Europe's Economy Looks East: Implications for Germany and the European Union

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
-2.6 1 7

-1.7 00

-17.1 .0 1

2 Mining, Quarrying

129

•Oft

-0.4

51

Table 2.8. Scenario C: CEC-EU free-trade area, tariffs and nonsensitive NTBs, sectoral effects on Poland ofCEC-EU integration (percentage change) Sector

1 Agriculture 310 Food 321 Textiles

Exports

Imports

Output

No. Firms

Czech.

Hun.

NAFTA

EFTA

EU-3

EU-9

Oth.

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

(13)

(14)

2.7

9.3

22.4

11.9

5.8

19.6

9.4

9.4

5.8

0.7

0.0

0.8

32.5

27.7

16.0

22.9

25.0

2.7

27.4

26.3

26.2

2.7

2.4

-1.1

-0.3

-1.6

6.6

11.7

28.6

29.1

1.5

26.1

15.8

15.9

1.5

3.5

-1.3

-0.4

-0.9

3.8

20.2

21.4

Bilateral Imports

Change in Employment Percent

1000s

15.2

12.6

10.8

8.6

21.5

8.6

-1.7

-1.7

-1.6

-2.8

323 Leather Products

55.1

-1.8

20.2

40.2

-17.5

3.2

5.7

5.4

-17.5

29.0

21.8

22.2

20.8

324 Footwear

44.1

15.3

23.7

28.7

-3.3

23.1

23.3

23.3

-3.3

8.3

2.7

4.5

322 Clothing

2.7

331 Wood Products

19.3

15.3

23.8

28.3

1.2

25.3

20.1

20.2

1.2

4.8

1.1

1.5

2.5

332 Furniture, Fixtures

11.5

19.5

19.1

24.2

3.7

28.5

23.3

23.5

3.7

0.0

-2.7

-1.6

-2.1

341 Paper Products

27.9

15.1

24.4

12.6

-4.9

20.6

18.9

18.8

-4.9

6.3

0.4

1.2

0.5

342 Printing, Publishing

12.2

15.6

19.1

8.0

0.2

22.9

19.0

19.1

0.2

-1.7

-4.9

-3.3

-1.2

21.1

20.4

13.3

27.1

25.3

-1.4

20.5

-1.4

0.0

-5.5

-4.6

-10.8

35B Petroleum Products

28.6

8.2

22.6

22.4

-5.3

19.4

19.5

19.5

-5.3

7.0

-0.1

-8.9

-16.4

355 Rubber Products

25.0

20.0

32.7

33.4

-1.7

30.0

27.7

27.6

-1.7

4.2

-1.8

-0.5

-1.1

36A Nonmetal Min. Prod.

27.7

10.5

24.7

21.8

-9.1

13.0

16.6

16.5

-9.2

8.5

2.0

3.4

1.1

362 Glass Products

33.0

7.8

13.4

17.2

-9.9

11.1

12.9

12.9

-9.9

9.7

3.8

4.9

3.1

25.6

35A Chemicals

7.8

3.8

14.5

24.3

1.2

18.2

21.8

21.9

1.2

4.5

-1.2

0.0

372 Nonferrous Metals

30.3

12.5

19.5

-4.3

-22

17.8

18.8

18.9

-2.2

11.2

8.1

9.8

7.7

381 Metal Products

26.6

11.6

22.3

23.3

-6.6

14.7

18.3

18.3

-6.6

10.3

5.1

6.5

11.8

382 Nonelec. Machinery

17.7

12.8

17.5

14.5

-0.8

23.7

16.3

16.4

-0.8

3.3

-1.2

0.3

0.9

383 Electrical Machinery

23.0

13.9

14.3

10.6

-1.3

20.6

19.2

19.2

-1.3

4.4

-0.9

0.3

0.6

384 Transport Equipment

26.9

14.2

21.1

20.5

-5.9

21.8

22.7

22.8

-5.9

6.1

-0.9

0.1

0.1

38AMisc.Mfrs.

25.1

13.0

22.5

16.0

-6.5

15.4

19.4

19.5

-6.5

7.5

1.4

2.6

21.1

8.7

31.2

24.2

-2.4

29.7

28.2

28.1

-2.4

7.6

1.1

2.3

9.3

4 Utilities

2.5

8.3

6.7

2.7

1.6

10.2

21.9

21.8

1.6

3.4

-2.1

-1.4

-6.2

5 Construction

3.4

0.1

3.0

2.2

0.0

-0.2

0.0

0.0

0.0

4.2

-2.0

-0.9

-10.8 -11.1

371 Iron, Steel

2 Mining, Quarrying

0.1

6.5

6 Wholesale Trade

-3.5

6.9

-0.1

-0.4

7.9

7.7

7.8

7.8

7.9

0.2

-2.3

-0.7

7 Transportation

-0.7

4.0

3.1

2.2

4.2

4.0

4.2

4.2

4.2

2.7

-1.5

-0.1

8 Financial Services

-4.6

7.1

1.2

-0.6

6.6

7.2

7.4

7.3

6.6

-0.1

-2.8

-2.0

-4.3

9 Personal Services Total

-4.9

7.1 106,

-0.4 122

0.6 10 A

8.6

8.4 14?

8.6

8.5

8.6

0.0

J1.2

15 5

-1 4

-1.9 02

-0.8 ft ft

-28.8 -0 2

160

-0 8

-2.2

52

D. Brown, A. Deardorff, S. Djankov, and R. Stern

While we have modeled the fixed and variable costs of monopolistically competitive firms as using the same proportions of direct capital and labor, we have allowed intermediate inputs to enter only into variable cost. As a result, when prices of intermediate inputs fall due to trade liberalization, marginal cost is reduced relative to average cost, and even without an increase in competition firms tend to expand in order to restore the optimal markup. Not reproduced here but available on request, the sectoral effects on the trade, output, and employment of the three regions of the EU are negligible. It appears therefore that it is the CECs themselves that are most affected by the reductions in tariffs and relaxation of NTBs that we have modeled in our various scenarios. It also appears, however, that if account were taken of the phasing in of the trade liberalization over a period of years, there would not be significant adjustment pressures experienced due to intersectoral shifts in labor and capital in the CECs and the EU. Even in the CECs, the negative sectoral changes in both employment and number of firms, both of which could be disruptive for the individuals involved, are no more than a few percent and could be accommodated easily within the normal turnover of workers and firms during a five- or ten-year period of implementation. Scenario D As mentioned earlier, the Europe Agreements have identified some sensitive sectors in the EU. These sectors, which are subject to NTB restrictions in the EU, include agriculture, textiles and clothing, chemicals, and iron and steel. Having collected information on the tariff equivalents of the various NTBs, we decided to conduct a hypothetical scenario involving a complete elimination of existing EU NTBs, along with the CEC-EU FTA presented in the Base Case. To conserve space we do not present the detailed sectoral results for scenario D, which are available on request, but we can report the major differences from the Base Case. As already noted, elimination of EU NTBs in these sensitive sectors was enough to reverse the deterioration of the terms of trade of the three CEC countries, and therefore to increase substantially the improvement in their welfare. At the sectoral level, this change was enough to cause notable improvements in employment in the sensitive sectors. In Czechoslovakia, for example, the additional hypothetical liberalization causes the employment increases to change from 1.4 percent to 13.1 percent in agriculture, from 0.2 percent to 6.9 percent in textiles, from 1.8 percent to 6.8 percent in chemicals, and from —1.7 percent to 12.1 percent in iron and steel. On the other hand, in the apparel sector, where employment declined by 2.7 percent, as shown in Table 2.6, additional liberalization of sensitive sectors including apparel caused Czechoslovakian employment to fall even more, by 6.8 percent. We presume that this result stems from the increased competitiveness of the EU clothing industries when given access to lower cost textile inputs. These results are typical of all three of the CEC countries modeled. These larger changes in em-

An economic assessment of integration into the EU

53

ployment in the sensitive sectors also led to a larger number of sectors in which output declines, in contrast to the Base Case where output expanded in almost all of the sectors of the CECs. 2.9

The roles of scale and variety

As explained earlier, the industry structure in all but the agricultural sector of this version of the Michigan CGE model is patterned after the Dixit-Stiglitz (1977) model of monopolistic competition. As such, both economies of scale and variety play a role in determining both the positive effects of policy changes and the welfare effects. In this section we report on two additional scenarios that we ran to decompose the results of the Base Case, separating out the effects of variety and scale. Table 2.9 reports the summary results of this exercise, repeating the results of Table 2.5 for the Base Case Scenario C, along with the new scenarios E and F. Scenario E repeats the analysis of multilateral removal of tariffs and nonsensitive NTBs among the CEC and EU countries, but it removes any effects of product variety from both the demand functions and the resulting levels of welfare. This is done by incorporating the number of varieties separately into the product aggregation function that enters utility and production functions and setting its parameter so as to exactly offset the effect of variety in the Dixit-Stiglitz aggregator. The results in Table 2.9 indicate that removing variety causes a small dampening of most trade and welfare effects of CEC-EU integration. The welfare improvement in Czechoslovakia falls from 4.5 percent in scenario C to 3.8 percent in scenario E. The dampening is smaller in Hungary, and welfare actually rises somewhat more in Poland. In all three countries the expansions of exports and imports due to integration are reduced somewhat by the removal of variety. These results are to be expected, since when variety effects are included, they increase the benefits to demanders in sectors where the number of products increase, and therefore increase both demand and the welfare derived from it. However, these results also indicate that the role of variety in our results is not very large, and that the major effects that we have identified as arising from integration are not due to the particular way that variety enters the model through the Dixit-Stiglitz aggregators. Economies of scale enter the model through the fixed and constant variable costs that are assumed for firms, following again the usual modeling of monopolistic competition in the new trade theory begun by Krugman (1979).27 If output per firm rises, fixed costs are spread over more units of output and average costs decline. As already noted in the sectoral results for scenario C, our results indicate that integration will increase output per firm in most industries 27

With more than one factor, in contrast to Krugman (1979), the factor intensities of fixed and variable costs could be different, leading to factor market effects from changes in scale. In our model, however, partly because we lack the data to make this distinction, fixed and variable factors of production are assumed to be used in the same proportions within an industry.

Table 2.9. Summary results of CEC-EU integration in the base case: CEC-EU FTA, tariffs and nonsensitive NTBs removed, decomposition of scale and variety effects Country/Region

Imports

Exports

Terms of Trade

$ Mill.

$ Mill.

(2)

(3)

Pet. Change (4)

Equivalent Variation

Wage Rate

Ret. to Capital

(5)

$ Millions (6)

Pet. Change (7)

Pet. Change (8)

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

Pet. Change

C. CEC-EU Free Trade Area, Nonsensitive NTBs Removed Czechoslovakia Hungary Poland NAFTA EFTA EU-South EU-Norfh Other

2,008.1 1,596.8 2,368.2 -8.5 1,230.3 151.3 4,061.5 15.6

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

E. CEC-EU FTA, Tariffs and Nonsensitive NTBs, with Scale Only Czechoslovakia Hungary

1,935.7 1,570.7

1,956.4 1,591.6

-0.1 -0.2

3.8 4.6

2,148.6 1,649.8

3.3 3.4

0.1 0.4

Poland NAFTA EFTA EU-South EU-North Other

2,348.7 9.1 1,213.6 157.1 4,002.9 29.2

2,427.8 5.1 1,099.0 174.0 3,889.9 -2.3

-0.5 0.0 0.0 0.0 0.0 0.0

3.9 0.0 0.3 0.1 0.1 0.0

3,061.9 -320.4 2,379.3 444.3 8,226.9 -332.2

3.8 0.0 0.1 0.1 0.0 0.0

-0.1 0.0 0.0 0.0 0.0 0.0

1,702.9 1,492.6 2,605.9 -615.4 1,800.4 203.7 5,895.4 -817 7

4.9 5.3 5.9 0.0 0.1 0.1 0.0

-1.1 -0.8 -1.8

00

00

F. CEC-EU FTA, Tariffs and Nonsensitive NTBs, with Neither Scale nor Variety Czechoslovakia Hungary Poland NAFTA EFTA EU-South EU-North Other

1,863.4 1,543.6 2,390.4 -85.8 1,081.7 123.8 3,553.2 -80 4

1,708.4 1,445.0 2,240.1 -35.4 1,001.1 145.8

0.9 1.0 1.0 0.0 0.0 0.0

3,723.9 -199

0.0

3.0 4.2 3.3 0.0 0.2 0.0 0.1

00

on

0.0 0.0 0.0 0.0

56

D. Brown, A. Deardorff, S. Djankov, and R. Stern

of the CEC countries, and therefore a portion of the effects that we have presented is due to such exploitation of economies of scale. To remove these effects, in scenario F we have held output per firm constant, requiring therefore that changes in industry output occur only through entry and exit of firms. Once again, comparing scenario F with scenarios C and E in Table 2.9, it is clear that the removal of scale effects causes a further dampening of the welfare benefits of integration for all three of the CEC countries. Trade too expands somewhat less for most of their exports and imports. Again, these results are to be expected, since scale economies increase both the incentives for and the returns to specialization and trade. However, here again the dampening effects only partially reduce the effects that were observed in the Base Case of scenario C, and the conclusions that we drew from the Base Case are largely unchanged. Thus we conclude that our results have not been particularly dependent on our modeling of the effects of scale and variety. There is one place where the removal of scale effects has increased rather than decreased the response to integration: real factor prices. In scenario C we reported increases in the real wage in the neighborhood of 3 percent for each of the CEC countries, and changes in the real returns to capital that were near zero, positive or negative. Thus the relative returns to labor increased about 3 percent compared to capital. In scenario F, however, real wages rose by almost 5 percent or more in these countries, while real returns to capital fell by about 1 percent or more, for a relative change of around 6 percent. Thus it appears that removal of the scale and variety effects has made the model behave somewhat more like the conventional Heckscher-Ohlin-Stolper-Samuelson model in terms of effects on factor prices. We should mention, however, that even with the removal of both variety and scale, the Michigan CGE Model is not a completely neoclassical model. The model still includes markup pricing, and these markups change with changing market conditions, even when scale effects are absent. Indeed, other scenarios that we have run with this model but do not report here indicate that even in the absence of variety and scale effects it is possible for us to get wages and rents moving in the same direction, presumably due to this remaining element of imperfect competition. 2.10

Conclusions and implications for research and policy

Our purpose in this paper has been to analyze the possible economic consequences of CEC-EU integration, using for this purpose the Michigan CGE Trade Model. To provide background for our analysis, we first reviewed briefly some of the important sources of potential benefits of EU-CEC integration, as well as some of the existing literature pertaining to EU-CEC integration. Since the EU has already negotiated a series of bilateral Europe Agreements with individual CECs, we then sought to provide a synthesis of the main features

An economic assessment of integration into the EU

57

of these agreements. We also discussed briefly the important features of the CEFTA that the CECs had negotiated among themselves. Most of the rest of the paper was devoted to a description of some important characteristics and assumptions of the Michigan CGE Trade Model, laying out our computational scenarios, and presenting and discussing the main aggregate and sectoral results. The version of the Michigan Model that we used has eight countries/regions and twenty-nine sectors. The sources of welfare improvement identifiable by the model include the traditional effects of changes in terms of trade and gains from specialization and exchange. They also include the effects of labor moving between sectors of different productivity. In addition, the presence of economies of scale, product differentiation, and imperfect competition among firms allows us to identify the contribution to economic well-being of the procompetitive effects of trade liberalization, together with the effects of increased scale and variety. Four liberalization scenarios were performed. The first scenario examined the formation of the CEFTA. The second scenario modeled a CEC-EU freetrade arrangement, eliminating tariffs both among the CEC countries and between them and the EU. In our final two scenarios, we made allowance for changes in NTBs as well as the removal of tariffs. The results suggest that the economic welfare of the CECs would be increased by the CEFTA, and that integration with the EU would bring even greater welfare benefits. The EU regions also gain from CEC-EU integration, but the gains are quite small as a percent of EU GDP. The effects on regions outside of Europe are negligible. Within the individual countries/regions, the distributional consequences are relatively small. It is interesting that the real returns to labor and capital rise in every country/region. The reason for this result is that there are large enough gains from the realization of economies of scale that accrue to both labor and capital so as to offset the negative effects that would otherwise be expected from Stolper-Samuelson logic.28 At the sectoral level, there are rather small effects on trade, output, and employment for the CECs associated with the CEFTA. The effects are more sizable when there is CEC-EU integration in the form of tariff removal and still larger when NTBs are relaxed to permit greater imports. In Czechoslovakia and Poland, output and employment tend to expand across virtually all sectors, whereas Hungary exhibited a greater degree of specialization, with some sectors expanding and others contracting. In our modeling, we assumed that the tariffs and NTBs would be changed at a single point in time. If allowance were to be made for the phasing in of the changes, it is unlikely that there would be any serious adjustment pressures felt in the CECs in connection with the integration process. It is especially noteworthy that the sectoral impacts of EU-CEC integration on the EU regions specified in the model appear to be negligible, 28

This does not mean that the scale effects themselves are particularly large in absolute terms, but only relative to the Stolper-Samuelson effects, which are also rather small.

58

D. Brown, A. Deardorff, S. Djankov, and R. Stern

and, consequently, adjustment pressures in the EU would be unlikely to be experienced. Our research on EU-CEC integration is by no means the last word on the subject. We have abstracted from the changes that CEC-EU integration might make in foreign direct investment and have not made any allowance for dynamic changes in efficiency and capital accumulation. Further, we have not included other aspects of the European Agreements besides tariffs and NTBs. Granting these limitations, our research provides some insights into the economic consequences of EU-CEC integration and confirms the conclusions of previous studies on EU-CEC integration that the EU has little to fear, while the CECs stand to gain significantly. REFERENCES Aghion, P., and P. Howitt (1992). "A Model of Growth through Creative Destruction," Econometrica 60, 323-53. Aghion, P., et al. (1992). "Towards the Establishment of a Continental European Customs Union," in J. Flemming and J. Rollo (eds.), Trade, Payments and Adjustment in Central and Eastern Europe, London: EBRD and the Royal Institute for International Affairs. Armington, Paul S. (1969). "A Theory of Demand for Products Distinguished by Place of Production," IMF Staff Papers 16,159-76. Bakos, Gabor (1993). "After COMECON: A Free Trade Area in Central Europe?", Europe-Asia Studies 45, 1025-44. Baldwin, R. (1994). Towards an Integrated Europe, London: Centre for Economic Policy Research. Brown, Drusilla K., and Robert M. Stern (1989). "Computable General Equilibrium Estimates of the Gains from U.S.-Canadian Trade Liberalization," in David Greenaway, Thomas Hyclak, and Robert J. Thornton (eds.), Economic Aspects of Regional Trading Arrangements, London: Harvester Wheatsheaf. Brown, Drusilla K., Alan V. Deardorff, and Robert M. Stern (1992a). "A North American Free Trade Agreement: Analytical Issues and a Computational Assessment," The World Economy 15, 15-29. Brown, Drusilla K., Alan V. Deardorff, and Robert M. Stern (1992b). "North American Economic Integration," Economic Journal 102, 1507-18. Brown, Drusilla K., Alan V. Deardorff, and Robert M. Stern (1993). "Protection and Real Wages: Old and New Trade Theories and Their Empirical Counterparts," Research Forum on International Economics, University of Michigan discussion paper, May. Brown, Drusilla K., Alan V. Deardorff, and Robert M. Stern (1995). "A Free Trade Agreement Between Tunisia and the European Union: Effects on Tunisian Trade and Foreign Direct Investment," in process. Brown, Drusilla K., Alan V. Deardorff, and Robert M. Stem (1996). "Computational Analysis of the Economic Effects of an East Asian Preferential Trading Bloc," Journal of the Japanese and International Economies, 10, 37-70. Brown, Drusilla K., Alan V. Deardorff, David L. Hummels, and Robert M. Stern (1994). "An Assessment of Extending NAFTA to Other Major Trading Countries in South America," University of Michigan, in process. Brown, Drusilla K., Alan V. Deardorff, Alan K. Fox, and Robert M. Stern (1995). "Computational Analysis of Goods and Services Liberalization in the Uruguay Round,"

An economic assessment of integration into the EU

59

in Will Martin and L. Alan Winters (eds.), "The Uruguay Round and the Developing Economies," World Bank discussion paper 307. Cadot, Olivier, and Jaime de Melo (1994a). "The Europe Agreements and EC-LDC Relations," CEPR discussion paper 1001. Cadot, Olivier, and Jaime de Melo (1994b). "France and the CEECS: Adjusting to Another Enlargement," CEPR discussion paper 1049. Cecchini, P. (1988). 1992, The European Challenge: The Benefits of a Single Market, London: Aldershot & Hants Publishers. Corado, Cristina (1994). "Textiles and Clothing Trade with Central and Eastern Europe: Impact on Members of the EC," CEPR discussion paper 1004. Deardorff, Alan V., and Robert M. Stern (1990). Computational Analysis of Global Trading Arrangements, Ann Arbor: University of Michigan Press. Dimelis, Sophia, and Konstantine Gatsios (1994). "Trade with Central and Eastern Europe: The Case of Greece," CEPR discussion paper 1005. Dixit, Avinash K., and Joseph E. Stiglitz (1977). "Monopolistic Competition and Optimum Product Diversity," American Economic Review 67, 297-308. Ethier, W. (1982). "Decreasing Costs in International Trade and Frank Graham's Argument for Protection," Econometrica 50, 1243-68. European Commission (1994a). "European Union Relations with the Countries of Central and Eastern Europe," Background Brief. Brussels, January. European Commission (1994b). "What is PHARE? A European Initiative for Economic Integration with Central and East European Countries," Brussels, June. Gasiorek, ML, A. Smith, and A. J. Venables (1992). "'1992': Trade and Welfare - a General Equilibrium Model," in L. Alan Winters (ed.), Trade Flows and Trade Policy after "1992," Cambridge: Cambridge University Press. Grossman, G., and E. Helpman (1992). Innovation and Growth in the Global Economy, Cambridge: MIT Press. Halpern, Laszlo (1994). "Comparative Advantage and Likely Trade Patterns of the CEECs," CEPR discussion paper 1003. Helpman, E., and Paul Krugman (1985). Market Structure and Foreign Trade, Cambridge: MIT Press. Hoekman, Bernard, and Gerhard Pohl (1995). "Enterprise Restructuring in Eastern Europe. How Much? How Fast? Where? Preliminary Evidence from Trade Data," The World Bank, in process. Krugman, Paul (1979). "Increasing Returns, Monopolistic Competition, and International Trade," Journal of International Economics 9,469-79. Krugman, Paul (1993). "The Uncomfortable Truth about NAFTA," Foreign Affairs, Nov./Dec, 13-19. Martin, Carmela, and Jordi Gual (1994). "Trade and Foreign Direct Investment with Central and Eastern Europe: Its Impact on Spain," CEPR discussion paper 1006. Messerlin, P. (1992). "The Association Agreements between the EC and Central Europe: Trade Liberalization vs. Constitutional Failure," in John Flemming and J. M. C. Rollo(eds.), Trade, Payments and Adjustment in Central and Eastern Europe, London: EBRD and the Royal Institute for International Affairs. Neven, Damien (1994). "Trade Liberalization with Eastern Nations. How Sensitive?", CEPR discussion paper 1000. Rollo, Jim, and Alasdair Smith (1993). "EC Trade with Eastern Europe," Economic Policy 16, 139-81. Romer, P. (1994). "New Goods, Old Theory, and the Welfare Costs of Trade Restrictions," Journal of Development Economics 43, 5-38. Stern, Robert M., Alan V. Deardorff, and Drusilla K. Brown (1992). "A U.S.-Mexico-Canada Free Trade Agreement: Sectoral Employment Effects and Regional/Oc-

60

D. Brown, A. Deardorff, S. Djankov, and R. Stern

cupational Employment Realignments in the United States," Appendix A in National Commission for Employment Policy, The Employment Effects of the North American Free Trade Agreement: Recommendations and Background Studies, Special Report No. 33 (October), Washington, D.C.: National Commission for Employment Policy. Winters, L. Alan (1994). "The Liberalization of European Steel Trade," CEPR discussion paper 1002. Winters, A., and Wang, Z. (1994). Eastern Europe's International Trade, Manchester: Manchester University Press.

PART II INVESTMENT PATTERNS

PART III LABOR MARKET ISSUES

PART IV THE PROCESS OF INTEGRATION

Index

acquis communautaire, 17, 18,318 agricultural sector, 30, 322, 347 in Eastern Europe, 240, 329 in Czechoslovakia, 52 in Poland, 48, 230 agriculture, 4-6, 15-17, 25-26, 28-30, 41^2," 68, 90, 116, 123, 196, 217, 259, 283, 316, 318, 336, 347, 349 in Bulgaria, 244 in Czechoslovakia, 48, 52, 65, 342 in Eastern Europe, 81, 89, 244-246 in Hungary, 48 in Greece, 342 in Poland, 48 in Portugal, 342 in Slovenia, 342 antidumping, 25, 27, 154, 336 apparel, 5, 42, 68, 86, 90, 343 Czechoslovakia, 52 Eastern Europe, 5, 25 exports, 65 asylum policies, 280, 283 asymmetry, 26, 348 Aussiedler, 278 bank lending, 171, 176 bankruptcy, 213, 230, 233, 237, 247, 318 banks, 210-211, 216, 220-221, 223-224, 228, 255, 261 in Czechoslovakia, 233 in Hungary, 232 in Poland, 232 351

bond financing, 171 border controls, 14, 279, 281 budget deficits, 10, 176, 195-196,223, 254 in Bulgaria, 215 in Hungary, 184,215,249 in Romania, 215 in Russia, 215, 249 Bulgaria, 1, 10, 14, 67-68, 160, 202, 230,238,260,319 Association Agreements with, 316 and the Common Agricultural Policy (CAP), 323 domestic savings, 174-176, 209 emigration, 271, 276, 278, 283, 325 employment, 244, 271, 276 European Union membership, 323, 333, 339 exports, 107 GDP, 214, 271, 325 GNP, 135 inflation, 183 investment in, 91, 167 privatization, 233, 255 Structural Funds, 322 trade with Germany, 101, 147, 150 business cycles, 163, 328 Canadian-United States Trade Area (CUSTA), 29-30 capital, 31, 52, 84, 98, 154, 172, 198, 216-217, 219, 230, 261, 269, 288, 291,293

352

Index

capital (cont.) accumulation of, 9-10, 39, 58, 167-170, 187-188, 192-193, 195-196, 214, 225-226 deepening, 237 efficiency, 10, 189, 191 flows, 5, 10, 171, 177-178, 193, 213, 248, 259, 329 foreign capital, 10, 171, 178-179, 186, 192-193,221 formation, 24, 212, 224, 226, 244, 247 human capital, 9, 12, 29, 41, 118-119, 135, 145, 150-153, 158, 162-163, 169, 222, 224, 226, 247, 301, 337 imports, 174, 259 income, 11,250 intensity, 3, 11,99, 106, 151-153, 163, 222, 227, 237-239, 249-252, 329 markets, 92, 183, 215, 220-221, 224, 227, 232-233, 235, 247, 254, 259, 329, 349 mobility, 2, 16, 157, 279, 284-285, 316,338,349 physical capital, 29, 64, 135, 162, 169, 190, 226 productivity, 151, 213, 222, 225, 235, 237 reallocation of, 168, 187 returns to, 4, 23, 39, 47-18, 56-57, 61, 226, 307 stock, 9-10, 64-65, 105, 162, 167, 171, 187-190, 192-193, 196-197, 222, 225, 235, 237-238,313,329 capital-labor ratio, 3, 118, 122 capital-output ratio, 167 CEC-EU Free Trade Agreement, 42-45 CEEC-6 Europe Agreements, 316 exports, 69, 74, 77, 79-81, 87, 89, 97-98 GDP, 99 and Germany, 69, 73-74, 76, 80-81, 91-92, 97-98 imports, 69, 74, 77 income, 88, 99

integration, 315-316 investment, 91-92 labor, 92-93 population, 87-88 revealed comparative advantage of, 76,81,98 trade, 68-69, 73-74, 76, 80-81, 83, 87-89,91,97 Central and Eastern European Countries (CEECs), 176, 178 CEFTA, 14, 16, 338-339, 345 and the Common Agricultural Policy, 15-17,154,167,197,319,321, 323,331,334-335,343,347 Europe Agreements, 2, 6, 9, 14, 65, 154,186,316,318,334 and the European Union, 1-3, 15-18, 123, 167, 170-171, 182, 191, 196-197,201,319-323,325, 327-329,331-335,337, 339-340, 342-343, 346-349 exports, 7-8, 81, 101, 105-107, 123, 150-151, 153-154, 164, 186 and Germany, 7-9, 18, 74, 76, 100-101, 105-107, 135, 145, 147, 150-154, 162-164, 342 GNP, 123, 135, 145, 147, 150 imports, 6, 8-9, 89, 101, 105-106, 150-151,153,163-164,204 income, 2, 6, 8, 145, 150-151, 153, 162, 167, 342, 346 integration of, 2-3, 14-15, 68, 202, 204-205,315-316,318-321, 326-327, 329-337, 340, 346-347, 350 investment in, 6, 9, 11-12, 167, 174, 179, 182-184, 186, 196 labor, 12-14, 18, 152, 167, 186, 321, 325-326, 330 privatization, 9, 11-12, 18, 189 productivity, 10-12, 195 reform, 11, 190-191,327 Regional and Structural Funds, 15, 17, 167,321-322,331,334,344 revealed comparative advantage of, 6, 81 trade, 2-3, 5-7, 81, 86-88, 100-101, 106-107, 116-117, 145, 147,

Index 150-153, 162-164, 168, 325, 330, 333, 348-349 transition process, 1, 9, 67-68, 88-89, 168, 170, 187-193 Central European Country (CEC), 30, 41 apparel, 5 CEFTA, 28, 42-48, 57 Common Agricultural Policy (CAP), 26 employment, 31, 38, 48 Europe Agreements, 5, 23, 25-26, 39 European Investment Bank, 25, 28 and the European Union, 4, 23-26, 28, 31-39, 42^6,48-53, 56-58 exports, 4-5, 27, 56 imports, 28, 56 Michigan Model, 29-30 unemployment, 31 wages, 38, 47-48 Central European Free Trade Area (CEFTA), 4, 14, 16, 23-24, 28-29, 42-43, 57, 338-339 chemicals, 81 Czechoslovakia, 52 employment, 52, 93 Europe Agreements, 52 and the European Union, 25, 343, 347 exports, 99, 106-107, 150 Germany, 106-107, 135, 150 imports, 8, 86, 89, 135 income, 9, 86 and nontariff barriers, 5, 42, 52 trade in, 5, 27, 68, 80, 90, 325, 349 clothing, 89, 122 Czechoslovakia, 48, 52 employment, 52 Europe Agreements, 25, 52 exports, 8, 86, 99, 151 Germany, 106, 151 imports, 80, 86, 123 income, 86 labor, 152-153 and nontariff barriers, 52, 123 and revealed comparative advantage, 81 Romania, 107 trade, 8, 27, 90 Cocom, 213, 248

353 cohesion, 248, 320, 322, 329, 337, 342, 348 COMECON, 168,325,339 command economy, 212, 214, 227 Common Agricultural Policy (CAP), 1, 2,41, 123, 167, 197, 331, 343-344, 347-348 and Eastern Europe, 26, 321-323, 332 and European Union membership, 15-17,42,319,334-335 reform, 16, 159, 329, 335-336, 340 communism, 1, 167, 189, 263 comparative advantage, 25, 43, 119, 150, 226 barriers to trade, 89 and Eastern Europe, 8, 9, 68, 76, 81, 89,92,98, 151,318,337,343 Europe Agreements, 9, 318 and the European Union, 318, 337, 343 Russia, 12 competition, 11, 46, 48, 65-66, 167, 237, 239,332 barriers to trade, 246 European integration, 15, 318-319, 336, 338, 348 Europe Agreements, 17, 25, 348 and Germany, 7, 151, 157 and Heckscher-Ohlin, 251 imperfect, 4, 24, 56-57 import, 160, 223, 248, 259, 284 monopolistic, 3, 30, 52-53 perfect, 116 Poland, 230 policy, 17, 27, 213, 223, 248, 254, 258,318,336,348-349 and privatization, 213, 216, 223, 227, 230, 246-248, 250, 258, 261 Romania, 233 state-owned firms, 214 unemployment, 12, 325 consumer goods, 117, 249 and Eastern Europe, 7, 8, 106, 153 and the European Union, 123, 135 Germany, 8, 106, 135, 151, 153 imports, 7, 123, 153 OECD, 123, 134 convergence criteria, 206, 333nll, 335 corporate governance, 12, 224, 259-262

354

Index

Council for Mutual Economic Assistance (CMEA), 135, 214, 238-239, 247, 255 agriculture, 240 capital, 237, 252 collapse, 24, 98, 100, 105, 187-188, 212,237 and East Germany, 101, 105, 212 employment, 234-235 GNP, 253 income, 147,226,248,251 privatization, 213, 221, 226, 229, 234, 254 productivity, 220, 234-235 trade, 101, 105 unemployment, 220, 234-235 current account, 65, 177-179, 184, 188, 192, 197, 210, 232 Czechoslovakia (CSFR), 23, 29, 47, 183 agriculture, 48, 240 apparel, 52 banking system, 177, 232-233, 261 and CEFTA, 4, 42, 57 emigration from, 278, 283-284 employment, 48, 52, 57, 235, 240, 283-284 Europe Agreements, 316 European integration, 53, 61, 65-66, 240, 260 foreign direct investment, 64-65 exports, 150 GDP, 342 and Germany, 101, 106-107, 135, 147, 150, 153, 260, 283-284 GNP, 135, 147 imports, 107 income, 215 inflation, 215 and nontariff barriers, 4-5, 42, 52, 57 Portugal, 63 privatization, 177, 229, 232-233, 235, 255 productivity, 234-235, 238, 212 services, 61 trade, 101, 106, 147, 153 Czech Republic (see also Czechoslovakia), 1, 10, 23, 67, 101, 211, 239, 271,316

CEFTA, 29 efficiency, 202-204 Europe Agreements, 14, 333 European integration, 315, 318, 333 GDP, 214, 230, 276 and Germany, 7, 68, 91, 252, 263, 281-284 growth, 183-184, 195, 219, 233-234 income, 167, 209 investment in, 91, 167, 179, 186, 232, 250, 252 labor, 189,219-220,234 Michigan Model, 4 privatization, 12, 177, 186, 230, 233, 238, 254, 264 productivity, 212, 219-220, 234, 254 savings, 174, 209 trade, 7, 68, 252 unemployment, 190, 220, 226, 234, 271,325 wages, 271 demographic development, 270, 276 disequilibrium, 290, 296, 300, 307 distance, 2-3, 5, 7-8, 84, 86, 90, 98, 116-119, 122-123, 134-135, 145, 147,150,186,229,263 Dixit-Stiglitz, 30, 38, 53 Eastern Enlargement, 14, 16-17, 320-321, 329, 331-332, 335-336, 347 East Germany, see Germany economies of scale, 48, 84, 319 Dixit-Stiglitz, 53 and Germany, 158 Heckscher-Ohlin, 5 investment and, 226, 252, 336-337 Michigan Model, 4, 29 productivity, 46, 222, 225, 240, 247 trade, 5, 24, 46, 53, 56-57, 158, 226, 252 ECU, 15, 26, 322-323, 328, 333, 348 education, 28, 64, 153-154, 222, 226-227, 258, 284 efficiency, 39, 168, 171, 198, 206, 210, 259 Bulgaria, 202

Index capital, 167, 187, 189, 197, 208, 226 competition, 246-247 Czech Republic, 202 and Eastern Europe, 11, 24, 58, 190, 193,319,349 European integration, 24, 58, 61, 223, 318-319,343 and the European Union, 10, 193, 318, 331,349 full efficiency, 172, 197, 203-204 Hungary, 202 labor, 167, 187, 189, 197,217 Poland, 203, 230 privatization, 11, 217, 220-221, 223, 230, 246, 262 productivity, 172, 193, 196, 203, 235 resources, 9-11, 24, 190-191, 238, 246 Romania, 202 trade, 46-47 efficiency frontier, 202, 203 efficiency wages, 217 elasticity, 48, 209, 217, 222, 225-226, 303-304, 328 of distance, 122, 134, 150 of exports, 119, 134, 150 of imports, 84, 86, 50 income, 122, 134, 150 of labor, 13,290-291,307,311 substitution, 38, 168, 191, 197-198 trade 134 electrical machinery, 106-107, 123, 135 employee ownership, 217, 230, 252 employment, 28 agriculture, 240, 244 Bulgaria, 244, 271 Czechoslovakia, 48, 52, 57, 65, 283 in Eastern Europe, 5, 31, 38, 48, 52, 57, 151, 168 and European integration, 4, 23, 31, 38, 64, 67 in the European Union, 6, 18, 48, 52, 271,283 France, 92, 94, 283 full, 9, 13, 18,31,38,203,249, 290-291,293,296-297 Germany, 92, 151, 157, 159, 283, 311 Hungary, 48, 219, 230, 235, 283

355 immigration, 13-14, 18, 94, 285, 293, 296-297,300-301,310-311 income, 159 Michigan Model, 4, 23 Poland, 48, 57, 219, 230, 235, 244, 283, 326 Portugal, 92, 95-96 privatization, 228, 235, 258 Romania, 244, 271 in "sensitive" sectors, 18, 52-53, 67, 92 in service sector, 65, 204-205, 208, 244 Slovenia, 235 trade, 31,84,92, 151, 157 wages, 38, 276, 285, 288, 297, 301, 308,310 endogenous growth, 167, 225, 259 energy Bulgaria, 238 Czechoslovakia, 63, 238, 255 and Eastern Europe, 30 efficiency, 63, 238 Hungary, 63, 227, 238, 255 investment, 227, 258 Poland, 63, 227, 238 Portugal, 30, 63 privatization, 227, 234, 238-239, 255, 258 productivity, 222, 238 Russia, 159, 234, 251, 260, 329 entrepreneurship, 214, 224 Europe Agreements, 4, 27, 64-66, 192 and Eastern Europe, 2, 6, 9, 14, 17, 23, 25,39,42,56, 154, 186,336, 338-339 France, 3, 26 Greece, 3, 26 labor, 39 and nontariff barriers, 5, 42 Portugal, 3, 26 sensitive sectors, 2, 5, 25, 52, 316 Spain, 3, 26 trade, 23, 25, 42, 316, 318, 338-339, 348 European Bank for Reconstruction and Development (EBRD), 25, 28 European Economic Area (EEA), 14, 16, 315,333-334,336-339,344

356

Index

European Investment Bank (EIB), 25, 28 European Union (EU), 11, 206, 226, 255 budget, 321-323, 329 capital, 10, 52, 187, 192, 197-198, 204, 253 CEFTA, 28,42-43, 338-339 Common Agricultural Policy of (CAP), 1,17,41-42, 167, 197, 319-320,322-323,336,343 competition policy, 17, 27 and the Czech Republic, 61, 67, 318 efficiency, 10, 193, 204, 319, 331 employment, 6, 18, 23, 31, 38, 271, 283 EU-North, 5, 29,43 EU-South, 5, 29, 43, 62 Europe Agreements, 14, 25-27, 39, 52, 56-57,186,316,318,339 exports, 5, 246, 253, 343 France, 2, 26, 213, 276 GDP, 57, 214, 248, 276, 322, 330 Germany, 3, 17, 213, 276, 278-279, 310-311 Greece, 17, 25-26,43, 182, 320, 322, 342 and Hungary, 61, 66, 182,318 immigration, 269-270, 276, 278-281, 283-285,300,310,325-326 imports, 27,42, 77, 81, 98, 157, 186, 247, 348 income, 2, 3, 167, 248, 322, 329 integration, 1, 2,4-6, 9, 14-18, 23-25, 29, 31, 4 2 ^ 3 , 56-58, 61, 65-66, 167,196,205,264,315-316, 318-319,329-332,335-337, 339-340, 342 investment, 182,248 Italy, 3, 213 labor, 10, 38-39, 52, 186-187, 192, 197-198,204,270,279,281, 284,310,316,326,330 membership, 14-16, 170-171, 196, 240,264,315-316,318-320, 328, 332-337, 343 Michigan Model, 29, 57 nontariff barriers, 4, 38-39, 42, 52-53, 57 PHARE, 27-28

and Poland, 61, 66, 182,319 privatization, 213 Portugal, 3, 17, 25-26,43, 182, 276, 320, 342 productivity, 10, 188, 196, 201, 209 reform, 326-328, 334, 340, 350 sensitive sector, 26, 52, 77, 81, 97-98, 167, 325, 343, 347 service sector, 10, 191-192, 204, 240 Spain, 3, 17, 25-26, 43, 182, 213, 271, 276, 320, 342 structural funds, 25, 167, 319, 321, 331,336-337,344 trade, 2-6, 14, 16-17, 23-26, 31, 38-39, 42, 48, 57, 61-62, 65-66, 252-253,269,315,325,338 Treaty, 16,27,281,315,318 unemployment, 269, 271, 279 wages, 38-39, 279, 343 White Paper, 18,205,318 exchange rates, 3, 162, 222-223, 228 in Eastern Europe, 3, 9-10, 31, 159, 164, 171, 177-178, 188, 195-196 in the European Union, 3, 10, 31, 323 Germany, 159, 164 policy, 9-10, 195-196,254 trade, 10, 31, 164 externalities, 11, 225, 228, 258, 260 factor endowments, 3, 162-163, 319, 343 factor prices, 39,47-48, 56, 92, 204, 250 factors of production, 11, 24, 47, 53, 151, 220-221,264 financial intermediaries, 176, 196 foodstuffs, 9, 107, 122, 151 foreign direct investment (FDI), 24, 39, 58, 64-65, 171, 179, 182-184, 186-187,210,216,219,221, 223-224, 232-233, 248-249, 252-255, 326, 330, 337 Former Soviet Union (FSU), 11, 101, 105-107, 168, 186,211,346 France, 16, 67, 123, 260, 262, 344 and Eastern Europe, 3, 6, 29, 69, 74, 76,91,97,183,283,326 employment, 92, 94, 283 Europe Agreements, 3, 26

Index immigration, 13, 276, 278, 280, 283, 326 imports, 74, 76 investment, 183 labor, 91, 213 productivity, 261 trade, 3,6,69,90,97, 331 free-trade area (FTA), 4, 29,42-43, 46-48, 52 furniture, 8, 107, 123, 150-151 Generalized System of Preferences, 316 Geneva Convention, 281 geographical distance, 84, 116-117, 119, 122, 134, 145 Germany, 25, 29, 208, 342 Bulgaria, 68, 147, 150, 283-284 capital, 151-154,227,329 Czech Republic, 68, 106, 147, 150, 153,252,263,281,283 and Eastern Europe, 3, 6-8, 17, 67-69, 73-74,76,81,90-92,97-98, 100-101, 107, 123, 134-135, 145, 147, 150-153, 158, 160, 162-164,331-332,344 employment, 13-14,94, 157, 159, 281,283-284 ethnic Germans, 13, 278-279 and European integration, 16-17, 240, 331-332,344 exports, 6, 8-9, 69, 74, 92, 101, 105-107, 134-135, 147, 150-151, 153, 157, 163 Hungary, 68, 106, 147, 150, 153, 252, 263, 283-284 immigration, 12-14, 270, 276, 278-281, 283-284, 289-290, 300,310-312,330 imports, 6, 69, 76, 80-81, 92, 98, 101, 105-107, 134, 147, 150-151, 153,157-159, 163 income, 150-151, 159,212 investment, 91-92, 161, 183, 227 labor, 13, 18,91, 151-153, 160,212, 281,283-284,325-326,330 Poland, 68, 106, 145, 147, 150, 153, 263, 283-284 privatization, 213, 239, 262

357 productivity, 152-153, 163, 212 Romania, 68, 147, 150,283 Russia, 147, 284 sensitive sector, 6, 76, 81, 90, 92, 97 Slovak Republic, 68, 106, 147, 150, 152 trade, 3, 6-8, 68-69, 73-74, 76, 90-92,97-98,100-101, 105-107, 135 unification, 212, 328 wages, 13,38,271 gravity model, 2-3, 5-9, 23, 68, 74, 84, 98-99, 101, 107, 157-159, 162 Greece, 4, 29, 90, 240, 332 and Eastern Europe, 14, 17, 25,43, 182, 254, 322, 332-333, 337, 342-343, 347 Europe Agreements, 3, 26 and European integration, 14, 17, 25, 43, 320, 332-333, 337, 342-343, 347 immigration, 276, 278, 330, 343 heavy industry, 10, 168-169, 191-192, 196 Heckscher-Ohlin Model, 5, 47, 158, 162, 249-251 Hungary, 10, 14, 23, 61-62, 160, 177, 183-184, 239-240, 276, 316, 328, 342, 346 capital, 49, 186, 202, 219, 232, 237 CEFTA, 4, 29, 42 efficiency, 202 emigration, 271, 278, 283-300 employment, 48, 230, 235, 283-284 European integration, 1, 48, 53, 61, 65, 240,300,315,318-319,333 exports, 107 and Germany, 7, 68, 101, 106, 147, 153, 252, 260, 263, 283-284 GNP, 135, 147 imports, 232 income, 160, 167,316 investment, 64, 91, 167, 179-182, 186,210,216,232,249,252, 254, 260 labor, 203, 219, 220, 235, 283-284, 30 Michigan Model, 4, 29, 47-48

358

Index

Hungary (cont.) privatization, 12, 186, 216, 227-228, 232, 235, 238, 254-255 productivity, 219, 234, 238 savings, 174, 215 trade, 4, 7, 42, 57, 66, 68, 101, 106, 147,150,153,252,316 wages, 47, 271 hyperinflation, 213, 215, 237 immigration, 13, 213, 269-270, 276-278,285-312 skilled immigration, 13, 288, 291-297, 309,312 unskilled immigration, 13, 288, 291-297 immigration policy, 12-14, 279-285, 301,309,311 income distribution, 250, 252, 259 industrial goods, 7, 84, 106, 186, 316, 325 inefficiency, 9, 167, 171, 187, 190, 220-221 inflation, 10, 172, 176-177, 183, 188, 215,223,254,276,332 infrastructure, 9, 208, 220, 222, 254, 347 capital, 221, 225, 227, 239, 247, 329 in the Czech Republic, 186, 239, 249, 258 in Eastern Europe, 159, 229, 239-240, 253,316,322,329,337 in Hungary, 186,239,253 in Poland, 186,239,249 privatization, 11, 217, 224, 227, 238-239, 258 innovation, 154, 157, 220-222, 224, 239-240, 248, 255 integration and Eastern Europe, 2, 14, 18, 23-25, 31, 42-13, 53, 56-58, 61, 65-66, 74,91,269,300,315-316,319, 328-329, 332-333, 335-339, 343, 346 effects of, 23-24, 4 2 ^ 3 , 56-58, 61, 64-66, 68, 338, 343 employment, 4, 23, 31, 64 and the European Union, 2, 14, 18, 23-25, 31, 42-43, 56-58, 61,

65-66, 213, 269, 300, 315-316, 319,328-329,332-338,343 gains from, 23-25, 56, 74, 65-66, 226, 334 and income, 2, 65 policy, 223 process, 2, 4, 6, 14, 16, 18,57, 315-316, 318-320, 328-329, 331-338,343,346 and trade, 2, 6, 23, 29, 53, 57, 65-66, 74,88-89,91,223,318,346 interest rates, 226, 250-252 in Eastern Europe, 10, 174, 176, 192, 195-196, 219, 237 and growth, 10, 172, 219, 254 and inflation, 10, 332 and savings, 172, 174, 176-177, 192, 195, 205, 209 Inter-Governmental Conference (IGC), 15-16,315,328 intermediate goods, 152, 221 intra-industry trade, 98, 160, 162 investment, 10, 222 and capital, 24, 29, 64, 167, 170, 192, 215, 217, 220, 224-227, 232, 235, 237-238, 246-247, 249 and Eastern Europe, 1, 7, 9, 12, 17, 24-25, 38, 58, 64-65, 91-92, 160-161, 167, 174, 179,182-184, 186-187,192,196,226-227, 232-233, 237, 244, 246, 249-250, 254-255 and the European Union, 17, 27, 58, 123, 187,330,337-338 financings, 167, 171-172, 174, 176-177,249 foreign investment, 2, 12, 24, 58, 64-65, 171, 179, 182-184, 186-187, 196,216,219,221, 223-224, 232-233, 252-253, 285, 326, 330 and Germany, 7, 91-92, 134, 151-154, 160-161,227 and growth, 167, 192, 215, 217, 221, 225, 227, 237, 244, 255 and infrastructure, 27, 217, 225-227, 238-239, 244, 247, 250, 254-255

Index investment goods, 7, 105-107, 117, 123, 134,151-153,217 investment patterns, 2, 9 portfolio investment, 179, 182 and privatization, 9, 215-217, 221, 223-224, 227-230, 232-234 productivity, 215-217, 221, 223, 225, 237-239, 244, 247, 249, 255 and trade, 123,225,237,230 Ireland, 29, 90, 145, 186, 240, 271, 276, 280, 322, 330, 337 iron, 5-8, 17, 25, 42, 52, 68, 77, 81, 84, 86, 88-90, 93, 98, 106-107, 123, 150 Italy, 3, 6, 29, 67, 69, 76, 90, 97, 123, 186, 213, 280, 330-331, 333, 343 labor, 31, 84, 116, 118, 207-208, 225 and Eastern Europe, 3-5, 9-10, 12, 23, 38-39,43,47,52,63-65,91, 100-101, 106, 150-153, 160-161, 163, 167, 171, 186-189, 191-193, 197, 204-205,212,219-220,227, 249-250, 270, 276, 279, 285, 310,322,326,330,334 and the Czech Republic, 189-190, 203, 219-220, 234-235 and employment, 13-14, 18,64-65, 189,203,205,219,235,254, 288, 350 and the European Union, 3-4, 10, 16, 18,23,38-39,43,52, 117, 171, 186-187, 191, 193,197-198,202, 270,276,281,284,310,318, 326, 334 and exports, 3, 99, 106, 123, 163, 251 France, 91,213 Germany, 7, 9, 13, 18, 91, 99-101, 106, 151-153, 157, 160-161, 163,212-213,251,283, 310-311,325,330 Hungary, 203, 219-220, 227, 234-235 and imports, 7, 106, 123, 152, 163 income, 11, 163, 211, 304 labor markets, 3, 12-13, 18, 38, 63, 91-92, 157, 160, 183, 186, 269-270,281,283-284,

359 288-289, 300, 307, 310-311, 318,326,330 labor mobility, 2, 16, 39, 157, 220, 270, 279, 284, 325, 330, 334 labor unions, 14, 18, 235, 288, 297 migration of, 12-14, 64, 250, 269-270, 285, 288-289, 296-297, 300-301, 304, 307, 309, 311-312, 322, 325-326, 330 Poland, 219-220, 227, 234-235, 240 Portugal, 91,330 privatization, 217, 219, 227, 235, 250, 261 and productivity, 57, 100, 152, 168, 190-191,202,212-213,219, 221-222, 234-235, 237-239, 247, 249, 261 returns to, 4, 23, 47, 56-57 Romania, 189,203,240 skilled labor, 4-5, 13-14, 18, 63-64, 92, 168, 186,249,285,288, 290-291,293,293,300-301, 308, 322 unemployment, 14, 18, 189-190, 220, 252, 288-289, 300, 322 unskilled labor, 4-5, 13, 63-64, 92, 249, 285, 288-291, 296-297, 300-301,304,308-309,311, 325-326 wages, 13-14, 18, 38-39, 47, 56, 64, 217,249-251,285,288-289, 291,297 language, 7, 105, 116-119, 122-123, 312,321,343 Laspeyres price index, 47 leather, 5, 8, 18, 48, 106-107, 122-123, 151-153 Maastricht Treaty, 11, 248, 281, 315, 327-328, 335, 337, 347-348, 350 manufacturing, 30, 48, 119, 122, 198, 216-217,234 in Eastern Europe, 7-8, 39, 150, 168-169, 196,207,244,246, 254,271 and Germany, 7, 18, 92, 271 and the Michigan Model, 4, 29 and trade, 7-8, 29, 118,252

