These policies had an effect on the national market supply but the government possessed a programme to buy and store crops if the above programme was not fully successful. The measures did lift the domestic prices of those crops above the world price, which meant that the government must interfere in the market both by not letting in designated crops from foreign nations attracted by those bloated prices and by causing the government to buy and then store in order to ‘sterilize’ their effects. This caused the US to adopt restrictive foreign trade methods such as embargoes and/or quotas. All of this further hurt the allocation of world resources and level of foreign trade.

3 Preparation for industrialization: Soviet development strategy The Soviet Union is the first case of a country that actually devised a deliberate strategy for economic growth. This was done in the 1920s in the course of what has come to be called the Soviet Industrialization Debate.3 The ensuing strategy was the result of intellectual debate, but it was also the consequence of unfolding events and political designs, which were largely divorced from abstract theorizing about economic development strategy. The issues of the industrialization debate The basic issues of the Soviet industrialization debate, which pitted the Bukharin wing of the Bolshevik Party against the Trotsky wing, were about property rights (the extent of private ownership in agriculture and trade insofar as much of large-scale industry had already been nationalized), the role of markets (private trade versus forced sales to state monopolies), sources of capital accumulation (personal saving versus forced saving), sectoral priorities (priority to heavy industry or to light industry and agriculture), and the role of foreign trade and foreign capital. The left wing favoured ‘superindustrialization’, heavy industry priority, complete nationalization, state trading monopolies, trade autarky, and savings forced from the peasant population. The right wing favoured balanced growth, a prosperous private agriculture, and heavy reliance on trade. The arguments of the left wing eventually won the day. Stalin initially allied himself with the right wing to rid himself of the political danger posed by the popular Trotsky, after which he turned against the right wing. Stalin’s solution was apparent in the party’s choice of the more ambitious variant of the first Five Year Plan in 1928 which called for ‘superindustrialization’ and in the March 1929 decision to force collectivization. The characteristics of the Soviet Development Model, which emerged from this debate, are well known.4 This chapter addresses those aspects of the model which allowed the Soviet economy to grow rapidly during the very period when the rest of the world was in a deep economic slump.

Paul R. Gregory and Joel Sailors 199

The new Bolshevik leadership of the Soviet Union, in their considerations of development strategy, were clearly uncomfortable with dependence upon Western product and capital markets. The desire to avoid further dependence was an integral plank of the left-wing programme put forward during the industrialization debate. The left wing, whose arguments were made by the vocal E.A. Preobrazhensky, made the following arguments in favour of autarky: 1 The Soviet Union was surrounded by political enemies, who could use economic leverage to obtain advantage. 2 The Soviet Union, having defaulted on tsarist debt in 1918, could not obtain credits in the West, irrespective of its political views, without entering into a costly amortization programme which would deprive the USSR economy of its domestic capital formation.5 3 Marx had emphasized the chaotic nature of market economies with their crises of overcapacity, underconsumption, and so on. To integrate the Soviet economy into the world economy would subject it to these crises. It would be better to have an autarkic planned economy which would avoid these disturbances. 4 The Soviet economy, like the tsarist economy, possessed a comparative advantage in agriculture. An integrated Soviet economy would therefore have to pay for imports via agricultural exports. Emphasis on agricultural exports would expose the Soviet economy to two dangers: One was the reliance on a private peasantry, which opposed Soviet power, and, two, the danger of a collapse in agriculture’s terms of trade in world markets. 5 The Soviet Union could raise its domestic saving rate sufficiently to finance its capital formation without foreign capital. The financing of a rising investment rate through domestic savings alone was to be achieved by what Preobrazhensky termed ‘non-equivalent exchange’ with the countryside, whereby a state purchasing monopoly would buy farm products at low delivery prices, sell at higher prices, and collect the difference as a tax for capital formation. It was Stalin who recognized that considerable force was required in the countryside to insure deliveries at such unfavourable terms of trade. Collectivization was his answer to this dilemma. Stalin, who had consolidated his power by the second half of the 1920s, came down on the side of the left wing (after personally sending them either to prison or exile). It is unclear whether he was with the left wing from the start or whether experience changed his mind. What he had witnessed was a deterioration of the USSR’s relations with Western Europe, a series of agricultural ‘crises’ in the USSR, and his own experience, which he regarded as successful, of forcible collection of grain from a recalcitrant

200 The Soviet Union

peasantry in the Urals and Siberia. By the end of the 1920s, he was ready to move forcefully to establish an administrative-command system that could carry out his goals. Characteristics of Soviet industrialization in the 1930s Table 8.1 shows that the actual Soviet industrialization pattern that emerged after 1928 bears a close resemblance to the industrialization programme proposed by the party’s left-wing: Soviet economic growth was heavily biased in favour of industry in general and of heavy industry in particular. Industrial production grew at an annual rate of 11 per cent, whereas agricultural production – historically a prime export product – grew at an annual rate of only 1 per cent between 1928 and 1937. The negative rate of growth of livestock graphically indicates the impact of collectivization upon agricultural performance. The same trends are apparent in the differential rates of growth of the agricultural and non-agricultural labour forces between 1928 and 1937; the former declined, while the latter expanded rapidly at an annual rate of almost 9 per cent. The structural transformations resulting from these differential sector growth rates are impressive. Agriculture’s shares of not national product and the labour force declined from 49 and 71 per cent in 1928–9 to 29 and 51 per cent in 1940, whereas the increase in industry’s product and labour force shares was from 28 and 18 per cent, respectively, to 45 and 29 per cent, respectively, during the same period. The most remarkable feature of the 1930s was the extent to which the pro-heavy-industry bias asserted itself (as Preobrazhensky said it should). Between 1928 and 1937, heavy manufacturing’s net product share of total manufacturing more than Table 8.1

Characteristics of Soviet industrialization, 1928 and 1937 (in per cent) 1928

Growth rates, 1928–37 GNP Consumption Investment Shares of GNP Consumption Investment GNP shares, sector of origin Agriculture Industry Socialist sector shares Capital stock Agricultural production

1937 4.8 0.8 14.4

80.0 13.0

53.0 26.0

49.0 28.0

31.0 45.0

65.0 3.3

99.6 98.5

Source: Paul Gregory and Robert Stuart, Soviet and Post Soviet Economic Structure and Performance (New York: HarperCollins, 1994), p. 88.

Paul R. Gregory and Joel Sailors 201

doubled, from 31 per cent to 63 per cent, whereas light manufacturing’s product share fell from 68 per cent to 36 per cent. The impact of this production programme on real consumption levels in the absence of significant foreign trade had already been foreseen by Preobrazhensky. Between 1928 and 1937, household consumption scarcely grew (at an annual rate of 0.8 per cent), and the share of consumption in GNP (in 1937 prices) declined markedly, from 80 per cent to 53 per cent. During the same period, gross capital investment grew at an annual rate of 14 per cent, and the ratio of gross investment to GNP doubled, from 13 per cent to 26 per cent. If we define total consumption expenditures to include both private consumption and communal services, and non-consumption expenditures to include investment, government administration, and defence, then total consumption fell between 1928 and 1937 from 85 per cent of GNP to 64 per cent of GNP. The changing institutional setting within which these transformations were occurring should also be noted: between 1928 and 1937, the socialist sector share of total capital stock, industry, agriculture, and trade expanded sharply, so that by 1937 the socialist sector totally dominated all economic activity. Consumer prices rose by 700 per cent between 1928 and 1937 and probably would have risen even faster without the extensive rationing of the period. Average realized prices of farm products, which are weighted averages of the extremely low state procurement prices, the above-quota state delivery prices, and collective-farm prices, on the other hand, rose by 539 per cent. If there were major disappointments in the execution of the Soviet grand design for industrialization, these disappointments would have been the decline in livestock herds as peasants resisting collectivization slaughtered their livestock rather than contribute them to the collective farms, and the failure of agriculture to serve as a source of net saving for industry. As it turned out, the slaughter of livestock forced a massive replacement of animal draft power with mechanical power, and the diversion of industrial resources to the production of agricultural equipment resulted in either a zero net flow of resources from agriculture to industry or, perhaps, a flow of net resources from industry to agriculture. Autarky and the administrative-command system We need not cover all aspects of the Stalinist administrative-command system – only those that shielded the Soviet economy from the external shock of the Great Depression and contributed to the rapid rise in its rates of domestic capital formation. From its very beginnings, the administrative-command economy was based upon a balance mentality. The job of planners was to balance the supplies and demands of various industrial and agricultural commodities by administrative means. Accordingly, planners naturally regarded

