In their transition from the legacy of Communism, Central and Eastern European countries (CEECs) are seeking to reduce ...
16 downloads
532 Views
14MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
In their transition from the legacy of Communism, Central and Eastern European countries (CEECs) are seeking to reduce the income gap that remains the major barrier to full European integration. The essays in this volume derive from a conference held at the American Institute for Contemporary German Studies in Washington, D.C., on May 15-16, 1995, and present general equilibrium calculations of the worldwide effects of trade liberalization between CEECs and the European Union (EU) on real wages and welfare; analysis of trade in "sensitive" sectors; and measurement of Germany's role in the transition. Simulations analyze the effects of CEEC macroeconomic policies on the transition process. Other essays examine the effects of privatization, labor migration from the East, and alternative approaches to integration of CEECs into the EU, including quick entry, variable geometry, and free-trade area. Economists and policymakers will value the collection's innovative quantitative assessments and presentation of distinct alternatives.
Europe's economy looks east
Papers from a conference sponsored by The American Institute for Contemporary German Studies The Johns Hopkins University Washington, B.C.
Europe's economy looks east Implications for Germany and the European Union
Edited by STANLEY W. BLACK University of North Carolina, Chapel Hill, and The American Institute for Contemporary German Studies, The Johns Hopkins University
CAMBRIDGE UNIVERSITY PRESS
CAMBRIDGE UNIVERSITY PRESS Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, Sao Paulo, Delhi Cambridge University Press The Edinburgh Building, Cambridge CB2 8RU, UK Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9780521572422 © Cambridge University Press 1997 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 1997 This digitally printed version 2008 A catalogue record for this publication is available from the British Library Library of Congress Cataloguing in Publication data Europe's economy looks east: implications for Germany and the European Union / edited by Stanley W. Black, p. cm. Includes index. ISBN 0-521-57242-8 1. European Union countries — Foreign economic relations — Europe, Eastern. 2. Europe, Eastern - Foreign economic relations - European Union countries. 3. Germany - Commerce - Europe, Eastern. 4. Europe, Eastern - Commerce - Germany. 5. Europe, Eastern Economic policy - 19896. Post-communism - Europe, Eastern. I. Black, Stanley W. HF3498.E852E97 1997 337.4043 - dc20 96-30247 CIP ISBN 978-0-521-57242-2 hardback ISBN 978-0-521-08823-7 paperback
Contents
List of figures and tables Preface List of conference participants List of acronyms 1 Introduction Stanley W. Black 1.1 Trade relations 1.2 Investment patterns 1.3 Labor market issues 1.4 The process of integration 1.5 Conclusions
page viii xiii xv xvi 1 2 9 12 14 17
1 Trade relations 2
An economic assessment of the integration of Czechoslovakia, Hungary, and Poland into the European Union Drusilla Brown, Alan Deardorff, Simeon Djankov, and Robert Stern 2.1 Introduction 2.2 The gains from integration, broadly conceived 2.3 Literature review of EU-CEC integration 2.4 The Stylized Europe Agreement Package (SEAP) 2.5 CEC-CEC integration 2.6 The Michigan CGE trade model 2.7 Computational results: aggregate effects 2.8 Computational results: sectoral effects 2.9 The roles of scale and variety 2.10 Conclusions and implications for research and policy Comments J. David Richardson L. Alan Winters
23 23 24 25 26 28 29 41 48 53 56 61 63
vi
Contents 3
Potential trade with core and periphery: industry differences in trade patterns Hari Vittas and Paolo Mauro 3.1 Introduction 3.2 Aggregate trade patterns 3.3 Sectoral trade patterns 3.4 Summary and conclusions Appendix A Appendix B
67 67 69 76 89 89 91
Comment Susan M. Collins
97
Impact on German trade of increased division of labor with Eastern Europe Dieter Schumacher 4.1 Introduction 4.2 Germany's trade with Central and Eastern Europe 4.3 Determinants of bilateral trade flows 4.4 Regression results 4.5 Estimates for Germany's trade with CEECs 4.6 Sectoral pattern of trade and structural changes in Germany 4.7 Summary and policy conclusions Comments Wolfgang Maennig Ellen Meade
100 100 101 107 119 135 150 153 157 162
I
Investment patterns
5
Investment and its financing during the transition in Central and Eastern Europe Stanley W. Black and Mathias Moersch 5.1 Introduction 5.2 Recent studies 5.3 Investment financing 5.4 The four stages of transition 5.5 The transition process 5.6 Conclusions Appendix A: Determination of capital and output Appendix B: Two-sector model Appendix C: Welfare analysis
167 167 170 174 187 192 196 197 198 198
Comments Douglas Todd Holger Wolf
201 207
Contents
6
Privatization, structural change, and productivity: toward convergence in Europe?
PaulJ.J. Welfens 6.1 6.2 6.3 6.4 6.5 6.6
Introduction Transformation and supply-side changes Experience with privatization Toward new economic structures Convergence issues Summary and conclusions
Comments Bruce Kogut Jan Mlddek III Labor market issues 7 Integrating the East: the labor market effects of immigration Thomas Bauer and Klaus F. Zimmermann 7.1 Introduction 7.2 Economic situation and migration potential 7.3 EU migration policies and options for the future 7.4 Theoretical framework 7.5 Calculating the gains from immigration 7.6 Conclusions Appendix A Appendix B Comments Barry Bosworth Robert LaLonde IV The process of integration 8 Joining the club: options for integrating Central and Eastern European Countries into the European Union Michael Koop 8.1 Transformation and integration 8.2 The state of European integration 8.3 Full EU membership 8.4 Partial EU membership 8.5 EEA membership 8.6 (C)EFTA option 8.7 Conclusions Comments Barry Eichengreen Hans-Jurgen Vosgerau Index
212 212 214 229 239 246 254 258 263
269 269 270 279 285 290 300 301 303 307 310
315 315 315 319 334 336 338 339 342 346 351
Figures and tables
Figures 3.1 3.2 3.3 3.4 3.5 3.6 4.1 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 5.10 6.1 6.2 6.3 7.1 7.2
Evolution of trade flows of Germany, France, Italy, UK, and Portugal with CEEC-6 Ratio of sensitive to total trade of Germany, France, Italy, UK, and Portugal with CEEC-6 countries Indices of revealed comparative advantage (adjusted for trade imbalances) Trends in employment in the manufacturing sector, Germany Trends in employment in the manufacturing sector, France Trends in employment in the manufacturing sector, Portugal Correlation of per capita income and human capital The four stages of transition Output choices Real loan rates CEEC savings rates Credit supply Real exchange rates Output per person in CEEC and EU Convergence of CEEC to EU productivity 1992 baseline Stylized view of the stages of transition Index of labor productivity in Eastern Europe Network effects of privatization and optimal privatization Paradoxes of capital intensity and structural change in an open economy Inflow of ethnic Germans to West Germany, 1950-93 Countries incorporated in the Schengen accords viii
page 73 77 82 93 94 95 146 169 170 176 177 178 179 194 195 202 203 219 229 251 278 280
Figures and tables 7.3 7.4 7.5 7.6 7.7 7.8
Theoretical framework Calculation of the gains from immigration Total immigration gains according to production factors at 10 percent inflow Immigration gains of natives and immigrant skills Immigration gains of natives according to production factors at 10 percent inflow The immigrant surplus
ix 289 291 292 293 296 308
Tables 2.1 2.2 2.3 2.4a 2.4b 2.5
2.6
2.7
2.8
2.9
3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8
Czechoslovakia: basic data, 1992 Hungary: basic data, 1992 Poland: basic data, 1992 Average tariff rates for CECs, European Union, NAFTA, and other trading nations, 1992 Average tariff equivalents of NTBs for CECs and European Union, 1992 Summary results of CEC and CEC-EU integration: changes in country imports, exports, terms of trade, welfare, and real returns to labor and capital Scenario C: CEC-EU free-trade area, tariffs and nonsensitive NTBs, sectoral effects on Czechoslovakia of CEC-EU integration Scenario C: CEC-EU free-trade area, tariffs and nonsensitive NTBs, sectoral effects on Hungary of CEC-EU integration Scenario C: CEC-EU free-trade area, tariffs and nonsensitive NTBs, sectoral effects on Poland of CEC-EU integration Summary results of CEC-EU integration in the base case: CEC-EU FTA, tariffs and nonsensitive NTBs removed, decomposition of scale and variety effects Trade with the CEEC-6 Relative importance of trade with the CEEC-6 in total trade Actual and potential trade with the CEEC-6 Composition of total imports from the CEEC-6 Imports from the CEEC-6 (average annual rate of change from 1989 to 1993) Estimated parameters of sectoral gravity equation Actual and potential imports from CEEC-6 by sector in 1993 Potential imports from the CEEC-6 under different scenarios
32 34 36 40 40
44
49
50
51
54 70 72 75 78 80 85 87 88
i
Figures and tables
3.9 4.1 4.2 4.3 4.4 4.5 4.6 4.7
4.8 4.9 4.10
4.11 4.12 4.13 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 5.10 5.11 5.12 5.13 5.14 5.15 5.16 6.1 6.2
Direct investment to CEEC-6 Trade of Germany with CEECs, 1989-94 Trade of West Germany and East Germany with CEECs, 1990-94 Commodity structure of German trade with CEECs, 1993 Commodity structure of German imports from five CEECs, 1993 Regression results for trade of 22 OECD countries with 70 countries (all variables) Regression results for trade of 22 OECD countries with 70 countries (five variables) Ranking of three-digit ISIC industries according to the value of the regression coefficients. Trade of 22 OECD countries with 70 countries (five variables) Regression results for trade of West Germany with 69 countries (all variables) Regression results for trade of West Germany with 69 countries (three variables) Ranking of three-digit ISIC industries according to the value of the regression coefficients. Trade of West Germany with 69 countries (three variables) Basic data of CEECs Estimated values of German exports and imports in trade with CEECs Factor content, exports relative to imports Investment requirements based on source-based approach Investment requirements based on needs-based approach Saving and investment Foreign direct investment in the CEECs Ratio of FDI to GDP in the CEECs FDI in selected EU countries Sources of FDI, selected countries Borrowing Macroeconomic indicators Real GDP CEEC capital stocks by country, 1992 baseline CEEC output with full employment CEEC output with full efficiency and full employment Alternative CEEC growth scenarios, 1992-2002 Productivity growth rates, 1992-2002 Capital/labor substitution in the EU, 1970-90 Selected economic indicators for Eastern Europe Change of labor productivity in industry and economic growth
91 102 104 108 112 120 124
126 136 138
140 144 148 152 172 173 175 180 182 182 183 184 185 187 188 190 191 193 194 204 215 218
Figures and tables 6.3 6.4 6.5 6.6 6.7 6.8 7.1 7.2 7.3 7.4 7.5 7.6 7.7 7.8 8.1 8.2
Share of private sector in GDP and employment, 1990-94 Excess employment in transforming countries, 1990-94 Degree of specialization and intensity of structural change Share of major sectors in Eastern Europe's output, 1990-94 Structure of employment by sectors, 1989-93 Sectoral structure in the Visegrad countries Economic indicators of Western, Central, and Eastern Europe in 1993 Demographic indicators Stock of foreign population in 1992 Asylum seekers in Western European countries Short-term employed workers from Eastern Europe in Germany Gains of immigration: full employment model Gains of immigration: full employment at 10 percent inflow Gains from immigration and the disequilibrium model Main economic indicators for selected CEEC and EU countries Budgetary effects of CEEC admission
xi 231 236 241 242 245 246 272 274 277 282 286 290 294 298 317 324
Preface
Economic relations between the European Union (EU) and the countries of Central and Eastern Europe (CEECs) underpin hopes for stability on the continent after the Cold War. The EU plays a central role in Europe's emerging architecture. In May 1995 the European Commission presented its White Paper, Preparation of the Associated Countries of Central and Eastern Europe for Integration into the Internal Market of the Union. The document specified the steps the CEECs were required to take to adapt to the acquis communautaire, focusing on twenty-three separate areas of legislative activity from free movements of goods, services, capital, and labor, to social policy and agriculture. Germany plays a central role in establishing the economic ligature between East and West by virtue of its location, economic magnitude, and experience with unification. Consensus reigns across all parties represented in the German Bundestag that CEEC membership is a vital German interest. In an effort to contribute to a better understanding of the forces at work between the two halves of Europe, the Economic Studies Program of the American Institute for Contemporary German Studies commissioned a set of papers that were discussed at the Institute conference "Europe's Economy Looks East: Implications for Germany and the EU," held in May 1995. This volume presents the papers, together with an introduction by Stanley W. Black, the Director of the Economic Studies Program. The American Institute for Contemporary German Studies at The Johns Hopkins University is committed to advanced research, study, and discussion of the Federal Republic of Germany. The Institute established its Economic Studies Program in 1994 to examine economic policy issues of central importance in Germany and for Germany's future economic development. The Institute's program provides a unique forum for research and analysis to be conducted and shared with policymakers, corporate leaders, and scholars who are engaged in the study of contemporary German affairs. This volume makes a valuable contribution to both research and dialogue among these groups on both sides of the Atlantic. The Institute would like to express its deep appreciation to Mr. Diethelm xiii
xiv
Preface
Hoener for his generous support of the Conference and the Economic Studies Program. We would also like to express our thanks to Lufthansa German Airlines for partial support of the conference. Carl Lankowski Research Director, AICGS Jackson Janes Executive Director, AICGS September 1996
Conference participants
Martin Baily President's Council of Economic Advisors Thomas Bauer Universitdt Miinchen Tamim Bayoumi International Monetary Fund Stanley W. Black American Institute for Contemporary German Studies Barry Bosworth Brookings Institution Susan M. Collins Georgetown University Jonathan Davidson Delegation of the European Commission Simeon Djankov University of Michigan Barry Eichengreen University of California at Berkeley Geza Feketekuty Office of the U.S. Trade Representative Bruce Kogut Wharton School, University of Pennsylvania Michael Koop Institutfiir Weltwirtschaft Robert LaLonde Michigan State University Wolfgang Maennig Universitdt Hamburg Paolo Mauro International Monetary Fund Ellen Meade Board of Governors of the Federal Reserve System Jan Mladek Czech Institute ofApplied Economics, Ltd. Mathias Moersch American Institute for Contemporary German Studies Steven Muller The Johns Hopkins University Wolfgang Reinicke Brookings Institution J. David Richardson Institute for International Economics Dieter Schumacher Deutsches Institutfiir Wirtschaftsforschung Robert Stern University of Michigan Douglas Todd Commission of the European Union Hari Vittas International Monetary Fund Hans-Jiirgen Vosgerau Universitdt Konstanz Paul J. J. Welfens Universitdt Potsdam L. Alan Winters World Bank Holger Wolf New York University
xv
Acronyms
CAP CEC CEECs CEEC-6 CEFTA CMEA CSFR CUSTA ECU EEA EIB EU EU-EFTA EU-North EU-South EBRD FDI FTA IGC NAFTA NTB s OECD PCI RCA ROW SEAP SOEs
xvi
Common Agricultural Policy Central European Country Central and Eastern European Countries Czech Republic, Hungary, Poland, Slovakia, Bulgaria, Romania Central European Free Trade Area Council for Mutual Economic Assistance Czech and Slovak Federal Republic Canadian-United States Trade Area European Currency Unit (^1.28 US $) European Economic Area European Investment Bank European Union Austria, Finland, Sweden Belgium, Netherlands, Luxembourg, France, Germany, Italy, Ireland, Denmark, United Kingdom Greece, Spain, Portugal European Bank for Reconstruction and Development foreign direct investment free-trade area Inter-Governmental Conference North American Free Trade Agreement nontariff barriers Organization for Economic Cooperation and Development per capita income revealed comparative advantage rest of world Stylized European Agreement Package State-Owned Enterprises
CHAPTER 1
Introduction Stanley W. Black
The end of the Cold War has had profound implications for the economic and political life of Europe. The Central and Eastern European Countries (CEECs, usually defined as Bulgaria, the Czech Republic, Hungary, Poland, Romania, and Slovakia) emerged from the legacy of forty-five years of Communism with a variety of economic pathologies ranging from misguided price incentives, distorted economic structures, mispriced resources, bankrupt state enterprises and financial institutions to inadequate legal systems and distrusted political institutions. The transition process is expected to be long and difficult. But it represents the most creative opportunity for raising standards of living in the industrialized world since the recovery from World War II. If grasped effectively, this opportunity has every prospect of generating great welfare gains, including substantial spillover benefits to the European Union (EU), not, however, without potential adjustment costs, in the form of increased competition in some EU industries and pressure on transfer programs including the Common Agricultural Policy (CAP). The primary beneficiaries of the transition process should be the CEECs themselves, although many of their residents have become impatient waiting for the benefits to arrive. The papers in this volume attempt to spell out the costs and benefits of many of the changes that CEECs will have to undergo to reap the gains. For members of the European Union, the benefits include building market economies and functioning democracies as neighbors and the chance to put an end to the East-West division of Europe. These developments reduce the security problems of Europe dramatically, as the CEECs have shifted from being members of an opposing alliance to applicants for membership in the EU and NATO. In the economic field, significant new trade, investment, and migration patterns are emerging. I would like to acknowledge the significant assistance of Mathias Moersch in planning and organizing the conference as well as completing the editorial work on this volume. Felecia Lucht was of great assistance in organizing the conference and Margaret Runyan-Shefa in preparing the manuscript and index for publication.
1
2
Stanley W. Black
Americans should be interested in this process for several reasons. While our direct trade relationships with Central and Eastern Europe are and will likely remain relatively small, we have important and growing foreign investments there. In addition there are significant ties of blood for many U.S. immigrants from the area. Perhaps most fundamentally, the United States paid a heavy price through both World War II and the Cold War for the liberation of Central and Eastern Europe from two different tyrannical regimes. Surely, we have an important interest in the peaceful evolution of the region into democracy and prosperity under a free market system compatible with the Western democracies. Nevertheless, the achievement of this objective will be very difficult. Incomes in the CEECs have fallen far behind EU levels as a result of the failures and collapse of the Communist system. The income gap is currently estimated at about 75 percent of EU average incomes and represents a formidable barrier to the integration of East and West, particularly if integration takes the form of full membership in the EU. The four freedoms of the EU entail free trade in goods and services between the high-wage EU and low-wage CEECs, as well as free movement of both capital and labor. To what extent will extension of the four freedoms create problems for workers and firms in the EU? Is there danger of a "giant sucking sound" of jobs moving east or a parallel flood of workers moving west? Will cheap labor combined with free trade overwhelm the declining industries of the West? The Europe Agreements already limit entry of CEEC products in "sensitive" industries. And EU redistribution programs such as the CAP and the Regional and Structural Funds would require vastly increased expenditures at current income levels in the CEECs. Quantifying some of these costs and benefits and examining the policies appropriate for the transition process is the goal of this study, which includes seven papers and thirteen comments by economists from Europe and the United States presented at a conference at the American Institute for Contemporary German Studies in Washington, D.C., on May 15-16, 1995. The papers are divided into four groups: trade relations, investment patterns, labor market issues, and the process of integration. They provide a clear picture of many of the key issues that must be addressed in each of these areas if the integration process is to move forward. 1.1
Trade relations
Trade relations between CEECs and the EU have been studied previously in a variety of different ways. Historical data from the interwar period have been used by Collins and Rodrik (1991) to predict future trade flows. Gravity models have been used by Baldwin (1994) to predict likely aggregate trade flows between countries at the same distance and relative income levels as the CEECs and EU. Partial equilibrium models of specific sectors by Winters and Wang (1994) have examined the effects of trade integration on CEECs and their EU
Introduction
3
partners. Effects on agricultural trade have been studied by Rollo and Smith (1993). The impact of the Europe Agreements on particular countries such as Greece, Spain, Portugal, and France have also been examined by various authors. What has not been done is to examine the general equilibrium effects of trade liberalization between the CEECs and the EU, allowing for changes in real wages and real exchange rates as labor markets and goods markets adjust to the changes in trade flows. And while the gravity models have been helpful in judging the possible levels of aggregate trade flows, they have not previously addressed the more important sectoral issues, where the possible displacement of workers is likely to be much larger than at the aggregate level, where gains and losses are netted against each other. It may be helpful to consider the distinction between general equilibrium and sectoral gravity models. The general equilibrium approach derives sectoral demand and supply functions for goods and services from utility and profit maximization in each country, subject to trade policy measures such as tariff and nontariff barriers. Changes in trade policies then generate changes in sectoral exports and imports, subject to the level of resources available and assuming that the exchange rate changes to maintain the balance of trade. By contrast, the gravity model may be thought of as the reduced form of a general equilibrium model, showing net sectoral trade patterns between pairs of countries as a function of each country's size, per capita income, and the distance between them. This has been rigorously demonstrated in a two industries by two factors context by Bergstrand (1989), assuming monopolistic competition with differentiated products and taking account of factor endowments and transportation costs. Bergstrand generalizes the 2 X 2 Rybczynski theorem to show that in a multi-industry world an increase in a country's endowment of capital (labor) tends to increase the output of relatively capital-intensive (labor-intensive) industries. In the gravity model, exporter's per capita income is a proxy for the capital-labor ratio of the exporter, while importer's per capita income influences the pattern of demand and distance is a proxy for transportation cost. Other more specific factors such as natural resources are, however, omitted from the model. Three papers in this volume examine trade issues directly. The paper by Drusilla Brown, Alan Deardorff, Serge Djankov, and Robert Stern is the first to examine CEEC trade issues using a general equilibrium world trade model. The Michigan model has been expanded to include four CEECs in a special study done for this conference. The paper by Hari Vittas and Paolo Mauro is the first to use a disaggregated gravity model to examine trade flows in eight sensitive sectors between the CEECs and several key EU trading partners: Germany, France, Italy, Spain, and Portugal. Finally, the paper by Dieter Schumacher looks in detail at current and likely future trade flows between the CEECs and Germany, using a gravity model disaggregated both by technology level and at the individual industry level.
4
Stanley W. Black
Assessment of the economic effects of entry The Michigan model includes micro-based demand and supply functions for twenty-nine individual tradable goods and services, including twenty-one twodigit manufacturing industries, agriculture, trade, transportation, mining, utilities, and construction. As customary in such models, the level of total employment and the balance of trade are assumed to remain constant in each country. As noted by the authors, there are several ways of interpreting the employment assumption, given the existence of substantial open and disguised unemployment in the CEECs. One interpretation is that the process of economic integration by itself will not affect the aggregate level of employment, in which case there is no problem. Alternatively, if integration does affect the level of employment, the model omits such effects. In any case, the model assumes, like all standard trade models, that the existence of unemployment does not interfere with the market processes built into the model. Less common is the model's careful allowance for differentiated products in manufacturing, along with imperfect competition and economies of scale. The world is subdivided into nine regions: three Central European Countries (CECs: Czech Republic and Slovakia - treated as one country for data reasons; Hungary; and Poland); three divisions of the EU (South: Greece, Spain, and Portugal; EU/EFTA: Austria, Finland, and Sweden; and North); NAFTA; other major industrialized and developing countries; and the rest of the world. Based on the 1992 structure of trade, tariffs, and nontariff barriers (NTBs), the authors examine a range of trade liberalization possibilities: (1) CEFTA (removal of all trade barriers including NTBs between Czechoslovakia, Hungary, and Poland); (2) CEFTA plus a joint CEC-EU FTA (maintaining all existing NTBs); (3) CEFTA plus a joint CEC-EU FTA including relaxation of selected nonsensitive NTBs (the base case, comparable to the Europe Agreements); and finally, (4) CEFTA plus a joint CEC-EU FTA including relaxation of all NTBs (comparable to entry into the EU). The initial tariff barriers between the CECs and EU, EFTA, and NAFTA average around 6 to 8 percent. Removal of these barriers leads to changes in domestic consumption and production and the pattern of trade, and in expanding industries leads to cost reductions as a result of economies of scale. The model's key finding is that all the potential trade liberalizations involving the EU yield positive welfare benefits for the CECs and all regions of the EU, at the cost of negligible negative welfare effects on NAFTA and other major trading nations. Complete liberalization is significantly more beneficial than partial liberalization. The second major finding is that there are no significant effects on the returns to capital or labor in any regions of the EU, given the assumption that all labor is treated as homogeneous. As noted by Winters in his comments, it would be desirable to be able to distinguish between skilled and unskilled labor, since the CECs are expected to export the products of relatively unskilled labor. And finally, while the return to labor in the CECs rises be-
Introduction
5
tween 3 and 5 percent, the return to capital either falls or is unchanged, depending on whether NTBs are liberalized or not. The benefits to labor are larger and capital loses when all NTBs are relaxed, as expected from the Heckscher-Ohlin model. The possibility that one factor gains while the other does not lose is shown to arise from economies of scale due to increased output per firm and the availability of increased varieties of goods. The largest CEC sectoral export effects, in the base case with only nonsensitive NTBs relaxed, come in the areas of leather and footwear, nonmetallic minerals and glass, mining, and food (all CECs), paper products and petroleum products (Czechoslovakia and Poland), and nonferrous metals and metal products (Poland). Eliminating all NTBs expands these effects sharply in agriculture, textiles, chemicals, and iron and steel. The largest effects on exports to the CECs come in the EU-North and EU-EFTA regions, averaging about 10 percent if all NTBs are relaxed, whereas the EU-South's exports rise only about 7 percent overall, although in many individual sectors exports rise much more. Trade liberalization tends to expand output in virtually all sectors. As the authors note, the relatively small sectoral impacts in the EU regions would be negligible if the impacts of the trade liberalization were to be phased in over time. Richardson in his comments points out that the model predicts very favorable outcomes for the CECs, with few problems for the EU. He wishes for more details on the (small) regional adjustment problems affecting the EU-South region. Winters' comments on the paper bring out some of the factors that the authors omitted from their analysis: allowance for differential effects on skilled and unskilled labor, migration, capital flows including technology transfer. While he views the results as generally convincing and the modeling strategy as appropriate, he does raise some interesting queries. The predicted decline in CEC apparel production may reflect the lack of special provisions for outward processing in the Stylized Europe Agreement Package simulated by the authors. The fall in service employment reflects the lower taxes on traded goods, but does not include the effects of the structural change that is taking place as the planners' bias against services is removed. Potential trade with core and periphery A different methodology is used by Hari Vittas and Paolo Mauro to focus on some of the same types of issues. Rather than constructing a complete world trade model, they adopt the increasingly popular gravity model of trade, which assumes that bilateral trade volumes depend directly on the economic size of trading partners and inversely on distance as well as trade barriers. This model, which has usually been estimated for aggregate trade data, is adopted at the sectoral level to examine potential trade in the "sensitive" sectors (agriculture, iron and steel, chemicals, textiles and apparel) subject to NTBs in the Europe Agreements. Vittas and Mauro begin by showing the rapid growth of trade between the
6
Stanley W. Black
six CEECs and Germany, France, Italy, Portugal, and the United Kingdom. Germany has the largest overall volume of trade (both exports and imports). Using Baldwin's aggregate gravity model, Germany has already achieved its medium-term potential level of trade with the CEECs (based on current income levels), while the other four countries have only achieved about 30 percent of their medium-term trade potential, on average. The Baldwin model indicates a large further potential for trade growth as the CEECs income levels gradually rise. Looking more closely at the "sensitive" sectors, Vittas and Mauro find that growth of imports from the CEECs was lower than the growth of total imports for each of the five countries they study, most notably for agricultural products and iron and steel. This appears to be also true for other (non-EU) OECD countries, suggesting that the limits in the Europe Agreements are not the only trade restrictions faced by the CEECs. Next they use indicators of "revealed comparative advantage" (RCA) to ask whether the CEECs actually are increasing their net exports of sensitive products. In several cases they are not, and in agriculture their advantage seems to be receding, suggesting the impact of EU restrictions. Vittas and Mauro then use their disaggregated gravity equations to look at actual versus medium-term potential imports in the "sensitive" sectors. In most cases they find the ratio of actual imports to potential imports in the "sensitive" sectors to be well below the comparable ratio for total imports. In Germany the ratio is 36 percent in the "sensitive" sectors compared to 105 percent for total imports. In France, the ratios are 15 percent and 53 percent, in Italy 28 versus 26, and in the UK 14 versus 28. Textiles, food, and agriculture (in that order) are the areas where protection seems to have held trade back the most. By implication, removal of trade barriers is likely to lead to especially large increases in these same areas and countries. A further simulation examines the effect of partial as compared to complete trade integration, by omitting the effect of the trade agreement dummy variable from the potential trade calculations. Potential trade between the six countries and the CEECs would be only about half as large without membership in the EU, according to Vittas and Mauro. This is considerably larger than the estimate of 20 to 30 percent gains in CEEC trade from the Michigan Model. Looking at factor market flows, Vittas and Mauro conclude that German direct investment in the CEECs is playing an important role in facilitating the growth of trade. On the other hand, employment trends in the EU do not seem to have been influenced negatively in the "sensitive" sectors. In most cases, normal trends do not seem to have been interrupted since 1989. The discussion by Susan Collins agrees with the findings of the paper, but doubts the strength of the evidence from revealed comparative advantage and sectoral gravity equations. In the case of the RCAs, Vittas and Mauro themselves point out that they can be misleading because of the distortions from central planning and therefore use the evidence from RCAs only to raise questions.
