The Economics of International Integration: Fourth Revised Edition

THE ECONOMICS OF INTERNATIONAL INTEGRATION Now in its fourth edition, The Economics of International Integration has established itself as an accessible introduction to the economics of regional integration among nation states. This new edition has been completely revised to reflect important theoretical developments of the past decade and current policy initiatives. The original focus on the analysis of customs unions, free trade areas, common markets, fiscal integration and harmonization, monetary integration and regional policy issues has been retained and updated, while two entirely new chapters deal with the ‘new’ economics of market integration and the role of transnational corporations. The book’s main focus is the theoretical framework that has been developed to understand regionalism and it provides a balanced account of contemporary mainstream analysis and current policy debates. The analysis reflects current initiatives in the European Union which have been the principal driving force for newer theories of integration and whose experience yields useful guidelines for other blocs. In the European context, the book assesses the impact of the Single Market programme, as well as discussing emergent budgetary problems in the light of monetary integration and enlargement into Eastern Europe. The Economics of International Integration provides an original and stimulating analysis that encompasses several distinct branches of economic thought. It is a balanced and accessible survey for students of economics and international relations of the theoretical and policy issues involved in international integration. Peter Robson is Honorary Professor of Economics in the University of St Andrews. He has held visiting professorships at the College of Europe and the Universities of Amsterdam and Clermont-Ferrand. He has served as the editor of the Journal of Common Market Studies. His recent publications include Transnational Corporations and Regional Economic Integration (Routledge, 1993).

THE ECONOMICS OF INTERNATIONAL INTEGRATION Fourth edition

Peter Robson

London and New York

First published 1980 by the Academic Division of Unwin Hyman Ltd This edition published in the Taylor & Francis e-Library, 2002. Second edition 1984, third edition 1987, third impression 1990, Fourth edition first published 1998 by Routledge 11 New Fetter Lane, London EC4P 4EE Simultaneously published in the USA and Canada by Routledge 29 West 35th Street, New York, NY 10001 © 1998 Peter Robson All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage and retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloguing in Publication Data Robson, Peter The economics of international integration/Peter Robson.—4th ed. p. cm. Includes bibliographical references and index. 1. International economic integration. I. Title. HF1418.5.R63 1988 337.1–dc21 97–29570 ISBN 0-203-01960-1 Master e-book ISBN

ISBN 0-203-20441-7 (Adobe eReader Format) ISBN 0-415-14876-6 (hbk) ISBN 0-415-14877-4 (pbk)

CONTENTS

List of figures and tables List of abbreviations and acronyms Preface

1

2

3

ix xi xiv

Introduction

1

The scope of the subject The emergence of the theory of international economic integration The present reach of international economic integration: a global perspective Plan of the book

2

9 15

The theory of customs unions and free trade areas

17

The theoretical context Customs unions and resource allocation The economics of a free trade area

17 18 28

Broadening the framework

37

The terms-of-trade effects of customs unions and free trade areas Economies of scale in customs union theory Conclusion

39 41 48

v

7

CONTENTS

4

5

6

7

Customs unions versus unilateral tariff reduction

51

The superiority of unilateral tariff reduction? The analysis of customs unions with public goods Public goods and the rationale for customs unions

52 55 59

The rationale for the integration of other economic policies

63

Policies to overcome allocational cross-border spillovers Policies to overcome stabilization spillovers Policies for cohesion and convergence The dynamics of policy integration

65 66 67 68

The theory of common markets

72

Integration and intra-bloc capital flows Foreign capital and the costs and benefits of integration Factor supplies, dynamic effects and convergence Conclusion

74 77 80 81

The new economics of market integration

82

Competition and the gains from integration Market integration: appraising the gains The progress and impact of the Single European market 8

9

83 92 101

Transnational corporations and regional integration 109 Internalization and the gains from integration Regional economic integration and the investment decisions of TNCs Public policy and TNCs TNCs and regional integration in Europe

110

Fiscal integration

123

Economic efficiency and the assignment of functions The relevance of fiscal federalism to regional economic integration The budget of the European Union EU budgetary issues in the longer term

123

vi

113 116 119

127 139 151

CONTENTS

10 Fiscal harmonization

163

The harmonization of indirect taxes The harmonization of direct taxes Tax harmonization in the EU Conclusion 11 Monetary integration

165 176 180 189 190

The benefits of monetary integration The costs of monetary integration Fiscal policy in a monetary union Strategies for monetary integration Monetary integration and the EU Conclusion 12 Equity, cohesion and convergence: regional issues The regional dimensions of economic integration The rationale of regional policy in an economic community The implementation of a community regional policy Regional policy in the European Union 13 Economic integration in a global perspective: experience and initiatives The ‘old’ regionalism among developing countries The ‘new’ regionalism Conclusion

193 197 212 215 217 228 231 233 243 247 250

268 270 277 295

References and select bibliography

299

Index

325

vii

FIGURES AND TABLES

FIGURES 2.1 2.2 2.3 2.4 3.1 4.1 6.1 6.2 7.1 7.2 11.1 11.2 11.3

The effects of a customs union The effects of a free trade area on a single country A comparison of free trade areas and customs unions, 1 A comparison of free trade areas and customs unions, 2 A customs union with economies of scale A customs union with economies of scale and public goods The impact of free intra-regional capital flows The costs and benefits of integration with foreign capital The effects of a reduction in trade barriers The gains from removing market segmentation The original Phillips curve trade-off between inflation and unemployment The expectations-augmented Phillips curve trade-off Rational expectations, credibility and time-consistency

21 29 31 32 43 56 75 78 86 88 201 203 205

TABLES 7.1 9.1 9.2 9.3

The economic gains from completing the internal market, 1985 EU budget revenue, 1988–97 EU expenditure, 1988–97 EU net budgetary transfers, 1995

ix

94 146 147 154

FIGURES AND TABLES

12.1 12.2 12.3 13.1

Effects of lowering trade barriers on the location of production Regional disparities in income, productivity and unemployment in the EU, 1983 and 1993 Regional disparities in income and unemployment in the EU by member state, 1983 and 1993 NAFTA’s macroeconomic impact on Mexico and the US, 1994–2003

x

236 254 256 291

ABBREVIATIONS AND ACRONYMS

APEC AFTA ASEAN BLNS CACM CAP CARICOM CARIFTA CEAO CEC CEEC CET CFA CMEA COMECON CU CUFTA EAC EAGGF EC ECB ECCM ECOWAS ECSC ECU EDF EEC EFTA EIB

Asian Pacific Economic Co-operation ASEAN Free Trade Area Association of South East Asian Nations Botswana, Lesotho, Namibia and Swaziland Central American Common Market Common Agricultural Policy Caribbean Community Caribbean Free Trade Area Communauté Economique de l’Afrique de l’Ouest (Economic Community of West Africa) Commission of the European Communities Commission of the European Economic Community Common External Tariff Communauté Financière Africaine Council for Mutual Economic Assistance (COMECON) See CMEA Customs Union Canada—US Free Trade Agreement East African Community European Agriculture Guidance and Guarantee Fund European Community European Central Bank Eastern Caribbean Common Market Economic Community of West African States European Coal and Steel Community European Currency Unit European Development Fund European Economic Community European Free Trade Association European Investment Bank xi

ABBREVIATIONS AND ACRONYMS

EMS EMU ERDF ERM ESCB EU FTA FTAA FDI GATT GSP

European Monetary System Economic and Monetary Union European Regional Development Fund Exchange Rate Mechanism European System of Central Banks European Union Free Trade Area Free Trade Area of the Americas foreign direct investment General Agreement on Tariffs and Trade Generalized System of Preferences (for developing countries) IMF International Monetary Fund LAFTA Latin American Free Trade Association LAIA Latin American Integration Association MERCOSUR Mercado Común del Cono Sur (Common Market of the Southern Cone) NAFTA North American Free Trade Agreement NAIRU non-accelerating inflation rate of unemployment OECD Organization for Economic Co-operation and Development PPS purchasing power standards SACU Southern African Customs Union TEU Treaty on European Union TNC transnational corporation UDEAC Union Douanière et Economique de l’Afrique Centrale (Customs and Economic Union of Central Africa) UEMOA Union Economique et Monétaire Ouest-Africaine (West African Economic and Monetary Union) UMOA Union Monétaire Ouest-Africaine (West African Monetary Union) VAT Value Added Tax WTO World Trade Organization

NOTE The 1992 Treaty on European Union established the European Union. The European Community is the pillar or dimension of the Union which is concerned with economic integration. The term European Union is in general use to refer to all activities undertaken xii

ABBREVIATIONS AND ACRONYMS

by the Union unless legal enactments concerning the institutions or individual treaties are in question. In this volume, for simplicity, this practice has been followed. Strictly speaking, there are three European Communities: the European Community (prior to 1992 formally termed the European Economic Community); the European Coal and Steel Community; and the European Atomic Energy Community. Since 1967 there has been a single Commission and a single Council for all three.

xiii

PREFACE

This book sets out to provide a concise and accessible introduction to the economics of international integration. Its main concern is with the analytical framework that has been developed for throwing light on the central policy issues of regionalism, both from the standpoint of the group, and from that of its component member states. Global and systemic implications of regional integration are largely left aside. At the core of the subject is the theory of customs unions that originated in the writings of Jacob Viner and others half a century ago. That theory remains basic, but it is only one part of the apparatus that is required today to analyse the many issues that arise in contemporary forms of regional economic integration. These go far beyond those addressed in the traditional theory of customs unions or free trade areas and concern not only those that arise in common markets but also others posed by policy integration in other key areas, notably taxation and fiscal and monetary policy. The analysis of the distributional aspects of integration in its spatial dimension, largely ignored in orthodox theory, has also become an important component of the subject. During the past decade or so, substantial changes have taken place in the analysis of regional integration. One of the most important is that, in its analysis of welfare gains, the traditional perfect competition approach has been supplemented by models that allow for imperfect competition, economies of scale and product differentiation. This development has considerably modified the views of economists both on the rationale for regional integration and its gains. Scale economies both internal to the firm and external are also central to newer models of the spatial aspects of integration. The embodiment of these and other, newer currents of thought into the analysis of international integration is not yet fully consolidated. This revised edition reflects central elements of those xiv

PREFACE

recent contributions that form part of the emergent mainstream and places them in historical perspective. It has often been remarked that the theory of international economic integration has been policy-driven. The initiatives of the EU have provided the principal driving force and inform much of the theorizing. The analysis that has emerged is nevertheless of far wider application. Few of the propositions of the modern theory of international integration are EU-specific, although some of the strategies that have been developed for the EU would not necessarily be transferable or of interest, elsewhere. The experience of the European Community (since 1992 the European Union) and its approaches to key issues in any case merit attention beyond its borders, since the EU is the world’s most highly developed example of regional integration and its policies and strategies are far and away the best documented and the most intensively scrutinized. Many of those policies will doubtless continue to furnish guidelines to other blocs as they have in the past. The EU context is discussed in each of the chapters dealing with the new economics of market integration, TNCs and fiscal, monetary and spatial issues. The literature of regionalism has grown very rapidly since the middle of the last decade, reflecting the renewed policy concern with the subject in Europe, the western hemisphere and beyond. In undertaking this revision to take account of recent analytical contributions, empirical evaluations and institutional progress, the summing-up has necessarily had to be highly selective. Lack of space has ruled out discussion of a number of important topics. It is believed nevertheless that a balanced appraisal is presented. Those who wish to penetrate more deeply into the subject will find a guide to more specialized theoretical and empirical studies in the list of references and the select bibliography. For this edition, the main structure of the book has been left unchanged, but two new chapters have been added and others deleted mainly to reflect new or neglected perspectives. The emergence of newer analytical perspectives on market integration which emphasize pro-competitive effects rather than comparative advantage is reflected in the chapter on the ‘new’ economics of market integration. The second new chapter deals with the significance for regionalism of the ability of transnational corporations, by internalizing transactions, to bypass the market on which the effects of integration in the orthodox model depend. Another substantially new chapter, which deals with regional integration initiatives in a global perspective, looks at the context xv

PREFACE

and objectives of the main new regional initiatives of the past decade, including the North American Free Trade Agreement. The chapter in the last edition dealing with integration among planned economies and COMECON has been deleted, since it is today of purely historical interest. In the course of writing successive editions of this book I have incurred many intellectual obligations to academic colleagues and others that are too numerous to acknowledge adequately here. I therefore limit myself to thanking specifically those who have made helpful comments on the major changes for this edition. In St Andrews, I wish first to record thanks to Christopher JensenButler. Outwith St Andrews, my particular thanks are due to Michael Keen, Anthony Thirlwall and Ian Wooton. In revising the chapter on monetary integration I have benefited from the collaboration of my colleague David Cobham. The analytical structure and thrust of that chapter remain largely unchanged, but the issue of time consistency is specifically addressed and the account of recent developments in the EMS has been completely recast, drawing in particular on studies by Cobham. He has agreed to allow his name to be associated with the revision of that chapter, for which we take joint responsibility.

xvi

1 INTRODUCTION

International economic integration, often termed regionalism, may be defined as the institutional combination of separate national economies into larger economic blocs or communities. Integration in this sense first took off in the late 1950s when the European Economic Community (EEC) was set up under the Treaty of Rome of 1956. Led by this example, regionalism soon spread throughout Africa, Latin America and other parts of the world in the 1960s, leading Haberler (1964) to characterize our era as ‘the age of integration’. Earlier examples of regional blocs can be found before the 1960s, but those established in the twentieth century were nearly all associated with either imperialism or colonialism. Soviet imperialism gave rise to COMECON, which was set up in 1949, while British colonialism and French colonialism were separately responsible for the creation of a number of African blocs. For most of the 1970s regionalism, though widespread, made little real headway outside Europe. From the mid 1980s, regionalism has developed a remarkable new impetus. This new phase was marked by the successive enlargements of the European Community (EC) and its policydeepening measures; by new initiatives in North America following the conversion of the US to the merits of free trade areas after years of hostility to regionalism; and by attempts to relaunch existing arrangements in Latin America, the Caribbean, Africa, South East Asia and Australasia and to establish new ones. The diverse initiatives of the new wave of integration differ markedly from earlier arrangements in their implications, approach and motivation, partly as a result of changes in the global policy context, and partly as a result of the identification of new sources of gain from regional integration. The outcome, in any case, is that today regional integration is just as striking an institutional feature of the world 1

INTRODUCTION

economy as it was when Haberler first drew attention to its spread more than thirty years ago. The claims made on its behalf and its effects merit close analysis.

THE SCOPE OF THE SUBJECT Economic integration is basically concerned with the promotion of efficiency in resource use on a regional basis. Necessary conditions for its fullest attainment include: the elimination of all barriers to the free movement of goods and factors of production within the integrated area; and of discrimination on the basis of nationality amongst the members of the group in that respect. In addition, where resources are allocated by the price mechanism, measures will be required to ensure that the market provides the right signals. Institutions will also be required to give effect to the integrating force of the market. The term negative integration was coined by Tinbergen (1965) to denote those aspects of regional integration that simply involve the removal of discrimination and of restrictions on movement, such as arise in a process of regional trade liberalization. This he contrasted with positive integration, which he saw as concerned with the modification of existing instruments and institutions and the creation of new ones, for the purpose of enabling the market to function effectively and also to promote other broader policy objectives in the union. Arrangements for international economic integration take a variety of forms. For analytical purposes the principal ones to distinguish are customs unions, free trade areas, common markets, monetary unions, and economic and monetary unions. Between customs unions at one extreme and economic unions at the other, a spectrum of forms of integration may be envisaged, corresponding to the degree of integration, harmonization or co-ordination of other policies that is adopted. All such arrangements for international economic integration rest on the conclusion of agreements among the member states that are intended to have a lasting character, and which limit the unilateral use of those instruments of economic policy that are harmonized. The customs union (CU) is a basic form of integration, and the one with which orthodox theory is mainly concerned. It is characterized by a common external tariff (CET), tariff-free trade amongst its members, and the integration of tariff policy. The core 2

INTRODUCTION

of the EEC was such a customs union. The nineteenth-century Zollverein amongst the German states is an important early example. The free trade area is a related form of integration that, like the customs union, is also exclusively trade-focused. Currently this form enjoys considerable support and indeed, in numerical terms, free trade areas outnumber customs unions. Like the customs union, the free trade area is characterized by tariff-free trade amongst its members, but each country continues to determine its own external tariff. The European Free Trade Association (EFTA), that was established in reaction to the formation of the EEC, is a leading example of this kind of arrangement. The Canada—US Free Trade Agreement (CUFTA) took this form, as does the North American Free Trade Agreement (NAFTA), which brings together the US, Canada and Mexico. A common market is a third form of regional integration in which there is not only a common external tariff and tariff-free movement of goods and services, but also freedom of movement for factors of production—labour, capital and enterprise. The EC was popularly termed a common market from its inception, but it only began to fulfil the requirements of a fully unified market thirty years later as a result of action taken after 1985, following the adoption of wide-ranging proposals to complete the internal market. The action taken to implement these proposals in a variety of spheres was a necessary precursor to the inauguration of the Single European Market in 1993. Monetary union is another form of regional integration. Its establishment requires the adoption of a single currency, central bank and monetary policy. Normally monetary union is found only in conjunction with a customs union or a common market. Conventional wisdom suggests that monetary union should follow and not precede the institution of arrangements for market integration. Few examples exist of monetary unions amongst independent states except in Africa, although multi-country currency boards which share some of their qualities are found in the Caribbean and elsewhere. Under the Maastricht Treaty on European Union of 1992 (CEC, 1992a), the EU is committed to the establishment of a monetary union and it is planned that a single currency will be introduced on 1 January 1999. Economic and monetary union is the most developed form of integration. This form of regional integration, which the EU is to implement under the Maastricht Treaty, involves both a common market and a common money and the integration or harmonization 3

INTRODUCTION

of a range of policies in other areas. Whether and to what extent the operation of an economic and monetary union also requires fiscal integration in the shape of a significant community budget and the co-ordination and convergence of national budgetary policies in relation to deficit financing and debt policies is a subject of debate. As far as the EU is concerned, fiscal criteria have been laid down in the Maastricht Treaty that not only define fiscal qualifications for entry into the monetary union, but will also govern subsequent national budgetary practice. Arrangements for international economic integration clearly involve the acceptance of constraints on national policy making. And the farther along the road between customs union and economic and monetary union that countries progress, the greater those constraints become. What induces countries to accept such constraints? Evidently, twentieth-centry initiatives have been influenced by both political and economic objectives and there may in some instances have been a significant trade-off between the two. COMECON, for instance, had political benefits to the USSR as a means of politically delinking Eastern Europe from the West, but it proved costly to the Soviet Union in economic terms (Marrese and Vanous, 1983). Likewise, the Southern African Customs Union (SACU) is of doubtful economic value to South Africa on its present terms, but it produced substantial political benefits. The EC itself was strongly rooted in both political and economic goals, but for most of its members there were advantages in membership from both points of view. The importance of the political element as a factor in understanding the actual progress of economic integration throughout the world can scarcely be overrated. From the policy standpoint, the role of economic analysis is to identify and to quantify the economic effects and issues, so as to provide a basis for judging whether and to what extent the economic impacts reinforce or run counter to the political considerations. What are the expected economic benefits that provide the motivation for forgoing policy independence in relation to trade, factor movements and other major policy areas? The orthodox approach looks at this question in terms of the effects of integration on efficiency and welfare, and almost entirely in terms of static resource allocation gains. It demonstrates that both customs unions and free trade areas can generate potential gains in terms of the national income of the bloc by encouraging specialization amongst the member countries on the basis of comparative advantage. A common market can generate additional 4

INTRODUCTION

gains in real income by overcoming the barriers to factor mobility that allow differences in factor productivities to exist in the bloc. Although factor mobility can be a major source of additional gains over and above those generated by trade blocs like customs unions, free factor movement within a bloc does entail important additional constraints on national policies by making it difficult or impossible for national authorities to maintain autonomous jurisdiction in key policy areas, including taxation. In this sense, it is ultimately unobstructed factor mobility—factor market integration—that creates an integrated economy out of separate national economic entities. Essentially the orthodox gains from product and factor market integration arise from increased specialization according to comparative advantage among countries with different economic characteristics. This will result in a growth of intra-bloc trade of an inter-sectoral or inter-industry character. If economies of scale are introduced, market size itself becomes a source of integration gains which may then be derived from specialization even where member countries possess similar economic characteristics. A further source of potential gain may arise from improvements in the terms of trade of the bloc with the rest of the world. One major problem with the traditional theory is that, like the orthodox theory of international trade from which it sprang, it fails to explain the phenomenon of trade in similar products between member countries. Yet this kind of trade, termed intraindustry trade, was from the outset a marked feature of the expansion of intra-EC trade that followed the establishment of the customs union. In the early 1980s, two main explanations were developed for the phenomenon. One centred on the interaction of scale economies with product differentiation (Krugman, 1979). The other focused on monopolistic competition and the incentives it provides for active market segmentation by producers to permit the charging of separate prices in different markets (Brander and Krugman, 1983). These theories have been applied to the economics of international integration and have substantially changed its thrust, first by identifying new sources of potential allocational gains from market integration that result from its pro-competitive effects, and second by providing plausible grounds for supposing that the gains from market integration in an imperfectly competitive situation would be much larger than the orthodox customs union approach might suggest. The allocational gains from market integration, even in the broader context just outlined, are only one aspect of the economics 5

INTRODUCTION

of integration. Further allocational gains may also be derived from regional integration in respect of activities in which there are significant cross-border spill-overs in private activities and public policies. Furthermore, integration may affect the location of production, and, through this and other channnels, may have important implications for the geographical distributional of the gains from integration. The traditional theory focuses on the benefits to the bloc as a whole. Since it cannot be assumed that, even though integration is potentially beneficial to a bloc, market forces would necessarily produce an accceptable outcome for all member states, the compensation issue cannot be ignored. Market integration may also have effects on capital accumulation and growth. Finally, monetary integration may have important implications for economic stability. An analysis of the effects of international economic integration must therefore extend far beyond the traditional theory of market integration. This is reflected in the evolution of the post-Vinerian literature, which now has several main pillars, although its main emphasis continues to be on allocational gains. The analysis of certain other aspects—notably the accumulation and growth dimension—remains relatively underdeveloped. One important implication of the broader scope of the subject is that it evidently requires the evaluation of integration arrangements by reference to a broader set of criteria or goals than that of efficiency in resource allocation, which alone is considered in orthodox customs union theory. This poses the issue of the relative weight to be attached to the attainment of the different objectives by member states and the bloc, and of a possible trade-off between the attainment of one or other of the objectives. Ultimately, international economic integration has to be viewed as a state or process for enabling its participants to achieve a variety of common goals more effectively by joint or integrated action than they could by unilateral measures. In this light, as Tinbergen’s pioneering contribution emphasized, it is concerned with the problem of policy optimization in a broad sense within the integrated area. The contribution of international economic integration to the more effectual attainment of policy objectives can thus be appraised only in terms of a cost—benefit analysis that reflects the weight attached to all relevant dimensions of welfare and the terms on which they can be traded off against one another. The role of economic integration in that context in any case must be viewed in its proper dimensions. Membership of an 6

INTRODUCTION

economic bloc cannot of itself guarantee satisfactory economic performance to a member state, or to the group as a whole, or even better performance than in the past, whatever the criteria employed. The purely static gains from integration, although they may be significant, can be—and often are—dwarfed by the operation of factors of domestic or international origin that have nothing to do with integration. Moreover, many of the more fundamental factors influencing a country’s economic performance are dynamic and these are unlikely to be affected by integration except in the long run. The experience of many countries throughout the world, some of them very small, makes it plain that it is not a necessary condition for economic success that a country should be a member of an economic community, although this does not exclude the possibility that some of those that have been successful without integration might have performed even better as members of a suitable group. Above all, regional integration is plainly no substitute for sound domestic economic policies, and is unlikely to succeed in their absence. Nevertheless, there are many sources of potential gain from regional arrangements amongst suitable partners that have been identified by economic analysis. There exists an expanding range of empirical studies that attempt to calibrate the orders of magnitude of those gains for specific regional initiatives.