202 The Soviet Union

imports as ‘good’ and exports as ‘bad’. Imports made the problem of balancing supplies and demands easier; exports detracted from available supplies and made the balancing task more difficult. Planners also rejected the notion of comparative advantage because the task of planners was to create the economy desired by political authorities unrestricted by ‘economic laws’. If political authorities wanted heavy industry and not agriculture, even if the economy’s comparative advantage was in agriculture, so be it. Note the similarities between the trade policies espoused by Soviet planners and pre-mercantilism. Pre-mercantilists made the Soviet-like distinction between ‘hunger for goods’, maximizing consumption, and ‘fear of goods’. During the Middle Ages, for example, the policy was to discourage exports and encourage imports. In France, for example, there was a general prohibition on exports, and export licences were required. No permission was required for imports (Heckscher, 1955, vol. II, p. 57). Soviet policy was also similar to mercantilism with respect to the role of the state. One writer describes mercantilism as ‘a group of ideas and practices … having as their purpose the building of strong political states and economic self-sufficiency (autarky) through the use of economic tools’ (Isaacs, 1948, p. 50). Mercantilism is not an unvarying mixture of policies, institutions, government ownership or regulation of production and personal life, but a varying combination of these, emphasizing national industrial and commercial prosperity. The balancing (mercantilist) mentality meant that foreign trade could not be left to chance. The balance mentality was similar to mercantilist thinking of the 1840s in which ‘balance of bargains systems’ were discussed whereby the import and export transactions of each merchant were to be balanced (Heckscher, 1955, vol. II, p.141). In order to complete certain balances, specific imports were necessary. What was to be imported had to be the decision of the central planners, not the decisions of individuals or enterprises. Exports were a necessary evil to obtain these needed imports. For these reasons, the administrative-command economy handled all its foreign transactions through a foreign trade monopoly. The foreign trade monopoly decided, on the basis of bilateral agreements, physical imports, physical exports, and the foreign exchange plan which would pay for trade deficits. Given the lack of supply of, and the lack of interest in, foreign credits, the Soviet economy was independent of world capital markets. Its own domestic capital market consisted of an administrative investment plan in which grants of capital were made from the state budget with no schedule of interest and amortization. Therefore, there was no domestic credit market, no market interest rates, and total independence of world credit markets. Domestic capital formation took place in the form of budget surpluses created by turnover taxes and enterprise profit taxes.

Paul R. Gregory and Joel Sailors 203

Retail prices contained a large differentiated turnover tax, which was the major source of state revenue. Money, prices, and exchange rates The major linkage, in this model, between the domestic Soviet economy and foreign markets was world prices. Transactions with the outside world were conducted in world market prices. The Soviet economy was not a large enough supplier of any product to affect its price, and the USSR was a relatively small buyer of products from the rest of the world, even for the machine tools noted below. Under the administrative-command system, exports, imports, and foreign exchange requirements were balanced by the administrative actions of a foreign trade monopoly, which managed all foreign transactions. Buyers and sellers were not permitted to engage directly in trade; the foreign trade monopoly conducted all buying and selling, making sure that these transactions were in accordance with the national economic plan. The value of imports was limited by the value of exports, in the absence of credits. The sole purpose of exports, therefore, was to earn foreign exchange to pay for imports. Individual buyers and sellers were prohibited from engaging in trade; trade was arranged through bilateral negotiations by state negotiators from the Ministry of Foreign Trade. These negotiations resulted, by and large, in barter deals which equated the values of exports and imports for each country. The Soviet economy’s scarce foreign exchange was also under the management of trade officials who used it to purchase goods in world markets according to material balance plans. These same officials received the scarce foreign exchange earned through the limited sale of Soviet export goods, such as grain, timber and metals. In this environment, flows of trade and capital formation were not affected by exchange rates, tariffs, and non-tariff barriers. Soviet export earnings were determined by world prices of grains, timber, and metals and by the physical quantities of exports that the plan permitted. All domestic transactions were denominated in domestic prices which were also set by planners. The allocation of capital was not affected by the cost of capital or by credit conditions because virtually all capital was distributed in the form of interest-free capital grants from the state budget. Companies that produced for export markets were required to sell their goods to the state foreign-trade monopoly at domestic prices. Any resulting profits achieved in world markets were deposited in the state budget. Soviet trade practices: 1928 to 1940 Under the Soviet development model, the foreign-trade monopoly would employ the Soviet Union’s limited foreign-exchange earnings to import advanced technology, which could be copied and serve as a base for the

204 The Soviet Union

industrialization drive. In particular, machine tools were to be an important instrument of industrialization. Exports had to be used to pay for these imports, irrespective of the terms of trade. Let us consider the actual results of Soviet trade activities during the period 1928 to 1940. As Figure 8.6 shows, Soviet exports and imports surged in real terms during the first three and four years of Soviet industrialization. After this initial surge, real exports and imports declined, real imports falling to 60 per cent of their 1928 level by 1937, and real exports returning to their original level. Current-price export and import figures show the Soviet Union’s declining terms of trade. By 1937, exports in current prices were 48 per cent of 1928, while exports in constant prices were 112 per cent of 1928; imports in current prices were 31 per cent of 1928 while imports in constant prices were 62 per cent of 1928. These figures imply a 15 per cent deterioration on the Soviet Union’s terms of trade during this period. Russia’s declining terms of trade meant that Soviet Russia had to expend increasing real quantities of exports (grain and minerals) in order to meet its planned import balances in support of industrialization. During the initial surge, the Soviet Union’s share of world exports and imports shot up from 1.3 per cent (imports) and 1.5 per cent (exports) to 2.6 per cent and 2.3 per cent, respectively. In fact, the USSR’s imports rose (in dollar terms) more than 20 per cent annually in 1930 and 1931, while other countries were experiencing enormous drops (League of Nations, 1933, pp. 214, 218). 250

200

150

Exports constant ps Exports current ps Imports constant ps Imports current ps

100

50

0 1928

1929

Figure 8.6 = 100)