Introduction
7
With respect to sectoral gravity equations, the key issue is whether the omission of natural resources and other specific factors biases the results for the sensitive products. Impact on German trade A detailed and thorough analysis of German trade with the CEECs is offered in the paper by Dieter Schumacher, which first reviews post-1989 German trade with six CEECs and the ex-USSR. The data show that Germany is by far the largest Western trading partner of the CEECs. The locus of that trade has shifted to western Germany from eastern Germany, where export competitiveness collapsed after exposure to Western competition and due to sharply higher wage costs after unification. Its focus is primarily the Visegrad countries (Czech Republic, Hungary, Poland, Slovakia), owing to the slower progress of reform farther east. The sectoral pattern involves mainly German exports of investment goods and imports of raw materials from the ex-USSR and labor-intensive consumer and industrial goods from the CEECs. To evaluate the potential level of trade between the CEECs and Germany, some very detailed and thoroughly disaggregated gravity models are estimated. To begin with, data for twenty-two OECD countries' trade with seventy partner countries are used to estimate export and import equations for (1) all goods trade, (2) manufactures, (3) manufactures disaggregated into three levels of technological sophistication: high, medium, and low, and (4) twenty-five three-digit manufacturing industries. Second, the analysis is repeated only for German exports and imports to sixty-nine partner countries, to discover whether the determinants of German trade differ from the "typical" OECD country. In addition to the standard variables of country size and distance, Schumacher includes measures of trade barriers such as membership in a customs union or free trade area, a common land border, common language, and colonial ties. The latter two variables are most often statistically significant, but do not greatly influence the basic results. At the aggregate level, for total trade and all manufactures, the gravity model works very well for exports of OECD countries, but not as well for imports, reflecting the greater importance of omitted resource endowments for OECD imports than for exports. In addition, exports appear to be more sensitive to distance than imports. The grouping of trade data by level of technology indicates higher income elasticities of demand and supply for higher technology items, as should be expected. At the three-digit level, Schumacher displays the distance, size, and income per capita coefficients, obtaining interesting differences across industries. Such factors prove important in estimating potential trade with CEECs as their income levels rise. Basic goods like mineral products, iron and steel, and nonferrous metals are very sensitive to distance, whereas exports of investment goods and imports of consumer goods are not. Supply elasticities rise with GNP
8
Stanley W. Black
and income per head. These appear to be large for mineral oil products, plastics, precision instruments, and transport equipment, as well as iron and steel, glass, and paper products. Demand elasticities appear largest in footwear, leather products, furniture, and clothing. Repeating the analysis for German trade reveals some interesting differences. The commodity structure of German exports tends to vary significantly with distance (low-tech goods only to nearby countries), while the commodity structure of imports varies more with the income level of the supplying country (high-tech goods from richer countries). While the three-digit results offer differences in detail, the general picture is the same as for all OECD trade. When Schumacher uses his gravity model estimates to project potential trade between Germany and the CEECs and Russia, his results are broadly similar to those obtained by Vittas and Mauro or Baldwin, namely that Germany has already reached its medium-term potential for trade with the CEECs. Several different estimates are offered: (1) assuming 1992 levels of income in the CEECs and 1989 levels of German trade, (2) assuming 1992 levels of CEEC income and inflating German trade to 1992 levels, (3) raising CEEC income levels by threefold to account for long-term recovery and growth, and (4) moving the economic center of Germany eastward from Frankfurt to Berlin to allow for the industrial recovery of eastern Germany. The whole exercise is also repeated using the OECD trade equations, as compared with those based on German trade only. Taking estimate (2) as the base comparison, German exports and imports to the CEECs and ex-USSR in 1992 were already at or above the medium-term potential as estimated by the German trade equations. The OECD trade equations give estimates of potential exports and imports to CEECs that are 56 percent and 62 percent higher for 1992, respectively, reflecting the larger role of distance and smaller role of GNP in the trade of the "typical" OECD country as compared to Germany. The CEECs, while poor, are rather close to Germany. Even these higher levels of potential trade have already been reached by actual trade in 1994 for the CEECs, but not for Russia. The impact of trebling income levels in the CEECs is roughly to treble trade levels, for both exports and imports. Moving the industrial center of Germany to Berlin raises them an additional 10 to 15 percent. The three-digit manufacturing sector gravity model results allow Schumacher to predict which industries are most likely to gain or lose market share as CEEC incomes rise. Both exports and imports with highest sensitivity to distance are likely to increase in importance as trade with the nearby CEECs grows. This includes on the export side clothing, wood products, mineral oil products, textiles, and shoes, and on the import side mineral oil products, iron and steel, other nonmetallic mineral products, wood products, and motor vehicles. Low-income countries such as the CEECs tend to buy large amounts of German iron and steel, industrial chemicals, and machinery. As CEEC incomes rise, they will buy more German consumer goods. On the import side, the CEECs' low income levels give them comparative advantage in basic goods
Introduction
9
such as wood products, pottery, iron and steel, and foodstuffs. As their incomes rise, they may specialize more in plastics, industrial chemicals, rubber products, shoes, and paper. An important calculation by Schumacher shows that German exports to CEECs embody more human capital than imports and relatively less raw labor. Thus, increased trade with the CEECs will raise the human capital requirements for the German labor force. As Germany also exports relatively high-tech products to the CEECs, maintaining future competitiveness of the German economy depends on increasing both the skills of the labor force and the technological prowess of industry. Schumacher faults the current mix of government and private spending for devoting too much effort to maintaining the status quo. In addition, the Europe Agreements provide too little scope for CEECs to expand trade based on their comparative advantage. Wolfgang Maennig in his comments raises some of the same questions about the gravity model of trade as Susan Collins. He points to the omission of factors such as exchange rates, subsidies, and infrastructure from the model and asks whether the model can reflect the actual historical experience of the CEECs. Ellen Meade, by contrast, argues for a time series approach to prediction of new trade patterns. 1.2
Investment patterns
A number of recent studies reviewed in the paper by Stanley Black and Mathias Moersch have examined investment behavior in the CEECs. Some have attempted to calculate how much Western investment might be available or how much might be needed to raise growth in the CEECs. Black and Moersch build on these studies by examining the domestic and foreign sources of finance in six CEECs, and then determining potential output growth based on a four-phase transition process including recovery, reform, restructuring, and capital accumulation. Others have studied the progress and requirements for privatization of the previously State-Owned Enterprises (SOEs), widely thought to be a prerequisite for vigorous investment and growth. Paul Welfens examines the various ways in which the privatization process may be expected to affect productivity in the CEECs. Investment and its financing The paper by Black and Moersch constructs a simulation model of the transition process for six CEECs based on accumulation of capital through domestic and foreign savings and a production function parameterized on the basis of EU experience. The model requires estimates of initial output, capital stock, and labor force in each country, domestic and foreign savings ratios, and the (in-) efficiency with which resources are initially used. The transition process consists of a ten-year period with gradual recovery of full employment, reduction of in-
10
Stanley W. Black
efficiency, reorientation of production from heavy industry to light industry and services, and replacement investment and accumulation of new capital. Focusing initially on domestic savings potential, Black and Moersch argue that relatively strong savings behavior in the CEECs prior to the inflation shock of 1989-90 can be restored if recently negative real interest rates become positive again. On the other hand, diversion of savings to finance government deficits and loss-making SOEs reduces the availability of finance from domestic sources. Foreign capital inflows have been substantial and growing, mainly in the Visegrad countries, and can be expected to increase if the CEECs follow appropriate exchange rate and macroeconomic policies. For the CEECs as a group over the ten-year period, the simulations assume that domestic savings rise from 23 percent to 29 percent of GDP, while foreign savings decline from 3 percent to 1 percent of GDP. Government deficits are assumed to decline from 5 percent of GDP to zero. The initial capital stock of the CEECs in 1992 is estimated, following Boote (1992), by assuming that CEEC capital resources were used with only 62 percent of the efficiency of EU capital of the same type and that labor resources were only 26 percent as efficient as EU labor. During the process of reform, these efficiency levels are raised to 100 percent and 73 percent, respectively. The recovery process involves raising the employment rate from 90 percent to 95 percent of the labor force. The restructuring process involves shifting the mix of output from 60 percent in industry to 65 percent in the service sector, comparable to that in the EU. The resulting improved allocation of resources is estimated to add 16 percent to productivity over the ten-year process. Combining these various factors, Black and Moersch find that productivity in the CEECs would rise from about $10,000 per worker to about $33,000 per worker, or about triple the original level. This is the same growth factor as assumed by Schumacher in his paper. Productivity would rise from 25 percent to 63 percent of the EU level. This favorable outcome depends on a variety of positive developments, including containment of budget deficits, maintenance of positive real interest rates and appropriate real exchange rates, restructuring of industry, reorganization of production and management methods, retraining of workers, and improved financial intermediation, as well as receptive markets for CEEC exports and capital inflows of some $15 billion per year. Less favorable conditions would lead to less economic growth in the CEECs, a pessimistic outlook leaving CEEC productivity at only $16,840 per worker after ten years, about half the optimistic outcome and only about a third the EU level. The key factor is the rate at which the efficiency of utilization of resources approaches Western levels. Individual CEECs vary around the average, depending on their starting point and individual savings, borrowing, and reform and restructuring capabilities. The Czech Republic and Slovakia start out ahead, and maintain their lead over Hungary and Poland. Bulgaria appears to gain the most rapidly, while Romania starts out far behind and remains behind until late in the simulation.
Introduction
11
The basic lesson of the Black and Moersch paper is that substantial progress in the CEECs is possible, but depends primarily on their own policies and behavior. Nevertheless, open markets and willingness to lend and invest are important Western contributions. Douglas Todd's comments recast the Black and Moersch production-oriented analysis into a factor-efficiency framework, showing the distance between the more advanced and less advanced CEECs. He points out the difficulties the less advanced will have in approaching full efficiency and suggests that the reform process will take considerably longer than ten years. He points out that the service sector in Western Europe has become more capital intensive in recent years and wonders if the same phenomenon may not occur in the East. While accepting the framework of the Black and Moersch analysis, Todd leans toward the pessimistic side of their outcomes, especially in light of problems within the European Union in achieving convergence on the Maastricht criteria for monetary union. Holger Wolf in his comments finds reasons for adopting the pessimistic version of the Black and Moersch simulations. He believes that not much is to be expected from recovery, restructuring, or reform and that much higher investment will be required to achieve output gains than is implied by Western production functions. He is also pessimistic about the savings potential in the CEECs, despite their earlier record. An aging population, newly available consumer credit facilities, and persistent government deficits are likely to drag down the savings rates. Rather than looking to financial institutions for intermediation, Wolf argues that most savings will have to come from retained earnings in the newly privatized or newly organized firms. Overall, he expects a Latin American rather than an Asian transition for Central and Eastern Europe. Privatization and productivity A key aspect of the transition process not emphasized in the previous paper is privatization, explored rather thoroughly by Paul Welfens in his paper. The paper focuses on the relationship between productivity growth and privatization and their relationship with structural change in the CEECs. Welfens begins by reviewing current productivity performance and related data for the various CEECs, including countries of the former Soviet Union, showing the important differences in performance between the leading and lagging participants in the transition process. The paper next turns to consideration of the sources of productivity gains. Privatization will significantly affect the efficiency of the utilization of the traditional factors of production. In addition, and highly relevant for the CEECs, are issues such as the degree of competition when state monopoly has been the dominant form of organization, externalities such as network effects when large numbers of firms are being created, and the availability of infrastructure. Welfens also argues that labor income could fall relative to capital income in
12
Stanley W. Black
the process of privatization, especially in Russia, where comparative advantage appears to favor capital-intensive industries such as shipbuilding and oil. Privatization has moved fastest in Hungary, the Czech Republic, and Poland, with favorable results on productivity in the newly private firms, but the techniques used in each country have varied significantly. Farther east, progress has been slower and results less favorable. The degree of privatization also seems related to the degree of structural change, as the growth of the service sector generates large numbers of new firms. It seems clear from the evidence in this paper that the ability of the CEECs to transform their economies and raise their income levels depends strongly on the ability to create a vigorous private economy. Bruce Kogut in his comments points out that in Russia, privatization has frequently been accompanied by relatively low valuation of assets, especially compared with Western standards. Of course this may reflect the high degree of uncertainty over future economic conditions. Nevertheless, privatization under such conditions may involve an asset transfer which, especially if it is to foreign owners, is politically unattractive. Furthermore, if foreign investment takes place for purposes of gaining control of markets instead of to improve productivity, there may be a rational basis to limit foreign ownership. Kogut accepts the argument that privatization has broken the power of the ministries in Russia, even if it has not led to higher productivity. While the key problem of corporate governance has not been solved, the basic underlying conditions for competition have been created. Jan Mladek in his comments focuses on the political economy of transformation. He notes the important distinction that must be made between Central Europe (Western Christianity) and Eastern Europe (Orthodox Christianity). He believes that privatization should stop at utilities, such as railroads and the health system. And in the Czech context, where many small firms have been privatized through vouchers, attempting to trade publicly the shares of such small concerns should be avoided. Such trading would place too great a disclosure burden on small firms. 1.3
Labor market issues
The low incomes and wages of the CEECs represent both their predicament and their opportunity, particularly in combination with their relatively high levels of human capital. The trade papers suggest that workers in the European Union do not have too much to fear from the competition of imported goods from Central and Eastern Europe. But direct competition in the form of immigrant workers competing for the same jobs could prove a more substantial threat, especially in the face of persistent high unemployment rates in Western Europe. The paper by Thomas Bauer and Klaus Zimmermann examines this question. The motivation for considering immigration a serious issue comes from the demographic structures of Eastern Europe - young, poor, and growing - ver-
Introduction
13
sus Western Europe - rich, old, and stagnant. Western Europe's need for new blood and cheap labor matches Eastern Europe's ability to supply both too well for migration not to be part of the solution. The match is particularly strong in the case of skilled labor, which can most efficiently be retrained at temporary employment in the West. Ethnic networks of compatriots are seen as influencing the direction of flow more than simple wage differentials. Recent moves to a border-free Europe are forcing the pace of common immigration policy in the EU. Coordinated asylum and short-term visa policies are just the first steps. Beyond these lie longer term policies on immigration, which differ according to national preferences and needs. Germany seeks to accommodate ethnic Germans from the East, France seeks to control the rate of immigration of ex-French colonials. Both seek to prevent illegal immigration. German policy toward her eastern neighbors takes the form of temporary work permits, which can be adapted to labor market conditions in Germany and which also facilitate training and experience of CEEC workers who will upon return home have increased rapport with German firms and interests. A major focus of the paper is to estimate gains and losses to the German economy from such migration. The key distinctions are between effects on skilled labor and unskilled labor and between a model with full employment and market-clearing wages versus a model based on union behavior to set wages of unskilled workers above the market-clearing level, generating persistent unemployment. A major assumption is that skilled and unskilled labor are complements in the production process. Under these assumptions, immigration of unskilled labor raises the demand for and wages of skilled labor, but drives down the wage of unskilled labor, which leads the market closer to equilibrium. Immigration of skilled labor reduces the skilled wage and raises the demand for unskilled labor, in the standard case also reducing the unskilled wage and leading the market closer to equilibrium. The logic of this model needs to be understood in the context of imperfectly competitive European labor markets. Adding to the supply of labor is going to lead the economy closer to the competitive case by undermining union power over wages. Of course the distribution of the resulting gains is going to depend sharply on the types of immigrants, since the skilled wage rises or falls depending on whether the immigrants are unskilled or skilled. This model is then calibrated to the German economy in 1993, with alternative immigration levels of 10 percent, 5 percent, and 1 percent, as well as alternative elasticity estimates. With 10 percent unskilled immigration, the gains to the economy run from zero to 4.36 percent of national income. Depending on how much the wage of unskilled labor falls, they could lose up to 5.53 percent of national income. This is the least favorable case for immigration. With 10 percent skilled immigration, gains to the economy run from 7.44 percent to 11.41 percent of national income, depending on how much the wage of unskilled labor falls. It may seem hard to believe that economic gains from immigration can be this high, and in a full employment economy they would
14
Stanley W. Black
not. By allowing fuller employment of the existing labor force as the union loses influence over wages, the immigration policy succeeds as a jobs program, according to Bauer and Zimmermann. The most effective way to achieve that goal, they find, is to concentrate on temporary immigration of skilled labor. Barry Bosworth in his comments contrasts the United States, where migration analysis focuses on distributional aspects, with Bauer and Zimmermann's results that there can be major aggregate effects on German GNP. He uses the standard Harberger triangle to argue that gains to migration must be small. Bosworth does not accept Bauer and Zimmermann's argument that migration, by removing a rigid wage barrier to employment, may have a first-order effect on total output and employment. Instead, he questions the formula. Robert LaLonde suggests that the largest impact of immigration into Germany should be on employment rather than wages, reversing U.S. experience, because wages are less flexible in Germany than in the United States. He finds the Bauer and Zimmermann results to be surprisingly large. LaLonde also points out the difficulties with enforcement of border controls and the problems with making skill-based immigration policies work effectively. 1.4
The process of integration
The previous papers have analyzed the costs and benefits of CEEC integration into the European Union, without considering many of the practical political aspects involved in the actual process. This is acceptable for economic analysis, but seriously inadequate in the larger sense of political economy, which is essential for understanding the motivations of the potential entrants as well as the EU. Like the previous Southern enlargement, which brought in Greece, Spain, and Portugal, all poorer and with less developed democratic institutions than the other members, the proposed Eastern enlargement can only be understood as a means of consolidating democracy and economic liberalization in the former Communist countries of Central and Eastern Europe. Despite the economic costs of admitting much poorer countries to the EU, the benefits of guaranteeing a permanent end to the East-West division of Europe furnish a powerful motive for the EU to move ahead. The paper by Michael Koop considers several options: (1) quick entry (i.e., by 2005 with a ten-year transition period) for the four most advanced CEECs, (2) partial membership (or what is sometimes called variable geometry), (3) membership in the EEA (European Economic Area) as a halfway house, and finally, (4) building on the existing CEFTA (Central European Free Trade Area) as a means of liberalizing trade with the EU. Koop also distinguishes three important groups of countries: (1) the advanced CEECs, many of whom already have bilateral Europe Agreements with the EU (the Czech Republic, Hungary, Poland, Slovakia, Slovenia, and possibly the Baltic countries), (2) a second group, which is making progress both economically and politically at a slower rate than the first group (Albania, Bui-
Introduction
15
garia, Croatia, Macedonia, and Romania), and (3) the European successor states of the Soviet Union (Belarus, Moldova, Russia, Ukraine). The major criteria for EU membership are the existence of a stable democracy, the rule of law, a decent human rights record, a functioning market economy strong enough to withstand international competition, and ability to accept the obligations of membership. The paper concentrates its attention on the first group of countries, all of whom "basically" satisfy these criteria. The second group is somewhat behind in its progress in many of these areas, while the third group is far behind in all. The solution to the problems of the Southern enlargement was quick entry with lengthy transition periods for agriculture and weak industries, cushioned by financial transfers to assist the new members with convergence toward the income levels of other members. This was feasible because the Southern members did not place too great a financial drain on the rest and because it was not yet appreciated what a burden their entry could place on decision-making procedures. The relative size, number, and low incomes of the CEECs make the financial and institutional problems much worse, along with the economic issues of trade and migration, analyzed in earlier papers in this volume. Institutional problems include the proliferation of languages and speakers at meetings, the feasibility of retaining unanimity on key issues with so many members, the issue of proportional representation for large countries, and the loss of influence of small countries. These issues appear to require constitutional solutions, which might possibly be reached at the 1996 Inter-Governmental Conference (IGC). The budgetary problems include the high cost of extending the Common Agricultural Policy and the Regional and Structural Funds to the CEECs. Taking account of prior restructuring in the CEECs and the reforms already made in the CAP suggests net budgetary costs of 20 to 25 billion ECU, which must be financed either by raising taxes or reducing benefits. These costs could be reduced by revenue arising from the economic gains to the EU and to the CEECs, as estimated by Brown, Deardorff, Djankov, and Stern, for example. Koop suggests that the difficulties with the quick-entry option may have been overstated. Expansion to fifteen members by inclusion of the EFTA countries Austria, Finland, and Sweden already requires institutional reform. Preparers of the 1996 IGC are considering a variety of reform options. One issue is restricting the size of institutions, which involves a trade-off between majority rule and minority rights. A bicameral legislature is one possibility, already implicit in the dual legislative roles of the Council of Ministers and the European Parliament. The unwillingness of national governments to strengthen the Parliament leads Koop to suggest a two-tier decision-making rule in the Council, based first on simple majority of votes and second on proportion of population. Such fundamental constitutional questions can be resolved only if the mem-
16
Stanley W. Black
bers agree on their objectives. To help bridge the gap between existing members, Koop suggests inviting the potential new members to participate in the IGC. Otherwise the outcome risks restriction to the "least common denominator" of integration. It seems unlikely, however, that current members would give potential members the vote ahead of time. The budgetary problem may be easier to solve, particularly if it is recognized that different members have different interests in enlargement, allowing for side payments in the form of a "package deal." Rapid transformation in the CEECs, as indicated for example in the paper by Black and Moersch, will reduce the scope of the problem, as would further desirable CAP reforms. Koop suggests temporary accretions to EU resources, either in the form of "integration bonds" or a temporary VAT tax with a higher rate for those who benefit more (i.e., Germany). These solutions may be blocked by those who perceive themselves as losers, especially the smaller and poorer countries, those who hope to deepen integration before widening (France), and those who fear that widening must be preceded by deepening (the United Kingdom). A further problem relates to the order of entry. Delaying entry may also delay the transition process in the CEECs, as vested interests see a chance to forestall change. Leaving out some members of Koop's first group on technical grounds could intensify the danger of political backsliding. Koop proposes giving reasonable timetables to both groups one and two to avoid this problem, say 2005 and 2010. After reviewing the other options, Koop finds them all to have serious drawbacks as compared with quick entry. Restricted or partial membership, as well as variable geometry, could save money on the CAP, but if applied to as important a field as agriculture, it represents a significant limitation on the "core" freedoms of the EU Treaty: free trade in goods and services, free movement of labor and capital. Membership in the EEA would guarantee the four freedoms and removal of trade barriers except in agriculture and would leave the CEECs out of the Regional and Structural Funds. While again a money-saver, the EEA offers no political or security benefits to the CEECs and limits their growth prospects by excluding them from important economic gains. The CEFTA option is already in place and hence offers relatively few additional benefits to CEECs, except for those not yet members of CEFTA. As Brown, Deardorff, Djankov, and Stern indicate in their paper, the trade benefits of CEFTA without EU membership are only a small fraction of the benefits of full EU membership for the CEECs. While concluding that quick entry in two phases is the best option for Eastern enlargement, Koop is under no illusion that it will be easy to obtain EU agreement to this process, given the sharp divisions among EU members themselves over the future direction of the Union. Barry Eichengreen in his comments points out that the relative difficulty of bringing the CEECs into the EU in the near future is not so different from the Southern enlargement of the European Community as many people think. The CEECs are proportionately poorer and more agricultural than the EU on aver-
Introduction
17
age today. But they are not worse off relative to the EU than Greece, Spain, and Portugal were relative to the EC in 1980. Eichengreen considers the potential enlargement from several viewpoints: the bureaucrat, the economist, and the political economist. From the bureaucratic perspective, EU enlargement will cause problems with voting in the Council and with financing the Common Agricultural Policy. But these problems already exist and have to be dealt with anyway. Neither migration nor trade in the "sensitive sectors" is actually a serious problem. From the perspective of an economist, the EU will gain from trade creation with the CEECs, while the CEECs need to worry about trade diversion with the EU, as well as onerous social obligations they are too poor to afford. From the political economy viewpoint, the Eastern enlargement is valuable both to solidify the commitment to reform in the CEECs and to provide them with a security blanket as members of the EU (if not NATO). Within the EU, German economic and security gains must be offset against losses of transfers and voting power to the Southern members. Taking these factors into account, Eichengreen compares the entry alternatives reviewed by Koop, finding early entry to be most costly but also most attractive to the CEECs (and Germany). Variable geometry is problematic because it undoes the complex bargain of the acquis communautaire. He predicts that early entry will be limited to the Visegrad countries, while the others may accept an EFTA-like arrangement that avoids the issues of the CAP and Structural Funds. Hans-Jurgen Vosgerau points out that the EU has insisted that the CEECs adopt EU competition policy as part of the Europe Agreements before giving them full access to EU markets. This is more demanding than the EU has been to its own members during the Single Market initiative and to the EFTAns in the European Economic Space. He also notes that the gathering clouds over the unfunded pension schemes in the EU will probably require the development of funded alternatives, thus providing an ample source of future investment funds for the CEECs. 1.5
Conclusions
The conference papers depict clearly the substantial potential as well as some of the difficulties associated with the enlargement of the European Union to the East. The Brown, Deardorff, Djankov, and Stern paper shows the substantial gains that would accrue to the CEECs and the smaller gains to EU members from further trade liberalization as well as the negligible costs. It is clear that reforming the CAP, while not analyzed by the Michigan model, would increase those potential gains. Winters and Richardson generally concur with the findings of the Michigan Model, while calling for further research on issues not yet addressed. The paper by Vittas and Mauro complements this analysis by focusing on the "sensitive" sectors of agriculture, textiles, leather, and iron and
18
Stanley W. Black
steel to show that significant adjustments must be expected in those areas, along with the overall gains. However, it does not appear that "sensitive" employment would be much affected. They also confirm that membership in the EU will make a significant difference for the CEECs in these areas. Collins agrees with their overall findings, while questioning their methods. Schumacher's paper, concentrating on Germany, indicates that much of Germany's adjustment, at least in the west, is already accomplished. Further changes will reflect the gradual rebuilding of manufacturing capacity in eastern Germany. He offers many detailed predictions of further industrial change. For Germany, as well as for the rest of the EU, opening to the East brings a further shift in the mix of labor demanded toward the higher skilled levels. Black and Moersch show through their simulation model that the CEECs are capable of achieving very substantial growth, if they choose the right mix of policies and face an open market for their products. On the other hand, they show that much slower growth will occur if there is inadequate saving, insufficient reform, and incomplete restructuring. Wolf and Todd think this is the more likely scenario. Welfens's paper shows how important privatization is to the reform process. He points out that it has proceeded fastest in the Visegrad countries, but that it may have different effects in Russia. Kogut and Mladek agree. European labor markets suffer from excess unemployment, owing to a variety of factors including union power, which keeps wages above marketclearing levels. Bauer and Zimmermann argue that importing labor from the East can break up this monopoly and lead to fuller employment and substantial gains for the West. Bosworth and LaLonde are not so sure of this outcome. Koop in his analysis of the alternative approaches to integration of the CEECs into the EU suggests that a relatively speedy entry offers the best chance to realize the various gains to all parties analyzed in the other papers at the conference. Eichengreen accepts this for the Visegrad countries and suggests an EFTA-like arrangement for the others. As the conference was concluding, the EU White Paper spelling out the gaps to be closed between CEEC policies and the acquis communautaire was published. The extensive range of areas covered leaves little doubt that the goal of entry, while clearly feasible for some of the CEECs, remains at the end of a lengthy and difficult process. But the economic and political benefits for both sides, many of them analyzed in these papers, will keep the parties engaged throughout, however long it may take. REFERENCES Baldwin, R. (1994). Towards an Integrated Europe, London: Centre for Economic Policy Research. Bergstrand, Jeffrey (1989). "The Generalized Gravity Equation, Monopolistic Competition, and the Factor Proportions Theory in International Trade," The Review of Economics and Statistics 71, 143-53.
Introduction
19
Boote, A. R. (1992). "Assessing Eastern Europe's Capital Needs," International Monetary Fund Working Paper WP/92/12, February. Collins, S., and D. Rodrik (1991). Eastern Europe and the Soviet Union in the World Economy, Washington, D.C.: Institute for International Economics, Policy Analyses in International Economics, May. Rollo, J., and A. Smith (1993). "The Political Economy of Central European Trade with the European Community: Why So Sensitive?", Economic Policy 16, 140-81. Winters, A., and Z. Wang (1994). Eastern Europe's International Trade, Manchester: Manchester University Press.
PARTI TRADE RELATIONS
CHAPTER 2
An economic assessment of the integration of Czechoslovakia, Hungary, and Poland into the European Union Drusilla Brown, Alan Deardorff, Simeon Djankov, and Robert Stern 2.1
Introduction
This paper is a study of the economic effects of the integration of the Central European Countries (CECs)1 into the European Union (EU). Our analysis of EU-CEC integration is based on a specially constructed version of the University of Michigan Computational General Equilibrium (CGE) Trade Model. We use this model to calculate the economic effects of EU-CEC integration on trade, output, and employment by sector as well as the real returns to capital and labor and the economic welfare of the CECs, the EU members, and the other major trading country aggregates included in the model. Our study is distinctive in two respects. First, we bring together the elements of the EU assistance to the CECs and construct a Stylized European Agreement Package (SEAP), rather than analyzing them as separate initiatives for each CEC country. This provides a more synthesized account of the scope and magnitude of EU-CEC integration than has been previously available. Second, by using a CGE model to evaluate EU-CEC integration, our paper complements previous studies of the EU-CEC agreements that have considered: (1) the effects of the Europe agreements on EU-CEC trade (Winters and Wang, 1994); (2) the positive welfare effects of improved access to the EU markets for the CECs (Aghion et al., 1992); (3) the reaction of the EU to changes in the trade policies of the CECs (Messerlin, 1992); and (4) the potential CEC trade patterns as reflected within a gravity model framework (Baldwin, 1994). The paper is organized as follows. Section 2.2 discusses in broad terms the various sources of gain that might arise from integration of the CEC into the EU. Section 2.3 is a brief review of the existing literature on EU-CEC integration. Section 2.4 examines the characteristics of the actual EU-CEC integration by describing a Stylized Europe Agreement Package. Section 2.5 similarly describes the integration that has occurred among the CECs, that is, the Central 1
The CECs include the Czech Republic, Hungary, Poland, and Slovakia. Prior to 1993, the term referred to Czechoslovakia, Hungary, and Poland. 23
24
D. Brown, A. Deardorff, S. Djankov, and R. Stern
European Free Trade Agreement (CEFTA). Turning to our formal analysis, Section 2.6 outlines some of the essential features of the Michigan CGE Trade Model. The various scenarios investigated using the model are then presented in Section 2.7, together with the aggregate results of the model simulations. The sectoral results are presented in Section 2.8. In Section 2.9 the roles of scale and variety are addressed. Section 2.10 contains our conclusions and implications for further research and policy. 2.2
The gains from integration, broadly conceived
Before narrowing our focus to those aspects of economic integration that we have been able to include in our CGE model, it is useful to look first at integration in broad terms. We therefore begin here with an overview of the many effects that integration may be expected to have for the participating countries. These include both general effects that are likely to arise in any preferential trade arrangement, and effects that depend on the special circumstances of the CEC countries. Integration with the EU creates both static and dynamic gains for the CECs. The static gains arise from increased efficiency of resource reallocation. The shift away from high-cost domestic producers to low-cost suppliers located in other European countries forces factors of production to undergo redeployment to sectors and firms that yield the highest returns. Structural market distortions (tariff barriers, subsidies, border formalities) are eliminated, further enhancing competition. Static gains also arise from product and process specialization of firms within the region, increased trade in intermediate products, and realization of economies of scale in imperfectly competitive industries.2 Dynamic gains from integration result in general from dynamic scale effects and easier transfer of technology, which increases the productive capabilities of member-country firms. Dynamic scale effects may in turn lead to factor pooling, which makes regional specialization self-reinforcing (Ethier, 1982). The flow of foreign direct investment (FDI) to the member countries may also increase, thus accelerating capital formation. In addition, there may be increased investment in human capital. Further, budgetary transfers from the structural funds of the EU may encourage additional investment. Last, integration enlarges the number of varieties of goods available in any CEC economy, and this may yield considerably larger welfare gains as compared to conventional estimates (Romer, 1994; Aghion and Howitt, 1992; and Grossman and Helpman, 1992). For the CECs in particular, there may be other benefits as well. Preferential access to EU markets may ease the process of trade reorientation, necessitated by the collapse of the Council for Mutual Economic Assistance (CMEA) and 2
Note in this connection that the overall static effects of the "1992" completion of the internal EU market were estimated at 2.5-6.6 percent of the EU's GDP (Cecchini, 1988).
An economic assessment of integration into the EU
25
the loss of the former USSR and East German markets. Second, the CECs are joining a regional bloc based on democratic principles,3 and this may reinforce democratization in the CECs. Worthy of note here is the ostensible success of the enlargement of the EU in enhancing democracy in Greece, Portugal, and Spain. Third, the PHARE4 program and the loan activities of the European Investment Bank/European Bank for Reconstruction and Development (EIB/EBRD) may result especially in increased investment in infrastructural and telecommunications development in the CECs, thereby reducing real production costs of CEC firms (Baldwin, 1994). 2.3
Literature review of EU-CEC integration
In this section, we briefly review a number of noteworthy studies of EU-CEC integration. Messerlin (1992) analyzes the Europe Agreements (EAs) of the (then) three CECs. His study focuses on the lack of trade concessions in the sectors most vital for these economies: agriculture, iron and steel, and textiles and apparel. Excluded de facto from the EAs trade liberalization package is a fourth important sector - chemicals - because it leaves unchanged the strict antidumping (AD) procedures against the CECs in this sector. Aghion et al. (1992) argue that, as a result of unilateral steps taken prior to the EAs, the CECs now have lower, more uniform, and more transparent protection than most of the OECD countries. Unless there is a revision of the EAs to grant additional concessions to the CECs in some sensitive sectors, they suggest that a reversal of the trade liberalization stance toward a more protective regime is possible. Winters and Wang (1994) discuss the principal components of the EAs and their economic effects. In their view, although it is constructive to establish an EU-CEC relationship in legal form, the EAs are disappointing in the degree of support they guarantee to the CECs. Their paper includes a synopsis of the 1993 Copenhagen Summit changes to the agreements, with detailed tables on the timing of concessions and removal of nontariff barriers (NTBs). Most of the attempts at quantifying the expected results of the EAs have been undertaken in a partial equilibrium framework. Winters and Wang (1994) employ a common conceptual basis in their study of the iron and steel, clothing, and footwear industries, although their three simulation exercises differ in detail. In each case, a different grouping of supplier-countries is used, to better reflect the comparative advantage in each sector. They assume that there is more direct competition, and hence greater substitutability, among suppliers within groups than between them. The standard Armington assumption is used, but in addition products are geographically disaggregated by place of sale. They con3 4
For a similar argument for NAFTA, see Krugman (1993). Poland and Hungary Assistance for Restructuring the Economy.
26
D. Brown, A. Deardorff, S. Djankov, and R. Stern
elude that the EU members have little to fear from opening up the sensitive sectors, while the CECs' gains from such opening are substantial. Winters (1994) focuses on the steel industry, making explicit allowance for the existence of excess capacity, non-marginal-cost pricing, and initial industry losses. He asks what would have happened if the EAs had existed in 1992 and allowed for complete steel liberalization. He shows that steel users everywhere would have gained substantially, with consumers inside the EU gaining 1.75 billion ECU as a result of the liberalization of steel trade. Rollo and Smith (1993) use a CGE model to analyze the effects of the EAs, focusing on agriculture in particular. Their framework is based on an earlier study by Gasiorek, Smith, and Venables (1992) that models the "1992" completion of the EU internal market. They show that if the CECs were included in the Common Agricultural Policy (CAP) of the EU, the net effects would be a 2 billion ECU sectoral gain for the CECs and an approximately equal welfare gain for EU consumers. A number of recent studies have also looked at the effects of the Europe Agreements on individual EU members: Greece (Dimelis and Gatsios, 1994), France (Cadot and de Melo, 1994b), Portugal (Corado, 1994), and Spain (Martin and Gual, 1994). In each of these cases the effects on the EU members were insignificant. Finally, Hoekman and Pohl (1995) use trade data to make some preliminary assessments of the extent, speed, and location of enterprise restructuring in the CECs. They conclude that there has been very significant reorientation of the trade of the CECs toward the EU since 1989, with the most noteworthy changes evident for the Czech and Slovak republics.5 2.4
The Stylized Europe Agreement Package (SEAP) The Stylized Europe Agreement (SEA)
While the EU has negotiated separate agreements with individual countries, these agreements have a number of features in common. It is useful accordingly to review some of their main features in what we will refer to as the Stylized Europe Agreement (SEA).6 The SEA consists of 124 Articles covering areas of both Community and Member State competence. The agreement is concluded for an unlimited period. It is intended to create a free-trade area within a transitional period of ten years, with a shorter timetable of liberalization on the EU side (asymmetry of concessions). "The European vocation" of the applicant is explicitly recognized: the CECs are treated as potential members although no accession dates are specified. Under the SEA, EU trade barriers will be progressively abolished 5
6
Other recent studies pertinent to EU-CEC trade and investment relations include Cadot and de Melo (1994a), Halpern (1994), and Neven (1994). The main elements of the SEA are treated in detail in Winters and Wang (1994, esp. pp. 32-52).