THE EMERGENCE OF THE THEORY OF INTERNATIONAL ECONOMIC INTEGRATION The branch of economic analysis that deals formally with the economics of regional international economic integration is of quite recent origin. The core of the subject is the theory of customs unions, which is commonly regarded as having taken shape with the publication of Viner’s pioneering study (1950), although important contemporary contributions were also made by de Beers (1941) and Byé (1950). This view of the late origins of customs union theory appears to be essentially well founded. The pre-Vinerian literature on customs unions in the twentieth century is sparse, and some of it suggests that, before the publication of Viner’s study, economists simply regarded customs unions as a step on the road to free 7

INTRODUCTION

trade. This is not to say that some of the more sophisticated ideas underlying current orthodoxy cannot be found in earlier writings. Since as many as sixteen customs unions were formed between 1818 and 1924, it would be surprising if this were not the case, for it is hardly to be imagined that the classical and neoclassical economists would have been wholly blind to their implications. O’Brien (1976), however, takes a more extreme position and has contended that the central features of the Vinerian analysis of customs unions—namely, trade creation and trade diversion—were a matter of essentially correct analysis by classical economists in the nineteenth century. The classical economists certainly discussed at length the effects of preferential commercial treaties such as the Methuen Treaty of 1703 (which admitted Portuguese wines into Great Britain on preferential terms in return for the removal of a prohibition on British woollen exports to Portugal) and the Cobden (Anglo-French) Treaty of 1860. In their turn, Adam Smith (1776), Ricardo (1817) and McCulloch (1832) each attacked the Methuen Treaty, essentially on the grounds of its trade-diverting effects. The Cobden Treaty also generated lively debate on its trade effects, and trade diversion was an issue in that debate. Moreover, when the German Zollverein, the most notable example of a customs union to be formed in the nineteenth century, was established in 1834, McCulloch and others subjected its provisions to detailed scrutiny in the course of which its diversionary trade effects were paid particular attention. In a different analytical tradition, the German economist List (1885) clearly viewed customs unions as protective devices for promoting infant industries, and in this way he can be said to have anticipated a stream of modern customs union theorizing initiated by Johnson (1965) and Cooper and Massell (1965). Relying on such contributions as these, O’Brien has claimed that: all Viner did was to start from the position of the classical economists and of Hawtrey and Lord Robbins, and then simply add to this the logical possibility of some trade creation, depending on the relative height of the preunion tariff and the common external tariff, and/or the possibility that both countries in the union were producers of the goods in question. (1976, p. 560) 8

INTRODUCTION

Strictly speaking, this may be true, but this apparently simple addition, embodying the concept of trade creation, is crucial to the orthodox theory of customs unions, since it is the sole source of benefit in that context. It is not clear that earlier economists appreciated its significance, as distinct from the idea of trade expansion. In any event, the discussions of customs unions by classical and neoclassical economists were largely incidental to their consideration of broader issues, and any theory was largely implicit. It is fair to say that historians of economic thought (Machlup, 1977) have failed to disclose any adequate treatment of the allocational effects of customs unions in the literature prior to the studies of Viner (apart from the work of de Beers, who was himself a pupil of Viner) and Byé. Those two economists together deserve to share the credit for their innovative theoretical insights, and for providing the first explicit formulation of a systematic analytical foundation for customs union theory. In any case, however significant the contributions of their forerunners are judged to be, it must be accepted that customs union theory is only one part of the economics of international integration. Not until the 1950s did there begin to develop, initially on the basis of policy-driven studies by economists such as Shoup (1953), Meade (1953, 1955a), and Tinbergen (1965), a major body of theoretical literature dealing systematically with the impact of other dimensions of policy integration, including its monetary, fiscal and spatial aspects, and much of this literature is in fact the product of the last fifteen years.

THE PRESENT REACH OF INTERNATIONAL ECONOMIC INTEGRATION: A GLOBAL PERSPECTIVE During the past four decades, many international economic blocs have come into existence. The most important of these, and indeed the largest regional bloc in the world, is the European Union (EU), whose foundation goes back to 1957. Under the Treaty of Rome of 1956, France, West Germany, Italy, Belgium, the Netherlands and Luxembourg set up the European Economic Community. The Community of the original six members was first enlarged in 1973 by the accession of the UK, Ireland and Denmark. In 1981 the southern enlargement of the Community began with the admission of Greece and was completed when Spain and Portugal were 9

INTRODUCTION

admitted in 1986. In 1992 the European Union was established on the foundations of the European Community, which constitutes the Union’s first and main pillar.1 The latest enlargement of the EU, in January 1995, brought in Finland, Sweden and Austria. East Germany automatically became part of the bloc following German reunification. One of the main current policy issues confronting the EU is how to proceed with the next enlargement that will extend to certain Central and Eastern European countries that have signed association agreements with the EU (including Hungary, Poland and the Czech Republic) and the Mediterranean states (Malta and Cyprus). The EU rests on a common market, with a common external tariff and trade policy, and freedom of movement for labour and capital. The EU conducts a number of common policies and also, in order to make the market effective, harmonizes a range of national policies over which member states retain jurisdiction. At the time of its first enlargement the EC decided to introduce its own regional policy, which was designed to complement national policies and to contribute to a reduction in economic disparities within the Community. As early as the late 1960s, proposals had been made and agreed by the Council that envisaged the development of the Community into a full monetary and economic union by 1980. Little progress was made towards this objective until the commencement of the present decade, partly as a result of the stresses accompanying successive enlargements and the impact of a succession of international monetary and economic crises. With the signing of the Maastricht Treaty of 1992, Economic and Monetary Union (EMU) returned to the centre of the stage and preparations are currently being made for the introduction of a single currency. This has already had important implications for national fiscal policies as a result of the criteria for macroeconomic convergence that must be satisfied if member states are to qualify for entry. There will be further important implications for national fiscal and possibly for EU budgetary policies when a single currency comes into operation. The European Free Trade Association is a second Western European grouping but is now of limited importance. The formation of a Europe-wide free trade association was originally proposed by Britain in the mid 1950s in order to allow those European countries that, like itself, were not prepared to commit themselves to a common agricultural policy, or to other political and economic objectives envisaged in the establishment of the 10

INTRODUCTION

EEC, nevertheless to enjoy the benefits of free trade in industrial products. The UK’s proposals proved to be unacceptable to the EEC, and the free trade association that was in the end established in 1960 was confined to the UK and the smaller countries of Western Europe, namely Switzerland, Austria, Denmark, Norway, Sweden and Portugal. Finland became associated with EFTA in 1961. EFTA is purely a trade grouping and has none of the supranational features of the EU. Following the enlargement of the EEC by the entry of the UK, Ireland and Denmark in 1973, EFTA’s importance diminished considerably. However, simultaneously with the enlargement, those EFTA countries that were left outside were offered links with the enlarged EEC through a series of free trade agreements that were aimed at creating a wider European industrial free trade area. Towards the end of the last decade, the potentially adverse trade implications for EFTA of the EC’s policy-deepening embodied in its Single Market programme led the Community to offer EFTA countries a more structured relationship involving the creation of the so-called European Economic Space. A requisite for membership was that EFTA countries would not only have to embody much of the existing body of EC legislation into their own systems but also to commit themselves to adopting future legislation, without their having any right to influence the character of that legislation. Confronted with that choice, Austria, Finland and Sweden opted for accession to the EC. In consequence, the geographical scope of EFTA today is limited to Iceland, Liechtenstein, Norway and Switzerland, of which all but Switzerland participate in the European Economic Space. Two other initiatives for regional integration amongst advanced market economies should also be noted. These concern North America and Australasia. In 1988 the Canada—US Free Trade Agreement came into effect. Second, building on the free trade area that has been in force between Australia and New Zealand since 1965, agreement was reached to develop still closer economic relations between these two countries. This is expressed in the 1983 Closer Economic Relations Agreement that has subsequently been strengthened. The instances of economic integration so far mentioned have all involved advanced market economies. In the past, the former planned economies of Eastern Europe also operated an integration bloc. ‘Socialist’ integration operated under the auspices of the Council for Mutual Economic Assistance (CMEA), 11

INTRODUCTION

usually referred to in the West as COMECON. This institution was established in 1949 and included the USSR, Albania, Bulgaria, Czechoslovakia, the German Democratic Republic (GDR), Hungary, Poland and Romania, together with the nonEuropean countries of Cuba, Mongolia and Vietnam. Its proclaimed objects were to accelerate economic development and to establish a more rational division of labour among its member countries. In its heyday its economic interdependence, as measured by the share of intra-bloc trade in total exports, attained a level reached by the EU only in the current decade. With the end of Soviet hegemony over Eastern and Central Europe, and the general move towards market economies, COMECON has disappeared and trade amongst the former m e m b e r s h a s u n d e rg o n e a s h a r p d e c l i n e . T h e n e w l y independent states that have succeeded the former Soviet Union itself have not for the most part adopted significant initiatives for integration amongst themselves. The most numerous instances of international economic integration are to be found outside Europe, in the developing world. In Africa alone, eight or nine groupings of this kind have been established. Most trade-based integration initiatives in Africa have so far made little or no positive contribution to trade or economic development. Efforts to improve the performance of some of the potentially important groupings are currently in train. One such reform involves ECOWAS. This is the largest African integration bloc and includes both the anglophone and the francophone countries which generate much of the GDP of middle Africa. It has made little real progress since its establishment in 1975, but a revised treaty that was signed in 1993 and the restructuring of certain of its operations may result in improved performance. A second initiative involves the West African Economic and Monetary Union, which was formally initiated in 1994. Its object is to transform the existing successful monetary union of West Africa into a full economic and monetary union. It intends to do this by undertaking operations previously carried out by the roughly coterminous but now abolished Economic Community of West Africa (CEAO), which was a purely francophone trade bloc. The third reform centres on the Southern African Customs Union, which has an unbroken and in its limited sphere a successful history which goes back some eighty years. It is the only African instance of a fully operational customs union. The agreement is currently being renegotiated with every prospect of a successful outcome. The largely parallel rand monetary 12

INTRODUCTION

area constitutes a long-standing example of successful integration in monetary affairs. In Latin America, apart from the vestigial Latin American Integration Association (LAIA) that was set up under the Treaty of Montevideo of 1980 in succession to the Latin American Free Trade Association (LAFTA) of 1960, three main regional blocs exist. The one with perhaps the greatest potential results from the major new initiative taken by Argentina, Brazil, Uruguay and Paraguay in 1991 in which the four countries, which account for a large part of the population and income of Latin America, committed themselves to establish a Common Market of the Southern Cone (MERCOSUR). The Central American Common Market (CACM), which was set up under the Managua Treaty of 1960, to include Guatemala, Honduras, Costa Rica, Nicaragua and El Salvador, is a second arrangement which initially achieved a considerable measure of success. CACM has recently taken steps to reactivate itself, and it has made substantial progress towards re-establishing the high level of intra-bloc trade that it had earlier achieved and towards reincorporating Honduras into the market. The third arrangment is the Andean Group, which was set up under the Cartagena Agreement of 1969 and which now includes Colombia, Venezuela, Ecuador, Peru and Bolivia. The establishment of MERCOSUR has reinforced the Andean Pact’s initiatives for the creation of a customs union and common market. Following the possibility held out by President Bush in 1990 of the establishment of free trade areas between the US and Latin American countries and blocs, and the subsequent formation of NAFTA, which incorporates Canada, the US and Mexico, the development of far-reaching free trade arrangements covering a large part of the western hemisphere has become a real prospect. Negotiations on the formation of a Free Trade Area of the Americas are to begin in 1988. A Caribbean Community (CARICOM) was set up in 1973 on the foundations of the Caribbean Free Trade Area (CARIFTA) by the English-speaking countries of the area, with the exception of some of the smallest countries of that group and the Bahamas. An important element of CARICOM is the Caribbean Common Market. A special regime exists within it for its less developed members in the shape of the Eastern Caribbean Common Market (ECCM). The West Indian Commission (1992) has recommended a move towards a Single Market and Economy. In Asia, policy-led institutionalized regional economic integration 13

INTRODUCTION

has so far made little progress. The only example is represented by the Association of South East Asian Nations (ASEAN), whose foundations go back to 1967. Although ASEAN’s origins lie in political and strategic considerations, it soon developed an interest in economic co-operation. Divergent trade policies have hitherto hindered the implementation of any significant preferential arrangements, but with the institution of the ASEAN Free Trade Area (AFTA) in 1992, the formation of a free trade area by 2003 has been agreed upon. Another new initiative for regional economic co-operation in Asia has involved the formation of the wider Asian Pacific Economic Co-operation (APEC) forum. This body includes the advanced countries of the Pacific Rim, the newly industrializing economies and the developing countries of the region. At the end of 1994 the forum adopted an agreement to institute free trade by 2020, with the industrialized countries reaching that goal by 2010. The agreement is, however, non-binding. In general, Asia has been unenthusiastic about formal regional integration with an institutional basis. What is favoured is sometimes described as ‘open regionalism’. This is characterized by informal efforts to widen markets without tariff discrimination. Until the last decade most regional groupings of developing countries operated behind highly protectionist tariffs and non-tariff barriers that were designed to stimulate import substitution. Almost everywhere their performance has been poor and frequently punctuated by severe stresses prompted by controversy over distributional issues. A satisfactory solution to those issues proved to be elusive, although a variety of ad hoc palliatives have been resorted to. The outcome has been dissolution in one or two instances (as in the case of the EAC, which broke up in 1978) and stagnation in most of the others. The new wave of integration among developing countries that began in the 1980s is taking place in the context of the widespread adoption of trade liberalization initiatives as those countries have moved to outward-looking trade policies. That context, together with a better understanding of the sources of benefit and of the operational issues of integration, suggests that the schemes of the new wave have the potential to perform far better than their predecessors. Whether this potential is achieved is likely to depend to a considerable extent on the ability of the new initiatives to implement policy-deepening measures of regional integration, on which their contribution to cost reduction and investment creation will largely hinge. 14

INTRODUCTION

PLAN OF THE BOOK This book sets out to provide a concise exposition of the economics of international economic integration, paying particular attention to important recent contributions that have substantially altered the orthodox perspective. A secondary concern has been to indicate the bearing of the analysis, wherever possible, on current policy issues in the EU and other groupings. Chapter 2 analyses the simple economics both of customs unions and of free trade areas on orthodox assumptions, in terms of resource allocation gains. In Chapter 3 some of the restrictive assumptions of the basic analysis are relaxed, specifically by introducing terms-of-trade effects and economies of scale. Chapter 4 discusses the question of whether a customs union is inferior to a policy of unilateral tariff reduction and considers whether the second-best nature of customs unions is affected by the introduction of public goods. Chapter 5 considers the rationale for going beyond customs unions and common markets to integrate other economic policies in terms of cross-border spillovers. Chapter 6 analyses the specific effects of instituting a common market for factors of production. Chapter 7 discusses the so-called ‘new economics’ of integration, which continues to focus on the allocational gains from market integration but does so in a context of imperfect competition and scale economies. Chapter 8 discusses the relationship between regional integration and the operations of TNCs. Chapter 9 considers the role of the budget and of public finance in economic groupings in terms of the efficiency considerations that have a bearing on the assignment of public economic functions from the member states to the bloc. Chapter 10 discusses tax harmonization, primarily from the standpoint of resource allocation. Chapter 11 analyses the costs and benefits of monetary integration and the problem of fiscal policy in a monetary union. Chapter 12 analyses the distributional and locational effects of regional integration and the rationale for regional policies in a common market. Chapter 13 discusses the experience of ‘first-wave’ regional initiatives outside Europe, and the context and objectives of the so-called ‘new’ regionalism. Throughout the book the terms ‘economic grouping’ and ‘regional economic bloc’ are used to refer generally to any of the forms of international economic integration distinguished in this chapter. The terms ‘customs union’, ‘free trade area’ and ‘common market’ refer specifically to those forms of integration. The term ‘economic community’ is used to refer broadly to groupings that, in addition 15

INTRODUCTION

to operating a customs union or a common market, attempt to harmonize or integrate a range of other economic policies. It will be apparent that the book is not primarily a study of the problems and policies of the EU. Nevertheless, the analysis provides a basis for understanding most of its key policy issues. With that object in mind, throughout the book, appropriate reference is made to its experience and policies.

NOTE 1

The Maastricht Treaty on European Union of 1992 (TEU) established the European Union, which rests on three ‘pillars’. The Union is founded on the European Community which is the well developed first pillar originating in the Treaty of Rome that is concerned with the common market and other common economic policies. The Treaty establishing the European Economic Community was amended in important respects by the Maastricht Treaty, notably by the addition of provisions on monetary union, subsidiarity and cohesion. The European Community (a term entrenched in the TEU) has the task of conducting the activities of the Union that are concerned with economic integration. Two other pillars were added by the TEU that deal respectively with co-operation on justice and home affairs, and with a common foreign and security policy. The term European Union is in general use to refer to all activities undertaken by the Union unless legal enactments concerning the institutions or individual treaties are in question. This practice has been followed, for the sake of simplicity, in this volume.

16

2 THE THEORY OF CUSTOMS UNIONS AND FREE TRADE AREAS

THE THEORETICAL CONTEXT The essential features of a customs union (CU) are: • the elimination of tariffs on imports from member countries; • the adoption of a common external tariff on imports from the rest of the world; • the apportionment of customs revenue according to an agreed formula. The establishment of a customs union will generally alter the relative prices of goods in the domestic markets of member countries, with repercussions on trade flows, production and consumption. The theory of customs unions analyses these effects and their implications for resource allocation and for the welfare of a participating country, for the group as a whole and for the world. Since the common external tariff can be set at any desired level, in principle a customs union can establish the tariff, and therefore union prices, at levels that will maximize the social welfare, however defined, of the participating countries as a group. The effects, and the gains and losses that a customs union may give rise to, result from its impact on: (1) the allocation of resources and international specialization; (2) the exploitation of scale economies; (3) the terms of trade; (4) the productivity of factors; (5) profit margins; (6) the rate of economic growth; (7) the distribution of income. The classical theory is exclusively concerned with the first three aspects for a customs union and for trade in final products. The fourth aspect is excluded from conventional economic analysis by assuming that technical progress is exogenous and that production is carried out by processes that are technically efficient. 17

CUSTOMS UNIONS AND FREE TRADE AREAS

Distributional considerations are for the most part disregarded by the focus on the group. The analysis is comparative static in nature. It is assumed that: • • • •

• • •

there is pure competition in commodity and factor markets; factors are mobile within countries but not between them; transport costs are ignored; tariffs are the only form of trade restriction employed—and for simplicity they are often assumed to be specific in form rather than ad valorem; prices accurately reflect the opportunity costs of production; trade is balanced (exports equal imports); resources are fully employed.

This restrictive list of assumptions is shared with the standard pure theory of international trade. Its limitations are not so damaging for the central importance of the analysis as may be thought. The theory provides insights of major importance, and has demonstrable explanatory power. Nevertheless, the character of modern production and trade, and, in particular, the importance of product differentiation, intra-industry trade and imperfect competition mean that the outcome of customs unions (and other trading blocs) may significantly differ from the ‘predictions’ of orthodox theory. The newer theories, discussed in Chapter 7, specifically address these issues and supplement the orthodox theory in crucially important respects.

CUSTOMS UNIONS AND RESOURCE ALLOCATION The basic theory analyses the effects of customs unions on resource allocation, specialization and welfare, for a member state, for the group as a whole and for the world. Not all customs unions would have allocative effects. For instance, if the prospective member countries initially enjoyed identical tariff rates on all commodities, and the tariffs were not redundant, no trade effects would follow from the introduction of the customs union if tariffs remained unchanged. Each country in the initial situation would be producing the output, for which the domestic supply price equalled the world price plus the tariff. Neither the world price 18

CUSTOMS UNIONS AND FREE TRADE AREAS

nor the tariff would be affected by the institution of the union, and consequently nothing would change. For a customs union to have allocational effects it is necessary that the tariff rates of prospective members should differ for at least some products, unless some of the tariffs are ineffective. The resulting harmonization of tariffs gives rise to the allocational effects of the customs union. Harmonization may take the form of averaging, or some other intermediate position between the highest and the lowest rate may be adopted. The orthodox theory of customs unions analyses the effects of customs unions on resource allocation in terms of the trade creation and trade diversion that would result. Trade creation refers to a union-induced shift from the consumption or higher-cost domestic products in favour of lowercost products of the partner country. This shift has two aspects: •

•

the reduction or elimination of the domestic production of goods that are identical with those produced abroad, the goods instead being imported from the partner country; increased consumption of partner-country substitutes for domestic goods that formerly satisfied the need at a higher cost.

The first gives rise to the production effect—the saving in the real cost of goods previously produced domestically; the second gives rise to the consumption effect—the gain in consumers’ surplus from the substitution of lower-cost for higher-cost means of satisfying wants. The two together constitute the trade creation effect of the union. Trade diversion refers to a union-induced shift in the source of imports from lower-cost external sources to higher-cost partner sources. This shift can also be regarded as having two aspects: • •

an increase in the cost of the goods previously imported from abroad, owing to the shift from foreign to partner sources; a loss of consumers’ surplus resulting from the substitution of higher-cost partner goods for lower-cost foreign goods of a different description.

Such shifts together constitute the trade diversion effect of a customs union.1 The merits of particular customs union situations are then 19

CUSTOMS UNIONS AND FREE TRADE AREAS

evaluated, using as the sole criterion the relative magnitudes of trade creation and trade diversion. A union that is on balance tradecreating is regarded as beneficial to welfare, whereas a tradediverting union is regarded as detrimental. In this context it is to be emphasized that the magnitudes of trade creation and trade diversion depend not merely on union-induced changes in the volumes of trade from different sources, but also on the associated price and cost changes. The terms ‘trade-created’ and ‘trade-diverted’ are sometimes used to refer to the volume changes alone. The rest of this section analyses in some detail, for several alternative cases, the resource allocation and other effects of customs unions, utilizing the methods and spirit of the orthodox analysis. The treatment is not intended to be comprehensive—for that purpose a multitude of cases would have to be examined, which would require a book in itself. The examples chosen nevertheless suffice both to illustrate the method of analysis and to bring out the essential nature of the central issues. The simplest analysis that is capable of eliciting the basic issues of customs union theory requires a three-country model, as opposed to the two-country model customarily employed in international trade theory. Accordingly, in the following analysis two countries are assumed to form a customs union, while a third country, which can be taken to represent the rest of the world, is excluded. Little of importance is lost by limiting attention to such a three-country world in the first instance. Returns to scale will be disregarded, and terms-of-trade effects for the union will be excluded by assuming a perfectly elastic supply price for its trade with the rest of the world. Some of the implications of relaxing these two limiting assumptions will be considered in Chapter 3. Figure 2.1 depicts the demand and supply conditions in the market for an identical product in two countries—H, the home country, and P, the partner country—contemplating the establishment of a customs union. The supply price of the same product from the rest of the world is also indicated. DH is country H’s demand curve for the product, SH is its supply curve, and (SH+ MP) is the supply curve in country H of the product originating in the customs union. It combines the supply curve of country H with the supply curve of imports from country P, assuming that the goods of the latter are admitted free of duty. PW indicates the supply price of the rest of the world’s product to countries H and P and is assumed to be constant. Market conditions in country P are similarly shown in the lower figure, DP being country P’s 20

CUSTOMS UNIONS AND FREE TRADE AREAS

Figure 2.1 The effects of a customs union

demand curve for the product and SP its supply curve. The specification of market conditions for both member countries permits the important aspect of tariff harmonization to be explicitly treated. Three cases will be considered, which are distinguished from each other only by alternative assumptions about the pre- and post-union tariff positions. 1. Suppose first that prior to customs union a tariff of PWTH was in force in country H and one of PWTP in country P. In that case 21

CUSTOMS UNIONS AND FREE TRADE AREAS

the domestic demand of each country would be entirely supplied by domestic production. If a customs union were formed on the basis of tariff averaging so that CET=1/2(TH+TP), the new common external tariff would be PWCET; but this would be ineffective, or redundant, since at price CET, supply would be greater than union demand. Price would therefore settle at CET’, where supply would equal demand. Domestic consumption in country H would increase to Q, and domestic production would decline from N to L. Country P would then produce T, consume R and export LQ (=RT) to country H. This situation for country H would be one of trade creation. The saving in cost on the initial domestic production in country H that was replaced by imports (the production effect—Vinerian trade creation) would be denoted by the triangle ABD; the gain in consumers’ surplus from the substitution of imports for other goods would be ADC. The total gain from trade creation would be the sum of these two areas, which would be approximately equal to half the product of the total change in imports (LQ) and the fall in price. The area above SH marked off by the lines TH and CET’ would represent an internal transfer from producers’ surplus to consumers’ surplus which, for the country as a whole, would cancel out. In the case of country P, which was also initially self-sufficient, the customs union would cause a rise in price. Consumers would suffer a consumption loss equal to d. There would also be a production effect, denoted by e, as additional resources were drawn into the industry to produce an additional amount ST of the product. These costs, however, would be more than outweighed by the extra income earned from the newly created export trade to country H (hatched rectangle). In that case country P would clearly also be in a more favourable position. As to the rest of the world, its trade with the customs union would be unaffected by the formation of the customs union, being zero both before and after. In these circumstances consideration of the welfare effects of the union could be confined to the countries that constituted it. The formation of the union would result in an improved allocation of resources, and each member would be in a superior position compared with the pre-union situation. 2. Alternatively, suppose that while the initial tariff in country P was again P WT P, in country H it was lower at P WT’ H (=P WCET). In this case prior to customs union the initial situation for 22

CUSTOMS UNIONS AND FREE TRADE AREAS

country P would obviously be as before. In country H, however, domestic demand at the pre-union price of T’ H would be met partly by domestic production and partly by imports. The quantity consumed would be P, domestic production would account for M, and imports from the rest of the world would make up the difference—MP. Total tariff revenue would amount to MP× P WT’ H. If in these circumstances a tariff-averaging customs union were formed with the common external tariff equal to P WCET’ the effect on country H would be as follows. The amount demanded would increase to Q, as in the first case, and domestic production would fall to L. There would thus be a saving on production cost, denoted by the triangular area marked a, in respect of the reduction in the initial domestic output and a gain in consumers’ surplus, denoted by the triangular area marked c, in respect of the expanded consumption. Together these magnitudes would represent a trade creation effect. However, in addition, the produce formerly imported from the rest of the world would in this case be shifted to a higher-cost source, country P, involving an increase in outlay in respect of the initial import amounting to MP×P WCET’. This magnitude would represent trade diversion. The magnitudes of trade creation and trade diversion would have to be compared in order to ascertain whether the customs union was beneficial or detrimental for country H. In this specific example, it would clearly be detrimental. The effects on country P of the formation of the customs union would clearly be identical to those that it would experience in the first case: it would thus be in a superior position compared with the pre-union situation. As far as the rest of the world was concerned, its trade with the customs union would be reduced. Since the world supply curve is assumed to be perfectly elastic, the impact of this reduction on the welfare of the rest of the world can be disregarded. However, the reduction in imports from the rest of the world would indicate that the union was, on balance, a trade-diverting one. 3. To conclude, a third case may be considered. It will be supposed that the tariff initially in force in country H is PWCET’, giving a tariff-inclusive price of CET’, while in country P the tariff PWTP continues to be in force, as in the previous case, giving a tariff-inclusive price of TP. It will now be assumed that the common external tariff of the union is established, not by tariff-averaging, 23