360

Index

Marshall Plan, 171,328 Metzler effect, 216 Michigan CGE Trade Model, 3-4, 17, 24, 29, 56-57, 65 migration {see also immigration), 5, 297 and Eastern Europe, 12-13, 15, 17, 39, 64, 248, 250, 255, 269-271, 276, 278-279, 300, 325-326, 330, 333, 343, 347 and the European Union, 17, 39, 64, 269-271, 276, 278-281, 284, 293, 300, 325-326, 330, 333, 343, 347 and Germany, 13-14, 157, 269-271, 276,278-279,300,301,330 migration policy, 270, 279-281, 300-301 monetary union, 11, 105, 201, 205, 315, 318, 331, 333, 335-336, 344, 347 monopoly, 11, 18, 30, 224, 248, 255, 325 monopoly union, 288, 301, 308 motor vehicles, 8, 66, 106-107 natural resources, 3, 7, 84, 119, 150, 163, 221-222, 238, 244 network effects, 11, 227-228 Newly Industrialized Countries (NICs), 214,219,223,240,253 nontariff barriers (NTBs), 3-5, 25, 29, 38-39, 41-43, 47, 52-53, 57, 62, 65, 123 nontradables sector, 213, 221, 225, 248, 250-251 North American Free Trade Agreement (NAFTA), 4, 25, 29, 31, 43, 46, 48, 61 oil, 8, 12, 106, 123, 135, 150,251 optimistic scenario, 172, 195-197 optimum plant size, 214, 220 Organization for Economic Cooperation and Development (OECD), 90, 171, 206,212,214,221,238,247 and Eastern Europe, 8, 80-81, 101, 123, 147, 150-151, 159, 163, 220, 235, 237 exports, 7-8, 117-119, 159

and Germany, 8, 101, 107, 134, 145, 147, 150, 159, 162-163 imports, 7-8, 76, 80, 117-119 trade, 6-8, 25, 81, 101,117-119, 122-123 per capita income (PCI), 3, 39, 84, 86, 98-99,118-119,122-123, 135, 145,150-151, 153,212,214,219, 244, 253, 322, 328, 330 pessimistic scenario, 172, 195, 210 Phare Program, 25, 27-28 Poland, 1,4, 10, 23,42, 48, 63, 67, 135, 202-203, 239, 300, 316, 322, 332, 346 employment, 57, 190, 230, 235, 244, 271,283 Europe Agreements, 14, 64-65, 316 and European integration, 27, 29, 53, 61,65-66,240,315,318-319 and the European Union, 14, 27, 42, 29,53,281,283 exports, 5, 7, 107 GDP, 214, 217, 238, 276, 342 and Germany, 68, 91, 101, 106, 147, 150, 153, 260, 263 growth, 183-184, 195, 212, 217, 234, 254, 276, 328 investment in, 91, 167, 179, 186, 232, 250, 260 labor, 220, 234-235, 276, 279, 281, 283-284, 326 privatization, 12, 227, 229-230, 232, 234, 238, 254-255 productivity, 190, 212, 219, 234-235, 254 savings, 174, 177 trade, 66, 68, 101, 106, 147, 153, 333 political instability, 214, 228, 253, 255 political stability, 186, 253, 210-21 \, 337 Portugal, 4, 14, 25, 30, 182, 186, 214-215, 240, 254, 260, 276, 280, 320, 322, 330, 332, 337, 342-343, 347 and Eastern Europe, 3, 6, 17, 43, 63, 67-69, 74, 76, 81, 89, 91-92, 97 and employment, 92, 95

Index and Europe Agreements, 3, 26 trade, 3, 6, 68-69, 74, 76, 81, 89-91, 97 potential trade, 4-8, 67, 74, 83-84, 86-89,98-99,116,135 preferences, 13, 169, 187, 192, 198, 237 preference zones, 116-117, 119 privatization, 18, 28, 174, 177, 189, 213, 237,319,424 and efficiency, 11-12, 217, 221-223, 246 insider privatization, 234, 246 and investment, 186, 215-216, 223-224, 227, 232-233, 247 and labor, 11-12, 217, 235, 250-252, 258, 326 privatization policies, 217, 221, 224-225, 228-230, 232-234, 238, 252, 254-255, 259-262 and productivity, 9, 11, 196, 212, 214, 221,223,234-235,238,246, 258, 260-262 voucher privatization, 12, 229, 232-234, 264 product differentiation, 30, 57 production function, 9, 64, 167-168, 171, 188, 197, 221, 225, 227, 259, 285, 301,303 productivity, 51, 153, 172, 189, 209, 248, 303 capital productivity, 64, 168, 190, 208, 213, 221-222, 225, 235, 237, 254 and Eastern Europe, 10-12, 64, 100, 152,190,195-196,212-213, 219-223, 234-235, 237-238, 246-247, 254-255 and the European Union, 10, 195-196, 223, 237 labor productivity, 57, 64, 100, 152, 168, 190, 195,213,219,222, 234-235, 237-239, 249, 284 and privatization, 9, 11-12, 212, 221-222, 227-228, 234-235, 238, 246, 254-255, 258, 260-261 productivity growth, 11, 207, 212-214,219-224,227, 234-235, 238-240, 244, 246-247, 249, 254-255

361 purchasing power, 118, 168, 187-188, 189,214 purchasing power parity (PPP), 171, 178, 189,342 quick-entry, 14-16, 320-321, 325-326, 328, 330-335, 339, 346-347, 350 R&D, 119, 134, 151, 153, 224-226, 235, 248, 252-253 raw materials, 7, 29, 84, 106, 122, 264 recovery, 1, 8-11, 105, 145, 168, 170, 187,189-190,192,203 reform, 170 in Eastern Europe, 7, 9-11, 17-18, 27-28,81,89,101,168,187, 190-192, 196, 204, 208, 261-262, 264, 316, 319, 330, 334 institutional reform within the European Union, 10, 15, 27, 159, 187, 316,319,326-329,334-335, 344, 347 Regional and Structural Funds, 2, 15-17, 24,28,167,319,321-322, 331-332, 334, 336-337, 344, 348 revealed comparative advantage (RCA), 6,76,81,98 Romania, 1, 10, 15, 67, 184, 202-203, 215,247,326 and European integration, 316, 322-323, 325, 339 exports, 107, 147 GDP, 214, 284, 235, 238, 240, 271 and Germany, 23, 38, 101, 147, 150, 281,283-284 imports, 147 investment, 38, 167 labor, 190, 235, 244, 271, 279, 281, 283-284, 325 and privatization, 233, 235, 255 productivity, 189-190, 195, 235 savings, 174, 176-177,209 trade, 23, 101 Russia, 15, 213, 239, 253, 264, 316, 329, 346 employment, 220, 235 exports, 106, 147,215,251 GDP, 214-215, 219, 230, 235, 261

362

Index

Romania (cont.) and Germany, 8, 135, 145, 147, 159 imports, 106 investment in, 232, 249 labor, 251 privatization, 12, 18, 229-230, 232, 234-235, 246, 255, 260-261 trade, 8, 106, 147, 159 Rybczynski theorem, 3, 249 safeguard clause, 27 savings, 10-11, 170-172, 174, 176-177, 187, 192-193, 195-196, 201, 205, 209-211, 215-216, 226, 229-230, 259-260, 309, 329 Schengen Accords, 279, 280 sectoral effects, 5, 39, 48, 52, 68 sectoral gravity model, 3, 98 sectoral trade patterns, 68, 76, 97 "sensitive" sector, 42, 316 and Eastern Europe, 2-3, 5-6, 25, 39, 43, 46, 52-53, 68, 76-77, 80-81, 87-89, 92, 99, 343, 347-349 employment, 53, 84, 94-95 and the Europe Agreements, 2, 5, 25, 52 and the European Union, 3, 52, 77, 80, 95-96, 321, 325, 343, 347-349 exports of, 6, 76-77, 80, 97, 99 service sector, 10, 197-198, 239, 279, 349 and Eastern Europe, 48, 61, 92, 168-169, 191-192, 196, 204-205, 207, 237, 240, 244, 316 employment in, 5, 31, 48, 65, 92, 196, 204-205, 207-208, 217, 244, 247, 325, 338 trade, 2-4, 16, 29, 51, 92, 284, 316, 338 Slovakia, see Czechoslovakia, Slovak Republic Slovak Republic (see also Czechoslovakia), 4, 7, 10, 14, 23, 29, 67, 91, 167, 174, 183, 214, 230, 271, 276, 283, 315-316, 328, 342 Slovenia, 14, 160, 214-215, 219, 234, 316, 320, 322, 328, 333, 339, 346 social conflicts, 251, 255

southern enlargement, 14-16, 240, 320, 328,330-331 Spain, 4, 14, 25, 213, 280, 320, 322, 326, 330 and Eastern Europe, 3, 17, 43, 145, 182, 186,240,254,271,276, 332, 337, 342 Europe Agreements, 3, 26 trade, 3, 90, 343, 347 spillover effects, 221, 225-226, 228 standstill clause, 27 state-owned enterprises (SOEs), 9, 214, 220-221, 224, 227, 229-230, 233-235, 254 steel, 18, 90, 93, 283 and Eastern Europe, 7-8, 25, 27, 29, 42, 52, 77, 81, 89, 101, 106-107, 150-151, 316, 318, 325, 343, 347 and nontariff barriers (NTBs), 5, 25, 41-42, 52 trade in, 5-8, 4 1 ^ 2 , 77, 81, 86, 98-99, 106, 123, 150-151, 316, 318, 325, 349 structural change, 5, 11-12, 198, 212-213, 235, 239-240, 244, 246, 258-259,334 Stylized Europe Agreement Package (SEAP), 5, 23, 26-27 subsidies, 9, 24, 105, 154, 159, 224, 226, 230, 235, 237, 244, 258, 261, 319, 322 tariff barriers, 4, 24 tariff equivalents, 41-42, 52 tariffs, 4, 24, 29, 38-39, 41—43, 46-47, 52-53, 57, 107, 123, 154, 262, 316 technical progress, 221, 238 technology, 3, 7, 24, 29-30, 46, 64, 84, 86, 105, 107, 154, 157-158, 204, 208, 221-225, 235, 239, 244, 246, 248-249, 252-253, 271, 288 telecommunications, 25, 159, 227, 239, 260 textiles, 5-6, 8, 17, 25, 41-42, 52, 68, 80-81, 89-90, 93, 95, 98, 106-107, 122-123, 150, 152-153, 316, 318, 325, 343, 347, 349 time series, 9, 164, 176

Index transition process, 1-2, 9, 11, 16, 30, 67, 69,88, 168, 170, 191-193, 196-197, 205, 254 transportation costs, 3, 84, 106, 116, 222, 239, 254 two-sector model, 191, 198 Ukraine, 15, 213-214, 220, 229, 233, 235,251,255,264,316,346 unemployment, 13, 189, 195, 247 and Eastern Europe, 4, 64, 190, 196-198, 220, 226, 234-235, 244, 248-249, 252, 269, 308-309, 322, 325-326 and European integration, 31, 64, 248, 308-309, 326, 330 and the European Union, 12, 18, 193, 248,269,271,279 immigration, 12, 269, 279, 285, 288-289, 296-297, 300-301, 308-311,325 and privatization, 213-214, 217, 222, 226, 234-235 unemployment insurance, 217, 297, 300-301 utilities, 4, 12,232,264 utility function, 192, 198-199 valuation, 12, 169, 171,259 variable geometry, 14, 16-17, 335, 344 variety, 1-2, 10, 15, 18, 24, 29, 38, 41, 46, 53, 56, 97, 118, 163, 167, 252 visa policies, 13,280-281 Visegrad countries, 7, 10, 17-18, 106,

363 147, 176, 209, 213, 219-220, 230, 237, 240, 246, 254-255, 315, 343, 346, 348 wages, 61, 151,217,237,329 in Eastern Europe, 3, 12, 18, 38-39, 47,56, 150, 186, 189,210,249, 255,276,300,310,330 and the European Union, 3, 38-39, 47, 64,276,300,310,330 and immigration, 13-14, 18, 285, 288-289, 297, 300-301, 304, 307,310-311 real wages, 3, 39, 47, 56, 105, 189, 237,255,276,301,330 welfare, 23, 38, 43, 47, 57, 64, 198, 285, 311 and Eastern Europe, 4, 23, 26, 43, 52, 161, 192,330-332,343 and European integration, 1, 24, 42, 46, 53, 57, 66, 330-332, 343 and the European Union, 4, 23, 26, 43, 53,330-331,343 and trade, 4, 23, 39, 43, 46, 159-160 welfare effects of trade liberalization, 43, 46 Werkvertagsarbeitnehmer, 283 West Germany, see Germany winner's curse, 216, 254 X-efflciency, 204 X-inefficiency, 221 Yugoslavia, 107, 278

Comment J. David Richardson

Discussing this paper is a pleasure, something like watching a movie. I'm spared the tedious screenwriting, casting, shooting, reshooting, and editing. I have only to focus on the finished product. Is it compelling? Is it enriching? How does it come out at the end? Computable general equilibrium approaches are, indeed, more than a little like a movie. They simulate life. And this one does so in a compelling, enriching way. Here's how it comes out at the end. Rationalization rules victorious. Factor adjustment burdens are overwhelmed. Almost everyone lives happily ever after. That is to say, the authors portray integration of Czechoslovakia, Hungary, and Poland into the European Union as leading to three important consequences. There are significant predicted gains in economic efficiency, some of the conventional kind and some because surviving firms in these countries economize on fixed costs more effectively in the face of larger markets. There are only mild predicted displacements of workers from sectors that shrink, and also only mild adjustments in wages and capital returns (though displacements, i.e., deaths, of "firms" are not so mild). And the predicted gains in the typical citizen's standard of living are so large (between 3 and 7 percent) that most workers, even when displaced, gain back enough to offset any mild temporary displacement losses (rarely double-digit). There are also some interesting subplots in this movie. For example, peripheral characters don't fare very well (one is even named, nefariously, "other"). The NAFTA region and the (other) rest-of-world region are almost always hurt from eastward expansion of the European Union, though by very small amounts. That is, trade diversion does seem to afflict. The EU-South region never gains very much. The large services sectors in Czechoslovakia, Hungary, and Poland don't fare very well either; all shrink, and by substantial amounts relative to other shrinking sectors. So what might this moviegoer have missed in this fine piece of art? Not much, only a few things. MORK (mobility of real kapital) might have played some role; he's an interesting character. Or it might have been nice to shoot a 61

62

J. David Richardson

periphery-vs.-periphery sequence: adjustment stories by sector in the EU-South region, for example. Though such adjustments would almost surely be small in magnitude, this viewer would like to see if their sectoral features were the opposite face (mask?) of the sectoral adjustments in Czechoslovakia, Hungary, and Poland. And, finally, I couldn't find anywhere that the authors gave me the account of how the sizable rents from nontariff barriers - slush that sloshes across borders and among characters until surgically lanced - affected their calculations of economic health from EU widening. But enough carping from this theater-chair critic. I give the movie a "thumbs up."

Comment L. Alan Winters

This is a nice application of a model that has now become a standard within our profession. The modeling is generally appropriate to the task in hand and the results generally convincing. In addition, the authors have a sensitive intuition about what is driving their numerical results. All this is good for the authors, but bad for the discussant - 1 find that I have no major arguments with the analysis. As an alternative, therefore, I want to make a series of comments on the modeling strategy. These amount to suggestions about how future exercises might be conducted and requests for the authors to speculate about how the changes suggested would affect their results. The first issue is one for which the authors explicitly apologize: the use of the Portuguese input-output table for 1980-81 instead of national tables for Czechoslovakia, Hungary, and Poland. This is clearly a limitation and one which, I am sure, the authors will wish to address at the earliest possible moment. l One cannot tell a priori whether such an adjustment will make much difference to the results, but for at least one reason I suspect that it might. That reason is that although Portugal was at a similar level of development to that which we think pertains to Czechoslovakia, Hungary, and Poland currently, it was not so egregiously energy-inefficient. Thus I suspect that reverting to national data would increase the share of energy in incremental output and this, in turn, could have serious implications for the balance of trade constraint in the modeling exercise. Indeed, even before the new input-output tables are available, the authors may care to experiment by, say, doubling the energy coefficients in the existing Portuguese table. A second area where it would be highly desirable to modify the model is in the labor market. The current approach is as if there were a single homogeneous kind of labor with different nonpecuniary rewards to working in different sectors. In fact, however, the distinction between skilled and unskilled labor is a Both the World Bank and the GTAP modeling consortium are working on input-output matrices for these countries. Thus it should not be too long before the Michigan exercise can be rerun on national data.

63

64

L.Alan Winters

very important one to both sides of the integration modeled here. The big fear in Western Europe is that integration will cause unemployment or wage erosion among unskilled workers. Most commentators expect it to be good for skilled workers. In the East the issue is slightly different: We would expect it to export unskilled labor-intensive goods and thus boost wages and/or employment, but the outcome for skilled labor is less clear-cut, depending on the exact way in which it interacts with unskilled labor and capital in production functions. A further issue that would become amenable to analysis, even if only crudely, through the disaggregation of labor would be migration. Again the fear in the West is of inflows of unskilled labor while the fear in the East is, or certainly should be, the outflow of skilled labor. CEPR (1992) suggested the possibility that Eastern Europe could get into a vicious cycle of decline if skilled labor and physical capital were complementary. If skilled labor emigrated, the productivity of physical capital could decline leading to a decline, ceteris paribus, in the capital stock, which would further reduce the returns to skilled labor, and so on. I have been surprised and a little alarmed at the lack of attention that has been paid to the danger of a brain drain in analyses of East-West integration. The features of the Europe Agreement that pertain to mobility of persons - for instance the mobility of key personnel, the availability of education in Western Europe, and the like - suggest that Western companies will have plenty of scope for screening immigrants from the East. In that way they can select the best of them for local employment and undertake the costs of obtaining for them permanent or semipermanent employment rights. The recruitment of skilled labor from abroad makes a great deal of sense so far as Western Europe is concerned, but could, as above, prove costly in the long run for the Eastern partners. It would be very useful to have a model that examined the sensitivity of output and welfare to such migratory flows. (See Bauer and Zimmermann, Chapter 7.) The authors have explicitly rejected the alternative of modeling the adjustment of the capital stock. I appreciate their reasons for feeling that current technology for modeling the medium-term gains from the accumulation of capital is rather crude, but I feel that it is sufficiently important to warrant some attention. The purpose of integration, and still more the transition, was not to reap Harberger triangles, but to improve rates of economic growth and levels of development. To be sure, economists have not worked out how to model the main components of these objectives, but the medium-term growth bonus is one way in which they can contribute to the debate. I should have liked to have seen how this operated in this model. Related to the capital stock is the issue of foreign direct investment (FDI). Once the capital stock is endogenous, it is less important to model who owns it. There are, however, two ways in which it is useful to reflect FDI in this sort of exercise. First, FDI is viewed as a means of technology transfer. If one were bold enough to postulate the nature of this relationship, it would certainly be helpful to see how allowing for FDI improved the returns to the Europe Agreements. Second, FDI is one of the ways in which the West expects to benefit from

Comment

65

increased integration with Czechoslovakia, Hungary, and Poland. With poor capital stocks, those economies are likely to offer significant opportunities for investment, and West European companies expect, for reasons of proximity and familiarity, to be the best able to exploit them. The potential importance of this is illustrated by the attention the Europe Agreements pay to issues surrounding FDI: for example, the clauses guaranteeing the current account transactions associated with the repatriation of profits, the establishment rules, the mobility of key personnel, and the rules associated with disinvestment. Modeling protection is frequently the most difficult element of the computable general equilibrium modeling of integration. The authors' approach of treating nontariff barriers as if they were binding quantitative restrictions (QRs) - and hence precluded all quantity responses for a proportion of trade - is interesting, but it raises a number of issues. First, are all NTBs binding QRs? Or are some of them weaker forms of restriction or nonbinding? How does the existence of a binding constraint on part of the import flow affect the pricing of imports and the market behavior of domestic firms? How are the rents on the bound imports allocated? Do they really accrue only in the importing country? I do not know if addressing these issues in a different way would change the results much, but it would be useful to have a fully articulated framework for thinking about them. I now turn to the results. My first comment is merely to express surprise at some of the details. In particular it is surprising that the apparel sector declines as a result of the Europe Agreements. One of the most remarkable features of the actual data from the last few years is the huge growth in apparel exports especially in outward processing trade, which entails inputs flowing from west to east, not east to west as the authors imply. I am also surprised that Czech agriculture gains while Czech engineering loses employment. Finally, I quite understand why in models such as this services employment falls with the trade shock. But taking the transition as a whole, we expect services to expand. The problem arises because we start from a base that reflects 1992, not how the CEECs would be in, say, 2000 with the transition but without Europe Agreements. This may be an important consideration, given that the model generates proportionate changes from a base position. If the base had double the employment in services that the authors assume, increasing services competition would release more labor and allow greater increases in other sectors. The main set of results in this paper also shows the terms of trade for the eastern partners declining when they integrate with Western Europe. In the presence of scale effects, however, this is neither surprising nor alarming: they are, apparently, just slipping along their average cost curve. I should find it interesting, however, to see this effect broken down into an export and an import price effect and into an intra-European and extra-European effect. It is interesting to note that although integration boosts European incomes, it does not do so sufficiently to offset the effect of trade diversion on welfare in the rest of the world. That is, according to the Michigan Model, the rest of the world los-

66

L.Alan Winters

es from deeper integration in Europe, albeit by a very small amount. It is difficult to extract from the tables the extent of trade diversion within the extended European economy, and I would have appreciated a little analysis from the authors, who have access to the detailed results. Trade diversion is a potential threat in Czechoslovakia, Hungary, and Poland in sectors where they have high protection against non-EU suppliers. In particular, this is true of the motor vehicles sector, in which EU suppliers are granted guaranteed market shares. The scope for diversion in this sector looks, at least from a casual viewpoint, quite significant. Given that the results of this model refer only to static benefits, the effects of integration are quite strong for Czechoslovakia, Hungary, and Poland. That is, these countries can expect significant benefits from integrating with Western Europe. The principal driving force behind this, however, is not, I suspect, the opening of the EU market, but rather the opening of their own markets to competition from outside. Bearing that in mind, they would be well to remember that such benefits are available in even greater measure from a mostfavored-nation liberalization. I recognize that many of the benefits of the Europe Agreements have not been modeled in this exercise and that there are also strong political reasons for deeper integration within Europe, but it is important for Czechoslovakia, Hungary, and Poland to exploit their ability to trade with the rest of the world as well as with Western Europe. Specifically, it would be deleterious to their economic well-being if they were to raise barriers any further on imports from the rest of the world, and ideally they should lower them. REFERENCE CEPR (1992). Is Bigger Better? The Economics of Enlargement, Monitoring European Integration No. 3, CEPR, London.

CHAPTER 3

Potential trade with core and periphery: industry differences in trade patterns Hari Vittas and Paolo Mauro

3.1

Introduction

This paper analyzes the trade patterns that have evolved between a few Western European countries, on the one hand, and a group of formerly centrally planned economies in Europe, on the other hand, since the initiation of the transition process in these economies in the late 1980s. The analysis has two aims: first, to ascertain whether there have been significant differences, across Western countries, in the speed with which the potential macroeconomic benefits from the opening up of Eastern Europe have so far been realized; and second, to assess whether there is any evidence to date to support the frequently voiced concern that east-west European economic integration might adversely affect output and employment in some "sensitive" sectors in Western Europe. The group of transition economies considered in this study, denoted as CEEC-6, includes Bulgaria, the Czech Republic, Hungary, Poland, Romania, and the Slovak Republic. To keep the paper manageable, these countries have been grouped together, even though they differ from each other in some significant respects, including the extent to which they had historical ties with the Western countries in our sample and the precise time at which they began their transition to market-based systems. As for the sample of Western European countries, it includes the four largest economies of Europe (Germany, France, Italy, and the United Kingdom) and Portugal. Germany is a natural choice given that it has had close historical ties with the transition economies and has quickly reestablished its position as a dominant trade partner for most of the CEEC-6 by virtue of its geographic location Earlier versions of this paper were presented in a seminar on "Western Europe in Transition," held in Trieste, Italy, in October 1994, and a conference on "Europe's Economy Looks East - Implications for Germany and the EU," held in Washington, D.C., USA, in May 1995. The authors would like to thank Heinz Handler, Susan Collins, Geza Feketekuty, and Georg Winckler, for many useful comments and suggestions. They also gratefully acknowledge the excellent research and/or secretarial assistance provided by Jolanta Stefanska, Fritz Pierre-Louis, Peter Kunzel, Kristy Pettie, and Gordana Rodic. The views expressed are those of the authors and do not necessarily represent those of the International Monetary Fund.

67

68

Hari Vittas and Paolo Mauro

and economic size.1 The other three large European economies have been included in the analysis primarily for the sake of comparison with Germany's experience. Portugal is probably a less obvious candidate for inclusion in the analysis given that it does not have any significant historical ties with the transition economies and is not geographically close to any of them. It is nevertheless considered because its economic structure and stage of economic development bear some similarity with that of the CEEC-6; this has led many observers to predict that the macroeconomic and sectoral effects of Portugal's integration with the CEECs would be large.2 Thus, comparisons between the experience of Portugal and that of the other countries in our sample could be expected to shed some light on the relative importance of economic, as opposed to geographic, proximity in determining the speed with which the trade effects of integration have so far been attained. Much of the analysis is based on simple intercountry comparisons of the East-West trade partners that have emerged in the first few years (1989-94) of the postliberalization period in the CEECs. Nevertheless, use is also made of more rigorous analytical tools, including the so-called gravity model of foreign trade. An interesting, somewhat surprising, by-product of the paper is the finding that the gravity model performs well in explaining not only the aggregate trade patterns observed in European countries in the recent past but also the sectoral composition of trade flows between these countries. The rest of the paper is organized as follows. In the next section, we briefly review developments in the aggregate trade flows of each of the Western European countries in our sample with the CEEC-6. We find that while in all cases trade has expanded rapidly, the rate of expansion has varied significantly, being the fastest in those countries that are geographically the closest to the CEEC-6. Also, in all the Western European countries considered here except Portugal, the trade balance vis-^-vis the CEEC-6 has improved, tending to support the presumption that Western Europe has benefited from the market liberalization process in Eastern Europe at least as much as the transition countries themselves. Section 3.3 concentrates on analyzing sectoral trade patterns and in particular imports of "sensitive" products.3 It finds that such imports have 1

2

3

In 1993, Germany accounted for about 30 percent of the foreign trade of the Czech Republic and Poland and a little less than 25 percent of the foreign trade of Hungary and Slovakia. Its relative importance as a trade partner of Romania and Bulgaria was, however, significantly lower (15 percent and 11 percent, respectively). This prediction has been supported by the results of some empirical work. Baldwin, for instance, has estimated that Portugal's potential exports to the CEECs in the long run were eleven times as high as the actual exports to the CEECs in 1989. For comparison, the ratio of potential to actual exports for Germany was estimated in the same study to be only 1.2 (see Baldwin, 1994). In line with previous literature, sensitive products have been defined to include food and agriculture, textiles and apparel, chemicals, and iron and steel. These are all sectors which suffer at present from considerable excess capacity globally and in which protectionist trade policies are still prevalent. They are also sectors in which many transition economies either possess a clear comparative advantage relative to Western European countries or have a large potential to ex-

Potential trade with core and periphery

69

generally grown at a much slower pace than the overall trade between Western European countries and the CEEC-6. While relative demand conditions may have been partly responsible for this, it is difficult to avoid the conclusion that in at least some of the sensitive sectors, restrictive trade policies and practices in Western European countries have constrained the ability of the CEEC-6 to fully realize their export potential. At any rate, the evidence presented in this paper suggests that the concerns in Western Europe about the potential dislocation effects of the opening up of Eastern Europe have been grossly exaggerated. 3.2

Aggregate trade patterns

Tables 3.1 and 3.2 and Figure 3.1 present some key data on aggregate trade flows between each of the Western European countries in our sample and the CEEC-6. Several features are worth highlighting. 1. For all the Western countries we consider, trade with the CEEC-6 has expanded very rapidly since 1989. The average annual rate of increase in such trade has varied quite markedly, however, ranging from a low of 14 percent in the case of France to as much as 24 percent in the case of Germany (Table 3.1). 2. In all cases, trade with the CEEC-6 has grown more rapidly than total trade. Thus, the share of the CEEC-6 in the exports and imports of each of the Western European countries in our sample has risen. Notwithstanding this rise, however, the share of the CEEC-6 in the foreign trade of Portugal has remained totally insignificant, and in the cases of France and the United Kingdom this share is still extremely small. Only for Italy and especially for Germany have the CEEC-6 already emerged as important trade partners (Table 3.2). 3. All the Western countries, except Portugal, have experienced an improvement in their trade balance vis-a-vis the CEEC-6. This finding is especially noteworthy, considering that economic activity and the level of income have at best stagnated in the CEEC-6 over the period covered by the analysis (while they have grown markedly in Western Europe). It strongly suggests that, at least at the macro level, the liberalization of East-West trade relations has been even more beneficial to the West than it has been to the East. 4. Intercountry differences in the rate of increase of exports to the CEEC-6 have been substantial. In particular, Italy and Germany have done considerably better than the other large European countries in expanding their sales to the CEEC-6. Interestingly, Portugal's export performance has been the weakest in our sample (Figure 3.1). 5. Cross-country differences in the rate of growth of imports from the CEEC6 have been even more pronounced than the differences observed on the export pand their exports in the early phase of the transition process on account of the productive capacity they inherited from the period of central planning.

71

70

Table 3.1. Trade with the CEEC-6" In millions of U.S. dollars at current prices

Change (in percent)

In millions of U.S. dollars at constant (1985) prices

1989

1994

1989

1994

At current prices

At constant prices

Germany Exports Imports Total Trade Trade Balance

6,884 5,658 12,542 1,226

19,137 17,517 36,654 1,620

4,426 4,224 8,650

11,117 12,106 23,223

178 210 192

151 187 168

France Exports Imports Total Trade Trade Balance

1,162 1,541 2,703 -379

2,760 2,169 4,929 591

796 1,193 1,989

1,751 1,608 3,359

138 41 82

120 35 69

Italy Exports Imports Total Trade Trade Balance

1,459 2,438 3,897 -979

5,379 4,164 9,543 1,215

1,000 1,994 2,994

3,467 3,455 6,922

269 71 145

247 73 131

United Kingdom Exports Imports Total Trade Trade Balance

933 1,223 2,156 -290

2,450 2,051 4,501 399

745 956 1,701

1,844 1,594 3,438

163 68 108

148 67 102

Portugal Exports Imports Total Trade Trade Balance

22 39 51 -17

48 74 122 -26

16 32 48

32 59 91

118 88 139

100 84 90

Source: International Monetary Fund, Direction of Trade Statistics and WEO Data Base. ' Bulgaria, the Czech Republic, Hungary, Poland, Romania, and Slovakia.

to

Table 3.2. Relative importance of trade with the CEEC-6 in total trade (percentage) Exports

Imports

1989

1991

1994

1989

1991

Germany

2.0

3.0

4.6

2.1

2.8

France

0.7

0.9

1.2

0.8

0.8

1.0

Italy

1.0

1.4

2.8

1.6

1.4

2.5

United Kingdom

0.6

0.7

1.2

0.6

0.5

0.9

Portugal

0.2

0.4

0.3

0.2

0.3

0.3

Source: International Monetary Fund, Direction of Trade Statistics.

1994 4.7

Potential trade with core and periphery

350-

73

Export Volume (1985-100)

300-

Germany 1/'^r

/

^ ^ ^

..-••

200-

150-

100.,.-"••"

^

France /

\

/

/

^

^

^

Portugal

.

50-

85

. 86

, 87

88

89

90

91

92

93

92

93

94

300 -,

Import Volume (1985=100) 250200-

150-

100-

50 85

87

Figure 3.1: Evolution of trade flows of Germany, France, Italy, UK, and Portugal with CEEC-6. Source: IMF, DOT data base. I/Data until 1989 are for West Germany only.

side. Specifically, Germany's experience stands out as quite exceptional, with its purchases from the CEEC-6 trebling within a span of only five years. Figure 3.1 shows that the strong upward trend has persisted well after unification. By contrast, France's imports from the CEEC-6 over this period have increased by only 40 percent.

74

Hari Vittas and Paolo Mauro

It is interesting to explore whether these cross-country differences in the growth of trade with the CEEC-6 reflect cross-country differences in the potential for trade unleashed by the opening of Eastern Europe as opposed to differences in the speed with which such potential has been exploited so far. To address this issue, we compare actual trade with estimates of potential trade developed by Baldwin (1994). The results are reported in Table 3.3.4 The table suggests that cross-country differences in potential trade do help in some cases to understand the comparative trade performance of Western European countries vis-a-vis the CEEC-6. For example, France's relatively slow growth of imports from the CEEC-6 appears attributable in part to France's comparatively limited scope for expanding such imports. Similarly, Italy's seemingly strong export performance appears less impressive when allowance is made for its comparatively large potential for increasing its exports to Eastern Europe. Notwithstanding these qualifications, however, the main result of the calculations is to confirm that Germany has made much more progress than the other countries in our sample in realizing the trade opportunities provided by the economic transformation of the CEECs. This is true for both imports from and exports to the CEEC-6. At the other extreme lies Portugal, which appears so far to have exploited only a small part of its potential trade with the CEEC-6. The next question that begs for an answer is what factors account for Germany's exceptional record. Although a complete analysis is beyond the scope of this paper, there are several, somewhat interrelated, factors that are likely to have boosted the growth of its imports from the CEEC-6, in the period under review. First, its average rate of economic growth has been comparatively high, largely reflecting the post-unification boom. Second, the deutsche mark has appreciated substantially, in both nominal and real terms, over the 1989-94 period, thereby shifting demand from domestic supply to imports (including imports from the CEEC-6). Third, there has been a substantial redistribution of income from the old to the new German States, which has boosted consumer spending and imports. However, none of these factors explains why German exports to CEEC-6 have also grown at a comparatively fast rate. Indeed, the appreciation of the deutsche mark should have worked in the direction of inhibiting Germany's export growth relative to that of the other countries in our sample. Thus, we are inclined to conclude that Germany's economic size, its historical ties with the CEEC-6, and, perhaps above all, its proximity to the new markets have made 4

Baldwin developed two scenarios of potential trade between Eastern and Western European countries. Both scenarios are derived from a gravity model of bilateral trade flows estimated using a group of countries that can be regarded as well integrated into the world trading system. The first is a medium-term scenario that assumes that CEECs become as integrated into European trade as the average Western European country was in the 1990s. In this scenario, the real per capita incomes and populations of all countries are held constant at their respective 1989 levels. The second is a long-run scenario that allows, not only for trade integration, but also for a substantial catching up of the standard of living in the CEECs. In Table 3.3 we make use of both of these scenarios, although only the former is really relevant for our analysis given that the catching-up process in the CEEC-6 has only just started.

Table 3.3. Actual and potential trade with the CEEC-6 (million US $ and percentage) Germany

France

UK

Italy

Portugal Imports

Exports

Imports

Exports

Imports

Exports

Imports

Exports

Imports

Exports

A. Actual (1993)

9,773

10,012

1,576

1,307

2,854

2,537

1,581

1,182

22

38

B. Medium-term potential

8,759

9,549

4,670

5,079

5,346

5,795

3,869

4,185

269

272

39,959

NA

21,927

NA

25,484

NA

18,173

NA

1,281

NA

D. Ratio of A to B (percent)

112

105

34

26

53

44

41

28

8

14

E. Ratio of A to C (percent)

24

NA

7

NA

11

NA

9

NA

2

NA

C. Long-run potential

Sources: International Monetary Fund, Direction of Trade Statistics and WEO Database, and R. Baldwin 1994.

76

Hari Vittas and Paolo Mauro

it comparatively easy for German firms to exploit quickly the new opportunities for trade provided by the change in regimes in the CEEC-6. By the same token, it is also reasonable to think that producers in the CEEC-6 will have accorded priority to penetrating the vast German market, before embarking on the effort of exploiting the scope for trade offered by small and relatively distant markets, such as the Portuguese one. 3.3

Sectoral trade patterns

Sensitive products account for a high proportion of imports from the CEEC-6 (Figure 3.2). This proportion ranged in 1989 from over 40 percent in Germany, France, Italy, and Portugal to about 30 percent in the United Kingdom and in most countries it was distinctly higher than the corresponding share of sensitive products in exports. Since 1989 the annual rate of growth of sensitive imports from the CEEC-6 has been erratic. In the cases of Germany and Portugal, on average it fell short of the corresponding rate of growth of total imports from the CEEC-6. Thus, their share in total imports has declined significantly. In the other three countries in our sample, the ratio of sensitive to total imports has, on balance, changed little since 1989, although after an initial pickup this ratio has recently also been on a slight downward trend. The tendency for the share of sensitive imports in total imports to decline might reflect a comparatively slow rate of liberalization of import restraints on these products by the Western countries. But it might also be a consequence of other factors, such as comparatively low income elasticities of the demand for sensitive products, or supply shocks in the transition economies. Alternatively, the decline in the share of sensitive imports might merely indicate that the usual presumption that the CEEC-6 possess a comparative advantage in these products is not correct. The remainder of this section attempts to shed some light on which of these hypotheses lies closer to the truth, although because of data shortcomings and the rather limited experience that has so far accumulated in the postreform period it is at best possible to draw only tentative conclusions. A number of indicators are used in the analysis. First, the composition of sensitive imports is examined to determine if the decline in their relative weight is widespread or concentrated in a few products. In addition, the behavior of sensitive imports (from the CEEC-6) in the industrial countries in our sample is compared with a broader measure of import demand for these products in OECD countries. Second, "indicators of revealed comparative advantage" are calculated for various commodity groups, to see whether the CEEC-6 really enjoy a revealed comparative advantage in sensitive products. Finally, gravity equations for various sensitive products are estimated and used to calculate potential imports of sensitive products. A comparison of the ratio of actual to potential imports for these products with the same ratio for total imports is then made to assess the progress so far made by the Western European countries in realizing their potential for imports of sensitive products from the CEEC-6.

77

Potential trade with core and periphery 60

Exports to CEEC-6 50

40

France

y'

Germany 30

20

10

85

86

87

88

89

90

91

92

93

60

Imports from the CEEC-6 **>*-% France

Italy

-:>^ Germany

30

20

85

87

89

90

91

92

93

Figure 3.2: Ratio of sensitive to total trade of Germany, France, Italy, UK, and Portugal with CEEC-6 countries (percent). Sensitive sectors are Food (0), Chemicals (5), Iron and Steel (67), Textile Yarn (64), Clothing and Apparel (85); Standard International Trade Classifications in parentheses. Source: TARS data base.

Composition of sensitive imports from the CEEC-6 The decline since 1989 in the share of sensitive products in total imports is accounted for mainly by imports of agricultural products and iron and steel (Table 3.4). In both sectors, the trade regime of the European Union countries has remained highly protectionist and in a few instances this has been reinforced by

Table 3.4. Composition of total imports from the CEEC-6 (percentage) France

Germany

Italy

UK

Portugal

SITC Code

1989

1991

1993

1989

1993

1989

1993

1989

1993

1989

1993

Sensitive Commodities

44.5

40.8

37.6

45.6

44.3

42.8

42.4

30.2

29.9

43.1

21.7

OFood

13.5

11.3

7.3

16.3

11.4

20.5

12.8

5.8

3.4

10.5

5.7

5 Chemicals

6.7

6.9

4.7

9.8

8.7

8

10.5

8.4

8.8

12.6

7.2

67 Iron and Steel

7.8

5.6

5.1

5.2

2.5

7.1

4.7

4.8

3.8

16.9

4.9

2.7

2.9

2.8

3.2

2.8

3

4

3.7

4.6

3

3.3

84 Clothing

13.8

14.1

17.7

11.1

18.9

4.2

10.4

7.5

9.3

0.1

0.6

65 Textile

Other Commodities

55.5

59.2

62.4

54.4

55.7

57.2

57.6

69.8

70.1

56.9

78.3

2 Nonfuel Crude Materials

7.9

6

5

7.5

3.6

10.1

12.7

11.7

3.1

1.2

10.4

3 Fuels, etc.

7.8

6.1

3.9

12.9

6

27.4

2.2

11.3

8.3

10

16.1

17

17.7

19.2

8.8

13.2

6.2

15.7

13.5

23.3

8.2

16.4

9

16.1

19.1

11.3

19.1

9.8

20

18.2

20.7

33.6

28.1

10.8

10.8

12.6

13.1

12.7

1.9

5.4

13.3

11

3.6

6.7

6-65-67 Other Basic Manuf. 7 Machines &Transport Equipment 8-84 Misc. Manufactured Goods Other

3.1

2.5

2.6

0.9

1.2

1.8

1.5

1.8

3.8

0.3

0.4

Total

100

100

100

100

100

100

100

100

100

100

100

Source: IBRD, Trade Analysis and Research System data base.

80

Hari Vittas and Paolo Mauro

Table 3.5. Imports from the CEEC-6 (average annual rate of change from 1989 to 1993, percentage) Germany

France

Total Imports

25.1

3.9

6.6

Imports of Sensitive Products

20.0

3.2

7.3

Portugal

Other OECD

7.7

8.0

10.9

6.4

7.5

-9.0

7.7

-5.1

-5.2

-5.6

-7.1

-1.9

14.4

1.0

14.0

9.0

-6.1

4.4

Iron and Steel

12.4

-13.5

-3.8

1.8

-20.8

-1.7

Textile, Yarn

26.4

-0.1

14.8

13.6

10.2

10.8

Clothing

33.2

18.8

34.1

13.7

77

26.7

Italy

UK

of which: Food and Live Animals Chemicals

Source: OECD, Trade by Commodities.

the imposition of new restrictions. By contrast, the weight of imports of textiles and clothing in the total CEEC-6 imports of most countries in our sample has tended to rise. This may be interpreted as prima facie evidence that market access in these sectors was not greatly impeded by trade restrictions. As for chemicals, it is difficult to discern any clear pattern across countries.5 Table 3.5 compares the rate of change (over the 1989-93 period) of "sensitive" and total imports of the industrial countries in our sample from the CEEC6 with the rate of change of the corresponding imports of all other OECD countries, taken as a group. The growth of "other OECD" imports is used as a crude indicator of the evolution of global demand for CEEC-6 exportables. In comparing such growth rates, however, it is important to note that the group of other OECD countries includes countries whose imports from the CEEC-6 were also subject to some trade restrictions. The comparison suggests that the weakness of French and Portuguese imports of sensitive products from the CEEC-6 cannot be ascribed to any significant extent to low elasticities of the demand for these products (or to supply shocks in the exporting countries). For almost all categories of sensitive products, the growth of "other" OECD imports from the CEEC-6 was markedly higher than that of French or Portuguese imports and the gap was more pronounced than that for nonsensitive imports. The same conclusion can also be drawn for Italian and UK imports from the CEEC-6, although the differences between sensitive and nonsensitive imports are modest. For Germany, the picture is less clear-cut as the growth of its imports has been strongly influenced 5

This is perhaps not surprising given that the chemicals sector is a very large and differentiated sector in many European countries. It points to the need for a more disaggregated treatment of trade in chemicals than we have attempted in this paper.

Potential trade with core and periphery

81

by unification effects. Nevertheless, even for Germany the behavior of at least a few categories of sensitive imports from the CEEC-6, notably food products and iron and steel, appears to have been discouraged by the restrictive trade policies of the European Union in these sectors. Thus, a possible interpretation of thefindingsreported in Table 3.5 is that the countries in our sample have protected the "sensitive" sectors to a higher extent than other OECD countries have. Indicators of revealed comparative advantage Indicators of revealed comparative advantage (RCA) are plotted in Figure 3.3.6 The technique, introduced by Balassa (1965), assumes that the actual commodity trade pattern "reveals" a country's comparative advantage, based on relative costs and nonprice factors. In interpreting the figure, it should be kept in mind that the RCAs are calculated on the basis of actual trade patterns, which have been significantly affected by restrictive trade policies and by the distorted structure of production that the CEECs inherited from the era of central planning. Because of these distortions, the RCAs can convey very misleading information as to where the true comparative advantage of the CEECs lies and should not be taken at face value. They are useful only to the extent that they may give some indication as to the sectors in which the CEECs may have the greatest potential to expand their exports until distortions are eliminated. Figure 3.3 suggests that the CEEC-6 have a substantial revealed comparative advantage vis-a-vis their Western trade partners only in two of the five sensitive sectors considered, namely agriculture and iron and steel. In one of these two sectors, that is, agriculture, the revealed comparative advantage (RCA) of the CEEC-6 appears to have been eroded somewhat since the beginning of the reform process, indicating that the restrictive impact of the agricultural support policies of the West may have increased. The CEEC-6 also display a positive RCA vis-a-vis OECD countries in the clothing sector; however, this is not the case as far as their trade with Portugal is concerned. In the other sectors that previous analysts have classified as sensitive, namely textiles and chemicals, calculated RCAs favor the OECD countries rather than the CEEC-6. Finally, it is interesting to note that the CEEC-6 enjoy a revealed comparative advantage in some nonsensitive sectors, notably fuels, other crude materials, and basic manufactures. All in all, the evidence provided by the RCA indices supports two tentative conclusions. First, that it is not unambiguously clear that the CEEC-6 possess a comparative advantage in all sectors identified by previous literature as sensitive; and second, that in at least some of these sectors, the expansion of CEEC6 exports to Western European countries may have been hindered in recent years by the persistence or intensification of protectionist policies. 6

The formula used is RCAt = [(X. IX) - (Af. / M)] / [(X. /X) + (M. / M)] where X. and A#. are exports and imports of sector i to and from the CEEC-6 and X and M are total exports to, and imports from, the CEEC-6.

Hari Vittas and Paolo Mauro

82 100

OECD versus CEEC-6

75 m 1988-89

CSS 1992-93

50 25 0 -25 -50 -75 Crude -100

100

Food

Chemcls Textiles

Iron Steel

.Clothing Beyerag. .Matrls , Access. Tobacco exclFuels Fuels Oil/Fat

Basic Machines Misc. UnclassMnf. 2 / Transp.Equ.Mnf. 3/ Goods

Germany versus CEEC-6

75 SB 1988-89

ess 1992-93

50 25 0 -25

it

-50

il

-75 -100

100

Food

Chemcls Textiles

Iron Steel

Crude .Clothing .Beverag. .Matrls Access. Tobacco excl.Fuels Fuels

Oil/Fat

Basic . Machines Misc. Unclass. Mnf. 2 / Transp.Equ.Mnf. 3/ Goods

Crude .Clothing .Beverag. .Matrls Access. Tobacco excl.Fuels Fuels

Oil/Fat

lasic . Machines Misc. . Unclass. Mnf. 2 / Transp.Equ.Mnf. 3/ Goods

France versus CEEC-6

75 EB 1988-89

BS 1992-93

50 25 0 -25 -50 -75 -100

Food

Chemcls Textiles

Iron Steel

Figure 3.3: Indices of revealed comparative advantage (adjusted for trade imbalances)!/. Source: IBRD, "Trade Analysis and Research System" and authors' calculations. 1/ For each SITC sector, a positive index indicates a revealed comparative advantage, in favor of the OECD countries, Germany, France, Italy, UK, or Portugal, respectively. 2/ Excluding textiles, iron, and steel. 3/ Excluding clothing and accessories.

83

Potential trade with core and periphery 100

Italy versus CEEC-6

75 SB 1988-89

ES 1992-93

50 25 0 -25 -50 -75 -100

100

FoodChemcls Textiles

Iron Steel

.Clothing Beverag. .Matrls . B 8 Access. Tobacco excLFuels Fuels

Oil/Fat

Basic .Machine? Misc. .Unclass, Mr\T. 2 / Transp.Equ.Mnf. 3 / Goods

Crude .Clothing ,Beverag. .Matrls Access. Tobacco excl.Fuels Fuels

Oil/Fat

Basic , Machines Misc. Unclass. Mnf. 2 / Transp.Elqu.Mnf. 3/ Goods

Oil/Fat

, Basic . Machines Misc. . Unclass. Mnf. 2 / Transp£qu.Mnf. 3 / Goods

UK versus CEEC-6

75 SB 1988-89

ess 1992-93

50 25 0 -25 -50 -75 -100

100

Food

Chemcls Textiles

Iron Steel

Portugal versus CEEC-6

75 SB 1988-89

ES 1992-93

i

50 25

Hi

0 -25 -50 -75 -100

Food

Chemcls Textiles

Iron Steel

Crude .Clothing .Beverag. .Matrls Access. Tobacco excl.Fuels Fuels

Figure 3.3: (cont.)

Sectoral gravity equations In this section we present estimates of potential trade between Eastern and Western European countries and compare them to actual trade flows. Following standard methodology, we first estimate "gravity" equations on the basis of bilateral trade flows among industrialized countries, and then use the estimated equations to "predict" bilateral trade flows between the Eastern and Western

84

Hari Vittas and Paolo Mauro

European countries in our sample.7 The gravity model assumes that bilateral trade between two countries can be explained by each country's size and income per capita and the distance between them. Theoretical foundations are provided in Helpman and Krugman (1985) and Bergstrand (1989). Baldwin (1994) offers an intuitive summary. Size and per capita income of the importing country determine its demand for goods, while size and per capita income of the exporting country determine its supply of goods. Distance is a measure of transportation costs. The model should work well for industrial goods, whose supply and cost depend upon the availability of general factors such as capital, labor, knowledge, and economies of scale, but not for raw materials dependent upon specific factors such as natural resources. Our departure from previous practice is that to estimate potential trade for the "sensitive" sectors we use individual-sector gravity equations, whereas previous studies have relied on equations for aggregate trade flows.8 There is no a priori reason to believe that the extent to which income or distance affects trade flows is the same for all commodity groups. One might expect, for example, that the elasticity of imports with respect to distance will be higher in sectors where transportation costs are high; that elasticities of import demand with respect to income will be higher for luxury goods than for necessities; and that the association with income per capita (as a proxy for the level of technology and wage levels) of the exporting country will be positive for high-technology goods, but could conceivably be negative for low-technology goods. The basic specification we use is the same as in Baldwin (1994):

**„ = Po + P,^ + Wj + P3< W

+ UY/PP + P ^ + M , + P ^

where all capital letters are in logarithms; Af.. is the flow of imports by country i from country/ Y. is gross domestic product in country /, P. is population in country /, D.. is the distance between countries i and,/; a., is a dummy that takes the value of one if countries i andy have a common border on land and zero otherwise, b.. is a dummy that takes the value of one if countries i and./ are members of the same trade agreement (EU or EFTA) and zero otherwise. As in Baldwin (1994), the adjacency dummy is used to allow for nonlinearities in the relationship between trade flows and geographical distance.9 We estimate the model for each sector, as well as for aggregate trade, for purposes of comparison. The results are reported in Table 3.6. 7 8

9

For details, see Appendix A. For example, Cadot and de Melo (1994) estimate the effects of trade liberalization on employment in the various sensitive sectors by using the Baldwin (1994) estimates of potential aggregate trade flows. We have also experimented with adding interaction terms of the common trade agreement dummy with all variables. The rationale for this is that an increase in distance or income might have a different impact, depending on whether two countries are members of the same trade agreement. The interaction terms are sometimes significant. Nevertheless, as the main results are broadly similar, we present those obtained using the conventional specification, in order to preserve comparability with previous literature.

Table 3.6. Estimated parameters of sectoral gravity equation (all variables in logs)

PoPi

PoPj

GDPj PoPj

Distij

ay

by

const.

R2

P<>Pi

GDP:

0 Food and live animals

0.790 (13.37)

0.677 (11.47)

1.906 (12.34)

1.018 (6.59)

-0.870 (8.70)

0.184 (0.66)

1.131 (6.58)

4.953 (5.23)

0.657

65 Textile yarn and fabrics

0.619 (11.97)

0.849 (16.42)

1.498 (11.08)

0.425 (3.14)

-0.916 (10.47)

0.352 (1.44)

0.613 (4.07)

7.282 (8.79)

0.662

84 Clothing

0.352 (4.85)

0.863 (11.89)

4.062 (21.40)

-0.695 (3.66)

-0.945 (7.69)

0.224 (0.65)

0.721 (3.41)

3.893 (3.34)

0.684

5 Chemicals

0.861 (19.24)

0.747 (16.69)

0.827 (7.06)

2.522 (21.55)

-0.954 (12.60)

0.263 (1.25)

0.315 (2.42)

4.898 (6.83)

0.789

67 Iron and steel

0.964 (13.41)

0.820 (11.41)

1.110 (5.91)

1.788 (9.51)

-1.221 (10.04)

0.300 (0.88)

0.178 (0.85)

6.014 (5.22)

0.598

Aggregate trade

0.722 (22.72)

0.750 (23.61)

1.404 (16.90)

1.513 (18.21)

-0.664 (12.34)

0.402 (2.69)

0.279 (3.02)

-0.050 (0.09)

0.838

T-statistics are reported in brackets. See Appendix A for data sources.

86

Hari Vittas and Paolo Mauro

We find that the "gravity" model performs quite well even using individualsector trade flows. The coefficients are typically of the predicted signs, are estimated fairly precisely, and are plausible (with the notable exception of the equation for clothing). The explanatory power of the equations is also satisfactory, with the coefficients of determination ranging from 0.54 to 0.71. 10 Gravity equations for the various individual sectors do indeed yield substantially different estimated coefficients. We illustrate this with a number of examples. The elasticities with respect to importing- and exporting-country income and population vary considerably, presumably depending in part on the extent to which the sector in question is technologically advanced. A 1 percent increase in income - for a given population - (a 1 percent fall in population for a given income) of the exporting country is associated with a 2.52 percent (1.78 percent) increase in imports of chemicals, but only with a 1.01 percent (0.34 percent) increase in imports of food.11 Increases in income or decreases in population of the importing country have a small effect on imports of chemicals and a large one on imports of food. Perhaps in the case of chemicals higher income implies higher demand, but also higher technology, so that the country will meet a substantial part of the additional demand with local production. The very large positive elasticity of imports of clothing and apparel with respect to importing country income and the negative elasticity with respect to exporting country income are surprising; if taken at face value, they would seem to imply that this sector consists mainly of luxury goods that nevertheless can be produced using relatively mature technology. This is implausible and we do not therefore include this sector in the calculation of potential trade flows. The "scale effects," by which we mean the effects of an equiproportionate rise in income and population (which maintains per capita income constant), differ less considerably across sectors. A 1 percent increase in both income and population of the importing country is associated with increases in imports by 0.79 percent for food, 0.86 percent for chemicals, 0.96 percent for iron and steel, 0.62 percent for textile yarn, and 0.35 percent for clothing and apparel. The coefficients on distance and on the dummy for adjacency are typically quite large and significant. More interestingly, the dummy for common trade agreements is always positive and highly significant and its magnitude is very large. For example, if i and j are members of a common trade agreement, imports of food by i fromy will on average be 113 percent higher than otherwise, other things being equal. The coefficient on the bilateral trade dummy is typically larger in the equation for individual sensitive sectors than in the equation for aggregate trade. This finding is consistent with the claim that the sectors we analyze were relatively highly protected in 1986-88. 10

11

As the dependent variable cannot take negative values, the appropriate technique would be a Tobit. We use ordinary least squares as in this case the results have typically been found to be very similar (Wang and Winters, 1991). The coefficients seem to vary even across subcomponents of the chemicals sector, as reported in Table 3.6. This suggests that further disaggregation of the sectors might be desirable.