An economic assessment of integration into the EU
27
in five years except for textiles, where it will take six years. The elimination of quantitative restrictions is linked to the results of the Uruguay Round. Apart from a few exceptions, these objectives will be attained after seven years for Poland and after nine years for all other countries. Further concessions are applied on a reciprocal basis. A distinction is made between "general industrial goods" and "sensitive" sectors, such as agricultural products, steel, coal, textiles, clothing, and chemicals, where EU liberalization is limited during the transitional phase. Trade in processed agricultural products and in fishery products is governed by specific provisions. Thus, while the SEA eases access for CEC exports to the EU markets, this is done largely by consolidating previous concessions. The SEA also includes common provisions such as a standstill clause prohibiting the introduction of new trade restrictions, a safeguard clause, antidumping provisions, and definitions of originating products. The CECs are allowed to derogate the standstill clause in exceptional cases to protect infant industries and industries under restructuring. In the area of competition, EU rules apply. However, the associated countries are considered as low-income regions where, in accordance with the EU Treaty, development-oriented state aids may be authorized. At the Copenhagen European Council Summit (June 22-23, 1993), the EU accelerated unilaterally the opening of its market to the signatories of the Europe Agreements. The five-year liberalization period for industrial products in the agreement was shortened by two years.7 Concessions on agricultural products outlined in the Interim Agreements were pushed up by six months. The deadline for the abolition of duties on direct imports of textiles was shortened to five years instead of six and the remaining duties on steel are also to be eliminated sooner than originally planned (four years instead of five). The PHARE program 8
PHARE is the second integral part of the SEAR It is the EU's aid program to support economic restructuring and democratic reform in the CECs. Assistance under PHARE is provided in the form of nonreimbursable grants. The PHARE funds are awarded through individual indicative programs under which each recipient country submits a proposal to the EU Commission for assistance on specific projects. In this way, the beneficiary countries decide on their own restructuring priorities. The 1991 Guidelines identify five core areas of assistance: (1) restructuring and privatization of state enterprises; (2) support for the private sector (notably to small and medium-sized enterprises), investment promotion (15 percent of PHARE funds are devoted to infrastructural investment support), and tourist promotion; (3) modernization of the financial system, from fiscal policies to financial services; (4) development of an affordable social 7 8
This gives free access by January 1, 1995, instead of January 1, 1997, for the CECs. For more detail, see European Commission (1994b).
28
D. Brown, A. Deardorff, S. Djankov, and R. Stern
safety system from active employment policies to antipoverty measures; and (5) support of the policy reforms, demonstrated through the establishment of regulatory and legislative frameworks. PHARE also has sectoral objectives, concentrating on agriculture, infrastructural development, energy, and communications. It does not grant direct financial support to private business ventures. Instead, contracts funded by PHARE are awarded under public procurement procedures, either by the competent authorities of the recipient state or by the European Commission. Besides national programs that meet each country's specific needs, there are also regional programs that involve transnational issues, such as environmental protection, education, and operation of joint ventures. Examples include public administration reform (SIGMA program), further education and training (TEMPUS), and economic and scientific research (ACE). In terms of framework and components, PHARE is similar to the Structural Funds within the EU, but an important difference is that funds are given prior to EU accession with the aim of assisting the CECs in fulfilling entrance requirements. The European Investment Bank (EIB) As a main financial institution of the EU, the EIB 9 has provided loans for the development of large public sector, infrastructural projects in the CECs. The EIB has also acted as a catalyst in forming the European Bank for Reconstruction and Development (EBRD), which aids the growth of the private sector in the CECs. Together, the EIB, the EU, and its member states are the largest shareholders in the EBRD, together holding more than 51 percent of the bank's assets. In addition to private sector lending, the EBRD and the EIB jointly finance several public projects in each CEC. Central among these projects are privatization of the national telecommunication systems and privatization and modernization of the banking sector. 2.5
CEC-CEC integration
Alongside the EAs with the EU, the CECs have concluded the Central European Free Trade Agreement (CEFTA).10 The CEFTA was a response to escalating import duties on products imported from other CECs. Signed bilaterally among the four countries in 1992, CEFTA envisages free trade among them by the year 2001. The agreement differs from the EAs in its symmetric tariff removal time schedule, which was first implemented in March 1993. The trade liberalization includes: 9 10
See European Commission (1994a). Based on Bakos (1993).
An economic assessment of integration into the EU
29
Step 1: Immediate dismantling. This covers industrial raw materials (Polish copper, Hungarian aluminum, etc.) and some industrial manufactures (e.g., Polish agricultural machinery, Hungarian Pharmaceuticals). Special tariff exemptions cover machinery imports subject to quota limits. Step 2: Tariff elimination by the end of 1996. This includes most industrial products. Import duties are to be cut by one third each year for three years, starting in January 1995. Trade liberalization in agricultural products falls into one of four categories. Some products receive a 20 percent total tariff reduction over two years. Others receive a 50 percent total tariff reduction over five years. In each of these two instances, the importer may levy quotas on a restricted number of products. Step 3: Elimination by 2001. The list includes "sensitive industries" (e.g., motor industry in Poland, textile production in the Czech Republic and Slovakia, steel in Hungary). As CEFTA was concluded bilaterally, the list between any two of the four countries is longer and product specific. 2.6
The Michigan CGE trade model
Ideally there are four essential components that should be captured in a model of EU-CEC integration: (1) reduction or elimination of tariffs and NTBs; (2) rationalization of the production process by capturing scale economies and increasing product variety; (3) reduction in real transaction costs (e.g., reduced transportation and communication costs, simplified border formalities, harmonized product and safety standards); and (4) facilitation of technology transfers and new investment in physical and human capital. Our CGE model captures the effects of only the first two components. It is an extension of the model first constructed by Brown and Stern (1989) to analyze the economic effects of the Canada-U.S. Trade Agreement (CUSTA), and later expanded by Brown et al. (1992a,b, 1994, 1995, 1996) to analyze NAFTA, the extension of NAFTA to some major trading countries in South America, an East Asian trading bloc, and a free-trade agreement between Tunisia and the EU. In its further elaboration for present purposes, we model the three CECs - Czechoslovakia, Hungary, and Poland - individually. The EU is divided into three groups: EU-North (Belgium-Luxembourg, Denmark, France, Germany, Ireland, Italy, Netherlands, and the United Kingdom); EUSouth (Greece, Spain, and Portugal); and EU-EFTA (Austria, Finland, and Sweden). We also include the three NAFTA countries (Canada, Mexico, and the United States) and an aggregate of another fifteen major industrialized and developing countries. All remaining countries of the world are consigned to a residual rest of world to close the model. The sectoral coverage in each country/region includes twenty-three product categories covering agriculture and
30
D. Brown, A. Deardorff, S. Djankov, and R. Stern
manufacturing and six categories covering services, including government, all of which are modeled as tradable.11 The agricultural sector in each country is characterized as being perfectly competitive, and it is assumed that the products of this sector are differentiated according to the place of production. The manufacturing and services sectors in each country are characterized as being monopolistically competitive with free entry, and the products that are produced and traded are assumed to be differentiated by firm.12-13 Since the more advanced CECs produce relatively sophisticated consumer and producer goods, the assumption of product differentiation and monopolistic competition is more appropriate for them than others. Also, entry and exit are likely to be important parts of the transition process. The reference year for the data base of the model is 1992. The input-output relations used in the model refer to different years, depending on the availability of national input-output tables.14 More complete technical details, including a full statement and description of the equations and parameters of the model, are available from the authors on request. The 1992 base data for the three 1
' We have recently constructed a bilateral matrix of international trade in services for the thirtyfour countries in the model's data base so as to be able to treat all twenty-nine sectors as tradable and to analyze the effects and interaction of liberalization of both merchandise trade and services in our model countries/regions. For some preliminary analysis along these lines for the Uruguay Round negotiations, see Brown, Deardorff, Fox, and Stern (1995). 12 It is thus being assumed that there are constant returns to scale in the agricultural sector and increasing returns to scale in the manufacturing and services sectors. The assumption of national product differentiation for agriculture means that the so-called Armington (1969) assumption that each country produces its own differentiated product is being applied and that nations will have some degree of monopoly power in trade in this sector. For the manufacturing and services sectors, product differentiation byfirmdispenses with the Armington assumption so that the potentially strong terms-of-trade effects associated with national monopoly power will be greatly diminished. But, as will be noted below, the realization of increasing returns may result in substantial scale effects. 13 Issues of the modeling of market structure are discussed in Brown and Stern (1989), where a variety of different imperfectly competitive market structures are used in analyzing the economic effects of the CUSTA. For the current model, as noted, we use a structure of monopolistic competition, following Helpman and Krugman (1985), for all of the manufacturing and service industries. There is free entry of firms, each producing a different variety of a good and producing it with a fixed cost and constant marginal cost in terms of primary and intermediate inputs. Varieties enter via a Dixit-Stiglitz (1977) aggregation function into both utility and production functions, with the implication that greater variety reduces cost and increases utility. 14 It is always a problem to use completely up-to-date input-output tables because of ongoing changes in technology and productivity that would alter the input-output coefficients for particular sectors. This applies especially to the CEC countries, which have been undergoing considerable restructuring. In the absence of current input-output tables for these countries, we decided to use the 1980 input-output table for Portugal as a proxy for the economic structure of the individual CECs. Once we are able to obtain more appropriate input-output tables for the CEC countries, we will then be able to revise our model simulations to see what difference it makes in having used the Portugal table. There is some concern that particular sectors of the CEC economies, such as energy or services, may be particularly inefficient compared to their Portuguese counterparts. In the absence of more information on the extent of such differences, however, we have not attempted to correct for such possibilities.
An economic assessment of integration into the EU
31
CEC countries are provided in Tables 2.1-2.3. Data for the other countries and documentation for the model are also available from the authors.15 There are several important assumptions that either are built into the model or are implemented by the model for the present analysis. It is important that these be understood in interpreting the results to be reported below. Full employment. The analysis assumes throughout that the aggregate, or economy-wide, level of employment is held constant in each country. The EUCEC integration to be analyzed is therefore not permitted to change any country's overall rates of employment or unemployment. This assumption is made because overall employment is determined by macroeconomic forces and policies that are not contained in the model and are not themselves to be included in a negotiated agreement. The focus instead is on the composition of employment across sectors as determined by the microeconomic interactions of supply and demand with the sectoral trade policies that an EU-CEC agreement will alter. There are several ways of interpreting this key assumption, in light of substantial open and disguised unemployment in many of the CECs. One is to argue that the aggregate level of employment is affected by economic integration, and that the computational results to be reported below assume that aggregate unemployment is constant. Alternately, to the extent that economic integration does have effects on aggregate employment and unemployment (see Black and Moersch, Chapter 5), these effects are not taken into account in our analysis. In any case, we must assume that the existence of unemployment does not interfere with the functioning of the markets for goods and services as characterized in our model. (For the effects of a different assumption, see Bauer and Zimmermann, Chapter 7.) Balanced trade. It is assumed that trade remains balanced for each country, or more accurately that any initial trade imbalance remains constant, as trade barriers are changed with an EU-CEC agreement. This assumption is intended to reflect the reality of mostly flexible exchange rates among the countries involved. It also, like the full employment assumption, is appropriate as a way of abstracting from the macroeconomic forces and policies that are the main determinants of trade balances.16 (For some of the effects of changes in the balance of trade, see Black and Moersch, Chapter 5.) 15
16
The main data used cover trade, production, and employment, and these data come primarily from United Nations sources and to a lesser extent from national sources. The model parameters are constructed from the trade and input-output data for the countries included in the model and from published studies of trade and capital/labor substitution elasticities. For a comprehensive discussion of the data and parameters, see Deardorff and Stern (1990, pp. 37-45). The results reported below for changes in total exports and imports may appear to contradict this assumption of balanced trade. This is because what are reported are measures of the changes in quantities traded, which are relevant for output and employment changes. They are not the values of trade, which undergo additional change due to changing relative prices. It is the values of exports relative to imports that are held fixed by the balanced-trade assumption.
32
33
Table 2.1. Czechoslovakia: basic data, 1992 Sector 1 Agriculture 310 Food 321 Textiles 322 Clothing 323 Leather Products 324 Footwear 331 Wood Products 332 Furniture, Fixtures 341 Paper Products 342 Printing, Publishing 3SA Chemicals 35B Petroleum Products 355 Rubber Products 36A Nonmetal Min. Prod. 362 Glass Products 371 Iron, Steel 372 Nonferrous Metals 381 Metal Products 382 Nonelec. Machinery 383 Electrical Machinery 384 Transport Equipment 38A Misc. Mfrs. 2 Mining, Quarrying 4 Utilities 5 Construction 6 Wholesale Trade 7 Transportation 8 Financial Services 9 Personal Services Total
Source: CEC data base.
Output (Mill. $)
Labor (000)
Imports (Mill. $)
4,744.50 4,762.10 1,832.20 875.20 934.10 658.70 1,118.60 979.90 1,032.80 687.00 2,197.00 2,408.40 2,304.00 1,285.70 1,655.50 4,275.50 1,538.50 4,913.20 3,725.90 3,222.80 2,148.40 3,823.70 2,857.20 8,078.50 5,452.30 7,171.40 11,021.70 1,655.70 14,985.00
795.43 253.39 164.47 58.74 72.60 33.37 107.25 88.96 57.80 56.63 152.24 82.20 50.75 82.18 101.51 247.61 133.14 263.77 363.22 184.70 184.34 233.40 201.48 297.64 702.91 690.09 726.23 293.44 1,350.35
651.49 706.81 276.71 164.83 164.47 71.74 328.80 540.13 144.62 246.72 701.81 380.82 941.37 602.97 645.71 722.30 313.83 941.34 809.73 709.74 434.19 1,360.02 551.45 332.01 362.15 832.90 1,490.26 1,105.02 588.25
102 14S SO
8 029 84
17 m. 19
Exports (Mill. $) 925.17 562.90 459.75 235.75 238.42 149.62 405.98 327.16 278.68 214.74 718.66 518.16 672.22 396.95 359.79 1,238.96 253.03 1,384.72 424.36 378.21 233.66 1,042.25 296.18 1,249.37 194.00 875.33 1,857.81 143.33 578.92 1661408
35
34
Table 2.2. Hungary: basic data, 1992 Sector 1 Agriculture 310 Food 321 Textiles 322 Clothing 323 Leather Products 324 Footwear 331 Wood Products 332 Furniture, Fixtures 341 Paper Products 342 Printing, Publishing 35A Chemicals 35B Petroleum Products 355 Rubber Products 36A Nonmetal Min. Prod. 362 Glass Products 371 Iron, Steel 372 Nonferrous Metals 381 Metal Products 382 Nonelec. Machinery 383 Electrical Machinery 384 Transport Equipment 38AMisc.Mfrs. 2 Mining, Quarrying 4 Utilities 5 Construction 6 Wholesale Trade 7 Transportation 8 Financial Services 9 Personal Services Total Source: CEC data base.
Output (Mill. $)
Labor (000)
Imports (Mill. $)
Exports (Mill. $)
4,717.30 6,540.80 1,440.30 856.10 767.40 356.10 741.80 726.40 789.80 1,177.80 2,283.40 3,100.70 2,254.00 790.70 971.00 2,129.50 1,517.30 2,198.20 2,228.60 2,701.30 1,188.90 3,182.40 1,066.40 3,181.40 1,940.30 2,499.90 2,593.50 1,141.40 8,619.60 63.702.30
459.90 274.92 69.01 98.34 34.31 26.36 69.31 51.65 26.21 73.26 101.15 88.86 65.27 40.21 64.07 46.08 37.62 107.16 125.89 131.39 80.64 147.63 95.04 108.22 216.71 597.21 346.48 210.63 984.34 4.777.87
855.26 716.47 311.97 188.58 168.01 82.94 309.47 327.47 196.83 233.88 901.28 696.43 677.62 371.24 367.28 695.85 274.57 804.51 471.79 423.28 261.70 1,096.33 530.87 447.28 332.03 802.07 1,410.27 1,731.83 601.99 16.289.10
1,129.99 1,016.17 370.99 216.36 275.42 90.49 241.46 197.93 82.00 81.96 559.45 240.42 511.25 213.07 194.11 430.58 223.12 707.90 268.23 235.11 151.15 863.51 110.49 410.91 205.81 236.39 418.97 126.75 632.42 10.442.41
37
36
Table 2.3. Poland: basic data, 1992 Sector 1 Agriculture 310 Food 321 Textiles 322 Clothing 323 Leather Products 324 Footwear 331 Wood Products 332 Furniture, Fixtures 341 Paper Products 342 Printing, Publishing 35A Chemicals 35B Petroleum Products 355 Rubber Products 36A Nonmetal Min. Prod. 362 Glass Products 371 Iron, Steel 372 Nonferrous Metals 381 Metal Products 382 Nonelec. Machinery 383 Electrical Machinery 384 Transport Equipment 38AMisc.Mfrs. 2 Mining, Quarrying 4 Utilities 5 Construction 6 Wholesale Trade 7 Transportation 8 Financial Services 9 Personal Services Total Source: CEC data base.
Output (Mill. $)
Labor (000)
Imports (Mill. $)
Exports
12,403.90 9,319.90 1,735.30 947.20 964.60 856.60 1,793.40 1,544.10 829.20 442.50 1,952.50 4,381.50 5,881.50 893.10 596.50 3,863.80 1,844.70 3,931.50 1,355.10 2,222.90 904.20 3,176.60 4,260.80 5,096.60 7,949.10 9,198.30 11,475.80 4,343.60 31,796.50 35 961 Ifl
4,027.88 511.00 245.00 178.00 93.52 61.33 168.11 126.32 36.54 37.84 236.25 183.50 238.41 33.48 62.58 120.00 78.47 182.50 286.49 201.20 129.48 247.42 407.50 436.42 1,242.00 1,658.20 1,736.12 213.48 3,743.47
1,650.46 1,746.16 322.42 168.98 146.22 62.32 286.01 456.39 229.09 290.71 1,512.89 907.02 792.91 547.47 552.68 762.29 269.85 869.54 703.57 638.95 398.80 1,508.24 741.13 470.82 468.08 1,009.69 2,314.85 1,714.58 739.65 22.281.77
1,380.22 911.52 551.18 308.18 379.07 105.36 370.11 311.43 185.96 134.13 564.86 701.86 595.17 297.62 205.69 1,227.51 572.76 1,513.27 320.24 337.20 181.74 959.60 549.33 605.90 124.61 269.54 857.13 287.70 438.12 15.247.01
16 922 51
38
D. Brown, A. Deardorff, S. Djankov, and R. Stern
Fixed relative wages. While the economy-wide wage in each country is permitted to adjust so as to maintain full employment, the wages across sectors are held fixed relative to one another. This permits the analysis to focus on the labor-market adjustments that an EU-CEC agreement might require, independent of any relative wage changes that may facilitate those adjustments.17 (For some of the effects of changing relative wages and changing labor supply in the EU countries, see Bauer and Zimmermann, Chapter 7.) Fixed labor supply. The total labor supply in each country is assumed to be held fixed in the analysis. This is not to say that changes in labor supply will not occur in the course of a phase-in of an EU-CEC agreement, but only that such changes are assumed not to be the result of such an agreement.18 Role of variety. The Dixit-Stiglitz aggregation function in its usual form uses a single parameter, the elasticity of substitution, to determine both the degree of substitution among varieties of a good and the extent to which an increased number of varieties adds to welfare of consumers and reduces costs of intermediate inputs. This effect on welfare and cost could be quite important in an analysis of trade liberalization, since reduced trade barriers provide greater access to varieties produced abroad and could increase welfare on that account alone. In Section 2.9 below we will explore the sensitivity of our results to this effect of variety.19 The policy inputs into the model are the tariffs and nontariff barriers (NTBs) that are currently (as of the early 1990s) applied to the bilateral trade of the CEC countries and EU regions modeled explicitly with respect to each other and to the other two aggregated regions included in the model.20 As will be noted below, in order to investigate the sectoral employment effects of an EU-CEC preferential trade agreement, it will be assumed that the existing bilateral tariffs will be removed and NTBs will be relaxed all at one time rather than in stages. NTBs are assumed to be binding over the fraction of the industry indi-
17
18
19
20
In effect then, we do not distinguish workers according to their skill characteristics and therefore how the wages and employment of different skill groups may be affected by an EU-CEC arrangement. In Stern, Deardorff, and Brown (1992), the U.S. employment changes that might result from NAFTA were decomposed by sector, occupation, and geographic location. See Stern, Deardorff, and Brown (1992) for analysis of the cross-border movement of labor between the United States and Mexico that may occur as a result of NAFTA. In earlier work we have noticed that the effect of variety in lowering costs can introduce an instability into the model, because an increase in demand for an industry can lead to entry, additional variety, lower costs to users, and hence additional demand. To avoid this happening in our model, we depart slightly from the Dixit-Stiglitz formulation, using an additional parameter to control these variety effects. In the results reported here, the effect of variety on welfare has been set to one half of what would occur in the Dixit-Stiglitz model. The data on tariffs and NTBs will be discussed further below.
An economic assessment of integration into the EU
39
cated by our NTB trade coverage data and to generate rents in the importing countries only.21 When the policy changes are introduced into the model, the method of solution yields percentage changes in sectoral employment and other variables of interest for each country/region. Multiplying the percentage changes by the actual (1992) levels given in the data base yields the absolute changes, positive or negative, that might result if the bilateral tariffs/NTBs were removed all at one time. More realistically, of course, the removal of tariffs (and NTBs) in an EU-CEC agreement would almost certainly be phased in over a period of years. If information were available for the different phases, the model could in principle be solved sequentially, taking into account the barrier reductions in each time period. In addition to the sectoral effects that are the primary focus of our analysis, the model yields results for changes in total exports, total imports, the terms of trade, the overall level of welfare in the economy measured by the equivalent variation, and the economy-wide changes in real wages and real returns to capital. Because both labor and capital are assumed to be homogeneous and intersectorally mobile in these scenarios, we cannot distinguish effects on factor prices by sector. Nor, as noted above, can we distinguish effects on different skill groups or other categories of labor. In particular we are unable to address the important question of how an EU-CEC agreement might affect the differential between the wages of skilled and unskilled workers. (But see Bauer and Zimmermann, Chapter 7.) Although the bilateral removal of tariffs and NTBs constitute the main changes in trade policies that would occur with an EU-CEC agreement, there may be other changes as well. These relate especially to changes in FDI and to the cross-border movement of workers as the result of changes in the rate of return on capital and changes in real wages. Changes in FDI have indeed occurred, and more may occur in the future as a result of an agreement. However, the changes so far have been insignificant and would not alter the basic results of our model. The Europe Agreements do not allow for free movement of labor between the EU and the CECs, and thus we abstract from migration issues.22 Also, as already noted, we do not make any allowance for dynamic efficiency changes or capital accumulation. Data on tariffs and nontariff barriers Average tariff levels and averages of the tariff equivalents of NTBs applying to the trade of the individual countries/regions in the model are summarized in Ta21
22
NTBs are assumed to apply at the level of the industry, not the firm, so that the pricing behavior of perfectly competitive and monopolistically competitive firms is unaffected. We recognize, however, that the large differences in per capita income that exist between the CECs and the EU create great incentives for such movement, and that reduction of trade barriers and increased economic interaction may well cause migration to increase.
Table 2.4a. Average tariff rates for CECs, European Union, NAFTA, and other trading nations, 1992 (percentage) Importing Country
Exporting Country Hungary
Poland
EFTA
EU-3
EU-9
NAFTA
OTH
Czechoslovakia
0.0
7.1
7.1
5.2
6.5
6.5
3.6
5.8
5.2
Hungary
8.1
0.0
7.6
6.3
8.5
8.5
6.3
7.6
6.9
Poland
7.5
7.9
0.0
6.9
7.1
7.1
61
8.3
8.3
EFTA
7.6
6.7
5.7
0.0
0.0
0.0
4.3
5.7
7.2
EU-3
6.4
6.2
6.9
0.0
0.0
0.0
3.5
4.5
4.4
EU-9
6.4
6.2
6.9
0.0
0.0
0.0
3.5
4.5
4.4
NAFTA
8.2
7.8
7.6
3.8
3.6
3.6
0.0
4.4
4.4
17.9
17.0
18.3
10.3
13.4
13.4
8.8
9.3
6.9
Czechoslovakia
Other
ROW
Source: CEC data base. The bilateral tariffs are own-country, import-weighted averages of pre-Uruguay Round MFN tariff rates, using 1992 imports for weighting purposes. The tariff rates are from the Uruguay Round Model data base; see Brown, Deardorff, Fox, and Stern (1995) for further details.
Table 2.4b. Average tariff equivalents ofNTBsfor CECs and European Union, 1992 (percentage) Exporting Country
Importing Country
Czechoslovakia Hungary Poland NAFTA EFTA EU-3 EU-9 OTH Czechoslovakia
0.0
1.0
1.1
3.6
4.4
3.9
3.9
2.6
Hungary
1.8
0.0
1.7
1.5
2.9
2.6
2.6
1.1
Poland
0.4
1.2
0.0
2.4
3
2.4
2.4
1.8
EFTA
9.9
9.9
9.9
4.8
0.0
0.0
0.0
2.4
EU-3
11.5
11.4
11.9
0.0
0.0
0.0
1.3
EU-9
11.3
11.4
11.9
0.0
0.0
0.0
2.4
Source: Simple averages of tariff equivalents obtained from a variety of sources. See text.
An economic assessment of integration into the EU
41
bles 2.4a and 2.4b. There we report the import-weighted average tariffs and simple average tariff equivalents applying to each pair of bilateral trade flows for the individual countries/regions of the model. The sectoral tariff rates and tariff equivalents are available from the authors on request. The bilateral tariffs were constructed by weighting the pre-Uruguay Round, Most-Favored Nation (MFN) line-item tariffs by bilateral imports so as to calculate the tariffs that each country/region applied bilaterally to its trading partners. Information on NTBs was collected in two forms. First, the percentage of trade subject to NTBs was calculated, based primarily on the NTB inventory data assembled by the United Nations Conference on Trade and Development (UNCTAD). These NTB measures are calculated by first making an inventory of existing NTBs classified by disaggregated import groupings, then determining the value of imports that are subject to any NTBs, and thereafter aggregating up to the sectoral level used in the model. Thus, a sector with a zero percent NTB trade coverage is taken to be completely exempt from NTBs, whereas, say, an NTB coverage of 25 percent is taken to mean that 25 percent of the imports in that sector are subject to one or more NTBs. It is important to emphasize that these measures of NTB trade coverage are not the same as the tariff equivalents of the NTBs. Nor are they used in the model in the same way as tariffs and tariff equivalents. Rather, they are used to dampen the quantity responses of sectoral imports, in response to changes in prices and other variables, below what would have occurred otherwise. For further discussion, see Deardorff and Stern (1990, pp. 23-25). Tariff equivalents of NTBs were also assembled for the countries/regions of the model from a variety of sources. The principal source was Rollo and Smith (1993), who focused on the Common Agricultural Policy but calculated and reported tariff equivalents for a variety of sectors, including especially agriculture and food products. Tariff equivalents on steel came from Winters (1994) and for textiles from Halpern (1994). All of these estimates were based on price comparisons. We also include some small tariff equivalents on trade among the CEC countries, taken from Bakos (1993).23 2.7
Computational results: aggregate effects The scenarios
It is possible to use our CGE model to analyze various combinations of country/region membership in a preferential trade agreement. What we did in our 23
There also exist some NTBs with small positive tariff equivalents focused on agriculture between EFTA and the EU that are not included here. These resulted from the exclusion of the EFTA countries from the EU's Common Agricultural Policy.
42
D. Brown, A. Deardorff, S. Djankov, and R. Stern
first scenario was to model the free-trade agreement that the three CECs had already formed among themselves as the CEFTA. Second, we looked at the effects of complete tariff removal among the three CECs and the EU. In this scenario, existing NTBs were assumed to remain in effect. Their role in the model is to limit the responsiveness of trade to prices, this limitation being in proportion to the data on the trade coverage of NTBs. One question that therefore arises is how to handle any reduction or elimination of these NTBs that might occur between the CEC and the EU. Thus in our base case scenario we looked at a preferential arrangement involving tariff elimination among the three CEC countries and the EU, plus allowing tariff equivalents of NTBs to be removed, except in sensitive sectors (agriculture, textiles and apparel, chemicals, and iron and steel), where NTBs remained intact. In our fourth scenario, we examined a hypothetical and more extreme case of liberalization in which the tariff equivalents of all NTBs were removed. The various scenarios that we ran were therefore as follows: A. CEFTA. Bilateral removal of all tariffs on trade among Czechoslovakia, Hungary, and Poland. Bilateral NTB trade coverage ratios set to zero for the three CECs. B. CEC-EU Free Trade Agreement. CEFTA plus bilateral removal of all tariffs between all three CECs and the EU. CEC-EU bilateral NTBs remain in place. C. CEC-EU Free Trade Agreement plus Removal of Nonsensitive NTBs. Same as scenario B, but also eliminating tariff equivalents of CEC-EU bilateral NTBs on all trade except EU imports of agriculture, textiles and apparel, chemicals, and iron and steel, where NTBs are assumed unchanged. This is our Base Case, coming the closest we can to the actual Europe Agreements. D. CEC-EU Free Trade Agreement plus Removal of All NTBs. Same as scenario B, but also eliminating tariff equivalents of CEC-EU bilateral NTBs on all trade. This corresponds most closely to the concept of "full membership" of the CECs in the EU, except for continued exclusion of agriculture from the CAR An overview of results on trade, terms of trade, welfare, and factor payments for each of the scenarios is reported in Table 2.5. Perhaps the single most important number to consider in evaluating EU-CEC integration is the impact on welfare, that is, the "equivalent variation" measure of the change in real gross domestic product (GDP).
An economic assessment of integration into the EU
43
Economic welfare In scenario A, the Central European Free Trade Agreement (CEFTA) can be seen to increase economic welfare in the three CECs to a small extent and to reduce welfare in all other regions but by an insignificant amount. With an EU-CEC FTA in scenario B, the welfare of the three CECs is increased, while the welfare gains are noticeable for the three EU regions, although as percentages of GDP the EU gains are still quite small. When the existing NTBs in nonsensitive sectors are eliminated in scenario C, the welfare effects for the three CECs and the EU regions are larger still. Removal also of sensitive-sector NTBs in scenario D creates the greatest gains for the CEC countries, and increases slightly the welfare gains for the EU. The negative effects for NAFTA and the Other Major Trading Nations persist but are minor throughout. Among the three EU regions, while there are small gains from integration with the CEC countries in all, these gains are noticeably smaller in the EU-South region (Greece, Portugal, and Spain) than in EU-North and EUEFTA. This reflects the fact that the EU-South includes the poorest parts of the EU, which therefore are likely to compete most directly with the CEC countries in the EU markets for especially labor-intensive goods. It should be noted that positive welfare gains are not inevitable when trade is liberalized on a preferential basis, although the presumption that each country will gain from joining an EU-CEC FTA is strong. Several different forces are at work determining the welfare effects of trade liberalization. On the positive side, consumers are free to choose the least expensive source of goods from countries within the FTA. In addition, by expanding trading opportunities, each country has the option of specializing production in the range of goods in which it has a comparative advantage. There are three other forces, however, that have an ambiguous effect on welfare. First, consumers are not able to choose freely among all foreign sources of goods because tariffs and NTBs are removed only on included partners. Hence, consumption choices may be distorted by the preferential nature of the trade liberalization. Second, a country's terms of trade could improve or deteriorate as a result of trade liberalization. If import prices rise and export prices fall, welfare gains stemming from specialization and exchange could be reversed. However, in most cases we expect that the terms-of-trade effects following liberalization by a small country will be too small to reverse other sources of gain. This is the case with an EU-CEC FTA, as can be seen from column (4) of Table 2.5. There tend to be negative terms-of-trade effects for the CECs as long as sensitivesector NTBs are excluded from the liberalization, but these negative effects are relatively small and have not led to a net fall in welfare, as seen in columns (5) and (6). It is interesting that removal of sensitive sector NTBs in the hypothetical scenario D is enough to make the terms-of-trade effects positive for the
Table 2.5. Summary results ofCEC and CEC-EU integration: changes in country imports, exports, terms of trade, welfare, and real returns to labor and capital Country/Region
Imports
Exports
$ Mill.
$ Mill.