CUSTOMS UNIONS AND FREE TRADE AREAS

but by the alignment of country P’s tariff with that of country H, which involves an increase in the average level of protection. The formation of a customs union on such a basis will leave production and consumption of the product in country H wholly unaffected. The effect of the customs union will be to bring about a replacement of its imports from the rest of the world by an identical quantity from country P. This would amount to a case of pure trade diversion from the standpoint of country H and the cost that it would thereby incur would be represented by its loss of tariff revenue. As far as country P is concerned, however, the effect of customs union on the quantity of its exports to country H and the favourable effect on its income will be identical to both the cases previously considered. As to the rest of the world, since its export supply curve is perfectly elastic, the welfare effect of the reduction in its exports can again be ignored. In terms of resource allocation effects, this third case is the least favourable of the three that have been considered, since trade creation would be altogether absent. Customs union would result in a deterioration in global resource allocation, accompanied by a redistribution of income from country H to country P. The three cases just considered involve the analysis of a specific market situation, which may be regarded as a special case. The method of analysis is, however, equally applicable to a large range of other situations. One such case often considered is that in which, in the initial situation, one of the partners already competes in the home country with the rest of the world. No significant points are neglected as a result of the choice adopted here, however, nor are any of the effects considered peculiar to these cases. In other relevant situations the effects can be expected to differ in direction and extent but will be similar in kind. There is, however, a more important point that bears on the generality of the analysis, which must be briefly mentioned. In the interests of simplicity, customs union theory has here been presented in terms of a partial equilibrium analysis that looks at the market for a single product. Is the generality of the analysis limited by this procedure? Specifically, can the analysis be applied to the practically important case of a customs union extending to many products? The requirement that must be satisfied in order that partial equilibrium analysis should not violate general equilibrium requirements is that the relative prices of all other goods should be unaffected by the changes in the market under consideration, so 24

CUSTOMS UNIONS AND FREE TRADE AREAS

that the price variable can be interpreted as the price of the good in question relative to the price index for other goods. Provided that this requirement is met—as effectively would be the case where the market being analysed is so small that changes affecting it would not have significant repercussions on supplies and demands elsewhere in the economy—the approach can be justified. Supply and demand analysis and figures relating to the market for a single product (or sector) can then specifically be given a general equilibrium interpretation, as is sometimes done (Corden, 1974; ElAgraa and Jones, 1980; de Melo, Panagariya and Rodrik, 1993). But a partial equilibrium analysis of a single market cannot appropriately be undertaken on this basis in circumstances in which the relative prices of other goods would be significantly affected. A fortiori, it would be inappropriate to seek to determine the consequences for the whole economy of the formation of a customs union that affects all traded products by simply aggregating outcomes for particular markets that rest on partial equilibrium assumptions. The alternative general equilibrium approach to the analysis of customs unions is well represented in the literature. The earliest post-Vinerian formulations in this vein were presented almost exclusively in terms of an application of the standard two-good general equilibrium model. The first significant advance is represented by the three-country two-product models developed by Vanek (1965) and Kemp (1969). An alternative model first put forward by Meade (1955a) deals with three countries and three products. The general equilibrium approach, and more specifically those variants mentioned here, manifests both strengths and weaknesses. In principle, in the interests of symmetry and generality, a three-product approach has substantial merit for the analysis of customs union issues. When more than two products are taken into account, however, the analysis becomes highly intricate and can allow only inadequately, if at all, for trade creation and trade diversion and the impact of union on intra-union terms of trade— central features of customs unions that can more easily be given prominence in simpler formulations. The issues—which can only be alluded to here—are explored in Berglas (1979), Collier (1979), where the choice of model is considered, Lloyd (1982), who compares several variants of the three-product approach, and Wooton (1986). The analysis of this chapter specifically assumes that the formation of the customs union does not affect the terms of trade of the member countries with the rest of the world. This means, as already 25

CUSTOMS UNIONS AND FREE TRADE AREAS

noted, that the effect of the customs union on the welfare of the rest of the world can be disregarded. Consequently, a given change in the welfare of the members of the union will mean a corresponding change in the welfare of the world as a whole, including the members of the customs union. In principle, the formation of a customs union can make the union in net terms either better off (if the trade creation effect predominates) or worse off (if the trade diversion effect predominates). It is evidently also possible for at least one of the members to be injured in a tradecreating union and, similarly, for at least one member country to benefit in a trade-diverting union. Thus even a trade-creating union can only be said to be potentially beneficial to all members. If the distribution of gains and losses among the member countries is not a matter of indifference, then, unless the gains and losses are appropriately distributed, compensatory transfers may be required to ensure that no country loses. In practice, the determination of the structure of the common external tariff is a principal means by which attempts are made to ensure an initially acceptable distribution of gains and losses, but this approach may not be the most efficient way of dealing with the distributional problem. The variety of models employed in relation to the analysis of customs unions can be bewildering, but one fundamental proposition is consistent with them all, namely the demonstration by Kemp and Wan (1976) that it is always possible to raise the level of welfare of the customs union as a whole, and of every member. Necessary conditions are (1) that the common external tariff of the group is set appropriately (for instance, by choosing the Vanek compensating tariff, which maintains the aggregate trade of the customs union members with the rest of the world at its preunion levels), and (2) that lump-sum compensatory taxes and transfers between the members of the union are made to ensure that each country benefits. In other words, a trade-creating union is always possible in principle, and so a customs union can guarantee that it can raise the welfare of each of its members. Strictly, this conclusion also requires that the rest of the world does not react.

The conditions for a trade-creating customs union A final issue for consideration is whether any general statements can be made about the actual circumstances that will determine 26

CUSTOMS UNIONS AND FREE TRADE AREAS

whether a particular customs union will be predominantly tradecreating rather than trade-diverting. If this could be done, it would obviously be very useful. Many economists have offered such generalizations; but analytical elaboration has shown that most of them depend on the circumstances of special cases. The few surviving generalizations may usefully be summarized at this point. Some follow directly from the previous analysis, but others could not be demonstrated without carrying the analysis of the present chapter somewhat further. The most helpful general statements that can justifiably be made seem to be as follows: •

•

•

•

The larger is the economic area of the customs union, and the more numerous are the countries of which it is composed, the greater will be the scope for trade creation as opposed to trade diversion. The relative effects can be related to the height of the ‘average’ tariff level before and after union. If the post-union level is lower, the union is more likely to be trade-creating; if it is higher, trade diversion effects may be more likely. Trade creation is more likely the more competitive are the economies of the member states, in the sense that the range of products produced by higher-cost industries in the different parts of the customs union is similar. Likewise, the smaller the overlap, the smaller will be the possibilities of reallocation, which is the source of trade creation. For a specified overlap, trade creation is more likely to predominate the greater are the differences in unit costs of protected industries of the same kinds in the different parts of the customs union, since these will determine the allocation gains to be derived from free trade among the members.

Intuitively these propositions appear to be reasonable. Nevertheless it must be emphasized, as Viner himself asserted, that: Confident judgement as to what the overall balance between these conflicting considerations would be, it should be obvious, cannot be made for customs unions in general and in the abstract, but must be confined to particular projects and be based on economic surveys thorough enough to justify reasonably reliable estimates as 27

CUSTOMS UNIONS AND FREE TRADE AREAS

to the weights to be given in the particular circumstances to the respective elements in the problem. (1950, p. 52) This judgement is as valid today as it was in 1950.

THE ECONOMICS OF A FREE TRADE AREA Although most of the basic theory of market integration is framed in terms of customs unions, a similar analysis may readily be undertaken for free trade areas. Two basic features distinguish a free trade area from a customs union: • •

The member countries retain the power to fix their own separate tariff rates on imports from the rest of the world. The area is equipped with rules of origin, designed to confine intra-area free trade to products originating in, or mainly produced in, the area.

The purpose of rules of origin is to limit trade deflection, that is the redirection of imports through the country with the lowest tariff for the purpose of exploiting the tariff differential. As in the case of a customs union, the formation of a free trade area may be accompanied by trade creation or trade diversion, but there are important differences in the operation of the two alternative forms of integration. In this section the economics of free trade areas will be briefly analysed. Their effects will be considered first for a single country and then, analogously with the previous section, in such a way as to make explicit the effects on each of the two members of which the free trade area is assumed to be composed. Some comparisons between the relative effects of a customs union and a free trade area will be made for specific situations. For this purpose a tariffaveraging customs union, that is, one in which the CET=1/2(TH +TP), will again be taken as the basis of comparison. In the light of these comparisons some propositions will be presented on the relative merits of customs unions and free trade areas from the standpoint of their resource allocation effects. It should again be emphasized that the treatment is not claimed to be comprehensive. The technique employed can, however, readily be applied to 28

CUSTOMS UNIONS AND FREE TRADE AREAS

illuminate the effects of free trade areas in the presence of any alternative assumed market conditions and tariff positions for a particular product.

The effects of a free trade area from a single-country standpoint Consider two countries, H and P, both of which produce wholly domestically an identical product, which is subject to a different customs tariff in each country, a lower duty PWTP being imposed in country P and a higher duty PWTH in country H. Assume that the two countries form a free trade area in which the rules of origin prevent inflows from the rest of the world to country H through country P. Between the markets of countries H and P only the areaorigin product enjoys tariff-free movement. This difference in treatment may or may not create a price differential between the area-origin product and non-area product. The effects of such a free trade area will now be analysed with reference to Figure 2.2. It is assumed for simplicity that country H’s tariff before integration is prohibitive, in the sense that it excludes all imports. Country H’s supply curve is S H, its tariff is P WT H, and its production is L at price TH. Country P’s tariff is P WT P, and its

Figure 2.2 The effects of a free trade area on a single country

29

CUSTOMS UNIONS AND FREE TRADE AREAS

supply curve is horizontally added to that of country H to arrive at curve SH+P. PW denotes the world supply price. If a free trade area is formed, the price of the area-origin product in country H can never fall below TP so long as the free trade area as a whole remains a net importer. At the same time, the price can never exceed TH, which equals PW plus PWTH, country H’s protective duty. Consequently, from the standpoint of country H, the effective supply curve of the product, including area and non-area product, is TPBFGK. In a free trade area the amount that country P will be willing to supply will depend on the price, and the price that is obtainable will depend on country H’s demand curve. Two possibilities are considered in Figure 2.2, which correspond to the alternative demand curves DH and D’H. If country H’s demand for the product were represented by DH (relatively inelastic at price TH by comparison with D’H), the price in country H would be TP, with country P supplying L’R at that price. In this case triangle a would represent trade creation, and triangle c would represent the consumption effect of the reduction in the price of product X in country H brought about by the free trade area. If, instead, country H’s demand were represented by D’H, the price in country H (PH) would be nearer to the upper limit TH above which imports would be supplied from the rest of the world. In this case country H would supply its own market to the extent of N, and country P would supply country H to the extent of NN’. Trade creation would then be denoted by the smaller triangular area below the intersection of D’H and SH and above the horizontal line at PH. In general, in a free trade area, country P would supply country H at any price above TP, up to its total capacity to supply, making up any consequential shortfall in its domestic market by imports from the rest of the world. In this way the price in country P’s market would be kept down to TP irrespective of the final price for product X in country H. The induced change in trade flows is termed indirect trade deflection (the substitution by country P of non-area for area products). It cannot be eliminated by the rules of origin of a free trade area. The impact of indirect trade deflection is further clarified for each member in the following section, which analyses the operation of a free trade area from the standpoint of both countries. A free trade area: two countries Figures 2.3 and 2.4 depict the demand and supply curves in countries H and P for a given product. P W again denotes the 30

CUSTOMS UNIONS AND FREE TRADE AREAS

Figure 2.3 A comparison of (a) free trade areas and (b) customs unions, 1

world supply price. Before integration, country P has a relatively low tariff PWTP, giving a tariff-inclusive price of TP. In the figures: triangle a represents trade creation (the production effect); rectangle b any excess outlay on the partner product (trade diversion) for the initial import; triangle c the consumption effect in country H, which is positive; triangle d the consumption effect in country P, which is negative; and triangle e the production effect in country P, which may be neutral or negative and is relevant mainly for customs unions. Two illustrative cases will be analysed, and for each a comparison will be made with the alternative of a tariff-averaging customs union. 1. In the first case (Figure 2.3) countries H and P are assumed 31

Figure 2.4 A comparison of (a) free trade areas and (b) customs unions, 2

CUSTOMS UNIONS AND FREE TRADE AREAS

to have similar demand conditions; but country H is a relatively inefficient producer, whereas country P’s supply curve is relatively elastic and competitive, although above world market price P for outputs in excess of L". Before the free trade area is formed, country P consumes and produces M at price TP, its tariff keeping out all imports. Country H produces L and consumes N, the difference LN being imported at price PW from the lowest-cost source, namely the rest of the world. Customs revenue in country H is denoted by LN×PWTH. If countries H and P formed a free trade area (Figure 2.3a), area supply at price TP (=OM+OL’) would clearly be less than area demand at that price (M+N’), but the difference (L’N’) would be less than country P’s capacity to supply at that price. In a free trade area that excluded the least-cost source of supply, country P would supply country H’s market with L’N’ (=L"M) at price TP, leaving an amount equal to OL" for its home market, its remaining requirement (L"M) being imported from the rest of the world at price PW. In this case after integration there would be a single equilibrium price in the free trade area that would be equivalent to the lower of the two members’ prices before the establishment of the free trade area. It can be seen that in country H the production effect a (Vinerian trade creation) plus the consumption effect c would outweigh the cost of trade diversion (stippled area b). The difference between the cost of trade diversion, b, and the initial customs revenue represents an internal transfer from the exchequer to consumers and not a loss of real income to the community. In country P the same amount would be produced and consumed as before, at the same price, but government revenue would increase by an amount that equalled the hatched rectangle. This would represent a national income gain to country P. As far as the rest of the world is concerned, its exports would clearly be larger than before (L"M>LN) because of the shifting of country P’s supply to satisfy country H’s demand. The free trade area would represent an improved position for both countries and also, presumably, for the rest of the world. This outcome may next be compared with that which would result if instead countries H and P formed a tariff-averaging customs union (Figure 2.3b). In this case it can be seen that union supply at price CET would be greater than demand, so that the common external tariff would set only the upper limit of the price. The equilibrium price would be PCU where supply equalled demand 33

CUSTOMS UNIONS AND FREE TRADE AREAS

(TM"=US). Once more, trade creation (the production effect a and the consumption effect c) would clearly exceed trade diversion b, although the trade creation effects would be smaller than in the case of the free trade area. The principal difference between the two alternative situations would arise for country P. In the case of the customs union its consumers would suffer a consumption loss, denoted by d. Although its producers would enjoy a net gain, there would be an adverse production effect, denoted by e. In the case of the free trade area there would be no loss from the production and consumption effects, but there would be a gain in government revenue that was larger than the net gain that would accrue to country P with the customs union. In the case of the customs union, moreover, trade with the rest of the world would be eliminated, whereas it would increase in the alternative of the free trade area. Taking these considerations into account, the customs union alternative can be said to be inferior to the free trade area arrangement if judged purely in terms of static efficiency considerations. The difference between the two alternatives results essentially from the indirect trade deflection that occurs in the free trade area case, which rules of origin cannot prevent. In limiting cases this may make a free trade area equivalent in its effects to a customs union that takes the lowest pre-union tariff as the basis of the common external tariff. Evidently, if there are transport costs, which have so far been ruled out, the more geographically dispersed are the members of a free trade area, the smaller is likely to be the indirect trade deflection that results. 2. A second case will now be considered in which, unlike the first, a price differential would arise for the product in the free trade area. In this case country P’s supply is again assumed to be relatively competitive and elastic, but it is now assumed to be incapable of satisfying country H’s demand (Figure 2.4). Before the free trade area is formed, both countries are assumed to have prohibitive tariffs. Country P produces and consumes M, and country H produces and consumes N. If a free trade area were formed (Figure 2.4a), country P’s supply at price TP would be incapable of satisfying the extra demand in country H, and the equilibrium free trade area price in country H would therefore be PFTA (L’N’=OM’). At the same time, the price in country P could not rise above TP, at which level imports from 34

CUSTOMS UNIONS AND FREE TRADE AREAS

the rest of the world would be available, so that two equilibrium prices would exist in the area. In this case country H would experience only a trade creation effect (a+c). Country P would incur no excess costs of consumption or production, but it would enjoy a gain in government revenue, equal to the hatched area, which would represent an increase in its national income. If instead countries H and P formed a customs union (Figure 2.4b), the common external tariff would be effective, with demand and supply approximately in balance at that level, the price in the union being a little higher than in the free trade area case. Country H would experience trade creation. Country P would benefit, on balance, from its ability to export to country H at a higher price, but at the cost of adverse production and consumption effects, denoted by d and e. A comparison of these two situations shows that in this second case, just as in the first, the customs union alternative is inferior to the free trade area arrangement. This conclusion appears to be generally valid for the alternatives of a tariff-averaging customs union and a free trade area, irrespective of the particular market conditions assumed. These comparisons of a customs union with a free trade area refer to trade and tariffs with respect to final products. If, as is possible in a free trade area, tariff disparities also exist on intermediate inputs, these differences may give rise to distortions in the area’s pattern of production at that level. If processing costs were identical, production in a free trade area would tend to concentrate in the country whose input tariff was lowest. In general, however, it is not possible to say whether such tariff disparities would encourage a concentration of production of inputs in more efficient or less efficient member countries. In comparing the merits of a customs union with those of a free trade area, it is clearly necessary to take into account any such production-distorting effects of a non-harmonized tariff at the input stage that may upset the relative advantage of a free trade area in comparison with a customs union.

NOTE 1

These particular definitions of trade creation and trade diversion are those of Johnson (1962). They differ from those initially employed by Viner (1950), Meade (1955a) and Lipsey (1960), who have used the

35

CUSTOMS UNIONS AND FREE TRADE AREAS terms to refer solely to production effects. Johnson’s usage has subsequently been endorsed by Viner (1965) in a letter to Corden. It seems preferable in that it rightly emphasizes that there are only two effects at work: the free trade (trade-creating) effect, and the protection (trade-diverting) effect of customs union.

36

3 BROADENING THE FRAMEWORK

The last chapter analysed the gains and losses that may arise from the formation of customs unions and free trade areas as a result of their impact on resource allocation within the group, and between it and the rest of the world. In a less restrictive framework of analysis than was there assumed, other effects of integration that may be of equal or greater importance also demand attention. In the first place, if the supply curve of the rest of the world cannot be assumed to be perfectly elastic, potential terms-of-trade effects may need to be considered. The assumption of a perfectly elastic supply curve of imports will often be appropriate for a group of small countries whose importance in world trade is minor, but for a bloc such as the EU it is less likely to be so. Second, the enlargement of the market that is brought about by customs union may have a variety of important effects that are disregarded by the classical Vinerian theory. These include: • • • •

•

effects associated with economies of scale; effects on the technical efficiency with which given factors are utilized within firms; the effects of integration on prices and costs through increased competition; effects on the rate of growth of output, through the impact of market enlargement on the rate of growth of productivity, technological advance, etc.; effects of market enlargement on the location of investment and its rate of growth.

These effects have become a prominent feature of later analysis. The effects of custom unions that are associated with economies of scale in an otherwise traditional framework are the main concern of this chapter. With regard to the other effects of 37

BROADENING THE FRAMEWORK

market enlargement that were noted above, its potential effects on the technical, productive efficiency of firms was an issue that attracted attention at an early stage. In seeking to justify integration, politicians in particular have habitually emphasized the beneficial effects on industrial efficiency of what a former British Prime Minister, Harold Macmillan, once described with respect to the EC as the ‘bracing cold shower’, rather than the allocational gains that classical theory points to. In terms of the academic literature, Pelkmans (1984) was one of the first to embody this effect on managerial slackness, interpreted as a downward shift in supply curves, into the analysis of customs unions, following Leibenstein’s (1966) ‘x-efficiency’ approach. He provides some justification for the emphasis given by politicians to the industrial efficiency aspect of integration by demonstrating that the effects that operate through this channel can be a much more important source of gains than those from resource reallocation, which are the only ones considered by orthodox analysis, where given technical efficiency is assumed. The connection between integration and productive efficiency (static or dynamic) is not, however, a simple one—if only because, as well as the cold shower effect which may result from a reduction in protection for certain sectors, there is also a ‘Turkish bath effect’ to consider in respect of other sectors that benefit from trade diversion. In any case, the crux of the cold shower argument rests on the impact of integration on competitive firm behaviour, and this sits uneasily in the framework of orthodox integration theory, with its assumption of perfect competition. The analysis of this effect can be better conducted in a perspective involving imperfect competition. It is in that framework that the variety of issues posed by the effects of increased competition have been pursued with such powerful effect in newer theories of integration. In that connection, Vinerian customs union theory not only disregarded economies of scale but, by its assumption of homogeneous products, ruled out any consideration of the implications of product differentation for integration and its gains. Economies of scale and product differentation imply imperfect competition and jointly give rise to the phenomenon of intra-industry trade, which constitutes a major element of international trade. In an imperfectly competitive framework, new potential sources of gain from integration arise in the shape of a reduction in profit margins and the better exploitation of economies of scale and, perhaps, greater variety. 38

BROADENING THE FRAMEWORK

As to the potential effects of integration on accumulation and innovation, sometimes termed its dynamic effects, these are almost certainly of great significance, since if they arise, even if they are small, they can easily come to dominate the purely allocational effects of integration in a short space of time. Not surprisingly, much effort has been devoted to this aspect in subsequent appraisals. So far, however, economic analysis has not yet fully succeeded in illuminating the impact of integration on growth in a policy context, notwithstanding the work of the late 1980s on technological innovation and the subsequent important attempts of Baldwin (1989, 1990, 1992) to incorporate accumulation and growth effects in a measurable way into analyses of integration that are referred to in later chapters. For analytical convenience, this chapter is limited to a discussion of modifications to the analysis of the previous chapter that are required when terms-of-trade effects and economies of scale are introduced while otherwise remaining within a competitive comparative static framework. With the exception of the economies of scale (and location effects), most of the effects of market enlargement cannot be handled within that framework. Indeed, as will be seen, even the analysis of scale economies in that framework is not free from difficulties. The other effects of market enlargement will therefore be discussed mainly within the framework of Chapters 7 and 8. Spatial polarization which may be encouraged in the presence of external scale economies is discussed in Chapter 12.

THE TERMS-OF-TRADE EFFECTS OF CUSTOMS UNIONS AND FREE TRADE AREAS If the possibility of terms-of-trade effects is allowed for, the analysis of the previous chapter may require modification in certain respects. In particular, where terms-of-trade effects arise with respect to the union’s trade with the rest of the world, the welfare of the world can no longer be assumed to change pari passu with that of the union. One of the most informative analyses of the terms-of-trade effects of preferential arrangements can be found in Mundell (1964). If the formation of the customs union does not affect the demand for imports from the rest of the world, the union’s terms of trade will be unaffected even if the supply from the rest of the world is

39

BROADENING THE FRAMEWORK

less than perfectly elastic. Otherwise, there will be a tendency for the union’s terms of trade with the rest of the world to improve. This effect will operate to reduce the loss that any trade diversion imposes, and it may suffice to eliminate it altogether if the fall in the price of the imported product is sufficient. With respect to free trade areas, the outcome in relation to termsof-trade effects is less clear. Whereas in a customs union trade with third countries is likely to be reduced after integration unless the average level of tariffs is reduced, in a free trade area imports will not fall below the (low-tariff) partner country’s requirements in the pre-free trade area situation, and they may even rise above them on account of the indirect trade deflection that may occur. Consequently, any improvement in the terms of trade will be smaller in a free trade area than in a customs union. Indeed, if indirect trade deflection is very large, it is conceivable that the terms of trade of the free trade area could deteriorate, but that is not a necessary outcome. Where terms-of-trade effects arise, however, a free trade area will inflict less harm on third countries than will a tariff-averaging customs union. It is possible that members of a customs union may be able to exploit their influence on the terms of trade more effectively than if they imposed tariffs separately. This possible effect, fully recognized by Viner (1950), necessarily involves corresponding injury to the rest of the world. Other things being equal, the greater the economic area of the tariff-levying unit, the greater is likely to be the improvement in its terms of trade with the outside world resulting from the tariff. Of course, in considering welfare effects, the volume of trade has to be taken into account. In this connection, Wooton (1986) points out that, in assessing the impact of customs unions on third countries in a broader framework, other factors have to be taken into account. In particular, the favourable effects of a customs union on the real income of the area can be expected to increase the demand for imports from third countries so that, on balance, those countries need not be net losers. The possibility that a customs union may provide a larger termsof-trade gain than would otherwise be possible has been elaborated by Arndt (1968). The difficulty with his analysis, however, is that a conflict of interest is generally implied and, in general, gains arise only if one member country can persuade another to pursue a nonoptimal policy in the interests of the former. A possible exception to this might be the case where a customs union was formed solely to impose a common tariff on a particular product that each country 40

BROADENING THE FRAMEWORK

continued to import. In that case each country might gain from the action of the others. Although any country could gain still more by not joining, the customs union itself might not be formed unless all agreed to join. The issues then are similar to those that arise in the case of an international commodity agreement in which a group of countries generates gains for itself by increasing the product’s price through export restriction. A non-participant would gain still more; but if none participates, no producer will gain. The terms of trade of a customs union with the rest of the world can be influenced not only by the union’s common external tariff but also by the tariffs of other countries. In general, the higher the tariffs of other countries on the export products of the union, the less favourable will be the terms of trade of the union with the rest of the world. Since the level of foreign tariffs can be affected to some extent by tariff negotiations, that aspect must also be taken into account. In this connection it seems likely that, the larger the customs union, the greater its bargaining power is likely to be. This consideration was thought by Viner (1950) to have been historically very important in accounting for the formation of customs unions, and it is one also stressed by Meade (1955a) and, in a different context, by Wonnacott and Wonnacott (1981), as noted in the following chapter. Although conflicts of interest may obviously arise in this case also in relation to the nature of the concessions to be sought in the course of tariff bargaining, the possibilities of detrimental effects from the standpoint of the individual members of the union, and globally, are perhaps smaller. The welfare effects in this context of tariff policy in a customs union if imperfectly competitive industries are assumed would be far more complex and cannot be pursued here. The kinds of consideration that are relevant are illustrated, in a purely twocountry framework, in Chapter 7. The policy relevance of strategic or aggressive trade policy for particular external tariffs on particular products is, however, unlikely to be great, for reasons that are noted in Chapter 8.