Potential trade with core and periphery

87

Table 3.7. Actual and potential imports from CEEC-6 by sector in 1993 (million US $ at current prices) Germany

France

Italy

UK

Portugal

Food and Agriculture A. Actual B. Potential C. Ratio of A to B (percent)

1,015 3,342 30

206 1,624 13

403 1,798 22

57 1,276 4

3 53 6

Textile Yarn A. Actual B. Potential C. Ratio of A to B (percent)

388 1,504 26

50 605 8

126 690 18

75 490 15

2 35 5

Chemicals A. Actual B. Potential C. Ratio of A to B (percent)

649 1,822 36

158 599 26

330 744 44

145 503 29

4 36 11

Iron and Steel A. Actual B. Potential C. Ratio of A to B (percent)

704 1,008 70

45 250 18

149 320 47

63 199 32

3 8 31

All "sensitive" imports A. Actual B. Potential C. Ratio of A to B (percent)

2,756 7,675 36

460 3,079 15

1,009 3,552 28

340 2,468 14

11 133 9

105 60

53 17

26 27

28 18

14 8

Total Imports: Ratio of actual to potential Baldwin's estimate Own estimate

Sources: OECD, Foreign Trade by Commodities, and authors' calculations.

The potential import flows of the Western European countries in our samples from each of the CEEC-6 for each sensitive sector are calculated by using the coefficients reported in Table 3.6. We follow Baldwin (1994) in defining medium-term potential trade flows as the predicted values obtained by using 1989 data on GDP (before the output collapse in the CEECs) and population, and assuming that these countries become as fully integrated with Western European countries as the latter are at present with each other. Table 3.7 reports the results. For the sensitive sectors, actual trade is well below potential trade.

88

Hari Vittas and Paolo Mauro

Table 3.8. Potential imports from the CEEC-6 under different scenarios (million US $ at current prices) Germany0

France

Italy

UK

Portugal

1986-88

2,885

1,057

1,233

857

45

1989

4,064

1,984

1,559

964

70

1993

3,943

1,393

1,644

1,069

73

1986-88

5,449

2,180

2,491

1,755

86

1989

7,675

3,079

3,552

2,468

133

1993

7,946

3,029

3,530

2,296

151

Incomplete Trade Integration

Complete Trade Integration

a

West Germany for 1986-88, sum of East and West Germany for 1989, Germany for 1993. Sources: OECD, Foreign Trade by Commodities, and author's calculations.

Furthermore, the ratio of actual to potential trade is lower for the sensitive sectors than for aggregate trade. This finding suggests that the degree of protection may have been particularly high in the sensitive sectors. The ratios of actual to potential trade show considerable variation across sectors. The lowest ratio in all three countries is that for textile yarn, and the highest is for iron and steel. Table 3.8 reports the estimates for potential trade in the sensitive sectors obtained by using 1986-88,1989, and 1993 income and population data, with and without full trade integration.12 This provides an indication of the extent to which changes in the assumptions about income levels and the degree of trade integration affect potential imports in the sensitive sectors. Using GDP and population data for 1993 (rather than 1989) leads the estimates of potential imports to fall by around 5 or 10 percent. This may provide an indication of the extent to which the slow growth in imports from the CEEC-6 may be attributed to the fall in output that coincided with the initial phase of the transition of these economies.13 The estimated resulting fall in potential trade seems too small to provide a full explanation for slow import growth, though the transition process might have temporarily disrupted the CEECs' ability to compete effectively on international markets. Alternative assumptions on the extent of trade integration have an even more crucial impact on estimates of potential trade in the sen12

13

The estimated effects of trade integration or of changes in output vary among the countries considered because of the different sectoral composition of their trade. Table 3.8 shows that the estimated effects of the output collapse on potential trade are even smaller in the presence of a bilateral trade agreement. This is due to the interaction of different elasticities with respect to income in the various sectors, with differences in the estimated 1989 composition of potential trade obtained with and without trade agreement.

Potential trade with core and periphery

89

sitive sectors: calculations that assume that the bilateral trade agreement dummy takes the value zero reduce the estimates for potential trade by roughly one half, thereby highlighting the importance of trade integration for these sectors. In summing up the analysis in this section, we would note that various approaches used tended to reinforce the view that the trade restrictions have so far hindered the growth of imports of sensitive products from the CEEC-6. The persistence of barriers to trade may increase the incentive for factor flows to take place to exploit comparative advantage opportunities. Appendix B reviews some limited evidence on the extent to which this may have already happened. 3.4

Summary and conclusions

This paper has reviewed developments in trade relations between the four largest economies of Western Europe and Portugal with a group of CEECs that are at a relatively advanced stage in their efforts to reform their economies on the basis of market principles. For all the countries considered, trade integration has proceeded at a rapid pace well in excess of the growth of world trade. Nevertheless, there have been significant variations across the Western European countries in the rate at which their trade with the CEEC-6 has increased. These variations have reflected to some extent cross-country differences in the potential for trade with the countries in transition. But, more importantly, there have been also significant cross-country differences in the degree to which such potential has so far been exploited. The evidence suggests strongly that to date the speed of adjustment has been much more closely correlated with geographic proximity rather than economic proximity. The paper has also documented that the pace of East-West European trade integration has varied significantly across economic sectors, being particularly slow in a few sectors such as agriculture and iron and steel where the CEEC-6 are believed to possess a comparative advantage. This appears to be related, at least in part, to a comparatively slow rate of liberalization of the trade restrictions that Western European countries apply on imports competing with domestic production in these sectors. In other sectors, usually considered to be sensitive, such as the chemicals and the textiles and clothing sectors, the experience of individual countries has varied, but overall there is not much evidence that imports from the CEEC-6 have been systematically discriminated against. Appendix A Data sources and definitions The trade data used in this paper are drawn from two different sources. Data on aggregate bilateral trade flows are obtained from the IMF publication Direction of Trade Statistics. Data on bilateral flows by sector are drawn from the World Bank's Trade Analysis and Research System (TARS) data base and are

90

Hari Vittas and Paolo Mauro

the same as those reported in Foreign Trade by Commodities, published by the OECD. The commodity groups analyzed in Section 3.3 are chosen so as to correspond closely to the sectors that previous literature has identified as "sensitive," namely food and agriculture, textiles and apparel, chemicals, and iron and steel.14 These are relatively protected sectors. Specifically, data on the following SITC commodity groups have been used: 15 0. 5. 65. 67. 84.

"Food and live animals" "Chemicals" "Textile yarn, fabrics, made-up articles, n.e.s. and related products" "Iron and steel" "Articles of apparel and clothing accessories"

The "gravity" equations for total trade and for the individual "sensitive" sectors are estimated on the basis of bilateral trade flows for a sample of twenty industrialized countries: Austria, Canada, Denmark, Finland, France, the Federal Republic of Germany, Greece, Ireland, Italy, Japan, Korea, the Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, Turkey, the United Kingdom, and the United States. Thus, each equation is estimated using 380 bilateral trade flow observations. This sample appears to be sufficiently diverse with respect to size, level of development, and distance to allow reasonably precise estimates of the parameters to be obtained. It is also fairly representative of the type of bilateral trade flows that we are ultimately interested in predicting. It is similar, though not identical, to the sample of countries chosen by Baldwin (1994). The cross-sectional gravity equations were estimated using 1986-88 averages of bilateral trade flows, population, and GDP data in order to reduce "noise" in the dependent variable and thus obtain more precise parameter estimates.16 The data on gross domestic product are drawn from the International Monetary Fund World Economic Outlook files. All trade and GDP (at purchasingpower parity) data are nominal magnitudes, in million dollars. Where bilateral trade flows in a given sector are zero, we substitute the (relatively small) value of one million dollars in the estimations.17 The distances between countries typically refer to distances between capital cities (however, for Germany distances 14

15

16

17

These are the sectors that have been identified as "sensitive" both by Cadot and de Melo (1994) and by Andersen and Dittus (1994). However, the definition of "sensitive" sectors in the Europe Agreements that the European Union has concluded with several CEECs appears to be narrower. The definitions of the commodity groups follow the Standard International Trade Classification (SITC) by the United Nations (Statistical Papers, Series M, No. 34, Rev. 3, 1986). Three-year averages are also used by Wang and Winters (1991). Baldwin (1994) obtains much lower standard errors of the estimated coefficients than we do, as he uses random effects panel data techniques. This is preferable to treating the observation as missing (which would lead to a downward bias in the absolute value of the coefficients).

Potential trade with core and periphery

91

Table 3.9. Direct investment to CEEC-6 Germany (DM millions) 1989

77

1990

218

1991

1,289

1992

1,573

1993

1,793

Sources: Deutsche Bundesbank, Monthly Report, various issues and information provided by the authorities. Data refer to total Central and Eastern Europe.

from Frankfurt rather than Bonn are used) and are drawn from a private software source for air travel (Official Airlines Guide). Appendix B Integration of factor markets Factor movements are one of the many channels through which the integration of the CEECs into the global trading and economic system can occur. Factor movements typically facilitate and reinforce the effects of trade integration, although they may also act as a substitute for trade when the latter is inhibited by protectionist policies or other barriers. In the following paragraphs, we discuss briefly developments in direct investment flows from Germany to the CEEC-6, before considering some preliminary evidence on the extent to which the labor markets in Germany, France, and Portugal have so far been influenced by the opening up of the CEECs. Direct investment flows German direct investment to the CEEC-6 was very low prior to the initiation of political and economic reforms in these countries, but has since increased by leaps and bounds (Table 3.9).18 Although its share in total German direct in18

Hungary and the Czech Republic have been by far the main recipients of these investment flows, accounting in 1993 for well over half of the total. By contrast, German direct investment to Romania and Bulgaria has remained very low while Poland has become a sizable recipient of such investment only in 1993 (and Slovakia has made a promising start).

92

Hari Vittas and Paolo Mauro

vestment abroad remains modest (about 9 i percent in 1993), it is already substantially higher than the corresponding share of the CEEC-6 in total German exports and imports. This is consistent with the view that integration of capital markets is not only faster than that of goods markets but may also be a precondition for full trade integration.19 Labor markets In the absence of rigidities, enhanced East-West trade would be expected to lead in Western Europe to a fall in the relative price of factors (such as unskilled and semi-skilled labor) that are used intensively in the production of exportable goods and services in the East. However, if (as in Western Europe) relative factor prices are sticky, the effects of trade expansion would be felt mainly on the level of employment. In particular, one might expect a relative decline of employment in those sectors in which Eastern Europe enjoys a comparative advantage and in which therefore import penetration would be comparatively high.20 Figures 3.4 through 3.6 show, however, that this expectation has not been borne out by the available evidence on employment trends in Germany, France, and Portugal. In Germany there was a small contraction of employment in the "sensitive" sectors (relative to total employment in manufacturing) in the few years preceding the opening up of Central and Eastern Europe. However, far from becoming any worse, this relative decline has since been reversed.21 Among the individual sensitive sectors identified in Figure 3.4, there is only a

19

20

21

Information on the investment flows between Portugal and the CEEC-6 is not readily available, but the presumption must be that such flows are probably insignificant. In any case, the interesting issue in Portugal's case is not the size of its direct capital flows to or from the transition economies, but rather the extent to which the opening up of these economies may have diverted foreign investment away from it. Net direct investment inflows into Portugal have declined sharply during 1992 and 1993, after peaking at a very high level in the three years to 1991. However, the decline has coincided with a pronounced deterioration in business cycle conditions throughout Europe, which weakened temporarily both the incentive to invest and the availability of internationally mobile capital. Without further investigation, it is not possible to tell if this decline has also been influenced by the resource reallocation effects of growing Eastern-Western European integration. A recent study by the Bundesbank supports the widespread belief that relative wage rates in Germany (as is the case in most other countries in Europe) remain insufficiently flexible. The study found not only that the dispersion in the rate of increase in pay rates across sectors has remained small in western Germany in recent years, but also that some of the largest increases have taken place in the structurally weak textiles industry. For details, see Deutsche Bundesbank (1994). It is worth noting that the conclusion that relative employment trends across manufacturing sectors have not been affected in any visible way by increased trade with Eastern Europe remains valid even if the analysis is confined only to the few sensitive sectors (food and basic metals) in which, according to the RCAs, the CEEC-6 appear to enjoy a comparative advantage.

Potential trade with core and periphery

93

140.

Sensitive Sectors 130-

120-

110-

100-

..Textiles 90-

Basic metals 80-

85

101.0

86

90

87

91

92

93

92

93

n

Index for Sensitive Sectors / Index for Manufacturing 100.5-

100.0-

99.5-

99.0-

98.5-

97.5

85

86

87

89

90

Figure 3.4: Trends in employment in the manufacturing sector, Germany (1985 = 100). Source: OECD, Indicators of Industrial Activity.

mild suggestion that the secular decline of the textiles sector may have accelerated in the recent past and some even weaker evidence that the same may hold true for basic metals (which includes the iron and steel industry). By contrast, recent employment trends in the chemicals and food sectors have been consid-

94

Hari Vittas and Paolo Mauro Sensitive Sectors

130 -

/ Food

/

120-

/

Chemicals

110-

'

100-

•— 6 ^

Total manufacturing

90-

^"^^^^flasic metals 80-

Textiles

70-

1

85

87

1

1

90

88

• ...

1

91

1

92

93

92

93

Index for Sensitive Sectors / Index for Manufacturing 101-

100-

99-

97-

96-

95-

85

87

88

89

90

Figure 3.5: Trends in employment in the manufacturing sector, France (1985 = 100). Source: OECD, Indicators of Industrial Activity.

erably at variance with what one might expect if import penetration had become a "problem." The experience of France bears similarities with that of Germany. In particular, the share of employment accounted for by the sensitive sectors had been on a steep declining trend in the period preceding market liberalization in the

Potential trade with core and periphery

95

Sensitive Sectors •••••..

100>

v

Textiles

' • • • •

^

^

• • • • • • • • • • •

^

'"""••••••••

"""^^ ^V^o
..

90 A

Basic metals Xv

80-

Chemicals ***«.,

N. ^ w

.

70-

60-

50

H

85

86

87

89

90

91

92

93

91

92

93

103.0 n

Index for Sensitive Sectors / Index for Manufacturing 102.0-

101.0-

100.0-

99.0-

98.0-

97.0 85

86

87

90

Figure 3.6: Trends in employment in the manufacturing sector, Portugal (1985 = 100). Source: OECD, Indicators of Industrial Activity.

transition countries. This trend has since been interrupted and indeed slightly reversed (Figure 3.5). Among the individual sensitive sectors, textiles is the only sector where the decline in the level of employment may have become more pronounced in the postreform period, and even for this sector the evidence is quite weak. In Portugal employment in all the sensitive sectors has been de-

96

Hari Vittas and Paolo Mauro

dining at least since the mid-1980s, but once again there is no clear worsening of this trend after the opening up of Eastern Europe. REFERENCES Andersen, Palle, and Peter Dittus (1994). "Trade and Employment: Can We Afford Better Market Access for Eastern Europe?", mimeo, Bank for International Settlements, Basle. Balassa, Bela (1965). "Trade Liberalization and 'Revealed' Comparative Advantage," The Manchester School 33,94-123. Baldwin, Richard (1994). Towards an Integrated Europe, London: Center for Economic Policy Research (CEPR). 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. Cadot, Olivier, and Jaime de Melo (1994). "France and the CEECs: Adjusting to Another Enlargement," CEPR Discussion Paper 1049. Deutsche Bundesbank (1994). "The Trend in Agreed Pay Rates and Actual Earnings Since the Mid-1980s," Monthly Report, August. Helpman, E., and Paul Krugman (1985). Market Structure and Foreign Trade, Cambridge: MIT Press. Wang, Z., and L. A. Winters (1991). "The Trading Potential of Eastern Europe," CEPR discussion paper 610.

Comment Susan M. Collins

This paper addresses topics of considerable concern to both policymakers and researchers. For example, how will trade flows from transitioning economies in Eastern Europe affect economies in the West? How important are restrictions on trade in "sensitive" sectors? The paper also presents a variety of interesting facts and figures about recent trade developments. I see two main conclusions of the paper. First, it documents that there has been a significant increase in trade between the CEEC-6 and its Western European trading partners, especially Germany. Second, the rise in Western European imports of "sensitive" goods has been quite a bit smaller than the overall rise in imports and the paper suggests that this is largely attributable to trade restrictions. The paper's two main findings are broadly sensible. However, I have some questions about aspects of the paper. I would also like to offer some alternative interpretations of some of the authors' findings. In my comments, I will follow the general outline in the paper. The first section of the paper provides a useful and informative overview of recent developments in trade flows between the CEEC-6 and five Western European countries. In particular, it documents that trade with the CEEC-6 has increased as a share of the West's total trade, and that Germany, France, Italy, and the UK (but not Portugal) have seen an improvement in their trade balances visa-vis the CEEC-6. As the authors note, these developments are worth stressing in light of the widespread beliefs in some circles that the transition economies would have little to export and in other circles that the opening of Eastern Europe would cause major dislocations in the West. However, I am somewhat less surprised than the authors appear to be that Germany's trade with the region has increased so much more quickly than has been true of other Western European nations' trade. Germany's unique history and location as well as the special links between the CEEC-6 and eastern Germany have surely played a central role. The next section of the paper looks at sectoral trade patterns. It shows that sensitive products are a large share of exports from the transition region, but that they have grown more slowly than total exports overall, and have fallen as 97

98

Susan M. Collins

shares of total imports, especially for Germany. Further, the decline is particularly evident for agricultural products and for iron and steel - sectors highly protected in the European Union. However, I am uncomfortable about drawing conclusions about the importance of trade restrictions in sensitive sectors from simple statistics such as the percent change in a sectors' trade flow between 1989 and 1993. To bolster their statements, it would have been helpful for the authors to provide additional information about what trade restrictions were actually in force during the relevant time period, and about any changes. The paper goes on to discuss comparative advantage among the transition economies. It uses a now familiar technique of looking at actual changes in net relative to gross trade flows so as to form an index of "revealed" comparative advantage. While this approach can be very useful in some contexts, I am not convinced that it is an informative approach here. The trade flows during 1992-93 are determined by a number of factors that arguably have little to do with the underlying comparative advantage of the CEEC-6. They still reflect the old "command" structure, the upheaval of transition (including political developments, as well as the breakup of the CMEA and move toward markets). Further, some special developments during the period, such as a drought, may account for the decline in agricultural exports from the region. In my view, this methodology - and thus Figure 3.3 - can tell us little about the sectors in which these countries are likely to enjoy a comparative advantage. The next section of the paper estimates gravity equations by sector to assess the potential trade flows between CEEC-6 and the Western European countries. Overall, the basic findings of this work are quite sensible. They are that actual trade flows are well below potential trade, even with Germany, and that this suggests that trade restrictions may be playing some role. At the same time, I had not seen gravity equations used to examine sectoral trade flows before, and found it interesting to think about how to interpret the estimates of potential trade flows that emerge. First, should we expect the types of gravity equations estimated here to provide good estimates of potential bilateral trade flows by sector more generally? We know that on aggregate data, the relatively ad hoc gravity model performs surprisingly well - better than most structural models of tradeflows.There have been some attempts to reconcile it with a theory of trade flows, based on models of intra-industry trade. It seems to me that there may be good reasons to be more suspicious of the sectoral version than of the aggregate version. Surely endowments must matter much more in explaining the composition of bilateral trade flows than they seem to in explaining the aggregate volume of trade flows. The sectoral gravity model says that per capita income (PCI), population, and distance should explain not just Germany's trade with Japan versus Germany's trade with Australia, but how much textiles and yarn Germany trades with Japan versus with Australia. This suggests a different interpretation of the results in Table 3.6. PCI may be acting in part as a proxy for the capital-

Comment

99

to-labor (K/L) endowment of the country. This could explain why PCI of the exporter enters the clothing equation with a negative sign. As K/L (PCI) rises, a country's exports of clothing tend to decline. More generally, the pattern of coefficients on the exporter's PCI suggests that the poorest countries tend to export clothing, moderate-income countries tend to export textiles (which use relatively more capital to produce), and the wealthiest countries tend to export chemicals, fertilizers, and steel (the most capital intensive). Again, this is not to say that the sectoral results are not interesting. Indeed, I am surprised that the equations seem to fit as well as they do. It would be quite a bit of work to put together additional data on endowments to see if the fit could be improved further, and similar data are unlikely to be available for the CEEC6. (However, I would like to see this exercise done for industrial countries, to assess how well gravity equations compare with structural models more firmly based in trade theory to explain sectoral trade flows.) But the paper would benefit from a broader discussion of how these results might be interpreted, and from a brief review of gravity models on aggregate data. A second issue is how to interpret these particular figures for the CEEC-6.1 am concerned that the estimates of potential trade might be too high because the 1989 GDP figures are known to be overestimates. In light of this, I was surprised that reestimating based on 1993 GDP made relatively little difference. Further, the authors may be asking for "out-of-sample" estimates and the fit may be quite poor out-of-sample. (1) The 1993 PCI figures for many of the CEEC-6 are well below PCI among most industrial countries. This suggests including at least some middle-income countries in the sample. (2) The estimation is done on a cross-country sample. It is not obvious to me that one would get the same relationship between trade flows and PCI over time for a sample of countries that experienced significant declines in PCI as one would get from a cross section. This problem is more difficult to remedy, but the issues seem to me to warrant some discussion. It is also worth noting that using the parameters from the estimation of sectoral trade flows between pairs of industrial countries to estimate potential trade between CEEC-6 and Western Europe presumes that the latter should have the same underlying structural determinants as the former. I would have thought that the gravity equations estimated would form reasonable estimates of the potential trade flows once the transition had taken place. But they may well be overestimates (or perhaps underestimates) of trade in the early 1990s, during transition. Thus, time frame becomes very important in this context. The implication of these thoughts is that I am worried about interpreting the difference between estimates of "potential" trade and actual trade as evidence that trade restrictions have "hindered the growth of imports of sensitive products" - even though I believe that trade restrictions may well have played such a role.

CHAPTER 4

Impact on German trade of increased division of labor with Eastern Europe Dieter Schumacher

4.1

Introduction

Following the recent political developments in Central and Eastern Europe, a transformation of the former centrally planned economic system into a marketbased economic order with decentralized decision making is underway. This implies radical social and economic changes in these countries affecting the economic interdependence between Central and Eastern European countries (CEECs) themselves, as well as their economic relations with Western industrial countries and developing countries. After the collapse of the Council for Mutual Economic Assistance (CMEA), trade relations between the former centrally planned economies declined dramatically, whereas their trade with Western industrial countries quickly expanded. One can assume that economic relations between Eastern and Western Europe will intensify further in the medium and long term. Regional proximity and the positive effects of an international division of labor on productivity, but also the political urge to decrease the economic and social gap between Eastern and Western Europe, point to this trend. Furthermore, for political reasons trade relations between East and West until the collapse of the old system remained considerably low in comparison to a "natural" level of economic interaction. The changes in CEECs will have substantial effects on other countries in general and, in particular, on Western Europe. Within Western Europe this is especially true for Germany and Austria. The pace at which these processes will occur, however, depends to a considerable extent upon how fast the transformation in the various Eastern European countries will result in economic growth. In the following analysis, we will investigate how the scope and sectoral structure of German trade with CEECs have developed since the beginning of The research carried out for this paper is part of a larger project on the impact on foreign trade of the transformation process in Central and Eastern Europe, financed by the Volkswagen Foundation. The computations were carefully carried out by Gerlinde Hopp-Hoffmann and Karin Hollmann. The original German draft was translated by Timothy Clements. 100

Impact on German trade of division of labor

101

the transformation process, which started at the end of the 1980s and beginning of the 1990s. The expected pattern of future bilateral trade flows will be estimated on the basis of a gravity model on the assumption that future trade between Germany and the CEECs will follow the same rules as trade among market economies. Conclusions will then be drawn regarding the effects of increased division of labor with Central and Eastern Europe on economic structures in Germany. 4.2

Germany's trade with Central and Eastern Europe

Germany is by far the biggest Western trading partner of the CEECs. Of the former Soviet Union's trade with OECD countries in 1993, approximately 30 percent fell to (unified) Germany; of the other CEECs' OECD trade - namely, that of Poland, CSFR,1 Hungary, Romania, and Bulgaria - over 40 percent (Table 4.1). This is considerably more than Germany's proportion of total OECD trade (approximately 16 percent). In trade with CEECs the significance of Germany is greater on the export side than on the import side, which is in line with Germany's general position in foreign trade. The difference has diminished, however, as German imports have risen sharply in the course of the unification process - both directly into East Germany and into West Germany due to increased production. In recent years, Germany's foreign trade with CEECs has been characterized by two divergent developments. Whereas West Germany's trade has grown considerably, East Germany's has plummeted (Table 4.2). In this way, West German imports and exports in trade with Central and Eastern Europe, excluding the former Soviet Union, have nearly trebled within five years. Trade in goods with the CSFR has seen the sharpest increase, followed by Poland and Hungary. A fundamental reason for this rapid intensification of trade relations with the three "Visegrad" countries is due to the fact that the reform process and trade liberalization have progressed the most in these three countries. In addition, geographical proximity and the fact that trade relations with these three countries were already very close at the time of the CMEA, certainly played a role. Furthermore, all three countries tried to support exports by a considerable depreciation of their currencies in real terms. The intensification of trade with Bulgaria and Romania has made slower progress, and West German trade relations with the successor States of the Soviet Union combined have scarcely grown. Initial hopes that East Germany would be able to maintain its high level of trade with the CMEA countries - its biggest trading partner - and, thereby, ease its necessary economic reform, were not realized. Although East German exports to CEECs at first increased following the introduction of the Economic For statistical reasons the Czech Republic and the Slovak Republic are treated as one country. Besides the ex-USSR, the five CEECs considered in the analysis, therefore, represent, in fact, six states.

103

102

Table 4.1. Trade of Germany with CEECs, 1989-94 Poland

CSFR

2.4 2.9 5.1 5.3 5.9 6.4

1.5 1.9 3.0 5.3 5.5 7.2

1.9 2.1 2.6 3.0 3.1 4.0

0.3 0.7 0.7 0.9 1.1 1.2

0.7 1.5

0.4 1.7

0.6 0.9

39.2

49.6

38.7

1.9 3.2 4.4 5.3 5.2 6.3

1.3 1.7 3.1 4.7 4.8 6.6

1.4 2.0 2.6 3.0 2.7 3.4

0.8 0.7 0.7 0.8 0.8 1.1

0.2 0.2 0.3 0.4 0.3 0.5

1989

0.7

1994

1.7

0.5 1.8

0.5 0.9

0.3 0.3

47.7

50.7

40.6

33.4

Hungary

Romania

Bulgaria

Five countries total

USSR

of which: Russia

for comparison: World

Exports (in US-$ billion) 1989 1990 1991 1992 1993 1994

6.9 8.1

6.1 6.4

341.2

11.9

10.7

402.6

0.6 0.5 0.7

15.0 16.1

8.9 9.6

19.5

10.1

6.9 6.6

0.1 0.3

0.2 0.2

2.0 4.6

1.8 2.4

1.6

27.8

29.6

40.7

33.9

0.8 0.5 0.5

398.4

430.4 380.1 424.0

Share in total German exports 1989 1994

100.0 100.0

Share of Germany in OECD exports (in %) 1992

16.6

Imports (in US-$ billion) 1989 1990 1991 1992 1993 1994

17.9

4.6 5.7 8.6 8.3 7.9 9.9

6.5 8.2

0.1 0.1

2.1 4.7

1.7 2.6

2.2

24.0

44.8

26.8

5.7 7.9 11.1 14.1 13.9

269.7 342.6 389.0 408.5 342.7 378.1

Share in total German imports 100.0 100.0

Share of Germany in OECD imports (in %) 1992

* 1989 and 1990 West Germany only, 1994 preliminary figures. Source: Own calculations based on data from the Federal Statistical Office and OECD.

15.1

Table 4.2. Trade of West Germany and East Germany with CEECs, 1990-94° (DM billion) Poland

CSFR

Hungary

Romania

Bulgaria

Five countries total

USSR

of which: Russia

for comparison: World

West Germany Exports

Imports

1990

4.7

3.1

3.4

1.1

0.8

13.0

10.4

1991

7.5

4.3

3.8

1.0

0.7

17.3

8.6

1992

7.7

7.5

4.5

1.2

0.8

21.7

8.4

1993

9.3

8.5

5.0

1.7

0.8

25.3

11.0

7.7

616.4

1994

9.9

11.1

6.2

1.9

1.0

30.1

11.6

8.4

673.1

1990

5.2

2.7

3.3

1.1

0.4

12.6

9.1

1991

6.5

4.4

4.0

1.1

0.5

16.5

9.9

633.1

1992

7.4

6.6

4.5

1.2

0.6

20.3

10.1

628.0

642.8 648.4 657.4

550.6

1993

7.9

7.2

4.4

1.3

0.5

21.4

10.6

8.4

557.8

1994

9.2

9.9

5.3

1.8

0.7

26.8

13.0

11.2

601.0

1990

2.9

3.4

2.7

1.5

1.4

12.0

17.8

1991

1.0

0.6

0.4

0.2

0.1

2.4

9.0

1992

0.5

0.8

0.2

0.1

0.1

1.6

5.5

1993

0.4

0.5

0.2

0.2

0.1

1.3

4.9

3.7

1994

0.5

0.6

0.2

0.1

0.1

1.4

3.4

2.4

1990

1.8

1.7

1.2

0.4

0.5

5.7

9.1

East Germany Exports

Imports

38.1 17.5 13.8 11.9 12.1 22.9

1991

0.8

0.7

0.3

0.1

0.0

1.9

4.3

1992

0.8

0.7

0.2

0.0

0.0

1.7

2.9

1993

0.7

0.7

0.1

0.0

0.0

1.6

2.4

2.3

8.4

1994

1.0

0.8

0.1

0.0

0.0

2.0

2.0

1.9

10.2

" 1993 and 1994 preliminary figures. Source: Federal Statistical Office.

10.9 9.6

Impact on German trade of division of labor

105

and Monetary Union with West Germany in the middle of 1990 - with the help of a guarantee by the Federal Government that the GDR's already existing trade commitments would be fulfilled and extensive subsidies continued - they collapsed after the disbandment of the CMEA at the beginning of 1991 just as domestic sales of East German firms had done so before. In this way, East German exports to CEECs, excluding the former Soviet Union, shrank to one tenth of their value in 1990. Exports to the successor states of the Soviet Union declined only to a level of 30 percent, but the absolute size of this collapse was of greater consequence to the East German economy than the loss of markets in other CEECs. Above all, production of agricultural and metal-working machinery as well as ships and railroad equipment was severely struck. The fall in East German exports to the CEECs is partly due to the difficulties experienced by the former state-trading economies during their economic transformation and to the effect of disintegration following the collapse of the CMEA, and later, of the Soviet Union. However, supply-side factors played an equally important role, above all, the unsatisfactory price and nonprice competitiveness of East German products under world market conditions as real wages rose; the lack of quality and poor standard of technology could not even be compensated for by extensive subsidies. Moreover, the factors that had earlier been seen as East Germany's "natural advantages," such as familiarity with the language, culture, and mentality, as well as the foreign trade system of the former socialist countries, quickly lost their importance (Losch and Wohlers 1994: 135). The regional differences in the size of export reductions are mainly a consequence of the preferential treatment of supplies to the former Soviet Union by means of the German export-guarantee system and other subsidies. Differences in the course of transformation processes have probably played a role too. In those economies that had already made much progress, East German products - mainly investment goods - were, with respect to quality and technological standard, less competitive than they had been in the Soviet Union, where the market-economic reforms and, thereby, also the modernization of the capital stock were advancing only slowly. Despite existing obligations to purchase goods, East German imports from the CMEA area collapsed immediately after the introduction of the Economic and Monetary Union with West Germany. Within a short period of time, imports from the Soviet Union and equally from other CEECs fell to below 30 percent of their value in 1990. The main reasons were the decrease of industrial production - caused by the transformation - which reduced the need for intermediate products drastically, the diminishing competitiveness of East European products under world market conditions as well as better access to West German goods (Losch and Wohlers 1994: 135). With the recovery of production in East Germany, total imports in 1994 increased moderately once again. However, the increasing volume of imports continued to come mainly from Western countries; the imports from CEECs remained repressed.

106

Dieter Schumacher

Whereas in trade with the former USSR the loss of East Germany's trade relations could not be compensated for by strengthened West German trade contacts, East Germany's loss of markets in the other CEECs was more than made up for by increased supplies from West German firms. Conversely, suppliers from CEECs (excluding the former Soviet Union) were able to increase their sales in West Germany considerably, more than equaling their losses in East Germany. Post-unification Germany maintains more extensive trade relations, in particular with the three Visegrad countries, than did the two parts of Germany taken together in 1990. It is remarkable that this trade increased further in 1993, although Germany's total imports and exports fell as a result of the recession in Germany. With the improved economic situation in 1994, German imports, as well as exports, increased again. Trade with Eastern Europe was particularly dynamic, with imports performing better than exports. Also, imports from Russia increased once again, whereas exports to Russia declined. The main impediments to any extension of exports to Russia are uncertain finance, high taxes and duties, a lack of legal security, and high transportation costs. Summarizing, one may note that the dynamics of German trade with CEECs has been (1) concentrated in Poland, the CSFR and Hungary and (2) restricted to West Germany. By 1994, West Germany's trade with CEECs expressed as a proportion of West Germany's total foreign trade had reached 6.5 percent in comparison to some 4 percent in 1990. East Germany in 1994 conducted 40 percent (1990: 75 percent) of all its foreign trade with CEECs, total trade still being at a low level. East German firms are still unable to supply internationally a wide and competitive range of goods. So far, they have experienced their major market economic successes at a regional level. Trade between Germany and CEECs is characterized by a large amount of intersectoral division of labor, that is, sectors that take advantage of exports are different from those that come under additional competitive pressure as a result of imports (Table 4.3). Above all, Germany supplies investment goods, in particular products from the mechanical engineering, electrical machinery, and transport equipment industries. Conversely, the former USSR supplies Germany with mainly mineral raw materials (with a proportion of some 50 percent). Among the industrial goods, raw-material-intensive product groups, such as mineral oil products, nonferrous metals, and wood, are the most prominent. The other CEECs mainly supply industrial goods, in particular consumer goods. With the exception of some basic industries (basic industrial chemicals, iron, and steel), all (West) German export industries have been able to expand their supplies to CEECs since 1990. In particular, exports of electrical machinery and motor vehicles have risen sharply. The same is true of textiles, clothing, and leather, owing to the extension of outward processing activities. Similarly, with the exception of mineral oil products, (West) German imports from these countries have risen in all product groups. Raw-material and capital-intensive production goods have shrunk in significance, whereas labor-intensive products

Impact on German trade of division of labor

107

have gained in importance. This is also true for a range of technology-intensive goods from the investment goods industry, above all in the electrical machinery and motor vehicle industries. More electrical goods were supplied mainly from Hungary and the CSFR and more products of the motor vehicle industry from the CSFR and Hungary, and then later from Poland. These trends may indicate where the competitive advantages of CEECs really lie under market conditions. By now, Romania offers the least diversified range of imports, concentrating on clothing and furniture, alongside Bulgaria, which provides mainly foodstuffs and textiles (Table 4.4). Clothing is also the most important item of supplies from Poland and Hungary, followed by furniture, foodstuffs, and metal products in Poland and by electrical goods and foodstuffs in Hungary. The CSFR supplies the most diversified product range - including, above all, metal products, mechanical engineering goods, clothing, iron and steel, electrical goods, and products of the motor vehicle industry. On the whole, the significance of the former Soviet Union and other CEECs as trade partners is still relatively small. In some sectors, however, they play a significantly more important role. In 1993, more than 10 percent of total German exports in railroad equipment and the ship and footwear industries went to the former USSR. For exports to the other CEECs, this is true for the textile and leather industries. On the import side, the market share of the former USSR is more than 10 percent only in nonferrous metals, while the other CEECs have, in the meantime, reached some 10 to 20 percent in a number of product groups (clothing, wood products, furniture, fertilizers, glass and other nonmetallic mineral products, metal products, and railroad equipment). The expansion of their import market shares in Germany did not affect the developing countries negatively, but was at the expense of the OECD countries and of former Yugoslavia. There is a tendency toward German export surpluses in merchandise trade with the former USSR, and equally in trade with other CEECs. These surpluses occur mainly in the mechanical engineering, motor vehicles, textiles, chemicals, and precision engineering industries. In contrast, the clothing sector suffers the largest deficits, followed a long way behind by furniture, wooden articles, and a range of other consumer and production goods. 4.3

Determinants of bilateral trade flows

In the assessment and projection of trade flows between Germany and CEECs, one has to imagine a "normal" pattern of trade flows. Here we can apply the gravity model, which represents bilateral trade volumes as a function of (1) the national product of the supplier country and of the destination country, showing the strength of supply and demand, and (2) factors hindering and stimulating trade. Of the factors restricting trade, transport costs and protection by tariffs or nontariff measures are most important. Factors that encourage trade

Table 4.3. Commodity structure of German trade with CEECs, 1993 (percentage) ISIC o

Poland, Hungary, CSFR, Romania, Bulgaria

Ex-USSR

Exports

Imports

Exports

1

Agriculture

2.0

3.1

2.2

3.0

2

Mining and Quarrying

0.4

3.0

0.0

40.1a

3

Manufacturing Industries

97.7

93.9

97.7

56.9

Food, Beverages, Tobacco

5.3

6.2

14.7

3.1

11.8

4.7

2.5

1.1

1.8

16.7

1.5

3.9 0.3

321

Imports

322

Textiles

323

Wearing Apparel

32

Leather and Leather Products

1.1

0.8

0.2

324

Footwear

0.6

2.6

1.5

0.6

331

Wood and Wood Products

0.5

3.6

0.2

2.7

332

Furniture and Fixture

0.8

6.2

0.8

0.9

341

Paper and Paper Products

2.3

1.2

1.0

0.7

342

Printing and Publishing

1.0

0.4

0.5

0.0

351

Industrial Chemicals

6.3

4.7

3.3

6.2

3511

Of which: basic ind. Chemicals

2.7

1.8

1.1

3.2

©

3512

Fertilizers and Pesticides

0.5

1.0

0.7

2.2

352

Other Chemical Products

6.0

0.6

5.2

0.8

3522

Of which: Drugs and Medicines

2.1

0.3

3.0

0.4

353+354 Petroleum Refineries and Prod.

0.9

1.8

0.1

10.3

355

Rubber Products

0.9

0.8

0.3

0.1

356

Plastic Products

1.6

0.9

0.4

0.0

361

Pottery, China, and Earthware

0.2

0.4

0.1

0.0

362

Glass and Glass Products

0.5

1.4

0.1

0.1

369

Structural Clay Products

0.9

3.3

0.4

0.2

371

Iron and Steel Basic Industr.

1.8

4.6

2.2

4.8

372

Basic Nonferrous Metals

1.2

3.2

0.3

15.6

381

Fabricated Metal Products

4.4

7.7

2.6

0.6

382

Machinery (exc. Electrical)

20.5

6.6

28.3

1.4

3821

Engines and Turbines

0.1

0.1

0.2

0.1

3822

Agricultural Machinery

0.6

0.6

1.9

0.3

3823

Metal and Wood Work. Machinery

2.6

0.9

4.3

0.3

3824

Special Industrial Machinery

7.1

1.8

12.5

0.2

Table 4.3 (cont.) ISIC

Poland, Hungary, CSFR, Romania, Bulgaria

Ex-USSR

Exports

Imports

Exports

Imports

3825

Office, Computing, Account. M

2.7

0.4

2.0

0.0

3829

Lifting, Domestic Equipm. etc.

7.2

2.8

7.3

0.5

383

Electrical Machinery

11.2

7.4

8.1

0.5

3831

Electrical Industrial Mach.

2.7

1.8

0.8

0.2

3832

Radio, Television, Communic.

4.7

1.2

5.2

0.2

3833

Electrical Appliances, Housew.

0.6

0.6

0.4

0.0

3839

Batt., Accumulators, Lamps etc.

3.2

3.8

1.6

0.2

384

Transport Equipment

11.4

6.2

20.1

1.9

3.3

0.7

3841

Shipbuilding and Repairing

0.1

1.2

a

3842

Railroad Equipment

3843

Motor Vehicles

3845

Air Craft

0.4

0.6

0.4

4.9

10.5

4.3

11.6

0.6

0.1

0.1

0.1

0.0

3844/49 Motorcycles, Bicycles etc.

0.2

0.2

0.1

0.2

385

Measuring, Photogr.,Optical etc.

3.3

0.8

3.0

0.2

390

Other Manufacturing Industries

1.5

1.1

0.5

0.8

00

All Products

100.0

100.0

100.0

100.0

001

High-Tech

19.4

9.1

16.2

1.2

002

Medium-Tech

42.9

20.9

49.5

24.8

003

Low-Tech

35.3

63.9

32.0

30.9

Mineral oil and gas imports underestimated due to secret items not shown separately by commodity and partner country. Source: DIW Foreign Trade Data.

Table 4.4. Commodity structure of German imports from five CEECs, 1993 (percentage) Poland

CSFR

Hungary

Romania

Bulgaria

1

Agriculture

2.4

2.6

5.2

2.0

6.7

2

Mining and Quarrying

4.6

3.6

0.1

0.0

0.7

3

Manufacturing Industries

93.0

93.8

94.7

98.0

92.5

321

Food, Beverages, Tobacco

7.5

2.3

11.1

3.1

10.2

322

Textiles

3.0

5.8

5.5

5.5

9.1

323

Wearing Apparel

19.6

8.6

14.7

45.8

30.5

32

Leather and Leather Products

0.3

0.5

0.4

0.4

1.6

324

Footwear

1.5

2.1

5.7

2.4

2.3 0.3

ISIC

331

Wood and Wood Products

5.7

3.1

1.3

1.5

332

Furniture and Fixture

9.0

3.1

3.1

18.5

1.6

341

Paper and Paper Products

1.0

2.0

0.6

0.3

0.3

342

Printing and Publishing

0.1

0.8

0.4

0.0

0.1

351

Industrial Chemicals

4.0

6.4

3.8

2.0

3.1

3511

Of which: basic ind.Chemicals

1.7

2.5

1.5

0.8

0.9

3512

Fertilizers and Pesticides

1.1

1.3

0.2

0.5

1.7

352

Other Chemical Products

0.5

0.8

0.6

0.4

3522

Of which: Drugs and Medicines

0.1

0.4

0.4

0.2

1.3

1.8

2.0

2.2

0.5

0.0

353+354 Petroleum Refineries and Prod.

2.0

355

Rubber Products

0.6

1.1

0.9

0.3

0.7

356

Plastic Products

0.7

1.1

0.9

0.4

0.5

361

Pottery, China, and Earthware

0.2

0.5

0.1

1.0

0.5

362

Glass and Glass Products

0.7

2.1

1.1

2.7

0.2 1.7

369

Structural Clay Products

3.3

5.3

0.9

0.2

371

Iron and Steel Basic Industry.

3.6

7.9

2.0

1.8

1.6

372

Basic Nonferrous Metals

6.1

1.0

2.0

0.4

5.0

381

Fabricated Metal Products

7.1

10.0

6.8

3.4

2.1

382

Machinery (exc. Electrical)

3.8

9.1

8.4

2.8

9.2

3821

Engines and Turbines

0.1

0.2

0.2

0.0

0.0

3822

Agricultural Machinery

0.3

0.8

1.3

0.0

0.2 4.0

3823

Metal and Wood Work. Machinery

0.4

1.3

0.7

0.5

3824

Special Industrial Machinery

1.1

2.4

2.3

0.4

1.1

3825

Office, Computing, Account. M

0.0

1.0

0.2

0.0

0.7

Table 4.4 (cont.) Poland

CSFR

Hungary

Romania

Bulgaria

3829

Lifting. Domestic Equipm. etc

1.9

3.4

3.7

1.8

3.2

383

Electrical Machinery

4.8

7.6

13.9

2.3

5.6

3831

Electrical Industrial Mach.

0.8

2.5

3.2

0.3

2.2

3832

Radio. Television, Communic.

0.6

1.5

2.1

0.2

0.5

3833

Electrical Appliances, Housevv.

0.8

0.5

0.6

0.0

0.3

3839

Batt., Accumulators, Lamps etc

2.7

3.1

8.1

1.7

2.5

384

Transport Equipment

6.2

7.6

5.7

1.3

1.9

3841

Shipbuilding and Repairing

2.3

0.6

0.6

0.1

0.0

3842

Railroad Equipment

0.2

0.7

0.2

0.6

0.0

ISIC

3843

Motor Vehicles

3.4

5.9

4.7

0.6

3845

AirCraft

0.1

0.1

0.0

0.0

1.5

3844/49 Motorcycles, Bicycles etc.

0.3

0.3

0.1

0,0

0.0

385

Measuring, Photogr., Optical etc.

0.4

1.1

1.3

0.3

0.6

390

Other Manufacturing Industries

0.6

1.8

0.9

0.6

1.7

00

All Products

100.0

100.0

100.0

100.0

100.0

0.4

001

High-Tech

5.5

10.1

15.9

2.8

9.7

002

Medium-Tech

19.3

25.0

20.9

7.3

18.8

003

Low-Tech

68.1

58.6

57.9

87.9

64.0

Source: DIW Foreign Trade Data.

116

Dieter Schumacher

include cultural similarities, a common language, a historical sense of belonging together, and regional preference zones. The gravity hypothesis advanced by Linnemann (1966) proved to be empirically very successful. A theoretical framework to complement the hypothesis was suggested by Anderson (1979) as well as Bergstrand (1985,1989). The approach has also been applied at a sectoral level in connection with factor endowment variables (e.g., Learner 1974,1993; Soete, 1987). The hypothesis has come back into favor in recent years (1) in the analysis of trends toward regionalization in world trade (e.g., Frankel, 1992, or Frankel and Wei, 1993; Saxonhouse, 1993; Dhar and Panagariya, 1994), and (2) for estimating potential trade with Eastern Europe after the political and economic changes that have occurred there (e.g., Havrylyshyn and Pritchett, 1991; Wang and Winters, 1991; Winters and Wang, 1994; Baldwin, 1994; Vittas and Mauro, 1994). An important conclusion drawn from earlier studies on the foreign trade flows of North American and Western European countries - from the beginning of the century until the middle of the 1970s - was the persistence of spatial structures in international trade (Haass and Peschel, 1982). This has been the case for three quarters of this century, a century characterized by extensive changes in the political and social structures of Europe, as well as by a considerable amount of change regarding borders. The foreign trade in the first half of the 1970s resembles the pattern before the Great Depression of the 1930s more so than that of the time in between. The fact that transportation costs and hence cif prices2 tend to increase with distance is one reason why the spatial pattern of trade depends on distance. This argumentation assumes relatively homogeneous goods and markets with perfect competition. The influence of distance, therefore, ought to be less, the more heterogeneous and imperfect the markets are (Brocker, 1980). Furthermore, transport costs as a proportion of production costs have become very low for most goods. However, distance can also have an effect on trade by means of the cost of communication taken in its broadest sense (Peschel, 1980). One cannot assume from the outset that all aspects of communication vary with distance in the same way as transportation costs. It is, therefore, necessary to distinguish between transportation and communication costs (Herrmann, Schmidtke, Brocker, and Peschel, 1982). A clear distinction is not possible, as certain aspects of communication costs may be embodied both in geographical distance and membership of preference zones. However, it was attempted in the study cited above to consider communication costs using cultural and language affinities as indicators. These affinities were portrayed in the empirical analysis by factors, such as differences in life-style and social division of labor (measured by the variations in the level of urbanization, importance of agriculture, and significance of women in employment), language barriers, religious differences, former colonial relations, and the trade relations in Western Europe and North America in 1900. 2

Prices including cost, insurance, and freight.

Impact on German trade of division of labor

117

The analysis of individual regression coefficients for total exports and for fourteen selected investment and consumer goods with the help of data from the beginning of the 1970s showed that geographical distance, former colonial ties, membership of EC, EFTA, or Commonwealth preference zone, as well as language connections are relatively significant determinants of spatial trade patterns. As for the indicators of life-style, mentality, and social division of labor - including religious affinity - as well as the foreign trade figures from 1900, the coefficients proved to be insignificant or of the wrong sign (Herrmann et al., 1982: 222). The greatest contribution to the explanation of the spatial structure of total foreign trade, after eliminating the impact of size, is provided by geographical distance, followed some way behind by the variables of "cultural and language affinities," which in turn make a more significant contribution than does membership of preference zones. The explanatory value of "cultural and language affinities" was greater for communication-cost-intensive groups of investment goods than for trade as a whole (Hermann et al., 1982: 5). On these grounds the following equation was formulated to estimate the "normal" pattern of trade flows between market economies: Y

i

Y

i

lnX. = a + bx In Y. + b2 In -j + b3 In Y. + b4 In ^ + b5 In Dtj + b6 ADJtj + b7 ECV + bs EFTAtj + b9 CUSTAy + bl0LANtj + bn COLtj where: Xr Y. Bt Y. B. Dr

= = = = = =

ADJ..

=

ECr EFTAr CUSTAr =

LANr

=

COLr

=

exports from country i to country j GNP of supplier country i population of supplier country i GNP of destination country j population of destination country j distance in miles between the economic center of country / and j dummy variable that has the value of 1 if both countries / and./ have a common land border; otherwise it is zero dummy variables with a value of 1 if both countries i andy belong to the EC, EFTA, or the Free Trade Agreement between the USA and Canada, respectively dummy variable that has a value of 1 or 0.5 if both countries / and j share the same language dummy variable that has a value of 1 or 0.5 if both countries have colonial relations with one another

The regressions were estimated for the average values of exports and imports between 1988 and 1990 of twenty-two OECD countries in trade with sev-

118

Dieter Schumacher

enty countries, namely the same twenty-two OECD countries and an additional forty-eight developing countries. These additional countries were selected according to their level of trade with OECD countries and then as to whether data were available for all remaining variables. Estimates were made for trade in all goods, goods of the manufacturing sector as a whole and of individual branches of the manufacturing sector. The industries are defined in line with the ISIC system at the three- and four-digit levels. Trade figures from the OECD reclassified from SITC Rev. 3 into ISIC Rev. 2 - were applied. Zero values were replaced by a very small figure and an OLS estimation procedure was used. The number of zero values increases when the product groups are more narrowly defined and, therefore, a Tobit estimation procedure would be more appropriate. OLS estimates should be sufficient, however, for the limited purposes of this paper, that is, to estimate flows of total trade with CEECs and to indicate the direction of likely changes in the sectoral pattern. The data on GNP per capita and population were taken from the World Bank's Development Reports. The distance between the countries was calculated as the shortest line between their commercial centers according to the degrees of latitude and longitude. We generally took the capital cities as commercial centers except for Canada (Montreal), United States (Kansas City),3 Australia (Sydney), West Germany (Frankfurt a.M.), Brazil (Rio de Janeiro), Pakistan (Karachi), and India (Bombay). The language dummy is 1 if both countries have the same first language, and it is 0.5 if the same language is the second language in one or both of them. The colonial dummy is 0.5 if such political ties existed around 1914, and it is 1 if these relations still existed in an institutional form between 1970 and 1990 (e.g., Commonwealth or Franc Zone). Considering the national product Y and population B in the equation, there are three possibilities, which are of equal value but allow for a different interpretation of the coefficients. All three variants are to be found in empirical literature and assume, in line with the hypothesis, that the national product represents the total strength of a country as a supplier and purchaser with a positive effect on the volume and variety of export supply and import demand, whereas the population represents the size of the country and has a negative effect, because larger countries tend to have a higher level of self-sufficiency. In the formulation chosen here, the impact of per capita income on trade will be quantified, as well as the impact on trade from the total purchasing power of a country and its economic strength as a supplier, respectively. The per capita income can also be interpreted as an indicator for the level of human capital or the capital-labor ratio and is the only variable in our approach that represents the factor endowment. Therefore, the coefficients that can be expected for the Y./B. variable are likely to vary largely according to sector. Alongside the endowment 3

Kansas City is a commercial center in the middle of the United States and, thus, is a geographical compromise between the centers on the East Coast and those on the West Coast.