(2)
(3)
(4)
198.2 117.0 168.6 -0.3 2.7 2.6 24.4 2.1
Terms of Trade
Wage Rate
Ret to Capital
Pet. Change
Pet. Change
(5)
$ Millions (6)
(7)
(8)
0.0 -0.1 -0.1 0.0 0.0 0.0 0.0 0.0
0.6 0.4 0.4 0.0 0.0 0.0 0.0 0.0
330.8 152.1 343.7 -20.7 -31.4 -8.9 -48.9 -28.9
0.4 0.3 0.3 0.0 0.0 0.0 0.0 0.0
-0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0
-0.6 -0.5 -0.9 0.0 0.0 0.0 0.0 0.0
4.0 4.0 3.1 0.0 0.2 0.0 0.1 0.0
2,227.8 1,424.2 2,491.9 -277.1 2,036.0 283.6 6,604.0 -332.1
3.3 2.7 2.9 0.0 0.1 0.0 0.0 0.0
-0.1 0.4 -0.1 0.0 0.0 0.0 0.0 0.0
2,076.7 1,616.0 2,443.9 -8.0 1,100.3 158.7 3,927.4 -10.6
-0.4 -0.2 -0.5 0.0 0.0 0.0 0.0 0.0
4.5 4.7 3.8 0.0 0.3 0.0 0.1 0.0
2,513.9 1,662.7 3,025.8 -377.6 2,352.4 321.5 8,032.2 -450.8
3.6 3.0 3.3 0.0 0.1 0.0 0.0 0.0
-0.3 0.1 -0.4 0.0 0.0 0.0 0.0 0.0
3,181.9 2,689.3 4,057.7 73.9 1,882.6 295.6 6,826.7 27.2
1.7 1.3 0.0 0.0 0.8 0.0 0.0
7.3 6.8 5.6 0.0 0.5 0.1 0.2 0.0
4,128.4 2,437.4 4,403.3 -933.0 4,224.3 1,035.8 16,115.1 -591.7
5.5 2.9 4.3 0.0 0.1 0.1 0.1 0.0
-2.1 -1.7 -2.8 0.0 -0.1 0.0 0.0 0.0
Equivalent Variation
Pet. Change Pet. Change
A. Intra-CEC Free Trade Area (CEFTA) Czechoslovakia Hungary Poland NAFTA EFTA EU-South EU-North Other
199.5 106.7 147.2 1.0 8.8 4.2 46 5.1
B. CEC-EU Free Trade Area, NTBs in Place Czechoslovakia Hungary Poland NAFTA EFTA EU-South EU-North Other
1,772.9 1,356.2 1,985.3 -10.9 1,096.1 137.4 3,612.9 23.1
1,876.3 1,404.6 2,114.4 -10.5 970.9 140.9 3,370.9 -0.7
C. CEC-EU Free Trade Area, Nonsensitive NTBs Removed Czechoslovakia Hungary Poland NAFTA EFTA EU-South EU-North Other
2,008.1 1,596.8 2,368.2 -8.5 1,230.3 151.3 4,061.5 15.6
D. CEC-EU FTA, All NTBs Removed Czechoslovakia Hungary Poland NAFTA EFTA EU-South EU-North Other
3,307.8 2,858.9 4,242.4 69.2 2,007.8 248.1 6,362.3 98.1
46
D. Brown, A. Deardorff, S. Djankov, and R. Stern
CEC countries, indicating that it is precisely in these sensitive sectors that the CEC countries have a comparative advantage. A third force determining the welfare effects of trade liberalization concerns the realization of economies of scale and pricing by imperfectly competitive firms. Tariff liberalization is expected to have a pro-competitive effect on import-competing firms in each country. Without tariff protection, domestic firms feel competitive pressure from imports and may charge a lower price in order to compete. This in turn causes expansion of output per firm and a lowering of average cost. In their export markets this lowering of cost and price contributes to a worsening of the country's terms of trade, although the net effect on welfare is almost certainly positive. In industries where there are significant economies of scale and, thus, declining average costs, the firm that charges a lower price may also have to increase output in order to break even. As the firm moves down its average total cost curve, the inputs required to produce a unit of output decline on average. If many of the firms in a country are forced by competitive pressure to economize on inputs in this way, then the country overall will be able to produce more than before the liberalization using the same inputs and technology. This gain from the realization of economies of scale enhances the more traditional gains from specialization and exchange. However, scale gains, while likely, are not inevitable. Trade liberalization is pro-competitive for import-competing firms. However, curiously, export firms experience an anti-competitive effect. As the trade partner lowers its tariffs, export firms now have easier access to foreign markets and, therefore, compete less vigorously at home. Such firms may respond by raising price and cutting back production, with adverse consequences for the economy overall. Scale economies will be discussed in more detail below. However, we find that for most countries, firm output tends to rise, so that scale gains are generally positive though not large. The complications that are introduced by the considerations just mentioned - trade diversion, terms of trade effects, and effects of changing competition, scale, and variety - could in principle dominate the results of the analysis. This does not appear to be the case, however, in our results. All the participants in the free-trade areas that we model here do experience increases in welfare, and this continues to be the case even when we take out the effects of variety and scale in two additional scenarios that we will report below. This is consistent with what we have found elsewhere in similar work on the NAFTA and other preferential trading arrangements. It appears that the fundamental driving force behind the welfare effects of such preferential arrangements is still the improvements in economic efficiency that are the basis for the classical gains from trade.24 24
We are indebted to J. David Richardson for noting and stressing this point.
An economic assessment of integration into the EU
47
Real wages and return to capital Having established that the welfare effects of an EU-CEC FTA are positive for all participants, we next turn to the distributional consequences. In particular, we are interested in which factors of production are likely to gain and which to lose with formation of an EU-CEC FTA. For this purpose, we have calculated changes in factor prices deflated by changes in a price index for consumption. These measures understate the true changes in real factor prices because the deflator is effectively a Laspeyres price index and does not take full account of the efficiency gains due to trade liberalization. For this reason, our changes in welfare reported above, which are based directly on the assumed utility functions of consumers, can report greater improvements in welfare than our reported increases in real factor prices. Since the size of this downward bias should be the same for both factors, however, the relative changes in factor prices are nonetheless accurate. The percent changes in the real returns to labor and capital are reported in columns (6) and (7) of Table 2.5. In most cases, real wages in the CEC countries rise and real returns to capital fall, with negligible changes in other countries/regions of the model. These changes are therefore consistent with what one would expect from the Stolper-Samuelson theorem of the Heckscher-Ohlin trade model. Presumably the CEC countries are relatively well endowed with labor, compared to capital, vis-a-vis the more advanced economies of Europe with which their trade is being liberalized. The Stolper-Samuelson theorem predicts a rise in the real return to the abundant factor and a fall in the real return to the scarce factor, exactly as found by the model for Czechoslovakia and Poland. However, the model includes features that are not part of the HeckscherOhlin framework and that can cause both real factor returns to rise. We have found this to happen repeatedly in other applications of the model, and we find it here, too, for Hungary and in scenarios B and C. That is, when the CEC countries and Europe reduce only tariffs against each other, and also to a lesser extent when the tariff removal is accompanied by removal of nonsensitive NTBs, Hungary experiences a rise in both its real wage and its real return to capital. This is possible because, in the context of differentiated-products model with increasing returns to scale, like the one used for this study, other forces may be at work undermining Stolper-Samuelson-type mechanics.25 Note, however, that in both of these cases labor gains more than capital, and this is consistent with Stolper-Samuelson. Scale effects work very much like the relative price effects articulated in the Stolper-Samuelson theorem to determine the implications of trade liberaliza25
For a further discussion of factor prices in a differentiated products model, see Brown, Deardorff, and Stern (1993).
48
D. Brown, A. Deardorff, S. Djankov, and R. Stern
tion for factor prices. Scale effects, like price effects, tend to accrue to one factor only. For example, it can be shown that an increase in output per firm in an industry raises the real return to the factor used intensively in that industry and lowers the return to the other factor. Price and scale effects differ, however, in one important regard. If scale gains emerge across the board in nearly all industries, then both factors may gain. This is apparently the case in our model.26 We turn next to consider the sectoral results. 2.8
Computational results: sectoral effects
The sectoral results for the three CECs for our Base Case Scenario C are given in Tables 2.6-2.8. For each country/region, the percent changes in total exports and imports are reported in columns (2) and (3). Imports are decomposed by trade partner in columns (4)-(10). The percent changes in industry output and numbers of firms are listed in columns (11) and (12). The percent changes in output per firm, which can be used to determine the extent to which economies of scale may be realized, are calculated by subtracting column (12) from column (11). Finally, the percent and absolute changes in employment are listed in columns (13) and (14). The results for the three EU regions, NAFTA, and the Other Major Trading Countries, as well as for all countries/regions in the other scenarios, are available from the authors on request. An EU-CEC FTA has substantial sectoral impacts on the three CECs, as is evident in Tables 2.6-2.8. For Czechoslovakia in Table 2.6, output increases in twenty-six of the twenty-nine sectors, the exceptions being clothing and two of the services sectors. The largest absolute employment increases in Czechoslovakia are in agriculture, leather, metal products, and mining. For Hungary in Table 2.7, there is slightly more specialization, with output expanding in twenty-four of the twenty-nine sectors. The largest increases in Hungarian employment are in agriculture, food, and leather products. For Poland in Table 2.8, output expands in twenty-six of the twenty-nine sectors. There is a sizable increase in employment in the Polish agricultural sector, leather products, and metal products. In all three CEC countries, there are rather large absolute (but small percentage) declines in employment in community, personal, and social services, as well as smaller reductions in most other services, which lose employment to those sectors where trade barriers are being reduced. Comparison of columns (11) and (12) for all three countries suggests that there are positive scale effects across all the manufacturing and services sectors, reflecting especially the increased competition in larger markets and the consequent increase in elasticity of demand faced by firms. This induces surviving firms to expand and thus lowers their average costs. In addition, a technical feature of our assumed production structure also contributes to this result. 26
As already mentioned, we do not distinguish workers according to skill groups, so that we cannot determine if skilled and unskilled workers will be affected differently.
49 Table 2.6. Scenario C: CEC-EUfree-trade area, tariffs and nonsensitive NTBs, sectoral effects on Czechoslovakia of CEC-EU integration (percentage change) Sector
1 Agriculture 310 Food 321 Textiles 322 Clothing 323 Leather Products 324 Footwear 331 WoodJ?roducts 332 Furniture, Fixtures 341 Paper Products 342 Printing, Publishing 3SA Chemicals 35B Petroleum Products 355 Rubber Products 36A Noranetal Min. Prod. 362 Glass Products 371 Iron, Steel 372 Nonferrous Metals 381 Metal Products 382 Nonelec. Machinery 383 Electrical Machinery 384 Transport Equipment 38AMisc.Mfts. 2 Mining, Quarrying 4 Utilities 5 Construction 6 Wholesale Trade 7 Transportation 8 Financial Services 9 Personal Services Total
Exports
Imports
(2) 6.4 21.3 9.1 4.1 39.9 48.5 19.5 11.9 31.6 15.7 13.6 29.0 17.8 27.8 25.0 2.0 23.7 21.5 14.6 17.5 19.5 20.0 37.7 6.5 3.1 -5.7 -0.1 -4.1 -6.3 12 5
(3) 15.2 17.5 14.1 24.1 10.4 13.5 18.6 16.1 11.5 17.8 7.7 12.7 11.6 9.6 16.0 13.6 19.6 13.6 19.2 19.0 16.1 14.1 0.3 1.7 1.1 7.9 4.1 6.5 8.2 11 7
Hun. (4) 30.4 38.7 21.9 20.0 36.0 29.6 25.4 12.8 14.5 4.3 18.2 9.4 20.1 11.8 12.9 17.1 17.7 14.7 17.7 14.2 18.5 23.1 19.1 1.2 3.2 0.1 2.3 -1.1 1.2 16 7
Pol. (5) 32.4 29.5 18.4 11.8 27.8 21.4 21.5 8.6 22.0 3.5 5.4 22.1 17.2 22.7 23.4 22.6 24.9 21.5 22.2 18.2 17.7 25.1 16.6 2.1 5.3 0.9 3.1 -0.3 0.6 IS?
NAFTA (6) 5.5 2.7 1.2 8.8 -7.2 -7.1 -3.7 0.7 -6.6 -0.3 -1.2 -7.7 -3.5 -8.3 -4.0 1.6 2.4 -2.2 3.0 1.0 -0.3 -3.4 -4.2 0.2 1.0 8.4 4.3 6.4 9.1
Bilateral Imports EFTA (7) 13.7 26.7 15.1 29.9 13.7 15.0 22.2 18.7 17.3 19.4 9.9 13.9 15.3 10.7 19.0 12.5 21.5 14.2 21.7 21.9 20.2 17.8 12.5 3.5 0.8 8.2 4.1 6.7 8.9 13
EU-3 (8) 16.8 18.0 16.5 27.4 12.6 15.4 21.9 18.5 10.0 21.5 8.1 11.9 13.8 12.0 18.9 14.2 21.0 15.3 21.6 21.7 18.7 15.7 14.7 3.7 1.0 8.4 4.3 6.9 9.1 in
EU-9 (9) 16.8 18.0 16.6 27.6 12.3 15.4 22.0 18.6 10.0 21.6 8.2 11.9 13.7 12.0 18.9 14.2 21.0 15.3 21.7 21.8 18.8 15.8 14.6 3.7 1.0 8.3 4.3 6.8 9.0 n 9
Output
No. Finns
Oth. (10) 5.5 2.7 1.2 8.8 -7.2 -7.1 -3.7 0.7 -6.6 -0.3 -1.2 -7.7 -3.5 -8.3 -4.0 1.6 2.4 -2.2 3.0 1.0 -0.3 -3.4 -4.2 0.2 1.0 8.4 4.3 6.4 9.1
(11) 1.3 2.8 4.1 -2.8 16.8 13.0 5.1 0.9 11.3 1.5 6.1 8.3 5.1 9.1 4.8 2.5 4.1 6.6 1.5 2.2 3.4 5.7 10.5 5.4 3.9 -0.3 2.6 0.3 -0.2
(12) 0.0 -0.8 -0.7 -2.8 10.6 8.0
01
M
0.9
-2.2 5.0 -1.8 0.9 0.6 -0.5 2.4 -0.9 -2.9 1.0 1.9 -2.4 -2.6 -1.9 0.2 3.5 -0.7 -1.7 -2.7 -1.7 -2.7 -1.9 -0 4
Chance in Employment Percent 1000s (14) (B) 1.4 11.5 -0.2 -0.1 0.2 0.3 -1.6 -2.7 7.9 10.9 9.6 3.2 1.3 1.2 -1.0 -0.9 5.9 3.4 -0.1 -0.2 2.7 1.8 -9.0 -7.4 0.4 0.7 3.1 3.8 0.1 0.1 -4.3 -1.7 3.5 2.6 8.3 3.1 -4.0 -1.1 -2.9 -1.5 -2.2 -1.2 2.9 1.2 9.6 4.8 0.0 0.0 -5.6 -0.8 -8.7 -1.3 -2.4 -0.3 -5.9 -2.0 -12.2 -0.9 ftO
no
50
Table 2.7. Scenario C: CEC-EU free-trade area, tariffs and nonsensitive NTBs, sectoral effects on Hungary of CEC-EU integration (percentage change) Sector
1 Agriculture
Exports
Imports
Output
No. Firms
Czech.
Pol.
NAFTA
EFTA
EU-3
EU-9
Oth.
Percent
1000s
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(U)
(12)
(13)
(14)
8.9
14.9
21.4
21.1
4.4
30.9
30.3
30.3
4.4
2.3
0.0
2.7
12.6
37.4
-0.5
3.0
Bilateral Imports
Change in Employment
310 Food
28.2
15.3
20.1
22 A
-0.4
34.2
37.5
5.0
1.5
321 Textiles
11.6
13.9
19.7
18.9
-1.7
16.9
18.8
18.9
-1.7
4.5
-0.2
0.8
0.6
322 Clothing
3.8
16.9
14.7
4.7
6.6
19.3
20.1
20.2
6.6
-2.9
-2.9
-2.7
-2.6
323 Lealher Products
59.5
-1.2
18.8
36.3
-17.1
0.5
0.7
0.4
-17.1
33.1
26.2
26.9
9.2
324 Footwear
52.3
9.1
22.3
18.1
-12.2
12.9
12.3
12.3
-12.2
14.8
9.3
11.5
3.0
331 Wood Products
23.3
9.5
11.9
9.0
-4.6
11.8
13.7
13.8
-4.6
7.0
3.3
3.9
2.7
332 Furniture, Fixtures
13.4
14.7
14.5
10.8
-0.2
19.0
19.8
19.9
-0.2
-2.0
-0.8
341 Paper Products
23.1
12.6
14.1
7.1
-0.4
16.3
16.0
16.0
-0.4
2.9
-1.5
-0.1
0.0
342 Printing, Publishing
13.4
16.0
13.0
14.3
4.7
19.7
19.8
19.8
4.7
-0.8
-3.0
-1.6
-1.2
3SA Chemicals
12.5
12.8
24.5
16.8
-1.5
21.0
20.0
20.1
-1.5
3.4
-1.1
-0.1
-0.1
35B Petroleum Products
21.7
7.3
27.0
23.5
-1.8
15.9
15.7
15.6
-1.8
3.9
-1.5
-7.9
-7.0
355 Rubber Products
24.0
11.9
14.9
17.5
-4.8
16.4
16.4
16.3
-4.8
5.8
0.7
2.4
1.6
36A Nonmetal Min. Prod.
28.7
7.6
19.3
17.7
-6.5
8.3
12.1
12.0
-6.5
8.1
3.1
4.6
1.8
362 Glass Products
28.6
13.3
17.8
23.2
-3.9
15.8
18.9
18.9
-3.9
5.3
1.0
2.0
1.3
0.6
8.2
-0.4
0.9
7.7
11.6
10.5
1.7
11.9
10.4
10.5
1.7
1.4
-2.5
-1.5
-0.7
372 Nonfenous Metals
20.7
15.5
16.1
16.4
2.4
22.3
23.5
23.6
2.4
3.2
0.9
2.4
0.9
381 Metal Products
22.1
12.3
20.1
25.1
-4.0
15.8
17.3
17.3
-4.0
5.5
2.0
3.2
3.4
382 Nonelec. Machinery
13.9
19.5
21.7
23.7
2.1
23.6
25.2
252
2.1
0,1
-2.8
-1.7
-2.1
383 Electrical Machinery
18.9
18.3
23.7
16.9
1.1
23.9
22.6
22.6
1.1
1.2
-2.5
-1.5
-2.0
384 Transport Equipment
23.2
12.4
27.1
27.7
-2.1
15.1
18.0
18.0
-2.!
3.6
-0.9
-0.2
-0.2
38AMisc.Mfrs.
2i.7
13.2
19.0
18.5
-4.8
17.1
16.8
16.8
-4.8
6.5
1.6
2.9
4.2
371 Iron, Steel
2.7
30.9
4.0
33.8
25.6
-2.3
20.6
20.9
20.7
-2.3
6.4
1.4
2.9
4 Utilities
2.5
4.5
7.1
4.0
2.0
5.1
3.9
3.8
2.0
3.0
-12
0.1
0.1
5 Construction
2.2
0.4
3.5
4.8
0.5
0.3
0.5
0.5
0.5
2.6
-2.0
-1.0
-2.1
6 Wholesale Trade
-5.8
5.6
-0.9
-0.6
7.0
6.8
7.0
6.9
7.0
-1.8
-3.8
-1.5
-9.1
7 Transportation
-1.1
4.0
3.1
3.0
4.2
4.0
4.3
4.2
4.2
1.7
-1.3
-0.1
8 Financial Services
-6.1
5.7
-0.1
-1.1
5.5
5.9
6.0
6.0
5.5
-2.5
-4.8
-3.6
-7.6
9 Personal Services Total
-6.4
5.0
-1.9
-1.3
7.0
10 5
7.1 10 2
7.1 14 4
7.1
100
7.1 -0 4
-1.3 7
v
Textiles
' • • • •
^
^
• • • • • • • • • • •
^
'"""••••••••
"""^^ ^V^o^ Sign. ** **
ISIC
Coeff.
ISIC
Coeff.
332
1.57
384
-0.45
-5.0
T-Test
Sign. ** ••
372
0.91
11.8
**
324
0.82
5.0
323
1.57
390
-0.46
-3.8
383
0.87
11.5
•*
332
0.73
4.7
*•
324
1.54
352
-0.49
-6.4
••
382
0.87
15.0
••
323
0.72
3.8
••
322
1.49
351
-0.53
-6.5
**
385
0.85
13.5
•*
390
0.61
5.8
••
361
1.40
361
-0.54
-4.6
**
342
0.85
10.0
••
331
0.58
4.4
•*
331
1.35
382
-0.55
-6.6
**
332
0.85
6.8
••
361
0.56
5.4
••
342
1.24
383
-0.59
-5.5
**
**
323
0.84
5.4
**
356
0.47
5.7
**
362
1.21
384
-0.67
-6.7
361
0.84
10.1
**
321
0.40
2.9
•*
390
1.21
369
-0.67
-6.6
**
351
0.84
14.5
••
362
0.39
5.3
*•
372
1.20
362
-0.68
-7.9
•*
o
o
o
m
o
n
c
o
CO
o
r CO
o
n
o
o
o
o
o
©
o CO
CO
o
m
~
O CO
^
r CO
~
CO
o
-'
en
o
o
~
-'
-'
( N en m
o
CO'
© ' © '
o
vq
o
m
o
CO
o
a m
\
o
o
N n en
CO
< en
CO
o
N
>
)
o
© ' © " © '
$3
r ( N t N n m c n c N c n e n c n c n c n c n e n e n
o CO
CO
m
o
n
CO
CO
o
o
o
c N u m m
o
o
n T j o o m en rn
o
n en
CO
o
csj r*; en oq rn co c^ vq cs rr TJON ON rn csj c n o p t ^ o o o o o ^ t ^ o p c K v S s o v o v S o o r ^
o
o
o
o
_.i
o CO
o
s o n m en
o
CO
o
vq
o
o
CO
o
m
o
m
t
o
c
< N o o t n c n e n e n c n
t>«
f CO
o
o m
I
os
9
m"
00 en O
p
9
^-3
o
t>;
r-«
o
oo
•^-
en
O
0
en
0
C
en
N
0
9
o
r-#
r->
en
0
0
9
o
t^
^-
en
0
0
as o
ON
n
fa
o
t>-
en
»
en
0
o
sq
^
< >
en
O
s o » - ' > n T t ^ - i c N » n e n c N e n ' - H e n
^ •-*
0.25
CN
o
en
0.71
323
2.13
2.9
**
342
0.62
1.5
384
2.72
322
-0.60
-1.2
351
1.83
3.0
••
352
0.34
0.6
381
2.35
362
-0.67
-0.6
352
1.76
3.5
**
332
0.22
0.3
342
2.12
383
-0.73
-2.3
*
332
1.72
2.8
**
321
0.21
0.4
352
2.10
382
-0.75
-3.4
**
390
1.69
3.3
**
390
0.18
0.3
383
2.08
341
-0.86
-0.8
381
1.62
4.5
**
384
0.18
0.3
382
2.05
372
-0.92
-0.8
342
1.50
4.6
**
322
0.07
0.2
385
1.99
381
-0.92
-1.8
321
1.42
3.3
**
372
0.07
0.1
332
1.94
342
-0.98
-2.1
383
1.33
6.0
**
369
-0.11
-0.1
390
1.88
384
-1.15
-1.9
31
1.31
3.1
••
31
-0.12
-0.2
331
1.77
331
-1.28
-1.1
382
1.28
8.4
**
371
-0.53
-0.5
321
1.63
369
-1.31
-1.2
385
1.15
6.4
**
361
-0.62
-0.6
31
1.19
371
-1.91
-1.6
322
1.09
3.2
**
331
-1.41
-1.3
322
1.16
353/4
-4.56
-3.3
Note: ** indicates significance at 99% level, * indicates significance at 95% level.
*
**
Table 4.11. Basic data of CEECs Poland
CSFR
Hungary
Romania
Bulgaria
Russia
GNP per capita (US-$) 1988 CIA
7270
10140
8660
5490
7510
10984
1988 CSFBa
2000
3500
3000
1000
1500
2065
1989 World Bank
1890
3450
2630
1730
2780
1992 World Bank
1910
2280
2970
1130
1330
2510
GNP (US-$ billion)
73.3
35.6
30.6
25.7
11.3
374.0
Population 1992 (millions)
38.7
15.6
10.3
22.7
8.5
149
Frankfurt a.M.
539
222
474
868
899
1226
Berlin
343
212
434
830
858
1019
Distance (miles)* between capital and
a
Credit Swiss First Boston. Calculated as the shortest geographical distance between the relevant towns according to their degrees of latitude and longitude. Sources: World Bank, World Tables 1994, Washington (D.C.) 1994; World Bank, World Development Report, various issues; LQsch and Wohlers (1994): 151; Wang and Winters (1991): 45; own calculations. b
Impact on German trade of division of labor
145
level of GNP of more developed countries such as Ireland, Spain, Israel, Hong Kong, or Singapore by the end of the 1980s. Our estimates for the level of German trade with CEECs were initially based on 1992 GNP levels in CEECs. The results reflect the production level of West Germany in 1989, the base year for all regression calculations (Variant I). For a comparison with the actual value of foreign trade, one has to "inflate" the results to the 1992 level for the whole of Germany. Therefore, the estimated values derived from the regression equations for Germany were increased by 51 percent on the import side and 26 percent on the export side (Variant la). This corresponds to the change in total imports and exports respectively in current US-dollars for the whole of Germany in comparison to the West German level in 1989. The updated estimates derived from the regression equations for OECD countries were determined using the level of income for the whole of Germany in 1992 (Variant Ib); in 1992 the total German GNP in current USdollars was 66 percent higher than the West German level in 1989, and the 1992 per capita income for the whole of Germany was about 18 percent higher than the West German level in 1989.6 As a second stage of analysis, a GNP three times as high as 1992 was assumed for CEECs (Variant II). This level of income is derived from the average relationship between per capita income and the supply of human capital in a country with a market economic system - human capital is measured on the basis of the mean years of schooling of the population.7 As Figure 4.1 shows, the CEECs' level of income in 1992 was well below the level one might expect, given the average relationship between levels of income and human capital. One reason could be that the high level of qualification exists only officially and that this knowledge is devalued under market conditions. A trebling of GNP represents an estimate for potential growth, which one can expect in the case of comprehensive economic transformation (comparable to the results of Black and Moersch in Chapter 5 of this volume). As a third stage of analysis, it was assumed that the focus of economic activity in Germany would tend, during the course of recovery in East Germany, to shift eastward in the long term (Variant III). To this end, geographical distance for a further estimate was measured using Berlin as a point of reference and not Frankfurt am Main. Correspondingly, the distance to Poland was reduced by around 36 percent and to Russia by approximately 17 percent. For other CEECs the reductions are only marginal (4 to 8 percent). The various estimates for total trade between Germany and CEECs are compiled in Table 4.12. The results - derived from the regression equations for Ger6
7
This increase arises only in terms of value due to the price increases in the DM and the appreciation of the DM against the US$. Measured at constant prices, the GNP for the whole of Germany in 1992 was about 22 percent higher and the per capita income approximately 6 percent lower than the West German level in 1989. Thefiguresrefer to 1990 and were taken from UNDP, Human Development Report 1993, New York, 1993, pp. 135-7.
11
10-
« c
9-
9-
8-
2L
6 8 Mean years of schooling
10
12
14
Figure 4.1: Correlation of per capita income and human capital. Source: Own calculations based on data from the World Bank and UNDP. Based on In GNP per capita 1989 and mean years of schooling 1990 in 66 Western industrialized and developing countries (R2 = 0.75). The GNPfiguresfor the CEECs refer to 1992.
Impact on German trade of division of labor
147
many and updated to the level of German foreign trade in 1992 - had already been exceeded by the real value of exports and imports in 1992 in trade with Poland, the CSFR, and Hungary (as given in Table 4.1). Only a trebling of GNP in these countries and, thereby, also a trebling of imports and exports would lead to a considerable further expansion in trade. This is valid for Poland too, if the focus of economic activity in Germany were to shift eastward. In trade with Romania and Bulgaria, the estimated level of exports for 1992 was reached in 1994, but that of imports not yet. In Russia the volume of trade in imports and exports remained well below its potential. The estimates derived from the regression equations for the OECD countries foresee a higher level of German trade with CEECs; the GNP of partner countries plays, on average, a less important role for OECD countries than for Germany, while distance plays a greater role for OECD countries than for Germany. These higher estimates for 1992 were, more or less, reached in 1994 in trade with Poland, the CSFR, and Hungary, whereas the potential for trade with Russia, in particular, has not yet been attained. The biggest CEEC trading partners of Germany today are, after Russia, the CSFR and Poland, followed by Hungary. This corresponds to the hierarchy, which also emerges in the estimates made according to GNP and distance. Before the political changes in Eastern Europe, trade with West Germany reached a "normal" level only for Hungary, whereas trade for Poland and, even more so, for the CSFR lagged behind. Correspondingly, Hungary ranked above the CSFR as a trading partner of West Germany, despite Hungary being smaller and further away. The marked orientation of Hungary toward the West, which was apparent then, has lost its significance totally over the course of political developments. Trade with Poland, which with nearly 40 million inhabitants is by far the largest of the three Visegrad countries, will only outstrip trade with the CSFR relative to the extent that economic activities in East Germany strengthen and that Germany and Poland move closer together. The estimates for trade with CEECs reflect their low level of production and income. Estimates based on the earlier and much larger GNP data from the CIA lead to much higher values, indicating a large trade potential beyond the low level of East-West trade before 1990. It may be, however, that the low level of trade between the OECD and CMEA countries is explained, not only by political opposition, but also by a de facto low level of real income in CMEA countries by market economic standards. As for any backlog, it was quickly made up for by German trade with the countries most advanced in the transformation process, that is, Poland, the CSFR, and Hungary. Further extension of trade relations now depend crucially on future growth in these countries. In contrast, trade with Russia remains far behind its market economic potential, so that considerable increases can be expected, if serious economic reform in Russia is continued. However, in view of the political difficulties, realization of market economic reforms will require a lengthy period of time. The range of the po-
Table 4.12. Estmated values of German exports and imports in trade with CEECs (billion US $) Poland
CSFR
Hungary
Romania
Bulgaria
Five countries total
Russia
Exports Based on regression results for Germany I
GNP in CEECs 1992, West Germany 1989
3.3
3.4
1.9
0.9
0.5
10.0
7.6
la
GNP in CEECs 1992, German trade level 1992
4.1
4.2
2.4
1.2
0.6
12.5
9.6
II
GNP in CEECs three times 1992, German trade level 1992
12.2
12.4
7.0
3.4
1.8
36.8
28.1
III
additionally Berlin instead of Frankfurt a.M.
16.3
12.8
7.4
3.5
1.9
41.9
31.7
Based on regression results for OECD countries I
GNP in CEECs 1992, West Germany 1989
3.4
4.4
2.1
0.9
0.5
11.3
6.3
Ib
GNP in CEECs 1992, Germany 1992
6.0
7.6
3.6
1.5
0.8
19.5
11.0
II
GNP in CEECs three times 1992, Germany 1992
17.1
21.8
10.4
4.4
2.4
56.1
31.5
III
additionally Berlin instead of Frankfurt a.M.
25.5
22.8
11.2
4.5
2.5
66.5
37.1
6.0
Imports Based on regression results for Germany I
GNP in CEECs 1992, West Germany 1989
2.6
2.4
1.5
0.8
0.4
7.7
la
GNP in CEECs 1992, German trade level 1992
3.9
3.7
2.2
1.2
0.7
11.7
9.1
II
GNP in CEECs three times 1992, German trade level 1992
10.5
10.0
6.1
3.3
1.8
31.7
24.6
III
additionally Berlin instead of Frankfurt a.M.
13.4
10.2
6.4
3.4
1.9
35.3
27.3
Based on regression results for OECD countries I
GNP in CEECs 1992, West Germany 1989
3.8
4.2
1.9
0.8
0.4
11.1
9.7
Ib
GNP in CEECs 1992, Germany 1992
6.5
7.2
3.3
1.3
0.6
18.9
16.8
II
GNP in CEECs three times 1992, Germany 1992
23.8
26.4
12.1
4.8
2.2
69.3
61.4
III
additionally Berlin instead of Frankfurt a.M.
35.7
27.6
13.0
5.0
2.3
83.6
72.5
Source: Own calculations based on the regressions with the reduced number of variables.