ECONOMIES OF SCALE IN CUSTOMS UNION THEORY The theory of customs unions presented in the previous chapter assumes that union supply price is increasing. The introduction of 41

BROADENING THE FRAMEWORK

economies of scale gives rise to a number of problems and calls for some modification of the basic concepts. This section explains, following Corden (1972b), how static economies of scale may be taken into account.1 The economies of scale to be considered are those that are internal to the firm, which result in decreasing unit costs as output expands. For this purpose, consider the case of a single homogeneous product, which is produced in the rest of the world and supplied to countries H and P at constant costs, but which is capable of being produced in countries H and P at declining average costs. Assume that the cost curves of countries H and P for the product are uniformly above import parity price over the relevant ranges and that in the pre-union situation neither country exports to the other. To avoid a number of problems that are not relevant to the point at issue, it is assumed that domestic prices are determined by the cost of imports from the rest of the world plus the tariff. With respect to tariffs, it is assumed that preunion tariff rates are fixed at levels designed to make the tariffinclusive import price just equal to average costs, including normal profit, thus avoiding excess profits to producers.2 In that case, if there is no domestic production, there will be no tariff. Figure 3.1 depicts demand and cost conditions in the domestic markets of the two countries—H, the home country, and P, the partner country. DH is the home country’s demand curve for the product, and ACH is the average cost curve. DP and ACP are the corresponding demand and cost curves in the prospective partner country. DH+P represents the combined customs union demand curve, PW represents the constant price at which the product can be imported from W, the rest of the world. Terms-of-trade effects are thus ruled out. In the pre-union situation, there are three alternative possibilities: production in both countries, production in one country only or production in neither. Each of these cases may be briefly considered. 1. If there is initially production in both countries, the initial equilibrium will be as follows. Prior to the formation of a customs union, the home country H produces and consumes quantity OM, which sells domestically at a price PH. A tariff of PWPH is required to make the industry viable. The more efficient partner country produces and consumes an amount ON at PP with a lower tariff PWPP. If the two countries form a customs union and production is undertaken wholly by the producer whose cost conditions are more 42

Figure 3.1 A customs union with economies of scale

BROADENING THE FRAMEWORK

favourable, country P’s producer captures the whole market. The average costs of country P’s producer when it supplies the whole market will be less than its costs when it supplied only its home market and less than the costs of the former producer in country H when it was supplying its own market. Thus the union domestic price can be less than the domestic price ruling initially in either country. Consequently, the common external tariff will be less than the initial tariffs, and consumers in both countries will gain from the establishment of the union. The combined requirement of the market, XU, would be produced by the partner country at a price PCU, the required union tariff being PWPCU. Consumption in the home country increases to M’ and in the partner country to N’. The effect of the customs union can be considered for each country. Country H’s relatively expensive domestic production is replaced by imports from country P, which are cheaper to produce. Hence there is a movement to a cheaper source of supply, through the opening up of trade between countries H and P—and so an orthodox trade creation gain for country H (using this term in the broader post-Vinerian sense to incorporate consumption effects). This trade creation has two components: the production effect, resulting from the replacement of dearer domestic production by cheaper imports from country P; and the consumption effect, resulting from the increased consumption induced by the lower domestic price. These are measured by areas a and c respectively. Country P obtains its domestic supplies at a lower cost of production. This has been termed the cost reduction effect (Corden, 1972b). Although this is a consequence of the creation of trade with country H, it is not an orthodox trade creation effect, since it results not from a movement to a cheaper source of supply elsewhere, but from the cheapening of an existing domestic source of supply. The cost reduction gain accrues to the consumers of country P. This effect also has a production and consumption component: the production effect is that the original amount of production sold domestically is now obtainable at a lower price; the consumption effect is that, at the lower price, an extra amount is purchased on which consumers’ surplus is obtained. In Figure 3.1, area e represents the production component of the cost reduction effect, and d represents the gain in consumers’ surplus. In addition, country P derives a gain from its sales to country H at prices in excess of world market prices. This is represented by the hatched area f. 44

BROADENING THE FRAMEWORK

2. If in the pre-union situation there is production in one country only, two main possibilities arise. If the established producer is the more efficient of the two, namely country P, the most probable outcome is that it will capture the whole of the union market. In that case, the pre-union tariff of country H will by assumption have been zero. If country P is to capture country H’s market, this must result not from the freeing of trade in the union but from country H imposing a tariff—that is, from the establishment of a common external tariff, which implies an increase in the average level of protection. In that event the price to domestic consumers in country H will rise. The effects for the two countries are then as follows. Country H replaces its imports from the rest of the world by imports from country P. The latter must be dearer than imports from the rest of the world, since otherwise country P would not have needed the formation of the union to compete in country H’s market. Consequently, country H experiences trade diversion, a dearer source of imports having replaced a cheaper source. The resultant losses can be divided into production and consumption components. The new lower amount consumed in country H will be obtained at a higher cost than before. In addition there is a loss of consumers’ surplus on the reduced amount of consumption induced by the higher price to consumers. As in the first case, country P obtains its own product at lower cost, so that there is a cost reduction effect, and it again derives gain from its sales to country H at prices in excess of world market prices. Where there is initial production in only one country, and that country is the higher-cost producer (H), the most probable outcome is that the established producer will be driven out of business, giving rise to a production reversal. In that event a further effect has to be recognized. There will be a trade-creation gain for H because that country obtains its requirements from a cheaper source, whereas country P loses as a result of the replacement of its cheap imports from the rest of the world by dearer domestic production. The costs of country P’s newly established producer when it supplies the whole union market must be greater than the cost of imports from the rest of the world, for otherwise it could have become established before the union was formed. Where, as in such a case, imports from the rest of the world are replaced by domestic production, there is an effect that Viner (1950) has termed a trade suppression effect. It is akin to the trade diversion effect in as much as a dearer 45

BROADENING THE FRAMEWORK

source replaces a cheaper source, but it differs from it in that the dearer source is a newly established domestic producer, and not a source in the partner country. 3. If there is initially production in neither country, the establishment of the union may permit production in one country to begin—say, in country P. That country’s costs must be above the costs of imports from the rest of the world, excluding duty, for otherwise it could have competed in country H’s market prior to the union. In this case the formation of the customs union generates a trade suppression effect for country P and a trade diversion effect for country H. In the presence of economies of scale, it is not possible on the basis of a comparative static analysis to predict which of several possible alternative equilibrium positions will be attained in a customs union. For instance, if in the initial situation the product whose production is subject to scale economies is produced in both countries under identical cost conditions, it is to be expected, if the product is homogeneous, that after the union is formed one firm will take over the whole market. However, the analysis cannot indicate which firm it will be. The outcome will depend on dynamic considerations, including the reaction paths experienced and, in situations in which there is more than one producer in each country, on the nature of oligopolistic competition (Corden, 1972b). In this context, one possibility that has received some attention is that the pattern of specialization that results from trade liberalization when economies of scale are present could be perverse. To illustrate the point at issue, consider the case in which there is initially production in both countries, and suppose that the home country, H, is the higher-cost producer in the sense that its cost curve is uniformly above that of its partner. Suppose also that its demand for the product is so much higher than that of its partner that in the pre-union situation its actual unit costs for the larger amount demanded are below those of its partner for the smaller quantity that it is producing. If then, after the liberalization of trade, consumers in P substitute the lower-cost product of country H for their domestic product, the output of H could expand further and its equilibrium price could fall, whereas the opposite tendency could occur in P. Ultimately an equilibrium might be attained where H produced the combined requirements of the market at higher unit costs than the same output could have been produced in country P. This is 46

BROADENING THE FRAMEWORK

the specific possibility that led Grubel (1967) to argue that trade liberalization would not necessarily produce an optimal pattern of specialization in the group. Whether or not such perverse outcomes are empirically important, their possibility appears to depend on different assumptions from those that underlie the orthodox model. For instance, in the example considered, it may be thought inevitable that the production must in the long run shift to country P if there is certainty, perfect knowledge and an absence of transport costs. However, uncertainty and the cost of transporting the product to country H, where most of it is assumed to be consumed, might operate to prevent that outcome. Transport costs must, of course, be taken into account in determining the optimal pattern of specialization.3 The ‘perverse’ outcome in question might also be produced if, in a situation in which innovation was proceeding, ‘learning by doing’ were an important determinant of costs and were related to output. In such an event this ‘dynamic’ factor in the establishment of comparative advantage might suffice to ensure the survival of the ‘higher-cost’ industry of country H. It should be noted that the issue in question need not prevent trade liberalization from generating gains from specialization, but may result in those gains being smaller than would be ‘technically’ feasible. In a discussion of the same issue in a similar vein, Kojima (1971) has gone so far as to argue that, in the presence of economies of scale, specialization may not occur at all, the price mechanism being ineffective to promote it and equilibrium being stable in the initial situation. Similar issues have been explored in other connections by Meade (1955b). One most important point that is brought out by the inclusion of scale economies in the analysis is that integration through specialization can yield efficiency gains even if there are no differences in comparative costs between the two countries. The point is developed for a three-country two-product model in the following chapter. It can be seen from this analysis that, in the presence of economies of scale, the orthodox concepts of trade creation and trade diversion remain relevant to an evaluation of customs unions, but they require to be supplemented to take account of two other effects, namely trade suppression and cost reduction. However, since those effects are an integral part of the process and are conceptually similar to the ones considered in the orthodox analysis, it is convenient simply to extend the concepts of trade creation and 47

BROADENING THE FRAMEWORK

trade diversion to include these phenomena—just as the original Vinerian concepts that were once limited to the production effects of customs unions have been extended to include consumption effects, as explained in the previous chapter. At the same time, the limitations of comparative static analysis in this context and the possible significance of the dynamic factors mentioned above should be borne in mind. If those factors should be empirically significant, it would follow that trade liberalization per se might not be sufficient to secure the gains from integration. The likelihood of oligopolistic behaviour in the presence of persistent and significant economies of scale would strengthen this reservation. Free trade among members of a customs union may then be a necessary condition for securing such gains, but not a sufficient condition.

CONCLUSION Irrespective of whether customs union theory incorporates economies of scale on the lines just considered or not, orthodox analysis can only consider the impact of integration on inter-industry or inter-sectoral specialization among the member countries. Early empirical studies of the trade effects of economic integration in Western Europe (Verdoorn, 1960; Balassa, 1966; Grubel, 1967) found that much of the intra-bloc trade expansion took the form of a kind of exchange until then hardly acknowledged in the literature, namely of the same or of similar products, implying the growth of specialization within industries. Orthodox customs union theory clearly cannot encompass the phenomenon of intra-industry trade since, on its terms, a country cannot both import and export the same product. The orthodox theory of the time clearly left many unresolved problems in relation to the issue of economies of scale into customs union theory. This was partly because of its inability to confront the issue of imperfect competition. That remained the case until the early 1980s, when newer analyses of international trade with imperfect competition and increasing returns were developed. In addition to incorporating economies of scale into the analysis, the new theories relax the assumption of homogeneous products (so formally recognizing product differentiation and the consumer’s demand for variety), and specifically address the issue of oligopolistic behaviour 48

BROADENING THE FRAMEWORK

(Helpman and Krugman, 1985). These theories were soon applied to the analysis of the effects of customs unions (Smith and Venables, 1988a, b; Helpman and Krugman, 1989). In terms of that newer framework, some of the problems mentioned earlier in this section can be addressed. In particular intra-industry trade is easily accounted for. The existence of similar and therefore competitive, as opposed to complementary, production structures is clearly a necessary condition for intra-industry specialization to arise. If there is also some similarity of demand conditions among the member countries, reflected in overlapping tastes, and if goods are produced with economies of scale, so limiting the amount of product diversity that domestic producers can accommodate profitably, there will be an incentive to horizontal specialization within industries in order to benefit from the economies of large-scale production. Of course, not all intra-industry trade takes the form of final products. A high proportion takes the form of intermediate products, which is a category of trade associated with vertical specialization. Like intra-industry trade, this kind of trade finds no place in orthodox customs union theory either, but it has a place in newer trade theories. Some aspects of the application of these newer theories of trade to the economics of integration are discussed in Chapters 7 and 12. These theories throw much light on crucial issues, but one of the characteristics of the approach is that typical models exclude the operation of comparative advantage by assuming that the gains from trade and integration stem purely from economies of scale and product differentation (symmetrical situations are assumed with respect to costs of production) in order to focus on the gains from competition. To that extent they do not address the issues of trade diversion and creation. This approach, though not inherent,4 is followed partly in the interests of making the analysis of highly complex situations more tractable, but it also appears to reflect empirical judgements concerning the respective roles of comparative advantage and imperfect competition in trade. At present, it is impossible to arbitrate properly between the two approaches, but issues of trade creation and trade diversion must clearly remain central to that wide range of situations where comparative advantage resulting from differences in relative factor endowments and technologies is a material factor in determining trade. The problem of the arbitrary nature of the pattern of specialization with scale economies is still a problem even when product differentiation and 49

BROADENING THE FRAMEWORK

oligopoly are incorporated into the analysis. That difficulty could be resolved only by including dynamic factors in the analysis.

NOTES 1

2

3 4

In addition to static economies of scale, there may also be ‘dynamic’ economies from experience or learning which are bound up with the scale of production and are reflected in a downward shift in cost curves (CEC, 1988a). If in each country there are potential competitors able to enter the market, and established firms have no cost advantage over those potential competitors, as the theory of contestable markets supposes, average cost pricing would be the norm and a higher tariff would have no effect on profits. See Chapter 12 for an example of the interaction of transport costs and production efficiency. Harris and Cox (1984) in their analysis of US—Canadian trade liberalization in the newer trade framework do allow comparative advantage to play a role in their model.

50

4 CUSTOMS UNIONS VERSUS UNILATERAL TARIFF REDUCTION

In Chapters 2 and 3 the basic economic theory of customs unions and free trade areas has been outlined. This chapter is mainly concerned with the implications of a problem that has been implicit in that presentation but that has not been addressed up to this point. On the assumptions of the orthodox model, the basic theory as developed by Viner, Meade and Lipsey fails to provide an economic rationale for the formation of customs unions. This is because, apart from the terms-of-trade argument, the grounds on which it can be shown that a customs union may be superior to a non-discriminatory tariff are precisely those on which customs unions and free trade areas can themselves be shown to be necessarily inferior to unilateral tariff reductions and free trade. It may be questioned whether this is a significant issue. The basic model is highly restrictive and it excludes a number of economic factors that may be expected to motivate the formation of an economic grouping. In any case, even if it should turn out— as Viner himself suggested—that customs unions have to be justified on non-economic grounds, their economic effects would still require analysis. The issue of the inferiority of customs unions to a policy of unilateral tariff reduction was initially explored in the mid-1960s (Johnson, 1965; Cooper and Massell, 1965), and has recurrently aroused the interest of theorists since that time (Wonnacott and Wonnacott, 1981, 1984, 1992; Berglas, 1979, 1983; de Melo, Panagariya and Rodrik, 1993). It has also coloured the policy stance of the World Bank, which is generally hostile to regional blocs. The main contribution of the debate has been to suggest fruitful new ways of viewing certain customs union issues, although some of its implications have greater significance for economic communities than for customs unions proper. 51

CUs VERSUS UNILATERAL TARIFF REDUCTION

THE SUPERIORITY OF UNILATERAL TARIFF REDUCTION? The basic issue may be illustrated by reference to Figure 2.1. Suppose that the home country (H) has an initial tariff of PWT’ H. If it enters a customs union with a partner country (P) and they adopt a common external tariff’ CET’, then, as has already been shown, trade creation that is equal to (a+c) will result. However, country H also has the option of unilaterally reducing its tariff to P WCET’ on a non-discriminatory basis. If it were to do so, it would enjoy exactly the same level of domestic production and consumption as it would in the customs union. The difference is that, with the non-discriminatory tariff reduction, imports would come from the rest of the world; and this would generate a net gain to country H that was equivalent to the difference in total outlay on imports from the two sources. This net gain would also be equal to the customs revenue that would be raised if country H were to adopt the lower tariff on a non-discriminatory basis. In the case of free trade areas it may likewise be shown that, although a free trade area may be superior to a customs union, a free trade area would itself be inferior to a non-discriminatory tariff reduction. With respect to the situation depicted in Figure 2.3a, for instance, a non-discriminatory tariff reduction to PWTP by country H would result in the same amount of consumption and domestic production as in the free trade area case, but again in such a situation import requirements would be met by the rest of the world. The additional gain to country H by comparison with the free trade area option would equal the revenue raised from import duties, which would equal the import duty earned in the free trade area situation by country P. In this particular case the adoption of a free trade area as opposed to a nondiscriminatory tariff reduction would merely result in a redistribution of import revenue (equivalent to a national income loss) from country H to country P. This conclusion is unaffected by the introduction of economies of scale into the analysis. It is also unaffected by the contention of Wonnacott and Wonnacott that it ignores the the possibility of gain to the home country from reductions in the partners’ tariffs on their exportables. This point can be illustrated for the case with economies of scale by Figure 3.1, from which it can be seen that country H would be made just as well off by a 52

CUs VERSUS UNILATERAL TARIFF REDUCTION

unilateral tariff reduction to PCU as by entering into a customs union with P at the indicated common external tariff. Such would not be the case for the partner country P. However, in its case, the additional benefit that it would derive from the formation of the customs union as compared with the alternative of unilateral tariff reduction is obtained entirely at the expense of the home country, which for its part would be still better off if it were to eliminate its tariff altogether. In a more comprehensive model (but without scale economies) Berglas (1979) effectively concluded that any gain to one partner is more than offset by even greater loss to the other. The conclusion thus remains that a customs union does not allow for any mutually beneficial gains not open to each of the countries separately, so that it remains inferior to unilateral tariff reduction. The model presented in this chapter does not allow for transport costs or for tariffs in the rest of the world. In their most recent contribution to the debate (1992), the Wonnacotts seek to rescue their proposition by introducing those two elements, but it is not clear that they have succeeded in producing more than the possibility of a curious special case unless strategic interaction with the rest of the world be admitted. The conclusion that a customs union is inferior to unilateral tariff reduction and to free trade is not of course inconsistent with the Kemp—Wan proposition referred to in Chapter 2, namely that it is always possible to raise the welfare of every member of a customs union while leaving the welfare of the rest of the world unchanged if the common external tariff is set appropriately and lump-sum compensatory transfers amongst the members of the union are possible. Within the framework of the orthodox analysis, therefore, even when it is extended to incorporate economies of scale, it must be concluded that none of the considerations so far discussed provides an economic rationale for the formation either of customs unions or of free trade areas if the alternative of unilateral tariff reduction is available. If a customs union is beneficial, it is because net trade creation represents a move towards free trade. Free trade would represent a still better position. This is the orthodox point of view, which inevitably follows from the orthodox position that any initial tariff (or any other form of intervention in trade) is arbitrary and non-optimal. The trouble with this point of view, as Johnson has written, is that it: 53

CUs VERSUS UNILATERAL TARIFF REDUCTION

puts the economist in opposition to dominant strands in the actual formulation of international economic policy, which have to be treated by definition as ‘irrational’ or ‘noneconomic’… At the same time, the economist is left without a theory capable of explaining a variety of important and observable phenomena, such as the nature of tariff bargaining, the commercial policies adopted by various countries, the conditions under which countries are willing to embark on customs unions, and the arguments and considerations that have weight in persuading countries to change their commercial policies. (1965, p. 257) One instructive approach to the resolution of the dilemma has been to seek an economic rationale for the formation of customs unions and regional economic groupings by the inclusion of ‘public goods’ in the welfare function. The resulting theory then focuses attention on the implications of this extension—and in particular on the question of the relative efficiency, from the standpoint of two or more countries acting together, of customs unions as opposed to optimal non-preferential tariff systems as a means of providing goods in which an element of ‘publicness’ is involved, in the sense that the benefits of consuming them do not accrue solely to the purchaser. If, in this context, it could be shown that a customs union might make both countries better off relative to individually optimal policies of non-preferential tariff protection, a possible rationale for such unions would have been provided. Approaches along these lines necessarily abandon the single-country/singleproduct analysis of customs unions, which is incapable of illuminating the issues. These issues were first investigated systematically by Johnson (1965) and by Cooper and Massell (1965), who developed an analysis that is relevant to customs unions in which a public good is included in the social welfare function and therefore in the evaluation criteria. Each of these analyses can also be interpreted to apply to a situation where welfare is defined solely in terms of the consumption of private goods and services but where externalities of various kinds give rise to domestic distortions, which have so far been ruled out of the discussion. Johnson himself was primarily concerned with developing a theory to explain why governments behave as they do in relation to a variety of commercial policies having a protective character. Cooper and Massell, on the 54

CUs VERSUS UNILATERAL TARIFF REDUCTION

other hand, were directly concerned with the customs union issue and more specifically with explaining how membership of a customs union may enable countries to achieve more economically the ends that can be served by protection. Each of those analyses thus rests on a prior rational argument for protection within which the case for protecting industry may be regarded as a special case.