1930

1931

1932 1933

1934

1935 1936

1937

USSR exports and imports, current and constant prices, 1928–37 (1928

Paul R. Gregory and Joel Sailors 205

The Soviet Union’s trade ratios fell from approximately 6 per cent of GDP in 1928 to less than one per cent in 1937, while remaining roughly steady during the export and import spurts of the early 1930s (Holzman, 1963, p. 290). It should be emphasized, however, that the USSR’s drop in real trade volumes was relatively mild compared with other countries’ and that the major factor behind the declining trade ratio was rapid growth of real GDP.6 Figures 8.7 and 8.8 show the strategic composition of Soviet imports and exports by year from 1928 to 1937. The Soviet Union concentrated its imports on machinery, which rose from one-quarter to well over 50 per cent of the total between 1928 and 1932, after which machinery imports fell back to the 1928 share. The burst of machinery imports highlights the Soviet strategy of importing machines to make machines, to lessen dependence on foreign technology. A lesser-known ingredient of Soviet import strategy was the substantial imports of metals to fuel the USSR’s industrialization drive. In 1913, Russia’s imports of metals were a relatively insignificant 8 per cent, while metal imports were more than 20 per cent from 1932 onwards. The rising shares of machinery and metals was offset by the declining import shares of consumer goods – textiles and tea. Figure 8.7 shows how the USSR paid for its strategic imports of machinery and metals. The first three years of imports for industrialization were primarily paid for by oil products, wood products, and grain. After the USSR had acquired the machinery it required for industrialization, it cut back on exports of those materials needed for industrialization, such as oil

30

25

20 oil products wood grain meat,milk sugar cloth

15

10

5

0

1928

Figure 8.7

1930

1932

1934

Composition of USSR exports, 1928–37 (per cent)

1937

206 The Soviet Union 60

50

40 machinery metals textiles tea

30

20

10

0

1928

Figure 8.8

1930

1932

1934

1937

Composition of USSR imports, 1928–37 (per cent)

products, and it shifted the focus to wood products, which were deemed less vital for industrialization. It should be noted that collectivization cost Russia its traditional exports of meat and milk products, after Soviet farmers slaughtered their herds during the first years of collectivization. Grain exports were used heavily in 1929 and 1930, but ensuing famines cut back on their use until the late 1930s. Capital formation Without access to world capital markets, and without interest in such access, Soviet planners had to raise capital formation rates through increased domestic saving. The plan for raising domestic saving rates was Preobrazhenky’s non-equivalent exchange. Collectivization was supposed to place the burden of capital formation on Soviet collective farms by way of compulsory deliveries at low procurement prices. How the USSR actually increased its investment and savings rates is subject to controversy. Several studies have shown that there was no net transfer of savings from agriculture to industry. The value of farm products delivered to the city equalled or even fell short of the value of products delivered by industry to agriculture, primarily mechanized farm equipment required to replace the loss of draft animals (Millar, 1974). The increase in the investment rate was a consequence of investment growing more rapidly than consumption. The increase in farm incomes was limited by the low delivery prices paid by the state to collective farmers. The increase in industrial wages was controlled by wage authorities.

Paul R. Gregory and Joel Sailors 207

4

Costs and benefits of Soviet autarky for the 1930s

During the 1950s, one of the strongest appeals of the Soviet administrativecommand economy was its performance during the Great Depression as well as its performance during the Second World War. This influence was felt in both the United States and in Europe in the growing appeal of communist and socialist ideas. In the United States, Roosevelt’s reaction to this growing communist threat was the New Deal. In Europe, further growth of the welfare state was the result. The Soviet Union appeared to be a country that had freed itself from the ‘anarchy of the market’ – an economy that had grown and transformed itself during a period of time when the other countries were in crisis. This perceived advantage caused Soviet scholars to be among the most serious empirical scholars of the business cycle, with a massive volume of scientific works published on this subject (e.g. Trakhtenberg, 1963). With the collapse of the Soviet Union and its administrative-command economy, we have considerable benefit of hindsight, which was obviously not available in the 1930s and 1940s. The ultimate collapse of the Soviet administrative-command economy has been attributed to its inferior performance relative to market economies, not to the fact that it could not have survived through a ‘muddling along’ approach. With this hindsight, we can now conclude that the benefits gained from avoidance of the Great Depression were of relatively short duration when compared with the longrun disadvantages of the economic system. The first three columns of Figure 8.9 bring home the tendency of economies to recover even from dramatic economic downturns, even though severe recessions and depressions may leave a lasting imprint on long-term growth rates. Figure 8.9 shows that, if we indiscriminately take the period 1928 to 1937 – the peak period of Soviet growth – we find that Soviet growth of real GDP was rivalled by Germany and the UK during this same period. It is only in the area of industrial production that Soviet economic performance overshadows that of its major European neighbours. The comparative wheat production figures show the lasting shock of collectivization on Soviet agricultural performance. The USSR was a country seeking to maximize its wheat production throughout the period whereas in other countries wheat production was restrained by poor market conditions. The fourth and fifth columns of Figure 8.9 show the impact of Soviet ‘superindustrialization’ on the USSR’s share of world manufacturing, which rose from 4 to 19 per cent of the world total between 1928 and 1937. The sixth and seventh columns of Figure 8.9 show the Soviet Union’s almost total withdrawal from world trade by 1937. The USSR began the period in 1928 with a 6 per cent trade ratio (exports plus imports divided by GDP) and ended in 1937 with a trade ratio equal to less than one per

208 The Soviet Union 300

250

200 Austria France Germany Italy UK USSR

150

100

50

0 gdp

industry

wheat

mfg share28

mfg share37

trade 28

trade 37

Figure 8.9 USSR and Europe: various indicators, 1928 and 1937 (for GDP, industry and wheat production: 1937 as a per cent of 1928; otherwise per cent share)

cent of GDP. In other countries trade ratios declined, but these declines were marginal when compared with the Soviet decline. The Soviet collapse in trade ratios, however, was unusual in the sense that it was as much the result of rapidly rising real GDP as of declining trade. The USSR’s decline in real exports, for example, was only 20 per cent more than the quantum index of world trade.7 The Soviet administrative-command economy was eventually judged not viable by its leaders primarily because of the long-term declining trend in its output and productivity growth. No one has attempted to determine the growth costs of the USSR’s autarky policies instituted in the 1930s. The experiences of the last three decades have shown the substantive positive effects of integration into world product and capital markets and the positive effects of international competition on productivity and technology. That the USSR was deprived of these effects for more than 60 years must have contributed to its declining growth rates. Study of the effects of the Great Depression on the economic performance of market economies is one way to speculate about the effects of persistent autarky on long-term Soviet economic performance. We all know the disastrous impact of the Great Depression on output and productivity in the major industrialized economies. Insofar as these effects were induced by the autarkic policies of the period, their effects can be extrapolated to the long-term performance of a country like the Soviet Union that deliberately chose autarky as a development model for a protracted period of time.

Paul R. Gregory and Joel Sailors 209

Between the 1880s and 1985, Russia had three quite different experiences with respect to world markets. The first period – 1880 to 1913 – saw Russia as a major player in world markets, as an exporter and importer of goods and services, as a capital importer, and as a legitimate member of the world economic community via Russia’s membership in the gold standard. Russia clearly benefited from this involvement: Russia’s growth would have been significantly less had it not been the recipient of so much international saving. The second period – the interwar years of 1918 to 1937 – saw the Soviet Union cut off from world product and capital markets and far removed from the unstable international monetary system of the interwar years. The Soviet Union’s lack of involvement in world capital markets was inevitable – the result of its default on tsarist period debt. Its isolation from product markets was to a major degree deliberate. The Soviet Union’s non-attachment to world markets was not particularly harmful during this period; it would not have received capital anyway, and international trade was shrinking, particularly in agriculture. It was during the third period – 1945 to 1985 – that the Soviet Union’s autarky and lack of participation in prevailing international monetary arrangement were most harmful. The Soviet Union missed out on the technology transfers, the capital transfers, the effects on competition, and the lowering of trade barriers that so impacted on the postwar economic history of the industrialized world. Whereas the Soviet Union’s ability to mobilize capital independently of international capital markets remained strong up to its very end, it ability to absorb and utilize capital became its principal weakness. Notes 1 2 3 4