150
Dieter Schumacher
tential is also possibly exaggerated, assuming a per capita income that is still overestimated. 4.6
Sectoral pattern of trade and structural changes in Germany
The sectoral structure of trade between Germany and CEECs is determined by geographical proximity, large disparities in the level of per capita GNP and wages, and the relatively good situation of human capital. Which product groups may have comparative advantages on these grounds can be derived from the elasticities of exports and imports respectively with respect to distance and income (see Tables 4.7 and 4.10). The CEECs should prove to be more important export markets for those German industries that have a high elasticity of exports with respect to distance and a low elasticity with respect to the income of the importing country. Those industries that have a high elasticity of exports with respect to the GNP of the importing country will increase their share with rising GNP in the CEECs. As for German imports, the CEECs should tend to achieve high shares in those product groups that have a high elasticity of imports with regard to distance and a low elasticity with respect to the income of the supplier country. Those product groups that have a high elasticity of imports with regard to the supplier country's income should increase their share with rising GNP in CEECs. The following analysis will be concentrated on the division of labor within the manufacturing sector and will be based - corresponding to the regression approach - on general comparative advantage as a function of the level of income and, in as far as it is connected to this, human capital endowments. Country-specific advantages, such as history and natural conditions, will not be taken into consideration; they may, however, be apparent in the present commodity structure of imports from the individual CEECs. As the exports of the former USSR are obviously based on the considerable natural resources of the country, the following analysis will be restricted to the other CEECs. It will draw, above all, on the regression results for German trade and will be supplemented by results for OECD countries in general when they are more significant. This applies, in particular, to imports. In accordance with the distance elasticity of German exports, proximity to the market has the greatest positive effect on German exports of clothing, wooden articles, furniture, mineral oil products, textiles, and shoes. Conversely, on the import side short distances between CEECs and Germany give an advantage, above all, to supplies of mineral oil products, iron and steel, "other" nonmetallic mineral products, wooden articles, and motor vehicle industry products. This is particularly true for trade with the CSFR, followed - in order of distance from West Germany - by Hungary, Poland, Romania, and Bulgaria. The German structure of exports varies less with the level of income of the market. Germany tends to sell to low-income countries relatively large amounts of iron and steel, industrial chemicals, and machinery. With increasing income
Impact on German trade of division of labor
151
German exports of mainly clothing, shoes, furniture, leather products, and wooden articles become more important. As for imports from CEECs, their low level of income tends to give them a comparative advantage in wooden articles, pottery, china and earthware, iron and steel, and foodstuffs, as well as "other" nonmetallic mineral products. With increasing per capita income, the significance of, above all, plastic products, industrial chemicals, rubber products, shoes, and paper may increase. Summarizing, one may say that Germany imports more investment goods and fewer consumer goods, the higher the income level of the trading partner is, whereas on the export side the situation is reversed. A similar pattern is apparent using the classification of industries according to R&D intensity: With increasing GNPof the trading partner the proportion of medium- and high-tech goods on the import side increases and on the export side decreases. Hence, the intersectoral division of labor decreases, whereas the intrasectoral division of labor becomes more significant. This trend also emerges from the estimates of the commodity structure for German trade with CEECs on the basis of the regression equations at alternative income levels. Within the next few years, however, a high proportion of intersectoral division of labor can be expected, so that the expansion of trade will induce changes in the sectoral pattern of production and employment in Germany. This has regional implications and affects the demand for the factors of production. On the one hand, employment will profit in Germany from high exports to CEECs. On the other, domestic production finds itself under increasing competition from imports. At their present level of incomes, CEECS represent, first and foremost, competition for other low-income countries. However, due to geographical proximity to Germany, additional possibilities for international division of labor exist that are not available for countries further away. According to the results of earlier studies, CEECs are placed between the developing countries and OECD countries with regard to the proportion of intrasectoral trade, labor content, and human capital content of their supplies in Western trade (Schumacher, 1989; Sapir and Schumacher, 1985). More up-todate calculations for this study, based on the commodity structure of German trade until 1993, confirm these results. To this end, the direct content of labor and human capital for the various commodity bundles of exports and imports were calculated using the average sectoral production functions in West Germany. Input of labor was measured in hours worked per unit of gross output and input of human capital on the basis of personnel expenditure per hour worked. This assumes that the sectoral differences in the hourly rates of pay reflect only the differing structure in the level of training of work forces. In fact, wages depend on a number of other factors too, such as the productivity of capital, gender and age of workers, or the influence of unions. One can assume for the majority of industries, however, that the differences in wages are essentially related to the differing intensity of human capital.
152
Dieter Schumacher
Table 4.13. Factor content, exports relative to imports Labor content
Human capital content
Poland
0.93
1.16
CSFR
0.98
1.08
Hungary
0.95
1.12
Romania
0.81
1.31
Bulgaria
0.85
1.28
Five countries total
0.94
1.14
OECD countries
1.01
1.03
Developing countries
0.85
1.25
Dividing the factor content of exports by that of imports, the relations shown in Table 4.13 emerge for the commodity bundles in German trade with various countries or groups of countries for 1993. This shows that the labor content of exports in trade with CEECs is smaller than that of imports. The opposite is true for human capital intensity. A uniform expansion of trade leads to a somewhat smaller need for workers in Germany, and the economy-wide productivity of labor increases due to the change in the sectoral pattern of production. The demand for human capital in Germany increases following additional division of labor with Eastern Europe, that is, the loss of jobs is concentrated on less qualified workers, whereas the demand for qualified workers increases. The effects of changes in the sectoral structure of the German economy are strengthened by the expansion of the intrasectoral division of labor, which follows the same basic pattern: Standardized, labor-intensive parts of production processes are relocated to CEECs. This is particularly true for economic branches where regional proximity is necessary for production. Hence, outward processing has expanded in the last few years, above all, in the areas of textiles, clothing, and leather, but also in electrical appliances (Mobius, 1995). In the motor vehicle industry the big German firms included Eastern Europe in their internal division of labor. Further investment goods industries follow this example, so that the division of labor in intermediate goods intensifies with Eastern Europe, with corresponding structural changes and productivity gains in Germany. Consequently, the structural changes arising from trade with CEECs call for an upgrading of the skill content of the German labor force in general. Hence,
Impact on German trade of division of labor
153
increased trade with CEECs tends to further the long-term trends of structural changes in Germany (and other Western industrialized countries), that is, increases both labor productivity and human capital intensity. CEECs will develop more quickly into middle-income countries, the more modern capital they accumulate themselves. Their education and industrial experience may offer better prerequisites than are present in many developing countries, with whom they rank equally, at present, in terms of per capita income. 4.7
Summary and policy conclusions
German trade with CEECs, in particular with the CSFR, Poland, and Hungary, who are nearest to Germany and most advanced in the transformation process, has expanded dynamically in the last few years. An important role has been played by the increased division of labor in textiles, leather, and clothing, in the form of outward processing, but also in the motor vehicle industry and in electrical engineering. The regional structure of German foreign trade has shifted in the direction of Eastern Europe, above all, to the disadvantage of the West. Whether CEECs will regain their former share in German trade of nearly 18 percent (in 1913) depends crucially upon the ability of CEECs to make up the gap in income. In any case, the analysis has shown that there is still considerable potential for further expansion of trade under the condition that the transformation process toward a market economy progresses further and will be successful in terms of growth. With regard to the commodity patterns of trade, one might expect that CEECs will achieve high shares of German exports in all industries: (1) in basic and consumer goods due to geographic proximity and (2) in investment goods due to their low income level. Consequently, the commodity structure of German exports to CEECs will be very similar to that of overall German exports. On the other hand, CEECs will achieve high import market shares in Germany mainly in consumer goods due to their low income level, and with rising GNP their share in investment goods will rise too. A strong economy like Germany's should, in principle, be able to cope with the structural changes arising from the increased division of labor with Eastern Europe. To a large extent, these changes have already taken place, and the effects of a further doubling or trebling of trade with CEECs would be distributed over many years and the adjustment costs would not be very high as compared to the size of German GNP. The German economy should, in fact, benefit from these new structures. On the whole, the economic performance of West Germany in the past can be assessed as good. This is particularly true when measured against key data such as the increase in real income and productivity, as well as the terms of trade. It is, however, worrying that, in recent times, Germany has been slackening in its efforts to make provision for the future. Spending on R&D and education
154
Dieter Schumacher
is declining in proportion to GNP, and the share of predominantly structureconserving subsidies in total support continues to be high in comparison to subsidies affecting technology and innovation. These developments could mean that Germany is living increasingly on its capital and that its still comparatively good performance is gradually deteriorating, since such omissions make themselves felt only at a very late date. Also, the changes that Germany will be facing as a result of development in Eastern Europe will require German policy to be more forward looking and long term. Another prerequisite for a further increase in trade is a more liberal stance from the EC regarding imports from Eastern Europe. The imports of agricultural products are hampered by the Common Agricultural Policy, and this protection is not reduced in the Europe Agreements. In principal, the Europe Agreements provide a framework for a progressive liberalization of industrial imports from the CEECs. The stipulations in detail, however, show that the EC again follows the traditional attitude, that is, is less liberal, the more competitive the foreign supply. The final benefit to the CEECs will crucially depend on the actual behavior of the EC within the provisions of the Agreements. There is room for less liberal behavior with regard to raising tariffs on "sensitive" products above the ceilings and the escape clause may or may not be applied to introduce import quotas or antidumping proceedings when increased imports "cause or threaten to cause serious injury." All in all, one may say that the Europe Agreements potentially offer a significant liberalization of imports from the CEECs, providing space for high additional supplies of industrial products. The trend toward liberalization is not, however, irreversible as there are various escape clauses in the Agreements. To refrain from high antidumping duties and to remove, or avoid reapplication of, quantitative restrictions on products where CEECs prove to be competitive is more important than any acceleration of tariff reductions. REFERENCES Anderson, J. E. (1979). "A Theoretical Foundation for the Gravity Equation," The American Economic Review, 69, 106-16. Baldwin, R. E. (1994). Towards an Integrated Europe, London: Centre for Economic Policy Research. Bergstrand, J. H. (1989). "The Generalized Gravity Equation, Monopolistic Competition, and the Factor Proportions Theory in International Trade," The Review of Economics and Statistics 71, 143-53. Bergstrand, J. H. (1985). "The Gravity Equation in International Trade: Some Microeconomic Foundations and Empirical Evidence," The Review of Economics and Statistics 67,474-81. Brocker, J. (1980). "Measuring Trade Impeding Effects of National Borders by LogLinear Interaction Analysis," Institut fur Regionalforschung der Universitat Kiel, discussion paper 19. Dhar, S., and A. Panagariya (1994). "Is East Asia Less Open than North America and the European Economic Community? No," The World Bank Policy Research Working Paper 1370.
Impact on German trade of division of labor
155
Frankel, J. A. (1992). "Is Japan Creating a Yen Bloc in East Asia and the Pacific?", NBER Working Paper 4050. Frankel, J. A., and S.-J. Wei (1993). "Trade Blocs and Currency Blocs," NBER Working Paper 4335. Haass, J. M., and K. Peschel (1982). Rdumliche Strukturen im internationalen Handel - Eine Analyse der Aufienhandelsverflechtung westeuropdischer und nordamerikanischer Lander 1900-1977, Munchen. Havrylyshyn, O., and L. Pritchett (1991). "European Trade Patterns After the Transition," The World Bank Country Economics Department Working Paper 748. Herrmann, H., W. D. Schmidtke, J. Brocker, and K. Peschel (1982). "Kommunikationskosten und internationaler Handel. Uberlegungen zum Marktverhalten von Exporteuren und empirische Untersuchungen zur Erklarung der Aupenhandelsverflechtung," Schriften des Instituts fur Regionalforschung der Universitdt Kiel 4, Munchen. Learner, E. E. (1993). "U.S. Manufacturing and an Emerging Mexico," North American Journal of Economics and Finance 4, 51-89. Learner, E. E. (1974). "The Commodity Composition of International Trade in Manufactures: An Empirical Analysis," Oxford Economic Papers 26, 350-74. Linnemann, H. (1966). An Econometric Study of International Trade Flows, Amsterdam: North-Holland. Losch, D., and E. Wohlers (1995). "Auswirkungen der Transformationsprozesse in Mittel- und Osteuropa auf die deutsche Wirtschaft," Wachstumsperspektiven in den neunziger Jahren, Beihefte der Konjunkturpolitik 42, 131-72. Mobius, U. (1995). Passive Lohnveredelung im Rahmen der Textilimporte der EU und der Bundesrepublik Deutschland. Gutachten im Auftrag des Bundesministeriums fur Wirtschaft anla(31ich der Import-Messe Berlin 1995. Mobius, U., and D. Schumacher (1995). "Analysis of Community Trade Barriers Facing Central and East European Countries and Impact of the Europe Agreements," European Commission, "The Economic Interpenetration Between the EC and Eastern Europe," European Economy. Reports and Studies, No. 6, Brussels/Luxembourg, 17-76. OECD (1992). Industrial Policy in OECD Countries. Annual Review 1992, Paris. Peschel, K. (1980). "On the Impact of Geographical Distance on the Interregional Pattern of Trade and Production," Environment and Planning, 13, 605 seq. Sapir, A., and D. Schumacher (1985). "The Employment Impact of Shifts in the Composition of Commodity and Services Trade," in Employment Growth and Structural Change, Paris: OECD, pp. 115-27. Saxonhouse, G. R. (1993). "Trading Blocs and East Asia," in J. de Melo and A. Panagariya (eds.), New Dimensions in Regional Integration, Cambridge, pp. 388-416. Schumacher, D. (1992). "A Note on the Human Capital Intensity of EC Trade," Cahiers Economiques de Bruxelles 133, 3-19. Schumacher, D. (1989). "Employment Impact in the European Economic Community (EC) Countries of East-West Trade Flows," ILO International Employment Policies Working Paper 24. Soete, L. (1987). "The Impact of Technological Innovation on International Trade Patterns: The Evidence Reconsidered," Research Policy 6, 101-30. Vittas, H., and P. Mauro (1994). "Implications of the Opening Up of Eastern Europe for Germany, Switzerland and Portugal," paper presented to a Seminar on "Western Europe in Transition: The Impact of the Opening Up of Eastern Europe and the Former Soviet Union," Trieste, October 10—11. Wang, Z. K., and L. A. Winters (1991). "The Trading Potential of Eastern Europe," CEPR discussion paper 610.
156
Dieter Schumacher
Winters, L. A., and Z. K. Wang (1994). Eastern Europe's International Trade, Manchester: Manchester University Press. Wood, A. (1994a). North-South Trade, Employment and Inequality. Changing Fortunes in a Skill-Driven World, Oxford. Wood, A. (1994b). "Give Heckscher and Ohlin a Chance," Weltwirtschaftliches Archiv 130,20-49.
Comment Wolfgang Maennig
Schumacher's well-founded work will be of great interest for the German academic public. Since this trade study has, for a wider audience at least, more of an instrumental character with regard to the central labor market questions, it seems appropriate to make a few preliminary remarks to put the discussion in a wider context. For Eastern Europe, trade with Germany and the resulting effects on employment, competitiveness, and growth are of considerable significance; around half of the Eastern European countries' exports go to the EU, and half of these in turn go to Germany. Their future growth will likely depend very much on the volume of trade, but also to a large degree on the trade-related transfer of technology and the effects on competition and innovation. For Germany, the effects might be much smaller. While Schumacher emphasizes the considerable dynamism of Germany's eastern trade, which exceeded in 1994 the trade volume with the United States for the first time, we are still talking about no more than 7 percent of German trade. Even if it is appropriate to differentiate between individual sectors, it must nevertheless be said that even for a relatively "open" country such as Germany (with export and import shares of 38 percent and 30 percent, respectively), a multiplication of trade with the Eastern European countries will only have a limited effect on the economic structure and macroeconomic variables in the foreseeable future. On the other hand it should be noted that effects on the German labor market are expected not only via trade but also by migration and capital movements. Thus the total effects could be much larger. Fundamental remarks on the gravity model The extent to which the gravity model can explain the dimensions of bilateral trade both here and in other studies is astonishing. Given R2 of 0.8 and more, the question arises whether it is at all worth teaching students what has now become a wide range of "old," "new," and abandoned trade theories. In some cas157
158
Wolfgang Maennig
es it is enough to point to the fact that where activity is greatest, trade, a typical economic activity, will "naturally" also be induced. The theoretical foundations of the gravity model are unclear. It has pretensions to general validity and does not differentiate between Ricardian and Heckscher-Ohlin forms of trade, trade under imperfect competition and other types of trade. The inclusion of the developing countries in the data base for the evaluation of the gravity model leads to a better representation of the former types of trade, yet to a worse depiction of the latter. The resulting concentration on comparative cost differences as the determinant of current and future trade, which we also find in Schumacher's argumentation, is surprising, particularly in the light of the significance of the "new trade theories" in contemporary research and teaching. "New trade theory" emphasizes the role of external and internal economies of scale for trade. Gravity models normally register economies of scale by including a size variable such as population as an exogenous variable. This size variable then is expected to have a negative effect on trade, because a large country needs the trade less to realize economies of scale. This effect probably also exists in Schumacher's study, although it remains hidden and indeed is "reinterpreted" with the aid of the income/population variable into an effect of human capital endowment according to the Heckscher-Ohlin pattern. Another area where new trade theory has made significant advances remains completely neglected in the gravity models. Indeed the new trade theory shows precisely that comparative cost advantages are neither constant nor exogenously given, but rather are dependent on experience. They show that international trade can under certain conditions run against "potential" comparative cost differences and is heavily influenced by history and accident. To this extent the argument that Eastern Europe has excellent human capital at its disposal and that the Eastern European countries' other comparative cost disadvantages could be quickly diminished by the transfer of technology that accompanies trade has only limited validity. The economic structure that has arisen due to the political mistakes of the last forty years places Eastern Europe way up on the learning and average-cost curves. Economies of scale in the trade between Germany and the mostly significantly smaller Eastern European countries could thus lead to unwanted effects for the latter. Hence, given the basis from which it is starting, Germany could specialize in production with economies of scale, which are typically connected with rents, while the Eastern European countries could specialize in other products typically involving no rents, in accordance with their "apparent" comparative cost advantages. Besides the doubts about the ability of the gravity model to correctly depict the volume of trade effect and the structure of trade effect, there should also be an examination of the capability of the approach to estimate correctly the direction of trade effect. The gravity model cannot distinguish between trade creation and trade diversion. The increase in the exports of an Eastern European country to Germany has already been linked to dramatically reduced exports
Comment
159
from the same country to other Eastern European countries (approximately 50 percent), something that will also be the case in the future. Moreover, the increase in German imports from Eastern European countries could be less of a burden on German import substituting production than on imports from OECD partner countries, as Schumacher has already determined for past developments. The income and welfare effects derived from the gravity model would thus be biased upward. Under certain circumstances a systematic downward distortion can also result. Hence the increasing supply pressure in the German agrarian sector will, for example, influence the situation of German agrarian production not only via bilateral trade but also via the increasing difficulties of being successful on world markets. Moreover, the changes in Eastern Europe make reform of the Common Agricultural Policy more likely in the medium term, with corresponding effects on income and employment in Germany. Other determinants that are of central importance in estimating imports and exports are not included in the gravity model. In this context, it should be clear that not only are the exchange rates between the DM and the Eastern European partner currencies of significance, but that the strength of the dollar and the yen will also have an influence on bilateral trade relations. Subsidies (in Russia energy and petrol prices are just a fraction of world market prices) will influence the Eastern European economies' ability to export. The German policy of the Hermes export credit insurance agency, which has tended to be more regressive in the recent past, particularly toward Russia, could also be of significance. And finally, social order policies and institutional influences must also be taken into account. The inadequate transport and telecommunications infrastructure, high instability in economic policy (and, in connection with this, the exchange rates, tax conditions, and demand for goods), and the underdeveloped financial markets in Eastern Europe will have a far greater effect on trade, due to its dependency on transport, than will the "normal" relationship between income and trade as depicted by gravity models. The errors associated with the gravity model may as a whole be well compensated for, and on average still provide useful explanations. Yet, overall, the model gives only an overview of trade to be expected at some time in the future, but by contrast has little to say about mechanics of transformation. Data problems Dieter Schumacher uses GNP and GNP/population as explanatory variables and argues that this is equivalent to the use of GNP and population. Yet it must be noted that Schumacher's procedure can create a multicollinearity problem and can lead to the misinterpretation already mentioned. The estimates for future trade depend to a large extent on income in Eastern Europe. Prognoses on future income developments are, however, very difficult at present. Schumacher assumes a tripling. This is of only limited use for two
160
Wolfgang Maennig
reasons: first, Schumacher says little on the period over which the increases in income are to be realized. Yet such a temporal framework is an indispensable factor for labor-market-oriented prognoses. Not least in view of the coming demographic changes, the implications for the labor market in Germany differ enormously when these adaptations take place in twenty to thirty years instead of ten to fifteen. Second, the assumption of a tripling of income seems somewhat too imprecise, even if no temporal framework is adopted. It is already clear at the present time that the successes in transformation of the twenty-seven Eastern European countries - from Bosnia-Hercegovina, Serbia, Tajikistan, Armenia, Belorus to Slovenia, Chechnya, and Hungary - vary enormously. The number of countries that have succeeded is less than those that have failed. A tripling may thus seem to be too optimistic even for a long-run view. Yet it may also seem to be too little, for even if attained it means that only 30 percent of the German level of income is reached, to take Bulgaria as an example. In any case a better foundation and greater differentiation of the income growth assumed in Eastern Europe and consideration of the fact that German income will also increase, leading in turn to additional trade-increasing effects, could improve the quality of the forecast. Interpretation of the results Schumacher's results lie withing the spectrum of estimates made by others. These estimates probably represent a realistic picture. Yet in the end this confirms the relativization undertaken above for Germany. Even if one day the potential should be fully realized, the volume of total German trade (without taking possible trade diversion from other countries into account) would only increase by 10 to 15 percent. Spread over the necessary adaptation phase of around twenty years, the annual rates of adaptation remain small. According to Dieter Schumacher, future East-West trade will be similar to current West-West trade. However, a strongly growing intra-industry trade causes less political resistance than the interindustry trade typical for North-South trade. The import competition will cost jobs in certain sectors, but the simultaneously arising export opportunities will in the end give rise to other jobs with a comparable qualification profile. The problems will thus be less than when whole industries are destroyed by import competition. Schumacher does not present any explicit welfare calculations. Yet it can be deduced from his line of argument that both sides will experience positive effects. Without fundamentally questioning this, two small caveats should still be mentioned for the Eastern European side. First, the implicit assumption that trade as a whole will be well-balanced is a dubious one, particularly in view of the lasting trade imbalances in the global economy. Second, even if it were balanced, 3 large part of coming trade growth will be induced by direct investments and transfers of originally "German" production units to the Eastern European
Comment
161
countries for the purpose of minimizing labor costs. German direct investment in Eastern Europe was twenty times higher in 1994 than in 1989 and made up 10 percent of all German direct investment. The welfare effects associated with obtaining Eastern European deliveries from German subsidiaries are more positive for Germany and less positive for Eastern Europe than those associated with deliveries from solely Eastern European enterprises. To illustrate the role of history and accident, there can hardly exist a more topical case than that of Eastern Europe. In view of the related possibility of a hysteretical or at least persistent structure of trade, the question of the necessity of an active industrial and trade policy should also be asked.
Comment Ellen Meade
Let me begin by saying that Dieter Schumacher is to be commended for undertaking this study. Getting reliable estimates of trade flows at an aggregate and industry level is difficult enough for countries in stable times, but predicting relationships during such a period of change is enormously challenging. My comments are mostly about the big picture, but I do have some remarks on the details of this paper as well. We would all agree I think that the circumstances in Central and Eastern European countries are changing so rapidly and have changed so rapidly in the past five or six years as to make empirical estimation of economic relationships tenuous at best. Fundamental changes in economic structure have led to tremendous swings in real incomes and exchange rates. Moreover, although we know that the swings are large, precise measurement of particular economic variables is unreliable and fraught with error - for instance, in measuring real incomes, how reliable are pretransition data, how can we value the existing capital stock, and so on. These considerations affect how one decides to deal with the question of Germany's trade with Central and Eastern European countries, the question that Dieter Schumacher addresses in this paper. My sense is that the author would ideally have liked to estimate a model based on endowments of labor, human capital, and physical capital, along with other variables such as geographical proximity and trade preference dummies across disaggregated industry data that is, an explicit testing of the Heckscher-Ohlin theorem. Instead, this paper relies on a "gravity model," in which income and income per capita in source and destination countries proxy for endowment factors. Another model worthy of empirical evaluation would be one based on intra-industry trade - it is worth noting that since equations are estimated over a sample that includes both developing and industrial countries, we would expect trade to be a mix of resourcebased and intra-industry trade. To be completely clear about the exercise, the gravity model is estimated for trade of OECD countries and then for Germany's trade with seventy or so trading partners, in cross-section regression analysis using average data for the 162
Comment
163
1988-90 period. Values for Central and Eastern European countries are then "plugged into" the estimated equations, yielding predictions for trade between Germany and CEEC countries. I'll return to some discussion of this in a minute. Upon reading this paper, I was troubled by this approach. In a period of such transformation, are results from this approach - or any other empirical approach - meaningful? Perhaps the author would answer me that factor endowments, at least of labor, human capital, and natural resources, are invariant to a first approximation. But since these endowments are proxied in this exercise by levels of income, and income has varied widely and is imprecisely measured, then what exactly are we getting? Moreover, the estimates are intended to proxy for long-run trading relationships. Thus, it seems that 1988-90 is too short a period over which to average the data. A fifteen- or twenty-year period perhaps is more appropriate to establish the average long-run relationships - and to remove particular variation due to business cycles from the data. What this analysis gives is a wide range of estimates. As Table 4.12 shows (it gives estimated values of German trade individually with Central and Eastern European countries, where a number of alternative estimates of CEEC income and other adjustments are plugged into the estimated equations), Germany's exports to thefiveCEEC countries may range from as low as $ 10 billion to as high as $66.5 billion. The range of predicted values for imports is even higher - from $7.7 billion to $83.6 billion. Thus the author himself would attach a wide range of uncertainty around any particular point estimate for German trade. I wonder how these estimates would differ if the empirical estimates were made based on the trading relationships between OECD countries or Germany with developing countries alone (that is, excluding trade between industrial countries). One argument in favor of this is that trade with the Central and Eastern European countries most resembles the trade with developing countries, and it is therefore that portion of the sample that one should pay attention to. Another interesting calculation is shown in Table 4.13. The column labeled "labor content" gives the hours per unit of output for German exports relative to German imports. As the estimates indicate, Germany is more productive in the output of its exports to CEEC countries than those countries are in their exports to Germany. Or putting it another way, capital intensity is greater in German exports to than imports from these countries. The second column - "human capital content" - shows the wage bill per hour worked, once again as a ratio between German exports relative to imports from CEEC countries. The author interprets the figures as indicating that human capital content is higher in German exports. A similar pattern emerges for trade with developing countries, but not for trade with OECD countries. These estimates support the notion that the trade of countries at equivalent stages of development is primarily intra-industry, whereas trade between industrial and developing countries is largely of the interindustry variety.
164
Ellen Meade
Returning to my more general comments earlier, my initial reaction was that the more appropriate way to examine trading patterns between Germany and CEEC countries was to estimate bilateral time series equations for exports and imports, based on real incomes and relative prices - the sort of relationship estimated in most macroeconometric models of trade. (I should note that such an approach does not preclude breakdown by industry.) It seems to me that this is a question I would rather know the answer to: that is, for different paths of income growth and relative price (or real exchange rate) changes, what is the path for German trade with CEEC countries? Unfortunately, I don't know that my question can be answered easily. Such predictions require a stable relationship estimated on historical data. As I noted earlier, the data are unreliable. Furthermore - even if data were reliable - in such a period of transformation, a relationship based on historical experience may not prove robust to such changes. It does seem to me that there is a role for relative prices in the work presented here. Just as incomes vary across countries, so do real exchange rates, and they should help to explain patterns of trade. In fact, the transformation of economies in Eastern European countries has led to significant declines in real exchange rates, helping to enhance the exports of those countries.
CHAPTER 5
Investment and its financing during the transition in Central and Eastern Europe Stanley W. Black and Mathias Moersch
5.1
Introduction
The end of the Cold War found the countries that spent the previous 45 years under Communism recovering from a legacy of misguided economic incentives and policies and a system built on false premises. The 75 percent gap in income levels between the Central and Eastern European Countries (CEECs) and the members of the European Union (EU) is a major barrier preventing the CEECs' entry into the EU, a key political goal. This gap not only raises the cost of including the CEECs in the EU's income redistribution programs such as the Common Agricultural Policy and the Regional and Structural Funds, it also forms the basis for low wage competition for the "sensitive" industries in the EU, which already face severe competition from Third World countries. This paper addresses some of the issues involved in closing the income gap, assessing the growth potential and associated investment demand in Bulgaria, the Czech Republic, Hungary, Poland, Slovakia, and Romania, collectively called the CEECs. Over the past five years, a variety of estimates have been made of the demand for investment in these transition economies. The most popular approach relies on a production function to estimate the starting level of capital stock required to produce the starting level of output, assuming a given degree of inefficiency in the initial use of capital and labor (Boote, 1992; Giustiniani, Papadia and Porciani, 1992; Holzmann, Thimann and Petz, 1994). Catching up to Western European levels of output depends on the elimination of these inefficiencies and the accumulation of capital. Similar approaches assume an incremental capital-output ratio; the increment of capital required depends on the desired increment of output (Begg, Danthine, Giavazzi and Wyplosz, 1990; Collins and Rodrik, 1991; Handler and Steinherr, 1993). Other approaches include adding the Eastern labor force to the West and assuming gradual reallocation of capital (Fitoussi and Phelps, 1990), and an endogenous growth analysis relating output per head to the investment share of GDP and Research assistance for this paper was provided by Felecia Lucht and Kellie Maske.
167
168
Stanley W. Black and Mathias Moersch
schooling (Borenstein and Montiel, 1991). To achieve a given output per head then requires a particular investment share. This paper also relies on a production function approach, but differs from previous work in several ways. Rather than postulating an output goal and deriving the financing need as a residual, potential sources of finance are estimated. Given these estimates, potential output levels are determined. In addition the paper considers two elements of the transition process that seem to be omitted from most other analyses, namely the negative initial shock to the CEECs' economies and the reallocation of resources from industry to services. First, recovery from the initial shock is likely to be more rapid than the following period of sustained growth. Second, reallocation of capital and labor from heavy industry to light industry and services potentially provides a substantial increment to productivity, even without additional capital. This is because services and light industry require a lower capital/labor ratio than heavy industry and may also have a higher elasticity of substitution between capital and labor. In addition, producing what consumers actually want to consume in itself generates a more highly valued output. The conceptual basis for these arguments is illustrated in Figure 5.1. Shown is the production possibility curve between manufacturing and services with all available resources used efficiently. Point B represents the initial production point with factors used inefficiently prior to the reforms in the CEECs. Point A represents the fall in output due to the initial shock to the CEEC economies. Trade between the CEECs and the Former Soviet Union (FSU) fell sharply with the collapse of "transferable ruble" trade among the members of COMECON. Some decline in demand for CEEC output was also due to the newly available Western goods and the resulting shift in demand away from CEEC goods. There was also a fall in the demand for CEEC military products with the end of the East-West confrontation. The future growth of the CEECs then consists of four phases: recovery (A-B), reform (B-C), restructuring (C-D), and capital accumulation (D-E). The recovery phase depends upon the reestablishment of external markets for CEEC outputs and the recovery of purchasing power by domestic residents. It is modeled by simulating a return of employment figures to precollapse levels. The reform phase depends upon the replacement of obsolete capital and organizational methods and the reorganization of management and improvement of incentives and skills of the labor force. This phase can take many years, depending on the speed with which reforms are adopted and take root in industry. We model this effect by adjusting efficiency parameters in the production function. The restructuring phase (Figure 5.2), while occurring simultaneously with the other phases, is conceptually distinct in that it involves a reorientation of the production structure away from the priorities set by the central planners and toward those determined by market choices. We assume that CEEC consumers
Investment and its financing during the transition
169
Y, - Manufacturing
Figure 5.1: The four stages of transition.
have the same preferences as Western Europeans, so we use Western prices as the standard for valuation of outputs. Associated with the initial production structure C, and its inefficient counterpart B, was a set of distorted prices px that can be thought of as tangent to the production possibility curve at point C. Due to the undervaluation of production of services as "nonproductive activities" and the bias toward heavy industry and military production, the share of services relative to manufacturing goods is initially low. Valuing the initial output at the set of Western prices p* consistent with the choice of point D would show point C to be worth considerably less at Western prices than point D. Furthermore, the value to consumers of the output at C is substantially overstated by standard GDP accounting, since point C reflects planners' choices rather than consumer preferences. Consumers gain substantially from the ability to choose their preferred bundle at D. We model the restructuring phase as a move from point C to point D along the production possibilities frontier toward services. This move of the economy toward the preferred composition of output will itself be a stimulus to the growth of output, measured properly in Western prices. Finally, the accumulation phase shifts the production possibility curve in Figure 5.1 outward from D to E, as both physical and human capital are accu-
170
Stanley W. Black and Mathias Moersch
Y,-Manufacturing
Figure 5.2: Output choices.
mulated through investment and training. The simulation of this phase is based on projections of domestic and foreign savings behavior. Section 5.2 provides a summary of previous studies. Section 5.3 evaluates the potential sources of internal and external financing. Section 5.4 seeks to quantify the recovery, reform, and restructuring phases in a static context, while Section 5.5 provides a simulation of the transition process including the capital accumulation phase, based on the data and assumptions introduced in Sections 5.3 and 5.4. Section 5.6 concludes with an evaluation of the CEECs' current prospects for a successful transition to eligibility for membership in theEU. 5.2
Recent studies Methodology
Source-based and need-based estimates are two commonly used methods to arrive at scenarios concerning the buildup of capital in the CEECs over the next decade. Source-based estimates identify the likely sources of financial flows into the CEECs over the short and medium term. They consider the overall availability of capital from private and public investors relative to the world-
Investment and itsfinancingduring the transition
171
wide demand for capital. Alternatively, source-based estimates have been built around comparable historical episodes of large demands for capital. Need-based estimates start by formulating a level of output, often given as a percentage of Western output levels, that the transition economies wish to achieve in future years. Simulation exercises are then used to determine the capital stock needed to reach this level of output. An important aspect of these simulations is that a number of assumptions are required concerning the desired level of output, the initial position of the economy, and the parameters of the production function. Not surprisingly the range of estimates obtained with needs-based methods is therefore rather wide. Obviously, assumptions about the desired target level of output will affect the estimates. The target levels are usually formulated with a political goal, for example membership of the CEECs in the EU, in mind. The valuation of current levels of capital, labor, and output causes a problem for two reasons. First, due to the lack of well-developed accounting systems, the data are not very accurate. Second, alternative methods of currency conversion affect valuations in dollar terms. Valuations are usually done either at current exchange rates or at purchasing power parity rates. Most studies assume an initial inefficiency in the production function. The assumed time path at which the inefficiencies are eliminated will strongly influence the estimates. Finally, the relationship between inputs and output needs to be specified. This involves choosing a particular production function and setting various parameters. Total investment, finally, can be financed either internally or externally. Subtracting domestic savings from the total capital requirement gives the amount of foreign capital needed. Empirical results Table 5.1 summarizes studies based on source-based estimates. Collins and Rodrik (1991) derive their estimates by adding up the potential capital flows that can be expected from major providers of capital. They distinguish between international financial institutions, governments of the developed countries as represented by the twenty-four OECD members and the European Commission, and private investors. International institutions and governments lend for political and economic reasons whereas private investors base their lending on risk and return considerations. While relatively reliable figures exist for official funding, private flows are much harder to estimate. Collins and Rodrik base their estimates on extrapolations of previous patterns and a survey of large international firms. They distinguish between three sources of private funds, foreign direct investment, bond financing, and bank lending. A variant of the source-based method underlies their second approach. Here Collins and Rodrik extrapolate from the experience of the Marshall Plan. They ask how much money would have to be transferred today in order to obtain the
172
Stanley W. Black and Mathias Moersch
Table 5.1. Investment requirements based on source-based approach Study
Country Coverage
Investment Requirements (in billion US-$ p.a.)