THE ANALYSIS OF CUSTOMS UNIONS WITH PUBLIC GOODS Although the basic point is a simple one, an illustration with a brief example provides a number of useful insights into the issues. For this purpose, in contrast to the analysis of previous chapters, it will now be supposed that welfare is generated both by the private consumption of goods and services and by the collective consumption of a variety of public goods that must be provided through government agency at the cost of sacrifices in private consumption. One such good, for instance, might be a level of certain kinds of industrial or agricultural production that was in excess of what would be commercially viable in the absence of protection. The following analysis allows for such public goods, the benefits of which are assumed to be country-specific. As in much of the analysis of Chapters 2 and 3, two prospective partners are considered, the home country, H, and the partner country, P. It is assumed that each is producing at least two products in the industrial sector that it is desired to protect. It is also assumed that, in the public aspect, the two countries are indifferent between the two industries—or any others—and that industrial diversification is not an object of policy or a source of benefit. The costs and benefits of any customs union are further assumed to arise solely in the course of trade, so that any possibility of redistributing benefits through fiscal transfers is excluded. All the products of the industrial sector in question are assumed to require protection and therefore cannot be exported to the rest of the world. Thus the level of domestic demand governs the level of sales in each industry. The gains from customs union can be illustrated in Figure 4.1, which is merely a reinterpretation of Figure 3.1 to apply to two countries and two products. The comments made in Chapter 3 about pricing apply equally here. Curves AC H1 and ACH2 represent the 55

Figure 4.1 A customs union with economies of scale and public goods: the gains from specialization

CUs VERSUS UNILATERAL TARIFF REDUCTION

average cost curves for the two products produced under conditions of economies of scale in the home country, for which the respective demand curves are DH1 and DH2. For simplicity of representation it is supposed that country P’s cost curve for industry 2 is represented by the cost curve of country H for industry 1 and that country P’s cost curve for industry 1 is represented by country H’s cost curve for industry 2. It is assumed that the private domestic demand for each product is the same and identical in each country so that DH1=DP2=DH2=DP1. In pursuit of its industrialization or public goods objective, it is assumed that prior to customs union country H imposes made-tomeasure tariffs on the products of the two industries. This would generate an amount of industrial production equal to OM of product 1 and ON of product 2, the respective tariffs being PWPH and PWPP. The forgone private income compared with importation, which is equivalent to the excess domestic cost of production, is equal to a+f on product 1 and d+h on product 2. Similarly in the pre-customs union situation country P imposes made-to-measure tariffs and produces OM of product 2 and ON of product 1, the respective tariffs being PWPH and PWPP, the forgone private income being the same as for H. If the two countries form a customs union and impose a common external tariff of PWPCU on each product and specialize respectively in product 2 in country H and product 1 in country P, exporting the excess of their domestic production over domestic consumption to the other country, each will be able to satisfy its preference for industrial products or public goods at a smaller cost in terms of forgone national income. In that situation, the right-hand part of Figure 4.1 then depicts the production costs of each product when an amount OXU of each is produced to satisfy the combined requirements of the market. The gains from union for each country will be identical and can easily be arrived at by a comparison of the relevant areas, as was done in the previous chapter. If the products protected are those that provide the specified amounts of industrial product at the least cost in terms of the combined national income forgone, as we suppose to be the case in this example, the tariff may be termed an ‘efficient’ common external tariff. This convenient outcome, in which an efficient common external tariff results in each country at least achieving its initial industrial production level, is not the only possible one. The industrial cost structure of the two countries may be such that the 57

CUs VERSUS UNILATERAL TARIFF REDUCTION

two least-cost industries are both located in country H. In that event, if a customs union were formed that protected the two industries, country H would produce the whole output of the two goods whose cost and demand conditions are depicted and P would produce none. Such an outcome would not be acceptable to country P, since in a customs union it must be assumed to produce at least as much industrial output as in the pre-customs union situation if its preference for industry is to be satisfied. A mutually advantageous customs union might still be feasible if a tariff were chosen that would enable each country to produce at least its pre-customs union or other specified level of industrial production in the cheapest possible way by, for instance, protecting some third industry in the partner country that could produce more economically than the same industry in the home country, even though its costs might be higher than either of the two indicated industries in the home country. Such a tariff may be termed a ‘quasi-efficient’ common external tariff. If, however, all possible industrial products in the partner country are capable of being produced in the home country at lower cost, and if exchange rates are fixed, a simple customs union would not enable the partner country to achieve its public goods or industrialization objective, although opportunities for potential gain from integration could still be present. To achieve those gains, protection within the customs union might be necessary (Cooper and Massell, 1965). This consideration suggests a rationale for the inter-union protective devices found in certain customs unions among developing countries where there is a preference for industry in participating countries, such as the Southern African Customs Union. The analysis of this and previous chapters has assumed that the national income effects of customs union on its participants take effect solely through trade. If fiscal transfer payments can be made from one country to another in order to redistribute benefits, as happens in Southern Africa, the analysis becomes much more complex. It is then conceivable that an efficient tariff could be chosen, thus maximizing the income gains from the union, even if all industry were located in one country, so long as the other country was compensated for its loss of industry by a sufficiently large income transfer. However, the amount of income compensation required by any country would depend on its particular preference function, and would not necessarily be compatible with the income gains enjoyed by the other country. It is possible, that is to say, that 58

CUs VERSUS UNILATERAL TARIFF REDUCTION

the country gaining industry (and income) would not gain enough to enable it to pay sufficient compensation to its partner to make a fully efficient tariff feasible. In that case, if a customs union is to operate, at best only a quasi-efficient tariff can be adopted. In any event, if one country is to get no industry, it can be concluded that it would presumably need more income compensation than was represented by the excess cost to it of its partner country’s products because it could effectively attain that situation by a unilateral move to free trade. On the approach just outlined the gains to a country from a customs union depend on what happens to both income and industrial output, which in this case is the public good. This analysis does not make the classical distinction between trade-creating and trade-diverting customs unions irrelevant but suggests that, for a gain to be enjoyed by each member, overall trade creation, though necessary, is not a sufficient condition for gain.

PUBLIC GOODS AND THE RATIONALE FOR CUSTOMS UNIONS In the context of a preference for public goods, the analysis of the previous section demonstrates that the formation of customs unions may be a more efficient means of satisfying that preference than individually optimal non-discriminatory tariff protection. In this way a rationale for the formation of customs unions appears to have been provided. The benefits would, of course, have to be demonstrated in each case, for even in this context a particular customs union may not be beneficial, any more than a customs union in a purely classical framework would invariably be trade-creating. However, with that qualification, does the public goods approach, or the recognition of a national economic case for protection, satisfactorily dispose of the difficulty raised at the outset? Simply to demonstrate that a customs union may be superior to unilateral tariff policies from an efficiency viewpoint when public goods or other rational considerations motivate protection still fails to provide an economic rationale for the formation of customs unions unless their formation can be shown to represent the best means of encouraging or providing the public goods in question. If governments have the option of providing direct production

59

CUs VERSUS UNILATERAL TARIFF REDUCTION

subsidies, orthodox theory would suggest that these, rather than customs unions, would be the most efficient protective mechanism to use—at least in the specific public goods case considered—on the grounds that they would avoid the consumption costs involved in the imposition of tariffs. This argument is not rebutted by the consideration that subsidies have to be financed by taxation, which will itself impose distortions, since it can be shown that, if optimal tax-subsidy policies are pursued, any such distortions are likely to be smaller than those generated by a tariff (Meade, 1955b; Corden, 1974). On this view, an economic rationale for customs unions can be established on public goods grounds only if political or other constraints rule out the use of direct production subsidies financed by lump-sum transfers. A variety of institutional factors, and in particular a country’s international obligations, do in fact often rule out resort to a range of possibly superior protection policies such as the institution of discriminatory tariff preferences amongst selected partners. The World Trade Organization set up in 1995 has taken over the established GATT principle that customs unions and free trade areas are the only permissible forms of preferential tariff reduction, except in the case of developing countries. Likewise, the payment of subsidies to make exports feasible to selected partners would be excluded under the rules governing world trade. International constraints of these kinds may effectively mean that customs unions may represent the only practicable means of achieving the gains in question. Nevertheless, Krauss has argued (1972) that the public goods argument fails to provide a general economic argument for customs unions and that it leaves substantially unimpaired the Vinerian view that customs unions are essentially non-economic institutions. This conclusion rests on his defining the institutional constraints on protection policies that would be superior to tariffs as political Stated in these terms, the issue is partly semantic. A more interesting question is whether customs unions and other forms of regional integration can sensibly be regarded as non-economic institutions that embody second-best policies in the context of a less restrictive interpretation of the context of customs unions. If the concept of public goods is broadly interpreted and extended to cover those where the effects spill over to other jurisdictions, a public goods rationale for customs unions and other forms of ‘discriminatory’ regional economic integration may be found that does not have to rest merely on ‘institutional’ or political 60

CUs VERSUS UNILATERAL TARIFF REDUCTION

constraints. In relation to the provision of a broad spectrum of public activities that are undertaken in complex modern societies, such as those discussed in the following chapter, the presence of cross-border spillovers may give rise to structural and policy interdependence that can furnish compelling economic arguments for increasing the size of jurisdictions beyond the boundaries of single countries. In principle, these considerations may sometimes suggest that a global extension of jurisdiction would be optimal. In practice, however, integration that finds its justification on such grounds must of necessity usually be limited to a regional scale and so be ipso facto discriminatory. The costs of organization, management, information and negotiation, coupled with the influence of uncertainty—all of which are disregarded by orthodox customs union theory—will in reality limit the scale of optimal jurisdictions to a less than global level. For instance, a significant motive for the formation of customs unions that orthodox theory neglects is to reduce the transaction costs associated with the imposition of trade and other indirect taxes and their attribution. These constitute an important barrier to trade and specialization but their elimination or reduction is an issue that can hardly be addressed except at a regional level. It is implausible for instance, that the wide range of barriers to trade and production that are addressed in the EC’s Single Market initiative could possibly have been overcome except in a regional context. The fate of discussions on most ‘new issues’ of trade policy under the WTO well illustrates the force of this caveat. Any policy-relevant model must also take account of the fact that states differ widely in their preference for public goods in general and for specific public goods because of political, social and perceptual differences and disparities in levels of economic development. This consideration also points to limiting the membership of international groupings to relatively homogeneous states in the interests of an optimal provision of public goods and in particular those that call for the harmonization of a range of economic policies to facilitate the removal of barriers to trade and production. Most instances of international economic integration are in fact found among groups of countries whose preferences for public goods are relatively homogeneous by global standards. There can still be little argument with the proposition that there is no general case, even when public goods grounds are admitted, that would justify customs unions or wider forms of international integration. At the same time, to dismiss the many important factors 61

CUs VERSUS UNILATERAL TARIFF REDUCTION

that may justify regional economic integration as political or noneconomic, or a policy of regional integration itself as necessarily ‘second best’, is an unhelpful categorization that harks back to the conditions of an extremely restrictive analytical model which excludes transaction costs, imperfect information and cross-border spillovers. In the end, the important points not to lose sight of are: that customs unions and other forms of regional integration are potentially capable of generating a variety of welfare gains; that to capture many of those gains through either unilateral action or global co-operation is not feasible; and that however useful the concept of ‘first best’ may be as an analytical touchstone, first-best policies cannot sensibly be defined without reference to transaction costs and imperfect information. The way in which these factors come into play in forms of integration that go beyond customs unions is the subject matter of several of the following four chapters.

62

5 THE RATIONALE FOR THE INTEGRATION OF OTHER ECONOMIC POLICIES

The last chapter sought to explain how policy objectives other than the maximization of private income can be formally introduced into customs union theory in the simplest manner. In reality not only are the objectives of public policy extremely diverse, but so are the policy instruments that may be used to attain them. Even if the objective is purely to secure resource allocation gains from market integration, tariff policy is not the only instrument available, nor is it necessarily, on its own, a sufficient means. The mere elimination of tariffs in a customs union will not necessarily bring about improvement in the allocation of resources if other policy-induced obstacles to a unified market then become the binding constraints. The gains to be secured from the reduction or elimination of these non-tariff barriers may be as great as, if not greater than, those to be derived from the elimination of such non-tariffs themselves. If states decide to seek efficiency gains through market integration, they may thus be induced to seek to expand those gains by the adoption of common measures that go beyond the institution of a customs union or a common market. Typically, some of the most important non-tariff obstacles to the improved allocation of resources in a regional grouping take the form of technical barriers in the shape of varying safety regulations and industrial standards for machinery and equipment, and varying health regulations in respect of foodstuffs, agricultural products and pharmaceuticals. The fundamental objection to unharmonized industrial standards is that they may make it necessary to produce different products for each member country, thus restricting intra-group trade, making it more difficult to achieve scale economies and, in general, hindering the reduction of costs

63

INTEGRATION OF OTHER ECONOMIC POLICIES

and prices that it is a major purpose of customs union to bring about. Likewise, if health regulations are not harmonized, trade in agricultural products and foodstuffs may be unjustifiably restricted. In the EC itself, despite thirty years of persistent attempts to bring about formal harmonization through common regulations, and the effects of the principle of mutual recognition of each others’ regulations that resulted from rulings of the European Court of Justice, varying industrial standards still constituted as late as the mid 1980s one of the biggest remaining obstacles to the creation of a single internal market, and they were without question often used for covert protection to fragment the market. In order overcome this and other obstacles to a fully integrated market the Community in 1985 adopted a far-reaching programme (CEC, 1985) which aimed at the removal of these kinds of technical obstacles as well as others of a physical and fiscal nature, in order to achieve a single internal market by 1992. It has made great strides towards enacting the measures in question, although even in 1996 much remained to be done (CEC, 1996). The target date for a fully effective single market is now January 1999. The elimination of tariffs and non-tariff barriers on intra-group trade is an instrument for procuring certain efficiency gains in a regional market. The institution of free movement for factors of production as well as products, involving the creation of a common market, is an obvious next step to take on allocational grounds after a customs union has been established. But neither regional trade liberalization in itself or free movement for factors can contribute much to the attainment of wider objectives of union policy. Indeed, the free movement of labour and capital may run counter to some of its other goals. The possibility must then be considered that the economic interdependence of markets that is fostered by customs unions and, still more, by common markets, may enable trade-based regional groups to procure further economic benefits by pursuing other economic goals in common through integrated policies in other fields. Economic analysis identifies three main areas where further policy integration may offer the prospect of gain, namely: • • •

policies to overcome allocational cross-border spillovers; policies to overcome macroeconomic and stabilization spillovers; policies to promote cohesion and convergence 64

INTEGRATION OF OTHER ECONOMIC POLICIES

The regulation in common of economic policies amongst the members of an economic grouping for these and other purposes may in principle be pursued by various means and to various degrees. In its broadest sense, the term ‘integration’ may be used to refer to any of these approaches. It is, however, often useful to draw a distinction between integration, in a narrower sense, harmonization and co-ordination. Integration in its most restrictive sense refers to the assignment of particular economic functions and instruments to the union or community and their exercise at that level rather than at the level of the member states. Harmonization is a second level of integration that refers to agreement on the manner in which each member state will exercise or utilize a particular instrument over which it retains control. Harmonization involves the adoption of legislation by the institutions of the community that is designed to bring about changes in the internal legal enactments of member states. Co-ordination is a third level of integration which refers to voluntary and largely unenforceable alignments of national policies and measures in particular fields. The context will normally make clear in which of these senses the term ‘integration’ is being used.

POLICIES TO OVERCOME ALLOCATIONAL CROSS-BORDER SPILLOVERS The elimination of tariff and non-tariff barriers among the members of a grouping can contribute to efficiency gains from the reallocation of resources with respect to private goods and services. Member states are also concerned with the provision of a range of public goods through the budget and by regulatory action. The effects of such measures may extend across the borders of member states. If such public sector activities or those of other economic agents in one member state create significant cross-border spillovers that are positive or negative, or if there are increasing returns to scale in the provision of public services, then, if the extent of provision is determined by the separate member states without taking account of these spillovers, the outcome will not be optimal. The bloc as a whole will fail to provide the appropriate level (or kind) of activity unless the externalities can be internalized, by suitable integration of policies. Resources will not be correctly allocated and there will be efficiency losses as a consequence of market failure.

65

INTEGRATION OF OTHER ECONOMIC POLICIES

There are many examples of policy areas where significant crossborder spillovers can be expected to be present. Measures to deal with pollution and environmental damage where substantial jurisdictional spillovers typically arise are a case in point. Certain aspects of transport provision and support for research and development are further instances. Tax competition amongst jurisdictions may also generate substantial spillovers. Unco-operative tax setting does not necessarily produce allocational distortions but, by driving down tax rates, it is apt to result in a less than optimal level of provision of tax-financed public services. Spillovers of any of these various kinds may justify policy integration in a narrow sense, or point to co-ordination if the costs of co-ordination itself are not too high and it can be made effective.

POLICIES TO OVERCOME STABILIZATION SPILLOVERS The issue of further policy integration is also posed in relation to macroeconomic policies. One important argument for moving towards policy integration in these areas is based on the desirability of facilitating the primary resource allocation objectives of economic integration. Neither intra-group trade nor intra-group investment will take place in an optimal way, it may be argued, if intra-union exchange rates are volatile or misaligned. A high degree of certainty concerning intra-union exchange rates could not be delivered without the co-ordination of macroeconomic policies. Just as, if not more, important are arguments for the coordination of monetary and fiscal policies that derive from macroeconomic spillovers. As the national economies of member states become progressively interdependent as a result of trade and factor movements, the justification for co-ordinating their monetary and fiscal policies on macroeconomic grounds becomes increasingly strong to the extent that such policy measures undertaken by one country to influence its macroeconomic variables spill over and significantly affect those of other member countries. In these circumstances indeed, for small countries, unilateral measures of financial policy may be largely ineffective. Financial co-ordination may be particularly important if there are constraints, such as may be found in a customs union or common market, on the other policies that member states may use

66

INTEGRATION OF OTHER ECONOMIC POLICIES

domestically to promote external balance. If, for instance, there should be an agreement to limit exchange rate fluctuations among the members in the interests of promoting intra-community trade and investment—such as is the case with the ERM—the remaining financial instruments open to any single country may not suffice to enable it to maintain internal and external equilibrium and to faciliate necessary adjustments. Co-ordinated action by both deficit and surplus countries may then be a requisite for bringing about and maintaining a satisfactory equilibrium. In certain circumstances the weight of argument in favour of policy co-ordination in the area of monetary policy may point to full monetary integration involving the establishment of a common currency. In any event, co-ordination of financial policies is likely to be desirable in any union in which a high proportion of trade is intra-community trade (the proportion is more than 50 per cent in the case of the EU) and where, for that reason, both structural and policy interdependence can be expected to be considerable.

POLICIES FOR COHESION AND CONVERGENCE The question of policy integration also arises in connection with distributive issues. Integration measures that promote the attainment of allocational efficiency will not necessarily benefit all members or produce distributionally acceptable results across the market. The operation of market forces alone may well result in a widening of economic disparities among member countries, at least for long periods of time, if not indefinitely. Integration may even make some member states absolutely worse off, according to their perceptions and criteria, than they would have been if they were outside the community, even though the community as a whole may enjoy gains that outweigh such losses. If the community is a significant determinant of the economic performance of the member states, such outcomes may seriously damage the prospects for cohesion of any economic community that lacks a supranational political authority. To avert such damage, one or more of several approaches might be pursued by common action. One approach would involve the differential harmonization of certain national policy instruments. Harmonization of instruments on distributional grounds would not require the effective uniformity that would be appropriate if considerations of competition and

67

INTEGRATION OF OTHER ECONOMIC POLICIES

resource allocation were alone in question, but rather the introduction of measures of agreed discrimination. For instance, measures such as regionally differentiated industrial development incentives might be used for the purpose or, indeed, the structure of the tariff itself might be employed, as was envisaged in the previous chapter. Any such measures should be designed to deal with distributional constraints and objectives at the least possible cost in terms of resource allocation and other considerations. A second approach to dealing with distributive issues would be to make provision for fiscal transfers among the member countries, either directly or through the medium of the community budget. This approach has the merit that it need not, in principle, interfere with efficient resource allocation. However, resort to fiscal transfers may not represent an acceptable solution unless the transfers contribute to the removal or amelioration of the underlying causes of difficulty, since otherwise permanent transfers would be needed. The third approach would be to develop policies that could contribute to a modification of the economic structure of any member state that was adversely affected in absolute terms, or that did not enjoy an equitable share in the gains from integration to enable its economy to perform more effectively. That might require the provision of conditional, performance-related and ear-marked transfers from the community to member states, under the aegis of a regional policy institution.

THE DYNAMICS OF POLICY INTEGRATION The view is often expressed that, sooner or later, the formation of a customs union will almost inevitably entail the extension of policy integration (or at any rate policy co-ordination or harmonization) to fields such as those outlined above. Orthodox static equilibrium analysis affords no support for any such conclusion. It may be the case that the static gains from improved resource allocation in an integrated area can be maximized only if, in addition to product market integration, there is also factor market integration and harmonization or co-ordination of certain other policies that directly affect competition and economic structure. Yet national welfare would not necessarily be enhanced by the adoption of such further measures of integration because the attainment of other ‘public goods’ objectives, such as domestic or regional balance, or of

68

INTEGRATION OF OTHER ECONOMIC POLICIES

domestic income distribution objectives may be impaired by doing so. To that extent further gains from efficiency in the use of resources have to be balanced against their cost in terms of the attainment of other important policy objectives. On such grounds it may be perfectly rational for the member countries of a bloc to wish to confine the scope of their integration mainly to tariff policy and product markets, as would be the case in a simple customs union or free trade area. It is, of course, evident—as has already been stressed—that, even in customs unions and free trade areas, some degree of harmonization of other trade and commercial policies and instruments will be indispensable if the basic objectives of the arrangements are not to be frustrated. If quantitative trade restrictions such as quotas were nationally determined, large differences in their scope or structure could undermine the very basis of a customs union by enabling any member unilaterally to alter the degree of effective protection enjoyed by an industry or activity, notwithstanding the existence of the common external tariff. Some minimum degree of harmonization of such measures may thus be indispensable, since they are so closely connected with the purposes of the customs union and are often, indeed, of equivalent effect to tariffs. However, the extent to which it would be desirable to go beyond this limited stage, and to harmonize or integrate other economic policies, should depend on the member states’ perception of their various policy objectives, on the weight they attach to each, and on the terms on which, in the union, one objective can be traded off against another by political bargaining. In purely theoretical terms, there is no reason why a game-theoretic static equilibrium should not be derived. Dynamic factors, on the other hand, may lend support to the view that policy integration that goes well beyond the elimination of tariff and non-tariff barriers is, in one sense, inevitable. Except in very favourable circumstances, customs unions and, still more, common markets that fail to concert certain other policies may not constitute a stable long-term equilibrium. A variety of dynamic and structural forces that such unions embody and strengthen may work powerfully either towards a more intimate form of integration, which may approach economic union in the limiting case, or towards separation. One such force is represented by the spatial effects of integration, which may be a crucial determinant of the distribution of the costs and benefits of integration. It is not clear how these spatial effects could be addressed adequately within the confines 69

INTEGRATION OF OTHER ECONOMIC POLICIES

of a simple customs union or a common market. Casual empiricism would indeed suggest that the history of almost every recorded customs union provides support for the view that integration cannot stop there and that the adoption of further measures of policy integration can be said to be inevitable—if a customs union is to survive. It is, of course, often said of the EU itself that, like a bicycle, it must keep moving if it is not to fall over. The systematic study of the dynamics of policy integration has emerged only recently, and then mainly in the application of public choice analysis to the purely institutional dynamics of integration. So far, general studies of the issue in an integration context are few, but there are several studies in these terms of specific policy areas in the EU context, including the Single Market programme, the European Monetary System and fiscal harmonization (Vaubel, 1994), which analyse the incentives that work in favour of policy centralization. In an analysis of the economic implications of extending economic integration to policy areas such as those outlined in this chapter, it is convenient to consider the issues in the first instance from the standpoint of a single primary policy objective or criterion; but a broader basis of appraisal involving multiple goals is ultimately desirable, incorporating considerations of the kinds referred to in this chapter. Thus, for instance, although tax harmonization may be partially evaluated in terms of its resource allocation effects, its impact on the attainment of the social welfare and stabilization objectives of the separate member countries and of the union itself must also be considered. Likewise, monetary integration may be evaluated from the standpoint of its almost certainly favourable microeconomic effects on allocational efficiency and transaction costs, but a balanced view of its merits can clearly be formed only if account is also taken of its possible impact on the macroeconomic objectives of member countries. Apart from Chapters 7 and 8, which are concerned with new perspectives on integration, the remaining chapters of this book discuss central issues that arise in a progression from a customs union to more developed forms of integration involving the integration of other key economic policies. Chapter 6 discusses the effects of instituting a common market for factors of production. Chapter 9 considers the role of the budget in economic groupings and the fiscal efficiency considerations that bear on the question of the optimal assignment of public functions from the member states 70

INTEGRATION OF OTHER ECONOMIC POLICIES

to the bloc or union. Chapter 10 discusses tax harmonization, primarily from the standpoint of resource allocation considerations. Chapter 11 considers monetary integration, with particular reference to its implications for macroeconomic policy objectives. Finally, regional policy, which has important structural and distributional objectives, forms the subject matter of Chapter 12.

71

6 THE THEORY OF COMMON MARKETS

The basic microeconomic theory of international economic integration consists of the static theory of customs unions and free trade areas, a central assumption of which is that factors of production are immobile both amongst the member countries and vis-à-vis the rest of the world. A common market, by contrast, involves not only the integration of product markets through the trade liberalization that results from customs union, but also the integration of factor markets through the elimination of obstacles to the free movement of factors within the bloc. The concept of a common market was probably introduced by the Spaak report of 1956, but in any case the term was in widespread use from the mid 1950s. The Treaty of Rome itself prescribed the establishment of a common market within twelve years, which would entail ‘the abolition, as between member states, of obstacles to freedom of movement for persons, services and capital’. In a narrow sense the basic concern of a theory of common markets is with the additional benefits that can be derived by going beyond a simple customs union to the establishment of a common market. At the very least such a market would require the removal of legislative restrictions upon the free movement of factors between the member states, but the effective integration of factor markets would in practice also call for the adoption of positive harmonization measures to ensure effective non-discrimination at the regional level in the markets for labour, capital and enterprise. At a purely static level, the benefits, if any, to be derived from the superimposition of a common market upon a customs union are allocational gains. If, within a customs union, differences subsist in the marginal productivity of the different factors in the various member states, a reallocation of factors that eliminates such differences can increase the income and welfare of the group. The 72

THE THEORY OF COMMON MARKETS

migration of factors from countries where such productivities are relatively low to those where they are higher will then be beneficial. It will be accompanied by a tendency for disparities in factor earnings among the different member countries to be reduced. In terms of orthodox theory, the operation of a customs union would itself, through its trade effects, reduce intra-union disparities in factor earnings and marginal productivities. If those effects could be relied upon to equalize intra-union marginal productivities of factors, resources would then be utilized in the most efficient way in the area, and a move from a customs union to a common market would not result in any further increase in allocational efficiency. In terms of that criterion alone, therefore, there would be no advantage in establishing a common market. The conditions required in order that trade alone should equalize marginal productivities are, however, extremely restrictive, and the empirical relevance of the factor—price equalization theorem is generally accepted to be extremely limited. In practice, various reasons such as differences in production functions between member states, or the existence of economies of scale in production, mean that further gains may be anticipated from the advance from a customs union to a common market. These and other sources of gain from factor movement have been analysed in a purely orthodox framework by Meade (1953, pp. 61–73). In this chapter the basic economic implications of common markets are discussed. For reasons of simplicity the analysis is conducted in terms of capital flows. The chapter first considers the additional benefits that may be derived by superimposing an integrated factor market upon an existing customs union that has adopted a suitable common external tariff. Second, it considers the modifications required to the orthodox criteria for evaluating the welfare effects of a customs union or common market in the presence of foreign factors and, specifically, of profit remittances. Third, it comments on some dynamic aspects of common markets. The analysis deals mainly with allocational and distributional aspects of common markets. If, however, a complete common market is introduced, and in particular if there is full financial integration (including substantially integrated bond markets), major constraints will be implied for the autonomy of national macroeconomic policies. Likewise, full factor mobility will imply major limitations on the effective jurisdictional autonomy of national policies in other major areas such as taxation and social security. All these considerations must be taken fully into account 73

THE THEORY OF COMMON MARKETS

in evaluating both the costs and the benefits of the introduction of a common market, and the issue of whether a common market can be a stable stage of economic integration. These further aspects are not considered at all in this chapter, but some are briefly discussed in Chapters 9 and 12.