The figures cited in Figures 8.1 and 8.2 are taken directly from Mitchell (1976). These findings are summarized in Gregory (1994). The Soviet industrialization debate has been best discussed in Erlich (1960). The Soviet Development Model was described by Kuznets, Gur Ofer, Holland Hunter, and a number of other specialists. For a brief description, see Gregory and Stuart (1995), chap. 4. 5 The negotiations that took place in Genoa and the Hague in 1922 to resolve the Russian debt indicated that Russia would be hard pressed to generate the surplus of exports over imports to amortize the debt. On this, see Pasvolsky and Moulton (1924). 6 The volume of real trade fell by about 13 per cent between 1928 and 1937, while real GDP increased by 163 per cent (in 1928 constant prices) or by 55 per cent (in 1939 constant prices). These figures are from Hunter (1992), p. 128 (citing data from Dohan), and from Bergson (1961), p. 196. 7 These figures are cited in Dohan and Hewett (1973), p. 24.

References Bergson, A. (1961) The Real National Income of Soviet Russia Since 1928 (Cambridge, MA.: Harvard University Press).

210 The Soviet Union Dohan, M. and E. Hewett (1973) Two Studies in Soviet Terms of Trade, 1918–1970 (Indiana University: International Development and Research Center). Erickson, R.E. (1990) ‘The Soviet Statistical Debate: Khanin vs. TsU’, in H. Rowen and C. Wolf (eds), The Impoverished Superpower: Perestroika and the Soviet Military Burden (San Francisco: Institute for Contemporary Studies): 63–92. Erlich, A. (1960) The Soviet Industrialization Debate, 1924–1928 (Cambridge, MA.: Harvard University Press). Gregory, P. (1982) Russian National Income 1885–1913 (New York: Cambridge University Press). ——– (1994) Before Command: An Economic History of Russia From Emancipation to the Five Year Plans (Princeton: Princeton University Press). ——– (1996) ‘Russia and Europe: Lessons from the Pre-Command Era’, in R. Tilly and P. Welfens (eds), European Economic Integration as a Challenge to Industry and Government (Berlin: Springer). ——– and R. Stuart (1995) Soviet and Post-Soviet Economic Structure and Performance, 5th edn (New York: HarperCollins). Heckscher, E.F. (1955) Mercantilism, transl. Mendel Shapiro, revised edition edited by E.F. Soderlund (London: Macmillan, 2 vols). Holzman, F. (1963) ‘Foreign Trade’, in A. Bergson and S. Kuznets (eds), Economic Trends in the Soviet Union (Cambridge, MA: Harvard University Press). Hunter, H. (1992) Faulty Foundations: Soviet Economic Policies 1928–1940 (Princeton: Princeton University Press). Isaacs, A. (1948) International Trade: Tariffs and Commercial Policy (Chicago: Irwin). Khanin, G. (1993) Sovetski ekonomicheski rost: Analiz zapadnykh otsenok (Novosibirsk: Eko). Kuznets, S. (1963) ‘A Comparative Appraisal’, in A. Bergson and S. Kuznets (eds), Economic Trends in the Soviet Union (Cambridge, MA.: Harvard University Press). League of Nations (1933) World Economic Survey (Geneva). Millar, J. (1974) ‘Mass Collectivization and the Contribution of Agriculture to the First Five Year Plan’, Slavic Review, 33 (December): 750–66. Mitchell, B.S. (1976) European Historical Statistics, 1750–1970 (New York: Columbia University Press). Pasvolsky, L. and H. Moulton (1924) Russian Debts and Russian Reconstruction (New York: McGraw Hill). Taussig, F. (1927) International Trade (New York: Macmillan). Trakhtenberg, I.A. (1963) Denezhnye krizisy (Moscow: Nauka).

9 ‘Afterword’: Counterfactual Histories of the Great Depression Barry Eichengreen and Peter Temin

As Dr Balderston notes in his introduction to this volume, history is necessarily written in terms of a model, whether implicitly or explicitly, and a model invariably suggests counterfactuals. In this note, we first review our model of the Great Depression (‘the ET model’ as it is referred to by Balderston) and then explore the counterfactuals that flow from its application to the monetary, macroeconomic and political history of the 1930s.

1

The Depression as it was

It is necessary to understand the causes of the Great Depression in order to answer the question of whether things could have turned out differently. The simultaneous fall in production and prices in the early 1930s strongly suggests that the initiating factor for the Great Depression was a series of negative aggregate demand shocks. But how could so many countries have experienced a negative demand shock at the same time? The answer is that all of these countries, faithful to the dictates of the gold standard, pursued deflationary policies at the same time. The essence of the gold standard was the free flow of gold between individuals and countries, the maintenance of fixed values of national currencies in terms of gold and therefore one another, and the absence of an international coordinating and lending organization like the International Monetary Fund.1 Under these conditions, when the United States and Germany adopted deflationary policies, other countries had little choice but to do likewise.2 As confidence in this system waned, central banks and governments swore their allegiance to the gold standard even more loudly. But their actions betrayed the words; fearing for the stability of the system they shifted their reserves out of US treasury bonds and British consols into gold.3 The international reserve backing for global money supplies collapsed between 1928 and 1932: the share of foreign exchange in reserves of 24 European countries fell from 42 per cent to 8 per cent.4 This was how deflation in two countries, which between them accounted for no more than a third of the world economy, 211

212 Afterword

could produce a vast deflationary shock that quickly engulfed the entire world economy. This recession began at the end of the 1920s in the United States and Germany. Their economies began to contract, partly as a result of centralbank pressure.5 But while their initial downturns had some independent roots, their economies were connected, allowing tightening by the Fed to make it harder for Germany to maintain its customary level of capital imports, which in turn forced the Reichsbank to tighten. In any case, it was clearly gold-standard policies that turned the downturn into the Great Depression and pulled down the rest of the world. The choice of deflation over devaluation was the most important factor determining the course of the Depression. Contemporaries clearly saw this as the key decision of the authorities, and they supported it wholeheartedly. Policy-makers in all industrial countries insisted that the way out of depression was not to ‘debase’ the currency but instead to cut wages, lower production costs, and reduce the prices of goods and services. Devaluation did not become a respectable option until much later – until after an unprecedented crisis had rendered the respectable unrespectable, and vice versa. For the time being, however, deflation remained the only accepted option. Governments and central banks could not easily deflate their economies in the early 1930s. Given the sacrifices they had made in the Great War, workers who once had mutely borne the burdens of financial stability now expected, indeed demanded, a voice in policy. The inability of economic policy-makers to force down wages was at the core of the period’s economic strains. As a result of the difficulty of forcing down wages, labour demand was depressed. Profitability was squeezed. Credit was tight (the real value of the global gold stock, which provided the backing for money supplies, was low because price levels were high).6 The political strains created by attempts to cut wages caused investors to fear for the stability of the gold standard even as policy-makers struggled to maintain it. And as those investors grew less confident about the depth of political support for and therefore the operation of the system, the destruction of international reserves quickly got under way, as central banks shifted out of foreign exchange in favour of gold. At its inception, the Great Depression was transmitted internationally as a result of the hegemony of the gold-standard ideology, a mentality that decreed that external economic relations were primary and that speculation, like that manifest in the booming stock market in New York, was a threat to economic stability.7 US, British and German policy-makers thus acquiesced in the deflation and liquidation that set in at the end of the 1920s. And as the US, British and German economies contracted, they depressed other economies through the mechanism of the gold standard. These countries reduced their imports as they contracted, reducing the