Impact on German trade of division of labor

119

of natural resources, human capital is the decisive determinant of the sectoral pattern of a country's comparative advantage (Wood, 1994a, b). This holds particularly true for the division of labor between industrial and developing countries; it can also be found, however, in the intersectoral division of labor between industrial countries (e.g., Schumacher, 1992). The basic idea behind the approach is to explain the long-term pattern of bilateral trade flows among market economies mainly by GNP, distance, and factor endowment measured indirectly by per capita income. Empirical evidence here and elsewhere shows that the approach explains bilateral trade at the aggregate level very well. At the level of individual product groups, it provides only a partial explanation unless a larger number of factor endowment variables are introduced to represent more accurately sectoral comparative advantages. It is, however, sufficient to derive the direction of changes in comparative advantage as a function of per capita income of the supplier country. 4.4

Regression results

The multiple regressions for the trade of OECD countries in all goods combined yield typical results with regard to sign and significance of variables (Table 4.5). Trade flows between two countries are all the bigger, the higher is their national product and the smaller the geographical distance between them. As for the preference zones, EC membership has a positive impact when the analysis is based on the export statistics. The same is true of a common language and of colonial relations. The impact of a common border recorded here is not always - as should be expected - positive and is less significant than in other studies. The explanatory power of the approach is very high with regard to exports of OECD countries (R2 = 0.84). It is less impressive for imports of OECD countries (R2 - 0.50). The omission of variables concerning endowment with natural resources is apparent. This plays an important role for imports from developing countries. In contrast to the export side, EC membership and colonial relationships are not significant for imports of all goods combined. By and large, the same pattern emerges for total trade in manufactured goods. As for individual industries within manufacturing, very different coefficients arise. When grouping the industries in accordance with their intensity of R&D - following the OECD's definition - into high-, medium-, and lowtech,4 the picture given in Table 4.5 emerges. Exports of high- and mediumtech products rise considerably faster in relation to national product of the exporting country i, than exports of low-tech products. The elasticity of exports with regard to the per capita income of the exporting country / is highest for high-tech products and lowest for low-tech products. Also, the demand for 4

See OECD (1992). High-tech = ISIC industries 3522, 3825, 383, 3845, 385. Medium-tech = ISIC industries 351, 352 without 3522, 355, 356, 372, 382 without 3825, 3842, 3843, 3844/9. Low-tech = all manufacturing industries not classified as high- or medium-tech. As distinct from OECD we classified ISIC 390 (other manufacturing) as low-tech instead of medium-tech.

Table 4.5. Regression results for trade of 22 OECD countries with 70 countries (all variables) Exports

to(y/B,)

AWL

EC

EFTA

CUSTA

LAN

COL

Constant

Degrees of Freedom

000 All goods A'-Coefficients StdErrofCoef. T-Test Significance

0.89 0.02 39.5 **

0.40 0.04 8.9 **

0.79 0.02 37.6 **

0.17 0.03 6.6 *•

-0.87 0.04 -22.8 **

0.01 0.19 0.0

0.31 0.12 2.5 •

-0.45 0.25 -1.8

0.07 0.76 0.1

0.89 0.11 8.2 **

1.16 0.23 5.1 *•

0.84

-13.16

1506

3 Manufactures A'-Coefficients StdErrofCoef r-Test Significance

0.87 0.02 36.5 **

0.49 0.05 10.4 *•

0.77 0.02 34.3 •#

0.17 0.03 6.2 **

-0.95 0.04 -23.4 **

-0.04 0.20 -0.2

0.28 0.13 2.2 *

-0.48 0.27 -1.8

0.17 0.81 0.2

0.78 0.12 6.7 *•

1.32 0.24 5.4 *•

0.82

-13.18

1506

001 High-Tech A'-Coefficients StdErrofCoef. T-Test Significance

1.10 0.07 14.8 **

2.11 0.15 14.3 **

0.95 0.07 13.5 **

0.18 0.08 2.2 *

-1.40 0.13 -11.1 **

-1.15 0.62 -1.9

0.20 0.41 0.5

-1.14 0.84 -1.4

-0.94 2.52 -0.4

1.48 0.36 4.1 •*

1.82 0.76 2.4 *

0.50

-32.32

1506

1.19 0.04 30.0

0.83 0.04 22.2 **

0.08 0.04 1.7

-1.20 0.07 -17.9 **

-0.23 0.33 -0.7

0.08 0.22 0.4

-0.29 0.44 -0.7

0.34 1.34 0.3

0.56 0.19 2.9 «*

1.45 0.40 3.6 **

0.69

-17.19

1506

••

0.60 0.08 7.7 •*

0.74 0.04 18.9 **

0.29 0.08 3.8 **

0.79 0.04 21.6 •«

0.22 0.04 5.1 **

-1.09 0.07 -16.5 **

-0.21 0.32 -0.7

0.18 0.21 0.9

-0.66 0.44 -1.5

-0.24 1.32 -0.2

1.05 0.19 5.6 **

1.30 0.40 3.3 •*

0.63

-10.19

1506

000 All goods A'-Coefficients StdErrofCoef. r-Test Significance

1.02 0.06 18.2 **

0.17 0.07 2.5 *

1.18 0.06 19.9 **

-0.24 0.12 -2.0 •

-0.93 0.10 -9.2 •*

-0.41 0.49 -0.8

0.11 0.33 0.3

0.10 0.67 0.2

-0.52 2.00 -0.3

0.97 0.29 3.4 **

1.05 0.60 1.7

0.50

-12.87

1506

3 Manufactures Jf-Coefficients StdErrofCoef r-Test Significance

1.20 0.06 18.6 **

0.24 0.08 3.1

1.24 0.07 18.1 **

-0.13 0.14 -1.0

-1.00 0.12 -8.6 ••

-0.74 0.57 -1.3

0.38 0.38 1.0

0.39 0.77 0.5

-0.79 2.32 -0.3

1.25 0.33 3.8 *•

1.20 0.70 1.7

0.49

-17.35

1506

001 High-Tech A'-Coefficients StdErrofCoef. r-Test Significance

2.08 0.15 14.0 **

1.25 0.18 7.1 •*

1.59 0.16 10.1 **

1.04 0.31 3.3 •*

-1.34 0.27 -5.0 •*

-2.03 1.31 -1.6

0.70 0.87 0.8

0.12 1.77 0.1

-3.38 5.33 -0.6

2.07 0.76 2.7 *•

4.09 1.60 2.6 •*

0.40

-53.00

1506

002 Medium-Tech A'-Coefficients StdErrofCoef. r-Test Significance

1.85 0.13 14.5 •*

0.77 0.15 5.0

1.76 0.13 13.0 **

-0.53 0.27 -2.0 *

-1.56 0.23 -6.8 **

-1.45 1.12 -1.3

-0.15 0.74 -0.2

0.83 1.52 0.5

-1.37 4.57 -0.3

0.42 0.65 0.6

3.10 1.38 2.3 •

0.39

-29.90

1506

003 Low-Tech JT-Coefficients StdErrofCoef. r-Test Significance

1.25 0.07 17.0 ••

0.14 0.09 1.6

1.29 0.08 16.6 *•

0.03 0.15 0.2

-1.03 0.13 -7.8 •*

-0.84 0.64 -1.34

0.41 0.43 1.0

0.39 0.87 0.4

-1.57 2.63 -0.6

1.37 0.38 3.6 *•

1.11 0.79 1.4

0.44

-19.81

1506

002 Medium-Tech A'-Coefficients StdErrofCoef. r-Test Significance 003 Low-Tech A'-Coefficients StdErrofCoef. r-Test Significance

Note: •• indicates significance at 99% level, * indicates significance at 95% level. Source: Own calculations; for methodology and data see text.

122

Dieter Schumacher

high-tech products increases most sharply when the national product of the importing country j rises. At the level of individual industries (classified according to three- and fourdigit ISIC numbers, respectively), the validity of the approach diminishes in most cases, with R2 values ranging from 0.3 to 0.8. In general, the explanatory power of the approach is weakest in raw-material-intensive and/or laborintensive sectors, such as foodstuffs, textiles, clothing, leather goods, wooden articles, nonferrous metals, and other manufacturing goods (musical instruments, toys, sporting goods, and jewelery). The main reason here is, that, besides the per capita income as an indicator for the capital-labor ratio, no variables for the factor endowment of the various countries are considered. At the level of individual industries, noticeably higher elasticities of income and distance emerge than at a higher level of aggregation. In the majority of industries, unexpected negative signs emerge for a common border, which are in some cases even significant. The interpretation of coefficients is difficult, however, because the explanatory variables are not completely independent of each other. For instance, a common language and common border coincide in a number of cases. A common language and colonial relations overlap even more often. There is also, to a certain extent, a correlation between income level and distance, as a large number of countries with a high level of income are concentrated in Western Europe. These connections do not call the validity of the approach as a whole into question, but complicate the process of distinguishing exactly the impact of individual variables. Alternative calculations considering solely national product, per capita income, and geographical distance yield results with only slightly less explanatory power (Table 4.6). This is valid for aggregate trade, as well as for individual sectors. The additional explanatory contribution gained by taking into consideration factors such as a common border, membership of a preference zone, a shared language, and colonial relations, is, therefore, very slight for all OECD countries on average. The ranking of industries according to income elasticities and elasticity of distance changes only marginally compared to the result derived using all variables. In Table 4.7 results of the reduced approach are compiled ranking the threedigit industries according to the estimated coefficient values of the respective variable. The figures give the number of the ISIC industry, the value of the coefficient, and the value of t. Distance and GNP are significant determinants of bilateral shipments in all industries, the same is true for GNP per capita in most cases. The coefficient values vary widely across industries, however, and there are also differences between exports and imports.5 In general, exports are more sensitive to distance than imports (b5). On the 5

This is due to the OECD trade with developing countries. Considering trade only among OECD countries, the results for exports and imports are very similar; here, the differences are due to divergences between export and import statistics of the same flows.

Impact on German trade of division of labor

123

export side, food, general machinery, electrical machinery, and precision instruments show the smallest (negative) coefficient for distance and rank highest on the list. On the import side, this is true for "other" manufactures, textiles, and plastic products. OECD trade that occurs over rather long distances tends to be concentrated on exports of investment goods and imports of consumer goods, reflecting the intersectoral division of labor between OECD countries and developing countries. On the other hand, basic goods like mineral oil products, iron and steel, "other" nonmetallic mineral products, and nonferrous metals are traded mainly with nearby countries. As for imports, this is also true for transport equipment and machinery. When estimating future patterns of exports from the CEECs as a function of their GNP one should refer to the coefficients of GNP (bx) and GNP per capita (b2) in the import equations, because they cover the broad range of OECD and developing countries as suppliers. With regard to likely future patterns of exports to CEECs as a function of their GNP, one should rely on the coefficients of GNP (b3) and GNP per capita (b4) in the export equation, because they represent the whole range of countries as importers. The coefficients show that exports of mineral oil products, plastic products, precision instruments, and transport equipment increase most quickly with per capita income of the supplier country (b2). In total (b{ + b2) the exports of iron and steel, glass products, and paper products increase most with GNP of the supplier country. On the other hand, the imports of consumer goods such as footwear, leather products, clothing, and furniture increase most quickly with per capita income (b4) as well as GNP as a whole (b3 + b4) of the importing country. Referring to the results from the regressions including all variables, we find a significant positive impact of EC membership on imports in a number of industries, although the EC dummy is not significant for the overall level of imports. It is significant for food and those industries that mainly produce consumer goods (ISIC numbers 321 to 342, 355 to 369, and 390). This may reflect the impact of EC protection, which is highest for agriculture and food, due to the Common Agricultural Policy, and for consumer goods, in particular, clothing, textiles, and leather products, which tend to be more strongly protected by tariffs and by nontariff barriers than producer goods and investment goods (Mobius and Schumacher, 1995). To a certain extent, however, the EC dummy may be significant because it covers part of the impact of distance. The outcome of the analysis also differs for individual countries. When the regression equation given above is applied to the trade of the six biggest industrial countries individually, then national product and distance prove to be the deciding determinants for Germany as well as France, Italy, Great Britain, Japan, and the United States. For Great Britain and France, however, language and colonial relations are also important factors. Furthermore, the exports and imports of Japan and the United States are more sensitive to distance than the trade flows of the biggest European industrial countries are. The explanatory power of this approach is also very high for the trade of individual countries -

125

124

Table 4.6. Regression results for trade of 22 OECD countries with 70 countries (five variables) in(Y)

ln(r/B,)

R2

Constant

Degrees of Freedom

Exports 000 All goods

3 Manufactures

001 High-tech

002 Medium-tech

003 Low-tech

A'-Coefficients StdErrofCoef. T-Test Significance ^-Coefficients StdErrofCoef. T-Test Significance ^-Coefficients StdErrofCoef. T-Test Significance ^-Coefficients StdErrofCoef. T-Test Significance ^-Coefficients StdErrofCoef. T-Test Significance

0.92 0.02 39.8 **

0.38 0.05 8.2 **

0.79 0.02 36.3 •*

0.17 0.03 6.5 **

-0.89 0.03 -27.1 **

0.82

-13.07

1512

0.91 0.02 3.72 **

0.47 0.05 9.7 **

0.77 0.02 33.4 **

0.17 0.03 6.0 •*

-0.96 0.03 -27.8 **

0.81

-13.11

1512

1.14 0.07 15.5 **

2.08 0.15 14.1 **

0.93 0.07 13.3 **

0.18 0.08 2.2 *

-1.30 0.10 -12.3 •*

0.49

-32.99

1512

1.22 0.04 31.0 **

0.58 0.08 7.4 *•

0.82 0.04 22.1 ••

0.08 0.04 1.7

-1.18 0.06 -21.2 **

0.68

-17.35

1512

0.77 0.04 19.7 *•

0.27 0.08 3.5 **

0.79 0.04 21.3 •*

0.23 0.04 5.1 *«

-1.07 0.06 -19.3 •*

0.62

-10.35

1512

Imports 000 All goods

3 Manufactures

001 High-tech

002 Medium-tech

003 Low-tech

A'-Coefficients StdErrofCoef. T-Test Significance A'-Coefficients Std Err of Coef. T-Test Significance A'-Coefficients StdErrofCoef. J-Test Significance AT-Coefficients StdErrofCoef. T-Test Significance A'-Coefficients StdErrofCoef. T-Test Significance

1.00 0.06 18.1 **

0.18 0.07 2.6 **

1.20 0.06 20.6 **

-0.24 0.12 -2.1 *

-0.90 0.08 -10.8 **

0.49

-13.14

1512

1.18 0.06 18.4 *•

0.26 0.08 3.3 ••

1.26 0.07 18.6 ••

-0.13 0.14 -1.0

-0.99 0.10 -10.2 **

0.49

-17.56

1512

2.03 0.15

1.27 0.18 7.2 **

1.66 0.16

-1.25 0.22 -5.7 *•

0.39

-53.59

1512

10.7 **

1.00 0.31 3.2 *«

18.1 0.13 14.4 **

0.77

1.79 0.13 13.5 •*

-0.54 0.27 -2.0 *

-1.44 0.19 -7.7 **

0.39

-30.69

1512

1.23 0.07 16.9

0.16 0.09 1.8

1.31 0.08 17.1 •

0.03 0.15 0.2

-1.01 0.11 -9.2 *•

0.43

-20.04

1512

13.8 **

0.15 5.1 **

Note: ** indicates significance at 99% level, * indicates significance at 95% level. Source: Own calculations; for methodology and data see text.

to

Table 4.7. Ranking of three-digit ISIC industries according to the value of the regression coefficients. Trade of 22 OECD countries with 70 countries {five variables) Exports ISIC

b,tfnYj Coeff.

T-Test

b2 (for Y/B,) Sign.

ISIC

Coeff.

T-Test

bt+b2 Sign.

ISIC

Coeff.

385

6.28

353/4

4.81

23.4

**

385

4.81

20.4

••

324

4.18

20.3

**

341

4.41

14.5

**

342

6.27

361

4.10

21.3

**

342

3.89

13.1

**

341

6.25

323

3.68

19.1

**

332

2.90

8.2

••

332

5.98

371

3.62

20.1

**

384

2.76

10.8

••

372

5.13

322

3.51

20.4

**

356

2.66

9.7

**

384

5.02

362

3.31

19.5

••

352

2.57

11.7

••

331

4.85

369

3.22

18.5

**

383

2.26

11.9

**

371

4.78

372

2.21

6.1

•*

356

4.63

355

3.11

20.1

**

332

3.09

17.5

**

381

2.08

9.7

**

390

4.27

372

2.93

16.2

**

382

2.03

14.4

••

355

4.00

3

p

m

ON

ON

VO*

O

R

00

en

ON

127

00

00

en

2 en

2

en

en

$

2 en


g

s en

c r cn

oo en

^H

9

^H

9

O

CN

00 O

p

00

9

§

9

en so

-T -T

s; s a

p ^

O

^*

ON

^*

00

^

f>;

^

Tf

^

^"

^j

Tt

^'

CN

_H*

»—"

o

VO

O o o e n e n t ^ s o v o c N * / ^ * / ^

<>i


128

1 I

£

3

.i

cs -H'

ors G K

CO

CO

CO

O


O

00

O

co

CN

CO


m

O

—« «n


00

CO

n

so

I-H CN

O

ON

o

ON

CO

CN

s en


C

>

GK

' v o 00 00 CN
CO

oo

CO

o o co

cs

CO

m u > C h co co co

CO

CO

H

r4

CO

1^' in

^H'

TJ-

ON

^H'

r

oo

^

ON

oo

T-5

-

^

CO

»

^H'

-

^H'

^

OS

CO

H

co

^

©

N

ON

C

oo

H

ON

^

^H

O

t^.

C

^H

O

O^

C

CO

S

c N « n r > - r > - s o c s c o f N CO

in ~

CO

so

en

so CO

^

so

en

CN

C>

oo

en

*-•

c>

o

en

^H

o

en

o

oo

OS

"-< CN

vo CO

en

^

00

t-» CO

en

en CN

CN

en

—J

r**»

S

CN

en

oo CN

C^

oo

cn

en

so

^^^

r*.

en

i-H CN

C^^

35

cn

Tf CN

^^^

00

t**»

^^^

CN

cn

en

^^

TJ-

oo CN

^^^

o

^^5

^^^

o

*^^T

^^^

o

o

o

^^^

o

o

o

^^N

o

o

-H

o

^^

m

o

o

m

o

CN

s o c N t ^ o o r ^ ^ - r ^ t ^ i n c N s o r ^ c N P ^ o o N o d t ^ O N c J o d o r ^ o ^ O N T f o

^

in

m

^

CN SO

^

^

^

^

^

^

^

^H'

O

O

O

i — l O o o s o s o i n e n o O N t - ^ i n o o i n o o en en CN CN CN CN CN CN O O O ON ON oo ^

ON

~* o CN

129

I

s

8 V 55

s 55

a

o so

O

O

^H'

O

«n

8

in

§. rf

85

= g s

^ H o o s q i n » n i n i n « n

^J
SO CO

ON

E:


O

Tl
»-H'

—4 Tf CO

CS

^H

P*» CO

p en

131

381

3.42

19.3

**

362

1.39

6.2

**

385

4.58

351

3.42

18.6

*•

352

1.36

6.0

**

342

4.43

352

3.31

17.6

**

351

1.31

5.9

**

324

4.26

353/4

3.27

15.9

*•

369

0.83

3.5

••

361

4.24

384

3.05

16.8

•*

381

0.73

3.4

**

381

4.15

*

332

3.03

16.4

**

390

0.44

2.1

382

4.11

342

2.92

16.8

**

322

0.34

1.5

383

3.87

385

2.73

14.9

**

321

0.32

1.6

323

3.45

321

2.55

15.1

**

361

0.25

1.0

332

3.05

382

2.53

15.9

**

31

0.13

0.8

390

2.93

390

2.49

14.3

**

332

0.02

0.1

321

2.87

383

2.47

15.2

**

324

0.01

0.0

331

2.75

322

2.39

12.8

*•

323

-0.07

-0.3

322

2.73

31

1.72

12.1

**

331

-0.89

-3.6

31

1.86

**

132 Table 4.7 (cont.) Imports

b3 (for Yj)

bt (for)

ISIC

Coeff

r-Test

Sign

ISIC

Coeff

T-Test

353/4

4.18

19.3

**

322

3.28

8.3

*•

332

5.06

13.5

**

332

7.0

»•

322

4.94

2.73

Sign.

ISIC

Coeff.

ISIC

Coeff

r-Test

Sign.

390

-0.76

-2.9

**

321

-0.89

-3.5

**

372

3.15

371

2.65

11.9

*•

361

2.65

6.3

*•

342

4.73

356

-1.07

-3.7

•*

341

2.48

11.6

**

342

2.42

6.6

**

361

4.69

361

-1.11

-3.7

**

369

2.46

11.8

•*

390

2.29

6.2

••

390

4.13

352

-1.11

-3.9

**

362

2.46

12.4

*•

324

1.69

3.8

•*

324

3.60

31

-1.11

-5.2

•*

351

2.41

12.4

**

331

1.56

3.5

**

331

3.55

323

-1.24

-4.2

**

381

2.39

12.7

**

383

1.06

3.1

••

383

3.07

322

-1.35

-4.8

*•

352

2.37

12.0

**

321

0.71

2.0

•

369

3.00

385

-1.38

-5.0

••

323

2.35

11.3

**

385

0.67

1.7

381

2.99

342

-1.40

-5.4

*•

384

2.32

12.1

**

381

0.60

1.6

385

2.92

381

-1.46

-5.5

**

133

332

2.32

11.9

**

369

0.54

1.3

356

2.62

362

-1.48

-5.2

**

342

2.31

12.6

**

355

0.51

1.2

353/4

2.59

383

-1.53

-6.3

•»

356

2.28

11.1

**

356

0.34

0.8

384

2.58

332

-1.55

-5.6

••

385

2.25

11.7

»*

384

0.25

0.7

352

2.48

331

-1.58

-5.0

»•

382

2.08

12.3

**

382

0.15

0.5

355

2.46

355

-1.58

-5.2

**

361

2.04

9.7

**

352

0.11

0.3

362

2.45

324

-1.63

-5.2

•*

383

2.01

11.7

**

362

-0.01

-0.0

372

2.42

372

-1.74

-5.3

**

331

1.99

9.0

*•

31

-0.26

-0.9

321

2.36

351

-1.75

-6.4

*•

9.2

••

323

-0.29

-0.7

382

2.23

382

-1.77

-7.4

*•

8.6

••

371

-0.53

-1.2

371

2.12

341

-1.79

-5.9

*•

10.0

•*

372

-0.73

-1.6

323

2.06

369

-1.84

-6.2

»•

11.9

••

341

-0.78

-1.8

341

1.70

384

-2.16

-7.9

**

8.4

*«

351

-1.33

-3.4

**

31

1.53

371

-2.55

-8.1

*«

9.3

**

353/4

-1.59

-3.7

**

351

1.08

353/4

-4.65

-15.1

**

355 324 390 31 322 321

1.95 1.90 1.84 1.79 1.67 1.65

Note: ** indicates significance at 99% level, * indicates significance at 95% level. Source: Own calculations; for methodology and data see text

134

Dieter Schumacher

with the exception of the United States - in all goods and all manufactured products on the export side as well as the import side (with R2 of 0.8 to 0.9). In contrast, the equation only explains 50 percent of the variance of trade in the case of the United States. It is possible that U.S. trade is oriented to such an extent toward the Pacific region that it cannot be explained by the level of income and distance alone. This can be taken into consideration by the use of a dummy variable for the membership of APEC, which according to the results from Frankel and Wei (1993), displays the highest significance of the various regional trading blocks. Selected results for Germany are shown in Tables 4.8 and 4.9. They demonstrate that German exports and imports are highly dependent on the national product of the destination and supplier countries, respectively, and on geographical distance. Restricting the approach to these variables reveals a substantially higher distance elasticity of trade than is derived from the equation when all variables are used. This shows that a part of the impact of distance is included in the variable for membership of the EC. As far as Germany's trading partners are concerned, a high GNP, geographical proximity to Germany, and membership in the EC often coincide. The pattern of coefficients according to R&D content of products is much more pronounced for Germany than for the average of OECD countries. Low-tech products are exported mainly to countries near to Germany, whereas medium- and high-tech goods are exported over longer distances. The income elasticity of exports of all three product groups is approximately 1, that is, in this categorization of goods the commodity structure of German exports varies only slightly with the destination country's level of GNP. In contrast, the GNP of the supplier country plays an important role on the import side: With increasing GNP of the supplier country - assuming a constant population - German imports of low-tech products rise at the same pace, whereas imports of medium-tech goods increase nearly 50 percent faster and high-tech goods at even double the rate. Distance has a similarly strong (negative) impact on the level of imports in all three product groups. In this subdivision the commodity structure of exports tends to vary with distance, whereas the commodity structure of imports varies more in connection with the GNP of the partner country. At the three-digit ISIC level, GNP and distance are significant determinants of German exports in all industries, and GNP per capita is significant in most cases (Table 4.10). The industries whose exports increase most strongly with changes in GNP of the importing country (b3 + b4) are the same consumer goods industries as for OECD countries on average. The ranking of German industries according to the distance elasticity of exports is different from the OECD average in a number of cases. Again, the exports of investment goods, however, tend to be less sensitive to distance than producer and consumer goods. As for German imports, the volume is determined significantly by the GNP of the supplier country in all industries, whereas GNP per capita is significant only for precision instruments, general and electrical machinery, plas-

Impact on German trade of division of labor

135

tic products, and industrial chemicals. Distance plays an important role for German imports of machinery and mineral oil products, as well as precision instruments, electrical machinery, and printing goods, whereas it is not significant at all in the other industries. All in all, the figures tend to show that German exports are more sensitive to distance than German imports. Referring to the regressions including all variables, the dummy for EC membership is not significant for German imports of any industry at the three-digit ISIC level. This may show the relatively liberal behavior of Germany, despite EC protection in general. For German exports the EC dummy is significant in a number of consumer goods industries. It remains unclear, however, to what extent this covers merely part of the impact of distance. 4.5

Estimates for Germany's trade with CEECs

The estimates for Germany's trade with CEECs are derived from the regression results of the reduced approach, which includes the impact of total GNP, the level of GNP per capita, and distance solely. We rely on the regression results for German foreign trade, so that the characteristics of Germany can be considered. Additionally, the results for the trade of OECD countries are applied to provide a comparison and wider basis for estimation. The crucial variable for estimating potential trade between Germany and CEECs is - besides distance - the level of GNP. As a Western-style national accounts system is just in the process of being developed, figures are still subject to large discrepancies. The World Bank's most up-to-date estimates for the values of per capita national income in 1992 are shown in Table 4.11. These are considerably lower than earlier estimates, which, in any case, vary substantially (see Losch and Wohlers, 1994). The lowest values are from the CSFB and correspond more closely to present estimates. The estimates of the CIA, however, are as much as four or five times as high. The present-day GNP must already be lower than the level attained at the end of the 1980s, as a large part of physical and human capital lost its value in the course of transformation to a market economy. Presumably, the latest figures from the World Bank still overestimate the state of development of the former CMEA countries - with the exception of the Visegrad group (Losch and Wohlers, 1994: 154). In particular the per capita income in Russia may be overestimated. By and large, however, one can assume that the estimates for GNP per capita in 1992 offer a realistic point of reference. According to these estimates Russia, Poland, the CSFR, and Hungary have reached about one tenth of the West German level of GNP per capita and are on a par with Turkey and middle-income developing countries (like Thailand, Iran, Panama, or Chile). Romania and Bulgaria have attained only a twentieth of the German level and rank among the lower third of middle-income developing countries (like Morocco, Ecuador, Colombia, Jamaica, or Paraguay). In contrast, CEECs should have already gained, according to CIA estimates, the

136

137

Table 4.8. Regression results for trade of West Germany with 69 countries (all variables) \n(Y)

ln(r/B,)

ln(i>,,)

ADJL

EC

LAN

R2

Constant

Degrees of Freedom

Exports 0.15 0.06 2.4 *

-0.48 0.10 -4.8 **

0.13 0.34 0.4

0.51 0.27 1.9

0.80 0.52 1.6

0.94

0.67

62

Significance

0.80 0.05 16.4 **

3 Manufactures

X-Coeff. Std. Err. of Coeff. r-Test Significance

0.79 0.05 16.1 **

0.17 0.06 2.4 «*

-0.47 0.10 -4.7 *«

0.11 0.34 0.3

0.56 0.28 2.0 *

0.84 0.52 1.6

0.94

0.48

62

001 High-tech

X-Coeff. Std. Err of Coeff. 7-Test Significance

0.85 0.06 13.0 **

0.22 0.08 2.7 **

-0.32 0.13 -2.4 *

0.05 0.46 0.1

0.82 0.37 2.2 •

1.12 0.69 1.6

0.91

-3.49

62

002 Medium-tech

X-Coeff. Std. Err. of Coeff. r-Test Significance

0.82 0.05 16.9 **

0.14 0.06 2.4 *

-0.40 0.10 -4.0 **

0.15 0.34 0.4

0.50 0.27 1.8

0.76 0.51 1.5

0.94

-0.77

62

003 Low-tech

X-Coeff. Std. Err of Coeff.

0.68 0.06 12.34 *•

0.20 0.07 3.00

-0.74 0.11 -6.54 **

0.05 0.39 0.13

0.46 0.31 1.47

0.79 0.58 1.34

0.93

2.00

62

000 All Goods

JT-Coeff. Std.Errof Coeff.

r-Test

r-Test Significance

*•

Imports 000 All Goods

X-Coeff. Std. Err. of Coef. r-Test Significance

0.76 0.07 10.5 **

0.11 0.09 1.3

-0.30 0.15 -2.1 *

0.45 0.50 0.9

0.65 0.40 1.6

0.59 0.76 0.8

0.85

-0.07

62

3 Manufactures

X-Coeff. Std. Err. of Coeff. r-Test Significance

0.89 0.12 7.7

0.17 0.14 1.2

-0.20 0.24 -0.8

0.44 0.81 0.5

1.29 0.65 2.0

1.31 1.22 1.1

0.77

-3.67

62

001 High-tech

Jr-Coeff. Std. Err. of Coeff. r-Test Significance

1.24 0.20 6.2 *•

0.66 0.24 2.7 **

-0.08 0.41 -0.2

-0.09 1.39 -0.1

2.05 1.11 1.8

2.52 2.09 1.2

0.74

-15.64

62

002 Medium-tech

JT-Coeff. Std. Err. of Coeff. 7/-Test Significance

0.97 0.18 5.4 *•

0.39 0.22 1.8

-0.23 0.37 -0.6

0.81 1.26 0.6

1.45 1.01 1.4

1.62 1.90 0.9

0.68

-8.18

62

003 Low-tech

X-Coeff. Std. Err. of Coeff. r-Test Significance

0.86 0.13 6.5 **

0.08 0.16 0.5

-0.36 0.27 -1.3

0.35 0.93 0.4

1.10 0.75 1.5

0.86 1.40 0.6

0.69

-2.13

62

Note ** indicates significance at 99% level, * indicates significance at 95% level. Source: Own calculations; for methodology and data see text

139

138

Table 4.9. Regression results for trade of West Germany with 69 countries (three variables) MY)

ln(r/B,)

R2

Constant

Degrees of Freedom

Exports 000 All Goods

JT-Coeff. Std. Err. of Coeff. 7/-Test Significance

0.81 0.05 16.5 **

0.17 0.06 2.7 **

-0.65 0.07 -9.2 **

0.93

1.83

65

3 Manufactures

Jif-Coeff. Std. Err. of Coeff. T-Test Significance

0.80 0.05 16.2 *»

0.18 0.06 3.0 **

-0.65 0.07 -9.1 **

0.93

1.69

65

001 High-tech

X-Coeff. Std. Err. of Coeff. T-Test Significance

0.86 0.07 13.1 **

0.24 0.08 3.0 **

-0.55 0.09 -5.9 **

0.90

-1.89

65

002 Medium-tech

Jf-Coeff. Std. Err. of Coeff. T-Test Significance

0.83 0.05 17.0 **

0.16 0.06 2.7 **

-0.57 0.07 -8.1 **

0.93

0.37

65

003 Low-tech

X-Coeff. Std. Err. of Coeff. T-Test Significance

0.69 0.05 12.6 **

0.22 0.07 3.3 »*

-0.88 0.08 -11.3 **

0.92

2.98

65

Imports 000 All Goods

X-Coeff. Std. Err. of Coeff. T-Test Significance

0.77 0.07 10.7 **

0.13 0.09 1.5

-0.54 0.10 -5.3 **

0.84

1.59

65

3 Manufactures

X-Coeff. Std. Err. of Coeff. T-Test Significance

0.93 0.12 7.9

0.21 0.14 1.4

-0.61 0.17 -3.6 **

0.75

-0.89

65

001 High-tech

X-Coeff. Std. Err. of Coeff. T-Test Significance

1.29 0.20 6.5 **

0.71 0.24 2.9 **

-0.61 0.28 -2.2 *

0.72

-11.99

65

002 Medium-tech

A'-Coeff. Std. Err. of Coeff. T-Test Significance

1.01 0.18 5.6

0.44 0.22 2.0 *

-0.74 0.26 -2.9 **

0.66

-4.63

65

X-Coeff. Std. Err. of Coeff. T-Test Significance

0.89 0.13 6.80 **

0.11 0.16 0.69

-0.69 0.19 -3.66 **

0.68

0.12

65

003 Low-tech

•*

Note: ** indicates significance at 99% level, * indicates significance at 95% level. Source: Own calculations; for methodology and data see text.

Table 4.10. Ranking of three-digit ISIC industries according to the value of the regression coefficients. Trade of West Germany with 69 countries (three variables) b, (for lft

Exports ISIC

371

Coeff. 0.92

T-Test

bAfoxY/B) Sign.

11.9

••

ISIC 322

Coeff 0.96

T-Test 6.0

ft, (for £>^ 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

<§

o

0.2

9

©

en

i-H

rs

i-H

cs

9

vq

en

S 3

*-H

*

in

^H

oi

^H

cs

m

*-H

en

«

n

•—i

en

«

n

CN

en

C

N

T

t

en

2 S

en

V

oo

c>

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*L
0< 3< 1

where a is a scale parameter, K. is the respective firm's capital stock, L. is labor employed, and a is the supply elasticity of capital outside the firm. This leads to the aggregate production function:

The aggregate output elasticity (P 4- a) of capital is bigger than the output elasticity 3 perceived by individual firms, so that profit maximization will lead to a suboptimal investment rate. Moreover, the marginal productivity of capital will not fall with individual investment, provided that other firms undertake sufficient investment and R&D with positive spillover effects. Consequently, government should tax consumption and use part of the tax revenue to subsidize capital accumulation and R&D appropriately. Models of endogenous growth as developed by Romer (1986; 1990) - emphasize that growth is generated by R&D and innovations, which partly represent public goods with positive

226

Paul J. J. Welfens

spillover effects in the whole economy. This suggests that R&D should be subsidized because the innovator generates a positive externality, which implies that - relative to optimum R&D - insufficient R&D expenditures are generated by the private sector. Lucas (1990) suggested that human capital is accumulated in a way that has positive spillover effects, so that private human capital accumulation would be suboptimal if it were not subsidized by government. If efficiency improvements can raise the return to capital, investment will be higher than otherwise and savings will be stimulated in parallel via higher real interest rates. The second-round impact from investment and trade can be substantial if both human capital and physical capital can be accumulated. Dynamic gains could stem from integration, which leads to economies of scale and positive spillover effects that prevent returns to capital accumulation from falling quickly. This explains empirical evidence showing that increased openness stimulates investment (Levine and Renelt, 1992). The World Bank has estimated such second-round impacts as twice as high as the first round impact for developing countries (World Bank, 1995, p. 60). For Eastern Europe a similar reasoning holds, a view reinforced by dynamic growth effects estimated by Baldwin (1989) for the EU. In a Rebelo-type (Rebelo, 1990) growth model, Y — zK and the growth rate g is given by (z - a)/ft, where a is the time preference rate and ft the inverse of the elasticity of the consumption function. Low time-preference rates in Eastern Europe could stimulate an intra-European convergence process. Intertemporal substitution elasticities in Europe also play a role. If one were to apply the new growth theory to Eastern Europe, it makes the subsidization of capital investment appear desirable. On the other hand, even if the social return on investment in some sectors exceeds the real interest rate, the subsidization of some sectors will stimulate massive rent-seeking activities by other sectors. Moreover, reducing subsidies in industries that lack comparative advantage would become even more difficult in a political climate of expanding subsidies. In the initial transition stage, the main task in the field of subsidies is to eliminate at first most of the subsidies in order to reduce Xinefficiencies, to discourage rent-seeking activities, and to promote exit of firms that are nonviable. High public debt/GNP ratios reinforce the antisubsidy case for investment in transforming economies. R&D subsidies should be granted only to private firms. Subsidization of human capital formation might be quite useful in the postsocialist countries, which, except for the Czech Republic, all suffer from high unemployment. Retraining, upgrading human capital skills, and improving the education system could help to stimulate a technological and economic catchingup process in the long run. In most ex-CMEA countries, people employed in the public education system have, however, suffered from a loss in real income positions. Providing free state-financed education in Eastern Europe is of limited economic significance if there is a wage gap vis-a-vis the private business sector; with respect to teaching personnel, an adverse selection process in grade schools, high schools, and universities will be the consequence. Privatizing part

Privatization, structural change, and productivity

227

of the university system thus is an interesting long-term option to raise productivity. Finally, positive external supply-side effects are associated with infrastructure investment. Munnell (1993) found output elasticities of 0.15 for infrastructure, 0.31 for private capital, and 0.59 for labor, so that output elasticities exceed unity - there are increasing returns to scale. The marginal product of public capital was found to be roughly equal to that of private capital. For various reasons there is an enormous backlog in infrastructure capital formation in Eastern Europe, so that one may assume that the marginal product of infrastructure capital is rather high.2 High investment in the energy sector, in telecommunications, and in the traffic system could therefore stimulate economic growth and productivity growth. Generally, there is no need to provide infrastructure capital through state-owned firms. On the contrary, providing infrastructure services is associated with high capital intensities and hence big investment projects. Privatization of energy generation and telecommunications as well as other infrastructure activities should be a priority if increasing economic growth ranks high on the political agenda. Privatization coupled with free competition or regulation and an efficient capital market as afilterfor investment decisions could stimulate growth indirectly. In reality, privatizing and competition in infrastructure investment play a very limited role in Eastern Europe. Instead of introducing infrastructure capital into a macroeconomic production function, one may analyze the effect of infrastructure investment with respect to the reduction of costs associated with other input factors. Seitz and Licht (1995) found the following infrastructure elasticities for Germany: private production costs -0.22, private investment in machinery 0.13, construction investment 0.22, and demand for labor -0.32. This means that infrastructure investment increases private sector productivity by allowing saving on costs (and input resources). Infrastructure also stimulates private production in machinery and construction, but there is a direct substitution effect with respect to labor. The net effect on labor demand could be positive if the equipment and machinery industry has a relatively high labor intensity. Privatization and network effects Privatization includes three elements of private ownership in industrial assets: (1) creation of private property rights so that state-owned firms can be sold to 2

Traffic infrastructure traditionally focused on East-East trade links within the CMEA, while postsocialist trade orientation is much more along the East-West axis in Europe. The use of telecommunications was restricted as a means to control private political activities and initiatives, but also because in the command economy there was limited need in industry for autonomous decision making, horizontal exchange of information, and flexible alliance building between firms. The spatial structure of the division of labor typically is more widespread in capitalism than in socialism; communication needs in a decentralized economy are enormous; and the rising aspiration level of people in Central and Eastern Europe reinforces the need for infrastructure investment. In energy generation only Hungary considers privatization; private road financing is considered in Hungary and Poland.

228

Paul J. J. Welfens

domestic or foreign investors; (2) allowing the transfer and sale of property in secondary markets, which is necessary for pricing of industrial assets on the one hand, and, on the other hand, to open the opportunity for assets to change hands in a way that maximizes profits and private real wealth, respectively; and (3) creation of new firms by individuals. The privatization of a single firm in a functioning market economy is different from the case of privatizing a whole economy for at least two reasons: (1) Privatization of many firms and industries has macroeconomic effects on output, employment, the budget deficit, the price level, and the exchange rate. (2) Privatizing a whole economy creates network effects among the newly privatized firms, which can be compared to positive network externalities that are observed in the expansion stage of a telephone network. Assume that government wants to privatize an existing stock of firms (in reality prudent privatization policies would raise the number of firms compared to the initial figure) and that government aims at maximizing short-term revenue from privatization (see Figure 6.2). In this case government takes into account the initial asset demand curve DDo (individual bidders' willingness to pay is a markup over the scrap value). The marginal cost of privatization is assumed to be falling. If the time horizon of government is rather short - for example, due to notorious political instability - it will not take into account the network effects of privatization. Network effects of privatization mean that, with the emergence of privatized and more efficient suppliers, the market value of downstream firms will increase so that the asset demand curve will become steeper in the medium and long term. A short-term-oriented government that wants to maximize short-term proceeds from privatization would thus set privatization prices too high and actually privatize too small a range of firms - compared to the social optimum defined by marginal costs of privatization = marginal benefit of privatization. In principle the privatization agency could capitalize the network effects of privatization. However, in the first stages of transition there might be several noncooperative privatization agencies (e.g., there were two such agencies in Hungary until 1995). Massive and rapid privatization can create positive spillover effects, which mean that firm f s factor productivity is increasing if other firms - with backward and forward linkages - are privatized. This can be interpreted as positive network effects (at least in the network expansion stage), thus implying that the demand for industrial assets is shifting outward over time. If government maximizes privatization net revenues, while facing marginal costs of privatization k', the number of "unit firms" privatized will be qo at asset price po as long as network effects are not taken into account. Assuming falling marginal costs of privatization and positive network effects, in the form of a shift from DDo to DDX the number of units privatized under revenue maximization will be qv Here we assume that creation of financial market institutions (banks, independent central bank, stock exchange) necessary for privatization and valuing industrial assets is characterized by falling marginal costs k' of privatization so that the social optimum is point A, at which marginal social benefits are equal

Privatization, structural change, and productivity

229

P

Qo q,

q2

Z

q

Figure 6.2: Network effects of privatization and optimal privatization.

to marginal social costs. Clearly, the social optimum cannot be attained if privatization is organized within an approach of maximizing the net sales proceeds from privatization. Since average privatization costs k exceed marginal costs, a deficit equal to AB times the number of privatized "unit firms" q2 has to be covered by government as the costs of setting up the market economy. The distance Z minus q2 is that part of the economy that remains state-owned. 6.3

Experience with privatization

Since the ratio of the stock of savings to the estimated value of state-owned firms was low in most ex-CMEA countries, voucher schemes were considered and actually implemented in the former CSFR and Poland as well as in Russia and Ukraine. Voucher privatization is taking place under very distorted and uncertain conditions in goods and factor markets. Privatization strategies that aim at reducing the risk of individual wealth losses require a two-stage approach for the majority of the population, namely ownership in investment funds that can pool risks, which in turn will increase market transparency and help get a market for corporate control started. Lipton and Sachs (1990) emphasized the usefulness of individual ownership in investment funds in the early transformation debate. By contrast, strategies of mass privatization, which effectively assign ownership in individual firms to employees, create large inequalities in wealth and thus very uneven initial wealth endowments. Moreover, selling off un-

230

Paul J. J. Welfens

profitable firms and assets not necessary for production will be delayed, as workers and management in depressed industries have an incentive to lobby for higher subsidies and soft credits from the state-owned banking system or from other firms. This only delays the recombination of industrial assets in accordance with efficiency and profitability criteria. The initial endowment of industrial property rights is crucial not only in terms of ownership structures and wealth effects but also with respect to incentives for investments and savings. Privatization in the Visegrad countries In Eastern Europe, commercialization of state-owned firms is the first step toward marketization and full privatization, which involves a complicated political and economic process (Schipke and Taylor, 1994; Jackson and Bilsen, 1994; Jackson and Biesbrouck, 1995; UNCTAD and Kopint-Datorg, 1994; OECD, 1995). Under commercialization, managers obtain a higher degree of freedom and start to pursue profit maximization goals within a new environment of gradually hardening budget constraints. Capital privatization - that is, a public sale - is the next possible step in privatization. Liquidation of the state-owned enterprise and transfer of property rights to employees or managers is an alternative step that has been particularly popular in Poland and Russia. Bankruptcy liquidation means that the firm goes out of business, so that new firms may be created on the basis of assets of the liquidated firms. Privatization also occurs effectively if government reduces its share in the equity capital of a firm. The private sector's contribution to output quickly increased in Poland, Hungary, the Czech Republic, and - with some qualifications - the Slovak Republic, where parliamentary elections slowed down privatization in 1994. In the Czech Republic, the private sector contributed 62 percent of GDP in 1994, up from 28 percent in 1992. In the Slovak Republic the increase in the period 1992-94 was from 22 percent to 40 percent, while Poland reached 50 percent in 1994. With 70 percent of GDP, Hungary was top of the list in 1994, but the share of employment was only 53 percent, 6 percentage points lower than in Poland. The share of the private sector in GDP was only 23 percent in Bulgaria, 35 percent in Romania, and 25 percent in Russia (see Table 6.3). Poland's progress in privatization of industry was not as impressive as aggregate figures suggest because these figures include the big agricultural sector, which in socialist Poland had largely remained in private hands. Privatization in Poland was slow in the beginning and characterized by a strong emphasis on employee ownership participation schemes. It took government two years to recognize that industry-wide privatization programs would be useful as a means to achieve both private ownership and competition. In industry about one third of the total workforce is in privatized firms, which are employee-owned partnerships that are often considered as not strongly oriented toward profitability and efficiency. While their wage level is comparable to that of employees in joint stock companies, the employee-owned partnerships might lack a consistent long-term business strategy and suffer from in-

Table 6.3. Share ofprivate sector in GDP and employment, 1990-94 (percentage) Employment

Gross Domestic Product 1990

1991

1992

1993

1994

1992

1993

Czech Republic

5a

9a

28

45

62

31

47

Slovakia

5a

22

26

40

17

28

Hungary

16

18

25

65

70

48

53

Poland

31

42

45

48

—

56

59

Romania

—

—

26

32

35

41

44

Bulgaria

—

5

15

19

23

18

28

Russia

—

—

14

21

25

18

28

22

23

Ukraine a

Figure for the CSFR. Source: UNECE (1995), p. 83.

232

Paul J. J. Welfens

sufficient investment funds (Chmiel and Pawlowska, 1995, p. 47). Given the absence of a strategic investor, a clear long-term strategy is difficult to develop. In addition, firms face the burden of installments to be paid for the state assets taken over by them, so that even high profits are insufficient for long-term expansion. This holds all the more since banks are unwilling to provide credits for investment projects of employee partnerships. Until 1994, capital markets were not efficient in Poland because the restructuring and privatization of banks had made limited progress. Within a few years, however, the Polish banking system could be part of a functioning financial market system. The restructuring of banks will be almost complete in 1996 and Poland's ailing banks also are finding foreign partner banks eager to enter the market. Hungary followed a similar procedure of restructuring its banks by looking for foreign banks willing to enter a partnership. Hungary's restructuring of banks was quite costly and not as successful as in the former CSFR. Hungarian privatization at first avoided voucher privatization and emphasized the need for strategic investors and for attracting foreign investment (Szekely and Newbery, 1993). The organizational approach to privatization was changed several times and, except for the field of foreign investors, progress in privatization was slow. Hungary attracted the bulk of the transforming economies' foreign direct investment inflows: 23 percent, well ahead of the Czech Republic with 13 percent, Poland with 12 percent, and Russia with 19 percent (IMF, 1995, p. 79). Hungary was also clearly leading in terms of per capita FDI. High FDI inflows could partly explain Hungary's rapidly rising current account deficit as foreign investors were driving up the value of the Hungarian currency (reducing export growth) and raising Hungary's imports, mainly in the form of capital equipment.3 In 1989-92 foreign currency accounted for 65 percent of the financial funds in privatization; in 1993 for only 34 percent as most interesting firms had already been offered to foreign investors. Domestic currency for acquiring firms and sales for privatization credits accounted for 21 and 11 percent of funds for privatization, respectively, in 1989-92; but to 20 and 29 percent, respectively, in 1993. The keystone of the Hungarian privatization program is a five-year program of public offerings aimed at listing of shares on the Budapest Stock Exchange. Public utilities will represent a large part of the overall public offerings from which some 500,000 original holders of compensation bonds - issued to those with restitution claims - and about 1 million small investors are expected to benefit (OECD, 1995). 3

In the Republic of Korea it took more than a decade until multinational companies were net exporters (Inotai, 1992). Given the high volatility of international capital flows, Hungary's high current account deficit represents a major risk in transition policies. Foreign investors can stimulate the privatization process, in the sense that a high inflow of FDI can raise productivity and export competitiveness such that new jobs are created. As regards the share of the number of announced projects in transforming economies, one third was greenfleld investment and 21 percent acquisition (46 percent joint venture), so that foreign investors contributed to capital formation and modernization of the capital stock (IMF, 1995, p. 80).

Privatization, structural change, and productivity

233

The most innovative privatization scheme was the voucher privatization applied in the CSFR, where about ten major investment funds (Investment Privatization Funds) - many controlled by banks and insurance companies - attracted about 50 percent of all shares emerging from the first round of coupon privatization. In this voucher privatization strategy, firms to be privatized were commercialized at first, and then transformed into state-owned joint stock companies. Then a proportion of shares was set aside for voucher privatization (on average 64 percent of basic capital was allocated via voucher privatization). Government selected some 2300 large companies for the first wave of voucher privatization, while the final number selected reached 1491. In the first wave, two thirds of assets were privatized via vouchers. Two-stage voucher privatization helped to pool risks and to establish a secondary market, but it is doubtful that economic growth in the Czech Republic can largely be explained by the success of voucher privatization (Dlouhy andMladek, 1994). Restitution, small privatizations, and large privatization by sale also played a role in the Czech Republic's privatization process. Investment funds cannot hold more than 20 percent in an individual firm, thus restricting opportunities to get a strategic investor involved in restructuring a company. However, the Czech Republic and the Slovak Republic enjoy the advantage that they have created a functioning capital market. Moreover, the restructuring and recapitalization of banks via a centralized strategy has been successful, so that neither weak banks nor excessive interenterprise credits are an impediment to functioning capital markets in the ex-CSFR. Voucher privatization could create problems of dispersed ownership, but it remains to be seen whether aggressive investors will not accumulate growing proportions of equity stock in major firms in the long run. Privatization in other ex-CMEA countries Bulgaria was slow in privatizing the economy. The private economy accounted for about one fourth of national output in 1994; in industry the share of private firms was about one fifth only. Instead of the planned privatization of 308 firms, only 171 firms could be privatized in 1994. A voucher scheme in combination with mass privatization and some big privatizations in connection with a strategic investor were brought under way in 1994. In Romania it took until 1995 to introduce a competition law and bankruptcy procedures and to open a stock exchange. Some 3000 firms are to be privatized by February 1996. Since 1989 foreign direct investment inflows of $1.3 billion occurred. In 1995 Ukraine had privatized more than 11,500 firms through voucher sales, auctions, or other procedures. Some 10 million of the 52 million citizens have picked up vouchers or are otherwise active in the privatization program. Shares in the first batch of already privatized firms will be put up for sale so that there is a secondary market for industrial property titles in which domestic and foreign investors can participate. There are some 8000 large and medium-scale

234

Paul J. J. Welfens

enterprises up for privatization in 1995, of which only a dozen are earmarked for strategic foreign investors. Russia had 46,000 privatized firms in 1992, of which only 336 were big companies. By late 1994 there were more than 109,000 privatized companies, which represent some 80 percent of all firms to be privatized. The number of small enterprises increased strongly from 450,000 to more than 1 million. However, major parts of the Russian economy are not on the privatization list: energy, telecoms, and the military sector. One third of all industrial assets had been privatized by January 1, 1994; this includes nearly 50 percent of manufacturing industry and 35 percent of construction firms. Voucher privatization in combination with transforming SOEs into stock companies played a major role: of the 146 million vouchers issued, about 60 million were allocated by the citizens to investment funds. About 13,000 firms were involved in this voucher privatization, which assigned 19 percent of nominal equity capital of these firms to private owners. Among the alternative variants for privatization, two thirds of all cases were characterized by selling 51 percent of stocks to the employees, so that insider privatization dominated. Performance of privatized firms was not excellent, but the fall in output was 10 percent less than in SOEs (Sigmund and Werner, 1995). Poor performance of this insider privatization has resulted in changes of the rule for future privatization: employees cannot obtain more than 25 percent of the equity capital via vouchers and not more than an additional 10 percent at a discount price. More emphasis will be on strategic investors important for effective ownership control and productivity growth (blocks of 15-20 percent of equity capital will be for sale). Experiences with privatization and productivity Poland took the lead in labor productivity growth and economic growth in the transforming economies in 1994-95, when economic growth reached 5 to 6 percent. Labor productivity growth also reached high figures in Hungary and the Czech Republic. While the former CSFR has favorable growth prospects in 1995, Hungary's growth rate is expected to be modest at best. Since double digit productivity growth rates can be expected in most ex-CMEA countries in the medium term, high growth rates would be required to prevent unemployment from rising. Unemployment is expected to fall only in Poland and the former GDR in 1995. In the Czech Republic the unemployment rate is rather low, but in all other transforming economies - except for Slovenia - mass unemployment is an economic and political problem. Poland's stabilization was the most radical among the transforming economies in 1990-91, although the privatization program could not develop momentum until 1995, when mass privatization was finally adopted in parliament. In Poland privatization has made gradual progress since 1990. A recent World Bank study (World Bank, 1994) confirms the view that productivity and profitability are higher in private firms than in state-owned enterprises (SOEs).