INTEGRATION AND INTRA-BLOC CAPITAL FLOWS The following analysis applies the orthodox partial equilibrium neoclassical analysis of capital flows to the case of a common market, utilizing for simplicity a two-country model in which a single aggregate product is produced. It is assumed that two countries, H and P, have established a customs union, but that, in the precommon market phase, obstacles to intra-regional factor mobility exist, as a result of which the customs union co-exists with divergences in the marginal social productivity of capita. Factor supplies in each country are given. A single aggregate product is produced. This approach suppresses any consideration of trade diversion and trade creation. A more general analysis resting on a three-country model with two or more goods is presented by Wooton (1988). Figure 6.1 depicts production conditions in each country. The lines MH and MP relate capital stocks in the two countries to the marginal product of capital, given the amount of the other factor, namely labour. Initially, in the customs union phase, the capital stock is assumed to be M in country H and Q in country P. In the pre-common market phase, since capital is assumed to be internationally immobile, all capital stock must be nationally owned. Taxation is ignored. In a competitive model, profits per unit of capital will equal its marginal product. Hence total profits in country H are q+t. Total output is p+q+r+s+t, and labour’s share is p+r+s. In country P, similarly, profits will amount to x+z, and the share of labour will amount to y. Despite the existence of the customs union, the rewards of capital are different in the two countries, being higher in P, reflecting partly that country’s more favourable productive ‘atmosphere’. Let it now be assumed that the two countries decide to superimpose a common market for capital upon the existing customs union, so that obstacles to capital flows between the two countries 74

Figure 6.1 The impact of free intra-regional capital flows

THE THEORY OF COMMON MARKETS

are eliminated. It is also assumed that the total stock of capital of the customs union is unaffected by the introduction of the common market for capital by virtue of exogenous capital controls. Following the establishment of the common market, capital will flow from H to P in search of higher rewards, and a new equilibrium will be established when the distribution of capital between the two countries is such that its marginal productivity is equal, giving H an amount N and P the amount R of the total stock. The outcome is the generation of unambiguous efficiency gains, and the common market is beneficial to both countries, though not to all factors. The domestically produced product of country H declines to p+q+s, but its national product, including inward remittances of profit on capital employed in P, namely v+u, will increase by v-r. In country P domestic product increases by u+v+w, but national product, after allowing for outward profit remittances of u+v, increases only by w. In each country the share of labour in national product is altered in favour of owners of capital in country H, and against owners of capital in country P. In models involving three countries and two or more goods there may be additional effects from intra-regional factor flows to take into account if, as a result of consequential trade flows, the composition of production is changed (Wooton, 1988). For the common market to produce an unambiguous gain may then call for an adjustment in the common external tariff. Moreover, in that framework, factor mobility may not necessarily benefit each member country, so that some intra-union redistribution may be required if all partners are to share in the gains produced by the move to the common market. Furthermore, if intra-market capital flows do not merely entail the application of known and existing technologies in the respective countries (represented by a move along given MP and MH curves), but are also accompanied by a significant transfer of new techniques and know-how, as is very often the case, the consequences of intra-group mobility would be more complex and the conclusions of the static neoclassical analysis would need modification. Qualifications on this account (and others) are discussed in MacDougall (1960) and Grubel (1982).

76

THE THEORY OF COMMON MARKETS

FOREIGN CAPITAL AND THE COSTS AND BENEFITS OF INTEGRATION Customs unions and common markets are unlikely to confront circumstances where capital is completely immobile internationally. Foreign capital has been an important element in the economies of most of the countries that have sought to establish customs unions and other forms of international integration during the past quarter of a century. Where capital is internationally mobile, the production cost of tariffs, which constitutes one element of the cost of trade diversion, will be larger than when capital is immobile (Neary, 1988). But does the presence of foreign capital itself call for a modification of the traditional criteria for evaluating customs unions? At the level of the orthodox neoclassical analysis, it can easily be shown that trade creation and diversion cease to be sufficient indicators of costs and benefits for a member country if foreign capital is present. Basically this conclusion follows directly from the point underlined in the previous section, namely that, when foreign capital is present, the impact of integration on a country’s national income is not identical to its impact on the country’s geographical or domestic income. Whenever foreign direct investment is present in an economy in the shape of foreign enterprises, the effects of integration will be determined partly by its impact upon the net economic rents earned by those enterprises from the use of their exclusive assets. These assets include superior technologies, and special administrative and entrepreneurial capacities. These and other factors permit foreign enterprises to produce at lower cost and thus to earn pure or even quasi-rents even in competitive industries. In a simple partial equilibrium neoclassical analysis, these rents are measured by the magnitude of producers’ surplus. In the presence of foreign rents, a consideration of gains and losses of integration cannot be limited to the orthodox trade creation and trade diversion effects. The reason is that additional gains or losses for the host country will arise from changes in the rents earned by foreign companies because these imply a redistribution of income as between the country of origin of the foreign capital and the host country. For instance, in the case of an importable commodity produced by foreign enterprises whose price falls after integration as a result of trade creation, the host country will gain from the reduction in foreign company rents. This additional benefit has been termed by Tironi (1982) the foreign profit diversion effect. Figure 6.2 illustrates the point. SH and DH represent the demand and supply curves for an importable good produced by country H. 77

THE THEORY OF COMMON MARKETS

Figure 6.2 The costs and benefits of integration with foreign capital

Prior to a customs union, an amount M is produced at a price PH. Assume that, if a customs union is formed, the price falls to PCU. If producers are wholly domestic, the gain to country H will be denoted by areas a+c, as was explained in Chapter 2. If, instead, the commodity is produced wholly by foreign enterprise and capital, there will be an additional national gain equal to area g, which represents the reduction in remittable profits and mutatis mutandis for intermediate cases. Similarly, if there were foreign enterprises in the home country that enjoyed a regional comparative advantage, then, in addition to the standard trade and welfare effects, a further effect, termed by Tironi the foreign profit creation effect, would have to be taken into account in evaluating the costs and benefits of forming a customs union or common market. This foreign profit creation effect is represented by the additional rents or profits that would, as a consequence of the customs union, be obtained by foreign enterprises from their sales in the host country and their exports to the preferential markets of the partner country. For the host country, the counterpart of these additional rents from home sales is a fall in consumers’ surplus, which in this case represents a national income loss from its point of view. At the same time, the additional rents earned on sales to the partner country do not, unlike the cases analysed in Chapter 2, represent a national gain from the host country’s standpoint. For the importing partner country, on the other hand, rents earned on its imports by the foreign enterprise in the host country are reflected in the trade diversion that it experiences. 78

THE THEORY OF COMMON MARKETS

This loss is already accounted for in the orthodox analysis, and no further adjustment on its account is therefore required. Typically host countries participate in rents or profits earned by both foreign and domestic enterprises through taxation. On one view, this is the principal benefit derived by a host country from foreign inward investment (MacDougall, 1960), because the host country gains from more advanced technologies and skills that are imported only to the extent that it can use them without fully paying for them in the form of rents and profit remittances. If a country participates in profits through taxation, the benefit from the foreign profit diversion effect would be reduced to (1-t) d, where t is the rate of profits tax and d is the profit diversion effect. In the limiting case, therefore, where the rate of tax on foreign profits was 100 per cent, the welfare effects of forming a customs union when foreign enterprises are present would be identical to those arising where there are only national enterprises. Similarly, if taxes are paid by a foreign firm that exports to a partner country, the loss to the host country will be smaller. The previous section has explained how the introduction of foreign investment affects the calculation of gain from integration to members of the group, assuming competitive conditions to prevail. In fact, much of contemporary foreign direct investment is undertaken by enterprises that possess a significant degree of market power and whose operations in many cases are conducted on a multi-country basis. These two characteristics of foreign investment pose a number of important questions for an analysis of integration. The basic question is whether the orthodox analysis linking domestic costs and prices with the direction, composition and extent of intragroup trade and the distribution of its costs and benefits continues to be valid when the operations of such transnational corporations are explicitly taken into account. This issue is posed in part by the ability of the transnational corporation (TNC) to bypass the operations of the market which, in the orthodox model, determines the effects of integration on the efficiency of resource allocation. Within a transnational enterprise the activities of subsidiaries can be regulated as an intra-firm matter in ways that, although they may be optimal from the standpoint of the global interests and objectives of the TNC, are not necessarily so from the standpoint of the individual enterprise considered as a profit-maximizing unit, or from the standpoint of the host country itself. Some of the issues that arise for regional integration in such a context are analysed in Chapter 8. 79

THE THEORY OF COMMON MARKETS

FACTOR SUPPLIES, DYNAMIC EFFECTS AND CONVERGENCE Up to now this chapter has been concerned mainly with the static implications of factor mobility for the gains from regional integration. However, just as the institution of a customs union is likely to affect cost and demand conditions and the supplies of factors, and possibly also the rate of growth of the economy partly through dynamic effects, so also may the elimination of restrictions on factor mobility and progress towards a common market. A number of insights into these impacts can be derived from the literature on trade and investment, but strong predictions are few and relevant empirical studies sparse. In general, the analytical literature suggests that taking these factors into account would merely reinforce the presumed static effects. Starting from different but plausible assumptions, however, crucial objections can be raised against certain of the qualitative predictions of static neoclassical theory. In static analysis, the intra-regional movement of factors (here specifically the movement of capital) is seen as promoting the convergence of real income. In a growth context, also, the intraregional movement of capital and other factors is orthodoxly viewed as a vehicle for distributing the fruits of technical progress and of productivity growth more evenly throughout a common market. If, however, there is a tendency for those countries whose growth is most vigorous to attract direct investment and other factors from the rest of the economically integrated area, it may produce contrary effects and have adverse consequences for geographical balance. Member states are unlikely to be indifferent to these effects, even if incomes per capita should thereby be rendered more equal, so confronting them with a trade-off between the goals of geographical balance and of income convergence. Some radical critics of the neoclassical theory argue, however, that there is no such trade-off, because the geographical redistribution of factors that is stimulated by the institution of a common market does not contribute to the convergence of incomes, still less to the convergence of growth rates of income. Instead, it is claimed, the impact of economies of scale and of agglomeration effects, together with the progressive adjustments of costs and demands that accompany the interregional migration of factors, will produce self-reinforcing dynamic effects—termed polarization — that accentuate rather than ameliorate regional imbalances of real incomes. 80

THE THEORY OF COMMON MARKETS

Some empirical support can be found for the existence of such effects, although it is seldom easy to determine to what extent the widening disparities can be attributed to a common market effect rather than to the impact of structural factors that would in any case have produced a similar result even in the absence of a common market. The supposed importance of these dynamic effects constitutes an important part of the general case for instituting a regional policy in a common market. Regional policy issues are discussed in Chapter 12.

CONCLUSION Even if macroeconomic policy issues are disregarded, an assessment of the effects of factor market integration involves many complex issues. Although important efficiency gains may be secured for the market as a whole by a move from a customs union to a common market, even in a purely static appraisal, factor mobility may not benefit each member country, so that some intra-union redistribution may be necessary if all partners are to gain. Additional policy constraints—for instance, on taxation—will also almost certainly be entailed. In addition, in a dynamic perspective, the need for an effective regional policy will almost certainly make itself felt if structural forces contributing to economic divergence and possible losses for certain member states are to be contained. Moreover, if the presence of monopoly is strengthened by a common market through increased opportunities for enterprises to exploit scale economies, there may be adverse allocational consequences to take into account, in addition to adverse distributional effects. Above all, if integration significantly expands the role of TNCs—at least of those based outside the common market and therefore less amenable to the influence of the common market authorities—the assessment becomes still more complex. In terms of policy analysis, the purely competitive allocational gains from a common market constitute only one element of an appraisal of the costs and benefits of moving to a common market that must also take account of those other crucial aspects of the process.

81

7 THE NEW ECONOMICS OF MARKET INTEGRATION

The orthodox theory of customs unions demonstrates that integration can provide expanded opportunities for countries to engage in inter-industry specialization within a bloc, in accordance with comparative advantage, thus bringing about a rationalization of its production. As long as the benefits from rationalization in the form of trade creation are not outweighed by reduced specialization brought about by trade diversion between the bloc and the rest of the world, as a result of the bloc’s discriminatory common external tariff, resource allocation will be improved. By that means a onceand-for all increase in the income and welfare of the member states as a whole may be brought about. The theory assumes a perfectly competitive framework in which homogeneous products are produced by firms that lack market power and incur few or no transaction costs. The core theory—customs union analysis—also excludes inter-country factor mobility. Although comparative advantage has always been a key concept in integration analysis, the thrust of much recent analytical and empirical work has been to question its relevance to contemporary trade and regional integration. The new emphasis has been prompted in part by the inability of the orthodox theory—even when scale economies are incorporated—to explain intra-industry trade in similar products. Yet a large part of modern world trade is made up of such trade. Indeed, much of the increase in intrabloc trade in manufactures that took place following the establishment of the EC took that form and it has become a major component of trade in manufactures. New theories of trade have been developed which are capable of explaining this trade. As already noted, one explanation centred on the interaction of scale economies and product differentation (Krugman, 1979; Dixit and Norman, 1980). The other explanation turned on imperfect 82

NEW ECONOMICS OF MARKET INTEGRATION

competition and the incentives it affords for engaging in active policies of market segmentation (Brander and Krugman, 1983). Both developments were responsible for a marked change in thinking about the gains from international trade and those from regional economic integration that became apparent from the middle of the 1980s.

COMPETITION AND THE GAINS FROM INTEGRATION A prominent feature of the new theories of trade is the contention that a significant part of the trade that is carried on among industrial countries and among the EC and EFTA, in particular, results from arbitrary specialization driven by scale economies and product differentiation (Helpman and Krugman, 1985). The idea that trade may be based on economies of scale is not new. It was explained in the last chapter how their effects can be analysed in the context of regional economic integration. If a good were produced under conditions of increasing returns and transport costs were not excessive, there would be gains from concentrating production in one place while serving other markets by exports. That would be so even if, in a two-product, two-country world, cost functions were identical so that, in that limiting case, trade could not be said to be driven at all by ‘classical’ comparative advantage and relative factor endowments. There would be gains even if the countries were identical, because a larger market allows longer production runs and therefore lower costs. If most cost functions for most goods were subject to increasing returns and were also largely similar, it would be permissible to regard the country chosen for the production of particular products as arbitrary and essentially the result of minor locational advantages, historical accident or government policy. The newer analyses of gains from trade liberalization and regional market integration, while not excluding the operation and importance of comparative advantage based on factor endowments for certain sectors, do in fact focus on situations in which the gains from trade stem purely from increased competition in the presence of scale economies and product differentiation, rather than from comparative advantage as such.1 In considering the role of the new framework it should be borne in mind, as Winters has tellingly 83

NEW ECONOMICS OF MARKET INTEGRATION

remarked (1992), that there is not much that can be said about the relative importance of factor endowments and imperfect competition in trade, or their interaction, so that it is currently impossible to arbitrate properly between the two approaches. There is no reason to suppose, however, that for a wide variety of situations, differences in the characteristics of countries are not a major determinant of the gains from trade and its pattern. In the new framework, at least four sources of potential gain from regional integration can be distinguished, in addition to any that derive solely from increased specialization according to comparative advantage. •

•

•

•

Integration by reducing trade barriers increases effective market size, giving rise to more competition, a decrease in oligopolistic mark-ups and reduced market segmentation. Larger market size encourages longer production runs and so reductions in cost. In the process, some higher-cost firms may be eliminated and the demand met by imports. (This is the cost-reduction effect analysed in Chapter 3.) The increased market size that accompanies trade expansion may enable more products to be produced profitably, so generating welfare gains from increased product diversity. Integration may permit firms to engage in more plant specialization, thus reducing the number of products produced in a given plant, with attendant cost reductions.

Some of these benefits—notably the first—can be secured through integration in the short run, on the basis of existing economic structures. Others, however, can be captured only in the longer term, after industrial restructuring has taken place on the basis of new investment and of divestment. Any such longer-term changes will hinge crucially on investors’ perceptions of the credibility of any market unity that is established. They will also depend in any case on the strategic responses of TNCs to market integration. The latter are discussed in the following chapter. In the context of regional integration, formal analysis has focused on the first two of these sources of gain and the innovations are most prominent in relation to the first, the so-called pro-competitive effect.2 In that context, consider specifically the effects of increased competition and reduced market segmentation that may be expected to accompany integration in partial equilibrium terms. Suppose that integration results in a reduction of tradebarriers that takes the form 84

NEW ECONOMICS OF MARKET INTEGRATION

of a reduction in real trade costs. Let us look first at the effects of increased competition in that context.

The effect of reduced trade barriers Product differentiation means that firms have varying degrees of market power in each of their markets. In typical models, market power and price—cost margins are directly related to market share. Firms are assumed to command larger shares of their home markets and smaller shares of export markets because of the operation of trade barriers and taste biases towards the home product. Prior to integration, firms will charge higher prices for domestic sales than for exports. Following integration, their share of the integrated market will determine their market power. As a result of increased competition resulting from the reduced trade costs of partner-country suppliers in the markets of the bloc, the shares of home firms will be below their initial shares in the home market and higher in the markets of their partners. The loss of market power in their home markets means that their domestic price—cost margins will fall. The fall in domestic prices will be accompanied by an increase in domestic sales. On the other hand, in their partners’ markets, exporters will charge higher prices, since their market shares there will exceed their previous shares. However, since each market is dominated by domestic firms, average prices in each market will fall. The conventional conclusion is that the net effect is unequivocally beneficial to consumers. The salient effects of a reduction in trade barriers that operate through these channels can be illustrated most easily for the simplified case of a single homogeneous product which is sold in segmented markets. Figure 7.1 relates to the market for such a product where two types of firm, home and partner, operate in the market. DH represents the demand curve. The behaviour of all firms in the industry conforms to the Cournot hypothesis—namely, in deciding how much to supply in any market, a producer treats the sales of other firms in that market as given. Home suppliers are assumed to produce at a constant marginal cost MCH and partner-country firms at a constant marginal cost of MCP. Part of the marginal cost of partner-country firms is made up of real trade costs such as frontier formalities and the cost of compliance with national product specifications. For these conditions of cost and demand, there is an initial equilibrium at price P0 and total sales M of which home firms sell N and partner firms sell NM. 85

NEW ECONOMICS OF MARKET INTEGRATION

Figure 7.1 The effects of a reduction in trade barriers

Suppose now that trade costs are reduced through integration, lowering the marginal cost of partner-country firms to MC1P. A new equilibrium will be established with a lower price, P1, and larger total sales (M1). The sales of domestic firms reduce to L and the sales of foreign firms increase to LM1. The welfare implications are as follows. Home consumers gain by the increase in their consumer surplus, equivalent to a+b+c. Home producers will lose, since their profits fall by a+d+h. Partner firms gain, since their profits rise by d+f+g-b. This is made up of the reduction in real trade costs at the initial level of imports (g), profits shifted from home firms as a result of the stronger market position of partner firms (d) and the net effect on profit of increased total sales (f -b). The net effect on the home country is b+c-d-h, which may be positive or negative. The net effect on bloc welfare is thus c+f+g-h. The element h in this case corresponds to a kind of trade diversion, since in the post-integration equilibrium more is supplied by higher-cost 86

NEW ECONOMICS OF MARKET INTEGRATION

producers. If this element were sufficiently large, the bloc at large would not benefit. However, if it is assumed that the excess cost differential is attributable solely to trade barriers, then, with trade cost reductions that are sufficiently large, the net effect on bloc welfare must be positive. The conclusion is that home consumers benefit, since they purchase more of the product at a lower price. Partner-country producers also should benefit, since they increase their sales and market share and they appropriate part of the reduction in real trade costs. Home-country producers lose, because of the fall in price, but also because of the reduction in their market share. It is not clear that the home country benefits from the reduced trade barriers. That will depend on whether the gain to consumers outweighs the loss to home-country producers. It should be noted that, in this particular case, more goods would be supplied in the home market by high-cost producers, but that is not a necessary outcome of a reduction in trade costs. The result is that exporting countries are likely to gain from reductions in real trade costs while importing countries may gain or lose. A reduction in real trade costs between the members of an economic grouping should therefore be beneficial to all countries involved since, although they could lose in their home markets for any product, each country will gain in its export markets. Export gains should outweigh losses at home, thereby giving each member a share in total gains.

The effect of eliminating price discrimination The second potentially important effect of market integration results from the removal or reduction of intra-bloc price discrimination that is a feature of segmented markets in the presence of imperfect competition. In typical models following the assumption made in the reciprocal dumping theory of intra-industry trade formulated by Brander and Krugman (1983), it is assumed that firms charge higher prices at home than abroad. This theory suggests that, because market shares are larger at home, the demand for their products is less elastic. Consequently, it is in the interests of home producers to exploit their domestic market power and to charge higher prices in the home than in the partner market, where they will face higher demand elasticities because of their lower shares. If market integration reduces or eliminates market segmentation, 87

NEW ECONOMICS OF MARKET INTEGRATION

and so the possibility of price discrimination, two effects will follow. First, firms will reallocate sales from sectors of the market where they initially had to charge low prices, namely the export markets, to sectors where they were able to charge a high price, namely the home markets. In addition, the market power enjoyed by all firms will be weakened and total output will increase. The first effect, namely the reallocation of output, is illustrated in Figure 7.2. Suppose that total output is given and that the firm seeks to allocate sales between the home and partner markets so as to maximize profits. The home demand curve measured from origin OH is DH and the partner demand curve measured from OP is DP; the corresponding marginal revenue curves are MRH and MRP. If the two markets are fully segmented, profit maximization requires the firm to allocate outputs between the two markets so as to equate marginal revenue in each. The firm will thus charge a price PH at home and PP abroad, and sell OHSS at home and OPSS in the partner country. If market integration eliminates the ability to discriminate, the

Figure 7.2 The gains from removing market segmentation

88

NEW ECONOMICS OF MARKET INTEGRATION

same price PSM, defined by the intersection of the two demand curves, will be charged in both markets. Home sales will amount to OHSSM and sales to the rest of the bloc to OPSSM. The welfare effects are as follows. Home consumers gain a+b from integration. Consumers in the partner country lose d+e. The firm will increase its earnings from export sales by e-f. In respect of home sales it will lose a but gain c+d+f. The net effect on the firm’s profits is thus c+d+e-a. This amount must be negative, since otherwise the firm would not have engaged in price discrimination in the first place. It can be seen that, for the bloc as a whole, a net gain of b+c accrues from the elimination of market segmentation. However, more than this amount is gained by the exporting country, whose aggregate net gain, taking into account both consumers and firms (b+c+d+e), is achieved partly at the expense of foreign consumers, who lose d+e. If, however, both countries are exporting such a product to each other, and trade is in balance, export gains should outweigh import losses and each country should enjoy net gains equal to b+c. If there is a trade imbalance on such intra-industry trade, the outcome will tend to favour the country that enjoys a positive balance. The last example assumes that total sales are given, in order to focus on the effects that market reallocation can be expected to have on incomes. In addition, however, market integration can be expected to have an effect on the total sales of each enterprise. The loss of ability to discriminate may induce enterprises to produce a smaller output, as would be the case with a monopolist unable to discriminate amongst different consumers in a single market. On the other hand, in an integrated market, the market power of any one producer will be smaller than if the market was segmented. Thus profit margins should be lower and total production higher. Overall, the direction of change is therefore ambiguous. If, however, domestic demand and foreign demand are equally sensitive to price changes, so that price discrimination derives solely from differences in market power at home and in the rest of the bloc, the effect of integration will be to reduce the market power of the dominant seller in each sector of the market. The outcome then is necessarily lower average prices and larger total sales. Haaland and Wooton (1992) point out, however, that the effects may be less clear-cut in the presence of consumer preference for home-produced goods or if trade costs remain positive after integration. The reduced incentive to export that is entailed in the reduction of market segmentation also reduces the competitive 89

NEW ECONOMICS OF MARKET INTEGRATION

pressure from imports in each home market. If this allows home firms to raise their home prices the conventional welfare conclusions concerning the beneficial implications of market integration could be upset. The relevance of this argument is an empirical matter which would have to be determined. Three important points should be noted in connection with the perspective that this analysis opens upon integration. First, in the perspective of such a model, for the industries or sectors in question, intra-bloc imports can be expected to fall as a result of the rise in their prices. Thus, if sectors with the assumed market structure predominate, there could be significant gains from integration even if it should not be accompanied by expanded intra-bloc trade. This is one of several important respects in which the predictions of the newer theory differ from those of the classical approach. Second, considering purely the effects of relative price changes, imports from the rest of the world may also fall, reflecting the favourable effects of the sector’s improved competitiveness relative to the rest of the world. There may consequently be negative trade effects for the rest of the world, but they cannot be equated with trade diversion. However, improved competitiveness may not have such effects if the favourable impact of market integration on the real income of the area is also taken into account. The income effect can be expected to increase the demand for imports from the third countries: if that effect is sufficiently strong, on balance, those countries need not suffer negative trade effects. Third, in the context of this approach, if integration took place between countries with different economic characteristics, and was accompanied by trade diversion, customs unions and market integration might nevertheless be beneficial if the pro-competitive gains from lower prices more than offset the costs of trade diversion. The theory of market integration in situations of imperfect competition and oligopoly is evidently complex. Outcomes, which are notoriously sensitive to the specific assumptions that are made about tastes, technology and entrepreneurial behaviour, can sometimes be paradoxical (Smith, 1987) and are often ambiguous. Moreover, standard models incorporate Cournot—Nash assumptions about the behaviour of oligopolists that are sometimes crucial but implausible and for whose policy relevance empirical support is lacking. A further problem is that the oligopolists whose market behaviour is relevant are typically TNCs. These organizations have the capacity to bypass the market which, even in the new approaches outlined above, is relied upon to determine the effects 90

NEW ECONOMICS OF MARKET INTEGRATION

of integration on resource allocation. The transnational mode itself may permit continued price discrimination among the national markets of a bloc even when the significant public policy obstacles to market integration have been removed. Some of the issues that are specific to TNCs in relation to new approaches to regional integration will be discussed separately in the following chapter. For these and other reasons, it is extremely difficult to generalize about the gains from integration in the presence of imperfect competion and product differentiation. Many of the insights that have been gained in the past decade about the effects of integration in that context have emerged, not so much from the theory, as such, as from policy-driven attempts to apply it to major initiatives on the basis of computable equilibrium models, utilizing relevant trade and production data. Apart from providing estimates of the effects of integration, these evaluations have contributed to an understanding of theoretical interactions in models that are too complicated to study analytically. In North America the initiative for a Canadian—US Free Trade Area generated a large number of official and unofficial studies that attempted to estimate the importance of the trade and income effects of free trade in terms of certain of the sources of gain emphasized by the new thinking. In Western Europe the principal contributions to the application of the new economics of market integration arose out of the EC’s own path-breaking study (CEC, 1988a) of the prospective effects of the completion of of the EC’s internal market by 1992—the so-called Emerson report. It was soon followed by a number of important academic studies which sought to address issues that the EC study left unresolved. Since 1996 a substantial number of additional research studies have become available on the impact of the Single Market programme, mainly as a result of the Commission’s obligation to present a review of the single market programme to the Council and Parliament. The studies in question formed the basis of the communication by the Commission in October 1996 (CEC, 1996c) and are currently (CEC, 1997) being published. The rest of this chapter is devoted mainly to a brief review of the initiative for completing the European common market and of the methodology underlying the ex ante estimate of the gains that was subsequently undertaken by the EU. The overall quantitative results of that study (CEC, 1988a) are also briefly noted. Some comments are also made on growth and inequality in the context of market integration. The chapter concludes with a brief discussion 91

NEW ECONOMICS OF MARKET INTEGRATION

of progress in adopting the single market measures up to 1996, and some tentative indications of the initial economic impact of the programme on trade, investment, market structure and competitiveness in the light of the studies made for the EU’s own evaluation (CEC, 1996c; Monti, 1996).