Barry Eichengreen and Peter Temin 213

exports of other countries. They also reduced their capital exports (in the case of the US and UK) in response to tightening credit conditions at the end of the 1920s. No country on the gold standard could escape the discipline of this harsh regime as the Depression progressed.8 Some found their prices falling as a result of the lack of demand for their products in export markets. Others compressed domestic demand, forcing prices to fall in order to maintain the value of the currency. In almost all cases, deflation was accompanied by depression as declining aggregate demand moved countries down along their aggregate supply curves. Banking systems collapsed under the weight of this deflationary pressure, further reducing money, credit and economic activity. As confidence in the gold standard weakened, the deflationary pressure intensified. When sterling’s devaluation raised questions in the minds of investors about the prospective stability of the dollar, the Federal Reserve felt compelled to raise its discount rate despite the fact that the US economy was contracting rapidly. The Fed was clearly not worried about inflation in a period when prices were collapsing. It was not worried about excessive speculation now that the Wall Street bubble had burst. Rather, it was worried about gold losses, prospective as well as actual, and their implications for confidence in the dollar. The gold standard, clearly, was a primary consideration shaping the policies that so compounded the US Depression. It follows that abandoning the gold standard was the only way of arresting the decline. Going off gold severed the connection between the balance of payments and the price level. It severed the connection between global gold stocks and global money supplies. Unless they took this critical step, countries could not reduce the level of interest rates or expand money and credit without precipitating a currency crisis. Changes in the exchange rate rather than changes in domestic prices could eliminate differences between the level of domestic and foreign demand without a painful deflation. Any single devaluation might beggar neighbours if it was not accompanied by a sharp expansion of domestic credit, but devaluation all round would have increased the value of world gold reserves and allowed worldwide monetary reflation and economic expansion.

2

The Depression as it might have been

What would have happened if countries had abandoned the gold standard sooner than they in fact did? Instead of a large literature on the Great Depression, might there be only a small literature on the 1929–30 recession and the collapse of the interwar gold standard, analogous to the literature on the late-1960s recession in the United States and the collapse of Bretton Woods?

214 Afterword

Britain abandoned the gold standard in 1931, avoiding some of the horrors of the slump. But this policy did not allow either it or the world to avoid the Depression. Britain, one of the most open of industrialized economies, could not insulate itself completely from the rest of the world, which was still in decline. Britain also was not large enough for its apostasy to change the direction of the world economy. It was not large enough to do so even if it had in fact used that apostasy as an opportunity to initiate aggressively reflationary policies, which it did not. With historical Britain as the base, we can ask what would have happened if other countries had also responded like Britain. Ferguson and Temin (2001) argue that the German currency crisis of 1931 was produced by political actions in the Weimar Republic, not by any iron laws of economics. If the German crisis was not economically inevitable – if it was not the result of the Austrian collapse or of changes in American capital markets – the situation there could have developed differently if politicians had made different choices. Let us assume that the Weimar government had been less bellicose in the spring and summer of 1931. Specifically, assume that the Weimar government had resisted the impulses to push the boundaries of the Versailles Treaty by calling for a customs union with Austria and denouncing reparations in March and June 1931. Foreign loans would have continued to fund the German government’s growing deficits, and the German banks would not have failed. This would have made life a little better in Germany. Even though conditions were very bad in Germany in 1931, they might not have continued to get worse as quickly as they in fact did if the banking system had not been in crisis. But if the German government continued to adhere to the gold standard, to maintain capital mobility and a fixed price of its currency, any improvement in economic conditions would have been minor.9 The course of German history would not have been very different because economic policy could not have been very different. The worst outcome, the currency crisis of 1931, might have been avoided, but the dictates of the gold standard would still have held sway. While output would not have fallen as fast or as far, it still would have fallen.10 German macroeconomic policy was deflationary in late 1931 even though Germany had slapped controls on currency transactions and could have used this shelter to dramatically stimulate its economy (without causing its currency to collapse). If these deflationary domestic policies had been held to in what we might term the first of our counterfactual worlds, then the German Depression would have been a bit milder, but it still clearly would have been the Great Depression. Other countries would have benefited only marginally, especially insofar as larger capital inflows into Germany, attracted by the government’s less bellicose foreign policy, would have meant larger capital outflows from other countries, such as the

Barry Eichengreen and Peter Temin 215

United States, possibly impelling the Fed to jack up interest rates and thereby worsen the US slump.11 Assume now a second counterfactual in which the Weimar government also changed its macroeconomic policy in 1931. Assume that Germany either had currency controls, as it did, or that it devalued with Britain, and in addition that Germany took advantage of the opportunity in the summer and autumn of 1931 to adopt much more stimulative macroeconomic policies.12 Almost certainly, economic conditions there would have begun to improve more quickly.13 Moreover, if there had been no currency crisis in the summer (recall that this second counterfactual builds on the first), German consumers and investors would not have been subject to worries about their banks and currency. Conditions in Germany would have been even better in this second counterfactual.14 What about the rest of the world? Eichengreen and Sachs (1986) showed that devaluations are harmful to other countries when the devaluing country fails to take the opportunity to expand its economy by pumping up domestic credit. Britain in 1931 fell into this category; not until the spring of 1932 did the Bank of England lower bank rate and begin to stimulate economic activity. At this point Britain’s devaluation stopped being beggar-thy-neighbour and was neutral or helpful toward the rest of the world. The vast majority of other countries behaved similarly, in that they used their new freedom to pursue more expansive policies only tentatively and after some delay. The Eichengreen and Sachs model applies to Germany as well. If Brüning had devalued but persisted in the deflationary policies he actually pursued, then the devaluation would have helped Germany while hurting other countries. If, however, Brüning had understood that devaluation allowed more expansionary policies and responded accordingly, then the German devaluation could have also eased conditions elsewhere in the world. The German cabinet discussed the possibility of following Britain off gold in September 1931, and the tone of their discussions suggests that they understood that devaluation could be a package of actions: devaluation itself and a change in macroeconomic policy to expand the economy. It was the latter component that frightened these veterans of the German hyperinflation. Their fear was that devaluation would be less of a positive influence on prices and demand than a negative shock to confidence (whatever changes in domestic policy accompanied the change in parity). It is thus reasonable to ask what might have happened had Germany devalued in 1931 (or taken seriously the effects of currency controls) and then (possibly after six months or so, as in Britain) begun to prudently and cautiously expand domestic credit, stabilizing prices and demand. Unless one believes that the negative shock to confidence would have dominated, something that was not historically the case in any other country that took this tack, the answer is that the depression in Germany