Privatization, structural change, and productivity

235

Employment in the ex-CMEA countries fell strongly in the first transition stage, when output was sharply reduced. However, the decline in employment in many countries was smaller than the decline in output in the period 1990-93. This implies "excess employment" as a hidden form of unemployment and suggests that there is a high degree of excess employment in most transforming countries. Only Hungary, Poland, and Slovenia recorded little excess employment (Table 6.4). In the faster transforming countries, excess employment started to fall with the acceleration of privatization after 1991, while in other transforming countries excess employment continued to rise. The strong position of labor unions and mismanagement, as well as incentives for workers to remain on the payroll to maintain non-wage benefits tied to the enterprise (social provisions tied to the company are still typical in Russia), could explain excess employment. The systemic trait of socialist overemployment - estimated to reach 20 to 30 percent in the 1980s in socialist CMEA countries - was preserved in many transforming countries and indeed has been reinforced in some countries. This points to a strong potential for raising productivity on the one hand, and on the other hand to a looming mass unemployment problem. Interestingly, employment in the Polish private sector increased after 1992 (Chmiel and Pawlowska, 1995). The age of the capital stock in privatized and newly created private firms was lower than in state-owned firms (Habuda and Jennewein, 1995). Rising labor productivity as a consequence of reducing overemployment was stimulated in Poland, Hungary, and the CSFR by both privatization and a reduction of subsidies. By contrast subsidies as a percentage of GDP had increased in Romania and Russia (and probably in Ukraine) in the period 1989-93 (EBRD, 1994). Since international trade taxes generally were reduced in Eastern Europe or fixed at relatively low levels in the first transition stage, there is high pressure in the tradables sector to strive for efficiency and productivity gains. One caveat is that Russia and Ukraine still apply quantitative restrictions in external relations. Development of capital productivity Capital productivity is related to management, to incorporated technological progress, and to gross investment. Part of technology is capital-embodied, so that a poor selection of investment projects in rather inefficient capital markets also will bring about a low rate of technological progress. Disembodied technological progress is mainly the result of R&D efforts by skilled personnel, so that appropriate R&D activities have to be organized both in the private sector and in the government sector. For OECD countries, Wolff (1995) found four equally important sources for total factor productivity growth (2.3 percent per year). International trade, research and development, catch-up effect, and structural change each contributed 0.5 percentage point to average annual productivity growth. In 1973-93 the impact of catching up and structural change fell.

Table 6.4. Excess employment in transforming countries, 1990-94 (difference between change in GDP and change in employment in percentage points) 1990

1991

1992

1993

1990-93°

1994"

Russia

1.6

10.8

17.2

10.2

33.2

13.0

Ukraine

2.0

10.4

9.7

11.9

28.4

17.0

Romania

7.2

13.2

12.4

-4.8

24.2

—

Czech R.

0.3

8.7

3.8

-2.1

10.0

-5.0 -5.0

Slovakia

-0.1

4.2

-0.5

3.7

6.2

Bulgaria

3.0

-1.3

-2.4

2.2

1.0

—

Poland

7.6

1.1

-6.8

-4.4

-1.6

-6.0

Hungary

0.2

2.3

-4.3

-2.7

-3.5

-2.0

Slovenia

0.8

0.3

-1.2

-4.3

-3.6

-7.0

Note: Countries ranked according to the total excess employment over 1990-93. " Cumulative annual changes. b Estimate based on the first three quarters. Source: UNECE (1995), p. 109.

Privatization, structural change, and productivity

237

Vis-a-vis Western Europe, transforming Eastern Europe is in a situation that is similar to OECD-Europe in the 1960s vis-a-vis North America. An Eastern European high-growth strategy could be built upon high investment/output ratios, rising trade, and R&D as well as structural change and catch-up effects from competition. Ex-CMEA countries are rebuilding their capital stocks. Capital deepening often goes along with employing more productive, but also more specialized, types of equipment. The firm, facing higher sunk costs and reduced ability to cover a broad range of production activities itself, will accelerate outsourcing. Thereby one may expect that capital deepening will create new opportunities for private business (especially services) in the region where the firm is located. Capital productivity was very poor in socialist countries, as suggested by the coincidence of very high investment/output ratios and low growth rates of consumption as in the late 1970s and the 1980s. Economic opening up has rendered part of capital obsolete, as have sharp relative price shifts in many areas, which reflect both international influences and the new role of consumer preferences in markets. Raising output via internal reorganization of the production process will improve both labor productivity and capital productivity. This reorganization is induced by privatization. In capital-intensive industries, the presence of foreign investors could have a particularly positive impact. Since subsidies, which typically favored capital-intensive production, were reduced relative to GDP in transforming economies, average capital productivity will increase in the course of contraction of unprofitable capital-intensive sectors. Since import tariffs have remained low in most transition countries, capital-intensive industries in which Eastern European countries are expected to have a comparative disadvantage were not sheltered via protectionism. Economic viability of capital-intensive industry has, however, been artificially restored by hyperinflation in several countries, because the resulting depressed real wages maintained the profitability of old capital vintages. The fact that excess employment is still widespread in transforming countries means that capital productivity is higher than otherwise. Since real interest rates had turned positive in the Visegrad countries by 1993, one may expect that investment projects realized since then will have a positive marginal product of capital. The collapse of the ex-CMEA and the bankruptcy of many firms in Central and Eastern Europe (including the GDR) implies gradually increasing problems with spare parts, which reduces average capital productivity and stimulates new investment. Investment/output ratios fell in the transforming countries to about 20 percent, which for the Visegrad countries with positive and rising growth rates suggests that the marginal product of capital has increased. Despite slow growth, Hungary is not necessarily an exception, since stagnation in 1995-96 is mainly caused by a decline of demand. In the medium term, a high investment/output ratio in transforming economies could rejuvenate the capital stock and thus realize productivity gains

238

Paul J. J. Welfens

along the vintage approach. A high investment/output ratio would also be required for faster capital deepening, which in turn raises labor productivity. Energy productivity and infrastructure Improving the efficiency of the use of natural resources is another element of productivity growth. Energy consumption per unit of GDP was reduced in Poland after 1992 (Herer and Sadowski, 1995, p. 31). Given the fact that energy per unit of GDP is still much higher in Eastern Europe, there should be big opportunities to raise energy productivity. Energy intensity of GDP in Central and Eastern Europe (Bulgaria, CSFR, Hungary, GDR, Poland and Romania) was more than three times the EC-12 figure in the late 1980s and water intensity as well as pollution intensity about twice as high (Zylicz, 1994, p. 410). High losses in electricity transmission in Eastern Europe point to opportunities for improving the productivity of the energy sector (see World Bank, 1994). In the transforming economies, energy prices that were below world market prices in the CMEA have strongly increased. This energy price shock is similar to that suffered by OECD countries in the 1970s, when energy-intensive products and processes became obsolete overnight. In the 1970s the OECD countries could not restore profitability of these industries by quick reinvestment in more energy-efficient machinery. The situation in Eastern Europe is different since machinery and equipment that have to be largely imported anyway typically embody energy-saving technical progress. Saving energy will result in a higher capital intensity provided that energy and capital are substitutes. Privatization of the energy sector is likely to raise factor productivity in the energy sector and thereby to reduce real energy prices; this, however, would not stimulate the saving of energy and a rise of energy productivity. Consequently, privatization of the energy sector should be coupled with a rise of energy prices via higher energy taxes - possibly reflecting the negative external effects of energy consumption. Restructuring and privatizing the energy sector will be an enormous challenge in the 1990s. Hungary has decided to allow private ownership in the electricity sector as well as in the telecoms sector, and the Czech Republic and Poland are likely to adopt a similar course. Privatization in the electricity sector in Hungary ran into the particular problem that station managements favored an overestimation of the value of capital employed in order to obtain high capital allowances, which will facilitate hiding of profits for the sake of the management and the employees. Applying the experiences of British electricity privatization (Newbery, 1994; Welfens and Yarrow, 1995), the Eastern European economies could expect to raise energy efficiency strongly over the long term. Efficiency means generation and transmission of energy at least costs, that is, whether power stations are run in correct order and run efficiently. In the long run the required amount of investment should be undertaken at least cost at the most favorable

Privatization, structural change, and productivity

239

places (minimizing transmission losses). Integration of power generation with West Europe's system could reduce the amount of investment required. Infrastructure investment and productivity In a market economy, each firm will use as intermediate inputs information and transportation; hence the infrastructure will play an important role for output. The lower information and transportation costs are, and the greater the choice among alternative suppliers of inputs, the broader the opportunities for cooperation and more intense competition will be. Infrastructure expenditures are difficult to increase because of high deficit-GDP figures in most ex-CMEA countries. The alternative to government financing of infrastructure is to privatize the capital-intensive infrastructure - telecommunications, railways, airports, and the energy sector - or at least to reduce government stakes toward a minority holding. Even major highway projects could be organized within the private sector provided that toll roads are accepted politically. Eastern Europe's infrastructure endowment is poor relative to Western Europe. Compared to Germany, road density in Hungary and Poland is one third, in the Czech Republic half, in Russia one tenth. There is also backwardness in air transportation, but not in railway density. The East-West gap is very big in telecommunications, which is crucial for communication, organization of firms, the growth of information-intensive industries (especially services), and competition in a spatial context. Massive investments are necessary to improve the quality of the network and to raise telecom penetration rates. One important aspect of infrastructure bottlenecks is that they impair market entry and thus reduce competition directly. Another indirect negative effect results from the smaller overlap of regional markets caused by high transportation and information costs. 6.4

Toward new economic structures

Productivity growth requires thorough structural change, in the sense that productivity laggards should shrink, while sectors with big opportunities for productivity improvements should expand. Within six years, labor productivity increased dramatically in the former GDR, namely from 29 percent of the West German figure to some 52 percent in 1995. However, producer prices in the GDR, which fell after unification by some 40 percent, are increasing only gradually. Improvements in the pricing position of privatized East German firms are mainly due to a high rate of product innovation, which exceeds that of West Germany. In 1993 the share of new products in East German firms' sales reached 46 percent, which is 16 percent higher than the West German figure. Interestingly, medium-sized firms with fewer than 500 employees were more innovative than big firms; small firms with fewer than 20 employees were leading in the field of high-technology innovations (Stifterverband, 1995).

240

Paul J. J. Welfens Structural adjustment and transformation

Socialist CMEA countries were characterized by overindustrialization and except for the CSFR - by a rather unproductive agricultural sector with a high share of employment. Services were much underrepresented in comparison to even poor EU countries or NICs. Table 6.5 shows that industry in Eastern European economies also was overspecialized in comparison to EC countries and NICs. The degree of specialization in EC countries increased in the 1980s in most EC member countries while at the same time the share of industry fell. The intensity of structural change was relatively high in Portugal, Spain, and Greece relative to the intensity in other EC member countries in 1990. Among ex-CMEA countries we find a relatively high intensity of structural change only in Hungary and Poland in 1990. In all other ex-CMEA countries, the figures for the indicator for structural change were lower than in Spain, Portugal, and Greece (and Ireland) in 1980 and 1990, and the indicators were also lower than in Taiwan and Korea. This all points to postponed structural adjustment and weak adjustment capability in the ex-CMEA countries, except for Hungary and Poland. The relatively strong westward trade orientation of these two countries and the beginning of internal liberalization already in the endgame stage of socialism - in the 1980s - could explain why one finds here a higher degree of structural change and less specialization than in other CMEA countries. Given the low share of private industry in the Visegrad countries and the enormous transformation problems, a potential future EC entry of Hungary, Poland, and the ex-CSFR would require long transition periods for full EC membership. The set of rules in the 1990s is also more complex and on average incorporates a higher degree of liberalization than the EU rules of the game did at the time of the southern enlargement. Transfers and technical support will in no case be as comprehensive as in the case of eastern Germany's integration into the EC, when western Germany massively helped to modernize the ex-GDR. Structural change is coupled with economic growth when sectors with high potential productivity growth are expanding faster than those with slow productivity growth. Structural change is therefore a key ingredient for growth and productivity change. The biggest productivity increases can be expected in those areas where high innovation dynamics and strong demand growth - facilitating the exploitation of scale economies - are coinciding. The share of agriculture has fallen in Eastern Europe and plays an important role in the ex-CMEA countries only in Romania (Table 6.6). In the Visegrad countries the share of industry in GDP has fallen by several percentage points in the period 1990-94. In Poland, the economic upswing of 1993 brought an expansion of the share of industry by one percentage point - measured at constant prices. The share of labor-intensive industry has not increased in the Visegrad countries, much in contrast to the former GDR (Welfens, 1995b), where massive infrastructure projects also stimulated an expansion of the con-

Table 6.5. Degree of specialization and intensity of structural change Share of industry in GDP (%)

Degree of specialization'

Intensity of structural change*

[1985]

(1989)

1980

[1985]

1990

[30.4] [16.3] [26.4]

26.7 15.3 25.8 25.6 28.4 23.3 15.8 20.2 21.3 23.2 20.4 31.1

11.2 10.9 8.4 14.9 37.4 12.5 14.4 15.4 10.4 10.1 11.1 12.1

[10.4] [11.8] [8.5]

6.5 12.1 9.8 21.0 24.3 13.8 14.6 15.2 11.4 10.8 12.1 15.4

5.3 3.6 5.5 4.5 3.1 2.6 3.0 2.8 3.7 2.8 3.8 2.7

East Germany Hungary Poland Czechoslovakia Romania Bulgaria USSR

48.2 24.5 39.1 47.9 — 58.1 37.5

13.2 9.0 11.7 15.9 20.5 11.4 18.1

14.4 8.9 10.6 15.7 20.8 13.4 19.6

2.0 5.3 6.0 3.0 2.0 2.9 1.8

1.6 6.8 9.3 3.1 1.5 3.1 1.0

Mexico Taiwan Korea

24.6 35.6 31.3

8.9 11.0 9.1

10.3 11.2 10.3

3.6 5.1 7.6

2.7 3.7 3.5

Portugal Greece Spain Ireland Luxembourg Belgium Denmark Netherlands France Italy United Kingdom West Germany

a

h=l00[\+(Ls *\nsAnJ];*m.x == ln(number of branches), 0 <, h 100 <.

Source: EC Commission, UNIDO.

"cos 8 =

1980

[1985] [7.0] [4.4] [3.4]

*X0 *

i

1990 5.2 4.1 4.9 4.1 5.8 2.2 2.2 2.3 2.6 2.1 2.0 1.9

= share of i-branch in total MVA

243

242

Table 6.6. Share of major sectors in Eastern Europe's output, 1990-94 (GDP = 100%) (figures in brackets are based on constant prices of 1992 for Bulgaria, Romania; 1990 for Poland) 1990

1991

1992

1993

7 50 8 32

6 55 6 31

6 45 4 45

6 40 5 50

SLOVAKIA Agriculture Industry Construction Services

7 50 9 30

6 53 7 28

HUNGARY Agriculture Industry Construction Services

13 27 6 55

9 29 6 54

7 28 6 59

6 26 6 62

POLAND Agriculture Industry Construction Services

8(8) 45(44) 9(9) 36(37)

7(9) 40(39) 10(11) 41(39)

7(8) 34(37) 7(10) 48(40)

7(8) 33(38) 6(10) 53(41)

ROMANIA Agriculture Industry Construction Services

24(21) 44(42) 6(5) 26(32)

20(21) 40(41) 5(4) 35(33)

20(20) 41(41) 4(4) 35(35)

23(22) 39(42) 5(5) 34(31)

BULGARIA Agriculture Industry Construction Services

18(10) 43(39) 7(7) 30(39)

14(12) 39(38) 4(4) 41(43)

11(11) 37(37) 5(5) 44(44)

9(10) 35(39) 5(5) 47(44)

9

RUSSIA Agriculture Industry Construction Services

17 38 10 36

13 40 10 37

10" 49° 8° 33"

9 37 8 46

7 30 9 54

CZECH REPUBLIC Agriculture Industry Construction Services

" Break in series. Source: UNECE (1995), p. 79.

1994

244

PaulJ.J.Welfens

struction industry. The expansion of construction could help to create new employment and therefore would allow proceeding faster with restructuring of industry and rationalization investment, where there is a political threshold for tolerable unemployment rates. The share of the service industry increased markedly in Eastern Europe, but measured in current prices this change was reinforced by rising relative prices of services. While the expansion of the long-neglected service industry is impressive, one should not overlook that a viable services industry will largely have to rely on orders from the manufacturing industry. Without profitable manufacturing industry, the opportunities for firms and entrepreneurs in the services industry will remain quite limited. The employment share of the service industry has increased in all major transforming countries in the period 1989-93 (Table 6.7). Employment shares in agriculture reach more than 20 percent in Bulgaria, Poland, and Romania, where resistance against liberalization could come from an influential political clientele in the countryside. A high share of agricultural employment could cause high subsidies for agriculture and thereby reinforce the call for subsidization and protectionism in the whole economy. High-productivity growth in agriculture would occur if these transforming countries thoroughly modernize this sector and switch toward higher capital intensities. This would cause even higher unemployment rates, which in turn could reinforce resistance against rationalization investment and structural change. Reindustrialization in the future? It is unclear which type of structural change is optimal for Eastern Europe. Simply to strive for the growth-maximizing sectoral structure might not be feasible, because this could imply high unemployment, which would render transition and economic convergence not sustainable for political reasons. An interesting question with respect to structural change is which structural patterns would have characterized the CMEA countries if they had been market economies with comparable per capita income figures? A relatively recent approach of Dohrn and Heilemann (1992; 1993) follows Chenery (1960) and Chenery and Taylor (1968). According to the Chenery approach, growth is explained by several country-specific factors on the one hand, and, on the other hand, by universal factors that are used to explain sectoral growth functions. While neoclassical models focus on input factors to explain aggregate growth, Chenery's approach is a mixture of supply-side aspects and demandside elements. Value added per capita in sector / of country j is explained in terms of per capita income - representing demand side influences as described by Engel's law (but possibly also the technology level) and the population of the country, which is a proxy for the size of the economy. Dohrn and Heilemann (1993) extend this approach to test for the impact of the economy's endowment of natural resources; this influence is indirectly captured by per capita net ex-

Table 6.7. Structure of employment by sectors, 1989-93 Share in total employment Industry

Agriculture

Services

1989

1993

1989

1993

1989

1993

Bulgaria

18.6

22.1

45.3

36.6

36.6

41.3

Czech Republic

10.6

6.9

49.2

44.6

40.2

48.6

Hungary

15.5

9.1

37.8

33.8

46.7

57.1

Poland

26.8

25.8

36.8

31.6

36.4

42.6

Romania

27.9

35.9

45.1

35.8

27.0

28.3

Slovakia

13.8

12.1

46.3

39.8

39.9

48.1

Russian Federation

13.5

14.6

42.8

39.0

43.7

46.3

Source: UNECE (1995).

246

Paul J. J. Welfens

Table 6.8. Sectoral structure in the Visegrad countries (percentage of GDP) Agriculture

Manufacturing

Services

6.3 13.1 14.4

57.5 52.4 38.4

36.2 36.4 32.7

9 11 12

30 30 29

44 41 43

6 8 8

32 33 33

45 43 44

1988, actual figures CSFR Poland Hungary 1988, hypothetical figures CSFR Poland Hungary 1998, projected figures CSFR Poland Hungary Source: Dfihrn and Heilemann (1992), p. 61.

ports of primary goods. Moreover, the share of gross capital formation in GDP is taken into account to test for the impact of investment. Within this modified approach, which is modeled on the basis of flexible elasticities, the results from cross-country regressions are as follows for the Visegrad countries: The actual sectoral structure was biased in favor of manufacturing and - except for the CSFR - agriculture. Therefore, transition will entail a reduction of industry and agriculture at first. If market economy patterns prevail, economic growth will lead to a renewed expansion of industry in the late 1990s (Table 6.8). This result is interesting since it suggests that export pressure on the EU could increase in the future, since ^industrialization and specialization within exports will coincide. Dohrn and Heilemann (1992) found no significant impact of the scale variable, possibly because the spreading of free trade under the GATT umbrella reduces the importance of the size of the domestic market (one may, however, suspect that this is not true for technology-intensive goods, where reliance on a big profitable domestic market could be quite important). 6.5

Convergence issues

High-productivity growth can be achieved in Eastern Europe only if privatization is combined with structural change and competition. Privatization without competition cannot achieve major efficiency gains (Vickers and Yarrow, 1988). This is amply born out by the negative results of insider privatization without competition in Russia (Welfens, 1994). Competition is necessary to achieve both static and dynamic efficiency. Open markets and hence reduction of barriers to entry and to exit are required for competition. Barriers to exit are equivalent to barriers to entry, so that postponed liquidation and bankruptcy distorts

Privatization, structural change, and productivity

247

the competition process at the expense of potential newcomers. It took several years for ex-CMEA countries to adopt bankruptcy legislation; in 1995, Romania was the last ex-CMEA country to enact such legislation. Bankruptcy problems are compounded in post-socialist economies in many cases because firms developed in the first transition stage a net of ever-growing intercompany indebtedness. Such developments indeed should be avoided, because proper supervision of investment projects through the capital market otherwise cannot be sustained. Furthermore, continuing soft budget constraints for firms imply that relative price signals will play only a reduced role for resource allocation, so that efficiency gains and growth are diminished. Catching up Economic catching up requires that sectors with high potential productivity expand. Sectors with slow productivity growth will not necessarily have to shrink. Expansion of high-productivity growth sectors could imply a high and politically critical rate of unemployment if there are not also industries in which productivity growth is slow. For decades the services industry in OECD countries provided many new jobs but recorded below-average productivity growth. Resistance against productivity augmenting rationalization investment will be prohibitive if unemployment cannot be kept within tolerable limits. It is fairly clear that in an open transforming economy, modernization should be foremost in the tradables sector. Eastern Europe can catch up with Western Europe if privatization is comprehensive and linked to competition, together with investment in infrastructure and human capital. The example of Korea is illustrative for successful catching up. The major supply-side sources of economic growth in Korea during 1963-73 and 1973-86 - with growth rates of 9.5 and 7.8, respectively - were the following (Song, 1990, p. 89): Contribution of inputs (capital and labor) 5.4 and 4.1 percent, while total factor productivity growth reached 4.1 and 3.8 percent, respectively. Advances in knowledge and economies of scale were particularly important for total factor productivity growth. Improved resource allocation contributed less than one percentage point. Improving resource allocation can raise productivity and growth in the medium term considerably in Eastern Europe, since the initial situation was quite distorted. However, it seems clear that capital formation, advances in knowledge, and scale economies will be the sources of sustained economic growth in Eastern Europe. Rising exports to a nonprotectionist EU could be an important ingredient in this strategy. In the 1960s, Western European economies were catching up with the advanced United States (Darwick and Nguyen, 1989). They successfully exploited the advantage of backward countries, whose firms can follow proven technological and marketing paths set by firms from the leading countries. Imitating technologies is also often rather cheap for firms from the follower countries. In addition to this benefit, Western Europe enjoyed the advantages of increasing-

248

Paul J. J. Welfens

ly free trade and capital flows. Leading countries will find it more difficult to realize rapid growth and productivity increases, because firms at the technology frontier will have to try risky innovative solutions in a process of shifting the frontier outward by R&D. Increasing competition in the EU single market reinforces the pressure to relocate production within Europe. Opening up in Eastern Europe comes just at the right time, in the sense that EU firms reconsidering their choice of major production locations in Europe are increasingly taking into account Eastern Europe, which could attract high FDI inflows in the future. Western European integration brought some economic convergence, as measured by per capita GDP in the EU. This convergence was brought about both by market forces, such as considerable regional mobility at the margin - due to foreign workers from third countries moving into excess demand areas - and rising income transfers at the supranational level. Under the heading of cohesion in the Maastricht Treaty, economic convergence in Western Europe has become a goal for the EU. In a European East-West context, convergence might be a goal for two reasons: (1) a big East-West income gap would imply a major financial burden for the EU in the case of EU widening toward ex-CMEA countries; (2) a sustained East-West gap could mobilize rising East-West migration pressure, which could destabilize the receiving countries in a period of rising nationalism and high unemployment, but which could also undermine economic growth in transforming economies, which would lose young skilled and entrepreneurial people as a consequence of massive outmigration. The biggest problem for privatization in Eastern Europe is that the initial structure of many industries is characterized by monopoly or tight oligopoly, so that simply privatizing existing big firms is itself insufficient to create private business competition. Without competition policy, private monopolies will lead to all sorts of inefficiencies, which in the tradables sector could only be somewhat remedied by progressively freer import competition. The nontradables sector in any case requires dismemberment of big firms if economic efficiency is to be achieved. Backwardness: opportunities and problems in an open economy The CMEA's economic backwardness implies enormous opportunities to move up the technological ladder by imitation and innovation. COCOM restrictions have largely been abolished, so that the potential for tapping world market technologies by importing advanced goods and intermediate inputs has increased. Intensified competition should help to accelerate the diffusion process, so that foreign technological progress can be adopted quickly and domestic innovations applied on a broader scale. This will result in higher factor productivity. If quality of products can be improved and if product innovations are generated, prices fetched in domestic and international markets should increase. Value-added productivity is thus improved.

Privatization, structural change, and productivity

249

A major problem of backwardness is that leapfrogs in technology are possible. For example, instead of a smooth increase in capital intensity, there can be a switch to a new technology with much higher capital intensities. If capital equipment is mainly imported, modernization of the consumer goods industry - with many jobs lost - will not be offset by expansion of the machinery industry, which uses mainly skilled and unskilled labor. Traditional trade theory suggests that the move from autarky to an open economy will stimulate those sectors that use the relatively abundant production factor. In Eastern Europe, this factor is unskilled and skilled human labor. According to the factor equalization theorem, relative and absolute factor rewards will be equalized by free trade. According to this neoclassical Heckscher-Ohlin approach, the income share of the relatively abundant factor will increase, so that labor income should increase relatively in Eastern Europe. This result, however, is doubtful in Eastern Europe. The high level of unemployment - in contrast to the full employment assumption of the Heckscher-Ohlin model - could prevent an expansion of labor- intensive production from raising the ratio of wages to the profit rate. If income growth benefits only a small part of the population in the initial transition stage, high and sustained unemployment rates could become a major impediment to productivity increases. Workers will resist restructuring and rationalization investment if unemployment figures suggest that there is a high risk of long-term unemployment. Except for the case of poorly organized rural labor, one may expect that sustained unemployment will become a barrier for high-productivity growth. For the neoclassical world, the Rybczinski theorem states that an exogenous increase (decrease) in factor endowment will - given output prices - stimulate (decrease) the output of those goods in which the increasingly available factor is used intensively, while the output of other goods will fall. Systemic transition rendered a considerable part of real capital obsolete in Eastern Europe, so that the transition shock could be in favor of a relative expansion of laborintensive industries. This does not rule out a sharp aggregate contraction of industry, as witnessed in transition economies. In the medium term, capital intensity could increase, since privatization brings about layoffs and modernization of machinery and equipment. Modernization of the capital equipment will bring about a positive vintage effect, which will help to increase labor productivity. With an increasing inflow of foreign direct investment, one may expect a rise of capital intensity in the tradables sector. This suggests that capital-intensive production will expand in the medium term. Capital modernization will allow upgrading the product assortment, thereby to fetch higher prices in world markets. Value-added productivity will increase, so that the terms of trade should improve. With respect to financing infrastructure investment, countries with high budget deficits (e.g., Hungary, Russia) face major disadvantages. In a regional context, there will be regional spillovers from infrastructure investment. Therefore

250

Paul J. J. Welfens

Poland and the Czech Republic will particularly gain from infrastructure investment in the former GDR. Convergence and income distribution Even in the absence of international factor mobility, an equalization of absolute factor rewards can occur, as is shown in the Samuelson-Stolper theorem, which argues that the relatively abundant factor will improve its income share in a post-autarky world. Free trade can substitute for international labor migration and lead to a convergence of relative factor prices, which in effect also implies equalization of absolute factor rewards. In the neoclassical world, countries will specialize in accordance with relative factor abundance, which is identified with capital intensities in a two-goods model with labor and capital as input factors. According to the Heckscher-Ohlin approach, competition in goods and factor markets - to be established by privatization cum competition in Eastern Europe - will stimulate the production (and export) of labor-intensive production in countries with a relatively low capital intensity. The ratio of wage income to capital income will improve, since the relatively abundant factor will gain after giving up autarky. Economic opening up and privatization in Eastern Europe provide an interesting test for the Heckscher-Ohlin-Samuelson-Stolper (HOSS) approach, and one may anticipate that the test results will only partly corroborate the HOSS model. One major problem is that, in the real world, one must distinguish between the tradables and the nontradables sectors (see Figure 6.3). The aggregate domestic capital intensity k could be smaller than foreignfc*(starred variables refer to abroad) despite the fact that capital intensity in the tradables sector A;r exceeds that in the foreign tradables sector kT*. k < k* will hold if the capital intensity in the nontradables sector kN is rather low, that is, is sufficiently smaller than in the tradable sector. It also will hold if kT < kT* and kN exceeds kN* (and kT). After privatization and economic opening up, the wage income/profit ratio could fall despite an increase of the wage/real interest rate ratio if private owners decide to relocate capital and labor in favor of one sector to such an extent that (kT/L2) + {kN/L^) is sufficiently increasing to offset the rising wage/interest rate ratio w/r.4 The relation k > &* normally implies that w/r > w*/r*, so that economic opening up and competition cum privatization should stimulate the expansion of the capital-intensive sector(s), but this is in contrast to the possible constellation that kT < kT*', if the tradables sector dominates overall wage behavior in 4

In a two-sector analysis with sector 1 and 2 the wage income/profit income ratio is given by v = [wLl + wL2]/(rKl + rK2)\ dividing by L,L2 yields r[(kJL2) + (k2ILx)}v = H>L/(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'
N

(c) 1

k > k'

N

'> k *

k>k*

1

N

k
k>k*

(d)

(e) 1

k'> k *

N

T

N

k
k<^k*

k'>kT-

k>k*

(i) 1

N

k'< k *

k >k' > ^ *

k<

kN
k'
k<

Figure 6.3: Paradoxes of capital intensity and structural change in an open economy.

the economy w/r < w*/r*, so that labor-intensive products will be exported.5 We disregard here problems of political risk that will affect r and r*. If the small ex-CMEA countries mainly expand labor-intensive tradables industries, the real wage rate will increase and the labor-income/capital-income ratio should increase. Thus results in accordance with Heckscher-Ohlin theory will hold only if kT < kT* and if the nontradables sector's wage rates are not dominating those in the tradables sector. By contrast, Russia's exports could be mainly in shipbuilding, energy, and oil products - industries in which capital intensities typically are high. In Russia, the capital-income/laborincome ratio could increase in the course of privatization and economic opening up, such that income differentials between capital owners and workers could strongly increase. This could lead to destabilizing social conflicts with negative destabilizing spillovers into the smaller ex-CMEA countries; these countries would face major politico-economic problems if outmigration from Russia and Ukraine occurs. not lead to With respect to the Leontief paradox, one has to check whether ^ U S A < ^ R O w w/rUSA< (w/r) R O W so that the United States would have an advantage in exporting laborintensive products despite k, y ^ A > ^ R Q W ^ ls n o t implausible that the relatively efficient U.S. nontradables sector dominated the wage/interest ratio in the overall United States, which is a relatively closed economy. By contrast in the UK, the Netherlands, and Germany, the degree of openness was much higher, so that the tradables sector largely dominated the overall wage/interest ratio in Europe. This would resolve the Leontief paradox.

252

Paul J. J. Welfens

FDI inflows into Eastern Europe - and outmigration - could raise capital intensities. Countries that can attract relatively large FDI inflows will record rising shares of net exports, in skill-intensive and capital-intensive industries. Hungary and the Czech Republic already show in their 1990-93 trade pattern with Germany developments that are consistent with this view (Klodt, 1993). Unemployment in both sectors in Eastern Europe represents, of course, a serious deviation from the HOSS world; and this deviation could indeed restrict the rise of the wage/interest rate ratio very much. While non-neoclassical aspects such as economies of scale and product variety have been treated analytically with respect to income distribution (Helpman and Krugman, 1985), the consequences of unemployment on income distribution in the course of economic opening up have not been analyzed. This, however, is crucial for Eastern Europe as well as for developing countries, because the point of departure in economic opening up typically is open or disguised unemployment. If distribution effects of economic opening up are adverse for labor in a situation of initial unemployment, one might indeed consider privatization schemes with specific employee ownership programs that make sure that capital gains associated with economic opening up will compensate workers for the risk of being laid off and for unemployment. Otherwise opening up and liberalizing externally will fail for lack of political support, and the whole transformation program might turn out to be unsustainable. Foreign direct investment: ambiguous role for convergence Even if trade and international direct investment flows should bring about equalization of factor rewards, there are three reasons for long-term divergences in per capita income. A first obvious reason is that in industries with economies of scale, gains from trade will differ in favor of those countries that quickly specialize in the production of goods with relatively higher income elasticities. If manufacturing airplanes allows appropriation of higher benefits from static and dynamic scale economies than producing locomotives, the country that is relatively fast in expanding the aircraft industry will enjoy an absolute income advantage, which implies - with similar population figures - a relative income lead. This could prevent convergence. Dynamic scale economies tend to be highest in technology-intensive industries, which will primarily be located where R&D personnel are available and where an advanced machinery industry is capable of developing (in touch with the industry using the equipment) customized equipment. CMEA countries have specialized mainly in low-technology capital equipment. Western Europe clearly enjoys an advantage in this respect vis-a-vis Eastern Europe. Moreover, R&D-intensive products can be financed more easily if firms are major exporters to the world markets. Spreading R&D costs over a large number of domestic and international customers is an advantage in R&Dintensive production. Again, since EU countries are more open in trade and are

Privatization, structural change, and productivity

253

traditionally export oriented, R&D-intensive production will mainly be located in Western European countries. An offsetting aggressive R&D policy by Eastern European governments to subsidize R&D expenditures of exportoriented industries is conceivable. But it is doubtful that Eastern Europe could generate enough political stability to sustain consistent long-term R&D policies, which are necessary for technological upgrading. In the neoclassical world, the assumption of factor immobility has for a long time clouded some interesting cases of real world asymmetries in foreign direct investment. In a two-country world with FDI, assume that both countries are of equal size as regards the number of consumers and the number of firms. If there is two-way FDI, the outward FDI position of country I could be much more favorable than that of country II. In an extreme case, the foreign country's capital stock K* could be wholly owned by capital owners of country I. This would bring about major divergences in gross national income - even if gross domestic products were equal in country I and II. If production functions were identical in the EU and Eastern Europe, namely Y = A^L ( i p ) and capital abroad (in Eastern Europe) were fully owned by EU firms, GNP in the EU would be much higher than in Eastern Europe even if output in the EU, Y, and output in the ex-CMEA area, F*, were identical. Let us assume that the population and capital stock in the EU and Eastern Europe are identical. Then the EU's GNP - which is Y plus net factor income from abroad - would be Y + (3*F*, while in the ex-CMEA area GNP would be equal to F* - p*y*. An illustrative case would be Y = F*, p = P* = 3, which would lead to a West-East income ratio in Europe of (1 + p)/(l — p) = 2. Clearly, despite convergence in factor rewards, an asymmetric international distribution of ownership could cause major international per capita income differentials. This in turn would cause diverging demand patterns and sectoral structures. This illustrative calculation is not meant to argue that Eastern Europe should restrict foreign direct investment inflows, which create important technology spillovers (especially in backward countries) and have pro-competitive effects. But one may infer how important it is that firms in Eastern Europe develop the firm-specific advantages that allow firms to become viable competitors with production abroad. The Asian NICs have accepted FDI inflows and subcontracting for decades, until they increasingly became important sources of foreign direct investment. Massive inflows of FDI could raise both the overall investment/output ratio and the marginal product of capital; FDI inflows frequently occur in industries with high R&D intensities. However, political stability in host countries is required, which may be difficult to achieve in some ex-CMEA countries. Convergence in Europe is possible if Western Europe keeps its markets open and if Eastern Europe embraces radical and sustainable transformation toward a market economy. A major obstacle is political instability in the smaller exCMEA countries, while in Russia both enormous economic and political problems could prevent sustained economic growth. The macroeconomic policy re-

254

Paul J. J. Welfens

quirements for high growth are clear from the analysis of developing countries. Competitive goods and factor markets in combination with positive real interest rates and market-clearing exchange rates, as well as low inflation rates and budget deficits, will lead to high economic growth. These lessons are not exclusively for developing countries, but hold for middle-income countries, too, as shown by the analysis of the successful catching-up processes of Spain and Portugal versus that of Greece in the 1980s (Larre and Torres, 1991). Without comprehensive privatization, competition policy, and prudent FDI policies, Eastern Europe is facing a Greek tragedy. One should not overlook the possibility that removing coordination failures by prudent government intervention also could stimulate economic growth (Rodrik, 1995). However, in Eastern Europe, the risk of massive rent seeking rules out optimal government intervention as a viable option. 6.6

Summary and conclusions

The analysis shows that privatized companies record higher productivity growth than state-owned firms in Eastern Europe, where Poland and the Czech Republic are leading the transformation process. Hungary's lead as a host country for foreign investors allows one to expect rapid progress in the restructuring of firms taken over by foreign investors. At the same time, it seems that Hungary's very success in attracting FDI has diverted attention from a comprehensive and consistent restructuring of the banking sector. Indeed, there might be a winner's curse with respect to the FDI race in Eastern Europe, where major FDI inflows reduce emphasis on competition policy and capital market policy. Efficient capital markets are crucial for productivity gains, as are efficient governance structures. Infrastructure investment is vital for the transition process, since it raises factor productivity directly and indirectly (via the pro-competitive effects of reduced transportation costs). High budget deficits imply the risk of suboptimal investment in infrastructure. The role of infrastructure investment is also important because it will stimulate additional private investment in equipment and construction - the latter is labor intensive and could absorb considerable parts of excess employment, which still characterizes the private and public sector. The stagnation or even fall of construction output shares in transforming countries partly reflects lack of infrastructure investment. Insufficient infrastructure investment - often a result of high budget deficits - reduces the prospects of high private investment in machinery and construction; and construction is important because due to its high labor intensity it can absorb a large number of people laid off in manufacturing and agriculture. Overall, the Visegrad countries lead the transition process in the ex-CMEA area and should benefit both from the link between the privatization process and nurturing capital markets on the one hand, and on the other hand from the productivity-enhancing link between external liberalization and privatization. There is only partial privatization in infrastructure, where Hungary and the

Privatization, structural change, and productivity

255

Czech Republic tend to adopt innovative privatization projects in telecoms and the energy sector. Hungary slowed down privatization for various reasons and suffers from a lack of voucher privatization schemes, which would allow a broader strata of society to capitalize on the gains from moving to a market economy. In Poland and the ex-CSFR, voucher privatization encouraged workers to accept steep falls in real wages, since offsetting capital gains were expected. Banks and insurance companies as well as big companies are an unresolved issue on the privatization agenda. Lack of infrastructure investment in Visegrad countries is a bottleneck for a sustained investment boom and rapid productivity growth. High investment/output ratios as well as factor productivity could be impaired by political instability, which reduces the amount of profitable investment and shortens the investment horizon of both capital and innovation projects. Russia, Ukraine, Bulgaria, and Romania have achieved much less progress in privatization than the Visegrad countries - and there is some risk that Eastern Europe could fall back into historical pitfalls (Welfens, 1992, 1995a; Tilly and Welfens, 1995). This could lead to Latin American misery scenarios, in which the private competitive sector - facing vested political and monopoly interests - cannot expand. Oversized firms, lack of regional mobility, and poor infrastructure imply small productivity gains. In combination with low FDI inflows and a possibly increasing capital-income/wage-income ratio, this implies slow growth and rising social conflicts. A massive and sustained productivity gap vis-a-vis the Visegrad countries could emerge, which could stimulate migration into the Visegrad countries and a politically aggressive policy stance of the transition laggards. These developments in turn could undermine the transformation success in the Visegrad countries and impair their catching-up process. There is the risk of a double economic divide in Europe, which means an economic divide between the EU and the Visegrad countries and a new economic divide between the Visegrad countries and other ex-CMEA countries. REFERENCES Arrow, K. (1962). "The Economic Implications of Learning by Doing," The Review of Economic Studies 29, 155-73. Audretsch, D. B. (1995). "Industrial and Trade Policies for Emerging Market Economies," in L. A. Winters (ed.), Foundations of an Open Economy, London: Centre for Economic Policy Research, 155-77. Baldwin, R. (1989). "The Growth Effects of 1992," Economic Policy 9, 248-81. Chenery, H. B. (1960). "Patterns of Industrial Growth," American Economic Review 50, 624-54. Chenery, H. B., and L. Taylor (1968). "Development Patterns: Among Countries and Over Time," Review of Economics and Statistics 50, 391-416. Chmiel, J., and Z. Pawlowska (1995). "Development of Private Sector in Poland, 1990-93," Research Bulletin, No.l: RECESS, Warsaw, 37-63. Collins, S. M. (1990). "Lessons from Korean Economic Growth," American Economic Review, Papers & Proceedings 80, 104-107. Darwick, S., and D. Nguyen (1989). "OECD Comparative Economic Growth 1950-85: Catch-up and Convergence," American Economic Review 79, No. 5.

256

Paul J. J. Welfens

Dlouhy, V., and Mladek, J. (1994). "Privatization and Corporate Control in the Czech Republic," Economic Policy 19 Supplement, 155-71. Dbhrn, R., and Heilemann, U. (1993). "Structural Change in Eastern Europe," in B. Heitger and L. Waverman (eds.), German Unification and the International Economy, London: Routledge, 83-112. Dohrn, R., and Heilemann, U. (1992). "Zur allokativen Dimension: Sektorale Entwicklungsperspektiven ftir Ungarn, Polen und die CSFR," Wirtschaftsreformen in Mittel- und Osteuropa, Beiheft zur Konjunkturpolitik, Heft 40, Berlin, 51-69. EBRD (1994). Transition Report, London: European Bank for Reconstruction and Development. EBRD (1995). Economics of Transition, London: European Bank for Reconstruction and Development. Englander, A. S., and Gurney, A. (1994). "OECD Productivity Growth: Medium-term Trends," OECD Economic Studies 22, 111-29. Habuda, J., and Jennewein M. (1995). "Unternehmensanpassung in den Transformationslandern," IFO-Schnelldienst, 12/95, 9-18. Heilemann, U., and Lobbe, K. (1995). "The Structural Renewal of Eastern Germany: Some Initial Observations," in P. J. J. Welfens (ed.), Economic Aspects of German Unification, 2nd ed., New York: Springer. Helpman, A., and P. R. Krugman (1985). Market Structure and Foreign Trade, Cambridge, Mass.: MIT Press. Herer, W., and W. Sadowski (1995). "Effects of Change in Production Structure on Raw Materials and Energy Barriers to Economic Growth in Poland, 1990-1993," Research Bulletin, No.l: RECESS, Warsaw, 3-36. Hoekman, B. M., and P. C. Mavroidis (1995). "Linking Competition and Trade Policies in Central and East European Countries?", in L. A. Winters (ed.), Foundations of an Economy, London: CEPR, 111-53. Hunya, G. et al. (1994). "Central and Eastern Europe: Uneven Recovery," Research Paper 204, Wien: WIIW IMF (1995). World Economic Outlook (Spring), Washington D.C. Inotai, A. (1992). "Foreign Direct Investments in Reforming CMEA Countries: Facts, Lessons, and Perspectives," in M. Klein and P. J. Welfens (eds.), Multinationals in the New Europe and Global Trade, New York: Springer Verlag, 129-38. Jackson, M., and W. Biesbrouck (eds.) (1995). Marketization, Restructuring and Competition in Transition Industries of Central and Eastern Europe, Aldershot: Avebury. Jackson, M., and V. Bilsen (eds.) (1994). Company Management and Capital Market Development in Transition, Aldershot: Avebury. Jasinski, P., and P. J. J. Welfens (1994). Privatization and Foreign Direct Investment in Transforming Economies, Aldershot: Dartmouth. Klein, M., and P. J. J. Welfens (eds.) (1992). Multinationals in the New Europe and Global Trade, New York: Springer Verlag. Klodt, H. (1993). "Perspektiven des Ost-West-Handels: Die komparativen Vorteile der mittel- und osteuropaischen Reformlander," Die Weltwirtschaft, 424-40. Koop, M. J. (1994). "Privatisierung und Effizienz," in J. Hoelscher, et al., Hg. Bedingungen okonomischer Entwicklung in Zentralosteuropa, Marburg: Metropolis, 287-323. Krueger, A. O. (1990). "Asian Trade and Growth Lessons," American Economic Review, Papers and Proceedings 80, 108-12. Larre, B., and Torres, R. (1991). "Is Convergence a Spontaneous Process? The Experience of Spain, Portugal and Greece," OECD Economic Studies 16, 169-98. Levine, R., and Renelt, D. (1992). "A Sensitivity Analysis of Cross-Country Growth Regressions," American Economic Review, September.

Privatization, structural change, and productivity

257

Lipton, D., and D. Sachs (1990). "Creating a Market Economy in Eastern Europe: The Case of Poland," Brookings Papers on Economic Activity. Lucas, R. E., Jr. (1990). "Why Doesn't Capital Flow from Rich to Poor Countries?", American Economic Review, Papers and Proceedings 80,92-96. Munnell, A. (1993). "An Assessment of Trends in and Economic Impacts of Infrastructure Investment," OECD, Infrastructure Policy for the 1990s, 21-54. Newbery, D. M. G. (1994). "Restructuring and Privatizing Electric Utilities in Eastern Europe," Economics of Transition 2, 291-316. OECD (1995) Mass Privatisation, Paris: Organization for Economic Cooperation and Development. Rebelo, S. (1990). "Long Run Policy Analysis and Long Run Growth," NBER working paper 3325, April. Rodrik, D. (1995). "Getting Interventions Right: How South Korea and Taiwan Grew Rich," Economic Policy 20, 55-107. Romer, P. M. (1986). "Increasing Returns and Long-Run Growth," Journal of Political Economy 94, 1002-32. Romer, P. M. (1990). "Are Nonconvexities Important for Understanding Growth?", American Economic Review, Papers and Proceedings 80, 97-103. Schipke, A., and A. M. Taylor (eds.) (1994). The Economics of Transformation, Heidelberg: Springer. Seitz, H., and Licht, G. (1995). "The Impact of Public Infrastructure Capital on Regional Manufacturing Production Costs," forthcoming in Regional Studies 29. Sengupta, J. K. (1991). "Rapid Growth in NICs in Asia: Test of New Growth Theory for Korea," Kyklos 44,561-80. Sigmund, P., and K. Werner (1995). "Privatisierung in Rupland: eine widerspriichliche Zwischenbilanz," IWH-Wirtschaft im Wandel, 4/1995, 11-15. Song, B. N. (1990). The Rise of the Korean Economy, Oxford: Oxford University Press. Stifterverband(1995). Wissenschaftsstatistik, Essen. Szekely, I. P., and D. M. G. Newbery (1993). Hungary: an Economy in Transition, Cambridge: Cambridge University Press. UNCTAD and Kopint-Datorg (1994). Privatization in the Transition Process, Geneva: United Nations Conference on Trade and Development. UNECE (1995). Economic Survey of Europe in 1994-95, Geneva: United Nations Economic Commission for Europe. Vickers, J., and G. Yarrow (1988). Privatisation: an Economic Analysis, Cambridge, Mass.: MIT Press. Welfens, P. J. J. (1994). "Theoretical Aspects of Transformation in a Big Country: The Case of Russia," paper presented at 3rd EACES Conference: "The Transformation of Economic Systems," Budapest, Sept. 8-10. Welfens, P. J. J. (1995). Economic Aspects of German Unification, 2nd revised enlarged edition, New York: Springer. Welfens, P. J J., and G. Yarrow (eds.) (1995). Telecommunications and Energy in Systemic Transformation, New York: Springer. Wolff, E. (1995). "Productivity Growth among OECD Countries: Historical Record and Future Prospects," Washington, D.C.: World Bank. World Bank (1994). Policies for Growth with Equity, Washington, D.C. World Bank (1995). Global Economic Prospects and the Developing Countries, Washington, D.C. Zylicz, T. (1994). "Implementing Environmental Policies in Central and Eastern Europe," in A. Jansson et al. (eds.), Investing in Natural Capital, Washington, D.C: Island Press.