Collins and Rodrik (1991)
CEEC + Yugoslavia
approach I: 12-24 approach II: 5-14
McKibbin(1991)
CEEC + Yugoslavia
30
CEEC = Bulgaria, former CSFR, Hungary, Poland, Romania.
same flows as in the 1950s, adjusted to today's environment. The adjustment factors are a correction for inflation, a rescaling to keep per capita transfers the same, and a correction for current GNP in both donor and recipient countries. McKibbin (1991) bases his estimates on a rule of thumb, proposed by Sachs (1991), who suggests that for a credible adjustment program, 5 percent of a country's GNP is needed from abroad. Needs-based estimates are summarized in Table 5.2. The policy goal, formulated as a growth rate of real GDP, lies between 7 percent and 12.5 percent. Investment requirements range from $23 billion to $628 billion per year, with an unweighted average of $244 billion per year. As pointed out above, the large variation in these estimates is due to the differing assumptions on which the simulations are based. The study by Holzmann, Thimann, and Petz (1994) illustrates this point. Their low estimate is $23 billion, while the high estimate is given as $599 billion. In their study they allow a number of parameters to vary. First, they allow for differences in the speed of adjustment between Eastern and Western productivity. The optimistic scenario foresees a complete elimination of the difference in efficiency within ten years, while in the pessimistic scenario only half of the efficiency gap is closed by that time. Second, two different domestic savings rates, 20 percent and 30 percent, are considered. Third, the real rate of interest at which foreign financing is provided is also varied: 5 percent, 3 percent, and 0 percent (a grant) are considered. Combining these pairs of assumptions gives 12 forecasts1 that fall into the range given above. More important than the absolute numbers are the trade-offs implied by the various scenarios. Assuming full efficiency in ten years reduces the capital transfers by a factor of five compared to the inefficient scenario. A grant, as opposed to a loan at 5 percent, reduces capital transfers by roughly 20 percent. Similar reductions are obtained if the high, rather than the low, savings rate can be achieved. 1
2 productivity scenarios X 2 savings scenarios X 3 interest rate scenarios.
Table 5.2. Investment requirements based on needs-based approach Study
Country Coverage
Policy Target, GDP Growth p.a.
Investment Requirements (in billion US-$ p.a.)
Begg, Danthine, Giavazzi, and Wyplosz (1990)
CEEC
7%
103-226°
Boote(1992)
CEEC
12.5%
259-628°
Collins and Rodrik (1991)
CEEC + Yugoslavia
7%
344-421b
Fitoussi and Phelps (1990)
CEEC
—
110-180°
Giustiniani, Papadia, and Porciani (1992)
CEEC
9%-10%
58.9°
Handler and Steinherr (1993)
CEEC
7%
169°
Holzmann, Thimann, and Petz (1994)
CEEC
12.5%
23-599*
° Total financing need. * Foreignfinancingneed. CEEC = Bulgaria, former CSFR, Hungary, Poland, Romania.
174
Stanley W. Black and Mathias Moersch
5.3
Investment financing Domestic savings
The ability of the CEECs to finance their investment needs depends primarily on their own capability of mobilizing domestic savings, as Adam Smith knew when he claimed that "As the capital of an individual can be increased only by what he saves from his annual revenue or his annual gains, so the capital of a society . . . can be increased only in the same manner" (Smith, 1937, Book 2). This is true despite the fact that the CEECs wish to import capital equipment from the West, for they may pay for such imports with increased exports to the West without necessarily borrowing large sums. The first question to address is the ability of the CEECs to increase domestic savings and channel them to domestic investment. Unfortunately, data on domestic savings in the CEECs is completely inadequate. However, it is possible to derive estimates of saving by residual from the National Income Accounts, as shown in Table 5.3, based on statistics from PlanEcon and IFS. According to these data, domestic saving capacities except in Poland have been rather strong but in the cases of Bulgaria, the Czech Republic, and Hungary have weakened recently. Furthermore, except for the Czech Republic the CEECs have been using increasing proportions of their total saving to finance government deficits rather than domestic investment. Thus, strengthened efforts to encourage private saving and to reduce government deficits appear necessary. In addition, positive real interest rates are essential for enabling the financial system to mobilize domestic savings. According to Figure 5.3, the Czech Republic, Hungary, Poland, and Bulgaria have after a number of years finally succeeded in raising their loan and deposit rates to positive levels.2 Romania and Slovakia still have some way to go in this regard. A simple regression equation3 illustrates the relationship between average savings rates of the CEEC economies and their average real deposit rates. Allowing for stronger savings behavior in Bulgaria and Romania, there is a clear positive relationship between real rates of interest and the average savings rates across countries. While the small number of observations is a limitation, this regression relating the savings rate to the real deposit rate and a dummy for Romania and Bulgaria does suggest that every additional 10 percentage points in the real interest rate adds .536 percent to the savings rate. Thus correction of the egregiously negative real interest rates in the CEECs should be expected to raise saving significantly. 2
3
Figure 5.3 displays only real loan rates. Real deposit rates are below real loan rates for all countries, but have developed similarly over time. S/Y = 26.738 + .0536 R + 10.256 D R2 = .96 (.0078) (1.59) D — 1 for Romania and Bulgaria
Table 5.3. Saving and investment (percentage of GDP) Investment
Gov't. Deficit
Net Foreign Investment
Saving
Hungary 1989
26.59
0.00
3.33
29.92
1990
25.39
0.06
2.61
28.06
1991
20.83
5.18
-2.91
23.10
1992
19.50
7.03
-1.17
25.35
1993
24.46
6.53
-9.04
21.95
Poland 1989
22.94
0.11
-1.73
24.27
1990
19.52
-0.42
1.38
16.23
1991
18.11
3.87
-3.10
18.13
1992
17.40
6.20
-3.79
17.62
1993
16.85
2.85
-5.72
13.99
1994
17.17
2.94
-2.88
17.23
Slovakia 1990
33.46
0.00
-9.24
24.22
1991
36.63
3.83
-8.15
32.31
1992
28.26
3.08
1.86
33.20
1993
25.16
6.83
0.59
32.58
Bulgaria 1989
33.11
0.00
-3.85
29.26
1990
30.40
0.00
-4.40
26.01
1991
27.93
0.00
-3.71
33.17
1992
20.51
6.01
-8.59
29.37
1993
15.81
11.89
-8.85
18.84
Czech Rep. 1991
30.87
2.07
6.95
39.88
1992
24.09
0.21
1.30
25.60
1993
17.02
-0.15
3.10
19.97
1994
25.61
-0.89
-2.48
22.24
Romania 1988
28.39
-5.83
9.82
32.38
1989
26.79
-8.23
2.73
21.29
1990
30.25
-0.93
-9.45
19.86
1991
28.05
-1.94
-3.93
22.17
1992
31.05
4.72
-7.67
28.10
1993
27.44
1.79
-5.78
23.44
Source: PlanEcon, 1994, except for Romania, IFS.
176
Stanley W. Black and Mathias Moersch Czech Republic
Figure 5.3: Real loan rates (percent).
Unfortunately, the time series data do not support the same conclusion (Figure 5.4). After the shock of extreme negative real returns in 1990-91, savings rates in the CEECs began to fall from their previously high levels. Eventually countries began to correct their interest rate policies. Thus the rising real interest rates have been accompanied by declining savings rates, as savings behavior appears to respond to interest rates with a considerable lag. Nevertheless, we prefer to believe the cross-sectional evidence. With positive real interest rates of around 5 percent, savings in the Visegrad countries could reach 27 percent of GDP, while it could reach 37 percent in Bulgaria and Romania. Financial intermediation In addition to the problem of achieving a sufficient domestic flow of saving, the task of transmitting it from savers to investors remains. This was handled by the state under central planning, but requires financial intermediaries under the market system. However, the financial intermediaries of the CEECs remain weak and underdeveloped, not only because of high inflation but also because of the portfolio of bad loans to state enterprises. In addition, the governments of many of the CEECs have resorted to borrowing from the banks to finance
Investment and its financing during the transition Bulgaria
177
Czech Republic
Figure 5.4: CEEC savings rates (percent of GDP).
budget deficits, further undermining the flow of funds to the private sector. Figure 5.5 shows the behavior of the flow of funds through the banking systems of Czechoslovakia, Poland, Hungary, and Romania over various periods since 1983. In 1991 and 1992, most credit in Czechoslovakia was going to stateowned enterprises. After the division of the country, credit flows in the Czech Republic shifted strongly toward the private sector as privatization proceeded rapidly. In Hungary, credit to the private sector was well supplied through 1990, financed significantly by foreign borrowing. Beginning in the late 1980s, the government began to take a larger share of bank financing, squeezing out the private sector. In Poland, the private sector only began receiving large credit flows after 1990, although the government has also been absorbing large amounts of credit since 1991. In Romania, excessive credit expansion to the private sector has helped keep the inflation rate high. Foreign savings Among the key factors enabling countries to mobilize domestic savings and attract foreign savings are the real interest rate and the real exchange rate. Coun-
178 12
Stanley W. Black and Mathias Moersch Hungary
Czech Republic
10 -
420 -2 -4 -6
Poland
^
Romania
75 -
50-
25 -
R
JUrf«Ji_ri.JllM
• I •
87 NFA m
DC •
JLi 111
GD |
Figure 5.5: Credit supply (percent of GDP). NFA = net foreign assets; DC = domestic credit; DCG = domestic credit to government; DCP = domestic credit to private sector; DCSE = domestic credit to state enterprises; GD = government deposits.
tries that need to attract foreign savings must maintain an appropriate relationship between the real exchange rate, which determines the rate at which capital needs to flow in (the current account), and the real interest and profit rates, which determine the willingness of foreign capital to flow in. To restore balance of payments equilibrium, the CEECs need to maintain an undervalued real exchange rate, or in other words a nominal exchange rate that is more depreciated than suggested by relative price comparisons. Figure 5.6 shows that most CEECs indeed have depreciated their real exchange rates since 1990. But in recent years, several have begun to appreciate back toward purchasing power parity. To the extent that this appreciation reflects a sustainable path toward equilibrium, there may be no problem. The initial depreciation would normally overshoot the new equilibrium exchange rate, and then be followed by a gradual appreciation as the current account deficit is reduced to a level that is sustainable based on the expected level of capital inflows. But if either capital flows are interrupted because of changed political or market conditions or if the current account worsens because of other factors, the path of gradual appreciation will become unsustainable.
Investment and its financing during the transition Bulgaria
179 Czech Republic
Figure 5.6: Real exchange rates (official/PPP).
Foreign direct investment and portfolio investment The most important sources of foreign capital for the future development of the CEECs are direct and portfolio investment. Three data sources are used to evaluate the flows of foreign direct investment (FDI) into the CEECs. First, balance of payments statistics published by the IMF, second, PlanEcon estimates, and third, a recent study by the OECD. FDI flows from 1989 to 1993 are summarized in Table 5.4. The three estimates show considerable variations. This is not too surprising, since a number of factors complicate a consistent assessment of FDI activity in the CEECs. First, some statistics comprise FDI commitments in addition to investments that have already been undertaken. Second, FDI estimates based on the capital account do not contain investments in kind and investments by resident firms directly or indirectly controlled by nonresidents. Third, reinvested profits are not treated uniformly. Fourth, it is sometimes impossible to distinguish between current account and capital account transactions. Keeping these data problems in mind, Table 5.4 allows a number of observations. The CEECs have drawn about $10 billion in FDI from 1989 until 1993. About half of this was attracted by Hungary, followed by the Czech Republic and Poland. As a percentage of GDP, the cumulative FDI flows amount to about
Table 5.4. Foreign direct investment in the CEECs (million US $) Source
89
90
91
92
93
Total
IMF"
—
4
56
42
55
157
PlanEcon*
—
4
56
43
56
157
OECiy
—
—
—
16
—
57
IMF"
257
187
586
1,073
—
2,103
OECDC
256
180
640
824
—
1,900
Czech Rep. PlanEcon'
—
120
510
983
516
2,129
Slovakia
—
52
80
72
120
324
Bulgaria
CSFR
Hungary
PlanEcon' IMF"
—
—
1462
1,479
2,339
5,280
PlanEcon*
—
354
1459
1,471
2,238
4,522
187
311
1460
1,465
—
3,423
-7
89
298
665
1,697
2,742
OECD< Poland
IMF"
—
10
117
284
580
991
OECiy
100
252
348
709
—
1,409
PlanEcoi/
Romania
IMP
—
-18
37
73
87
179
PlanEcon*
—
0
37
73
—
110
8
—
112
156
269
—
537
—
112
156
269
—
537
PlanEcon
OECir
° Net, IMF, Balance of Payments Statistics Yearbook, 1994. * Net, BOP reported. ' OECD, Foreign Direct Investment in Selected Central and Eastern European Countries and New Independent States, 1993. d Figures do not add up to total. ' Net, made in cash through the banking sector. 'Net, narrow Polish BOP definition. 8 Total foreign direct investment received, broad definition.
182
Stanley W. Black and Mathias Moersch
Table 5.5. Ratio of FDI to GDP in the CEECs Country
GDP in 93a (billion US-$)
FDI, sum 89-93* (million US-$)
Ratio of FDI to GDP
Bulgaria
33.9
157
0.0046
Czech Republic
75.0
2,129
0.0284
Slovakia
31.0
324
0.0104
Hungary
57.0
5,280
0.0926
180.4
2,742
0.0150
63.7
537
0.0084
Poland Romania a
At purchasing power equivalent, estimated. World Factbook, CIA, 1994. * From Table 5.4. Highest estimate for each country.
Table 5.6. FDI in selected EU countries Country
Greece 0
Portugal Spain
GDP 93a (billion US-$)
FDI, sum 89-93" (billion US-$)
Ratio of FDI to GDP
73.8
5.0
0.07
84.0
8.4
0.10
478.4
36.0
0.08
a
International Financial Statistics, IMF, March 1995. Local currencies converted into US$ at current nominal exchange rates. b Balance ofPayments Statistics Yearbook, IMF, 1994. c GDP for 1992.
9 percent in Hungary. Again the figures for the other countries are significantly smaller (Table 5.5). Table 5.6 puts the foreign investment activity in the CEECs in perspective by comparing it to that in selected EU countries during the same period. Greece attracted about as much FDI as Hungary. Portugal received almost as much FDI as Hungary and Poland combined. Spain attracted about $36 billion,4 far more than the CEECs combined. Relative to GDP, the investments amount to between 7 and 10 percent. These comparisons confirm the commonly held view that so far, with the possible exception of Hungary, the magnitude of FDI flows into the CEECs is disappointing (OECD, 1993; Halpern, 1995). 4
According to Gual and Martin (1995) this is a conservative estimate. They find that FDI estimates taken from the balance of payments are lower than those from registration data.
Investment and its financing during the transition
183
Table 5.7. Sources ofFDI, selected countries Country
Austria
FDI from 89-92* (million US-$)
Share (of countries listed)
1,374
0.26
Denmark
31
0.01
Finland
37
0.01
France
595
0.11
2,248
0.43
229
0.04
60
0.01
United Kingdom
147
0.03
United States
545
0.10
5,267
1.00
Germany Italy Sweden
Total a
International Direct Investment Statistics Yearbook, OECD, 1994b. Local currencies converted into US-$ at current nominal exchange rates.
Table 5.7 shows sources of FDI. It lists the so-called crescent countries that surround the CEECs, as well as France, the United Kingdom, and the United States. The majority of FDI originates in two of the crescent countries, Germany and Austria, who between 1989 and 1992 provided $3.5 billion US-$, roughly two thirds of all FDI. The United States and France also report substantial direct investments in the CEECs. Use of international financial markets by the CEECs only amounted to 0.4 percent of worldwide borrowing in 1994 and has virtually been restricted to Hungary and the former CSFR. The most important categories are bonds and syndicated loans. From 1990 to 1994 the CEECs borrowed $17.4 billion. More than 64 percent ($11.2 billion) went to Hungary (Table 5.8). The small amounts of borrowing by the CEECs are not expected to have an effect on world capital markets. The determinants of FDI are briefly discussed next. The most important factors are macroeconomic stability, market access, regulatory stability, and labor market conditions. The estimates and projections of the macroeconomic conditions presented in Table 5.9 are based on OECD (1994a). Overall conditions have improved markedly. After a steep decline in the early phase of transition, output is now expected to grow in all six countries. The economies of the Czech Republic, Slovakia, and Poland in particular are expected to grow strongly over the next years. Inflation rates have been reduced considerably, but remain too
184
Stanley W. Black and Mathias Moersch
Table 5.8. Borrowing (billion US $ and percentage) '90
'91
'92
'93
'94
434.9
536.0
609.7
818.6
954.5
CEEC
4.3
1.6
1.2
6.5
3.8
% CEEC
1.0
0.3
0.2
0.8
0.4
229.9
308.7
333.7
481.0
426.9
CEEC
1.7
1.5
1.3
5.8
2.4
% CEEC
0.7
0.5
0.4
1.2
0.6
CSFR
0.4
0.3
0.0
0.0
0.0
Czech Republic
0.0
0.0
0.0
0.7
0.4
Slovakia
0.0
0.0
0.0
0.2
0.2
Hungary
0.9
1.2
1.2
4.8
1.7
Other
0.4
0.0
0.1
0.1
0.1
124.5
116.0
117.9
136.7
202.8
CEEC
3.0
0.1
0.2
0.6
1.2
% CEEC
2.4
0.1
0.2
0.4
0.6
Hungary
0.0
0.1
0.2
0.3
0.8
Other
3.0
0.0
0.0
0.3
0.4
Overall Volume of Borrowing Worldwide
Bonds Worldwide
Syndicated Loans Worldwide
Source: OECD, Financial Market Trends, 60, February 1995.
high, especially in Bulgaria and Romania. Concern has grown, especially in Hungary, but also in Bulgaria and Poland, over large external imbalances. In Hungary, both the current account deficit and the budget deficit currently stand at 7 percent of GDP, and a significant fiscal tightening is expected in the next years. The Czech Republic on the other hand has a current account surplus and no budget deficit. The size of the domestic markets provides a strong incentive for a direct involvement in the CEEC countries. The combined population is 95 million people; Poland alone has 38 million citizens. Even more important may be the ad-
Table 5.9. Macroeconomic indicators '93
'94
'95
'96
Output Growth (pei'centage change over previous year) Bulgaria
-4.2
1.0
2.0
2.0
Czech Republic
-0.3
3.0
4.0
5.0
Slovakia
-4.1
3.5
4.0
5.0
Hungary
-2.3
2.5
0.5
1.0
Poland
3.8
4.0
5.0
5.0
Romania
1.3
1.0
1.5
2.0
Inflation (percentage change over previoiis year) Bulgaria
64.0
120.0
60.0
30.0
Czech Republic
20.8
11.0
9.0
7.0
Slovakia
23.2
16.0
12.0
8.0
Hungary
22.5
20.0
17.0
12.0
Poland
35.3
30.0
23.0
18.0
295.0
130.0
45.0
35.0
Romania
Budget deficit (percentage of GDP) 11.0
7.0
6.0
6.0
Czech Republic
0.0
0.0
0.0
0.0
Slovakia
6.8
5.0
4.0
3.0
Hungary
6.0
6.5
6.0
5.0
Poland
2.9
4.0
3.5
2.5
Romania
0.7
3.5
2.0
2.0
Bulgaria
Current account balance (in billion US$) -1.4
-1.3
-0.8
-0.7
0.3
0.6
0.4
0.2
Slovakia
-0.7
-0.4
-0.4
-0.3
Hungary
-3.5
-3.5
-2.8
-2.5
Poland
-2.3
-1.3
-1.5
-1.9
Romania
-1.2
-0.7
-0.5
-0.2
Bulgaria Czech Republic
Source: OECD Economic Outlook, 56, 1994a. OECD Secretariat estimates and projections.
186
Stanley W. Black and Mathias Moersch
vantageous geographical location of the CEECs, which allows companies to set up an export base toward both the West and the East. The potential of exports is currently limited, however, by two factors. First, the uncertain situation in the FSU limits trade with the East. Second, in the West, the Europe Agreements currently restrict access to markets in the EU. The legal and regulatory framework is discussed extensively in OECD (1993). Considerable progress has been made toward regulation conducive to FDI, but a number of obstacles remain. Some industrial sectors continue to be barred from foreign investment. Repatriation of profits and foreign wages remains incomplete. The lack of guarantees against unknown liabilities caused by previous owners, especially with respect to environmental damages, creates uncertainty. In the Czech Republic, foreign firms have only limited access to the privatization process. Finally, there are restrictions on the structure of ownership and the conduct of branch business. Labor market conditions appear to be favorable in the CEECs. Winters and Wang (1994) find that the CEECs are ranked below Western Europe but above Southern Europe with respect to labor force skills and measures of educational achievement. They conclude that the CEECs have the potential to become a producer of sophisticated industrial goods. In addition the labor cost gap between the EU and the CEECs remains significant (Gual and Martin, 1995). At least part of Hungary's initial success in attracting foreign capital can be attributed to the openness with which privatization was conducted. OECD (1993) estimates that by the beginning of 1993,40 percent of FDI into Hungary was linked to privatization. In the Czech Republic, on the other hand, preferential treatment was given to domestic citizens and employees in the privatization process. Consequently foreign participation has remained limited. Szanyi (1994) reports on a recent survey of firms investing in Hungary. Survey participants were asked to compare a number of countries with respect to political stability and economic conditions, potential domestic market, infrastructure, FDI incentives, and cultural distance. In the survey, Hungary compares favorably with Poland and the Czech Republic, but is not competitive with EU countries. It outscores the Czech Republic in most categories and Poland in every category except for the size of the domestic market. A comparison with selected other countries, namely Ireland, Italy, Austria, Spain, Portugal, and Turkey, shows that the CEECs are not regarded to be in a position to compete with them. Scores in all categories are substantially higher for the nonCEEC countries. In sum, it appears that good labor market conditions and an improved regulatory environment make the CEECs an interesting target for FDI. A stabilization of the economic and political situation in the FSU and further opening of EU markets would help to make the CEECs more attractive as a base for exports. Finally, attainment of macroeconomic stability remains of prime importance, as indicated by the recent worries about developments in Hungary. In light of the overall progress over the past years toward a stable market
Investment and its financing during the transition
187
economy, the small amount of FDI is somewhat surprising. It may of course be that investors need more concrete and lasting evidence of stability and that FDI activity will pick up in future years. 5.4
The four stages of transition
The four stages of transition outlined in Section 5.2 are estimated in four steps. First, existing data show the size of the initial drop in output in each of the CEECs, which therefore becomes the potential for the recovery phase. Second, estimates are made of the initial inefficiency and the initial stock of capital and labor prior to the shock. Removal of the existing inefficiency then provides an estimate of the potential for the reform phase. Third, the initial structure of the economy in the CEECs is compared with the typical structure in member countries of the EU. Reallocation of capital and labor from the initial structure to the desired structure provides an estimate of the potential gains from the restructuring phase. Finally, accumulation of new capital is estimated based on domestic savings capacities and potential foreign investment from the West described in Section 5.3. Here the recovery, reform, and restructuring effects are modeled separately and are assumed to occur in a timeless fashion. In Section 5.5 the effects are combined and simulated over the ten-year period 1992-2002. Collapse Table 5.10 indicates the scale of the recorded collapse in output in the six CEEC economies during the period 1990-92, roughly 20 percent. Three exogenous shocks caused the collapse in output; first, the change in preferences that came with the end of central planning; second, the end of ruble-based trade among the members of the CMEA in January 1991; and third, the loss of purchasing
Table 5. W.Real GDP (1989 = 100) Year
Bulgaria
Romania
Czech Rep.
Poland
Hungary
Slovak Rep.
1989
100
100
100
100
100
100
1990
90.9
92.6
98.8
88.4
96.5
97.5
1991
80.3
78.6
84.8
81.7
86.9
83.4
1992
75.7
68.0
78.8
83.8
82.5
77.5
1993
72.5
68.0
78.5
87.0
80.9
74.3
1994
1\A
68L0
79.7
90.9
8L7
74.3
Source: IMF, World Economic Outlook, October 1994b. 1994 estimated by IMF.
188
Stanley W. Black and Mathias Moersch
Table 5.11. CEEC capital stocks by country, 1992 baseline CEEC
Czech
Slovak
Hungary
Poland
Bulgaria
Romania
V (billion US-$)
418.70
78.73
31.48
54.23
159.27
43.13
55.47
L (million US-$)
41.76
5.09
2.22
4.65
15.94
3.27
10.58
V/L (thousand US-$)
10.03
15.47
14.16
11.67
9.99
13.18
5.24
K (billion US-$)
1310.24
208.99
88.32
134.92
496.03
124.23
259.22
K/L (thousand US-$)
31.38
41.10
34.73
29.03
31.11
39.06
24.50
g
0.62
0.74
0.71
0.69
0.62
0.69
0.46
/
0.258
0.396
0.380
0.313
0.257
0.335
0.131
Note: V= output, L = employment, K = capital, g = capital efficiency,/= labor efficiency. Source: Authors' calculations.
power due to the sharp inflation that came with the decontrol of prices and devaluation of exchange rates. The first factor combined with convertibility of domestic currency for current account transactions allowed a major shift in demand away from previously produced output. The second factor removed the purchasing power of previous export markets in the CMEA. The third factor reduced the purchasing power of domestic residents with excess money balances and limited the real availability of bank credit to firms. The recorded drop in output may, of course, be overstated, due to the shift of workers into the unrecorded or secondary economy. The initial position Estimation of the initial position of the CEEC economies is rather difficult, given the substantial changes in economic structure and organization that have taken place since 1989. We follow Boote (1992) and McDonald and Thumann (1990) in estimating the capital stock of the CEECs by inverting a production function V(K,L) to obtain K(V,L). The parameters of the production function to be inverted are established from Western European data for 1992, on the assumption that the CEECs are seeking to approach a production structure similar to the EU. Table 5.11 provides a baseline estimate for the CEECs in 1992. According
Investment and its financing during the transition
189
to the baseline, total output of $419 billion evaluated at purchasing power parity (PPP) exchange rates is produced using a work force of 42 million persons5 and a capital stock of some $1300 billion. The efficiency with which capital is used is assumed to be 62 percent of that in the West on average, while labor's efficiency, taking account of the lower capital/labor ratio in the East, is only 26 percent.6 Table 5.11 also provides estimates of the distribution of the baseline capital stock, output, and labor over the individual CEEC economies. Output per worker varies from $15,000 in the Czech Republic to $5,000 in Romania, and the capital/labor ratio from $41,000 per worker in the Czech Republic to $25,000 per worker in Romania. Recovery The question is to what extent the reform process can permit the recovery of output from the depressed levels to which it has fallen. Clearly none of the three exogenous shocks is going to be reversed. Therefore recovery depends on the adaptation of the CEEC economies to the new conditions they impose, as well as the reincorporation of the workers in the secondary economy. The main microeconomic elements of that adaptation are, first, reorientation of output to the new pattern of demand, second, development of new external markets (and recovery of previous markets), and third, reestablishment of domestic purchasing power through improvements in productivity and real wages, while maintaining monetary and fiscal stability. Obviously, privatization of the means of production plays a very large role in accomplishing all three of the microeconomic objectives, as new management (and often new firms) are required to achieve the changes in product mix, marketing, and organization of production that can find and serve markets. Without addressing the privatization issue in this paper, let us merely note that the process is occurring at a rapid rate in many of the CEEC economies, but is being delayed in several important cases. With respect to the reemployment of the human and physical resources that became unemployed as a result of the collapse of Communism, the process depends upon their adaptability, which in most cases is inversely related to age, and the availability of alternative sources of income for older workers who wish to retire. The standard assumption of most analysts (e.g., Boote, 1992) has been that all workers and capital are reemployed, although some assume that all of the capital is discarded (Fitoussi and Phelps, 1990). Our working assumption will be eventual full reemployment of resources. Table 5.12 provides estimates of the increase in output that could be obtained 5
6
This is equal to the labor force minus recorded unemployment and most likely overstates the total labor force. The alternative of taking recorded employment, on the other hand, would arrive at a number that is too small. See Appendix A for the methodology. For labor efficiency,/represents the entire expression on the right-hand side of Appendix A equation (2).
190
Stanley W. Black and Mathias Moersch
Table 5.12. CEEC output with full employment CEEC
Czech
Slovak
Hungary
Poland
Bulgaria
Romania
V (billion US-$)
431.24
77.90
32.66
56.11
167.24
45.30
55.92
L (million US-$)
44.00
4.99
2.38
4.95
17.38
3.58
10.72
V/L (thousand US-$)
9.80
15.62
13.71
11.35
9.62
12.65
5.22
K (billion US-$)
1310.24
208.99
88.32
134.92
496.03
124.23
259.22
K/L (thousand US-$)
29.78
41.90
37.08
27.28
28.53
34.68
24.18
g
0.62
0.74
0.71
0.69
0.62
0.69
0.46
/
0.254
0.398
0.354
0.309
0.252
0.328
0.131
Source: Authors' calculations.
by reemploying the unemployed labor resources with the capital stock as estimated in Table 5.11. Only about a 3 percent increase in output is estimated. This means that the recovery phase is only of limited importance in the overall growth process. In fact, the other effects yield significantly larger increases in output. Recorded unemployment falls from 10 percent of the labor force to 5 percent, although in the Czech Republic unemployment is assumed to rise slightly. Most of the gain in output comes from Poland, where unemployment has been particularly high. Reform Simulation of the reform process involves reducing the degree of inefficiency in the production structure of the CEECs relative to Western practices. No additional resources are added, but those available are now assumed to be utilized so as to produce with the same degree of productivity as in the West. The assumptions underlying the adjustment of the efficiency coefficients are described in Appendix A. Table 5.13 provides the estimated results. Output rises from $431 billion to $972 billion, more than double what can be produced with the initial low levels of efficiency. Poland and Romania, with the largest work forces and low initial productivity, are the source of the largest potential gains. This is without adding any more physical capital or labor and does not involve
Investment and its financing during the transition
191
Table 5.13. CEEC output with full efficiency and full employment CEEC
Czech
Slovak
Hungary
Poland
Bulgaria
Romania
V (billion US-$)
971.70
131.22
58.84
104.43
375.62
85.51
212.75
L (million US-$)
44.00
4.99
2.38
4.95
17.38
3.58
10.72
V/L (thousand US-$)
22.08
26.31
24.70
21.12
21.61
23.87
19.85
K (billion US-$)
1310.24
208.99
88.32
134.92
496.03
124.23
259.22
K/L (thousand US-$)
29.78
41.90
37.08
27.28
28.53
34.68
24.18
g
1.00
1.00
1.00
1.00
1.00
1.00
1.00
/
0.662
0.728
0.701
0.648
0.655
0.688
0.631
Source: Authors' calculations.
restructuring the output mix. Productivity more than doubles, because both capital and labor are used more efficiently. Restructuring The third stage of the transition process consists of the reorientation of the production structure away from heavy industry toward services and light industry. As explained earlier, this can be expected to yield additional productivity gains beyond those already discussed, for three reasons. First, responding to Western prices and producing what consumers actually want will in itself generate a more highly valued output, as illustrated in Figure 5.2. Second, the capital/labor ratio in most services is distinctly lower than in heavy industry, thus allowing a better use of the scarce factor. Third, there is some evidence that the elasticity of substitution between capital and labor is lower in heavy industries than in services and light industry, thus enabling an easier process of structural transformation as capital is released from the less substitutable sector. To estimate the impact of structural reform, we have constructed a twosector model of the CEEC economies, estimating the parameters of each sector from the available data from the EU.7 For example, in the EU 65 percent of 7
See Appendix B for the methodology.