MARKET INTEGRATION: APPRAISING THE GAINS The Treaty of Rome sought to establish a common market in the sense of a customs union, coupled with the removal of restrictions on the free movement of labour, on the free movement of capital, on the right to establish enterprises, and on the right to provide services—the so-called four freedoms. Thirty years later, in the mid 1980s, it was painfully clear that the integration of the European economy was still severely limited by barriers to trade and investment between the member states. In the view of many key players in industry and finance, the weak international performance of the EU in many crucial economic sectors was bound up with the continued policy-induced fragmentation of its economic space, caricatured by unsympathetic critics as ‘the uncommon market’. It was in this context that the Commission set out the case for a large-scale, comprehensive assault on the limitations of the common market. The eight-year programme, which was to be completed by the end of 1992, was set out in the Commission’s White Paper (CEC, 1985), Completing the Internal Market. Unlike some other EC initiatives, the plan was not launched in a political vacuum. The aim of completing the common market was included in the Single European Act of 1986, which provided an important political framework for the succesful implementation of the legislative programme. The Act set a precise deadline for the introduction of the single market, namely 1 January 1993, and it reformed the decision-making process by introducing qualified majority voting for matters to do with freedom of movement and the completion of the internal market.

Market barriers The objective of completing the internal market by the end of 1992 required the elimination of all barriers to the free flow of

92

NEW ECONOMICS OF MARKET INTEGRATION

goods, services, capital and labour. Since tariffs and quantitative restrictions on trade had already been largely eliminated, the remaining barriers consisted essentially of a range of impediments that can be classified into two categories, namely cost-increasing barriers and market entry restrictions. •

•

Cost-increasing barriers are the target of many of the specific measures proposed in the White Paper. Delays at customs posts are in this category, whether they are related to the assessment of value-added taxation and excises, the collection of statistics or the verification of technical regulations. In addition, complying with different national technical regulations and norms in production and distribution increases costs. Some of these costs affect only that part of output which is exported, while others, like the cost of complying with technical regulations in production, affect the whole of production. Market entry restrictions are also targeted by many of the measures specified in the White Paper. Entry restrictions include, notably, government procurement policies, the right of establishment for certain service industries and professions, restrictions in some service sectors that prevent or limit direct trading across frontiers (e.g. insurance and electricity) and restrictions on entry into some regulated markets (e.g. civil aviation).

Calibrating the gains from a single market Some time after the 1992 programme had been set in train, the EU launched a comprehensive quantitative assessment of the prospective economic effects of the programme based on actual trade and production data for the major EU economies. The overall report was published in 1988 (CEC, 1988a; Emerson et al., 1988). It was supported by sixteen volumes of studies, of which volume 2, Studies in the Economics of Integration (CEC, 1988b), is central. The overall report (Emerson, 1988) suggested that the completion of the market would raise GNP by 5 or 6 per cent in terms of 1985 data. Although the precise figures may be debated, the broad thrust of the study remains unchallenged, and the methodology itself is of outstanding interest. The outcome is summarized in Table 7.1. Four categories or stages of economic loss from the fragmentation of the market are identified and the gains from overcoming them are quantified. Categories 1 and 2 relate to the direct short-run effects 93

NEW ECONOMICS OF MARKET INTEGRATION Table 7.1 The economic gains from completing the internal market, 1985

Source: Derived from CEC (1988a, table 10.1.1). Notes a Estimates are based on partial equilibrium methods and relate to seven member states (Germany, France, Italy, the UK and Belgium, the Netherlands and Luxembourg). b The figures are based on 1985 data at 1985 prices. c Scaling up to totals for the 12 member states to represent the same GDP share gives a range of ECU 173 billion to ECU 257 billion in terms of 1988 prices.

associated with the elimination of the resource costs of barriers. Categories 3 and 4, termed indirect, are associated with increased competition and, to the extent that they depend on restructuring, would take longer to manifest themselves. The estimates for the first two categories relate to the potential static welfare gains that would accompany the full implementation of the internal market programme, based on the partial equilibrium analysis of Cawley and Davenport (1988). Essentially this amounts to applying the orthodox analysis that was described in Chapter 2 to the specific case of cost-increasing barriers. Such barriers are similar to tariffs in their operation. However, since the barriers are not revenue-raising but, rather, unnecessarily resource-using, the potential for gain from reducing the barriers is greater than the gain from an equivalent tariff. The basic cost savings are based on primary surveys, and these are translated into price, trade and welfare effects in the conventional way. For the trade barrier elimination, extra-EU trade falls by 2.6 per cent while intra-EU trade 94

NEW ECONOMICS OF MARKET INTEGRATION

expands by 4.5 per cent. The cost of the pure trade barriers is relatively modest, and the net improvement in EU welfare from overcoming them is put at ECU 9 billion, or 0.3 per cent of GDP. Stage 2 considers the effects of eliminating barriers affecting all production, that is to say, internal cost-increasing restrictions that affect all production, including technical regulations and standards. The largest single beneficiary is the financial sector, but significant gains are identified throughout industry. The net effect on welfare is put at ECU 71 billion, or 2.4 per cent of GDP, which is accompanied by a further reduction in extra-EU imports of 7.7 per cent. The total direct cost of barriers is thus ECU 80 billion, or 2.7 per cent of GDP. These may seem to be modest gains, but they are nevertheless huge by comparison with the estimates of the gains from trade creation in the original EC that were made by Scitovsky, Balassa and others. In terms of the economics of integration, the most interesting part of the analysis relates not to the foregoing category of gains but to its method of evaluating those gains expected to be produced by market integration as a result of the effects of increased competition in imperfectly competitive markets and better exploitation of scale economies. The estimates of gain from categories 3 and 4 are derived from the evaluations undertaken by Smith and Venables (1988b), which rest on Brander and Krugman’s (1983) reciprocal dumping model and Krugman’s (1979) trade model, which centres on the demand for variety. The methodology abstracts from some important aspects of inter-market competition (Baldwin and Venables, 1995). The sources of potential gains from integration in such a perspective have already been outlined. Esssentially, reducing barriers increases competition, reduces profit margins, reduces average costs of production and increases intraEU market shares. In terms of such a framework, the Smith— Venables partial equilibrium study estimates on two alternative bases and for ten major industries the effects on trade, output, production costs and welfare of a reduction in trade barriers equivalent to a reduction in the costs of intra-EU trade of 2.5 per cent. On the first basis, it is assumed that EU markets continue to remain segmented after ‘1992’, so that firms can continue to charge different prices in each EU market. On the second basis it is assumed that markets are completely integrated, so that firms have to charge the same price to all buyers in the EU. The results vary from industry to industry but, for example in the case of office machinery, on the first basis, which assumes continued market segmentation, the 95

NEW ECONOMICS OF MARKET INTEGRATION

reduction in barriers would result (with a fixed number of firms) in an increase in intra-EU imports of 45 per cent, in extra-EU exports of 6 per cent and a decline in extra-EU imports of 25 per cent. These changes reflect the favourable effect of market integration on the EU’s competitiveness and they indicate a substantial amount of a kind of trade creation. Overall, for the industries as a whole, if the first basis is employed, only modest welfare gains are suggested. If the second basis is employed, the trade effects are considerably reinforced, and the estimates suggest that the policy yields large welfare gains. The largest gains are naturally found in sectors that are most concentrated and which have the greatest unexploited economies of scale. It is noteworthy that, in sectors where these characteristics are most in evidence, the indirect gains can be several times higher than the direct gains. The Commission’s report extrapolates the results from the tenindustry study to the economies as a whole on the basis of ratios of the welfare gains obtainable from market integration (stages 3+4), assuming full integration, to those of the static direct (stages 1+2) gains. The ratios vary from unity for unconcentrated industries with few scale economies to six for highly concentrated industries where scale economies are important. All industries are then classified according to scale and concentration and the appropriate multiplier is applied to the static welfare estimates. This procedure yields a total of about ECU 62 billion for the welfare benefits of stages 3 and 4. The EU report goes on to disaggregate the indirect effects into the competition effects and the economies of scale effects. The competitive effects are calculated from Smith and Venables by comparing the gains with fully integrated markets and variable industry size with those yielded by segmented markets and fixed industry size. On that basis, the bulk of the gains in the Smith— Venables model can be attributed to the purely competitive effect on profit margins, namely ECU 46 billion, while scale economies account for only ECU 16 billion. The report then adds to those competitive gains a separate, much larger and quite differently based estimate of the gains from exploiting the economies of scale which puts them at ECU 61 billion, or nearly four times what the Smith—Venables method suggests. The outcome on that basis is a figure for total direct and indirect welfare gains of ECU 180 billion, or 6 per cent of GDP, compared with ECU 142 billion, or 5 per cent, if the smaller estimate of gains from scale economies derived from Smith and Venables is employed. These figures of the 96

NEW ECONOMICS OF MARKET INTEGRATION

estimated gains from the partial equilibrium analysis are summarized in Table 7.1. The modest once-and-for-all gains from such a radical programme may seem not to warrant the great public interest in the completion of the single market. The response of policy makers to that criticism has been to emphasize that the greatest benefits of 1992 are to be found not in these once-and-for-all gains but in its dynamic effects on innovation, productivity improvement and investment, and so on growth. Even small increases in the growth rate, if sustained, would indeed soon result in large increases in real income. The Emerson report explicitly recognized the importance of these dynamic aspects but did not bring them into the estimates because they are difficult to quantify.

Appraising the estimates In appraising the Commission’s estimates of potential gain from completing the market and, in particular, their overall magnitude and significance, a number of points should be borne in mind. 1. Like the orthodox analysis discussed in the initial chapters of this book, this analysis is essentially equilibrium analysis. Adjustment costs and the period of adjustment are both disregarded, and it is assumed that all resources released in the process are re-employed. In the case of the restructuring of firms, the Emerson report suggests that the larger part of adjustments would be completed in five years. 2. Estimates of net welfare gains require that reductions in economic rents should be allowed for in computing net gains, as was explained for the orthodox theory in the comment on Figure 2.1. This has not been attempted for all sectors, and for some sectors, notably financial services, it could, the report suggests, be a significant source of overestimation. However, any loss of foreign rents (CEC, 1988a, does not consider ownership) would operate in the other direction, as noted in the following chapter. 3. Bias and errors may be introduced as a result of the summation of partial estimates, rather than as a result of the estimation of gains within a general equilibrium framework. For instance, the GDP effect is not reflected in the trade effects. A later study within the framework of a computable general equilibrium model (Smith et al., 1992) does in fact broadly confirm the partial equilibrium analysis cited earlier, but does not entirely support the size of the official estimates based on them, though it must be emphasized 97

NEW ECONOMICS OF MARKET INTEGRATION

that the latter study does not allow for the deregulatory gains of stage 2. 4. It is arguable that the Emerson report is misleading because it gives too much weight to gains from economies of scale. Geroski (1989) contends that EU firms are not particularly small relative to the EU market. (The EU’s own data suggest that most industries would support twenty or more firms at about the minimum efficient scale.) In addition, the diversity of tastes would exclude production runs that could fully exploit potential economies of scale even in a fully integrated market. Pelkmans and Winters (1988) also caution against relying too heavily on economies of scale in estimating potential gains.

Growth effects It might on the contrary be argued that the gains from market completion may have been underestimated by the Commission’s studies because of their disregard of dynamic and growth effects. The Emerson report recognized the importance of dynamic factors and discussed a number of channels through which, for example, the completion of the market might have favourable effects on innovation, productivity and investment, and so on output growth. However, the estimates are purely static and do not embody any of these effects which, particularly for innovation, are extremely difficult to quantify. Baldwin has argued (1989) that it is possible to calibrate some of the dynamic effects of 1992, particularly in relation to the induced effects on savings and investment. This leads him to contend that the Cecchini report (Cecchini, 1988) that was published in parallel with the Emerson report significantly underestimated the impact of the single market programme. The static efficiency gains from the single market programme amount to producing an upward shift in the production function. If it is a proportional shift, the marginal productivity of capital will rise. There will thus be an incentive for firms to undertake additional investment until the marginal productivity of capital is once more equal to its cost. In the simplest of neoclassical growth models, the rate of growth of the capital stock is equal to the savings ratio multiplied by the output—capital ratio. Baldwin assumes that the output effects of 1992 do imply a proportional increase in the output—capital ratio. This leads to an increased rate of growth of capital stock and so, with a constant savings ratio, to an increase in 98

NEW ECONOMICS OF MARKET INTEGRATION

the rate of growth of income per head. However, with constant returns to scale, which orthodox growth theory assumes, the increased rate of growth and output lasts only over the medium term while the additional investment is being undertaken. The longterm growth rate is unaffected so long as technology is given. On the basis of estimates of production functions for several European countries, Baldwin suggests that what he terms the ‘medium-term growth bonus’ of the single market programme would amount in total to as much again as the static efficiency gains calculated by the Commission. The welfare implications of the growth bonus are, however, quite different from those of the static efficiency gains and, as Baldwin points out, much smaller since, unlike the latter, they do not increase the output that can be produced with given factors of production. Some ‘new growth’ theorists contend that, economy-wide, there are increasing returns to scale. On that assumption, any increment in the rate of growth resulting from the static efficiency gains would not only be larger but would also last longer. It is even possible that the growth rate could be permanently raised. For instance, relying on a specific application of a model by Romer (1986) for a case in which the capital—output elasticity is unity, Baldwin finds that the static gain from 1992 would increase the EU’s growth rate permanently by more than 0.5 per cent. That would imply an extra 5 per cent of real income not just once but every ten years.

The external impact of completing the Common Market The studies underlying the European Commission’s report (CEC, 1988a) contain estimates of the trade effects of 1992 on particular sectors. These can be used as a basis for looking at the external impact of 1992. The external trade effects broadly result from a number of developments. First, imports from outside the EU will be stimulated as a result of the predicted expansion in EU income. Second, the completion of the market leads to price and cost reductions and so to increased competitiveness for EU producers. The main result will be a displacement of imports from outside the EU by EU producers. This reduction should not be equated with diverted trade. Third, the reduction in barriers (for example, fiscal barriers) may imply some additional discrimination against imports and hence some import reduction. But a number of the changes— 99

NEW ECONOMICS OF MARKET INTEGRATION

for example, the harmonization of standards—may be beneficial to all traders, and some, like the substitution of EU quotas for national quotas, would be expected to benefit foreign traders. Fourth, external trade effects may cause changes in the terms of trade, as has been discussed in Chapter 3. A variety of studies have appeared which have looked at the impact of completion on third countries through these channels, including the US (Hufbauer, 1990), the Nordic countries (Haaland, 1990) and developing countries (Davenport and Page, 1991). On the whole, these studies suggest that the impact on third countries would not be of very great significance one way or the other. For developing countries they could be favourable, but only modestly so. For industrial countries they could be negative, but insignificantly so.

The overall picture The new economics of market integration addresses unconventional gains from regional integration. One aspect of these gains is concerned with the direct benefits to be expected from policies designed to reduce certain trade costs among the members of a group (but not the important transaction costs that arise from the existence of multiple currencies) and it has general application irrespective of the relative roles played by factor endowments, on the one hand, or imperfect competition, on the other, in generating intra-regional trade. The other aspect is concerned with gains from integration that are produced by increased competition for sectors where product differentiation, economies of scale and imperfect competition are significant. In its specific application to Europe, as manifested in the 1992 programme, significant gains are predicted in terms of the new approach. The studies outlined point to short-term gains from eliminating non-tariff barriers of 2.7 per cent of GDP and longerterm gains up to an additional 4 per cent. It is important to note that the achievement of these gains would not be at the expense of the trade and welfare of the rest of the world. Not much attention was paid to the possible redistributive effects of completion, but a number of later studies (Neven, 1990) have focused on this aspect. As to the size of the static gains suggested by the ‘official’ estimates, even if these appear over optimistic in the light of subsequent ex ante academic studies in the same 100

NEW ECONOMICS OF MARKET INTEGRATION

mould, on the whole those later studies have not seriously undermined the previous estimates of potential gains. Perhaps the most striking feature of the empirical analysis of regional integration in terms of the ‘new’ economics is that the predicted gains are several times larger than the integration gains (of less than 1 per cent) from orthodox trade creation commonly estimated by a range of studies at the time the customs union was established and in the following two decades (Balassa, 1975) for the EU and other blocs. Although the new results for the EU cannot be simply transposed to other groupings, many of the factors that generate the results are without doubt of widespread relevance. To that extent they have an important bearing on the rationale for integration throughout the world, and not least for integration among developing countries. A second striking feature is that a large part of the gains arises when firms are assumed to treat the market on a Europe-wide basis so that the sources of price differences between different national markets in the EU are removed. These potential gains will be achieved only if the behaviour of national governments and private enterprises does not frustrate the adjustments that are called for. In the following chapter, some aspects of the behaviour of transnational enterprises will be considered in the context of its possible impact on regional integration from this and other points of view.

THE PROGRESS AND IMPACT OF THE SINGLE EUROPEAN MARKET How far has the EU progressed towards a single market since the inception of the Single Market programme? What are the remaining gaps? What impact has the programme had? The most comprehensive review of these matters to date is to be found in the Single Market Review undertaken by the European Commission (CEC, 1996c, d) and presented in the Monti report (1996) and the background studies (CEC, 1997) on which it is based. The timing of this review, at least so far as concerns the studies of the programme’s impact, is perhaps premature since many measures, in particular those relating to the important area of technical trade barriers, were not implemented until 1994, and it will take some years before the effects of these and many other measures are fully felt, although some have been anticipated by businesses. 101

NEW ECONOMICS OF MARKET INTEGRATION

Nevertheless, the Monti report provides an unrivalled overall review of progress towards the adoption of the many measures involved in the Single Market programme and its preliminary assessments of impacts merits scrutiny. The report also provides a valuable statement on shortcomings, including those of implementation and transposition into national legislation and practice that have yet to be overcome if the single market is to become fully effective. In the light of the report’s findings, the Commission adopted in early 1997 an action programme designed to ensure that the single market would be fully operative by January 1999, when the single currency is scheduled for introduction.

Implementing the single market Progress towards a single market can be evaluated in terms of the implementation of measures to overcome obstacles to the free movement of products, services and factors of production.

Free movement of products In contrast to the situation that prevailed a decade ago, substantial progress has been made towards the goal of a single market for manufactured products. Changes in customs and fiscal procedures, discussed at length in Chapter 10, have abolished physical checks at borders and reduced the differential costs of cross-border trade. The replacement of the present transitional system by an originbased system of VAT could eliminate differential costs altogether but poses other problems which are likely to r emain insurmountable for some time. Technical trade barriers associated with different national industrial standards and health and safety regulations have been the most prevalent impediment to the free circulation of products in the EU. They strike at the heart of the single market if they require products to be re-engineered or tested to local specifications, probably at considerable cost. The mutual recognition principle has been central to the EU efforts to overcome regulatory technical barriers. It derives from the ruling of the Court of Justice in the Cassis de Dijon case in 1979 that member states must allow market access to products manufactured in accordance with partner-country specifications that 102

NEW ECONOMICS OF MARKET INTEGRATION

embody ‘equivalent’ levels of protection for attaining the regulatory objectives of the country of destination. Some 75 per cent of intraEU trade is potentially subject to regulatory barriers. The mutual recognition principle has been relied upon to overcome barriers affecting 25 per cent of intra-EU trade. In situations where national approaches are so divergent as to rule out the application of the principle, EU legislation harmonizing technical specifications is the only way to remove the barriers. This has involved two distinct approaches: the so-called old approach, and the new approach, dating from 1985. The ‘old’ approach involves issuing directives that incorporate detailed technical specifications. For many years progress was painfully slow, but legislation is now in force covering motor vehicles, chemicals, pharmaceuticals and foodstuffs. These measures substantially overcome barriers affecting sectors that account for 30 per cent of intra-EU trade. It is estimated that the motor vehicle directive will reduce the costs of developing a new car by 10 per cent (Monti, 1966, page 24). Under the ‘new’ approach introduced in 1985, EU legislation is limited to prescribing the requirements that are seen to be essential to meet health, safety and other objectives. The task of laying down the technical specifications needed to meet the requirements is left to European standardization bodies. They remain voluntary but producers who adopt the standards are presumed to conform with the essential requirements. Seventeen new approach directives had been adopted by the time of the Single Market Review, covering both consumer products such as toys and equipment such as lifts. The products in question account for 17 per cent of the value of intra-EU trade in manufactured products, but a number of the directives are only now coming into force. One area where obstacles to the free movement and supply of goods have patently not been overcome is in relation to public procurement. Freedom of movement applies in principle to the supply of goods to public purchasing bodies, but whereas member states meet a high proportion of their requirements of private consumption by supplies from partner states, the share of public procurement purchases sourced from partner countries is only 10 per cent. In the interests of transparency, tender notices are required to be published and contracts awarded to the most economic tender, but member states are delaying in incorporating procurement directives into national legislation (only three have done so completely) and do not enforce them effectively. 103

NEW ECONOMICS OF MARKET INTEGRATION

Free movement of services Freedom to provide services across the internal frontiers of the EU is provided for in the Treaty of Rome but, until the Single Market programme, little progress had been made, not only in relation to traditional services such as transport, insurance and banking, but also in relation to services associated with the development of newer technologies such as audio-visual services and certain forms of telecommunications. Prior to the Single Market programme, markets for these services were segregated by outright prohibitions on competition from non-domestic suppliers or quantitative restrictions, as in the road freight industry, or privileged treatment for national suppliers, as in air transport. In other sectors, national regulations of a prudential nature increased the cost of entry or establishment, as in financial services, or deterred the provision of services, as in insurance. The Single Market programme sought to open up the market on a non-discriminatory basis by separating the issue of market access from technical rules on supervision, market stability, safety and consumer protection, by setting minimum EU licensing requirements for operation, by harmonizing general interest provisions and by pursuing the mutual recognition approach where the right of establishment to practice was conditional on qualifications. The Single Market programme is widely perceived to have removed most of the major obstacles in most but not all sectors. In road freight, air transport and telecommunications, for instance, services can now be freely provided across frontiers, and new alliances and market entrants in air transport and telecommunications are increasing competition and efficiency. In banking services, the introduction of the ‘single banking licence’ has substantially reduced the cost of establishment in partner countries. In certain areas, however, such as insurance, the Single Market programme has not yet succeeded in opening up national markets. In relation to energy (gas and electricity) as well, although important steps were taken in the early 1990s towards market liberalization, these efforts have not yet come to fruition and the regulatory framework needs further attention.