Comment Bruce Kogut

The paper by Professor Welfens is an analysis of three related issues: privatization, structural change, and productivity. The study of the relationship between structural change and productivity is an ambitious topic. To add in privatization is a necessary piece of the understanding of the new landscape of economies in transition. Professor Welfens does an admirable job of pulling these issues together, and in the course of this journey avoiding most traps and, despite the complexity, arriving at a number of important insights and recommendations. His most important observation is that privatizing an entire industry poses fundamentally different implications than privatizing a firm. There are macroeconomic implications due to the effect on the budget, financing, and employment. There are also important network externalities that are generated through establishing suddenly large numbers of private firms. Welfens's assessment of privatization is largely positive. He sees privatization increasing productivity, as long as a number of pitfalls are avoided. He places particular emphasis on the effect of increased competition on productivity and warns against the policy that privatizes without establishing competition. His comments on the dangers of subsidies are particularly apt. Noting first that the potential for externalities suggests a policy of intervention, he cautions that a program of subsidization is likely to open the doors to rent seeking. It would be better to restructure first, close down inefficient operations, and then consider subsidies for particular activities. For particular sectors, he notes that government finances are inadequate to correct potential gaps between private and social returns. This perspective leads him to argue for the privatization of university education, since the current wage gap between the public and private sectors has led to the loss of the more qualified teachers. One can quibble that the distinction between privatization and competition is frequently lost. Much of the listed gains to privatization appears to be the consequences of deregulation or of resolving financial constraints. Certainly, the massive investments required in energy and infrastructure are beyond the financial resources of these governments. Yet, given the sizes of these countries, 258

Comment

259

a policy that allowed for deregulation, import competition, and entry would provide itself a powerful engine for change. While not assessing the relative contribution of privatization versus other policies, Welfens is certainly aware that structural change alone will increase productivity through the reallocation of resources across sectors. In particular, he notes the declining role of agriculture and industry and the growth of the service sector. To a certain extent, the growth of the service sector is not so much simply an expression of increased efficiency, but a necessary element of a capitalist system. Insurance and financial derivatives are good examples. It would be inaccurate to say that people and property were inadequately insured previously, or that there were inadequate mechanisms by which to manage many kinds of risk. Rather, many of these risks have been newly introduced into these economies. The final part of Welfens's paper looks at ^industrialization. He notes that the effects of opening an economy to trade poses complex implications for income distribution, especially given current distortions. His discussion appears to refer to the kind of analysis associated with Jagdish Bhagwati on immiserizing growth and trade with distortions. There is a difference, insofar as Welfens notes that these income effects will influence the political acceptability of reform. Similarly, he notes that direct investment may provide the capital flow required in a period of inadequate savings, but that higher net claims of foreigners on resident assets implies a persisting inequality in income distribution between transition and lending economies. The discussion could apply to either portfolio or direct investment flows. Welfens's discussion of direct investment implies that the issue is not so much income inequality as ownership. It is also unclear whether a policy that permits capital imports coupled with wage restraints could not increase the level of income sufficiently to compensate for foreign interest and dividend reimbursement. After all, the persistence of net claims of the United States on European and Asian countries was not long-lived. Moreover, if one moves away from the production function described in these pages to the endogenous growth models described earlier in the paper, a policy of borrowing for human and capital investment might have more than compensating effects. The reason these issues get messy is that there is no good theory to separate out the effects of portfolio from direct investment. The issues involved are not far removed from the overall literature on corporate governance, whereby different types of owners are poised to influence the productivity of firms. Does it matter who owns the firms? From an economic standpoint, the issues are cloudy. There are two types of considerations. The first is the question of valuation in a period of highly imperfect and incomplete capital markets. It is easy to dismiss the occasional outcry over privatization as nationalistic. At the same time, the valuations placed on assets can be surprisingly low. A study conducted by CS First Boston (1994), in their efforts to market the Russian economy, pointed out that in sectors such

260

Bruce Kogut

as telecommunications, energy, or metal, the capitalized value of capacity was a fraction of Western equivalents; only the tobacco industry, in which foreign investment has been large, shows Western valuations. These underevaluations might represent the shortage of domestic savings, or the discount applied to firms with poor governance oversight. In fact, astute analysts of this problem seem to swing between these interpretations (see Boycko et al., 1993, 1994). Yet there is no doubt that this period of transition presents unusual opportunities to those with the requisite wealth. Often, this wealth is foreign. The second consideration has to do with whether the subsequent investment behavior of foreigners differs from nationals. Cantwell (1989) has presented a formal description and some evidence to suggest that foreign investment can enhance productivity if the domestic sector is already enjoying positive externalities; but productivity may be hurt if investment is used for market access and the activities that are responsible for externalities, for example, research and development, are closed down. Tyson's (1992) argument for an activist trade policy seems to be based upon similar considerations. Of course, in the context of an integrated European economy, these considerations dim, for why should externalities respect national borders? Czechoslovakia, Poland, and Hungary are closer to the industrial heartland of Europe than Portugal. (See Sachs, 1993, for data.) The German investment in these countries is a reflection of the disregard of national boundaries. The fate of Bulgaria is partly its lower level of development, but also its location. The large share of Greek investment in its economy might, one can hypothesize, have lower externalities than that of German or Austrian. A simpler explanation rests on the politics of these economies. There is a tendency, one feels, to treat the politics of these countries as an expression of why economic development is stalled. But there is another side to all this, namely, that the nature of the future state and society is very unclear. People are naturally trying to influence their destinies in very profound ways. What is at stake is the creation of "bourgeois" societies, and I use this expression in the context of what it means in Europe. Balzac's observation of capitalist development in nineteenth-century France that behind every great wealth lies a great crime was exaggerated but still insightful. It would seem to hold true with regard to the accumulation of wealth in transition economies today. It is a great puzzle how to deal with the conflict between the need for the educated Old Guard and their belief in the earlier system. The rapidity of business creation and growth, which has particularly propelled the Polish economy, is explainable partly by the capitalist opportunities created by misvaluations and access to specialized resources. For this reason, the process of privatization is a deeply political issue, about who will belong to the future class of owners. Welfens is not correct in stating that Russian privatization has been less extensive than that in Central Europe, but he is right if he means that privatization has not resulted in the creation of Western-style governance mechanisms and has not displaced managers. The

Comment

261

Polish privatization program has not been especially impressive, but it has come to be an increasingly less interesting question. For with the rapid development of a private economy inhabited by new start-ups and by foreign firms and the atrophy of the public sector, the Polish economy is more Schumpeterian than neoclassical. And if the Hungarian economy looks stagnant by comparison to its Czech neighbor, it is because the resolution of competing claims between labor and new interests is far from complete. The link between privatization and productivity is, one suspects, a real phenomenon, but I question the generality of its significance. The nationalizations of France following Mitterand did not palpably affect French productivity, even though the most important enterprises were affected. Some might argue that this change in ownership was efficient from the perspective of Olson; it broke up the inertia that had accrued in the extensive intercorporate ties of the French economy. If privatization looks more promising now, it is because the French economy, and financial markets, are significantly different. The case of Russia is instructive. One accepts the argument of Boycko et al. (1993) that privatization in Russia succeeded above all in breaking up the power of the ministries and fostering a climate of competing claims on government resources. This result had the ironic effect of both weakening the traditional ministries and increasing the strength of government agencies responsible for the disposition of state property and for fiscal relief. Boycko et al. like this outcome, because reform interests dominate these agencies. Privatization has not, as we all recognize, led to an obvious gain in productivity, but it has laid the foundations for a balance of political interests. In all of these countries, there has been a search for mechanisms by which to exercise corporate governance in order to discipline managers. Despite the interest in equity markets, these markets are relatively small. In any event, external equity financing has been historically minor in most industrial countries; retained earnings has been the most important source of financing. If motivated owners are unlikely to be generated by equity markets, then it is natural to look toward banks as corporate overseers. But by and large, the banking systems of transition economies are very weak, and remain fragile despite major infusions of capital and writing down of bad loans. It is also ironic that many of these banks are still state-owned. (Here, Russia is a telling exception.) In Czechoslovakia, where a few banks are important owners of mutual funds, the state through its bank holdings remains an important owner of many newly privatized companies. Yet there will remain two important influences on enterprise behavior. One is simply the force of product market competition. Welfens notes the importance of hard budget constraints, but of course, private firms, as he observes, can also wield tremendous power in acquiring subsidies. Russian subsidies still amount to 6 to 7 percent of GDP, even though 45,000 firms have been privatized. The second influence is other enterprises, who, through supplier and buyer credits, are heavily indebted to each other. In Russia in particular, there is the

262

Bruce Kogut

suggestion that the wealthy firms, who have large profits from exporting commodities to the West, have become major sources of funding for other firms. In other words, we have a recapitulation of the industrial enterprise group that has been a common feature of the capitalist development of Germany, Japan, Korea, France, and other countries. There has been an overemphasis on the efficiency properties of privatized companies and the creation of corporate governance mechanisms. Competition and entry are important properties to any system insofar that they reward change. These properties can be achieved through a policy of low tariffs and deregulation. The Polish case is a good example of change that is driven by the dynamic of small and medium-size firm creation. Privatization is, above all, a political policy, one that may be required to render reform efforts irreversible, but political nevertheless. The real task is creating conditions of competition and new firm entry; privatization of existing enterprises is of secondary importance when such conditions are met. REFERENCES Boycko, Maxim, Andrei Shleifer, and Robert Vishny (1993). "Privatizing Russia," Brookings Papers on Economic Activity. Boycko, Maxim, Andrei Shleifer, and Robert Vishny (1994). "Voucher Privatization," Journal of Financial Economics 35, 249-66. Cantwell, John (1989). Technological Innovation and Multinational Corporations, Oxford: Basil Blackwell. CS First Boston (1994). "Russian Capital Markets," Boston: CS First Boston (private distribution). Sachs, Jeffrey (1993). Poland's Jump to the Market Economy, Cambridge, Mass.: MIT Press. Tyson, Laura (1992). Who's Bashing Whom?, Washington: Institute of International Economics.

Comment Jan Mlddek

After listening to Professor Welfens's somewhat technocratic presentation, I must think about the issue of which is the right profession to handle the transformation from a Soviet-type economy to a market economy. Should it be a (technocratic) economist, political economist, banker, businessman, accountant, or perhaps a philosopher or sociologist? Being an economist, I would not opt for the last, but I do believe that for transformation, Schumpeter and Hayek are much more relevant than Milton Friedman or Sachs. The process is really a political and economic one. Please allow me to tackle three problems. First of all, it seems that this general approach to all the post-communist countries is not acceptable. Postcommunist countries are unified only by a common fate in the years after the Yalta treaty. This fate was the attempt to introduce the same system of economic and social life - the Soviet model of communism. The countries differ for many reasons. First of all, one must differentiate between European and non-European post-communist countries, in other words, between those that share a common cultural tradition with the West and those that do not. Western civilization is thought by some to be universal - it is, in fact, not universal. Western civilization is built upon Catholic, Protestant, and Jewish traditions. This means that countries that share these traditions with the West have smaller problems with reintegration. In Europe this borderline runs through Helsinki, Tallinn, Riga, Vilnius, Lwow, Budapest, Zagreb. In my opinion this line at the moment divides the countries that are coping somehow with transformation and those that are not. Weber's theory that Protestantism is better for a market economy than Catholicism should be extended by the statement that the Eastern Orthodox tradition is even worse than Catholicism for a market economy. For those who do not like cultural and religious explanations, I have another reason why it is necessary to differentiate among the post-communist countries, namely the distance of those countries from the wealthy West. The proximity of Poland, the Czech Republic, and Hungary to Germany can in itself explain a lot. The clearest case is Hungary. West of the Danube is the land of prosperity. East of the Danube is the land of economic problems. 263

264

Jan Mladek

A country's endowment of raw materials and physical factors of production is also very important. Raw materials put Russia into a favorable position (despite all the problems with reform), particularly in comparison to countries like Belarus or Ukraine. Physical factors explain the favorable position of the Czech Republic. Communist industrialization meant in fact reindustrialization and expansion of already existing facilities (which had initially been allocated by market forces). Concentration of these companies was only administrative, not physical, and this is why it was possible to quickly create hundreds of new independent companies from the old communist ones. Last, but not least, is the relation of the region to the European Union as a political and economic block. It seems there are some eleven countries that do have the chance to become, sooner or later, members of the EU. On the other hand, there are countries that have virtually no chance, particularly Russia. Russia will eventually be forced to establish its own economic bloc or union, which should at least be competitive if not completely hostile to the EU. The second problem I would like to address is the issue of privatization frontiers. In Professor Welfens's paper it is suggested that there should be full privatization of utilities, railways, and so on. In the Czech Republic, we have just reached the frontiers of privatization. At the moment the debate centers around the privatization of hospitals, and the general public seems to be hostile to this idea. The outcome will probably be a silent change of privatization into municipalization of big hospitals. Originally, British-style privatization was planned for the railways. Recently, the government decided to privatize only a small percentage of railways (30 percent of the length, but only 10 percent of the value and the people of the current railways) and to keep the core railways as a state-owned company. Utilities, so far, are only partially privatized. Progress, with the exception of the telecom, will probably be slow. In fact, if the regulation of privatized utilities is not done properly, the performance could be worse than before, or even counterproductive to the development of other private companies. For these reasons, I do not believe that privatization of utilities should be the priority for reforms in post-communist Europe. Privatization of utilities was done in the West only recently and the results so far have been "mixed." Besides these reforms, Central and Eastern Europe has plenty of other targets for urgent privatization. The third issue on which I would like to comment is the issue of disclosures of companies that are tradable on the stock exchange. In the Czech Republic we have at the moment 1700 companies that are tradable. This is an absurd number for such a small country. The majority of them should be classified as "Ltd.," meaning they are not freely tradable by law, or as closed joint stock companies. To require detailed information from such companies could be counterproductive. In my opinion, the government should support so-called secondary privatization of smaller voucher privatized companies. The strict rules of disclosure should be applied only to the large companies. This secondary privatization is, however, not without risks. There is the danger that the concen-

Comment

265

tration of ownership needed from an economic point of view will be done at the expense of minority shareholders. If someone concentrates 51 percent of the shares and then changes the company into a closed joint stock company, those locked in will have no real chance to sell or collect any dividends. This tradeoff is tough. To introduce, for example, the rule that everybody with a stake higher than 35 percent is obliged to suggest a buy-out to all shareholders would be moral, but could cause the slowing down of the concentration process due to higher costs and logistical troubles. Without introducing those kinds of rules, hundreds of thousands of shareholders could finish with nothing. Risk is political.

CHAPTER 7

Integrating the East: the labor market effects of immigration Thomas Bauer and Klaus F. Zimmermann

7.1

Introduction

East-West migration is a central theme for the EU (European Union) integration process of central Europe. There is no doubt about the attractiveness of the political vision of integration. In the long run, there is also the economic vision of larger markets in a prosperous larger union of states. The issue is speed: speed of convergence of economic conditions, and priority in the flexibility of trade, capital, and labor. This paper argues that investing in Eastern Europe or liberalizing trade with these countries will not act quickly enough to relieve current demand for economic development and current pressure to migrate. On the contrary, a selective (at first temporary) immigration policy of the EU concerning Central Europe might be useful. It is suggested that East-West migration offers the alternative of skilled or unskilled migration and therefore introduces a new (qualitative) component to a European migration policy. In the short run, the issue is dominated by the problem of economic transition to a market economy in the East and from high levels of unemployment to lower levels in the West. In spite of the current economic crisis in the West with its 19 million unemployed in the EU in 1993, there are substantial economic East-West differences. However, it is unclear to what extent these differences cause out-migration. Previous experience with inner West European migration suggests that labor mobility is rather slow in adjusting wage and unemployment differentials. The experience so far with the anticipated mass migration from the East confirms this view. Nevertheless, the high unemployment rates in the West, larger than 10 percent in the EU in 1993, feed the emotional debates on migration in the West and the reluctance of policymakers to consider economic migration policies. If the developed economies do not become jobless societies, and given the We wish to thank both discussants and the participants of the AICGS Conference on May 15-16, 1995, in Washington, in particular John Haisken-De New, Andreas Million, Regina Riphahn, Ralph Rotte, Christoph M. Schmidt, and Johannes Wiegand for helpful comments on earlier drafts, and Mr. Winger from the Statistisches Bundesamt for providing some of the data used in this study.

269

270

Thomas Bauer and Klaus F. Zimmermann

stability of the current population projections, the long-run immigration issue of the EU may well be dominated by demographics. The most important demographic challenges for Europe in the next decade are (1) worldwide migration pressures from the less-developed regions of currently about 80 to 100 million people and the East-West migration potential, which was estimated between 5 and 40 million people; (2) a decline of Western European population and fast growing countries in Eastern Europe; and (3) a considerably aging labor force in Western Europe. From these developments it seems easy to predict that the next decade will see a large migration pressure from Eastern Europe that badly requires a policy response. Migration policies therefore have to consider jointly short-run and long-run issues. There is also a potential choice between South-North and East-West migration. The analysis in this paper concentrates on economic (labor) migration from the East to the West. Is migration potentially harmful or beneficial? Should a migration policy be selective? Section 7.2 discusses the current situation in the labor markets, the demographic development in the EU and Eastern Europe, as well as past migration flows from the East to the West since the collapse of the socialist regimes in Central and Eastern Europe. Section 7.3 summarizes current developments in migration policies in the EU and discusses some options for the future. Section 7.4 presents a theoretical framework to study the labor market effects of immigration. In Section 7.5, this framework is used to provide some rough calculations of the net benefits the West can expect from East-West migration using Germany as a benchmark case. Section 7.6 concludes. 7.2

Economic situation and migration potential

It is hardly possible to estimate the potential migration flow from Eastern to Western Europe. Several studies that have tried to estimate the extent of this migration wave have arrived at very different results. Many newspapers and politicians have speculated that about 20 to 40 million Eastern Europeans will emigrate. Estimates based on opinion polls in the sending countries suggest that between 13 and 27 million people are planning a move to the West (Coleman, 1993), whereas more modest predictions expect about 5 million people to migrate to Western Europe (see IOM 1991). Assuming that 5 to 40 million would come within the period 1994-2000 to the EU, this would imply an average inflow of 0.2 to 1.6 percent of the population size of the EU in 1994. Immigration to Germany in the last decades had always been around 1 percent. [See Schmidt and Zimmermann (1992) for a discussion of the German migration experience since World War II.] Hence, at the lower end of the predictions, migration would not seem to be a potential burden. The future of East-West migration largely depends on political stability in the East and the economic and demographic development in Eastern and Western Europe. Furthermore, the migration policies of Western European coun-

Integrating the East: labor market effects

271

tries, described in the next section, play an important part. With respect to the political situation in the East, it can be observed that democratic structures and human rights have been developed very quickly, but their stability in most countries (perhaps with the exception of Poland, the Czech Republic, and Hungary) is still questionable. Therefore, politically motivated migration may gradually be disappearing, but it cannot be excluded. Economic theories of migration conclude that the economic conditions in the sending country relative to the receiving country are important determinants of the migration decision. [An overview of economic theories and their empirical evaluations is provided by Bauer and Zimmermann (1995).] Table 7.1 exhibits some economic indicators of Western and Eastern European countries in 1993. The process of transforming the previously centrally planned economies of Eastern Europe into market economies does not progress at the same rate in all transition countries. In Eastern Europe, reductions in employment over the three years 1990-92 varied widely. Whereas the decline was less than 5 percent in the Commonwealth of Independent States (CIS) and Romania, it reached nearly 30 percent in Bulgaria. Employment decline in the EU countries, with the exception of the United Kingdom, was lower than in all Eastern European countries. In all Eastern European countries the transition to a market economy was followed by a sharp increase in unemployment, which was not existent, at least officially, under central planning. [Schmidt (1996) provides a recent analysis of unemployment in Poland.] In 1993, the unemployment rates in Eastern Europe varied widely, from 1 percent in CIS and 3.5 percent in the Czech Republic to 25 percent in Albania. In 1993 none of the EU countries had less than 5 percent unemployment, and only Albania had a higher unemployment rate than Spain and Ireland. Thus, there seems to be little potential for large-scale movements from the East to the West. Despite the high unemployment rates in the West, there is demand for high-skilled technical and professional workers, especially in information technology, as well as for low-skilled service workers. For instance, CIS specialists in space technologies, lasers, low-temperature physics and superconductivity, some fields in medicine, and computer software may be in great demand in the West (Coleman, 1993). If these high-skilled workers want to live in the West permanently, their emigration could lead to a brain drain in the Eastern European countries, with negative consequences for their future economic development. However, in the case of temporary migration these high-skilled workers would get acquainted with Western techniques and therefore could help their source country after returning. Another important factor encouraging migration in the short term is the persistently high income differential between the East and the West. For instance, in 1993 an average worker in the manufacturing industry in western Germany earned US $ 2058 per month. In Hungary, the average worker earned 14.0 percent, in Poland 10.7 percent, in the Czech Republic 9.7 percent, in Slovakia 8.3 percent, in Bulgaria 5.5 percent, in Romania 3.5 percent, and in the CIS 2.5 per-

273

272

Table 7.1. Economic indicators of Western, Central, and Eastern Europe in 1993 Country

Employment Change (%) (1990-92)"

Unemployment

Inflation Rate (%)"

Average Monthly Earnings (US $)<

GDP Real Growth Rate (%)*

2.8

1,667

-1.3

(%f

Western European Countries Belgium

-0.7

9.4

Denmark

-1.0

10.4

1.7

2,211

1.1

France

-0,5

10.8

2.2

1,306

-0.7

Germany

2.6

5.6

3.3

2,058

-1.9

Greece

-0.9

9.8

13.7

660

-0.2

Italy

0.7

111

5.1

—

-0.7

Ireland

-0.1

18.4

2.0

1,316

2.5

Netherlands

4.9

8.8

2.1

—

0.3

Portugal

-3.7

5.0

6.8

—

-1.2

Spain

-1.7

21.5

5.1

1,134

-1.0

United Kingdom

-5.4

10.5

3.5

1,383

1.9

EU

-1.4

10.5

3.8

—

-0.5

Albania

-21.7

25.0

117.0"

—

11.0

Bulgaria

-28.7

16.4

70.0

114

-5.0

Central and Eastern European Countries

Czech Republic

-8.8

3.5

20.0

199

-1.0

Hungary

-12.1

12.1

23.0

288

-2.0 4.0

Poland

-12.6

15.7

35.0

220

Romania

-4.5

10.1

260.0

73

0.0

Slovakia

-13.5

14.4

25.0

170

-6.0

CIS

-4.0

1.0

900.0

51

-12.0

" Percentage change over the period. * Percentage change over previous year. c Western European Countries: Average hourly wages in manufacturing multiplied with hours worked per week according to the trade union wage agreements multiplied by 4. d From January to September 1993. Sources: Economic Commission for Europe (1994); Statistisches Bundesamt (1994,1995); Deutsche Bundesbank (1994); EuropSischeKommission (1994); DIW (1994a, 1994b); own calculations.

274

275

Table 7.2. Demographic indicators Country

Total Population 1994 (millions)

Total Population Growth 1994 2025 (%)

Age Structure — 1990 15-64

2025

65+

15-64

65+

Western European Countries Belgium

10.0

-1.0

67.0

14.9

62.6

21.7

5.2

-1.9

67.2

15.6

63.1

21.7

France

57.4

5.9

65.7

14.0

62.2

21.2

Germany

81.2

4.1

68.7

14.6

64.1

20.5

Greece

10.2

-1.0

66.9

13.8

62.2

22.2

Italy

57.8

-2.8

68.6

14.1

63.3

22.3

Ireland

3.5

2.9

61.3

11.4

61.7

17.8

Netherlands

15.3

15.7

69.0

12.7

64.2

19.8

Portugal

9.9

2.0

66.0

13.1

64.3

18.9

Spain

39.2

3.6

66.9

13.4

63.7

20.2

United Kingdom

57.8

4.3

65.3

15.7

63.6

19.4

347.5

3.3

67.3

14.5

63.4

20.7

3.3

36.4

62.0

5.2

66.9

11.4

Bulgaria

8.9

-1.1

66.5

13.0

64.6

17.8

Former CSFR

15.7

14.0

65.3

11.7

64.6

16.4

Hungary

10.5

-1.0

66.2

13.2

64.5

18.0

Poland

38.5

13.8

64.7

10.0

63.3

16.3

23.4

12.4

66.0

10.4

65.2

14.3

284.5

21.1

64.9

9.3

64.1

14.1

Denmark

EU

Central and Eastern European Countries Albania

Romania

CIS

Sources: Eurostat (1991), United Nations Population Fund (1994), United Nations Population Division (1992), own calculations.

276

Thomas Bauer and Klaus F. Zimmermann

cent of the German earnings. However, costs of living are much lower in the East than in the West. One of the main problems of the transition countries has been inflation. The third column of Table 7.1 shows the substantial differences in inflation rates between Eastern and Western Europe. Due to high inflation rates, real wages in the East increased only slightly, or fell in 1993, leading to increasing real wage differentials between the transition countries and Western Europe. Furthermore, policy measures against inflation may affect employment negatively and therefore increase the migration pressure. The GDP growth rates in 1993 indicate that the economic situation in some Eastern countries improved significantly and therefore lowered the migration pressure. On average, GDP in Eastern Europe fell between 4.8 percent in 1990 and 16.9 percent in 1992. In 1993, Albania and Poland exhibit GDP growth of 11 percent and 4 percent, respectively. In Hungary and the Czech Republic, GDP fell by 1 to 2 percent, which was also observed in some EU countries like Belgium, Germany, Portugal, and Spain. The GDP in Romania remained constant, while it fell 5 percent in Bulgaria, 6 percent in Slovakia, and 12 percent in the CIS (Economic Commission for Europe, 1994). In the long run, differences in demographic development may be an important driving force behind the upcoming era of push-migration (Zimmermann, 1995a). Stagnating, aging populations tend to attract migrants, while young and large populations generate more mobile individuals. As Table 7.2 predicts, the share of the population over the age of sixty-five will rise in all countries of the European Union. This aging process is the slowest in United Kingdom, where the size of the over-sixty-five group will increase from 15.7 percent to 19.4 percent over the 1990-2025 period, and fastest in Greece, where the size of the same group will increase from 13.8 percent to 22.2 percent in the same period. Conversely, with the exception of Ireland, the working-age (fifteen to sixtyfour) population share declines in all EU countries by 2 to 5 percentage points over the same period. Table 7.2 indicates large differences in the demographic situation and future development among Eastern European countries. Whereas Bulgaria and Hungary show a similar demographic development relative to Western European countries, all other Eastern European countries are characterized by relatively smaller age groups beyond age sixty-five and relatively larger cohorts for ages birth to fourteen. This difference in the demographic pattern between most of the Eastern European countries and the EU countries may offer a migration potential for young people in the East, due to labor shortages in the West, especially in occupations usually filled by young people (Coleman, 1993). An interesting question in the East-West migration discussion is, which countries are likely to attract which migrants? This will be largely a question of ethnic networks. Table 7.3 shows the residents in the European Union according to nationality, which offers some clues as to the existence of ethnic networks. About 5 million people from the EU live in other member states, most of them ending up in Germany, France, or the United Kingdom. In 1992, about

Table 7.3. Stock of foreign population in 1992 (thousands) Total

EU

Africa

Belgium

922.5

554.6

Denmark

169.5

28.4

France"

3,596.6

Germany

6,066.8

Central and Eastern Europe

Poland

Romania

Former USSR

188.6

24 A

6.9

8.2

40.0

7.3

4.8

—

0.9

4.9

0.9

1,311.8

1,633.1

227.0

0.6

63.0

47.1

5.1

4.7

1,487.3

236.4

553.4

550.4

271.2

92.1

51.4

Greece

213.3

61.5

20.9

39.1

35.7

11.4

4.6

12.1

Italy

537.0

111.2

170.2

85.8

20.8

9.1

5.2

2.3

Netherlands

732.9

176.1

197.7

56.8

10.2

4.6

2.0

1.2

Portugal

114.0

30.0

48.0

4.5

0.6

0.1

0.0

0.2

Spain

a

Asia

360.7

158.3

62.9

32.3

3.1

3.1

—

—

UK

2,012.4

800.5

195.4

500.1

49.3

29.4

—

14.8

EU

10,041.9

—

2,762.6

1,564.1

747.3

385.8

109.9

88.2

Austria

517.7

77.5

8.5

25.7

62.3

18.3

18.5

2.1

Finland

37.6

5.5

3.2

4.0

12.1

0.7

0.2

10.5

Sweden

493.8

72.5

22.5

83.0

31.8

16.1

5.5

3.4

1990. Source: Eurostat (1994).

278

Thomas Bauer and Klaus F. Zimmermann

300000

250000

200000

150000

100000

50000

1965

—•—

1970

Romania

1975

1980

—*--

1985

1990

former USSR

Figure 7.1: Inflow of ethnic Germans to West Germany, 1950-93. Source: Waffenschmidt(1994).

10 million people or 2.9 percent of the total EU population was from outside the EU. About 3.2 million are from Turkey and the former Yugoslavia, 2.8 million from Africa, 1.6 million from Asia, and 0.7 million from Central and Eastern Europe. People from areas outside the European Union predominantly go to Germany, and even Asia has a larger group in Germany than the United Kingdom. (Many Asians, however, carry UK passports.) Among the major European immigration countries, Germany attracts Turks and people from the former Yugoslavia, France receives Africans, and the United Kingdom attracts mainly migrants from the EU member states. According to Table 7.3, about 74 percent of all immigrants from Central and Eastern Europe are in Germany, followed by France, the United Kingdom, and Greece. Due to historical connections with Hungary, the former CSFR, and Bulgaria, Austria seems to be another important receiving country for Eastern European emigrants. However, the numbers in Table 7.3 do not include immigrants of German origin from Eastern Europe called Aussiedler who automatically become German citizens. [An extensive discussion of the migration of ethnic Germans is given by Schmidt (1994).] Figure 7.1 exhibits the immigration of ethnic Germans to West Germany since 1950. As a consequence of the collapse of the socialist regimes, the inflow of Aussiedler in West Germany jumped from 78,498 in 1987 to 202,645 in 1988,377,042 in 1989, and 397,067

Integrating the East: labor market effects

279

in 1990. Consequently, the German government altered the entry procedures for Aussiedler in 1990, requiring them to apply for entry before arrival. This measure led to a reduced immigration flow: 221,974 in 1991, 230,489 in 1992, and 218,882 in 1993. In the period 1988 to 1993,51.2 percent of the Aussiedler came from the former USSR, 35.7 percent from Poland, and 12.2 percent from Romania. At the end of 1992, 3.5 million ethnic Germans were still living in Central and Eastern Europe. To summarize, it is unclear to what extent the economic differences between Eastern and Western Europe will cause out-migration in the short run. Previous experience with migration within the EU suggests that labor mobility is rather slow in adjusting wage and unemployment differentials. In the long run, the shrinking population and the aging labor force in Western Europe and the growing population in Eastern Europe may encourage migration from the East to the West. Assuming that future migration streams will flow along existing ethnic networks, East-West migration is likely to be mainly a German problem whereas South-North migration will affect other EU member countries. 7.3

EU migration policies and options for the future Common policies of the EU

In this section we discuss the migration policies in Western Europe after the collapse of the socialist regimes in Central and Eastern Europe in 1988 and afterwards. [A survey of European migration policies after World War II is provided by Zimmermann (1994a, 1995a).] EU migration policy is marked by two different developments. Migration within the EU member states has been liberalized steadily since the original Treaty of Rome of 1957. This development has found its conclusion in Article 8a of the Single European Act, which requires the achievement of free movement of people, capital, goods, and services since January 1, 1993. This implies the abolishment of controls at the interior borders of the EU. In contrast to the policies regarding internal migration, a collective policy of the European Union with respect to immigration from outside the EU is just in its infancy. [See Zimmermann (1994, 1995a, 1995b) for a comprehensive discussion of the immigration policies of the EU.] The necessity of a common EU migration policy was coupled with the plan of a common European market, as the abolition of interior borders results in a dependency of each member state on the immigration policy of the other states. Once a foreigner enters the territory of the EU, further migration of this person cannot be controlled. The first steps toward a joint EU migration policy were the Schengen Accords of June 1985 (Schengen I) and June 19, 1990 (Schengen II) and the accord of Dublin from June 15, 1990. The main objectives of the Schengen initiatives are the elimination of internal border checks, consistent and tighter external border controls, a unified visa policy, the coordination of different national asylum

Thomas Bauer and Klaus F. Zimmermann

280

I

Countries signal Schengen and realized the agreements on 26 may 1995 Countries signed Schengen but have

! not yet realized the agreements

j EU-Countrws not signed Schengen I Other Countries

Figure 7.2: Countries incorporated in the Schengen accords.

policies, and the installation of a common information system called SIS (Schengen-Information-System). The harmonization of the visa policies only covers visas for foreigners who want to stay in the EU for no longer than three months. To obtain a visa, foreigners have to prove that they are able to bear their costs of living. The long-term visa policy was left to the national governments because it is assumed that a person cannot guarantee self-financed living for longer than three months without working in the country of destination. The agreements of Schengen II with respect to the coordination of the asylum policies are nearly identical to the accord of Dublin, which regulates the responsibility for asylum applications. If an asylum seeker enters the EU illegally, his case must be reviewed by the member state he first entered. Each contracting party has to readmit persons of third countries to their territory if they have illegally migrated to another contracting state. On the other hand, the state in which a person resides illegally can either repatriate this person to the country that the individual last entered legally or, if he has transited through another state of the Schengen group, to this state. Figure 7.2 shows the countries that signed the Schengen Accord. The initial member countries were the Benelux States, France, and Germany. Meanwhile Italy, Portugal, and Spain have joined the Schengen group, whereas Ireland, the United Kingdom, and Denmark refused to participate. Austria signed the accord in April 1995. On March 26, 1995, the Schengen accord became effective in the Benelux States, France, Germany, Portugal, and Spain.

Integrating the East: labor market effects

281

An important step toward a common EU immigration policy can be seen in the Maastricht Treaty of 1992. The main progress has been the definition of an active common immigration policy including the harmonization of visa policies, common measures against illegal immigration, and a new framework for the harmonization of the different national (political) asylum laws. With respect to the visa policy, the Maastricht Treaty is not as far reaching as Schengen II. Neither the condition for giving a visa permission, the time of validity of a visa, nor the reciprocal recognition of visa between the EU member states has been regulated. However, the EU commission gets the right to propose to its member countries which sending countries fall under visa obligations. Regarding asylum policy, it was decided to implement the accords of Dublin and Schengen. In addition, the harmonization of political asylum acceptance criteria and repatriation policies, as well as the installation of a central information agency similar to the SIS, were agreed upon. The intention of considering labor market aspects in the development of a common immigration policy has been formulated. In spite of these developments, the European Union has no explicit collective labor immigration policy. Rather there is a tendency to leave the immigration issue to the national governments. The current debate has not many economic perspectives. National immigration policies Recent national immigration policies are characterized by the realization of the agreements of Schengen II, the accord of Dublin, and the Maastricht Treaty on the one hand and measures that try to cope with the phenomenon of increasing numbers of illegal immigrants and asylum seekers on the other. For instance, in 1991 Germany reported 23,508 seized illegal migrants. This number increased to 45,675 in 1992 (Presse- und Informationsamt der Bundesregierung, 1993.) Table 7.4 shows the development of asylum seekers in the period 1988-93 and their distribution over selected European countries. In the period 1988-91, the numbers of asylum seekers increased sharply in almost all Western European countries. Table 7.4 also indicates that asylum seekers from Eastern Europe go predominantly to Germany. It is evident that the number of political asylum applicants is smaller in relatively small and poor countries and larger in countries with liberal asylum laws like Germany and Sweden. The following measures have been adopted by most of the EU members to control immigration flows more effectively: tightening of border controls, extension of visa requirements coupled with an increased number of checks on employers, and swifter processing of asylum applications (see SOPEMI, 1994). Furthermore, most EU countries do not accept asylum applications from persons originating in countries that are parties to the Geneva Convention on Refugees. In 1991, the Schengen group and Poland signed a repatriation agreement that puts the Schengen II agreement into force. Furthermore, Germany has signed similar treaties with the Czech Republic (May 1993) and Romania (No-

Table 7.4. Asylum seekers in Western European countries (thousands) Country

1988

1989

1990

1991

1992

1993

From Eastern Europe 1988-92

1993

Belgium

5.1

8.1

13.0

15.2

17.7

26.9

1.0

7.7

Denmark

4.7

4.6

5.3

4.6

13.9

14.3

2.2

1.0

France

31.7

58.8

49.8

45.9

26.8

27.6

12.7

3.5

Germany

103.1

121.3

193.1

256.1

438.2

322.8

257.3

120.4

Greece

8.4

3.0

6.2

2.7

2.0

0.9

—

—

Italy

1.3

2.3

4.8

23.3

2.5

1.5

—

—

Netherlands

7.5

13.9

21.2

21.6

17.5

35.4

8.8

3.2

Portugal

0.4

0.2

0.1

0.3

0.7

2.1

—

—

Spain

3.3

2.9

6.9

7.3

12.7

13.8

—

—

United Kingdom

5.3

15.6

25.3

44.8

24.5

22.4

—

—

Austria

15.8

21.9

22.8

27.3

16.3

4.7

49.5

0.5

Sweden

19.6

30.4

29.4

27.4

83.2

37.6

5.4

1.1

Source: UNHCR (1994,1995).

Integrating the East: labor market effects

283

vember 1992), as has Austria with the former Czechoslovakia and Poland. The agreements between Germany and Poland, the Czech Republic, and Romania include technical and financial help to compensate for the costs that arise owing to these treaties. [For instance in 1993 and 1994 Germany paid DM 120 million in financial help to Poland (Bundesministerium der Finanzen, 1994).] On July 1, 1993, Germany passed a new asylum law that states that individuals from so-called safe countries (EU-member countries, countries that are members of the EFTA, Poland, the Czech Republic, Bulgaria, Romania, the Slovak Republic, Hungary, Gambia, Ghana, and Senegal) can no longer claim political asylum. Most of the other European countries also refuse asylum seekers from these "safe" countries. As Table 7.4 demonstrates, these more restrictive political asylum policies resulted in a slight decrease of the number of asylum seekers, especially in Austria and in Germany in 1993. Whether or not asylum seekers are allowed to work during the processing of the application is handled in a different way. Whereas Germany permitted asylum seekers to work in order to reduce the high costs of financial support to these persons (in 1990 the expenses for social help and accommodation of asylum seekers was estimated to reach DM 9 billion in Germany), France prohibits asylum seekers from working in order to discourage economically motivated applications (SOPEMI, 1994). Recently, a more selective immigration policy of Western European countries can be observed. Several bilateral agreements have been signed between Western European and Eastern European countries regulating seasonal work (e.g., the 1991 agreement between the former CSFR and Germany or the 1992 arrangement between France and Poland) and reciprocal employment (e.g., between Germany and Hungary or the former CSFR and the Commission of the European Union in 1991). In most cases, these agreements deal with short-term employment only, limiting the number of workers and requiring special characteristics of the applicants. For example, there exist four possibilities for Eastern Europeans to work temporarily in Germany (see Bundesanstalt fiir Arbeit, 1994). First, it is possible to work as so-called Werkvertragsarbeitnehmer. According to these agreements, Eastern European firms are allowed to employ their own workers in project-linked work arrangements coordinated under contracts with German firms. The number of workers who can work under this treaty is usually limited by the agreements. Each year, these quotas are adapted to the labor market situation in Germany. The wage of the Werkvertragsarbeitnehmer must be the same as that of German workers. Furthermore, the validity of these work permits is restricted to three years. These kinds of agreements are mostly used in construction, iron, and steel industries. Second, if an insufficient number of German workers is available, workers from Poland, the former CSFR, Bulgaria, Romania, and Hungary can work in Germany as seasonal workers for a maximum of three months. These workers are mainly employed in agriculture and by hotels and restaurants. Furthermore,

284

Thomas Bauer and Klaus F. Zimmermann

Germany started guest employee programs with Albania, Bulgaria, Latvia, Poland, Romania, Hungary, the former CSFR, and the Russian Federation. The aim of these programs is to improve the professional and linguistic skills of the participants. The participants have to meet the requirements of completed vocational education, basic knowledge of German, and being between eighteen and forty years old. These work permits are also limited by quotas and are restricted to eighteen months. Individuals in Poland and the Czech Republic living near the German border can work in Germany as so-called Grenzarbeitnehmer if they return each day or if they work in Germany for a maximum of two days per week. Table 7.5 demonstrates that the number of workers who entered Germany through one of the first three agreements was remarkably high, 176,740 persons in 1991, 311,412 persons in 1992, and 256,534 persons in 1993. Therefore, the number of Eastern European workers who made use of one of these agreements in 1992 reached 15.3 percent of all foreign workers in Germany. Most of these workers came from Poland. In 1993 a total of 166,224 Polish worked on a temporary basis in Germany, nearly as many as Italian guest workers (200,319) and more than Spanish (55,282), Greek (121,787), and Portuguese (50,276) guest workers in September of the same year. Options for the future As a result of free labor and product markets within the EU, the member countries are unable to follow independent migration policies without potentially harming others. Therefore, the European Union should consider a unified migration policy. [A general analysis of migration policy issues is given by Straubhaar and Zimmermann (1993).] Several strategies could be pursued. First, it might be tempting for industrialized countries to consider a selective immigration policy to attract highly qualified workers needed in innovative industries. So far, foreigners in Europe are more attracted to those industries that employ lower qualified workers (Zimmermann, 1995a). To reach the goal of a selective immigration policy the Canadian and Australian immigration models, which have been shown to be very efficient in selecting migrants, could serve as a starting point. Cost-benefit considerations are at the center of the economic analysis of immigration. A country should allow immigration if the marginal productivity of the foreigners for the country is higher than their marginal costs of integration. If there are substantial immigration costs for the receiving countries, they can be compensated by imposing financial constraints on migrants to share the burden, for example, an entry fee. Moreover, these fees might be an efficient measure to control and smooth the immigration flows. Some economic theories conclude that free trade and free capital mobility could replace free labor mobility. However, in spite of the fact that a substantial part of goods and services are nontradable, import competition and foreign

Integrating the East: labor market effects

285

investments due to cheap labor in Eastern and Central Europe may also crowd out native workers. Furthermore, free trade and free capital mobility are not useful in stopping short-term immigration, because it may take a decade until free trade and the necessary capital investments become effective. Available evidence in developing countries suggests that economic transfers and development policies supporting economic growth in the sending countries are not adequate policy options, because these kinds of measures often destabilize their economies and create new economic pressures for short-term emigration. In the following we will present and calibrate a model that demonstrates that immigration might be beneficial for society as a whole even in the presence of high unemployment, because it may erode institutional constraints on the labor market. The results will demonstrate that the European Union should consider a unified policy of selective immigration: to channel and coordinate the flows of immigrants and to secure the competitiveness in international markets with high innovative activity by attracting highly qualified workers needed in these innovative industries. 7.4

Theoretical framework

This section outlines our model on which the simulations of Section 7.5 are based. If labor is homogeneous, the standard competitive framework predicts an increase of total welfare at the expense of labor, because the wage rate is lower after immigration. However, wages may not be downwardly flexible, perhaps due to the behavior of unions. [See Schmidt, Stilz, and Zimmermann (1994) for a theoretical treatment of this issue.] If union behavior remains unaffected by immigration, unemployment may rise substantially. On the other hand, unions' wage-employment choice may be affected by the pressure of immigration. If labor is heterogeneous, the key issue for the evaluation of the wage effects of immigrant labor is whether foreigners are substitutes or complements to skilled or unskilled natives. To simplify the analysis, assume that there are only two types of labor, qualified or educated workers, and less-qualified or lesseducated workers. We will call the former skilled and the latter unskilled workers. One reasonable assumption is that skilled and unskilled workers are complements. Then one scenario is that immigrants are substitutes to unskilled natives and complements to skilled natives. Hence, increased immigration may depress wages and (possibly) increase unemployment of unskilled workers and may induce the reverse effects for the skilled natives. The reverse may happen in a scenario with skilled immigration. While a formal treatment is left to Appendix A, we briefly outline the framework and provide the intuition. The economy is assumed to produce a single output according to a constant-returns-to-scale production function with capital, skilled labor, and unskilled labor. Output price is considered to be predetermined and both types of labor are complements (the standard case). Natives supply input factors at fixed levels. Immigrants are perfect substitutes either to

Table 7.5. Short-term employed workers from Eastern Europe in Germany Bulgaria

Former CSFR

Hungary

Poland

Romania

Total

Total in%

486 0.9

3,808 7.4

9,109 17.8

26,468 51.7

1,478 2.9

51,240 100.0

1992 Total in%

1,891 2.0

10,550 11.2

12,186 13.0

51,011 54.3

7,341 7.8

93,912 100.0

1993 Total in%

3,829 5.3

5,559 7.6

13,563 18.6

21,424 29.5

13,895 19.1

72,734 100.0

12,600 10.2

4,100 3.3

75,700 61.4

Werkvertragsarbeiter: 1991

Seasonal workers: 1991 Total in%

123,300 100.0

1992 Total in%

:

28,000 13.2

7,200 3.4

136,900 64.5

2,900 1.4

212,400 100.0

70P3 Total in%

—

19,800 10.9

5,300 2.9

143,900 79.5

3,900 2.2

181,000 100.0

1991 Total in%

—

400 18.2

1,400 63.6

400 18.2

—

2,200 100.0

1992 Total in%

—

2,000 39.2

2,000 39.2

750 14.7

—

5,100 100.0

7PP3 Total in%

—

1,600 27.6

1,400 24.1

900 15.5

—

5,800 100.0

Guest workers:

Source: Bundesanstalt fur Arbeit (1992, 1993,1994).

288

Thomas Bauer and Klaus F. Zimmermann

unskilled natives or to skilled natives. They do not carry any capital with them and have no effect on the demand side of the economy. The level of immigration relative to the native population is fixed by governmental rules, and we will concentrate here on pure labor immigration. A monopoly union sets the wage in the market for unskilled labor, and employers then choose the level of employment in this market, whereas the market wage of skilled labor is determined by competitive forces. Nevertheless, the union cares about the wages of the skilled workers, which are affected by the employment level determined in the market of the unskilled workers. This spillover is generated by a standard neoclassical production technology. The consequences of skilled or unskilled labor immigration in such a model can be studied by use of Figure 7.3. The upper panel considers the case of immigration of unskilled labor. While the labor market of the skilled is controlled by competitive forces (Ao in Figure 7.3a), the monopoly union sets a higher than equilibrium wage in the market for unskilled labor (Bo in Figure 7.3b). This causes unemployment at level L - L for the unskilled. The union is concerned about the earnings of both the skilled and unskilled workers. Upon unskilled immigration (see the shift of the labor supply curve in Figure 7.3b), it therefore accepts a lower wage level for the unskilled (Z?7). Since both types of labor are complements, the increased unskilled employment (L7) shifts the demand curve for skilled workers upward (see Figure 7.3a), and the wage rate of skilled workers is increasing (Ay in Figure 7.3a). As a result, the union unskilled wage falls and drives the economy in the direction of the equilibrium point of a competitive labor market. In general, native unemployment may rise or fall. However, according to equation (A8) in the parametric framework chosen in Appendix A, the employment effect for unskilled natives is negative. The case of immigration of skilled labor is even more obvious. The increasing stock of skilled labor (see the shift of the supply curve in Figure 7.3c) drives the equilibrium point down from Co to Cr The demand for unskilled labor increases due to complementarity (see the shift of the demand curve in Figure 7.3d), and there will be a higher level of employment of unskilled workers, whether or not the union decreases or increases the unskilled wage. While it does not seem plausible that the union increases the unskilled wage strongly so that native employment falls, the theoretical model in Appendix A even predicts that wages will fall [Dj in Figure 7.3d and equation (A9) in Appendix A]. The increased level of unskilled employment again shifts the demand curve for skilled labor upward (C2 in Figure 7.3c). Hence, immigration of skilled workers will likely cause a decrease of the wages of the unskilled and a decline of native unemployment. This analysis suggests that there are complicated issues that determine whether one can expect gains from immigration and which groups will receive them. In a competitive (equilibrium) framework in both labor markets, natives in total will receive gains, but those workers who are substitutes to immigrants will lose. In the union model outlined here effects are similar in nature. If un-

Integrating the East: labor market effects

289

Immigration of Unskilled Labor (a) Skilled

(b) Unskilled Wiige i

w age i

Immigration

V

7

\

\ v

o

Ao

w

Sn

i

\

Y

L. I

Labor

Immigration of Skilled Labor (c) Skilled

(d) Unskilled Immigration

Wage

Wj

So

Sj

Labor

L, L

Labor

Figure 7.3: Theoretical framework.

skilled labor immigrates, there will be gains for skilled natives, but unskilled natives receive lower wages and face higher unemployment. To what extent natives still receive gains in total depends on the concrete situation. In the case of skilled labor immigration, both wages and unemployment will decline, and total income of natives will increase. How important are the derived effects? For this purpose one has to calibrate the model. The next section will carefully study the polar cases and provide some estimates of the potential effects. Since the previous sections have forcefully stated that Germany is the major European immigration country, and most

290

Thomas Bauer and Klaus F. Zimmermann

Table 7.6. Gains of immigration: full employment model (billion DM) Gains of Natives

Gains of Migrants

Total

Skilled

Unskilled

Skilled

Unskilled

91.822

17.078

51.420

-61.888

0

74.743

A = 0.5

132.044

3.372

-6.763

-29.479

80.708

47.965

A =1.0

156.934

4.998

-64.053

21.971

151.936

0

A=0

50.181

4.270

28.101

-38.886

0

45.911

A = 0.5

66.865

0.843

-2.974

-16.243

40.418

25.604

A =1.0

79.716

1.249

-33.826

11.160

78.467

0

A=0

10.719

0.171

6.003

-9.048

0

10.548

A = 0.5

13.508

0.034

-5.530

-3.489

8.094

5.380

A =1.0

16.143

0.050

-7.053

2.260

16.093

0

Total Inflow =10% A=0

Inflow = 5%

Inflow = 1 %

Notes: The calculations assume that national income equals DM 2,108 billion, that the share of national income accruing to unskilled workers is 14 percent, that of skilled workers 56 percent, and that of capital 30 percent. The income shares are assumed to remain unchanged during immigration. The elasticity of factor price for the unskilled is -0.85, that for the skilled is -0.45. The elasticity of the wage of skilled workers with respect to a change in the quantity of unskilled is 0.15; that of unskilled workers with respect to a change of skilled workers is 0.55. Furthermore, it is assumed that 27.1 percent of the native labor force is unskilled and 72.9 percent is skilled. These numbers are all derived from German data for 1993 as explained in the text. The fraction of skilled migrants is denoted by A.

Eastern European migrants move to Germany anyway, we use German data to calibrate the effects. 7.5

Calculating the gains from immigration

In this section we evaluate the benefits from immigration. At first, a simple equilibrium model with full employment is used, which is described in more detail in Appendix B and follows closely the work of Borjas (1995). In a second step, the calculations are then modified to deal with a situation where the unskilled labor market is in disequilibrium. Here we draw on the theoretical model outlined in the last section. The calibration is done using the following assumptions based on 1993 German data: The national income in 1993 equals DM 2.108 trillion. The share of national income accruing to unskilled workers is 14 percent, that of skilled workers 56 percent, and that of capital 30 percent. These numbers are derived from the German microcensus and the national accounts of the German statistical office. Assuming a Cobb-Douglas approximation, we calculate from these numbers the factor price elasticity for unskilled labor as

Integrating the East: labor market effects

291

Immigration of Unskilled Labor (a) Skilled

(b) Unskilled

Wage k

Immigration

\N

\ \

a

\ \ \ i

\

\

b

^C \

Labor

Immigration of Skilled Labor (c) Skilled

(d) Unskilled Wage

So

S/

Labor

LQ

Labor

Figure 7.4: Calculation of the gains from immigration.

—0.85 and for skilled work as —0.45. The elasticity of the wage of skilled workers with respect to a change in the quantity of unskilled workers is 0.15, and the elasticity of the wage of unskilled workers with respect to a change of skilled workers is 0.55. It is further calculated that 27.1 percent of the work force is unskilled and 72.9 percent skilled. These numbers are kept fixed throughout the analysis. Table 7.6 contains the calculated gains from immigration using the simple equilibrium model with full employment and different levels of immigration. Figure 7.4 demonstrates how these gains are calculated: If only unskilled migrants are accepted (X = 0), the total benefit of immigration is given by the sum

292

Thomas Bauer and Klaus F. Zimmermann

200

0

0.1

0.2

03

0.4 05 0.6 Shoe of SkflkdMgrwtt

• Skilled

•

Unskilled

Capital

Figure 7.5: Total immigration gains according to production factors at 10 percent inflow. For notes, see Table 7.6.

of the areas A, B, and C, where area B is allocated to immigrants and areas A and C to natives. It should be noted that the benefits of the immigrants refer only to their income in the receiving country. To calculate the net benefits from migration for the migrants, their forgone earnings in the sending country and the migration costs have to be subtracted from the numbers calculated in this study. As Table 7.6 indicates, the total gain is calculated to be DM 91.822 billion or 4.36 percent of national income in 1993 if 10 percent of the native work force (2.758 million persons) immigrates. If 50 percent of the immigrants are skilled (X = 0.5), this gain increases to DM 132.044 billion (6.26 percent of the national income) and reaches a maximum of DM 156.934 billion (7.44 percent of the national income) if all immigrants are skilled (X = 1.0). In the case of skilled immigration, the gains of natives in Figure 7.4 are the areas E and F, and the income of immigrants is area I. Table 7.6 further allocates the total gain to skilled and unskilled natives and

293

Integrating the East: labor market effects 18

Inflow 10%

0

0.1

0.2

0.3

0.4 0.5 0.6 Share of Skilled Migrants

0.7

0.8

0.9

1

Figure 7.6: Immigration gains of natives and immigrant skills. For notes, see Table 7.6.

immigrants, respectively. Inspection of Figure 7.5 shows that irrespective of the skill composition of the immigrants, most of this gain goes to skilled workers, followed by capital and unskilled workers. Due to the higher average wage of skilled workers, the highest immigration gain for immigrants is reached when only skilled persons move. Figure 7.6 reveals that the gain from immigration for natives in the full employment model reaches a maximum of DM 17.078 billion (0.81 percent of the national income) when only unskilled immigrants are accepted and reaches a minimum when 70 percent of the immigrants are skilled. In a similar study for the United States, Borjas (1995) estimated a maximum immigration gain for natives of 2.4 percent of the GDP. This implies that the gains of migration are much smaller in a European setting. Table 7.7 exhibits the calculated gains from immigration if alternative elasticities are used. The results indicate that the simulated effects of Table 7.6 are rather stable. Figure 7.7 shows the distributional effects of immigration in the equilibrium model, which could be quite dramatic. It appears that capital always bene-

Table 7.7. Gains of immigration: full employment at 10 percent inflow (billion DM) Gains of Natives Total

Gains of Migrants

Total

Skilled

Unskilled

Skilled

Unskilled

e^ = -0-45; elL = -0.85; eSL = 0.15; eLS = 0.55 1=0

91.822

17.078

51.420

-61.888

0

74.743

A. = 0.5

132.044

3.372

-6.763

-29.479

80.708

47.965

1=1.0

156.934

4.998

-64.053

21.971

151.936

0

10.046

55.358

-74.968

0

88.808

e s s = -0.5; eLL = -0.5; e SI = 0.15; e t s = 0.55 1 =0

98.854

1 = 0.5

133.663

1.752

-5.578

-32.768

80.430

51.481

1 = 1.0

156.378

5.553

-63.253

21.893

150.825

0

24.111

47.482

-48.808

0

60.679

€SJ = -0.2; eLL = -1.2; eSL = 0.15; e t i = 0.55 1 =0

84.790

1 = 0.5

130.980

4.436

-8.747

-26.112

0.039

44.449

1=1.0

159.710

2.221

-68.051

22.359

0.075

0

ess = -0.45; eLl = -0.85; eSL = 0.01; € IS = 0.06 1 =0

91.822

17.078

51.420

-61.888

0

74.743

1 = 0.5

130.084

5.332

-5.769

-27.923

78.616

46.135

1=1.0

156.934

4.998

-64.053

21.971

151.936

0

17.078

51.420

-61.888

0

74.743

€ M = -0.45; eLl = -0.85; e i t = 0.3; 6^=1.1 1 =0

91.822

1 = 0.5

134.192

1.224

-7.801

-31.232

82.948

50.019

1=1.0

156.934

4.998

-64.053

21.971

151.936

0

Notes: See Table 7.6. essand e,, denote the elasticities of factor prices of skilled and unskilled workers, respectively. eSL denotes the elasticity of skilled wages with respect to a change in the quantity of unskilled workers, and eLS denotes the elasticity of unskilled wages with respect to a change in the quantity of skilled workers.