216 Afterword

would have been eased and that the depression elsewhere eventually would have been eased as well. A third counterfactual is quite different, both in the location of the change and its probable effect. Assume now that German history had followed its actual path to crisis and beyond in the summer of 1931. Let Britain also follow its actual path of policies in the fall of 1931. But assume that the US Federal Reserve did not follow its historical path when the United States suffered gold losses in September and October of 1931. As everyone knows, the Fed followed the dictates of the gold standard and raised its discount rate in two jumps a week apart, amounting to the largest rise in the discount rate in the Fed’s history. As everyone from Friedman and Schwartz (1963) to Romer (1993) has noted, this action was widely applauded by the US financial community. Of course it also was an extremely harsh blow to an already depressed economy. The alternative would have been to follow England off the gold standard in the autumn of 1931. Had the US floated the dollar, the currency likely would have sunk with the pound, leaving the dollar–pound exchange rate more or less unaffected. This would have been a shock to the world financial system. As we know from other crises and from open-economic macroeconomic theory, exchange rates tend to overshoot their eventual equilibrium in response to changes in the policy regime. Assume, as we think reasonable, that this overshooting would not have been worse than the pound’s overshooting, which was about 20 per cent, and that exchange rates then would have stabilized. In other words, we think there is no reason to hypothesize the kind of exchange-rate chaos that the defenders of the gold standard anticipated at the time.15 In this case, a simultaneous devaluation of the world’s largest currency area and of the world’s most extensive trading area would have provided much needed liquidity to the world, assuming that the Fed and the Bank of England had followed up by expanding domestic credit. Devaluation would have allowed the Federal Reserve to consider domestic conditions instead of international strains, as it would do two years later under Roosevelt’s guidance. Instead of taking an unprecedented deflationary action, policy would have turned expansive – even if only gradually, as in England. In this third counterfactual, the world after 1931 would not have been saddled by uniformly deflationary policies. Given that it typically took six months following the abandonment of the gold standard for central banks and governments to convince themselves that abandoning gold did not threaten inflation and to begin to cautiously reflate, and up to another year for monetary policy to begin to have effects on the real economy, 1932 would not have been a good year even under this counterfactual. Other countries would have mainly felt a loss of competitiveness. Only toward the end of the year would they have felt some relaxation of credit stringency and upward pressure on demand from

Barry Eichengreen and Peter Temin 217

the stimulus applied by the US and British central banks. The positive response might have materialized more quickly if households and firms had anticipated the favourable effects of the reorientation of US and British policies as soon as the latter were initiated. But most of the evidence is consistent with the view that time was needed for the change in policy to work.16 Our final fantasy is a combination of the second and third counterfactuals described above. In this fourth counterfactual, Britain, Germany and the United States all devalued in the summer or autumn of 1931 and gradually changed their policies. Instead of being devout champions of the gold standard, they would have turned their attention more to domestic economic conditions. Perhaps they might even have encouraged each other to move faster and sped up the process relative to that actually observed in the English turnaround. The French of course would have been hostile to all this expansion, predicting chaos and disaster from these new policies. It is particularly fantastic to think of the French devaluing in 1931; it took five years of economic decline and bitter political debate to work them around to this position. But if the Germans could have taken the chance of expanding, perhaps even these faithful guardians of gold-standard orthodoxy could have joined the crowd (Franco-German rivalry being what it is). Whether or not France joined in, many other countries in the US, British and German spheres of influence would have presumably changed their policies, following these three countries, in the present counterfactual. This final counterfactual thus effectively envisages a change in policy in the autumn of 1931 by almost the entire industrialized world. Recovery would have begun almost immediately as investors and consumers realized that the policy regime had changed. The bottom would have been reached in 1931, and conditions in 1932 would have begun to brighten. There still would have been widespread unemployment and idle resources, to be sure, but the mood would have been altogether different as economic activity showed signs of picking up. We might remember the early 1930s only as a major recession, like the recession of the early 1980s, instead of the Great Depression.

3

Conclusions

It is of course great fun for economic historians to speculate on ways in which the Depression might have been moderated and shortened by the adoption of different economic policies. But there also is good reason for the general historian to take these speculative counterfactuals seriously. The most compelling such reason of course is the momentous historical discontinuity that occurred when Hitler became German chancellor in January 1933. If conditions had been different enough to avoid the scourge

218 Afterword

of the Nazis and perhaps also the Second World War, the history of the twentieth century would have been profoundly different. Rivers of ink have been spilled over the causes of the Nazi rise to power. Studies have championed and refuted competing hypotheses about the relationship between the German economy and the votes for the Nazi party. At some level, however, there cannot be any doubt that the Nazis were the party of the Depression. They were a fringe group in the 1920s and grew to electoral prominence only in 1930 when economic conditions deteriorated. They gained even more seats in the Reichstag in the first election of 1932, but lost seats in the second election later that year as economic conditions appeared to improve. Had that improvement come earlier, a new study using panel data shows clearly that the Nazi vote would have been smaller.17 We do not have a model of the political process that tells us how weak the electoral support for the Nazis would have had to be to significantly affect the political manoeuvring among the leaders of the Weimar Republic. But almost any model would say that better economic conditions would have decreased political support for the Nazis and therefore the probability that Hindenburg would have asked Hitler to be chancellor. The point on which we are insisting is that economic-policy counterfactuals are more than a game for economists. Economic policy is an ingredient of politics. Had economic policy been different in the Great Depression, one can well imagine that the horrors of Nazism and the Second World War might have been avoided. The sequence of counterfactuals we have described show the probable effects of progressively larger changes in economic policies. Each successive counterfactual considers a more radical change in policies and predicts a shorter depression. It is hard to know how many elements in this sequence would have been required to stop the Nazis’ political gains. Unquestionably, the fourth and most ambitious counterfactual would have changed economic conditions sufficiently to foil the Nazi grab for power. It seems to us a minor question whether the second or third counterfactual would have been enough to do the job. The ET theory of the Great Depression, as Balderston labels it, has two essential attributes. It provides a unified account of the macroeconomic history of the interwar years (as the reader will not be surprised to hear us say). And because it is based on an explicit model, the ET theory can also do more than this. It can provide a basis for considering history as it might have been. It can help us uncover the political actions that affected its course. It has led us to argue that it was the poor choices of the economic policy-makers that exposed the citizens of the world to violence and even genocide for more than a decade. As we argued in Eichengreen and Temin (2000), the officials and politicians in question failed to appreciate this fact. They continued to believe in the wisdom of their misguided policies even as the world descended into

Barry Eichengreen and Peter Temin 219

depression and chaos. It was only when they were replaced that the economic policy regime changed. They thought after the fact that they had been run over by a steamroller of suffering people.18 Ultimately, the question, a definitive answer to which lies beyond the scope of this essay, is how economic policies and economic policy leadership come to be changed. The Nazi acquisition of power is one example of how this can come about, but in the least desirable possible way. Notes 1 The large and melancholy literature on central bank cooperation in the 1920s proves how poor a substitute this turned out to be. 2 The adjustment mechanism for a deficit country was deflation rather than devaluation – that is, a change in domestic prices instead of a change in the exchange rate. Lowering prices and possibly production as well would reduce imports and increase exports, improving the balance of trade and attracting gold or foreign exchange. This is the price-specie-flow mechanism first outlined by David Hume in 1752. 3 Forcing the reserve-centre countries, finding themselves to be losing gold, to further raise interest rates, applying more deflationary pressure to their economies. 4 The data and the argument are both presented by Nurkse (1944). 5 France, meanwhile, was not in recession (having stabilized much later than the others, in 1926, and was only beginning to recover from its post-stabilization recession in 1928–9), but it was accumulating large amounts of gold due to what was in effect a tight monetary policy. The impact on the rest of the world was thus much the same as that exercised by US and German policies. 6 The last is the mechanism emphasized by Nobel Laureate Robert Mundell in a series of articles (see Mundell, 2000) and by Johnson (1997) in his variation on our theme. 7 There are really two strands of the financial orthodoxy of the time: one emphasizing excessive speculation and the need to liquidate those financial excesses, and another emphasizing the need for policy to hoe firmly to the dictates of the gold standard. Our argument is that these two strands were intertwined; they are both aspects of what we called the ‘gold standard mentalité’ in Eichengreen and Temin (2000). 8 Hsieh and Romer (2001) question whether the United States was in fact subject to this discipline in the key year of 1932. We are inclined to argue that it was, at least psychologically if not also technically. 9 Brüning’s famous price-reducing decree of December 1931 would have depressed the economy, in other words, even if there had not been a crisis in July. 10 Bernanke and James (1991) use panel data for several dozen countries and relate the evolution of output to the extent of deflation, the presence or absence of a banking crisis, and additional control variables. They find that output suffered both when deflation was more rapid and when a country experienced a banking crisis. Under the present counterfactual, Germany would have experienced the first of these adverse shocks in 1931 but not the second. Hence our conclusion that output would have still fallen, but by less. 11 Thus, a different political stance in Germany, by itself, would have mainly redistributed the incidence of the Depression internationally. This underscores our