192
Stanley W. Black and Mathias Moersch
output is in the service sector, produced by 61 percent of the work force and 55 percent of the capital. Thus, the capital/labor ratio in services is only 78 percent of that in the goods sector. Using these data to set parameters for production functions in each sector allows us to trace out the production possibilities for the CEECs, given their capital and labor resources. Since the CEECs began the transition process with only 40 percent of their output in the service industries, there is a significant potential for gain due to restructuring. Using the production possibility curve together with a Cobb-Douglas utility function based on Western preferences to estimate the increases in welfare due to restructuring yields about a 16 percent increase, including both consumers' and producers' gains.8 Capital accumulation The success of the capital accumulation phase depends on the ability of the CEECs to generate domestic savings as well as to attract foreign capital. Beginning with the fundamental identity
net/
=SP
+ S'+ Ss
-hK_v
net investment is calculated as the sum of domestic private saving and foreign saving, less the government deficit, less depreciation of the existing capital stock. We begin from the historical data for 1992 in Table 5.3. Based on the analysis of Section 5.3, domestic savings are projected to rise linearly from 23.8 percent of GDP to 29.3 percent, closer to the historical norm, assuming that real interest rates are raised to a positive 5 percent. Current account deficits are a source of foreign saving and are projected to decline from 2.72 to 1.27 percent of GDP, on the assumption that real exchange rates appreciate only gradually. Government deficits are projected to fall from 4.74 percent of GDP to zero. Depreciation is calculated at 6 percent of the capital stock. Together, these assumptions give us net investment over the ten-year transition period from 1992 until 2002. 5.5
The transition process
Relying on the analysis provided above, we can estimate the feasible growth of the CEEGs over the ten-year period 1992-2002 by combining investment with the other three elements of the transition process, recovery, reform, and restructuring. The growth path over the ten-year period is calculated, first on the assumption that all elements of the transition process contribute to growth, and then under the more pessimistic assumption that various elements fall short. Table 5.14 gives the various simulation results. The high and low estimates for output per person compared with the expected 2.5 percent growth in the EU 8
See Appendix C for the methodology.
193
Investment and its financing during the transition
Table 5.14. Alternative CEEC growth scenarios, 1992-2002 V (billion US-$)
V/L (thousand US$)
Annual Growth
in V(%)
Annual Growth in V/L (%)
418.70
10.03
1450.57
32.98
13.23
12.64
916.33
20.84
8.15
7.59
Only 1/2 of restructuring gains
1326.70
30.17
12.22
11.64
No gain in employment
1402.91
33.60
12.85
12.85
No improvement in budget deficit
1390.08
31.61
12.75
12.16
Only 1/2 of growth in savings rate
1415.45
32.19
12.95
12.37
703.13
16.84
5.32
5.32
1992 Optimistic scenario Only 1/2 of efficiency gains
Pessimistic scenario Source: Authors' calculations.
are also displayed in Figure 5.7. The following assumptions are made. All changes to ratios are projected to occur linearly over time. The unemployment rate is projected to decline from 10 percent in 1992 to 5 percent in 2002. Efficiency is assumed to improve linearly over time from the estimated initial level to full equivalency to the average EU level in 2002. Restructuring of the economy gradually adds an additional 16 percent to the value of output by the end of the period. Capital accumulates according to the assumed savings behavior, less depreciation of 6 percent per year. Labor forces are assumed to remain constant. Output under these assumptions rises from $10,000 per worker in 1992 to $33,000 per worker, as the capital stock rises from $1310 billion to $2409 billion. It is noteworthy that most of the capital accumulation is financed by domestic saving. Of the $1099 billion increased capital stock, $950 billion comes from net domestic savings and only $149 billion from foreign saving. Foreign capital supplies 13.5 percent of the total. This requires an average capital inflow of $15 billion per year over the ten-year period, which is at least twice as high as recently observed inflows. Output per worker rises from 25 percent of the EU level in 1992 to 63 percent by 2002. These results must be described as somewhat optimistic. Alternative simulations have been run with less favorable assumptions about the transition process. The key factor is the speed with which the CEECs approach Western productive efficiency levels. Half the rate of efficiency improvement takes away half of the productivity gains, leaving output per worker in 2002 only at 40 percent of the projected EU level, or $20,000 per worker. Losing half the gains from restructuring cuts productivity by 8.5 percent. Elim-
194
Stanley W. Black and Mathias Moersch
56
-
48
-
40
-
32
-
24
-
16
-
CEEC, fast growth CEEC, slow growth
1992
1994
1996
Figure 5.7: Output per person in CEECs and EU.
Table 5.15. Productivity growth rates, 1992-2002
Optimistic
Pessimistic
EU
2.5
Czech Republic
9.2
4.9
Slovakia
10.6
5.8
Hungary
10.6
5.3
Bulgaria
11.7
6.2
Poland
11.8
5.5
Romania
20.3
10.0
Source: Authors' calculations.
1998
2000
2002
Investment and its financing during the transition 80
-
70
-
60
~
50
-
40
~
30
195
Cze Slo Bui - Hun Pol
20
-
10
-
Rom 1
I 1992
1994
I 1996
2000
2002
Figure 5.8: Convergence of CEEC to EU productivity.
inating the reduction in unemployment lowers output by 3 percent, but raises output per worker. Failure to restrain government budget deficits cuts productivity another 4 percent. Finally, if only half the improvement in the savings rate occurs, output per worker is reduced by 3 percent. Including all of the negative factors together reduces output per worker in 2002 to only $16,840, or about half the level projected under the standard assumptions. Simulations for individual countries are presented in Table 5.15 and Figure 5.8. With the exception of Romania, growth rates for all countries fall within a narrow range. In the optimistic scenario they vary from 9.2 percent for the Czech Republic to 11.8 percent for Poland. Romania's growth of over 20 percent per year comes from the artificial assumption that it eliminates its extremely unfavorable productivity gap within the same ten-year period as the other countries, combined with its assumed strong savings behavior. A more reasonable interpretation would suggest that it will take Romania considerably longer to close the productivity gap than the other CEECs. In the pessimistic scenario growth varies from 4.9 to 6.2 percent, while Romania still achieves a growth rate of 10 percent. These projections clearly show the ability of the CEECs to make the transition. They also show the requirements for achieving strong positive growth. From a macroeconomic point of view, containment of budget deficits, maintenance of positive real interest rates, and appropriate real exchange rates are essential to permit the mobilization of domestic and foreign savings sufficient to finance the requisite capital accumulation. From a microeconomic viewpoint,
196
Stanley W. Black and Mathias Moersch
continued restructuring of the economy away from heavy industry toward services and light industry, continued reorganization of production and management methods, and continued retraining of workers must go hand in hand with new investment. Also, improvement in the process of financial intermediation is essential to permit the effective channeling of savings to investors. The role of foreign investment is not to supply the bulk of savings to the CEEC economies, which is infeasible in any case. Rather it is to supplement domestic savings and, more importantly, to provide the catalyst for industrial restructuring that foreign investors bring, in the form of new management methods, improvements in marketing, and effective servicing of customers in the new, market-oriented economies. 5.6
Conclusions
We believe that our study throws light on the determinants of growth in the CEEC countries and the conditions under which these economies will be ready by the year 2002 to enter the EU. The most important source for future growth is the elimination of inefficiencies in the production process. If only half of the assumed efficiency gains are obtained, for example, output growth is reduced from 13.2 percent to 8.2 percent. The importance of this effect is also found in previous studies. Restructuring of the economy away from manufacturing toward services also promises sizable gains in the value of output. A return to low levels of unemployment also adds modestly to output gains. The ability to accumulate capital depends largely on the mobilization of domestic savings. This in turn depends on stable macroeconomic conditions, namely a return to stable and positive real rates of interest, but also the ability of the government to control budget deficits. Foreign sources are also important, but will not be sufficient to sustain the needed buildup of the capital stock. Our optimistic scenario implies growth of 13 percent per year, an unusually high rate. This growth rate allows the CEECs to raise their output per worker from 25 percent to 63 percent of the EU level by 2002. Such a strong convergence in output clearly would facilitate membership of the CEECs in the EU. On the other hand, achievement of the pessimistic outcome of only 33 percent of EU productivity would presumably leave the CEECs far short of eligibility for EU membership. The difference between these two outcomes depends primarily on domestic policies in the CEEC economies, relating to interest rate and exchange rate policy, tax and expenditure policy, reform of financial intermediaries, restructuring of industry, reorganization of production, and privatization. Western European countries must be prepared to make room in their markets for the products of rapid growth in the CEECs, just as they may hope to take advantage of the strong markets the CEECs will provide for their own products and services. We are unable within the scope of this study to address such issues as the impact of the transition process in agriculture in the CEEC economies. Nor do
Investment and itsfinancingduring the transition
197
we examine the impact of the EU's Common Agricultural Policy on their agricultural economies. These and other limitations will be easily recognized. Our optimistic estimates of the possibilities for the future are offered with some caution. Nevertheless, we believe that even if the future turns out less favorably, this paper helps to identify the key policy issues whose resolution will determine the success or failure of the transition process. Appendix A. Determination of capital and output The basic assumption, following McDonald and Thumann, is that Eastern output is produced with the same production function as Western output, but with less capital per worker and with both capital and labor used inefficiently. A constant elasticity of substitution (CES) production function (1) is adopted, with elasticity of substitution a = 1/(1+X) equal to one half.9 V = [AL * L5~ A + AK * KS~X]-(VX)
(1)
The parameters AL and AK are chosen by reference to output, capital, and labor used in the EU in 1992, assuming that the EU is producing on its production frontier with full efficiency. In Eastern Europe, it is assumed that the output of 1992 is produced using the actual labor force and an imputed capital stock, both of which are used inefficiently. Capital services are represented as KS = g * K, while labor services are represented as
The coefficients g and h correspond to the assumed efficiency levels with which capital and labor are utilized. The third factor is based on the theory that a lower relative capital/labor ratio in the East lowers efficiency of labor use compared to that in the West. It is assumed that g = h and that/ = 0.5. Initial values of g are chosen to be consistent with the capital stock estimates of Boote (1992). Table 5.11 provides baseline estimates for 1992 for the CEEC countries as a group and individually. Tables 5.12 and 5.13 sequentially adjust unemployment to 5 percent and the efficiency parameter g to unity. The optimistic scenario in Table 5.14 starts from the baseline capital stock figures for 1992 and adds net saving to the capital stock, assuming a 6 percent depreciation rate, over the period 1992-2002. The saving rate rises linearly from the 1992 level of 23.8 percent to 29.3 percent. The current account as a percent of GDP falls linearly from the 1992 level of 2.74 percent to 1.27 per9
This figure is consistent with numerous studies of output behavior, such as Torres and Martin (1990).
198
Stanley W. Black and Mathias Moersch
cent. The government deficit as a percent of GDP falls linearly from the 1992 level of 4.74 percent to zero. The unemployment rate falls from 10 percent to 5 percent. Finally, structural change from industry to services is projected to gradually add up to 16 percent to the value of output by the end of the period. Appendix B. Two-sector model Assume there are two sectors, Manufacturing and Services, each producing output with homogeneous capital and labor using CES production functions with elasticity of substitution a = 0.5 as in Appendix A. Using the allocation of capital and labor to each sector in the European Union, we estimate the parameters AL and AK for Manufacturing and the analogous parameters BL and BK for Services. From the first-order conditions for firms' demand for factors, the slope of the isoquant in Manufacturing is ^ \2
w
AK\L
Equating (1) to the equivalent expression for the slope of the isoquant in Services tofindthe equilibrium allocation of capital and labor between the two sectors, / ^ £
2
(2)
The multiplicative factor in (2) is greater than unity if Manufacturing is more capital-intensive than Services. The slope of the production possibility frontier is then 3
dY,
A
L
\ y 1 /L 1 ;
Pl
Appendix C. Welfare analysis Valuing the initial constrained output C reflecting planners' preferences x = (JCpjy at the set of Western prices p* = (px*,p2*) consistent with the choice of the optimal point x at D would show point C to be worth considerably less than D. Thus restructuring the economy toward the preferred composition of output will itself lead to a gain in output, measured properly in international prices, equal to the producers' efficiency gains p* • (x — x). In addition, there is a gain in consumer welfare as consumers move from point C to point D. The price line p 2 tangent to the indifference curve passing through point C defines the expenditure level e(p2,U0) where e(p,U) is the expenditure function associated with utility function U(xvx2). The gain in consumers' welfare attributable to the ability to choose their preferred bundle of
Investment and its financing during the transition
199
goods can be measured by the increase in expenditure from e(p*, Uo) to ^(p*,f/j) = p* • x, where Ul is the utility level associated with the indifference curve tangent to the Western price line p* that passes through the initial constrained production point x at C Let x' = (xx\x2r) = e (p*, Uo) be the point on Uo chosen when prices are p* and x" = e p (p*, Ux) be the point on Ul chosen when prices are p*. Then the consumer gain is p* • (x" - x'). The total gain in welfare due to restructuring can then be measured as the sum of the consumer gain and the producer gain, or p* • (x - x'). REFERENCES Begg D., J. P. Danthine, F. Giavazzi, and C. Wyplosz (1990). "The East, the Deutschmark and EMU," in Monitoring European Integration: The Impact ofEastern Europe, London: Centre for Economic Policy Research, 31-69. Boote, A. R. (1992). "Assessing Eastern Europe's Capital Needs" IMF Working Paper 92/12 February. Borenstein, E., and P. J. Montiel (1991). "Savings, Investment, and Growth in Eastern Europe," IMF Working Paper 91/12, February. CIA (1994). World Factbook, Washington, D.C.: U.S. Government Printing Office. Collins, S.M., and D. Rodrik (1991). Eastern Europe and the Soviet Union in the World Economy, Institute for International Economics, Policy Analyses in International Economics, May. Fitoussi, J. P., and E. S. Phelps (1990). "Global Effects of East European Rebuilding and the Adequacy of Western Saving: An Issue for the 1990s," Rivista di Politica Economica, December. Gual, J., and C. Martin (1995). "Trade and Foreign Direct Investment with Central and Eastern Europe: Its Impact on Spain," in R. Faini and R. Portes (eds.), European Union Trade with Eastern Europe: Adjustment and Opportunities, Centre for Economic Policy Research, 167-200. Guistiniani, A. F, F. Papadia, and D. Porciani (1992). "Growth and Catch-up in Central and Eastern Europe: Macroeconomic Effects on Western Countries," Princeton University Essays in International Finance 186, April. Halpern, L. (1995). "Comparative Advantage and Likely Trade Pattern of the CEECs," in R. Faini and R. Portes (eds.), European Union Trade with Eastern Europe: Adjustment and Opportunities, Centre for Economic Policy Research, 61-85. Handler, H., and A. Steinherr (1993). "Capital Needs and Investment Financing in Eastern Countries," in D. E. Fair and R. Raymond (eds.), The New Europe: Evolving Economic and Financial Systems in East and West, Norwell, MA: Kluwer Academic Publishers, 363-89. Holzmann, R., Ch. Thimann, and A. Petz (1994). "Pressure to Adjust: Consequences for the OECD Countries from Reforms in Eastern Europe," Empirica, 1-55. IMF (1995). International Financial Statistics, March, Washington, D.C. IMF (1994a). Balance of Payments Statistics Yearbook, Washington, D.C. IMF (1994b). World Economic Outlook, October, Washington, D.C. McDonald, D., and G. Thumann (1990). "Investment Needs in East Germany," IMF Occasional Paper 75, 71-7. McKibbin, W (1991). "The New Europe and Its Economic Implications for the World Economy," Brookings Discussion Papers in International Economics 89, August. OECD (1995). Financial Market Trends, 60, February, Paris. OECD (1994a). OECD Economic Outlook, 56, Paris.
200
Stanley W. Black and Mathias Moersch
OECD (1994b). International Direct Investment Statistics Yearbook, Paris. OECD (1993). Foreign Direct Investment in Selected Central and Eastern European Countries and New Independent States, Paris, OECD/DG(94)3. PlanEcon (1994). PlanEcon Report, various issues, Washington, D.C. Sachs, J. (1991). "From Rubles to Reform," Washington Post, May 12, Outlook Section, C3. Smith, A. (1937). An Inquiry into the Nature and Causes of the Wealth of Nations, ed. Edwin Cannan, New York: Modern Library. Szanyi, M. (1994). "A Transitional Assessment," Institute for World Economics working paper 42, HAS, November. Torres, R., and J. P. Martin (1990). "Measuring Potential Output in the Seven Major OECD Countries," OECD Economic Studies 14, 127-49. Winters, L. Alan, and Z. Wang (1994). Eastern Europe's International Trade, Manchester: Manchester University Press.
Comment Douglas Todd
The paper by Stanley Black and Mathias Moersch is an adventurous and readable effort to bring some orthodox economic analysis to bear on what is clearly a major structural and hence nonmarginal adjustment problem - moreover, a structural adjustment that is considered as taking place over a relatively short time scale. For this discussant, reading the paper induced a distinct air of nostalgia since as a then civil servant in the UK, it reminded me of official and other debates concerning the British Labour Government's National Plan of the mid-1960s, when what now seem to be long lost expressions such as "investment requirements" and "incremental capital/output ratios, or ICORs" and the like readily tripped off the tongue. I find myself, therefore, in a kind of time warp but one which is nonetheless interesting for that. On this occasion, Rip Van Winkle has awakened to find that not much has changed! The basic starting point of the paper is to examine what the feasible sources of capital accumulation are likely to be and then see what in principle can be achieved with these resources. As the paper notes, this is in contrast to the more typical approach, which is to choose a target level, or growth rate for real output, translate this into an investment requirement, and then make some inferences about what is implied for the supply of savings. As it is structured, because the paper reads almost as if it is in two parts - a real side and a financing side - I want to concentrate my remarks primarily on the physical productive side of the exercise, namely, the nature and scale of the adjustment process across the six CEEC members, as envisaged by the authors and, in particular, what this implies for productive performance set in terms of the starting position of the individual CEEC countries with respect to both themselves and that of the average of EU member states. I will also make a few comments on some other specific issues that are raised in the paper. One point that should be made here in passing at the outset is that the EU currently is experiencing a number of shocks and stresses, making its use as a basis for comparison problematic. Against the background of relatively depressed economic conditions, the process of convergence of the member states and the path to a monetary union via the formal requirements of the Maastricht 201
202
Douglas Todd K/V
2.5 .
a
. -B CEEC Baseline Frontier
0.1
Figure 5.9: 1992 baseline.
Treaty is proving to be difficult for even some of the stronger participants. Although this is really a separate subject and is mentioned only briefly later, it is raised nevertheless, because it should be borne in mind as a part of the background against which the serious endeavours of the Central European Countries to raise their economic performance and their aspirations to be a part of a future enlargement of the EU should be assessed. To see more clearly the relative nature of the magnitudes involved, I found it helpful to cast the figures for real output and factor inputs of capital stocks and labor, which are given and derived by Messrs Black and Moersch in Tables 5.11 to 5.13 of their paper, in a slightly different way. For each of the six CEEC countries, the observations on labor and capital per unit of real output {LIV and K/V) respectively, are shown for the 1992 baseline year in Figure 5.9. Although we have only a very small number of points, one can imagine a rough form of technical efficiency frontier for this particular group of countries, as shown by the simple convex hull in Figure 5.9. The fact that all of these countries are emerging from a long period of regulation and control under broadly similar regimes, as compared with the Western industrialized countries, is at least one justification for such a comparison. In other words, it is assumed that there is enough homogeneity to proceed on this basis. Looking at Figure 5.9, we observe that as of 1992, there is a wide disparity in productive performance, with the Czech Republic and Hungary making up what can be thought of here as a crude sort of "best practice" frontier within the CEEC bloc. We see also that Poland and particularly Romania, are located at some distance from this 1992 baseline frontier. Simple radial efficiency indices computed with respect to the implied frontier suggest that relative to "best practice" Czech Republic and Hungary, Bulgaria has a technical efficiency deficit of about 10 percent, Poland 19 percent, with Romania a long way behind, being only one half as efficient as the two leaders. In other words, Romania for example, being the extreme case, would have to improve its productive efficiency by around 50 percent to emulate the performance of Hungary and the
203 CEEC (1 992) K/L = 31.38
H
CEEC 1992 Baseline CEEC Full Employment
Figure 5.10: Stylized view of the stages of transition.
Czech Republic simply as they are now. Poland requires an efficiency uplift of about 20 percent. By any reasonable standards, these are large numbers indeed. Even if they are less than half right, when the baseline alone for the exercise is interpreted in this way, the scale of the structural transition leading to full convergence by the year 2002 for the group as a whole under investigation by the two authors becomes that much more transparent and, one might venture to say, optimistic. This is after all, only the starting point. What Black and Moersch do next is to postulate a shorter term recovery phase that brings the six countries to full employment. They then estimate what feasible reforms could yield in efficiency gains, given the existing stock of physical capital, labor, and so on. Best utilization of all resources at full employment would bring the six countries to a state of "full efficiency," that is to say, technical efficiency at the old set of market prices. In terms of the above approach, this means that all countries in the group are brought to lie on the same technical efficiency frontier. This is stage II, being the movement from point B to point C in Figure 5.1 of their paper. Indeed, the figures in Table 5.13 of their paper yield precisely this result. No country alone is the technical leader since all have a radial technical efficiency score of exactly unity. The only thing which differs is the capital/labor ratio, the Czech Republic having the highest and Romania the lowest. Not only do Hungary and the Czech Republic greatly increase their productive performance with output per head growing at an annual average rate of around 10 percent, but the other four countries must do much better than this and, when appraised in terms of productive efficiency, are seen to catch up fully with the two leaders. This is what is implied by "full efficiency" in all six countries. In this sense, within the bloc, full economic convergence is assumed to have been achieved. Figure 5.10, which is not drawn to scale, provides a stylized view of these two stages. As is noted by the authors, there is little difference between the 1992 baseline and the initial recovery movement to full employment of resources, with the two frontiers being defined by the Czech Republic and Hungary. We
204
Douglas Todd
then have the calculated huge reform-induced adjustment that moves the CEEC frontier in the direction of the EU average and where all CEEC countries are seen as having achieved full efficiency by the end of the decade. Of course, as the paper makes clear, the stages considered are not strictly sequential; the reform phase, which embodies the move toward market prices, will be taking place also. However, it is clear that the big gains are taken to come from initial shakeout, particularly in the use of existing capital stocks. In the stylized move from B to C, relative factor prices remain the same and there is no change in the capital/output ratio and hence no shift around the production frontier. The gains are therefore of an X-efficiency type. With all of these efforts and successes assumed over a decade, the six countries naturally are still placed a very long way behind the EU average. In considering the scope for the restructuring of the production mix, the authors place a good deal of faith in the ability of the services sector to absorb capital released from the more out-dated heavy goods sector. It is stated, for example, that in the EU, the capital/labor ratio in services is much lower than that in the production of goods. In levels, this may well be a reasonable view to hold. However, it should be noted that the rate of growth of capital/labor substitution in the services sector over the past two decades in the EU has been at least as great as that in manufacturing, as the figures, which are based on a simple growth accounting calculation in Table 5.16 below, indicate. In graphic terms, the production frontiers in the EU have been drifting in a capital-intensive direction, and the usual growth-accounting estimates suggest that there has been comparatively little movement inwards. In other words, assuming that the CEEC imports and adopts the current technology and not that in the West of twenty years ago (one must assume that the spread of computers will be very rapid, for example), the scope for employment growth via the services sector may not be as great as Messrs Black and Moersch would hope. Indeed, it is of some interest to note that in the EU, it is newer and hoped-for peripheral tertiary activities such as health and home care services that are tending to receive attention now as potential new sources of employment growth. In the
Table 5.16. Capital/labor substitution in the EU, 1970-90 (annual average growth percentage) Manufacturing
Services
Germany
1.1
1.5
France
1.2
2.1
Italy
2.1
3.9
United Kingdom
L4
U
Source: Eurostat Sectoral Data, Services of the European Commission.
Comment
205
more traditional services activities, such as banking, insurance, and financial services, we have witnessed considerable shedding of labor, the consequences for employment of which have been articulated in European Commission's 1993 White Paper, Growth, Competitiveness, Employment as well as in other recent statements. In the next stage of the transition process, which is capital accumulation, the supply of savings from domestic sources supplemented by those from abroad are taken to be sufficient to raise output per employee to within almost 60 percent of the EU level by 2002. But to do this, growth rates characteristic of Japan in the 1960s or even greater are required over a long period by even the most favorably placed of the group. The paper makes a good deal of the valid point that the marginal product of capital in the CEEC bloc will almost certainly be high, and with real interest rates of the order of 5.0 percent, one would expect this to be an attractive haven for world savings. In this respect, one potential external source that is not mentioned by the authors rests in the fact that the EU as a whole, and unlike the CEEC, is faced with a certain future demographic shock (the more extreme case elsewhere is that of Japan). The financing requirements to meet the needs of aging populations mean that sooner rather than later it makes sense for the EU to begin running external account surpluses and to become a net exporter of capital. In this respect, given the anticipated high real rates of return in the CEEC bloc, additional capital may be expected to flow to these countries. This is certainly an option for future policy. The overall impression of this commentator is that while the paper does a respectable and interesting job in presenting a framework for the analysis and sets out well the lines of approach, the final quantitative picture is less than convincing. Perhaps my being a civil servant for too many years has made for an overcautious approach. But this sentiment is justified best where the paper makes the strong statement that "these projections show clearly the ability of the CEEC's to make the transition" (my italics). Earlier we were told, however, that these results should be taken as being optimistic. Golden age-style annual average growth rates of real output of 9.0 percent plus, over a decade or more and in the current world economic climate, is a tall order. When one reflects on the longstanding difficulties faced by the existing members of the EU in achieving convergence and on how far there is still to go, the mountain confronting the CEEC hopefuls is a big one. Countries in transition are always vulnerable to unanticipated asymmetric shocks, and there has been no shortage of these in the last two decades. Moreover, the authors note the sensitivity of their results to the rate at which the transition proceeds. Severe penalties are to be faced by the six countries if they fall short in their performance. The paper makes no mention of the important and increasingly acknowledged risk that, should monetary union be achieved or be seen as a feasible practical proposition within say the next decade, the likelihood is that the core of participants in European Monetary Union will open up a gap with the outer
206
Douglas Todd
group of the Union, and possibly an even greater one between themselves and the aspiring Central European members. Catchup will be that much harder, with even greater demands being made for internal improvements in economic efficiency. This is what underlies the convergence criteria and what in due course presumably one would expect aspiring participants to embrace. In this respect, one can only underline the need noted in the paper, to pursue stable macroeconomic policies and to control public budgets. This requirement, however, places much faith in the success of supply-side policies, an issue under close scrutiny at the present time both in the EU and more widely in the OECD countries. One thing we have learned in the process of both formulating and implementing macroeconomic policy is that when countries are opened up and exposed to the vagaries of world capital markets, credibility becomes of crucial importance. It is hard to achieve, as the CEEC countries are now discovering. Moreover, such credibility is very difficult to win back if lost. Witness the current difficulties being faced in the advanced industrialized EU, which in some ways could be regarded as a test case. Moreover, this is seen even with the accumulated benefits in Europe of continuous and accelerating growth over the long period 1983-90. Financial markets are indeed hard to win over and convince. However, in this area of policy nothing is certain, and at least the paper emphasizes the possible requirements even if the potential to achieve these over the time scale considered appears, to this commentator, to be somewhat overstated.