296

Thomas Bauer and Klaus F. Zimmermann

0

0.1

0.2

0.3

0.4 0.5 0.6 0.7 Share of Skilled Migrants

0.8

0.9

Figure 7.7: Immigration gains of natives according to production factors at 10 percent inflow. For notes, see Table 7.6.

fits from immigration and that these benefits increase with the share of skilled immigrants. The gains of capital reach DM 47 billion or 2.2 percent of national income if only skilled immigration of 10 percent of the total native labor force occurs. Skilled native workers show a positive immigration gain as long as no more than 44 percent of the immigrants are skilled, and unskilled native workers benefit from immigration only if more than 80 percent of the immigrants are skilled. Table 7.6 and Figure 7.7 also reveal that both types of labor could lose very much through immigration, depending on the share of immigrants that substitute for them. For instance, if 10 percent of the native work force immigrates and all immigrants are skilled, skilled native workers lose 5.4 percent of their initial income. The maximum loss of unskilled native workers is calculated to reach DM 62 billion or 21 percent of their initial income in the case of unskilled immigration (see Table 7.6). Table 7.8 examines the change of calculated immigration gains from the full employment model if unemployment of unskilled natives and different reactions of the union to immigration are taken into consideration. (Exact formulae to calculate the gains of immigration in the disequilibrium framework are available from the authors on request.) Here we refer to the theoretical Section 7.4 and Appendix A. First, we consider the case that 10 percent of the native work

Integrating the East: labor market effects

297

force immigrates and that all immigrants are unskilled. If the union lowers the wage of unskilled workers such that native unemployment remains constant, the immigration gain is the sum of the areas A, B, and C in Figure 7.4 (scenario Alia in Table 7.8). The immigration benefits in this situation are the same as in the full employment model. Scenario Allb in Table 7.8 represents the situation where the union keeps the wages of unskilled labor fixed, and immigration leads to an equal increase in native unemployment. Hence, natives lose area D in Figure 7.4 or DM 116.670 billion (calculated as w 0 multiplied by the number of immigrants), which equals the income of the immigrants. In this extreme scenario the total immigration gain is zero and the losses of natives are maximized. To obtain more moderate solutions, or a partial crowding-out of unskilled natives, one has to accept further assumptions. Based on the model outlined in Section 7.4 and Appendix A, the strength of the wage response of the union, and hence the increase in unskilled employment depends on p, the weight of native unskilled employment in the union's objective function. Considering different values of p (0.3, 0.5, 0.7) in Table 7.8, it becomes clear that the immigration gain is decreasing the closer P is to one. The theoretical basis for this finding is that the larger p, the more modest union wage policies are. Unions therefore react less responsively under migration pressure. If only skilled immigration is considered, the union also has several possibilities to react to the resulting increased demand for unskilled workers. One extreme reaction is described by Scenario Blla which considers the case that the union increases the wage in such a way that native unemployment remains constant. Figure 7.4 reveals that in this case the gains of natives remain to be the same as in the full employment model (sum of area E and F). The other extreme case (scenario Bllb in Table 7.8) is caused by a wage reduction resulting in zero unemployment. In this case the gains of natives from immigration are calculated as the sum of the immigration gain when only skilled persons immigrate (areas E and F in Figure 7.4) plus the total immigration gain if only unskilled immigration occurs that equals the number of unemployed natives (areas G and H in Figure 7.4). This case, which describes the maximum benefit from immigration for natives and in total, is estimated to result in a total immigration gain of DM 240.561 billion (11.41 percent of the national income) of which DM 88.627 billion (4.20 percent of the national income) is received by the natives. Similar to the case of unskilled immigration, the immigration gains decrease with p, the union's weight for unskilled employment. The last column of Table 7.8 exhibits the resulting increase of native immigration benefits if tax and the unemployment insurance payments are considered. These numbers include the tax and unemployment insurance contributions of immigrants as well as the costs and benefits of the unemployment insurance system that result from the variation in native employment levels. The resulting gains of natives are substantial. They reach DM 15 to 17 billion in the case of unskilled immigration and DM 41 to 44 billion if only skilled immigrants are accepted.

Table 7.8. Gains from immigration and the disequilibrium model (billion DM) Natives

Migrants

Total

Tax and Unemployment Insurance Gains"

A. Unskilled Immigration (I) Equilibrium

17.078 (0.81)

74.743 (3.55)

91.822 (4.36)

16.623 (0.79)

17.078 (0.81)

74.743 (3.55)

91.822 (4.36)

16.623 (0.79)

-116.670 (-5.53)

116.670 (5.53)

0 (0.00)

15.325 (0.73)

P = 0.3

-89.057 (-4.22)

113.100 (5.37)

24.043 (1.14)

16.963 (0.80)

P = 0.5

-97.773 (-4.64)

114.584 (5.44)

16.811 (0.80)

16.542 (0.78)

p = 0.7

-107.955 (-5.12)

115.910 (5.50)

7.954 (0.38)

15.941 (0.76)

(II) Disequilibrium a) Constant native unemployment

b) Immigration equals increase in native unemployment c) Partial crowding-out of unskilled natives

B. Skilled Immigration (I) Equilibrium

4.998 (0.24)

151.936 (7.21)

156.934 (7.44)

40.931 (1.94)

4.998 (0.24)

151.936 (7.21)

156.936 (7.44)

40.931 (1.94)

88.627 (4.20)

151.936 (7.21)

240.566 (11.41)

44.430 (2.11)

P = 0.3

96.138 (4.56)

151.936 (7.21)

248.074 (11.77)

44.430 (2.11)

P = 0.5

61.200 (2.90)

151.936 (7.21)

213.136 (10.11)

42.913 (2.04)

P = 0.7

43.962 (2.09)

151.936 (7.21)

195.897 (9.29)

42.262 (2.00)

(II) Disequilibrium a) Constant native unemployment

b) Zero native unemployment

c) Partial expansion of unskilled employment

Notes: See Table 7.6. Immigration of 10 percent of the native work force is assumed. (5 denotes the weight of native unskilled employment in the objective function of the union (see Appendix A). " These numbers include the unemployment insurance contributions of immigrants calculated as 0.0384 times their immigration gain, the tax contribution of immigrants calculated using the average tax rate for skilled and unskilled workers, respectively, and the costs or benefits of the unemployment insurance system resulting from the variation in native employment calculated as average unemployment benefit in 1993 (1,421 DM) in Germany times the variation in unemployment.

300

Thomas Bauer and Klaus F. Zimmermann

7.6

Conclusions

This paper has studied the perspectives and implications of East-West migration in Europe. It is argued that Western Europe, especially Germany, has seen significant inflow of migrants, and that this experience was mostly beneficial. Further migration is seen as largely unavoidable, at least from countries such as Poland, the Czech and Slovak Republics, and Hungary. There is substantial political pressure for integration into the European Union. Relying on the phenomenon of network migration, most of the migration pressure is expected to be directed to countries like Germany and Austria. However, the currently high unemployment rates in Western Europe seem not to permit liberal immigration policies. Consequently, a survey of the migration policies of the European Union has not revealed much flexibility, although one important member country (Germany) is executing some significant temporary immigration policies directed to Eastern Europe. The issue is whether immigration in the face of unemployment automatically causes problems for the labor markets of the receiving country. The conclusion here is that this is not the case. This result is obtained in two steps: First, a theoretical framework provides a setting with heterogeneous labor where skilled labor is in a competitive equilibrium and unskilled labor is in a disequilibrium with unemployment. Unskilled immigration may cause an increase in the unemployment of natives, but also a decline in the wages of the unskilled, and hence increase total employment. Skilled immigration will likely cause a decline of both types of wages and decrease unskilled native unemployment. Second, the framework is calibrated to obtain some feeling for the size of the potential gains and losses. Since Germany is likely to take a larger share, the calibration is done using the most recent available German data (from 1993). The traditional full equilibrium model serves as a benchmark case. Such a border case reveals gains for the receiving country; however, they are much smaller than often expected. A 10 percent increase in the labor force (about 2.758 million immigrants) would lead to direct gains for the natives of about 0.24 percent of national income in the case of skilled immigration and to about 0.81 percent (or about DM 17 billion) in the case of unskilled immigration. However, total benefits of natives and foreigners are much larger (7.44 percent) for skilled migration than for unskilled migration (4.36 percent). Also, tax payments and unemployment insurance contributions of migrants are much larger for skilled migration, making this option more valuable in practice, at least if one considers only temporary immigration. In a disequilibrium framework for unskilled labor, unskilled immigration is a rather risky strategy if one has no safe predictions about the employment effects. Losses could be up to 5 percent of national income. However, skilled immigration seems to be a valuable option even in the face of unemployment. If both types of labor are complements as assumed in the present analysis, there could be substantial gains due to the improvement of the employment possi-

Integrating the East: labor market effects

301

bilities of unskilled natives. Gains could be up to 4 percent of national income at current unemployment rates, which would add to the 2 percent gains from tax payments and unemployment insurance contributions of migrants. Migrants themselves would receive an income of about 7 percent of national income. Given the current excess supply of qualified workers in Eastern Europe and the need for further improvements of their human capital in the process of transformation, a temporary immigration policy as executed currently by the German government seems to be a valuable option for both the East and the West. Appendix A The model assumes an economy that produces a single output according to a constant-returns-to-scale production function with capital, skilled labor 5, and unskilled labor L. The output price is considered to be predetermined and both types of labor are ^-complements (the standard case). Natives supply input factors at fixed levels. The level of immigration M is fixed by governmental rules. To simplify the analysis it is assumed that immigrants do not carry any capital with them and have no effect on the demand side of the economy. Two polar cases are considered: Migrants are perfect substitutes either to the unskilled or to the skilled workers. A monopoly union sets the wage w on the market for unskilled labor and employers then choose the level of employment in this market. The wage v of skilled workers is determined by competitive forces. Nevertheless, the union cares for them. Employed unskilled natives are N = ctL, where a = N/(N + M), and M = (1 - a)L. The objective function of the union is given by: ^ f t = vS + wN*9

0 < p < 1,

(Al)

where S and N are the fixed levels of skilled and unskilled natives, and P is a weight for the employment of unskilled workers. Profit maximization of the firm implies that real wages are equal to marginal productivity. Suppressing the equation for capital it follows: v = v(S,L)

(A2)

L = L(w, S).

(A3)

S is predetermined to the model, w is predetermined by the monopoly union, and v is fixed by a competitive market. vs, Lw<0 and vL, Ls > 0. Second derivatives are assumed to be zero so that (A2) and (A3) are linear. The union's problem is to maximize ft with respect to w. Hence, the firstorder condition implies dL VV dv 1 — S + W*(P W*(P - ll )) ^^" 11 + S aw 1 6L \

// ddL \ [L + w — Q^LP- 1 = 0. [L \ dw)

(A4)

302

Thomas Bauer and Klaus F. Zimmermann

Considering the situation before immigration (a = 1), (A4) could be expressed as: L^VL?

NLJ

= 0,

(A5)

where dL w

=

^ ^r

dv L

and e =

- az: v •

Optimality requires 1

+ Ni,,, > 0.

(A6)

At first we want to assume unskilled immigration, which affects (A5) by a variation of a. The comparative statics lead to: - ( 1 4- BTK ) * - ^

w — =

> 0,

(A7)

w = dw/w and a = da/a. From 1

3a and ^ = —

(with 3M = M)

it follows: w - =

M

(1 +

BTI, ) ^-^^ < 0. W 2 - 1^(1 - P)]

(A70

One also obtains: - ( 1 - TI, ) i ! ^ ! _ < o. M 2 - -nLvv(l - 3) Skilled immigration can be modeled by a variation of S. Hence: N

=

± I

=

(i + N ^ ) - N t 5 + 0 - P ) N U ^ (1 p)] P J 2

(A8)

Integrating the East: labor market effects

303

In general, equation (A9) may take any sign. Reasonable sizes of the elasticities, however, imply: (A10)

so that w/§ < 0 in equation (A9). Straightforward derivations lead to: N

1°

1 + 15%,, + N u

»P-V.O-M-

„ > 0

The sign of (All) follows directly from the optimality condition (A6). Appendix B To calculate the gain from immigration in the equilibrium framework we follow closely the work of Borjas (1995). We assume a concave and linear homogeneous production function: Y = f(K, 5, L) = f(K, pN + XM, (1 - p)N + (1 - X)m),

(Bl)

where Y refers to the output, K to capital, S to skilled workers, L to unskilled workers, and M to the immigrants, p and X give the fraction of skilled workers among natives and immigrants, respectively. If the wage of each production factor is determined by the respective marginal productivity, the increase in national income through immigration accruing to the natives is: i

(B2)

Defining e r = d log q. I d log X as elasticity of factor price, using the restriction that 2 . e.. = 0 (Hamermesh, 1993, p. 37) and converting equation (B2) in percentage terms one obtains (see Borjas, 1995): - X)2m2

Jc€c,X(l - X)m2

- X)m2 2t f

sL

(B3)

where ys and yL are the shares of national incomes accruing to skilled and unskilled workers, m is the fraction of immigrants to the total labor force, and ts and tL are the shares of the work force that are skilled and unskilled, respectively.

304

Thomas Bauer and Klaus F. Zimmermann

If only skilled (X = 1) or unskilled (X = 0) immigration is considered, (B3) is reduced to:

1

Ll

1

Ll

(B4)

s

and (B5) L

respectively. It is evident that the immigration gain of both types of native labor is the higher, the higher their initial share of national income, the higher the absolute value of the elasticity of factor price, and the higher the fraction of migrants to skilled or unskilled native workers (m2 I t2v with / = S, L), respectively. In the simple case of only one type of labor, (B3) is reduced to: yem2

AYAT

which corresponds to triangle C in Figure 7.4. The wages qt with i = v, w after immigration can be calculated as:

*<• - i f (' + '•• r )

< B7 >

If we assume that the immigrants bring no capital with them and use (B2), the income accruing to immigrants can be calculated as: |

yL{\ - \)m

where the first three terms show the income accruing to the skilled immigrants and the second three terms the income accruing to the unskilled immigrants. The total effect of immigration on the production of the receiving country is: AY _AYN

+

AYK

Y VcXm vceccX2m2 v c € c / X ( l — X)m 2 ys s ys SL = ^— + \ + — + ,(1 - X)2m2 -^

n

^/e,oX(l - X)m2 — — •

(B9)

Integrating the East: labor market effects

305

If we assume that the income shares of the production factors is not changed by immigration, the immigration gain of skilled and unskilled natives can be calculated by subtracting the corresponding income of immigrants from the total gain. Because of the Euler Theorem the rest of the immigration gain is owned by the capital.

REFERENCES Bauer, T., and K. F. Zimmermann (1995). "Modeling International Migration: Economic and Econometric Issues," in R. van der Erf and L. Heering (eds.), Causes of International Migration: Proceedings of a Workshop, Luxembourg, 14-16 December 1994, Luxembourg: Office for Official Publications of the European Communities, 95-115. Borjas, G. J. (1995). "The Economic Benefits from Immigration," Journal of Economic Perspectives 9 No. 2, 3-22. Bundesanstalt fur Arbeit (1994). Amtliche Nachrichten der Bundesanstaltfiir Arbeit: Arbeitsmarkt 1993, Niirnberg: Bundesanstalt fiir Arbeit. Bundesanstalt fiir Arbeit (1993). Amtliche Nachrichten der Bundesanstaltfur Arbeit: Arbeitsmarkt 1992, Niirnberg: Bundesanstalt fiir Arbeit. Bundesanstalt fiir Arbeit (1992). Amtliche Nachrichten der Bundesanstaltfiir Arbeit: Arbeitsmarkt 1991, Niirnberg: Bundesanstalt fiir Arbeit. Bundesministerium der Finanzen (1994). Finanzbericht 1994, Bonn: Bonner Universitats-Buchdruckerei. Coleman, D. A. (1993). "Contrasting Age Structures of Western Europe and of Eastern Europe and the Former Soviet Union: Demographic Curiosity or Labor Resource?", Population and Development Review 19, 523-55. Deutsche Bundesbank (1994). Devisenkursstatistik Februar 1994, Frankfurt/Main: Deutsche Bundesbank. Deutsches Institut fiir Wirtschaftsforschung (1994a). "Die Lage der Weltwirtschaft und der deutschen Wirtschaft im Friihjahr 1994," DIW Wochenbericht, 16-17,229-62. Deutsches Institut fiir Wirtschaftsforschung (1994b). "Die wirtschaftliche Lage Ruplands: Fortsetzung des Niedergangs ohne hinreichenden Strukturwandel," DIW Wochenbericht, 47-18, 805-27. Economic Commission for Europe (1994). Economic Survey of Europe in 1993-1994, New York: United Nations Publication. Europaische Kommission {1994). "Statistischer Anhang," Europdische Wirtschaft 58, 113-94. Eurostat (1994). "Auslander machen iiber 4% der Gesamtpopulation der Europaischen Union aus," Eurostat Schnellberichte: Bevolkerung und soziale Bedingungen 7, 1-11. Eurostat (1991). Bevolkerungsstatistik 1991, Luxembourg: Eurostat. Hamermesh, D. S. (1993). Labor Demand, Princeton, N. J.: Princeton University Press. International Organization for Migration (1991). "Ninth IOM Seminar on Migration: South-North Migration," International Migration 29. Presse- und Informationsamt der Bundesregierung (1993). "Die Kriminalitat in der Bundesrepublik Deutschland," Bulletin 40, 349-88. Schmidt, C. M. (1994). "The Economic Performance of Germany's East European Immigrants," Miinchener Wirtschaftswissenschaftliche Beitrdge, Nr. 94-09. Schmidt, C. M. (1996). "Cohort Sizes and Unemployment: Lessons for Poland," in H. Lehmann and J. Wadsworth (eds.), Labour Markets by Design? ifo Studies of Eastern Europe and the Economics of Transition, No. 21, 126-54.

306

Thomas Bauer and Klaus F. Zimmermann

Schmidt, C. M , and K. F. Zimmermann (1992). "Migration Pressure in Germany: Past and Future," in K. F. Zimmermann (ed.), Migration and Economic Development, Berlin: Springer-Verlag, 201-30. Schmidt, C. M., A. Stilz, and K. F. Zimmermann (1994). "Mass Migration, Unions, and Government Intervention," Journal of Public Economics 55, 185-210. SOPEMI (1994). Trends in International Migration: Annual Report 1993, Paris: OECD. Statistisches Bundesamt (1995). "Lohne, Gehalter und Arbeitskosten im Ausland," Lbhne und Gehalter, Fachserie 16, Reihe 5, Stuttgart: Metzler-Poeschel. Statistisches Bundesamt (1994). Statistisches Jahrbuch 1994 fur das Ausland, Stuttgart. Straubhaar, T., and K. F. Zimmermann (1993). "Towards a European Migration Policy," Population Research and Policy Review 12, 225—41. UNHCR (1995). Die Lage der Fluchtlinge in der Welt: UNHCR-Report 1994, Bonn: J. H. W. Dietz Nachfolger. UNHCR (1994). Populations of Concern to UNHCR: A Statistical Overview 1993, New York: UNHCR. United Nations Population Fund (1994). The State of World Population 1994, New York: United Nations Population Fund. United Nations Population Division (1992). World Population Prospects 1992/1993, New York: United Nations. Waffenschmidt, H. (ed.) (1994). "Zahlen, Daten, Fakten," Info-Dienst Deutsche Aussiedler, 57. Zimmermann, K. F. (1995a). "Tackling the European Migration Problem," Journal of Economic Perspectives 9,45-62. Zimmermann, K. F. (1995b). "Immigration Policies in Europe: An Overview," in H. Siebert(ed.), Migration: A Challenge for Europe, Tubingen: J. C. B. Mohr, 227-58. Zimmermann, K. F. (1994). "European Migration: Push and Pull," Proceedings of the World Bank Annual Conference on Development Economics 1994, Supplement to The World Bank Economic Review and The World Bank Research Observer,

313-42.

Comment Barry Bosworth

The authors are to be commended for this effort to quantify the labor market effects of immigration. I am somewhat surprised, however, by the magnitude of the gains to the recipient country that they obtain. In understanding the results it may be helpful to step back from their more sophisticated model and start with the simplest formulation. Most prior analyses have concluded that the net effects on the native population are small: that the immigrants receive nearly all of the increment to output in wages; and the implications for the natives are largely redistributional. Existing workers lose through a fall in their wage rate, but this is offset by an increased return to capital, which has become relatively more scarce. The effects on the resident population are not completely redistributional, however, because the gain to capital exceeds the loss to labor. It is this net gain, the immigrant surplus, that the authors find to be significant in some of their disequilibrium analysis. The analytics can be illustrated graphically as an outward shift of the labor supply that lowers the wage rate (Figure 7.8). The gain to immigrants is given by the area BNN'C, and the loss to resident workers by the area WW'BA. The gain to capital is given by the redistribution away from resident workers plus the triangle ABC, the immigrant surplus. At the level of the total economy, the size of the immigrant surplus can be defined as S = \yem2,

(1)

where y is labor's share, e equals the elasticity of the wage with respect to a change in the quantity, and m is the share of immigrants in the total labor force. Labor's share can be taken to be about 0.7, and typical values for the elasticity of labor demand range between 0.2 and 1.0. Thus, immigration equal to 10 percent of the total workforce would offer a net gain to residents of 0.14 to 0.35 percent of GDP, a trivial surplus. Furthermore, a large surplus can only come at the cost of a large distributional effect, a large value for e. Thus, it is not surprising that the debate over immigration has centered around the distributional issues, not the net surplus. Bauer and Zimmerman alter this analysis in two respects. They divide the 307

308

Barry Bosworth

w

N

N'

Figure 7.8: The immigrant surplus.

labor force into skilled and unskilled workers, and they try to model the behavior of a labor monopsonist that sets the wage for unskilled labor above the market clearing level. In considering the case of two types of labor, they rely heavily on some work by Borjas. The mathematical expression in the appendix seems complex, but it can be simplified by considering a situation in which all of the immigrants fall within one class of workers. The formula for the immigration surplus is very similar to that above, but for the addition of the proportion of the native workforce in the category (t) in the denominator: = (yem2)/2t2.

(2)

It appears to imply that concentrating the immigration in a small segment of the labor force will substantially increase the net gain to the resident population. While the income share term will decline relative to that used in the simple case, the expression is dominated by the change in the denominator. Reduced values for t will give rise to large increases in the value of the net immigration surplus. The paper would benefit from some discussion of why this should be so. I am concerned that it may be a property of some of the approximations used to derive the basic formula. The second extension of the paper is the inclusion of monopoly union wage setting for unskilled workers. Again, the specific formulation seems somewhat unconventional in that the union preference function involves a maximization of the sum of the wage bills for skilled and unskilled workers, but with a discounting of the employment of unskilled labor [equation (Al) in Appendix A]. Because skilled and unskilled labor are complements, immigration in one segment increases demand in the other. Thus, if the union sets the unskilled wage above the market clearing level, the resulting unemployment can be eliminat-

Comment

309

ed by increasing the immigration of skilled workers sufficiently to raise the demand for unskilled workers to a level consistent with the monopsony-set wage rate. As the authors state, the conclusion that immigration of skilled workers could solve Europe's unemployment problem is very dependent on the assumption that the two types of labor are strong complements. One might be a little doubtful of its actual policy importance, yet it is an interesting result. With these two modifications the authors have been able to sharply raise the net benefits to the resident population of immigration. It is also interesting that the benefits of admitting skilled immigrants have nothing to do with the usual emphasis on the savings in educating additional resident skilled workers. Yet the model still produces very large distributional consequences, and they, I believe, lie at the center of the debate over immigration. While the authors' construction of a model that provides substantial net benefits to the recipient country is quite ingenious, the simple analysis may provide a better guide to the issues that actually underlie the public debate over immigration policy in both Europe and the United States. The lasting value of the paper lies in the sensitivity analysis that the authors provide with respect to variations in the magnitude of some of the elasticities, a subject for which our empirical knowledge remains very limited.

Comment Robert LaLonde

Thomas Bauer and Klaus Zimmermann have written a very thoughtful paper on the labor market consequences of increased immigration from Eastern Europe into the EU. Although the United States has always referred to itself as a nation of immigrants, its status as an immigrant nation is not unique among Western economies. For example, foreign-born persons in Germany now account for 7.4 percent of the population, a percentage that is comparable to the corresponding percentage in the United States. Not surprisingly, in recent years debates about immigration policy have intensified throughout Europe. As has been the case historically in the United States, the principle concern underlying these debates is not about the impact of increased immigration on the economy as a whole, but its impact on different groups of natives within the EU. In the United States, the evidence indicates that increased immigration has had its largest impacts on wages, particularly the wages of other immigrants. The impacts on employment and unemployment rates appear to be nonexistent, though some researchers report that natives' mobility among cities is sensitive to immigration rates. Overall, however, the impact of immigration on wages is relatively small, even on the wages of relatively unskilled natives who would seem to be the closest substitutes to immigrants in the labor market. Instead, the main redistributive impact of immigration appears to be its effects on government budgets and transfer programs. Overall, natives gain from this calculation, though the gains are substantially less than one percent of national income. The character of German economic and labor market institutions, as well as those in several EU countries, suggests that an analysis of the economic impact of immigration in Europe may differ from a similar analysis of immigration to the United States. Because German wages are less flexible (at least in the short run) than those in the United States, the impacts of increased immigration are more likely to be observed on the "quantity" side of the labor market. The prospect of increased unemployment rates for natives as a result of increased immigration underscores an additional redistributive concern. Namely, even among the unskilled or those workers who compete most closely with new immigrants in the labor market, the "costs" of immigration fall disproportionate310

Comment

311

ly on a few "displaced" natives. Such costs would be spread out among a larger population if the primary adjustment to increased immigration was lower wages. In principle, the redistributive issues underlying the debates over free trade and immigration should be the same. However, in the United States the fervor and focus of the debates on these issues has differed. The prospect of heavily skewed losses for unskilled German natives makes the debate over German immigration policy different from that in the United States, and more closely parallel to recent U.S. debates over trade liberalization. As is usually acknowledged in these debates, the benefits to trade are spread over a wide segment of the population and are small on a per person basis. But the costs of trade are borne by a relatively small number of persons and are enormous on a per person basis. Under these circumstances the strong opposition mounted against free trade agreements is understandable. In Germany it would seem that the debate over immigration policy should turn on the same issues. Those strongly in favor of substantial tightening of immigration flows, especially to the extent that greater controls will stem the increase in the number of new unskilled workers, may have good reason for their position. Therefore, the potential impacts of immigration on unemployment and employment rates of particular groups within the native population warrant careful study in Germany and in similar economies. In addition to the foregoing general observation on immigration to Germany, I also have three specific comments about some of the issues raised in Bauer and Zimmermann's paper. First, the authors base their calculations on an assumption that the elasticity of demand for unskilled labor is —0.85. They also experiment with higher and lower elasticities. But these elasticities generate wage effects of immigration that appear too large. Research on this question using U.S. data suggests that the wage effects in an economy with more flexible wages may be as little as one tenth the size implied by the authors. The U.S. results suggest (1) that the adverse redistributional effects of immigration are small, and (2) the gains to natives who do not compete with immigrants in the labor market also are small because "prices" have not changed much in response to increased immigration. Not surprisingly, researchers such as George Borjas have estimated that labor market effects of immigration improve native welfare by only a few billion dollars instead of the approximately 3.0 percent of GDP (or $200 billion) suggested by the authors' calculations. (See also the comment by Barry Bosworth in this volume.) A second comment I would like to make is to ask a question. How effectively can German authorities control the flow of low-skilled immigrants into the country? The U.S. experience illustrates the point that it is one thing to legislate immigration controls, it is quite another to enforce them. Enforcement requires resources, and despite increased expenditures since the 1986 reforms, illegal immigration primarily from Mexico and Central America remains substantial. The fact that Germany shares a geographic border with several

312

Robert LaLonde

countries in Eastern Europe suggests that it may be more vulnerable to illegal immigration than other countries in Western Europe. Finally, Bauer and Zimmermann write sympathetically about the skilledbased immigration policy in place in Canada and Australia. Under this policy, officials grade immigrants seeking to enter these countries legally based on their skills, English and French language fluency, and the desired destination of the immigrant. Those whose scores are above a certain threshold are allowed to immigrate. By adopting such a policy, Germany might lessen the adverse effects of immigration on unskilled natives. However, one variable that might frustrate policymakers' efforts to implement this practice effectively is the distribution of income in Germany compared with other immigrant destinations. To illustrate this point we consider the skill composition of immigrants to the United States and Canada. When we compare immigrants from the same source country, those who immigrate to the United States are on average more skilled. This difference occurs despite the substantially stronger emphasis that Canadian policy places on the skills of its immigrants. The explanation for this fact is that the distribution of earnings is significantly wider in the United States than in Canada. As a result, the economic returns to skills are greater in the United States. Therefore, there are strong economic incentives for the most skilled immigrants to enter the United States, and these incentives appear to have a greater impact on the skills of the immigrant population than does the difference between U.S. and Canadian immigration policy.

CHAPTER

Joining the club: options for integrating Central and Eastern European Countries into the European Union

Michael Koop

8.1

Transformation and integration

The transformation from plan to market puts a heavy burden on the countries in Central and Eastern Europe. Most of this burden these countries will have to carry on their own as Western support will be limited. Aside from technical and some financial aid, the most important support the European Union can offer is access to its markets and particularly access to its economic and political integration, that is, full EU membership. The Visegrad countries unequivocally expressed their desire to join the Union at the turn of the millennium: Poland and Hungary submitted their applications in 1994, Slovakia followed in 1995, the Czech Republic in January 1996. Although practical integration is already taking place, the EU is still undecided on the route to integration, preconditions, and a timetable for accession, or as Baldwin (1994, p. xv) put it: "PanEuropean integration is proceeding, but no one seems to be in charge." At a very early stage of the transformation process, in September 1989, a then very popular policy proposal (Kostrzewa and Schmieding, 1989) was for the Eastern European countries to join the EFTA. This was intended to create a large free trade area without making the reform states part of the Western European ambitions of building a political union. In 1995, this integration strategy is no longer a major option as most EFTA countries have either merged with the European Union or converged to the European Economic Area (EEA) with much closer ties with the EU. This paper will therefore discuss the options now available for integrating the Eastern European countries into the EU. In doing so, it has to be kept in mind that the EU itself is currently in a state of flux, with the Maastricht Treaty up for revision at the Inter-Governmental Conference which started in 1996 and monetary union to be introduced in 1999. 8.2

The state of European integration

The integration issue cannot be dealt with by discussing the options for the Central and Eastern European countries (CEECs). Proper differentiation between 315

316

Michael Koop 1

CEECs is necessary. Using some basic economic indicators (Table 8.1), CEECs can be divided into three groups. The first group of advanced reform countries consists of the Czech Republic, Hungary, Poland, Slovakia, Slovenia, and possibly the Baltic States. The second group is significantly behind the first group with respect to income levels and macroeconomic stability but also with respect to stable democracy. It mainly consists of Albania, Bulgaria, Croatia, Macedonia, and Romania. The third group covers the Commonwealth of Independent States (CIS), European successor states of the Soviet Union, that is, Belarus, Moldova, Russia, and Ukraine. To define the starting point for further integration, it is useful to take a look at how far the integration process has come. Between 1988 and 1990, the EC negotiated Cooperation Agreements with CEEC-4 that were extremely restrictive in their trade provisions. In addition, the EC voluntarily eliminated some quantitative import restrictions specific to Central European exports. The benefits of preferential tariffs under the Generalized System of Preferences were granted to Hungary, Poland (1990), and Czechoslovakia (1991). Sensitive goods such as agriculture, textiles, iron, and steel were excluded from all early trade liberalization measures. Finally, from 1991 on, the EC negotiated Association Agreements (Europe Agreements) with Poland, Hungary, and Czechoslovakia at first and extended the group of associated countries to include CEEC-6. Europe Agreements with the Baltic States were signed in the first half of 1995 and a Europe Agreement with Slovenia has been initialed but not yet signed due to Italian-Slovenian disputes over restitution and the right of foreigners to buy property in Slovenia. The Europe Agreements with CEEC-4 are identical in their major provisions, but since they have been negotiated individually, each contains clauses not present in other agreements. The provisions of the Association Agreements with Bulgaria and Romania, in contrast, are not as far reaching. But even for CEEC-4, the Europe Agreements do not include a binding promise of future EU membership. The goals of all Association Agreements are to support economic and political transformation and the reintegration of the reform countries into the world economy. The major parts of the agreements deal with creating a free trade area with respect to industrial goods over a period of ten years.2 Moreover, the Agreements contain provisions for the mutual rights of establishment for firms, the opening of EU government procurement to CEECs firms, progressive liberalization of trade in services, and partial liberalization of capital movements. To promote closer ties, the EU and CEECs will cooperate in areas such as industrial policy, infrastructure, research and development, and the monetary system. Finally, the associated countries agreed to adopt laws on economic and re1

2

CEEC-4 comprises the Czech Republic, Hungary, Poland, and Slovakia. CEEC-6 is CEEC-4 plus Bulgaria and Romania. This integration aim does not even put CEEC-4 on an equal footing with Turkey, for which a customs union is envisaged in the respective Association Agreement (Wysokinska, 1994).

Table 8.1. Main economic indicators for selected CEEC and EU countries

a

GDP per Capita (US-$) 1991

Growth 1994

Inflation 1994

Budget (percent of GDP) 1994

Unemployment (official) 1994

Czech Republic

7,570

2.7

10.0

+1.0

3.2

6

Hungary

6,080

3.0

20.1

-6.6

10.5

5

Poland

4,720

5.0

29.5

-3.5

16.5

6

Slovakia

3,790

2.5

13.4

-7.5"

14.6

9

Estonia

8,090

3.8

45.0

-0.1

8.0

7

Slovenia

10,800

4.5

20.0

+0.4

14.5

5

Bulgaria

1,840

-1.5

80.0

-7.0

18.1

1

Romania

1,390

3.5

180.0

-3.5

11.0

1

Russia'

3,220

-15.0

224.0

-8.8

2.4

1

Ireland

11,430

5.5

2.4

-2.3

14.7

Greece

7,680

1.0

10.9

-13.1

9.7

Portugal

9,450

1.5

5.2

-7.1

6.8

Germany

19,770

2.9

2.7

-2.7

8.3

c d 1991 in parentheses. * 1993. USSR prior to 1992. 1992. Sources: Baldwin (1994), EBRD (1994), World Bank (1994), FAZ Informationsdienste (1995), OECD (1995).

S Ag

318

Michael Koop

lated issues that are compatible with EU laws. For all that, the EU promised to provide technical and financial aid. The Europe Agreements constitute a major step toward EU integration. Their trade parts, however, have been criticized mainly on two counts (Langhammer, 1992; Messerlin, 1993). First, they contain "flexible" safeguard clauses that can easily be used to restrict trade in any product. Second, a number of goods, particularly those for which CEECs enjoy a comparative advantage, will either be liberalized at a slower pace (cement, coal, shoes, steel, textiles) or mostly excluded from liberalization (agriculture). Moreover, the EU's labor market will stay basically closed for the associated countries. This corresponds to the Base Case analyzed in Chapter 2 of this volume. Finally, at the Copenhagen summit in June 1993, the EU declared that the associated countries of Central and Eastern Europe may ultimately become EU members once they fulfill certain conditions. It was also agreed to start a "structured dialogue" between the EU and the associated countries as a system of multilateral consultations on a wide range of topics concerning mutual interests. At its Essen summit in December 1994, the EU announced a "Preaccession Strategy" for its Eastern European applicants. This "strategy" includes a White Paper that was completed for the Cannes summit in June 1995. This White Paper clarifies and defines the conditions the applicants have to meet prior to membership, including those parts of the acquis communautaire which they must adopt before they will be granted access to the EU. However, the White Paper neither contains a timetable for accession nor the proposed linkage between harmonizing laws and dropping nontariff trade barriers (MethCohn, 1994/1995). Just as the EC has never bothered to install an exit procedure, it has also never defined a clear set of conditions that applicants have to fulfill to become members. Formally, the EU Treaty, through Article 0, states that every European country may apply for EU membership. The decision on accession to the Union requires unanimity on the side of the EU. As laid down in the Copenhagen declaration, 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 that is strong enough to withstand international competition, and the ability to take on the obligations of membership including adherence to the aims of political, economic, and monetary union. With respect to these criteria, the associated countries of the first group would basically pass the test. Democracy has been in place for some years now and will have more time to prove its stability before the countries finally join the EU. The main institutions of a market economy such as a two-tier banking system, a liberal trade regime, private enterprises, bankruptcy legislation, and competition policies are also in place, even if the enforcement of some of the new rules is still lacking efficiency (Briistle and Db'hrn, 1994). Moreover, with respect to adapting the legal system to EU standards, significant progress has been made. The Czech Republic, Hungary, and Poland all made compatibility

Options for integrating CEECs into the EU

319

with EU legislation a legal requirement for passing new laws. In addition, practical integration is on the way, such as the harmonization of technical standards. For the second group of countries, there is a clear lack with respect to democracy, human rights, and the rights of minorities. Above all, the establishment of a market economy is far from being complete. Privatization has not made much progress, and macroeconomic stability is weak. Therefore integration into the EU will take much longer than for the countries of the first group, and other integration strategies have to be designed. Finally, the countries of the third group (the CIS successor states of the Soviet Union) are way behind with respect to democratization as well as economic reforms. They will not be acceptable applicants for decades to come. The only imaginable steps toward a closer integration of these countries are cooperation agreements that could ultimately lead to the establishment of a free trade area. 8.3

Full EU membership

There is little doubt that at least CEEC-4 will eventually become EU members. The gains from Eastern EU enlargement would be significant for the EU as well as CEECs. Full membership would give CEECs unrestricted access to EU markets, the Common Agricultural Policy (CAP), and Structural Funds spending. For the EU there would be political and economic benefits as well. As transformation from plan to market generates social and political frictions, EU membership could help by improving economic opportunities as well as Structural Funds spending. Since ex-communist parties have returned to power in Bulgaria, Estonia, Hungary, and Poland early accession would be preferable over long transitional periods. The fact that these new governments did not turn back the clock and in most cases do not deviate too much from previous reformist governments could be based on the expectation that doing otherwise would jeopardize EU membership. Postponing membership for a long time could make this reform incentive less powerful. Stabilizing the political and economic situation in CEECs should be particularly important to EU countries with direct CEEC borders. The economic gains of CEEC accession would mainly stem from the removal of barriers between countries with relatively different factor endowments. EU membership would allow for an improved allocation of resources (increased specialization, exploitation of economies of scale) and fiercer competition in general. Again, the sooner CEECs join the European Union the larger these efficiency gains will be. 3 Therefore the first integration 3

Another line of thought argues that firms base their location decisions on comparisons of protection and subsidies offered in different countries. Martin (1994) suggests that CEECs should be admitted early on, requiring rather too little than too much convergence. As CEECs converge to EU levels it might otherwise be optimal for them to stay outside the Union and attract industries by offering them more generous subsidies. Thus, all European countries would end up in a suboptimal Nash equilibrium with inefficient protection or subsidies. This reasoning, however, completely neglects the strong political motives for EU membership on the side of CEECs.

320

Michael Koop

option to be analyzed is full membership at a very early entry date. As the pros and cons for this quick-entry option are also valid for slower accession strategies, it will be discussed in some detail. The quick-entry option The Southern enlargement of the eighties was mainly driven by political considerations. Democratic reform and economic liberalization had made Greece, Portugal, and Spain acceptable members in principle and the EC wished to stabilize these countries by admitting them to the EU notwithstanding some mainly economic deficiencies.4 Opposition from Mediterranean areas, agricultural lobbies, and poor countries was appeased by an increase in EC spending programs and by long transition periods. Because the EC neglected its philosophy of only admitting countries "of a similar state of economic development," it had to arrange for financial transfers to prevent the member countries from drifting apart (Seidel, 1995). Thus the Single European Act introduced cohesion as one of the EC's guiding principles justifying structural spending programs for the poorest member countries. The reasoning and strategy of the Southern enlargement could be applied to Eastern enlargement as well (Bakos, 1993). Although not all economic preconditions are currently fulfilled, EU membership would certainly stabilize democracy and help in building prospering market economies. Therefore negotiations with CEEC-4, Slovenia, and Estonia could start immediately and these countries could be admitted as early as possible with some transitional rules in place. If the length of the Spanish membership negotiations is taken as an indicator for Eastern enlargement, membership does not seem possible before 2005 with transition periods easily reaching into the year 2015. This timetable is particularly likely since opposition to Eastern enlargement will be significantly fiercer and more powerful than in the Spanish case. Moreover, an earlier entry date does not seem advantageous as it will take time to train personnel of CEEC government agencies, public administration, and regulatory bodies. Since full membership includes the rules of the single market, trained and experienced experts on health, safety, and environmental standards will be required just as much as government personnel for designing, interpreting, applying, and enforcing laws and EU directives. Baldwin (1994) estimates that it could take up to a decade before this "human factor" will no longer pose a problem for EU membership. Again, 2005 looks like a possible date. Problems with the quick-entry option Contrary to the case of Southern enlargement, however, the European Union is reluctant on Eastern enlargement. The reason for the EU's half-hearted en4

Between the introduction of democracy and EC membership, seven years passed in the case of Greece, nine in the case of Spain, and twelve in the case of Portugal.

Options for integrating CEECs into the EU

321

largement policy is that the CEEC applicants are poorer than previous applicants and that they are also undergoing transformation from plan to market. In addition, the EC underwent tremendous change in the last ten years, so that the EU of the nineties is no longer the EC of the eighties. Four major problems of Eastern enlargement can be identified: • EU institutions and decision making, • budgetary burdens, • trade in sensitive products, • migration. Currently, the institutions of the EU of fifteen are deemed inefficient, its bureaucracy inflexible and overly expensive. With respect to decision making and administration, the Union has reached the size at which the quality of the goods it can provide to its members deteriorates as the number of members increases.5 For example, if only CEEC-4 were to be admitted, there would be approximately 755 members of the European Parliament (as compared to the current 626), speaking 15 languages. By the same token, the number of judges at the European Court of Justice would increase from 15 to 19, making it too big for ruling efficiently. Moreover, following previous practice would create four additional commissioners' positions, raising the number of commissioners to 23. The most important problem would arise in the European Council, where the number of votes would increase from 87 to as many as 109. This raises three important questions: Is unanimity still a feasible decision mechanism; are big countries willing to accept the widening gap between the size of their populations and the number of votes;6 and would the poor countries accept the breaking up of their current blocking coalition? As the latter issue could not be resolved for the most recent enlargement (which was secured only through the Ioannina compromise), it is clear that the latter question will be a key issue in future accession negotiations. The second major obstacle to quick admission of CEECs is the effect the entry of comparatively poor countries would have on the EU budget. Two policies stick out, as they absorb 80 percent of total EU spending: Structural Funds and the Common Agricultural Policy. Therefore per capita incomes and the size of the agricultural sector are the key variables for estimating the budgetary impact of CEECs admission. As Table 8.1 indicates all 5

6

Leipold (1994) provides some examples for institutional problems of further EU enlargement. After admission of CEEC-4, Council meetings could easily last six hours even if each country explains its position in only a twenty-minute statement. When controversial issues are to be decided, spontaneous discussions, repeated statements, breaks, and casting votes may lead to Council meetings of several days. In addition, raising the number of official languages from 9 to 15 would increase the number of language combinations from 72 to 210 with a respective increase in the number of interpreters and translators. In the Council, Luxembourg (0.1 percent of EU population) marshals 2.3 percent of Council votes as opposed to 11.5 percent German votes (21.9 percent of EU population). Admission of CEEC-4 would, under current practice, worsen these numbers to 0.09 percent Luxembourgians enjoying 1.8 percent of Council votes and 18.6 percent Germans enjoying 9.1 percent of the votes.

322

Michael Koop

CEECs (with the exception of Slovenia) are poorer than the poorest EU member Greece and are on average not even close to 75 percent of the EU's average GDP of US$ 16,800. Even though the income figures have to be interpreted with great care, especially when they are denominated in US$, all countries (including Slovenia) would be fully eligible for funds under Objective 1. The same holds true for Objectives 2 (declining industry regions), 3 (long-term unemployment), and 4 (youth training) as unemployment of young adults has been identified as a major problem in Eastern labor markets (Heinrich et al., 1995). In addition, the new Cohesion Fund provides funds for infrastructure and environmental projects. It is limited to countries with less than 90 percent of the EU's average per capita income, currently Greece, Ireland, Portugal, and Spain. It would not be justifiable to exclude much poorer CEECs from EU funds for improving their infrastructures or cleaning up their environments. Additional Structural Funds spending has been estimated to be in the range of 7.2 billion ECU (CEPR, 1992) and 18.8 billion ECU (Ludlow and Gros, 1992). Eser and Hallert (1993) estimate additional Structural Funds spending to be in the range of 8.6 to 10.3 billion ECU, with 70 percent going to Poland. Direct comparisons in Ludlow and Gros (1992) and CEPR (1992) show that admission of Bulgaria and Romania would increase Structural Funds spending by roughly 50 percent. As Cohesion Fund spending is not included in all calculations, any estimate of additional Structural Funds spending of less than 15 billion ECU seems very optimistic.7 The EU's biggest individual budget item is agricultural spending. Table 8.1 indicates that CEEC-4 are less agricultural than Ireland and Greece but that the share of agriculture in GDP is well above the 3 percent EU average. In addition, the significant fall in the importance of the agricultural sector between 1991 and 1993 is partly due to transformation-related cuts in agricultural subsidies and country-specific factors such as the Hungarian crop failure. Based on some crude assumptions regarding the area additionally available for agriculture, agricultural production, and consumption patterns, Baldwin (1994) argues that CEEC-4 enlargement would expand the EU's food surplus. Therefore the costs of maintaining the CAP price system would increase even if CEEC-4 farmers did not get CAP subsidies at all. If CEEC-4 were be fully integrated into the CAP, subsidization would lead to increased production, which would have to be exported entirely. Since agricultural production in CEEC-4 is concentrated in the livestock sector, the expansion would be particularly large. This would require substantial export subsidies. Dumping agricultural products on the world market could lead to a decline of world market prices and therefore require higher export subsidies for 7

In 1992, per capita Structural Funds transfers amounted to 182.2 ECU in Greece, 195.7 ECU in Portugal, and 289.4 ECU in Ireland (Leipold, 1994). As per capita incomes are lower in CEEC4, 300 ECU could be a plausible value for CEEC-4 per capita transfers (Welfens, 1993). Applying this number to the CEEC-4 population of 65 million yields an estimate of 19.5 billion ECU in additional Structural Funds spending.

Options for integrating CEECs into the EU

323

incumbent EU countries. Using a multicommodity dynamic simulation model of world food markets and a 1990 data set, and assuming that the CAP will not be changed beyond the MacSharry reforms, Anderson and Tyers (1993) estimate that the overall cost of including CEEC-4 in the CAP would cost as much as US$ 47 billion. Even using the current exchange rate of 1.34 US$ per ECU this would amount to 35 billion ECU and therefore double current CAP spending. Other studies put a significantly smaller price tag on including CEEC-4 in the CAP. Based on 1992 agricultural production levels, Ludlow and Gros (1992) estimate the additional CAP costs at 3.4 billion ECU. With partial adjustment of production levels until 1997, the same authors calculate the additional CAP costs to rise to 17.6 billion ECU. Including Bulgaria and Romania would again raise the costs by an additional 35 to 50 percent. Using 1989 data, the CEPR (1992) estimated additional CAP costs at a rather low 2.4 billion ECU. Even doubling agricultural output levels leads to only an extra 4.7 billion ECU. The huge differences in additional CAP cost estimates can partly be explained by the level of agricultural production used in the respective calculations. Whereas Andersen and Tyers's 1990 data set includes pretransformation crisis agricultural production levels, which do not take into account the significant absolute and relative (to GDP) reduction in agricultural production, the partially adjusted 1992 figures used by Ludlow and Gros start from a more adequate production level. Despite all methodological differences and different assumptions on the development of income levels in general and agricultural production in particular, the studies reported in Table 8.2 suggest that the total cost of admitting CEEC-4 would be in the range of 25 to 30 billion ECU. Subtracting CEEC-4 contributions of 2 to 5 billion ECU, the absolute minimum of annual net costs to the EU budget will be between 20 and 25 billion ECU. Entry of Bulgaria and Romania as significantly poorer and more agricultural countries would tremendously increase budgetary burdens. Financing an additional 20 billion ECU or more than 30 percent of the EU's 1992 budget would produce more pressure to exclude CEEC-4 from the EU. Who will actually oppose membership depends on how the financial burdens will be shared. One way would be to increase contributions across the board. The other alternative is to reduce the benefits across the board. As most EU transfers go to agricultural producers and poor countries, however, the structure of receivers would change. Neither strategy is likely to generate widespread political support not to mention unanimity as required for the acceptance of new members. Although tax increases as a means of financing the accession of CEECs seem unlikely, it might be the only feasible solution, as it puts the burden on the biggest possible number of shoulders. If the second alternative were chosen, mainly agricultural producers and poor countries would suffer and certainly use their substantial political clout to delay CEEC admission for a long time.

Table 8.2. Budgetary effects ofCEEC admission (billion ECU) Countries Admitted

CAP Cost

Structural Fund Cost

Total Cost

Contribution

Net Contribution

Ludlow and Gros, 1992

CEEC4 (1992)

3.4

11.7

15.1

4.7

10.6

Ludlow and Gros, 1992

CEEC4 (1997)

17.6

18.8

36.4

4.9

31.5

Ludlow and Gros, 1992

CEEC6 (1997)

23.4

29.0

52.4

7.7

44.7

Eser and Hallet, 1993

CEEC4

n.a.

8.6-10.3

n.a.

n.a.

n.a.

Andersen and Tyers, 1993

CEEC4 (2000)

47.0 (US-$)

n.a.

n.a.

n.a.

n.a.

CEPR, 1992

CEEC4

2.4

7.2

9.6

1.8

7.8

CEPR, 1992

CEEC6

3.7

11.6

15.3

2.4

12.9

Baldwin, 1994

CEEC4

n.a.

n.a.

14.4

2.7

11.7

Study

Options for integrating CEECs into the EU

325

The collapse of the COMECON and the CEECs' Eastern markets as well as the abolishment of the states' foreign trade monopoly led to a remarkable reorientation of CEECs' foreign trade. In 1989, the EC's share in CEEC-4's foreign trade was a little less than 30 percent. In 1993 it easily reached 60 percent. By the turn of the century, the Association Agreements will have liberalized EU-CEEC trade in industrial goods to a large extent. Only products that the EU deems "sensitive" will be fully (agricultural products) or partly (steel, coal, chemicals, textiles) excluded from liberalization. One reason for these exceptions could be that all sensitive industries are declining industries. Because they are often geographically concentrated, low-cost competition from CEECs could lead to a surge in region-specific unemployment. As these sectors have acquired considerable lobbying power, politicians might prefer to keep prospective competitors out. Another reason could be that CEECs exhibit a high degree of specialization in a narrow range of products. In that case firms in incumbent member states could be threatened by CEEC low-cost competitors. However, careful studies of actual trade patterns reveal - with the exception of agricultural products - that there is no rational explanation for the EU's sensitivity with respect to trade with CEECs (Chapters 2,3,4 of this volume; Smith and Rollo, 1993; Messerlin 1993). Adjustment problems would be in the range of normal economic change. Moreover, the continuous decline of the EU's sensitive industries should sooner or later be reflected in diminishing lobbying powers. This indicates that trade in industrial goods is not likely to be a serious impediment for the quick-entry option. Before the first CEEC joins the EU, the Association Agreements could be freed of their safeguard clauses and the room for nontariff trade barriers could be narrowed to reduce uncertainty for CEEC producers. The fourth possible impediment to the quick entry option could be migration. Table 8.1 indicates that German (and Austrian) GDP per head is about three to five times higher than in CEEC-4 and ten to thirteen times higher than in Bulgaria and Romania. With the exception of the Czech Republic, unemployment rates are between 10 and 18 percent and a quick reduction is not expected for the near future. At the same time, there is demand for unskilled labor and craftsmen in Germany and Austria and to some extent also in health care and household-related services. In addition, the social safety net that emigrants would have access to when working in the EU is better than anywhere in Eastern Europe. Taken together these factors might constitute a strong economic incentive to migrate from CEEC-4 and particularly from Bulgaria and Romania to the EU. Estimates of the extent of future migration in the case of free mobility of labor are extremely vague. One popular assumption is that generally 5 to 10 percent of the population are prepared to migrate in response to wage differentials of approximately 3 to 1 (CEPR, 1992, p. 86). Adjusting for differences in the cost of living, this wage differential may to some extent represent the situation along the EU's eastern border. Therefore the migration potential from CEEC-4

326

Michael Koop

could be in the range of 3.25 to 6.5 million migrants.8 Other studies estimated the migration potential at 3 percent of the population (1.95 million) (Blanchard, Dornbusch, Krugman, Layard, 1992). Although the numbers for CEEC-4 do not appear to be especially large, some problems could arise. Of particular importance is that Eastern Europeans would most likely not spread equally across Europe but settle at a few locations in Austria, Germany, and to a lesser extent France, because networks of previous emigrants as well as relatively close historical and cultural ties already exist in these countries (Chapter 7 of this volume; Angenendt, 1995). The structure of migration could cause additional problems. Unskilled workers will be hit particularly hard as privatization and firm restructuring move ahead in CEEC-4. In addition, the agricultural sectors in CEEC-4 and particularly in Poland will significantly reduce employment of unskilled workers. This will exert pressure on the low-skill labor market. Migration of low-skilled labor to a few places along the border might create tremendous tensions in regional labor markets and therefore be politically unacceptable. On the other hand, highly educated workers and young workers are more mobile internationally. Because youth unemployment is a major problem in CEECs and because wage differentials will continue to exist for a long time, it can be expected that young and/or skilled workers will choose to migrate and CEECs will experience a brain drain due to early access to the EU's labor market. Baldwin and Venables (1994) argue that this brain drain could be amplified: Because cheap and skilled labor is a major reason for foreign direct investment in CEECs, large-scale emigration of skilled workers would reduce the incentives for foreign direct investment. Quick entry reconsidered The problems with quick entry of CEECs seem to indicate that full membership will not be a feasible strategy for decades to come. However, the previous section may have overstated the problems of quick admission. The following section will take up previous arguments, relativize them, and put forward some ideas of how they can be resolved. It has been argued that EU institutions are not able to cope with around twenty member states. This, however, was partly true for EC-12 and the negotiations for EFTA enlargement made clear that it will be increasingly difficult to solve the EU's institutional problems. In other words, institutional reform is inevitable with or without CEECs. As CEECs will sooner or later join, any fundamental reform should take this into consideration and design institutions and decision and representation rules that can handle twenty or more member states. 8

Assuming that wage differentials three times as big as the 3 to 1 ratio would also increase the migration potential three times, possible migration from Bulgaria and Romania would be on the order of 4.8 to 9.6 million people.