220 Afterword

12

13 14 15 16

17

18

previous point that more radical reorientations of policy would have been required to actually end it. We return to this ‘beggar-thy-neighbour’ aspect of the problem below. This is something of a heroic assumption. Germany had after all experienced one of the most extreme hyperinflations of the early 1920s. This made policymakers and the public recoil, almost instinctually, from arguments for aggressively reflationary monetary policies (even after strict exchange controls were applied to bottle up some of the consequences). This observation of course just serves to highlight our essential argument. The use of ‘almost’ means that there is one important qualification, as we explain below. The Bernanke and James (1991) model described above now predicts a much more favourable evolution of output. Their expectations are documented in Eichengreen and Temin (2000). The preceding sentence alludes to the regime-change argument of Temin and Wigmore (1990), which the authors use to explain the early recovery of industrial production in the US following Roosevelt’s abandonment of the gold standard. But the gold bloc countries did not experience a sudden rebound in industrial production in anticipation of any eventually favourable effects of US policy in 1933, which is our point here. Stögbauer (2001) combined the clear time-series evidence and the previously ambiguous cross-section evidence to show that unemployment was a major spur to the Nazi vote. ‘They still think it wicked that this steamroller came along’ (Edie, 1934, p. 227).

References Bernanke, B. and H. James (1991) ‘The Gold Standard, Deflation, and Financial Crisis in the Great Depression: An International Comparison’, in R.G. Hubbard (ed.), Financial Markets and Financial Crises (Chicago: Chicago University Press): 33–68. Edie, L.D. (1934) Dollars (New Haven: Yale University Press). Eichengreen, B. and J. Sachs (1986) ‘Competitive Devaluation and the Great Depression’, Economics Letters, 22: 67–71. Enchengreen, B. and P. Temin (2000) ‘The Gold Standard and the Great Depression’, Contemporary European History, 9: 183–207. Ferguson, T. and P. Temin (2001) ‘Made in Germany: The German Currency Crisis of 1931’ MS. On line at http://papers.ssrn.com/sol3/papers.cfm?abstract_id= 260993. Friedman, M. and A.J. Schwartz (1963) A Monetary History of the United States 1867–1960 (Princeton: Princeton University Press). Hsieh, C.-T. and C. Romer (2001) ‘Was the Federal Reserve Fettered? Devaluation Expectations in the 1932 Monetary Expansion’, NBER Working Paper No. 8113 (February). Johnson, H. Clark (1997) Gold, France and the Great Depression, 1919–1932 (New Haven: Yale University Press). Mundell, R. (2000) ‘A Consideration of the Twentieth Century’, American Economic Review, 90: 327–340. Nurkse, R. (1944) International Currency Experience (Geneva: League of Nations). Romer, C. (1993) ‘The Nation In Depression’, Journal of Economic Perspectives, 7: 19–40.

Barry Eichengreen and Peter Temin 221 Stögbauer, C. (2001) ‘The Radicalisation of the German Electorate: Swinging to the Right and the Left in the Twilight of the Weimar Republic’, European Review of Economic History, 5: 251–80. Temin, P. and B. Wigmore (1990) ‘The End of One Big Deflation’, Explorations in Economic History, 27: 483–502.

This page intentionally left blank

Index Aldcroft, Derek H. 155 Argentina and the gold standard 3, 22, 162 Ashwin, Bernard 181 Australia and the gold standard 3, 13 Banking system and the crisis see Financial crisis Bernanke, Ben S. 44–5, 219 Blackett, Sir Basil 152–3 Borchardt, Knut 106, 107, 119, 120–1, 123, 133–4 Boyer, R. 61 Brazil and the gold standard 3, 22 Bretton Woods System and the gold standard 100 Briand Plan 11 Brown, Williams Adams 145 Brüning, Heinrich 11, 124, 130, 132–4, 215, 219 Bukharin, N. 198 Canada and the gold standard 3, 30–2 and the banking system 44–7, 53 business cycles, dating of 32–3 gold stocks and the Great Depression 13 recovery and 47–9 and the sterling bloc 186–7 and the US slump 8, 33–6, 52 Capital flows, international, and the Great Depression 2, 3, 20, 30, 92, 123, 128, 130 Cassel, Gustav 21 Chamberlain, Neville 159 Coates, Gordon 177, 179, 181, 187 Condliffe, J.B. 186 Consumption and the Great Depression 19, 59, 61, 69, 82, 86, 95, 193, 201, 205–6 Copland, D.B. 180 Counterfactual histories of the Great Depression 11–13, 48–9, 132–4, 213–17

Cunliffe, Walter (Lord), Cunliffe Report 148, 164 Curtius, Julius 11 Davidson, A.C. 180 Dawes Plan see Germany, foreign policy Depression of 1920–3 see Gold standard Dollar reserve assets see USA Douglas, C.H. (Major) 178 Eichengreen, Barry (other than in association with Temin) 63–4, 186 Eichengreen–Temin theory of Great Depression 1–7, 12, 14–16, 32, 58, 92, 98, 161–2, 178, 185, 211–3 Employment effects of Great Depression 2 Exchange-rate depreciation (floating) and recovery 7, 8, 14–15, 44, 95–8, 179–85, 213 regime see Gold standard, monetary regime Ferguson, Thomas 11 Fetter, Frank D. 163 Financial crisis (1931), and the gold standard 4–6, 10–12, 21, 213–14 absence of banking crisis in Canada, France, UK 8, 44–6, 81–2, 86, 94 central bank functions and 4, 14, 47–8 First World War and the Great Depression 2 Fisher, Allan 181 Floating exchange rates, effects of 2, 17 Forbes, George 177, 179 Foreign borrowing see Capital flows, International and under individual countries Foreign policy see France; Germany Franc see France