Comment Holger Wolf
The future of Eastern Europe remains murky. Some observers expect the transition economies to spawn the next generation of tigers, with high investment and growth rates enabling a rapid catching up to Western Europe. Others take a more pessimistic view, pointing to the inherited social burdens and unsettled politics to argue that Eastern Europe is more likely to follow the populist path of Latin America in the 1980s, and is thus destined to remain the poor cousin of Western Europe for decades to come. Which of these scenarios will ultimately play out depends on the investment rate the transition economies can reach. The necessary rates to achieve rapid catchup have been the subject of a lively recent debate. The answers vary widely, depending on the assumptions regarding the starting level of the capital stock and the extent of "free" disembodied productivity growth yielded by the switch to market-based allocation mechanisms. Alas, most studies stop here, devoting little attention to the not entirely unimportant question whether the derived investment paths can in fact realistically be financed. The paper by Stanley Black and Mathias Moersch goes the entire route, looking both at likely growth rates and at the financing of the underlying investment, and is thus a much needed addition to the literature. My comments will concentrate on their predictions, arguing that there are solid grounds for taking a less optimistic view on both the growth impact of a given investment path and the scope for obtaining financing, and thus, by implication, on the prospects of Eastern Europe achieving rapid catchup. The authors begin by presenting their own estimates, ending up firmly in the optimistic camp, based on an assumed rebound from the low income levels in the initial phases of the transition, a substantial rate of "free" growth from technological updating and an assumed low incremental capital/output ratio reflecting a predicted output shift from capital-intensive manufacturing to laborintensive services. How realistic are these assumptions? Viewed historically, sharp declines in output following social-political unrest or war are indeed frequently followed by sharp rebounds - a feature already observed by Ricardo. A particularly vivid example of this "rubberband growth" 207
208
HolgerWolf
occurred in postwar western Germany. Following destruction of key infrastructure linkages and monetary upheaval, output in the immediate postwar period stood at close to 20 percent of prewar levels. Yet to a large extent, the decline reflected a temporary retraction inside a largely unchanged production possibility frontier (PPF). Following - fairly inexpensive - repairs to the infrastructure and monetary reform, output rapidly snapped back toward the PPF the miracle was born. Alas, the situation in the transition economies is quite different: trade links are permanently lost, relative prices have permanently shifted, and as a result, a not insubstantial fraction of the capital stock has become permanently obsolete. In short, in the transition economies the PPF itself has contracted, sharply reducing the scope for a rubberband rebound. The second supporting rod for the authors' optimistic forecast is the assumption of substantial free growth through the shift to market-based allocation systems and Western technology. The optimism is shared by most studies in this field but is buttressed by remarkably little evidence. To be sure, improvements in input deliveries and the like will improve the productivity of the inherited industrial behemoths. But how important are reliable inputs if the production technology itself is uncompetitive? Alas, the outdated industrial complexes littering the landscape of the transition economies will not become stateof-the-art CAM facilities merely by virtue of shifting from plan to market: The evidence is overwhelming that technological progress is embodied, it will come with investment, and it will only come with investment. The relation is clear in cross-section: It is high investment Asia, not low investment Latin America, that is competing with the old industrial world in the high-tech areas. Thus while there may well be productivity progress on the margin, there is no reason to believe - as the authors' methodology implicitly does - that even without any investment, the productivity of the existing stock of capital will catch up to the efficiency of capital in Western economies merely through exposure to a market economy. In this light, one also wonders whether it is realistic to assume, as the authors do, that the existing physical capital is malleable. On the margin, reorientation of some enterprises toward producing consumer products is likely, yet again, the refocusing will generally require additional investment. In the end, it has to be recognized that a very sizable fraction of the existing capital stock is bound for the scrap heap. The authors' final ground for optimism is a low predicted incremental capital/output ratio, based on an assumed shift from capital-intensive manufactures to labor intensive services. To be sure, the service sector in many transition economies is underdeveloped. However, a closer look suggests that the deficiencies are at the high rather than the low end of services. Health care, child care, and so on were provided by enterprises as part of the nonmonetary remuneration, though, as "nonproductive" activities, underreported in the statistics. The evidence suggests that in fact employment in these service sectors is likely to decline as replication is reduced and as fiscal and budgetary constraints lead to reductions on both the demand and supply side for some of the services,
Comment
209
day care in particular. The high-end services - legal services, accounting, finance - are alas not obtainable without a quite high investment, in human rather than physical capital, and thus the authors' assumption of a sizable expansion of the (actual, not measured) service sector might be on the optimistic side. In sum, these caveats suggest that the authors' derived investment-growth link is very much on the optimistic end of the spectrum; adopting a more cautious view on the scope of disembodied productivity catchup and the likely size of the rebound suggests a considerably higher incremental capital-output ratio and hence the need for significantly higher investment ratios to attain any given output target. The second part of the paper deals with the financing of investment in both its quantitative and its qualitative aspects. Savings come in two forms, domestic and foreign. The authors begin with the former, concluding that "with positive real interest rates of around 5 percent, savings in the Visegrad countries could reach 27 percent, while it could reach 37 percent in Bulgaria and Romania." This prediction, which would place the transition economies squarely into the ranks of the Asian tigers, is based on a cross-sectional regression of the savings rate on ex post real interest rates. A closer look at the regression reveals, however, that the prediction seems to rest almost entirely on the constant (26.7) and a dummy for Romania and Bulgaria (10.2), rather than the real interest rate elasticity. There are strong reasons to place considerable confidence intervals around these predictions. Cross-country evidence on the determinants of private savings identifies three factors as being particularly important: income growth, demographics, and the development of consumer credit. Looking forward, the transition economies will benefit - hopefully - from faster income growth, which will buttress savings. However, on the negative side they will experience the same aging of their populations as the Western European countries, with a strong negative effect on savings. Indeed, using cross-country estimates of the elasticity of savings to the dependency ratio and to income growth suggests that just to offset the effects of aging, the growth rate of income per capita in the Czech Republic would have to increase by 11 percent - hardly a credible scenario. Furthermore, financial liberalization, in particular the introduction of consumer credit and a mortgage market with less than 100 percent down payment will, judging from cross-country evidence, reduce savings significantly. In short, to the degree that experiences of other countries can be transplanted to the transition economies, a sharp decline in household savings to the low double digits appears the most likely scenario. This leaves public and corporate savings to make up the balance. Alas, as the authors note, the public sector, with the exception of the Czech Republic, constitutes a net drain on national savings in the transition economies. Given an implied net savings of household and the public sector in the low teens, simple back-of-the-envelope calculations readily reveal that the rate of return required to generate corporate savings of 15 to 20 percent of GDP - and thus to
210
HolgerWolf
lift national savings to the level estimated by the authors - is quite outside the range of normal experience. In the end one is thus left to conclude that, to the degree that international experience is relevant for the transition economies, national savings rates are quite unlikely to reach the levels predicted by the authors; indeed, a final figure of not much more than half their estimate - 15 to 20 percent of GDP - appears a more likely outcome. Foreign savings are - again judging from cross-country evidence - unlikely to make up the difference. While capital inflows surged in the immediate aftermath of the transition, these flows largely represented a one-time stock adjustment to the opening of previously closed markets. Once the stock adjustment is over, there is little reason to believe that the sustainable deficit in the current account in Eastern Europe will exceed the typical 3 to 4 percent range observed in other countries at similar development stages. While wages are low and skills fairly abundant, this does not set Eastern Europe apart in the 1990s and is at least partly offset by large remaining uncertainties. To take just one example, Hungary continues to suffer from unsustainable external and fiscal deficits, severe problems in its banking system, an unsustainable pension scheme, and no clear political direction, hardly a climate conducive to sustained, above normal FDI. In sum, there are numerous reasons to believe that a given rate of investment will have lower growth effects than the authors predict, and that the attainable investment will, due to a smaller savings potential, be lower than the authors predict. In conjunction, higher ICORs and lower investment decisively diminish growth potentials and render the authors' pessimistic scenario a good case outcome. Of course the volume of savings is only part of the story; the efficiency of allocation matters as well, a point to which Black and Moersch turn next, though only briefly. The importance of allocative considerations may, however, deserve more extensive attention. Investment projects can be financed internally - by retained earnings - or externally. The latter in turn can take place via collateralized and noncollateralized loans or via debt or equity issues. The former is by far the dominant source for small and medium-sized enterprises in Western market economies. Alas, continued difficulties in creating well functioning real estate markets preempts the main type of collateralized loans. Noncollateralized loans require banks to devote considerably more effort to assessing the quality of loans; this is precisely where banks in Eastern Europe have difficulties: risk-return assessments require experienced loan officers, a scarcity in the transition economies. While foreign banks can to some degree ease this constraint, they, just like domestic banks, are likely to focus initially scarce resources on lending to the blue chip companies, leaving the small and mid-size firm sector, which is currently severely underdeveloped but is crucial for development, without sources of external finance. In the short- or medium-term future, it is doubtful that banks will play the role of an efficient intermediary between private and public sav-
Comment
211
ings and investment. Equity and debt markets are slowly emerging, but as yet there is little evidence that they will play a major role in enterprise finance, particularly not for the small and medium-size firms. In consequence, efficient investment allocation in the transition economies will require reliance on retained earnings. This is also desirable for a different reason: the existing banks to a large degree are either directly tied to particular enterprises (in particular in the FSU) or are burdened with inherited nonperforming loans. The quality of these loans is subject to significant uncertainty, depending on the future evolution of the economy. In consequence, it stands to reason that banks will continue to allocate credit preferentially to existing large customers regardless of their actual profitability. Reliance on retained earnings at least rewards profitability, and may thus be a second best outcome even though venture capital needs are still unmet. To conclude, on present trends the Latin American rather than the Asian scenario seems to be the more likely destination of the transition economies - with the possible exception of the Czech Republic. Politics in Eastern Europe remain mired in distributional conflicts, depressing both public and corporate savings, while household savings are set to decline under the continuing pressures of an aging population and freer consumer credit markets. In combination, the likely savings volume is insufficient to finance the rapid convergence scenario. That being said, the policy implications are straightforward: the combination of a redistribution of income from labor to capital, pro-business policies, and an end to fiscal profligacy is the price of admission to the Asian growth scenario.
CHAPTER 6
Privatization, structural change, and productivity: toward convergence in Europe? PaulJ.J. Welfens
6.1
Introduction
With the demise of the command economy, the economic East-West division of Europe became visible, and it turned out to be much bigger than anticipated. With German unification occurring in 1990, it quickly became apparent that the GDR - widely presumed to be the leading economy in terms of per capita income of the ex-CMEA (the former Soviet bloc Council for Mutual Economic Assistance) - did not match the consensus estimate of 50 percent of West Germany's per capita income and labor productivity. One third was the true figure. With the fall of the Berlin wall and the collapse of the socialist CMEA integration scheme and of the USSR, the wide East-West income gap became fully apparent. The Eastern European transformation process as well as economic opening up rendered part of real and human capital obsolete in the ex-CMEA area so that the European East-West economic divide became transitorily aggravated. Output levels in 1995 will about match those of 1989 in the ex-GDR, in Poland, and in the Czech Republic. The GDR transformation was the most radical to date in terms of structural change and productivity growth. The share of mining in industry fell by two thirds in the period 1990-94, the share of investment goods by more than 10 percentage points, while the construction industry - supported by special investment incentives - almost doubled its share. Overall, except for an overextended government sector (20.9 percent compared to 13.2 percent in western Germany) and an underdimensioned service industry (27.7 percent compared to 36.2 percent in western Germany), eastern Germany's economic structure had become rather similar to that of West Germany (Heilemann and Lobbe, 1995). While transformation brought sharp declines of output in the first stage in Eastern Europe, there is a potential for fast catching up in the second stage, provided that the rate of capital formation and of productivity growth are high. Factor productivity could quickly improve, as shown by the example of the former GDR and some transforming economies and also the experiences of Japan and OECD-Europe countries in the 1960s. In the OECD countries, the weighted av212
Privatization, structural change, and productivity
213
erage annual growth of total factor productivity in the decade prior to 1973 reached 2.7 percent. This is composed of 4.3 percent growth of labor productivity and -0.8 percent of capital productivity, weighted by their shares in total output (Englander and Gurney, 1994, p. 116). Japan reached labor productivity increases - on an output per worker hour basis - of 9 percent, Germany, France, Italy, Belgium, Denmark, and the Netherlands of roughly 6 percent, which was almost three times as much as in the technologically and economically leading U.S. economy. With respect to systemic transformation, the small ex-CMEA countries have made considerable progress in the period 1990-95. Major institutional changes, macroeconomic stabilization, economic opening up, and privatization have been achieved. Moreover, competition laws and bankruptcy procedures have been enacted in these countries. So far, Russia and Ukraine are two negative examples of transformation, since hyperinflation, falling output, and accelerating impoverishment of the population is taking place without visible signs of a broad economic upswing within a new market system. The economic divergence within the post-socialist ex-CMEA area has increased in the 1990s, and there are prospects for a double economic divide in Europe in the future: EU-Eastern Europe and in Eastern Europe, Visegrad countries versus others. With the EU facing the dynamics of the single EU market and major privatizations in France, Italy, Spain, and Germany (while the UK had already massively privatized under the Thatcher government), the EU is moving toward intensified integration plus sharper competition, which should lead to economic gains, so that Western Europe's lead vis-a-vis Eastern Europe could widen. Massive privatization in Eastern European countries allows one to expect that the post-socialist countries could mobilize enormous growth momentum, since post-socialist countries enjoy the potential benefits of economic catching up of low-income economies and since the creation of a private sector in Eastern Europe coincides with radically improved access to Western markets. The end of the Cocom list as well as the new opportunities of free trade and capital flows in Europe bring about new growth opportunities. Whether the economic potential of increased productivity and growth can fully be exploited is an open question. For Western European countries with fear of rising immigration in a period of high unemployment, the topic of catching up and economic convergence is as important as for Eastern Europe, where people expect market economic forces to allow rapid increases in economic well-being. The small transforming economies are gradually catching up with Western Europe. Indeed there could be prospects for overcoming Europe's economic East-West divide within two or three decades if sustained high productivity growth could be achieved. Creating a private business sector and the institutional framework of a market economy are widely held to be the keys to productivity growth. However, this paper will emphasize that privatization cum competition and structural change - changes in the output mix (e.g., in the ratio of tradables output to nontradables output) - are the real issues in the debate
214
Paul J. J. Welfens
of economic catching up. Section 6.2 will focus on supply-side changes and the links between privatization, structural change, and productivity growth. Section 6.3 discusses the experience with privatization, while Section 6.4 deals with structural change. In Section 6.5 convergence issues are raised and policy implications evaluated. It is argued that economic convergence in Europe can be achieved only if privatization and structural change are realized, which is likely to meet five barriers: (1) political instability in many countries; (2) high and sustained unemployment rates; (3) dominating impact of the nontradables sector and of capital-intensive sectors in some countries; (4) lack of structural change; and (5) lack of long-term approaches in the business sector. 6.2
Transformation and supply-side changes
Transformation of socialist systems means a new institutional network and major supply-side changes - privatization being the most important of the latter. Efficient factor allocation should both take place within companies and be organized through arm's-length transactions in markets. The rather inefficient socialist economic systems thus could be succeeded by dynamic open-market economies that could quickly catch up with Western Europe; ideally they could mimic the economic and political catching-up process that Asian Newly Industrialized Countries (NICs) launched vis-a-vis Japan in the 1970s and 1980s. However, the initial distortions in Eastern Europe are enormous: (1) decades of a command economy system have undermined the drive for entrepreneurship and private capital accumulation; (2) accepting major income differentials is incompatible with socialist ideology; and (3) the initial size distribution of firms in post-socialist countries is very much skewed toward big companies with more than 500 employees. By contrast, the majority of employees in Western Europe is employed in small and medium-sized firms with fewer than 500 employees. Many West European firms are below minimum optimum plant size, but the dynamic pro-competitive effect of many small newcomers constantly entering the markets more than outweighs the forgone benefits of optimum plant size. Even oligopolistic markets in OECD countries are found to be open enough to accommodate a constant stream of new entrants (Audretsch, 1995) - few of which survive and grow in the long term. (4) State-owned firms have not been exposed to competition for decades. (5) The military industry played a strong role in some ex-CMEA countries, most notably in the former USSR where conversion is a special problem. (6) Relative input and output prices diverged much from world market price structures. Gross domestic product per capita reached less than 50 percent of the EU average and differed markedly in Eastern Europe in 1994. The Czech Republic, Slovenia, Hungary, the Slovak Republic, and Poland are in a leading position; Russia, Romania, Bulgaria, and Ukraine are behind (Table 6.1). No country in Eastern Europe reached Greece or Portugal in terms of per capita income evaluated at purchasing power in 1993, and it is also remarkable that Portugal's
Privatization, structural change, and productivity
215
Table 6.1. Selected economic indicators for Eastern Europe Per capita income $ (1993)"
Investment/GDP 1993
Export/GDP 1993
Hungary
6,260
20
30
Czech Republic
7,700
17
55
Poland
5,010
19
19
Slovak Republic
6,450
25
67
Bulgaria
3,730
20
50
Slovenia
6,310
16
60
Romania
2,910
27
23
Russia
5,240
26
37
Ukraine
1,910
-
-
8,360
20
23
Memorandum Greece Portugal
9,890
29
29
Spain
13,310
23
18
Belgium
18,490
19
69
Netherlands
18,050
21
52
Germany
20,980
21
33
" At purchasing power parity exchange rates. Source: World Bank (1995), The World Bank Atlas; and EBRD (1995), Economics of Transition.
investment/output ratio was higher than that of any transforming country. Investment/output ratios fell in all major transforming countries to below 20 percent, except for the dynamic Slovene economy and the laggard Russia. Export/GDP ratios increased in the course of economic opening up. Except for Hungary, the CSFR, and Slovenia, transforming economies initially faced hyperinflation and thus major redistributions in income and wealth. High inflation rates are not conducive to capital market development and longterm investment. Budget deficits have been reduced after the first transition stage except in Hungary, Bulgaria, Romania, and Russia, where government absorbs a high share of private savings. If private households' savings rates remained low and if firms' profits and reinvested earnings were small, there would be no prospects for economic growth. Privatization, macroeconomic development, and growth There are important links between privatization, macroeconomic development, and growth: the basic assumption here is that the investment/output ratio
216
PaulJ. J. Welfens
in combination with the marginal product of capital determines economic growth: • The more comprehensive and competition-enhancing the privatization process is, the lower the future budget deficit will be. Higher future tax receipts from privatized companies should reduce prospective government debt and indeed enable government to run a budget surplus. With a higher overall savings rate, a higher investment/output ratio and higher growth will be possible. Higher growth in turn will stimulate savings, where we assume that the positive income effect is stronger than the Metzler effect associated with rising wealth (with a given wealth target, higher real per capita wealth would dampen per capita savings - an argument to be modified if the wealth target is not exogenous but part of a gradually increasing aspiration level). • Privatization will lead to a high marginal product of capital if privatization is competition enhancing, that is if unbundling of big state firms and the creation of new firms are considered to be important elements of a broader privatization strategy. If privatization heavily relies on foreign direct investment (FDI), this should raise the marginal product of capital in the tradables sector quickly. However, if an influx of foreign investment dampens progress in privatizing and regulating banks in an efficient manner, there could be an "FDI winner's curse": With inefficient credit allocation to the nontradables sector's newly privatized firms, their performance might indeed show only small progress. Hungary seems to be an interesting case in this respect, although one should not overemphasize the FDI winner's curse problem in the case of a small open economy in which - in contrast to a large country - the tradables sector dominates. • If privatization brings about a rather equal endowment of real capital and in effect allows only socially acceptable inequality to emerge typically associated with windfall profits and supernormal profits in international markets (known to be a hard turf for newcomers) - political pressure on redistributive and growth-retarding policies will remain low. To constrain rent-seeking activities that imply a low overall marginal product of capital would also require preventing the dominance of big firms in important sectors or regions of the economy. • Involving broader strata of society in the privatization process should facilitate political support for a market economy, provided that capital gains are recorded by most new owners of equity capital, land, and housing. Clearly, for the majority of people, only a mixed portfolio can combine a high rate of return and limited risk. Developing housing and land markets in combination with an efficient banking sector will generate sufficient collateral for investment in manufacturing industry and the service industry. This gives privatization of housing property and
Privatization, structural change, and productivity
•
•
•
•
•
217
land a special role; it also points to the strategic role of infrastructure investment, which will raise the real value of private property. Combining privatization policies and prudent macroeconomic policies will generate massive capital gains for owners of equity capital. Anticipating such capital gains will make transitory real wage concessions easier to accept and allow older capital vintages to be employed for an extended time - thus yielding higher output and a longer transition period. Privatization could stimulate the motivation of workers in the case of employee ownership schemes, so that the marginal product of labor is raised. Higher efficiency wages would be warranted in such a situation. Privatization could stimulate massive imports if viable private supplier industries in the tradables industry are not created early on. In the massive import scenario, privatization leads to massive devaluation in a flexible exchange-rate system. In the long term, privatization should lead to efficiency gains that will make exporters more competitive. Privatization of industry and agriculture will lead to massive unemployment in the short term as hidden unemployment is transformed into open forms of unemployment. This in turn could reduce expected real income growth in the consumption goods sector and thus cause reduced investment as the consumption goods industry will order fewer investment goods. Privatization of manufacturing industry could lead to long-term unemployment of unskilled workers as high-wage earners in the capital-intensive manufacturing sector will hardly find well-paid jobs in the less capital-intensive (and less skill-intensive) services industry. High unemployment rates and long-duration unemployment will raise resistance against technological modernization and rationalization investment such that both the investment/output ratio and the marginal product of capital - both important for economic growth - could fall. Unemployment insurance schemes therefore should strongly discourage workers from extended search periods, which means that unemployment benefits should be limited to several months only and possibly should give a once-and-for-allbenefit if a new but less well paid position is accepted in a long term contract. Privatization will increase the price elasticity of aggregate demand, since private owners of firms will spend their budget more carefully than state-owned firms whose managers operate under a soft budget constraint. This will have a dampening effect on the aggregate price level.
After a fall in real GDP in 1992-93, there was considerable economic growth in Eastern Europe, where Poland's 6 percent marked the highest growth rate
Table 6.2. Change of labor productivity in indusry and economic growth (percentage per annum) Labor productivity
Projection
Economic growth
1992
1993
1994
1992
1993
1994
1995
1996
-0.3
-14.4
-2.3
-1.2
5.2
-7.0
-0.9
2.6
3.8
4.9
Hungary
-3.9
-10.2
-1.4
13.4
15.0
-5.0
-2.3
2.6
0.2
1.5
-20.3
-5.2
13.7
12.6
17.0a
1.0
3.8
6.0
5.0
5.0
Slovak Republic
-1.1
-14.6
-10.2
-10.4
8.0
-6.0
-4.1
5.3
4.0
4.0
Slovenia
-7.5
-1.4
-3.3
6.4
13.2
1.3
5.0
5.0
5.0
Bulgaria
-12.0
-5.5
-7.0
3.7
14.1
-6.0
-5.0
-2.0
Rumania
-4.4
-15.0
-13.4
9.0
10.0
-15.0
0.0
0.0
Russia
3.4
-6.4
-16.5
n.a.
n.a.
-19.0
-12.0
-10.0
Ukraine
2.6
-2.3
-3.4
n.a.
n.a.
-14.0
-9.0
-10.0
Poland
a
1991I
Czech Republic
1990
Own estimate. Source: For productivity growth Hunya et al.,WIIW (1994), 9 and WIIW update; growth rates: IMF (1995), Table 10.
Privatization, structural change, and productivity
219
130-
120-
110CR HUN —A—POL
100-
-B-SR
—*— SLOV —•—BUL ROM
50 1982
1984
1986
1988
1990
1992
1994
Figure 6.1: Index of labor productivity in Eastern Europe (1989 = 100). Source: WIIW data base incorporating national statistics.
among the Visegrad countries in 1994 (Table 6.2); in 1995-96 Poland, the Czech Republic, and Slovenia are expected to record a growth rate of 5 percent. Only in Russia the GDP growth continues to be negative. Five percent growth certainly is an achievement, but this is still several percentage points below economic growth in Asian NICs, which recorded annual growth rates of 7 percent (or more) in the 1970s and 1980s, such that per capita income doubled within a decade. After sharp falls in productivity due to dramatic output decline in 1991-92, labor productivity started to increase in some transforming economies, most notably in the Czech Republic, Hungary and Poland in 1994-95 (Figure 6.1). Leaders in transformation were also leaders in industrial labor productivity growth, and there is little doubt that privatization and foreign direct investment inflows were key elements for productivity growth in industry. Negative interest rates in Poland and Hungary are likely to have left more capital employed and hence labor productivity and employment to be higher than is feasible in the long term. There is, however, no doubt that labor productivity is rapidly increasing in the Visegrad countries. In absolute terms, Hungary's labor productivity in 1994 was 21 percent higher than in 1989. In 1994 Poland and Slovenia were the other two countries in which the absolute level of labor productivity was higher than in 1989. Productivity growth in the service industry could also be considerable in
220
Paul J. J. Welfens
Eastern Europe, since many new firms have been established and modern capital goods - mainly computers and telecommunications equipment - can be obtained at falling relative prices in the world market. As with industry, it takes time to reorganize firms in the state and private sector. Productivity growth increased partly in parallel with unemployment rates, which reached high rates except for the Czech Republic (and Russia plus Ukraine). Double-digit productivity growth rates recorded in or expected for several ex-CMEA countries indeed point to the problem that there is growth and mass unemployment at the same time. While part of the unemployment problem could be eliminated by a greater role of part time jobs and a reduction in labor force participation ratios, mass unemployment can only be avoided if layoffs in contracting sectors are matched by vacancies in expanding sectors.1 Since contracting and expanding industries often are not located in the same region, functional infrastructure and labor mobility are important. Reduction in unemployment can be expected only if new firms are attracted into depressed regions or if workers are sufficiently mobile within the country. In terms of regional mobility, Poland, Ukraine, and Russia pose major problems because of the size of the respective countries and the rather underdeveloped infrastructure links in combination with excess demand in the housing sector. Factors ofproduction and sources ofproductivity gains Many studies find that empirical evidence from OECD countries confirms that private firms are more efficient than state-owned firms (see discussion in Jasinski and Welfens, 1994). Privatization is expected to stimulate economic growth for several reasons. Privatization allocates full property rights to the owner, so that there is an incentive to maximize the residual profit by reducing X-inefficiencies (Koop, 1994), and to move toward optimum plant size so that static efficiency criteria are met. Moreover, private companies have an incentive for innovations - that is, dynamic efficiency - in the form of both product innovation, which raises average revenues at home or abroad, and process innovation, which improves profits and allows a greater production volume. Private companies suffer from principal-agent problems which could result in major X-inefficiencies. Functioning capital markets as well as some form of insider control by banks are means to minimize the problem of X-inefficiencies. To ensure the efficiency of new investment, and to stimulate firms to strive for dynamic efficiency and optimum innovativeness, capital markets are required. This points to the need to create private banks and investors on the one hand. On the other hand, privatization strategies that nurture the expansion of a com1
Labor force participation rates in Hungary were 10 percentage points higher than the OECD average; the participation rates in Poland and the Czech Republic of 64 and 66 percent, respectively, in 1992-93 also were still substantially higher than the OECD total of 60.5 percent in 1991; in the Visegrad countries there is increasing regional mismatch and rising unemployment duration.
Privatization, structural change, and productivity
221
petitive capital market will bring about positive side-effects. Not only is the efficiency of privatized firms raised, but productivity-enhancing effects are generated in the whole economy. Hence, efficiency in industry can be expected only if economic policy brings about foundations for financial market stability and if banks and insurance companies already have been privatized. However, in all ex-CMEA countries, banks typically are the last group of firms to be privatized because the issue is politically sensitive and because achieving viable domestic banks via mergers and restructuring takes time. Productivity growth has two sides to the same coin. First, physical productivity growth means increasing the ratio of output to input; second, the marginal value added can be increased by product innovation, better quality, or better reputation of the firm, which allows increased prices. With respect to tradables, higher value-added productivity will show up in improved terms of trade in the long term. Raising physical productivity of the factors of production refers to six basic input factors in the firm's production function: labor (L), capital (K), technology (Z), natural resources, especially energy (/?); the fifth factor, which is often supplied as a public good, is infrastructure capital (KG). In the long term, the type of technical progress adopted will be of particular interest, since laborsaving technologies are easily available from OECD countries in which firms face rising labor-capital costs. The sixth factor considered here is imported intermediate goods (X*), so that trade as well as foreign direct investment directly affects the supply side. Both can be expected to generate positive spillover effects, in the form of learning effects that will benefit domestic firms and the nontradables sector in a way that is equivalent to induced technological progress. Defining capital K to comprise both domestic capital H and foreign capital A'**, one may restrict the sixth source of growth to trade (X*). Since trade and investment growth stimulate productivity gains, privatization strategies that encourage economic opening up will yield secondary benefits. Countries that pursue privatization strategies that lead to monopolistic firms and big business structures - with managers well tied to the political system - indirectly reduce the prospects for productivity increase and sustained growth. Privatization and efficiency Privatization of formerly state-owned firms in post-socialist countries can be expected to bring about seven types of efficiency gains at the microeconomic level in the medium and long term: (1)
To minimize inputs of production factors for a given level of output, which means elimination of X-inefficiency; overmanning will be eliminated and idle capital goods sold off. If labor and capital are scarce in other expanding sectors, factor productivity of complemen-
222
Paul J. J. Welfens
(2)
(3)
(4)
(5)
(6)
(7)
tary factors in these sectors can be increased by intersectoral and regional factor mobility. Improving the use of natural resources and infrastructure is also important. To adjust the output level in accordance with profitability requirements and competitive pressure from the market. In depressed industries this could mean further reduction of output and employment. Within a neoclassical context this will raise labor productivity. If many industries react in this way, rising unemployment rates and increasing duration of unemployment could result therefrom. In industries with scale economies, increasing output leads to productivity growth. To choose new and more efficient input mixes as capital is replaced and as new investment is undertaken; costs are thus lowered relative to sales proceeds. At the same time, adjusted product assortments could raise the average sales revenue at given factor cost. Raising capital productivity obviously should take longer in industries with a high capital intensity. Even with no process innovation in the whole economy, a more careful choice among existing production technologies by private owners should help to improve productivity. The productivity of labor, capital, energy, and other factors can be improved thereby. But switching to new technologies and rebuilding the capital stock will take time. To gradually approach the minimum optimum plant scale can increase factor productivity; optimum outsourcing is a complementary aspect that is important for Eastern Europe. Socialist firms rarely used specialization gains from outsourcing, as they were afraid of the delivery risk, which was a systemic by-product of monopolistic supplier structures. To locate investment in space optimally - an aspect that is long term and related to transportation costs, the size of markets, and the degree of external liberalization and integration. To accelerate technology transfer, which necessarily involves either foreign investment partners or attempts to become a multinational company, as most technology trade is intracompany trade or trade among multinational companies (Klein and Welfens, 1992). To upgrade human capital by training on the job; countries with a decent education system and firms (and countries) with a high share of young employees are well positioned to improve human capital productivity in the long term.
Privatization raises the elasticity of supply and therefore will increase the response to relative price changes, including the response to real exchange rate changes, playing a role in the context of the Robinson condition for devaluation policies. In addition to raising productivity by adjustment at the level of the firm, macroeconomic influences can affect total factor productivity and eco-
Privatization, structural change, and productivity
223
nomic growth, respectively. Macroeconomic stabilization and liberalization are associated with high growth in outward-oriented newly industrialized countries. High-growth countries were characterized by rather stable macroeconomic variables and market-oriented policies, that is, low inflation rates and small budget deficits, as well as small divergences between market exchange rates and unofficial exchange rates. This leads to the question to what extent privatization facilitates liberalization and stabilization. Privatization has important political effects, since it reinforces resistance against a return to the socialist system. Privatization generates sales proceeds in the short term and, more importantly, higher tax revenues in the future. Strategic aspects ofprivatization - competition Privatization not only means assigning private property rights for existing firms. If productivity gains and competitiveness are goals of transformation policy, dismemberment of big firms and unbundling of assets will also be important in the context of privatization. Moreover, this type of privatization from above has to be complemented by creation of new firms, that is, privatization from below, so that competition and innovativeness are encouraged. Unprofitable firms that cannot quickly be restructured should face liquidation or go bankrupt to open up the market for newcomers or takeovers. Barriers to exit are, of course, barriers to entry. This holds all the more if it is more difficult for newcomers to get loans from banks in Eastern Europe, where many banks had extended loans to the politically well-connected big firms. Privatized firms themselves can be expected to exploit fully the potential of productivity growth only if several conditions are met: (1)
Competition policy should keep markets contestable. Privatization in combination with competition will yield efficiency gains on a broad basis, since only with competition are firms induced to adopt best practice technologies and to pursue innovations. For Eastern Europe, both at the national policy level and - via the Europe Treaties - at the European level, competition policies rules are relevant (Hoekman and Mavroidis, 1995). Privatizations that are combined with dismemberment of big firms and spin-offs are pro-competitive. Many new firms have to be created to achieve competitive markets in which incumbent firms face pressure to behave efficiently. Competition is mainly an industry phenomenon, so that an industry-oriented approach to privatization is appropriate. (2) Integration policy with its focus on trade and foreign investment flows is important for productivity growth. Import competition and export promotion expose private firms directly to international competition, so that static and dynamic efficiency are achieved; this is clearly brought out by Asian NICs (Collins, 1990; Krueger, 1990; Sengupta,
224
Paul J. J. Welfens
1991). Privatization policies aiming to support integration could start in the tradables sector. Inward foreign direct investment (FDI) will stimulate international technology transfer, and in the case of big firms, also bring better human capital formation viafirm-specifictraining programs. Only with a competitive private supplier network in host countries can major FDI inflows be expected. Outward foreign direct investment allows firms to tap the pool of foreign technologies by observing market and technology trends abroad, so that a retransfer of foreign technological progress to the parent company will bring productivity growth at home. From a theoretical point of view, outward FDI requires that firms develop firm-specific or ownershipspecific advantages, which is possible only in competitive domestic markets with increasing R&D orientation. (3) R&D policies can stimulate innovations and help to realize first-mover advantages as well as to earn extra profits, which can be invested in risky productivity-augmenting investment projects by private firms. Competing private firms make sure that adoption of technological progress will be fast and the rate of innovation high. (4) Functioning capital markets can provide a market for allocating investment capital efficiently as well as a market for corporate control. Privatization of banks and insurance companies and the creation of a stock market are among the priority tasks of a growth-oriented privatization policy. Although capital markets hardly are efficient in reality, they are crucial to make sure that private firms operate efficiently. (5) Corporate governance is important for productivity growth in the firm, since both innovation and entrepreneurship can flourish only if firms are organized in a creative, flexible, and market-oriented manner. Lack of managers is a transitory problem in Central and Eastern European countries. In some countries the poor remuneration of university teachers stimulates an adverse selection process among the teaching and research personnel. Corporate governance depends upon privatization to the extent that strategic domestic or foreign investors will not emerge under all forms of privatization. From a strategic point of view, privatization policy is not only the task to find private owners for state-run firms; from a productivity-oriented perspective it also is important that it be designed in a way that the top five policy requirements for growth and productivity are supported. For example, if stateowned firms become private monopolistic firms that set up their own banks to give loans to inefficient firms, privatization is neither growth oriented nor procompetitive. Moreover, if big private monopoly companies influence policymakers in such a way that high subsidies are provided for the sake of restructuring, R&D funds will be diminished. If big suppliers for state infrastructure projects favor state-run monopolistic telecoms network operation and thereby entail
Privatization, structural change, and productivity
225
higher network user prices and hence reduced opportunities for newcomers in information-intensive industries, privatization policy has not realized a strategy consistent with private-sector expansion and economic growth. Economic growth will lead directly to productivity gains whenever it facilitates the exploitation of scale economies and stimulates R&D. With higher economic growth, higher R&D expenditures that will raise factor productivity can be spread over bigger sales volume. This will be particularly important in the nontradables sector. Exploiting scale economies in the tradables sector at home could also be important for productivity growth in imperfect international markets; the bigger sales volumes are at home, the more easily dynamic scale economies can be realized as a means to promote exports aggressively and capture Schumpeterian economic rents abroad. Some firms thereby will even become multinational companies with new options to tap the technology pool abroad and retransfer part of new knowledge to the parent company. Finally, providing better infrastructure and human capital improvement also can stimulate growth and productivity, as was emphasized by the new growth theory. Externalities: new growth theory and infrastructure investment An early dynamic growth theory in which the investment/output ratio is not the only determinant of rate of growth was developed by Arrow (1962), who argues that learning economies depend on cumulated investment. According to the new growth theory, positive external effects of capital accumulation play an important role at the macroeconomic level. Moreover, public goods properties of R&D expenditures can generate positive spillover effects at the industry level. If part of infrastructure capital or of private capital (or cumulated R&D expenditures) have positive spillover effects on the individual firm, the output level of the firm is given by: Y. = atK*LL/(L,L 2 ); hence v = (w/r)Z, where Z = (fcj/L2) + {kJLx). Assuming that labor can shift from one sector to the other within the country, a rise in the wage/real interest ratio as a consequence of opening up will not necessarily lead to a rise of the wage income/profit ratio in the overall economy. The Samuelson-Stolper theorem is only one possible outcome.
251
Privatization, structural change, and productivity (a)
(b)
k' k *
N
T
N
k k' > ^ *
k