Options for integrating CEECs into the EU

327

The major goal of institutional reform must be to restrict the size of institutions. As the top positions in EU institutions are basically political in nature, denationalization is not a feasible choice. Therefore restrictions of national representation will have to be introduced.9 Seidel (1995) provides detailed suggestions for this kind of institutional reform. The number of commissioners could be reduced to fifteen or sixteen by abolishing the principle that each member states appoints at least one commissioner. Instead, the number of big country commissioners could be reduced to one and the small countries could appoint the other commissioners according to a preset rotating scheme. The same method could be applied with respect to the European Court, where full representation still exists in principle but has been abolished in most cases already. The establishment of two senates instead of one would be the obvious alternative with a "Joint Senate" for leading decisions. The rules governing the decision making of the Council of Ministers would have to be at the core of any institutional reform. Here two aims clash: democratic representation and the protection of minority rights. Currently, the protection of minority rights is given high priority through requiring unanimity for important policy areas, such as accession of new member states and taxation. On the other hand, it has been shown that representation in the Council of Ministers is not democratic as voters from small countries exert stronger influence than voters from big countries. Leaving things unchanged, the admission of CEEC-4 would further worsen the underrepresentation of big countries. Various institutional alternatives have been discussed in the literature (Leipold, 1994). The number of votes in the Council could be redefined to give big countries more influence and avoid that the majority of the population can be outvoted by the minority. Going further, unanimity as a decision rule could be abolished and the number of votes could be assigned to each member state according to its population. As there is no exit option, however, unanimity should not be abolished, even though decision making would be made more complicated, as announcing a veto would be attractive to more countries and the veto premia could be higher. For all decisions made by simple or qualified majority, a two-tier decision rule could compromise the aims of democratic representation and the rights of minorities. A decision would pass only if the simple (qualified) majority of votes in the Council and of the EU population would be reached. Although this mechanism would reduce the power of small countries, it would not lead to an overrepresentation of big countries. Moreover, there would be no need to solve the tricky problem of defining the blocking minority because that would be implicitly done. For instance, more than one third of the EU population or one third of the Council votes would block a decision that requires a qualified majority. Whatever the specific interests of the incumbent member states are, they will 9

As Article 126 of the Maastricht Treaty calls for the principle of full member state representation, constitutional instead of simple procedural changes would be necessary.

328

Michael Koop

have to reform the EU's institutions. The Inter-Governmental Conference for the revision of the Maastricht Treaty, which began in 1996, provides a good opportunity for this undertaking. A major problem for this conference will be that incumbent members do not agree on the aims and means of European integration. For instance, it is controversial how far political integration should go, what the degree of organizational centralization should be (centralized versus federal structure), and in which ways political powers should be distributed (more executive powers for the Commission versus stronger influence of the European Parliament). In this context, it is the EU which is not ready to be joined by CEECs. Since the new institutional structure will be valid for CEECs, the associated countries should be invited to the Inter-Governmental Conference. Although this would most likely slow down negotiations, it seems worthwhile as fundamental decisions on further integration will have to be made. The costs to incumbent EU states have been identified as another major problem of the quick-entry option. Although budgetary burdens of admitting CEEC4 were estimated to amount to as little as 20 billion ECU, the budget problem is not as severe as it may seem. An additional 20 to 30 billion ECU would significantly increase the EU's budget. However, the sums are by no means large when compared to the resource transfers under the Marshall Plan after World War II or in the aftermath of German reunification. Budget neutrality has never been a prerequisite for EU membership, and the countries of the Southern enlargement heavily depend on EU funds after ten years of EU membership. As a matter of fairness, budget neutrality should not be required now. Baldwin (1994) calculates the minimum number of years it will take until CEECs reach the state of budget neutrality under different assumptions on the development of per capita income and importance of the agricultural sector. Assuming a high agricultural share elasticity and annual average growth rates of 6 percent, budget neutrality would be obtained by Slovenia in eight years, and in the Czech Republic, Hungary, Poland, and Slovakia in eleven, twelve, thirteen, and fourteen years, respectively. Growth rates of 6 percent may seem quite optimistic. "Normal" business cycles tend to produce growth rates in the range of - 1 to - 2 percent in recessions and around 3 to 4 percent in recoveries. The fluctuations can be caused by Keynesian-type fluctuations in aggregate demand or by supply shocks. The transformation from plan to market may be seen as an extreme form of a positive supply shock. Unraveling many of the distortions that characterized the pretransformation economies should allow for growth rates substantially higher than those caused by simple fluctuations in aggregate demand. The Asian and particularly Chinese experience indicates that higher growth rates are possible and sustainable for quite some time. Even the CEECs' growth rates of 2.5 to 5 percent reported in Table 8.1 support this view once it is taken into consideration that unknown but most likely significant growth is taking place in the private sector without being captured by official statistics. One study of the Polish economy estimates that the second economy accounted for an additional

Options for integrating CEECs into the EU

329

output of 20 percent of GDP (Rajewski, 1993). Since there is reason to believe that growth in the second economy exceeds growth in the official economy, this would lead to higher growth rates than those reported in Table 8.1. In addition, measurement problems tend to underestimate growth rates for transformation economies as long as pre- or early reform prices are used to calculate real growth rates (Berg, 1994). Because prices were not market clearing, increases in the output of goods that were in short supply before growth will be underestimated. Another factor not accounted for when using prereform prices and therefore leading to lower growth rates concerns the improvements in the quality of goods.10 Capital and infrastructure are major bottlenecks for increasing output in all transformation countries. Progress in the liberalization of the financial systems together with improved macroeconomic stability and capital inflow from abroad could improve the capital stock in a relatively short period of time and enhance economic growth. In addition, production will be less capital-intensive because no Western wages have to be paid in the transformation countries. Therefore less capital will be needed and comparisons with eastern Germany are invalid. Moreover, there is ample space for savings in the EU budget. The effectiveness of regional aid has to be scrutinized, and a streamlining of the EU's bureaucracy seems possible. The CAP needs fundamental reform beyond the MacSharry plan. Reductions in agricultural price levels could reduce budgetary burdens of CEEC admission. There are hardly any economic objections to this kind of reform. The expectation that this reform would most likely be stopped by the agricultural lobby indicates that the cost argument is not economic but political in nature. Should neither convergence of CEEC nor reductions in EU spending suffice in solving the budget problem, the additional financing requirements could be met through a temporary increase in EU revenues. By issuing "integration bonds" the EU could increase general revenues or attract funds for particular projects such as infrastructure programs within the framework of the Cohesion Fund. Interest payments could be financed by all incumbent members and repayment could start once the integration of CEECs has been completed. A bondfinanced Eastern enlargement would require the EU to intervene directly in international capital markets. Earlier plans for autonomous debt financing by the EC were rejected and do not appear to be have much political backing. Alternatively, EU funds could be increased along the lines of the German solidarity surcharge for financing unification. For instance, the VAT-related contribution could be increased for a period of five years. To make this solution more acceptable in the political arena, burdens could be differentiated; for example, rich 10

Household surveys as well as data on energy consumption, retail turnover, and other indirect measures indicate that the output decline in the transformation countries is less severe than indicated by official statistics (Berg, 1994). For a detailed analysis of output decline in Russia see Gavrilenkov and Koen (1994).

330

Michael Koop

countries or countries that will benefit the most from CEEC admission could be asked to pay relatively more. Migration was identified as another possible impediment to the quick-entry option. However, the migration potential of CEEC-4 admission was not found to be overwhelmingly large. First of all, it should not be forgotten that migration tends to equalize wage differentials, thereby creating welfare gains for the EU. Moreover, remittances that migrant or guest workers transfer to their families in Eastern Europe will increase income there and possibly help promote local suppliers. Second, historical experience from earlier enlargements teaches that even significant differences in GDP per head did not lead to massive migration. In the sixties it was expected that workers from southern Italy would overrun Northern Europe in response to the 3 to 1 difference in per capita income. However, observed migration from southern Italy was below the average migration rate for the rest of the EC. To protect the EC from migrants from Greece, Spain, and Portugal a seven-year transition period was agreed upon in the Southern enlargement. Afterwards, there was a net return of Greek workers to Greece and the numbers for Spain and Portugal do not indicate an increase in the numbers of migrating workers (Angenendt, 1995). Migration decisions are based on expectations about the future situation in the home and the guest countries. If people expect that the situation at home will improve they are less likely to migrate. It seems the surge in unemployment in the wake of transformation has reached or passed its peak, and real wages are rising if only slowly. Therefore the push factor of migration will lose in importance. Quick entry to the EU could further brighten expectations of a domestic upswing by increasing the credibility of political and economic reforms in CEECs. Moreover, it would improve economic opportunities not least through encouraging trade, and thereby relieve labor-market pressures. In addition, quick entry would also foster foreign direct investment because any CEEC location could serve as the single production location for supplying all Eastern markets. By allowing for more trade and investment, quick entry would enable people to stay at home. Despite this, migration may still be opposed from two sides. The first one could consist of direct neighbors such as Austria and Germany, which might feel threatened by large imbalances in regional labor markets. The second opposition group may consist of those countries currently providing cheap labor in the EU, for example, Ireland, Portugal, and Spain. If it is necessary to achieve full membership by the year 2005, the mobility of labor could be restricted for a transition period. For instance, a five-year transition period with progressively widening migration quotas could be installed. The political economy of quick entry The previous discussion acknowledged that there are substantial problems with the quick-entry option. However, they seem solvable and quick EU entry would

Options for integrating CEECs into the EU

331

be possible if it were supported by the incumbent states. As CEEC admission would generate tremendous gains for the EU in terms of political stability in neighboring countries and economic prosperity, support should be strong. However, aggregate welfare gains would coincide with substantial redistribution of income and political powers. In addition, welfare gains would be spread unevenly across the EU. For instance, a major increase in welfare would come from intensified trade between the EU and CEECs (Wang and Winters, 1991; Collins and Rodrik, 1991; Chapters 2, 3, and 4 of this volume). Baldwin's (1994) estimates support this general result. They indicate, however, that individual countries will benefit to a different extent. The ratio of potential to actual exports is highest for Belgium, Denmark, Finland, France, Portugal, and the United Kingdom. As actual trade levels were low in all these countries with the exception of France, the advantage of Eastern enlargement in absolute terms is less pronounced. In absolute terms, the major beneficiaries would be Austria, France, Germany, Italy, and the Netherlands. The redistribution of income and political power will lead those who lose from Eastern enlargement to oppose it. First, small countries will not be in favor of early enlargement because it will become more difficult to reconcile the interest of a larger number of countries for a blocking coalition. In addition, the principle of full representation of every member state in EU institutions will have to be given up in order to restore efficiency, which again will hit small countries harder. Second, budgetary burdens will prevent a number of countries from supporting Eastern enlargement. This problem goes beyond the fact that if burdens are spread equally every member state will have to pay more or receive less. Because cuts in spending programs would mainly concern regional aid and agricultural spending, the poor and largely agricultural countries would suffer the most. This is aggravated by the fact that the prospective members are poor and agricultural so that competition for Structural Funds and CAP spending would increase. Third, Eastern enlargement might be blocked by the "deepeners" in the Union. As, for instance, France views the deepening of integration as one way of restricting German powers, it is feared that the enlargement would slow down the deepening of integration. Taken together, small, poor, and agricultural countries and those striving for deeper integration will not find it in their immediate interest to support Eastern enlargement. The EC's Southern enlargement in the eighties was the starting point for the quick-entry option. Although there are parallels between Southern and Eastern enlargement (low income levels of applicants, strong agricultural bases, recent return to democracy), the situation has changed markedly, as the EU of the nineties is no longer the EC of the eighties. The major changes are that integration is deeper now (Single Market, Monetary Union) and that the structure of member countries has changed with Southern enlargement. The change in the structure of members coincided with a policy change toward more redistributive policies, which in turn makes it even more difficult to finance the accession of poor and agricultural new members.

332

Michael Koop

Support for Eastern enlargement can come only from those countries that will be the main beneficiaries of CEEC admission. These countries are most likely Austria and Germany, the Scandinavian countries, and possibly the Benelux states. They will make sure that their vital interests in Central and Eastern Europe are not harmed by opponents to CEEC admission. For Northern European countries, gains from Structural Funds spending or CAP spending in Poland or the Czech Republic are bigger than when used in Greece or Portugal. As these countries happen to be the major net contributors of EU funds, they will have to provide additional funds. They most likely will also use their financial means and political powers to "convince" other EU members to forgo some of their financial privileges in return for CEEC admission. In case any country vetoes CEEC accession, the Northern countries will have to reevaluate integration that goes beyond the Single Market Program. Executing quick entry So far it has been argued that EU membership of the most advanced CEECs would generate welfare gains for both sides and that the sooner accession can be completed the bigger these welfare gains will be. One popular counterargument is that CEECs are not ready for entry. Liberalization, deregulation, privatization, and restructuring at the firm level did not make sufficient progress for Eastern economies to withstand competition from Western countries. Spain's and Greece's experiences with integration are quoted as examples for the prospective difficulties. Although the Spanish economy was far more liberal than CEEC economies a few years ago, Spain stayed at the associated level for sixteen years. When it entered the EC and the Single Market, it was granted long transition periods for numerous measures and generous financial support from the incumbent members to cushion the radical economic changes. If these changes were radical at that time, they appear relatively modest compared to the changes in the wake of transition from plan to market that all countries of the first CEEC group successfully completed. As CEECs are completing the transformation of their economies, they could skip steps that other countries under normal conditions chose to take. In other words, as CEECs see EU membership as a "modernization anchor" for their economies (Inotai, 1994), quicker liberalization seems possible. Greece is probably the example that exhibits most clearly the problems of early integration. Although Greece was granted long transitional periods and substantial EC funds as well, it still has not taken off economically. Whereas the Spanish and Portuguese economies grew consistently faster than the average EC countries after EC membership, Greek growth was consistently lower than EC average growth. Moreover, the Greek public sector deficit is high, its public debt is over 100 percent of GDP and rising quickly, and consequently inflation and interest rates are high. The Greek example indicates that money is not a cure-all if economic policies are bad. The liberalization track record of the

Options for integrating CEECs into the EU

333

most advanced CEEC is far better than Greece's and gives rise to the assumption that wherever CEECs are not yet ready for the EU and its Single Market, they may be so in time. Prospective EU membership could just work as an additional incentive to speed up liberalization and restructuring. Another question that arises when full membership at an early date is discussed is who should participate in the first wave. Different scenarios are possible. On the one hand, CEECs should be admitted in groups, as this lowers agency costs for the EU. On the other hand, the diversity of CEECs should be taken into consideration. One such scenario is to admit CEEC-4 first, Slovenia and the Baltic states in a second step, and Bulgaria, Romania, and others at a still later date. Independent of a specific timetable, this would probably be the best enlargement strategy from a political point of view, as the EU expands slowly eastward. The political economy of the quick-entry option suggests a somewhat different structure. In a first step only the most liberal, richest, and smallest countries should be asked to join so as to minimize frictions and opposition from EU incumbents. Following this strategy would put the Czech Republic, Estonia, Hungary, and Slovenia first.11 The annual budget costs of this enlargement would most likely be in the range of 5 to 7.5 billion ECU; the migration potential would be restricted to a few hundred thousand. Some economic as well as political problems could arise if this option were to be accepted. From an economic point of view, trade diversion might be a problem for those CEECs not included in the first enlargement. However, as intra-CEEC trade is relatively unimportant and the share in overall trade shrinking, the trade diversion problem might not be serious. Moreover, other countries could move up the integration ladder at the same time. For instance, a second group of countries could be admitted to the EEA, removing the remaining nontariff trade barriers. In this case the major problem seems to be a political one, if Poland were left outside whereas Hungary were in.12 To a lesser extent this would be true for the Czech and Slovak Republics and Estonia and the other Baltic States. To solve this problem it could be helpful to define the conditions and a tentative timetable not only for the first Eastern enlargement but also for the second at an early date. The year 2005 has been suggested as a possible date for the quick-entry option; in 2010 the second group of countries could join. This schedule somewhat mutes the quick-entry notion and would certainly not meet CEEC requests. But it would provide them with a preset timetable. Moreover, it would give appli1]

12

Opposition to this strategy could also come from CEEC membership in the monetary union. At least for the Czech Republic, Estonia, and Slovenia it is conceivable that they will immediately fulfill the convergence criteria for monetary union upon EU entry, whereas for EC founding members such as Italy or Belgium there is a fair chance that they will not meet the requirements in the foreseeable future. Since the EU negotiates the Accession Treaty on a bilateral basis, this strategy would possibly lead to a later entry date of, for instance, Poland as the problems that have to be solved are more difficult than in the case of the Czech Republic.

334

Michael Koop

cants sufficient time to complete the transformation process, attain educated and experienced personnel, and adopt EU legislation. 8.4

Partial EU membership

The feasibility of the quick-entry option rests upon a number of optimistic assumptions. Growth in the East has to be significantly faster than in the West, structural change has to make Eastern countries less agricultural, the EU has to reform its institutions, and EU special interest groups must be persuaded to restrain from blocking enlargement. If one or more of these prerequisites turn out to be overly optimistic, full membership will not be a realistic option for decades to come. Since there are no more formal steps between the Association Agreements and full membership, the integration could come to a halt and gains from integration could be lost. The question therefore is what interim steps could be taken before full membership would finally be achieved. One such strategy is partial EU membership, of which two versions can be identified: restricted and variable membership. Restricted EU membership The quick-entry option concedes that transitional phases are unavoidable in certain policies. These transition phases give the new members extra time to adjust or protect incumbent states from new competitors. Restricted membership, that is, membership without full membership rights, would go beyond transition periods for individual measures. Instead, new members would be completely excluded from specific policies. For instance, CEECs could be excluded from Structural Funds spending if opposition from poor countries could not be appeased. Other examples would be restrictions on labor mobility or the Common Agricultural Policy. This is the type of membership analyzed in Chapter 2 of this volume. Compared to full membership with transitional phases, restricted membership would have the advantage that the opposition against it might be a little more modest. This may, on the other hand, also be seen as its major disadvantage, because it most likely deprives the new member states of their most important immediate benefits: access to agricultural markets and Structural Funds spending. Moreover, the pressure on the EU to reform its policies and institutions would also be lower. A tricky question of restricted membership is whether voting rights should be curtailed along with the respective membership rights. In other words, would a country that is excluded from CAP or Structural Funds spending be excluded from voting on these issues as well? If restrictions in certain parts of membership rights were to be matched with restrictions on voting powers, it could happen that Iceland or Norway as EEA members have more rights in the EU than CEECs. This scenario can hardly be called a membership option and it

Options for integrating CEECs into the EU

335

would be inferior to the quick-entry option. Whether it is nevertheless superior to no immediate integration steps remains an open question. Alternatively, restricted membership could coincide with full voting rights. In this case, however, CEECs would have a say in matters that do not affect them. For example, they could vote on any CAP reform, even if they were excluded from CAP. This introduces additional uncertainty in the EU decisionmaking process as the voting behavior of new and restricted members would be difficult to predict. On the one hand, they may vote with the poor-country coalition to increase redistributive spending programs because they will ultimately benefit from that as well. Especially if the periods for which restrictions apply are sufficiently long, new members would have strong incentives to vote for cuts in those programs from which they are excluded. Since the voting strategies are not easily anticipated by those incumbent countries that benefit the most from redistributive programs, it can be expected that the restricted membership option is politically not feasible. Variable EU membership Similar to restricted membership is the concept of variable membership or variable geometry. This concept, which was first proposed with respect to deepening integration within the existing Union as well as Eastern enlargement (Koop and Siebert, 1993), negates that all member states have to move on toward closer integration at the same time. Although the Union is officially still marching in step, in practice the EU is already moving in another direction. Britain opted out of Maastricht's social chapter, the Schengen Agreement excludes a number of member countries, the Danes voted "No" on monetary and political union, and the European Monetary Union will welcome only member countries that fulfill the Maastricht convergence criteria. This multitrack European integration could also be applied toward Eastern enlargement. Two aspects are involved. The first one is that membership in a specific common policy is bound to preset conditions that prospective members would have to fulfill. For instance, if any CEEC fulfills the Maastricht convergence criteria by the year 2000 - which seems possible for some of them - they would be allowed to join the monetary union. This would, however, not imply that they are automatically accepted in the CAP, structural spending, or anywhere else. The second aspect is that the fulfillment of the membership conditions does not lead to automatic acceptance and the obligation to join the common policy. Instead, CEECs must be allowed to opt out of the specific common policy. Only in this case, variable membership could help speed up Eastern enlargement. The variable membership strategy might be one possible way of advancing integration within the existing Union because countries that are willing and fit for closer integration could move on without being slowed down by others. It does not, however, appear to be a feasible option with respect to the EU's East-

336

Michael Koop

ern enlargement. First, there are currently no accepted rules for this strategy. When it was applied, it was mainly because it was the only way of avoiding a complete breakdown of further integration, as in the case of Denmark and the UK. Letting the multitrack option govern Eastern enlargement would require a clear delineation of different policy areas. Neither does this delineation exist nor would it be easily devised. Second, new rules for accepting new members would have to be agreed upon. This would certainly raise fundamental constitutional issues, as integration has so far required unanimity. In addition, severe practical problems would occur if some member states participate in some common policies and not in others. 8.5

EEA membership

Instead of restricting membership partially and arbitrarily, integration could be pursued in a more consistent way if CEECs joined the European Economic Area (EEA) instead of the EU. 13 The EEA extends the four freedoms to all its members. EEA nationals are allowed to seek jobs and establish businesses anywhere in the EEA and all capital controls are abolished. The mutual-recognition principle for product norms and licences is accepted and most trade barriers are removed. The major exception to this is agriculture. Government procurement is open to all firms and trade-distorting state aid to industries is generally prohibited. The EEA's basic competition rules are identical with the EU's and no antidumping duties can be imposed on intra-EEA trade. Moreover, the EEA created new institutions for surveilling and enforcing the rules of the EEA. There are some major differences between EEA and EU membership. EEA members are excluded from the EU's Common Agricultural Policy. They do not have access to Structural Funds spending and cannot participate in the future monetary union. Moreover, they cannot vote on EU laws concerning economic matters even though they have to adopt this legislation.14 The EEA option has a number of distinct advantages. The most obvious one is that it is more likely to be accepted by the EU, as the major problems of full EU membership (CAP, Structural Funds spending, shifts in voting powers) would not arise. Joining the EEA would also make the potential members more equal, which could speed up future EU membership negotiations. In the first place it would tear down barriers between CEECs that they put up for political and historical reasons but that also exist due to the bilateral nature of the Europe Agreements between individual CEECs and the EU. This "hub-and-spoke bilateralism" (Baldwin, 1994) hurts the CEECs because it leads to trade diversion and discourages foreign and domestic investment in CEECs. For industries for which 13 14

The EEA's current non-EU members are Iceland, Liechtenstein, and Norway. The non-EU EEA members have only the choice between accepting new laws or having the whole relevant part of the EEA Agreement suspended. They cannot decide whether or not to accept individual pieces of legislation (Baldwin, 1994).

Options for integrating CEECs into the EU

337

economies of scale can be exploited, the investment deterrent in CEECs gives Western locations along the border a head-start as they can, better than CEEC locations, supply all CEECs from one production location (Stehn, 1994). The EEA Agreement founded new political and legal institutions. Although the legal systems of EU and non-EU members are formally separate, they are intended to create a common legal system in practice. This joint system would be advantageous for CEECs as it allows them to improve their EU-specific human capital and become acquainted with the practical working of EU principles such as mutual recognition of product norms, trade policies, and supervising state aid. From the point of view of CEECs the EEA option is less desirable than full EU membership for a number of reasons. First, it would miss one important aim of integration: greater political stability. Having emerged from dictatorship and Soviet dominance, CEECs are eager to join the Western security system. For the military part this involves NATO membership. However, as the Maastricht Treaty calls for common foreign, security, and defense policies, EU membership would provide a deeper rooting in the Western European security system. Considering the EU's recent track record on these issues, that might not mean much with respect to external pressures. It could, however, help fight internal pressures for plan- or Eastern-oriented changes. Second, exclusion from the Union's agricultural and food markets prevents CEECs from exploiting their current comparative advantage, thereby inhibiting growth prospects. Thirdly, the exclusion from Structural Funds spending is a twofold disadvantage. On the one hand, it simply deprives the countries of funds they could spend if they were full members. On the other hand, there is a funds-related strategic disadvantage of not joining the EU. As anything that makes other locations and companies more competitive is bad for any location or firm, the exclusion from all EU spending programs would constitute a serious disadvantage for CEECs compared to the poor EU members. For instance, Cohesion Fund spending will improve the infrastructure in Ireland, Greece, Portugal, and Spain, making these countries more attractive for foreign direct investment. This investmentrelated problem adds to the trade-related strategic disadvantage. Because full integration lowers transactions costs within the EU, all firms that do not enjoy equal access or a level playing field face a competitive disadvantage (Krugmann, 1988; CEPR, 1992). The EU's recent enlargement made clear that the EFTA countries viewed this competitiveness issue as so important that they accepted the EEA only as a transitional phase before full EU membership. The superiority of EU over EEA membership led them to apply for full EU membership shortly after becoming EEA members. Taken together the arguments for the EEA option are ambiguous. The major reason why at least the most advanced CEEC should not join the EEA as a separate integration step is that it would deprive them of their major EU membership benefits. With the exception of political feasibility, not much will be won by this intermediate step. One reason put forward for a slower integration pace

338

Michael Koop

is that CEECs are currently not able to withstand the blow of Western competition (Stehn, 1994). Since both the EU and the EEA options include the Single Market Program, countries are either fit for both options or for none. Therefore, countries that can cope with the competitive pressures should join the EU without delay. For less advanced CEECs previous EEA membership could be a good preparation for the EU, which they could join later on. 8.6

(C)EFTA option

The EEA option for integration opens the Single Market for CEECs. Since CEEC governments as well as firms are still in the process of adapting to the rules of market economies, this might be seen as too much of a challenge. In this case, other intermediate integration steps should be introduced. The obvious way would be to create a free trade area and restrict liberalization to trade in goods and services and possibly capital movements. A substantial number of proposals for arranging a free trade area have been made. Early on in the transformation process it was proposed that the most advanced CEECs join the EFTA before the EFTA members join the EC (Kostrzewa and Schmieding, 1989; Baldwin, 1992). As there is not much left of the original EFTA, CEECs cannot join. Therefore it was proposed that they could adopt the original Treaty of Rome for themselves and start economic integration in the same way the EC did in 1958 (Messerlin, 1993). Alternatively, CEECs could introduce EFTAlike rules through deepening the provisions of the Central European Free Trade Area (CEFTA) established in the early 1990s. This corresponds to the first case examined in Chapter 2. Yet another alternative proposes a multilateralization of the Europe Agreements (Inotai, 1994) or an Association of Association Agreements (Baldwin, 1994). Finally, a "European Continental Common Market" was proposed as a concentric set of free trade zones between the EU and CEEC with ever-stricter rules the closer countries move to the provisions of the Single European Market (Messerlin, 1993). These proposals have two things in common. First, they were put forward at a time when there was significantly more uncertainty about the political and economic developments in Central and Eastern Europe and therefore were intended to keep the CEECs at some distance. Second, they all focus on liberalizing trade. Most prominent among the free trade proposals is the CEFTA option. As CEECs and the EU are natural trading partners, a CEEC free trade area as an intermediate integration step would allow CEECs to exploit the trade potential with the EU and stimulate economic growth. Moreover, abolishing intra-CEFTA trade barriers could create additional trade potential among CEECs and lead to cost reductions for CEEC firms. It would also reduce the trade- and investment-related disadvantages of the current hub-and-spoke bilateralism. Finally, it is possible that speaking with one (CEFTA) voice the political clout of CEECs in EEA- or EU-membership negotiations could be stronger than in the case of strictly bilateral relations.

Options for integrating CEECs into the EU

339

The CEFTA option and similar proposals are not too popular in CEECs. For one thing, it is seen as a way of keeping CEECs out of the EU (Inotai, 1994). Moreover, in history close cooperation has never existed among the CEECs as most of them always relied on the different dominating powers in Europe (Bakos, 1993). The only exception was the COMECON, which was a period of forced and inefficient cooperation that the CEECs are not eager to repeat even if the "partner" this time would be the EU. Moreover, for CEEC-4, Estonia, and Slovenia creating a free trade area of sorts would not help a lot. This is clear from the results of Chapter 2. Once the transitional periods provided for in the Europe Agreements elapse, a free trade area with the EU exists. Since the trade potential with other CEFTA countries would be relatively small compared to the EU, removal of intra-CEEC trade barriers would be beneficial but not overwhelmingly important. Therefore the most advanced CEECs would have to move on and introduce more and more Single Market measures. This would be a slow entry into the EEA option. As the EC's harmonization efforts of the seventies indicate, this piecemeal approach could hamper integration. For the less advanced CEECs the CEFTA option could be an intermediate integration step worth taking. 8.7

Conclusions

The paper discussed a number of options for integrating Central and Eastern European countries into the EU, namely the quick entry option, partial integration, and EEA and CEFTA membership. Two conclusions were derived. First, the quick entry is a promising option for integrating the transformation countries. The year 2005 appears to be the first entry date possible for the most advanced CEECs. Not all CEECs should be admitted at the same time. According to the respective states of their transformation processes, their income levels, and economic policies, CEECs should enter in three or four enlargement "waves." From a purely economic point of view, the difficulties of the quickentry option can be overcome. However, strong political opposition can be expected from poor regions, small countries, and agricultural producers. Whether this opposition can be made more agreeable to quick entry will very much depend on the willingness of those countries that will gain the most to provide additional funds for appeasing opposition. Experience with previous enlargement suggests that this will not be a cheap undertaking. For less advanced CEECs the other options could be attractive intermediate integration steps. Those countries currently somewhat behind the advanced CEECs could be granted access to the EEA by the year 2005 with the perspective to become full EU members five years later. For countries such as Bulgaria or Romania, which significantly lag behind in transformation and general economic development, an EFTA-like free trade area seems to be an appropriate first step toward closer integration. The second conclusion that can be derived from the discussion is that not

340

Michael Koop

only will the CEECs have to complete their transformation processes, but the EU has to make some fundamental decisions about the aims and means of future integration. Currently the EU is not capable of dealing with more member states. First, EU institutions are still in the process of coping with the EFTA enlargement. CEEC admission would finally make them inefficient. Second, the CAP needs further reforms, as it paralyzes EU policies by absorbing more than half of the EU's spending. The fundamental conflict between deepening and widening mirrors the different views on the aims of integration. It is not the economic shortcomings of the CEECs that makes their accession so difficult. Those problems have been solved in earlier enlargement waves as well. The fact that the EU does not have a clear idea of where it is going is the major problem for integrating Central and Eastern European countries. REFERENCES Anderson, Kym, and Rod Tyers (1993). "Implications of EC Expansion for European Agricultural Policies, Trade and Welfare," CEPR Discussion Paper 829, London. Angenendt, Steffen (1995). "Freiziigigkeit: ein Hindernis fur die Osterweiterung der Europaischen Union?", List-Forum fur Wirtschafts- und Finanzpolitik 21/1, 36-52. Bakos, George (1993). "After COMECON: A Free Trade Area in Central Europe," Europe-Asia Studies 45/6, 1025^4. Baldwin, Richard E. (1994). Towards an Integrated Europe, London: CEPR. Baldwin, Richard E. (1992). "An Eastern Enlargement of EFTA: Why the East Should Join and the EFTAns Should Want Them," CEPR Occasional Paper 10, London. Baldwin, Richard E., and Anthony Venables (1994). "International Migration, Capital Mobility and Transitional Dynamics," London School of Economics mimeo; quoted in Baldwin (1994). Berg, Andrew (1994). "Does Macroeconomic Reform Cause Structural Adjustment? Lessons from Poland," Journal of Comparative Economics 18/3, 376-409. Blanchard, Oliver, Rudiger Dornbusch, Paul Krugman, and Richard Layard (1992). "East-West Migration: the Alternatives," London School of Economics, Working Paper 101, London. Briistle, Alena, and Roland Dohrn (1994). "Systemtransformation in Ostmitteleuropa," RWI-Mitteilungen 45/2, 177-203. CEPR (1992). Is Bigger Better? The Economics of EC Enlargement, Monitoring European Integration 3, London. Collins, Susan M., and Dani Rodrik (1991). Eastern Europe and the Soviet Union in the World Economy, Institute for International Economics, Policy Analyses in International Economics, No. 32, Washington, D.C. Eser, Thiemo W, and Martin Hallet (1993). "Der mogliche Beitrag der EGRegionalpolitik bei einer Ost-Erweiterung der EG: Hilfe oder Hindernis?", Osteuropa-Wirtschaft 38/3, 195-217. European Bank for Reconstruction and Development (EBRD) (1994). Transition Report, October issue, London. FAZ-Informationsdienste (1995). Ldnderanalysen, various issues, Frankfurt. Gavrilenkov, Evgeny, Vincent Koen (1994). "How Large Was the Output Collapse in Russia? Alternative Estimates and Welfare Implications," IMF Working Paper 94/154, Washington, D.C. Heinrich, Ralph P., Michael J. Koop, et al. (1995). Sozialpolitik im Transformationsprozefi: Ziele, Erfahrungen und Perspektiven sozialpolitischer Reformen in Mittel- und Osteuropa, Kiel: Institute of World Economics, mimeo.

Options for integrating CEECs into the EU

341

Inotai, Andras (1994). "Die Beziehungen zwischen der EU und den assoziierten Staaten Mittel- und Osteuropas," Europaische Rundschau 111 3, 19-35. Koop, Michael J., and Horst Siebert (1993). "Institutional Competition versus Centralization: Quo Vadis Europe?", Oxford Review of Economic Policy 9/1, 15-30. Kostrzewa, Wojciech, and Holger Schmieding (1989). "Die EFTA-Option fur Osteuropa: Eine Chance zur wirtschaftlichen Reintegration des Kontinents," Kiel Discussion Paper 154, Kiel. Krugman, Paul (1988). "EFTA and 1992," EFTA Occasional Papers 23, Geneva. Langhammer, Rolf (1992). "Die Assoziierungsabkommen mit der CSFR, Polen und Ungarn: wegweisend oder abweisend?", Kiel Discussion Paper 182, Kiel. Leipold, Helmut (1994). "Die EU im Spannungsverhaltnis zwischen Vertiefung und Erweiterung," in Helmut Leipold (ed.), Ordnungsprobleme Europas: Die Europaische Union zwischen Vertiefung undErweiterung, Marburg: Marburger Gesellschaft fur Ordnungsfragen der Wirtschaft, 39-78. Ludlow, Peter, and Daniel Gros (1992). "The European Union and the Future of Europe," CEPS Working Document 73, Brussels. Martin, Philippe (1994). "A Sequential Approach to Regional Integration: The European Union and Central and Eastern Europe," CEPR Discussion Paper 1070, London. Messerlin, Patrick A. (1993). "The EC and Central Europe: The Missed Rendez-vous of 1992?", Economics of Transition 1/1, 89-109. Meth-Cohn, Delia (1994/1995). "Getting into the EU: Harmonizing Laws. Talking Turkey, Business Central Europe," December 1994/January 1995, 21-3. OECD (1995). Main Economic Indicators, April 1995, Paris. Rajewski, Zenon (1993). "National Income," in Leszek Ziekowski (ed.), Results of the Polish Economic Transformation, Warsaw. Rollo, Jim, and Alasdair Smith (1993). "The Political Economy of Eastern European Trade with the European Community: Why so Sensitive?", Economic Policy 16, 140-81. Seidel, Martin (1995). "Reform der institutionellen Strukturen der Europaischen Union als Vorbedingung einer Osterweiterung," List-Forum fiir Wirtschafts- und Finanzpolitik 21/1, 22-35. Stehn, Jurgen (1994). "Stufen einer Osterweiterung der Europaischen Union," Die Weltwirtschaft 4, 195-219. Wang, Zhen Kun, and Alan L. Winters (1991). "The Trading Potential of Eastern Europe," CEPR Discussion Paper 610, London. Welfens, Paul J. J. (1993). "Die Integration der mittel- und osteuropaischen Staaten mit der EG: Ausgangsbedingungen, Anpassungsprobleme und Chancen," in Erhard Kantzenbach and Otto G. Mayer (eds.), Europaische Gemeinschaft - Bestandsaufnahme und Perspektiven, Berlin: Verein fiir Sozialpolitik, 135-88. World Bank (1994). World Development Report 1994, Infrastructure and Development, Washington, D.C. Wysokinska, Zofia (1994). "Assoziierungsabkommen zwischen Polen und der Europaischen Gemeinschaft - Richtungen der Liberalisierung der Handelsumsatze," Osteuropa-Wirtschaft 19/1, 55-70.

Comment Barry Eichengreen

Michael Koop has written a fine paper on the accession of the Eastern European economies to the European Union. It is an excellent summary that I have no hesitation in recommending to other readers. If I have a quibble, it concerns the data used to show that assimilating the CEECs into the EU poses more formidable problems than did the accession of Portugal, Spain, and Greece. Koop compares the relative position of these two sets of countries today, where the more appropriate comparison is Eastern Europe today and Southern Europe fifteen years ago. Fifteen years ago the share of agriculture in GDP was 16 percent in Greece, 13 percent in Portugal, and 8 percent in Spain; according to the World Development Report, comparable figures today are 11 percent for Hungary, 7 percent for Poland, 6 percent for the Czech Republic, and 6 percent for the Slovak Republic. It is not obvious, in other words, that dependence on agriculture is more of a problem than in Southern Europe fifteen years ago. Nor do per capita GDPs provide a clear picture of the extent of the "cohesion problem." According to the Penn World Tables, real per capita GDP as a share of West German levels was 42 percent in Portugal, 49 percent in Greece, and 62 percent in Spain in 1980. For 1990 the Penn World Tables give ratios to Germany of 42 percent for Czechoslovakia, 37 percent for Hungary, and 27 percent for Poland. Again, the proportionate discrepancies are not significantly larger than for Southern Europe fifteen years ago. There are reasons to suspect that real incomes were artificially inflated under the old regime, of course; 1992 figures, again from the World Development Report, suggest ratios to Germany of only 10 percent. But these figures were artificially depressed by grossly undervalued exchange rates (only partially corrected by the World Bank's purchasing power parities) and the exceptional difficulties of the transition. I cautiously conclude that the cohesion problem is somewhat more serious than it was with the Southern European enlargement, but not overwhelmingly. The substance of Koop's paper can be read in several ways. One is through the eyes of an EU functionary. Most EU officials acknowledge the desirability of integrating Eastern Europe into the Union. Today's CEECs were once fully 342

Comment

343

integrated into the European economy; trade models suggest, given their factor endowments and proximity to Europe's economic core, that there is a logic to this happening again. But integrating the CEECs into the EU threatens to strain the bargains upon which the political economy of integration is based. To a non-European, none of the obstacles is insurmountable. Koop notes, for example, that Council meetings in an enlarged EU would last six hours if each country had twenty minutes to state its position. But it is already clear that, in an EU of fifteen member states, procedures will have to be streamlined. A solution will have been found well before the CEECs are ready for admission. Then there is the fact that enlargement to the east will increase the cost of operating the Common Agricultural Policy (CAP). But the admission of Sweden and Finland has already done so. As with the procedures governing Council meetings, it is already clear that the CAP will have to be reformed before the CEECs are finally granted admission. Migration is another worry raised by enlargement to the east, given income and wage disparities between the CEECs and the EU's founding member states. But this same worry was raised by the inclusion of Italy in the European Economic Community and by accession of Greece, Spain, and Portugal. In each case migration turned out to be surprisingly small, perhaps reflecting the significance Europeans attach to language, geography, and local culture. The same is also likely in the present instance. A final worry is the impact on the EU's sensitive sectors (steel, chemicals, textiles, apparel), in which Eastern European economies plausibly have a comparative advantage. Academic studies (including Chapters 2,3, and 4 of this volume) are unanimous, however, in concluding that this advantage is small and that enlargement will add little to the pressure already felt by the sensitive sectors. A second way the paper can be read is from the point of view of an economist - in terms of whether enlargement would enhance efficiency. In part, this is the question of whether extending the EU's customs union to the east would result in mainly trade creation or trade diversion. Much of what Western Europe would import from the East (agricultural goods, steel, chemicals, autos, machine tools) are not things that it presently imports from the United States or the newly industrializing countries; rather, these are goods in which European trade is restrained by various overt and hidden barriers. Because Western European trade would not be diverted from low- to high-cost suppliers, the West would primarily experience trade creation. The benefits to the East are less clear. The Visegrad Four have relatively open trading systems. When forming a customs union with Western Europe, they have reason to worry about trade diversion, since they would be tempted to import EU manufactures tariff free instead of lower cost, duty-ridden substitutes produced elsewhere in the world. In addition, admission to the EU may saddle them with expensive welfare states that they can ill afford. The EU is not likely to allow the Visegrad countries to opt out of the Social Charter as it did the UK, which had considerably more bargaining power.

344

Barry Eichengreen

A third way to read the paper is as a political economist - that is, in terms of whether EU membership can provide a "commitment technology" to solve the credibility and commitment problems of the transition economies. Can putting the CEECs on a glide path to admission lock in their commitment to reform? Can it influence the domestic debate by providing Structural Fund transfers in return for continued progress? Can making the CEECs part of the EU allay their security fears? A fourth and final way of reading Koop's paper is as a political scientist, who would ask which EU member states stand to gain and lose from enlargement. Here the author rightly points out that Germany stands to gain the most since it borders on Eastern Europe and has the largest markets there. In contrast, the relatively small, low-income EU countries of Southern Europe stand to lose in terms of both veto power and Structural Fund receipts. In combination, these approaches suggest that neither quick admission nor indefinite EFTA status is an obvious winner. Early admission is attractive to the East because it locks in access to Western European markets, but not insofar as it saddles the countries of the region with expensive social and agricultural programs. Early admission is attractive to the West because it locks in reform, but not insofar as it raises the costs of operating the Structural Funds and the CAP. An EEA- or EFTA-like arrangement is attractive because it allows the transition economies to opt out of expensive programs, but not insofar as they will lack a credible commitment of entry. The question, then, is whether compromises are feasible. This brings me to the most provocative part of Koop's paper, where he recommends "partial admission." The transition economies, he suggests, might become members of the EU but not participate in the CAP, the Structural Funds, or various social programs. But this proposal raises additional questions. If the CEECs can opt out of the CAP, for example, should they be allowed to vote on issues of CAP funding and administration? Surely the answer is no. Otherwise they would face a moral hazard problem and become a solid bloc of votes for eliminating funding and price supports. If the CEECs are offered admission, but receive no CAP funding and have no voice in the program's administration, is this a better deal than an EEA- or EFTA-like arrangement? Again, I suspect that the answer is no. Koop's "partial admission" proposal is an application to the East of the concept of "variable geometry" that is so fashionable in Europe today. As the preceding discussion suggests, I have doubts about the feasibility of variable geometry. The EU is based on an interlocking set of bargains. Germany and France will have to be in both the monetary union and the EU security community, for example, insofar as the core bargain is that Germany concedes a common currency in return for an expanded security role in the context of an EU foreign policy, while France makes the opposite trade. When one excuses countries from particular parts of the larger bargain, the entire arrangement can be jeopardized. For these reasons, I doubt the feasibility of partial admission. If I am cor-

Comment

345

rect, the only viable alternatives are early admission on equal terms and some kind of EFTA-like arrangement. It seems likely that the first option will be more attractive for the Visegrad Four, the second for the remaining transition economies.

Comment Hans-Jurgen Vosgerau

This is a lengthy but well-structured, balanced, and informative paper. It is a variation on the old theme of gains from trade or more precisely gains from trade liberalization and integration. Integration is of course more than liberalization of trade in goods because it includes not only factor mobility, common institutions, and policy coordination but also some redistribution. The general problem is complicated by the fact that both partners consist of several countries and that one of them is undergoing a difficult process of transition from centrally planned to market economy, while the other too is changing. The main question is how to organize the process of adjustment and how to organize the adaptation of the institutional framework, given the assumption that in the long run there are positive gains for both partners, and taking into account that these gains will be distributed very unevenly, comprising even negative elements. To answer this question a number of distinctions are introduced. The most obvious is that between the different members of the group of Central East European Countries (CEECs). The criterion of distinction is the level of development as measured for instance by income per capita and the functioning of institutions necessary for market economies. The groups of countries are: •

CEEC-4 or Visegrad countries (Poland, Czech and Slovak republics, Hungary), in addition sometimes Slovenia and the Baltic States • Bulgaria, Rumania, Albania, Croatia, Macedonia • the Community of Independent States (CIS) or the former Soviet Union, including Russia, Belarus, Ukraine and Moldova This distinction is connected with another double distinction, viz. that between different lengths of the transition period or time spans until full membership is reached on the one hand and different degrees and qualities of integration on the other hand. It seems plausible that the fittest CEECs should reach the status of full member more rapidly (approximately by the year 2005) than the less advanced and less adapted ones, which might reach this later status only via a number of intermediate steps, taking considerably more time. The bulk of the paper (more than half) is devoted to discussion of quick en346

Comment

347

try and full membership of CEEC-4. This concentration is certainly warranted in terms of relevance and realism. Problems of other CEECs, as well as partial and later accession, are treated as modifications. In the following comments I shall concentrate on quick entry. The quick-entry option is discussed against the background of the Southern enlargement of the EC during the eighties. Here are indeed striking parallels between Greece, Portugal, and Spain at that time on the one hand and present-day CEEC-4 on the other hand: •

In both cases democratic reforms preceded the discussion about accession to the EC or EU. • In both cases GNP per head in the economies seeking accession was much below that of the EC or EU. The competitive position of these countries was definitely inferior to that of Western and Central Europe. • Opposition to new members by some of the old members was not negligible, especially with regard to the agricultural sector. Reasons for admission were and are to a large extent political, aiming at stabilizing newly established democratic orders. The problems mentioned were attacked by methods aiming at improving the infrastructure in order to raise competitiveness. This was done by the establishment of new funds, which partly served in addition the purpose of compensating potential losers from the accession. Another means of easing the difficulties was the negotiation of relatively long transition periods. Thereby an asymmetric market access was established, which gave protection to the newcomers for several years. These transition periods in addition permitted the adaptation of rules, the training of officials, and the like to prepare the application of EC regulations in the new countries. The parallelism between Southern and Eastern enlargement is of course not perfect. The following differences seem to be the most important. • •

•

• •

CEECs are still poorer than the Mediterranean countries, and transition from central planning to market economy is still on the way. Therefore redistribution demands are very high, especially with regard to the Common Agricultural Policy (CAP), with resulting budgetary burdens of considerable degree. In some sensitive sectors such as agriculture, steel, textiles, and partly chemicals, CEECs have comparative advantages that threaten the corresponding EU industries, which are already declining. Threatening East-West migration complicates the picture considerably. Additional difficulties arise from the fact that the EU has meanwhile grown considerably; the Southern and Northern (EFTA) enlargement brought the number of members to fifteen, with the consequence that decision making becomes increasingly difficult. Results of Maastricht II are still uncertain and controversies about the Monetary Union are far from being resolved.

348

Hans-Jiirgen Vosgerau

I shall review the main points of Michael Koop's paper starting from the traditional gains from trade argument. This permits in some cases a view from a slightly different angle. I shall add particular comments in two fields, that of competition and that of migration. The traditional gains from trade argument consists in the assertion that by liberalizing trade, both countries will gain, provided there are no essential distortions. Under this heading, one should expect that sectors with obvious comparative advantages in CEECs such as agriculture, textiles, and steel should export their products to the EU. Even the simple textbook arguments predict substantial advantages for both partners. But just these products are classified as sensitive and denied general access to the EU. Why? One answer could be that the situation in the CEECs is distorted because of different conditions under the central planning regime. Another answer would rest on excessively high adjustment costs in the EU, which necessitate long transition periods to organize the necessary intersectoral factor movements. Perhaps the most important answer stems from the fact that gains are distributed very unevenly and may be negative, thereby inducing political objections. This line of argument along public choice lines is rightly stressed by the author. Therefore compensation schemes were established that are of course much more complicated than the textbook lump-sum payments. These schemes have different names such as CAP, various regional and structural funds, and, since Maastricht, cohesion funds. The redistribution schemes not only require institutions for their administration but in addition cause heavy budget problems when new members are ante portas. These budget problems sometimes seem to dominate the discussion about accession and enlargement. In the case of CEECs they are sometimes estimated at 20 billion ECU. It is important to note that an increase of the budget is usually viewed positively by politicians and bureaucrats, because it increases their own power. Institutions are necessary not only to handle compensation payments, but more positively - also to overlook the functioning and enforcement of rules to make markets work and to correct failures (public goods, external effects, etc.). This leads to my additional remarks on competition policy. Regarding the adaptation of the legal framework in the CEECs to that of the EU there is one particular area that merits special attention, viz. that of the competitive order. In the Europe Agreements originally concluded with the Visegrad countries and meanwhile extended to other CEECs, there are special provisions that in practice extend EU competition policy measures to these countries. They call for the establishment of antitrust authorities, which have instruments similar to the Directorate IV for Competition of the European Commission at their disposal. At a recent conference in Prague on European Competition Policy, I learned that the relevant institutions are practically in place and can start working. The problem remaining is essentially one of asymmetry in a twofold sense.

Comment

349

First there was criticism that the EU asks and demands from the CEECs stricter pro-competitive policies than from many of its old and long-time members. Second and more important it is to be remarked that the buildup of a competitive order in Eastern Europe is not in line with the development of EU trade policy vis-a-vis these countries. I refer to the barriers of trade still in effect with regard to the sensitive sectors (agriculture, steel, textiles, chemicals). In this way the parallelism between the removal of trade barriers and the establishment of an EU-wide competition policy that was practiced in the late eighties and early nineties within the EC and more recently with the former EFTA countries in the European Economic Space is not observed vis-a-vis the CEECs. This diminishes the efficiency gains otherwise associated with the substitution of a second-best trade policy regime by a first-best competition policy regime. In addition this procedure causes tensions between the EU and the CEECs, which could be eased if the EU made a firm commitment to admit at least the more advanced CEECs as full members within a defined and reasonable time span. The EU is not only a tariff union but has also liberalized factor markets with corresponding institutions to guarantee the free movements of goods, services, capital, and people. Full membership gains and losses therefore consist of many more components than those resulting from liberalized goods trade. Under this heading a few remarks on the connections between age structure, retirement funds, and migration between CEECs and EU will be added. With respect to migration from the CEECs to the EU, one should not confine the analysis to the next ten years. An extension of the time horizon to the year 2030 or 2040 might change the scenario in a way that opens possibilities for additional gains from capital mobility. I refer to the deteriorating age distribution in Western Europe, which by 2030 will either lead to a doubling of social security contributions or a halving of retirement payments (pensions) within the currently practiced pay-as-you-go systems of old-age insurance payments. As most EU governments are too shortsighted to switch in time to a fully funded system, it is to be expected that Western Europe will either have to encourage immigration when the time comes in approximately thirty years, or envisage other solutions. One such scenario assumes that private life insurance and pension funds will soon start to expand and possibly boom because of the growing demand for oldage pensions in addition to the insufficient social security systems. The accumulated funds could then be channeled into Eastern European capital markets and thereby enhance investment in the CEECs. This might contribute to an acceleration of growth. Similar ideas have recently been advocated by Reisen (1994) with respect to the more advanced developing countries, which already have functioning capital markets such as some South American and Asian countries. It seems interesting to apply these ideas to the CEECs.

350

Hans-Jiirgen Vosgerau

In concluding, I repeat that I have concentrated on the quick-entry scenario and the full member scenario, leaving aside the possibilities of intermediate forms of associations, some of which might ease some of the problems. When discussing these one should never forget the ultimate aims of all these discussions: • • •

• •

One aim is to establish a stable general order for Europe: political, social and economic. This cannot be done by force but rests on decisions by the countries concerned. These decisions are taken rationally in a positive way only if there are gains in a wide sense, either directly or indirectly via compensation payments. It is to be recognized that CEECs have to solve the difficult transition problems. At the same time the change of the EU itself has to improve the institutions (Maastricht II) so that they can cope with some twenty members and at the same time stay open to the rest of the world to prevent contradictions between a free world trade order and the regional European block.

By way of conclusion I repeat that I am in full harmony with the normative position of the author. When adding that in my opinion he underestimates the difficulties, this means that the efforts to realize a stable European order have to be increased. REFERENCE Reisen, Helmut (1994). "On the Wealth of Nations and Retirees," OECD Development Center, May.