223

224 Index France absence of banking crisis 81–2, 86 Bank of France policy and the gold standard 72–7, 80–1, 164 consumption in see Consumption economic policy, principles of 59, 70–1, 83 exchange rate (real) 59, 61–3, 68 explanations of entry into Great Depression 9, 61–8 fiscal policy and the gold standard in 8, 58, 63, 67–8, 70–2, 80, 83, 85 foreign policy and the Great Depression 4, 11, 130–1 foreign protectionism and French exports 61, 77, 84, 85 gold accumulation and the Great Depression 3, 9, 72–7 (see also Gold flows/reserves) interest rates in 60, 63–5, 68, 74, 77–81, 84 investment (real) in 63, 68–9 macroeconomic regime in 60, 68–9, 77, 83–4 money supply regime 9, 58, 72–7 new issuing of securities 63, 68–9, 74 profit rate in 64, 69 real exchange rate 59, 61, 62, 68 real supply in 60, 63, 67–8 stabilization (1926) 59, 63–4, 70, 76 trade balance of, with abroad 63, 67–8 war debt arrangements 129 Friedman, Milton, and Schwartz, Anna J., theory of Great Depression 1, 45, 47 Germany deparliamentarization of, and elections 124, 130–1, 218 counterfactual effects of fiscal reflation 132–4 counterfactual effects of float after 1931 11, 12, 214–17 exchange rate (real) 124–6 export performance of 125–6 financial crisis (1931) in, and effects of 4–5, 10ff, 21, 131, 214 fiscal policy/deflation and its effects 10, 107–11, 129–34

foreign borrowing and capital balance, determinants of 10–11, 21, 123–31 foreign debt and autarky policies in the 1930s 134–5 foreign policy, reparations policy and arrangements, and the Great Depression 4, 10–12, 106–7,123–31 gold reserves, stocks in the 1930s 13, 23 gold standard policy, reasons and effects 2, 7, 17, 116–19, 215 growth in international comparison 207 investment, determination of real, in 106, 115–16 Keynesian theory and 111–14,135–6 origins of Depression 212 recession of 1925–6 107, 124 reparations and see Germany, foreign policy sovereign debt theory see Germany, foreign policy wage–cost constraints 106, 120–3, 136 Glass–Steagal Act of February 1932 6, 70 ‘Gold bloc’ of the 1930s 5–6 Gold exchange standard 14 Gold flows, reserves/stocks and the Great Depression 13–17, 20, 72–7 Gold reserve requirements 3, 15 Gold production 14 Gold standard as a ‘commitment mechanism’ 119 as a ‘contingent rule’ 5 counterfactuals to see Counterfactual histories credibility of 5, 98, 119–20, 211–12 ‘deflationary bias’ of 2, 9, 17, 58, 70, 72, 90–2, 212–13 and demand for gold/foreign reserves 14, 17, 29, 211–12 and the demand for money 89 depression of 1920–3 and 5 domestic political conflict and 5, 7, 14, 212 essence of 211 and the financial crisis see Financial crisis

Index 225 and financial integration 30–2 as an ideology/intellectual construct 7, 18, 19, 28, 42, 50, 161, 162–5, 172–3, 212–13, 218ff international cooperation/conflict and international policy coordination under 4, 18, 44, 98, 161–2 and international transmission of disturbances 1, 8, 10, 48, 32–6, 89, 92, 212–13 monetary policy, money supply and 9, 10, 21, 22, 28–9, 49, 58–9, 72–3 (see also under individual countries) as a monetary regime 7, 47–8 relation to the ‘quantity theory of money’ 29 and shortage of gold 14–17, 22, 212 ‘Golden fetters’ and monetary policy 6, 18, 48, 106, 116 Goldsmith, Raymond W. 144 Great Depression economic historians and 49–50 origins of 1–6, 8, 27, 32–44, 49–50, 92–3, 212 and reparations see Germany transmission of see Gold standard Greece and the gold standard 5 Hawke, Gary 183, 186 Hawtrey, Sir Ralph 153, 169 Heston, A. 144–5, 151 Hitler, Adolf 11, 217–18 Hugenberg, Alfred 11 Hume, David 52 India determinants of gold demand of 146–8, 157 effects of the Great Depression on 143–6, 156, 158–9 First World War and the silver crisis 148–51 gold hoarding, dishoarding and deflation 13, 151, 154–6, 158–61 Hilton–Young Commission (1926) 153 rupee, postwar overvaluation 151–4 rupee exchange-rate policy: British control of 145, 150–4, 157, 159–61, 165

and the world economy; contemporary models 146–7, 157 Interest rates and the Depression see USA Irwin, Douglas I. 186 James, Harold 219 Johnson. H. Clark 155, 219 Keynes, John Maynard 42, 146–8, 150, 163, 165 Kindleberger, Charles P. 44 Kuznets, Simon 192 Lefeaux, Leslie

186

Maastricht Treaty, implications for 85 Maddison, Angus 145 Maier, Charles S. 163 Marseille, J. 60 Müller, Hermann 124 Mundell, Robert A. 219

84,

Nash, Walter 186 National Bureau of Economic Research 32 Nazi movement, Party 11, 131, 218 Netherlands, The, and the gold standard 5, 10, 20 New Zealand banks and economic policy 18, 173–4, 179–80, 182–3 devaluation and 18, 172, 174, 179–85 British empire preference and recovery of 184–5 economic diagnoses and remedies 172, 175–9, 185–7 and imperial economic cooperation 185–8 monetary system, monetary policy of 18 transmission of Depression and recovery to 174–5, 184 Niemeyer, Sir Otto 186 Norman, Montagu (Lord) 153, 157, 165 Paraguay and the gold standard 3 Park, A.D. 181 Phillips, Frederick 160 Poincaré, Raymond 60, 63, 67, 70–1

226 Index Portugal and the gold standard 5 Preobrazhensky, E.O. 199–201, 206 Price expectations in the early Great Depression 37–40, 42, 44 Price rigidity and the Great Depression 2, 89–90 ‘Price-specie-flow mechanism’ see Gold standard, international transmission of disturbances Primary producing countries, capital balances of 2, 3 Reparations and the Great Depression see Germany Reserves, international see Gold reserves Roosevelt, F.D. 162, 207 Rupee see India Russia see USSR Sachs, Jeffrey 215 Sauvy, A 58, 61, 85 Schacht, Hjalmar H.G. 106, 124, 128, 130 Sivasubramonian, S. 144–5, 151 Social Credit 178–9 Soviet Union see USSR Stalin, Josef V. 198–9 Sterling bloc 18, 186 Stewart, Downie 181 Stresemann, Gustav 105 Strong, Benjamin 129 Sweden and the gold standard 5 ‘Tardieu programme’ 70 Temin, Peter (other than in association with Eichengreen) 1–2, 7, 27, 32, 41, 50, 69, 101, 115, 136 Tripartite Agreement (1936) 7, 48 Trotsky, Leon 198 Unemployment and the Great Depression 2, 52, 84 UK absence of banking crisis in 94, 100–1 and the ‘deflationary bias’ of the gold standard 90–1, 100 effects of recovery policy, including sterling float 95–8, 101, 214–15 fiscal policy in 97–8 gold standard policy of 5, 9, 90–1, 100, 162–5

gold accumulation by 13 (see also India, rupee exchange-rate policy) growth in international comparison 96, 207 and imperial monetary cooperation 161–2, 185–7 national specificities of Depression in 89, 93–6, 100 and the real exchange rate 90, 99 transmission of Great Depression to 88, 92–3 wages (real) 90, 95, 120–1 Uruguay and the gold standard 3 USA, the gold standard and monetary policy 1–6, 8, 12, 20, 42, 47, 212–13, 216 business cycles, dating of 32, 52 counterfactual effects of 1931 devaluation 216–17 deposit insurance and recovery 48 dollar devaluation 1933–4 and effect on policy 6 dollar reserve assets 5, 6, 13, 14, foreign lending and 2, 3 gold stocks of 13, 15, 16, 20 interest rates during Great Depression 42, 44 ‘New Deal’ in international perspective 197, 207 USSR (Russia) and the Great Depression 18, 19, 191 autarky and world-economy relations 196–9, 201–6 industrialization strategy and characteristics 198–201 interwar growth in comparative perspective 192–3, 207–9 prewar growth of under gold standard 194–6, 209 Soviet and Nazi economic policies 197 Wages, rigidity of 53, 89–90, 212 Wall Street boom and crash and the Great Depression 2, 10, 20, 36 Ward, Sir Joseph 177 World Economic Conference, 1933 48 Wyplosz, Charles 63–4 Young Plan policy

see Germany, foreign