Free movement of capital Almost all impediments to the free movement of capital have been removed. Complete liberalization of capital movements was introduced 104

NEW ECONOMICS OF MARKET INTEGRATION

from 1990. In 1993 controls on exchange and financial transactions were removed. New treaty provisions prohibit all restrictions on capital movements and payments between member states. The remaining restrictions derive from the right of member states to restrict capital flows because of differences in the tax treatment of certain items such as income from savings or mortgage interests payments, or to ensure the application of prudential controls in the financial services field. The most significant restrictions affect investment undertaken by pension funds, where six member states currently impose restrictions which hinder the maximization of returns.

Free movement of labour Freedom of movement for persons and the abolition of discrimination by nationality is required by the Treaty of Rome and has long been a reality. Supporting measures include the right to transfer social security entitlements. Within the framework of the Single Market programme, provision was made for the aggregation and transferability of acquired pension rights. The principle of free movement for persons is, however, not yet free of tax barriers. Despite substantial differences in levels of wages and social security benefits among the member countries, there has been no large-scale migration of labour within the EU. The principal adjustment mechanisms have been through increased trade and capital mobility. Surveys nevertheless show a slow but steady increase in EU nationals working in other member states but significant migration is confined to the professional classes and to specialist technicians. Contract working, in building for instance, is also important.

Enterprise A framework has been established over the period 1968–89 that enables companies to conduct cross-border business with security and to ensure the freedom of establishment. Nevertheless it cannot be said that free movement has been achieved for companies in the single market. The degree of harmonization of national approaches that has been achieved is limited. Cross-border mergers are still hampered by legal problems. Tax problems hinder countries from viewing the EU as a single market. There is no common system 105

NEW ECONOMICS OF MARKET INTEGRATION

for EU-wide consolidation of losses, for example, and problems of double taxation have not yet been resolved. Some of the differences confronted do not present insurmountable obstacles to companies wishing to exploit the potential of the single market but they do impose costs on companies that are reflected in complex structures which limit flexibility and may impair competitiveness. The Regulation on the European Economic Interest Grouping is one innovation that is designed to encourage co-operation amongst enterprises from different member countries in the development of joint projects. Beyond that, the Commission has argued that a European company statute is required to enable individual companies to adapt their corporate structure to exploit the single market. There is little sign that this proposal will be soon adopted.

The impact of the Single Market programme The Single Market programme can be expected to produce three types of economic effects: allocation effects, producing efficiency gains; accumulation effects, contributing to higher growth rates; and location effects on the geographical allocation of resources among member countries and/or regions. The Emerson report (CEC, 1988a) attempted an ex ante evaluation of some of these impacts, notably the first. The Single Market Review attempts to provide an evaluation of the ex post effects. This is not easy, for various reasons. First, the period is too short. Many of the measures were not adopted until recently. Statistical data stop in 1994 or earlier. Second, the Single Market programme is not the only factor affecting developments, since it has been taking shape against the background of other changes of considerable significance, namely the enlargement of the EU, German reunification and the progressive globalization of the world economy. There are thus major methodological problems in gauging the impact of the single market programme, since this requires an estimate of how the EU variables would have developed in the absence of the programme. A comparison of observed developments since the programme was introduced would indicate its effects reliably only if no other factors significantly influenced the outcome. With that proviso, the following paragraphs look briefly at the impact on specialization, trade, cross-border investment, market structures, competition and income. The Single Market programme does not seem to have resulted in increased sectoral specialization as a result of concentration on 106

NEW ECONOMICS OF MARKET INTEGRATION

particular activities according to comparative advantage. There is, however, evidence of increased intra-industry trade involving specialization by countries within sectors on different price—quality ranges. With respect to the trade effects of integration, one common measure due to Truman (1975) uses changes in the share of apparent consumption in member states (defined as domestic production plus imports minus exports) supplied from each of three sources: domestic production (net of exports), intra-EU, and extra-EU imports. If no other event has significantly affected these shares, a decrease in the share of domestic production is indicative of ‘internal’ trade creation if the share of intra-EU imports has increased, or of ‘external’ trade creation if the share of extra-EU imports has increased. Conversely an increase in the share of domestic production is indicative of trade diversion. Calculations by Sapir (1996) for the period 1988–92 suggest that the internal market programme has resulted in both internal and external trade creation. There is no evidence that 1992 has operated to the detriment of third countries. There is some evidence that the single market has made the EU more attractive to foreign direct investment (FDI). In the early 1990s the EU absorbed 45 per cent of global foreign investment flows, by comparison with 28 per cent in the mid 1980s. The single market also had substantial positive effects on intra-EU flows which increased sevenfold between 1984–5 and 1992–3. In recent years the EU has seen a rapid growth in mergers and acquisitions, many of which have involved cross-border transactions, reflected in the previously noted expansion of FDI flows. This has been accompanied by substantial increases in the concentration of European industry. Nevertheless, a number of indicators suggest that the Single Market programme has had a favourable effect on the degree of competition. Price—cost margins for manufactures appear to have undergone a significant reduction. Increased competition in services has resulted in substantial and quite general reductions in the prices of telecommunications and selective reductions in sectors of air transport, banking and road haulage. As to the overall macroeconomic effects of the Single Market programme, a tentative assessment (CEC, 1996a) puts the gain in GDP for 1994 at 1–1.25 per cent higher than it would otherwise have been without the Single Market programme and average inflation at 1–1.25 percentage points lower. These gains are much 107

NEW ECONOMICS OF MARKET INTEGRATION

smaller than those predicted in the Cecchini report (Cecchini, 1988), but most countries have still not implemented all single market measures. Moreover, the estimates may not adequately allow for the adverse effects of recession and of German reunification. Details of these impact studies are to be found in the Economic Evaluation of the Internal Market (European Economy No. 4, 1996) and in the thirty-nine volumes of the Single Market Review (CEC, 1997).

NOTES 1 2

For the theoretical relationship between the two approaches see V.D. Norman’s contribution to the discussion of Neven (1990) in Economic Policy, No. 10, p. 50. A fuller treatment will be found in Bliss (1994), chapter 6.

108

8 TRANSNATIONAL CORPORATIONS AND REGIONAL INTEGRATION

The relationship between cross-border business integration through the operations of TNCs and regional economic integration is an intimate one (Dunning and Robson, 1988). First, regional and corporate integration both seek to overcome forms of market failure. In the case of regional integration the primary motivation is to reduce the allocative and efficiency costs of market distortions induced by national policies. In the case of TNCs, a primary motive for adopting this form is to avoid the transaction costs of using markets to organize particular activities by internalizing intermediate product markets. When the TNC form is adopted, it may also enable the enterprise to reduce the allocative and efficiency costs of distortions produced by public policies in the context of its global operations. Since TNCs typically account for a significant proportion of regional production, investment and intra-bloc trade in most blocs, it is important to pose the question of how, if at all, the effects of regional integration are modified, qualitatively or quantitatively, by the existence of the transnational form and the behaviour of TNCs in that context. Newer models of multinational enterprise in the context of trade theory (Horstman and Markusen, 1992) which relate the decision to go multinational to avoiding barriers to trade are not sufficiently specific to illuminate this issue. Recent policyoriented analyses undertaken for blocs in which TNCs are of dominant importance, e.g. the EU and the Canada—US Free Trade Area, have not attempted to consider the issue at all. The studies undertaken by the European Commission in connection with the completion of the internal market that were discussed in the previous chapter virtually ignore TNCs in any of their relevant aspects. Some of the studies undertaken for 109

TNCs AND REGIONAL INTEGRATION

the Canada—US Free Trade Agreement do grapple with the interaction of integration and TNC activity, but only partially. Within the framework of orthodox theory there is no scope for a formal analysis of the interaction between regional economic integration and TNCs, since firms, ownership and internalization considerations are entirely disregarded. Moreover, the orthodox paradigm could not explain intra-industry trade which, though not identical with the intra-firm trade of TNCs, is closely bound up with it. The development in the 1980s of newer theories of trade and integration and of industrial organization have opened the way to a systematic consideration of the interaction of regional and corporate integration, but a fully articulated model has yet to emerge. The basic question is whether the analysis linking domestic costs and prices with the direction, composition and extent of intra-group trade, and the distribution of its costs and benefits, continues to be valid when the operations of TNCs are explicitly taken into account. There are three important ways in which the operation of TNCs may affect the costs and benefits of regional integration. First, there may be efficiency effects to consider in the context of internalization. The capacity to internalize may permit a TNC to overcome public policy distortions and transaction costs to be reduced in its own interests. This behaviour need not necessarily be beneficial from the standpoint of the objectives of the bloc. Second, the existence and conduct of TNCs call for explicit consideration of the distributional effects where foreign TNCs are involved. Third, a TNC’s strategic behaviour in pursuit of market power has to be considered, since it may run counter to any efficiency gains engendered by its capacity to internalize.

INTERNALIZATION AND THE GAINS FROM INTEGRATION A crucial characteristic of a TNC from the standpoint of integration is that its operation involves intra-firm trade and the creation of internal markets for goods and factors. The price which applies to intra-firm trade (the transfer price) is an internal price. From the standpoint of the TNC such prices have a twofold significance: first, they may be shadow prices, designed to bring about an international allocation of resources within the TNC that is efficient in the sense 110

TNCs AND REGIONAL INTEGRATION

that it maximizes its global profits; second, they may be a means of determining the international location of profits in pursuit of the goal of tax minimization. Both aspects are relevant to an appraisal of regional economic integration in the presence of TNCs. The key questions concern the effects that such transfer pricing by a TNC in its own interests will have on global welfare, on the welfare of the bloc, and on individual countries within it. On one view, transfer pricing reduces the inefficiency caused by government intervention designed to segment national markets by arbitrating differences in (for instance) tariff and tax rates and interest rates (Aliber, 1985). To the extent that it does, where TNCs are dominant, the prospective efficiency gains in a bloc would be reduced. In other words, TNCs would already, in advance of formal regional integration, have brought about a substantial measure of regional cross-border integration through their capacity to surmount certain publicly imposed market distortions, notably tariffs. For that result to be achieved, however, TNCs would have to choose what Eden (1985) has termed the efficient transfer price, or internal market-clearing shadow price. This is the price that produces an allocation of TNC resources closest to what would occur under free trade. But it is only one of several options which a TNC may rationally choose, even in a model which excludes oligopoly. With respect to tax-minimizing transfer pricing, in the simplest of static models with given cost and demand conditions, it can be shown that if a TNC is the sole producer of a particular product in an integration group that is subject to a tariff in the importing country, pre-integration tax-minimizing transfer pricing has no effect on allocative efficiency, since output and price remain unaffected. However, transfer pricing makes possible an international transfer of income from the tariff revenue of the country of importation to increased monopoly profits for the producer. With integration, there is price convergence and a net gain in income and welfare in the bloc as a whole. The incentive to engage in tax-avoiding transfer pricing is eliminated. Consumers in the importing country gain from a fall in the price of the import but the overall net benefit to the importing country may be positive or negative. In the case of multi-plant production by a monopolist the results are more complex. The pre-integration efficiency cost of tariffs can be partly offset by the TNC’s ability to determine transfer prices. Integration will result in still more efficient production. However, it may not be sufficient to ensure an increase in aggregate consumer 111

TNCs AND REGIONAL INTEGRATION

welfare, because of induced price and profit changes, particularly if most potential efficiency gains have already been captured as a result of internalization in the preintegration situation (Robson and Wooton, 1993). In neither case, however, is the outcome qualitatively different from that to be expected without internalization. The incentive to engage in tax-avoiding pricing is not necessarily fully removed by customs union to the extent that other taxes are operative, but if market integration is also accompanied by appropriate fiscal harmonization the incentive could be largely eliminated. But, when competitors exist, other issues arise. A TNC’s capacity to internalize markets facilitates not only practices aimed at efficiency gains or tax minimization, but also strategic behaviour aimed at market control. Strategic responses to regional economic integration may affect its gains quantitatively, but it is even possible that they may affect the results qualitatively. If transfer prices are used for such purposes as cross-subsidizing a subsidiary in order to eliminate local competition, maintaining a position of market dominance, or hindering potential competitors from entry, their manipulation must be expected to affect not only trade patterns but also the character of industrial production, static and dynamic efficiency and the intra-regional distribution of benefits from integration. The simpler static models that attempt to deal with such issues are certainly open to the objection that they neglect crucial aspects of TNCs’ operations. But broader organizational analyses of TNCs sometimes reach similar conclusions. Starting from such approaches, a number of economists (Buckley and Casson, 1976; Dunning, 1977) have contended that a TNC’s ownership links, which generate economies in transaction costs and benefits from co-ordination, do not themselves significantly affect the locational distribution of its physical production activities or the scale of its plants—a proposition that Caves (1980, p. 322) has termed the ‘separation’ theory. These analyses again suggest the conclusion that the international location of production should not, merely as a result of the transnational character of enterprises conducting it, impede the attainment of the pattern indicated by considerations of regional cost minimization. An important objection to some of these further analyses is that they too disregard the interdependence of enterprises in arriving at their conclusions. Typically, the operations of TNCs take place in an oligopolistic context. In such situations it is well established that 112

TNCs AND REGIONAL INTEGRATION

patterns of oligopolistic rivalry among TNCs do affect their investment decisions. Knickerbocker (1973), for instance, found that US TNCs had an imitative pattern of behaviour that implies an exaggerated shift of output towards the national markets entered by imitative rivals, and investment in smaller-scale production facilities in such markets than if no such imitation occurred. In the same way, where TNCs seek to preserve a balance of interest visà-vis their rivals, it appears that their investment in production facilities may not serve to minimize the costs of production. In situations of market interdependence and oligopoly it is difficult if not impossible to make a priori predictions about the impact of TNCs in the context of integration. Predictions of the effects of trade policy on the behaviour of TNCs in oligopolistic situations are notoriously sensitive to the assumptions and a multiplicity of outcomes is possible.

REGIONAL ECONOMIC INTEGRATION AND THE INVESTMENT DECISIONS OF TNCs An alternative approach to analysing the interaction of TNCs and regional economic integration follows the industrial organization orientation of the theory of international production rather than the theory of international trade. This approach focuses directly on how a decision on FDI on the part of a TNC is affected by integration and in this connection it emphasizes the distinctive organizational characteristics of the TNC. A widely accepted model of the theory of international production is Dunning’s (1989) eclectic paradigm. He argues that the propensity of a firm to undertake foreign production depends on the combination of the following three elements: • • •

its ownership-specific advantages in the form of proprietary intangible assets vis-à-vis its actual or potential rivals; its locational advantages in the foreign market; internalization opportunities that derive essentially from market failure.

This is the so-called OLI paradigm. If only two of those advantages are present, the firm will serve the foreign market by means of licensing agreements or by exports rather than by investment in 113

TNCs AND REGIONAL INTEGRATION

foreign countries. Foreign direct investment will be the chosen option only if the firm can exploit all three advantages at the same time. An analysis of the interaction between TNCs and regional economic integration from that perspective calls for consideration of the specific responses of TNCs to the various changes brought about by market integration in particular, and by other relevant forms or aspects of integration, and their significance for integration. The changes brought about by integration can be expected to enhance the locational advantages of production inside the regionally integrated area vis-à-vis third countries for both domestic and foreign TNCs, giving rise to increased FDI and related factor inflows, and to alter the relative competitiveness of countries and regions within the bloc, thus affecting both the intra-bloc location of inward FDI and that of intra-bloc FDI itself. In a pathbreaking attempt to analyse these issues in relation to the EC, Kindleberger (1966) originated the concepts of investment creation and investment diversion to characterize the response of foreign TNCs to regional economic integration. Investment creation is the expanded FDI that is generated by trade diversion. Its modus operandi is that new plants are established inside the common market by outsiders to serve the market that would otherwise be cut off by tariffs and non-tariff barriers. Trade creation on the other hand was seen as providing opportunities for a reorganization of production within the common market to take advantage of scale economies and specialization. Investment diversion is the term used by Kindleberger to describe the effects of trade creation on already established foreign TNCs. Kindleberger’s dichotomy does not capture the full complexity of the relationship between the formation of a customs union and FDI. A more promising approach is to try to link each of the static and dynamic effects of economic integration with the strategic responses of firms, both foreign and domestic, that are engaged in international production. Four types of investment response may be identified (Yannopoulos, 1990; UNCTAD, 1992): (1) defensive import-substituting investment; (2) reorganization investment; (3) rationalized investment; (4) offensive import-substituting investment. Each of these represents different strategic responses to economic integration in terms of its effects on the locational and organizational decisions of firms. Defensive import-substituting investment, sometimes termed bridgehead investment, is a response to the trade diversion effects 114

TNCs AND REGIONAL INTEGRATION

of integration and represents a firm’s attempt to maintain its market share in the bloc by shifting from a trade-based to an investmentbased strategy. Reorganization investment results from pressures generated by the trade creation effects of integration. This encourages a reallocation of activity in accordance with competitive advantage, perhaps in fewer locations. The elimination of trade barriers within a bloc should reduce production costs inside the area, making it a more attractive place for international sourcing by TNCs. This will encourage rationalized investment, that is, investment in response to international differences in production costs. Finally, the effects of integration on efficiency, growth and innovation give rise to offensive import-substituting investment or export-oriented investment to take advantage of new markets. Defensive import-substituting investment or bridgehead investment is essentially trade-replacing. By contrast, rationalized and reorganization investment are likely to encourage trade. Each of the different strategies will have an effect on the level of FDI into the region as well as on the intra-bloc distribution of that investment. This fourfold classification essentially represents an attempt to explain how integration affects the strength of OLI advantages, and hence the investment responses of TNCs when cost and efficiency considerations alone are considered. Evidently, however, the extent and character of a TNC’s investment responses to integration will depend (given the specific form of regional integration) not only on the impact of integration on OLI advantages that rest on efficiency considerations, but significantly on its impact on those that are related to strategic behaviour proper. Such behaviour, motivated by considerations of market power, is reflected in strategies such as strategic pre-emption, rival absorption and the creation of excess capacity (Knickerbocker, 1973; Smith, 1987; Acocella, 1991). On several grounds integration might be expected, a priori, to intensify the presssure to pursue such strategies. Acocella suggested that the upsurge of mergers witnessed in the EC during the 1980s was a reflection of those pressures. The benefits of such behaviour for the bloc are problematical. The concepts and the taxonomies of investment response just outlined are useful ways of making sense of a complex reality and for organizing ex ante or ex post case studies of the effects of integration at the production level. Whether they have any value for prediction or for the ex post empirical measurement of observed FDI effects or, in isolation, for the evaluation of gains, is more 115

TNCs AND REGIONAL INTEGRATION

doubtful. For the approach to have predictive value, for instance, the relevant attributes would have to be specified in operationally measurable terms and the thresholds or values that would trigger particular responses would also have to be specified. As yet, there is no sign of such developments in the studies that employ this approach. Rugman and Verbeke (1990), recognizing the limited value of existing international production approaches, have attempted to provide an alternative matrix that would be useful for prediction. Its application to the US and Canada on the basis of intensive industrial analysis, case studies and surveys is revealing and significant. Nevertheless, the precise analytical and theoretical links between the various categories of the matrix and the industrial surveys are not made explicit. Ultimately, studies employing the international production approach need to refer to cost, demand and production data if they are to throw light on the key variables that are of primary concern to policy makers and indeed to economic theorists, in predicting ex ante the effects of integration or in ex post evaluation.

PUBLIC POLICY AND TNCs The connection between the activities of TNCs and the goals and policies of regional blocs and their member states is interactive and dynamic. It is complicated by the fact that the goals of blocs are concerned not merely with economic efficiency, stability and growth, but also with distributional, political and cultural aims. The justification for the formulation of a TNC-specific policy as a separate and distinct component of industrial or competition policy, if one can be found, must be based primarily upon the impact of TNC activity as such on the efficiency and the character of bloc policies (Hymer, 1970; Hufbauer, 1974; Panic, 1991). Even if justification exists, the feasibility of any such policy would in any case be severely limited by a number of constraining factors. Among these, the economic characteristics of the bloc and its economic weight are evidently important, since they determine its intrinsic attractiveness as a location for trans-border investment and thus, ultimately, the bargaining power of the bloc. Distributional considerations apart, the issue in terms purely of economic policy considerations can be considered in the first place against the two primary economic goals of regional economic 116

TNCs AND REGIONAL INTEGRATION

integration, namely to generate static efficiency gains by rationalization and to promote dynamic efficiency and growth and international competitiveness by improving technological and entrepreneurial capacity. Integration-induced cross-border integration may have both favourable and unfavourable effects on the attainment of these objectives. On the one hand, where the transaction costs of using the market as an exchange mechanism are reduced by intra-firm transactions, or where policy distortions are offset, a more efficient allocation of resources can be expected to result. On the other hand, if corporate integration leads, through investment and divestment and production integration, to a concentration of economic power and the promotion of oligopolistic or anti-competitive behaviour, it may work against those goals. The orthodox neoclassical view of TNCs that underlies early appraisals of the policy intervention issue points simply to the necessity and sufficiency of a vigorous bloc competition policy towards all enterprises whether multinational or not and without discrimination between bloc and foreign TNCs. More recent analyses have likewise for the most part concluded that, to secure the potential ‘unorthodox’ gains from market completion, a strong procompetitive thrust on the part of the bloc’s authorities will suffice, subject only to a theoretical question mark over the optimality of such a policy in terms of its possible adverse effects on the number and variety of goods and services produced. That judgement may be well founded as far as home-based TNCs are concerned. But where cross-border industrial integration is undertaken by foreign corporations, important policy issues may arise that competition policy cannot itself necessarily resolve. In particular, the distribution of the gains from rationalization-induced FDI inflows between group and non-group actors may need to be explicitly addressed if, as is possible, they have implications for the distribution of rents and for the dynamic competitive efficiency of the bloc. This consideration may conceivably provide grounds for a specific policy towards TNCs that would not accord ‘national’ treatment—in order to try to offset the adverse effects of distortions introduced from outside (Dunning and Robson, 1988) that cannot be countered by more suitable means. One aspect that has received some attention is the impact of inward FDI undertaken for the purpose of sourcing domestic technology. Some restriction of investment of this kind may be capable of shifting profits towards bloc corporations, essentially by strategic intervention. Limiting inward investment undertaken for technology sourcing may also 117

TNCs AND REGIONAL INTEGRATION

have a favourable effect on growth (Neven and Siotos, 1993). The ability of TNCs to circumvent through internalization the constraints of macroeconomic, financial or other policies has been perceived as a further ground that may justify specific policy intervention (Panic, 1991). The theories of strategic trade intervention that have been developed during the 1980s point to other possible grounds for adopting bloc policies towards TNCs. Specifically these theories may justify policies to support the operations of regional TNCs in external markets. The underlying motive is a distributional one in the first place, namely to switch rents towards them and away from their foreign competitors (Brander and Spencer, 1985; Dixit and Kyle, 1985), for instance by export subsidies. Such a policy has been discussed in its application to the EC’s aerospace industry and the Airbus by Klepper (1990). Neither it nor intervention to limit technology sourcing falls within the scope of competition policy. It must be said that, even if the initial conditions formally prevailed for exercising strategic intervention for such a purpose in favour of a bloc TNC, the surrounding circumstances in which intervention would be worthwhile and feasible are likely to be very limited if not altogether absent. Four points deserve emphasis: • • •

•

The outcomes are very sensitive to the underlying assumptions. Rent-switching intervention potentially invites retaliation by other jurisdictions, and if that occurs, all parties may suffer. Legal domicile and beneficial ownership do not necessarily coincide; in practice, some of the surpluses of foreign firms will accrue to domestic actors. In the few instances in which the effects of strategic trade policy have been calibrated (Smith, 1987), the gains appear to be relatively small.

On those grounds alone, the theory itself does not appear to provide a solid ground for generalized strategic trade policy intervention on behalf of TNCs in a regional bloc. In any case, even to design an advantageous policy based on these considerations would require information of a kind that is simply not available to the authorities. In the EU, TNCs as such have never been subjected to policy regulation. There was a brief period from 1973 when there were signs that the Commission wished to target the activities of TNCs 118

TNCs AND REGIONAL INTEGRATION

specifically, over and above their subjection to competition policy, but the initiatives were soon abandoned. Since the mid 1970s TNCs have been seen as crucial allies in achieving the goals of EU industrial policy and the Commission has adopted a flexible approach to their behaviour. Thus, although the overall thrust of policy is still designed to protect competition, the authorization of trans-border mergers that are potentially anti-competitive is not ruled out if solid grounds can be made out in terms of their contribution to the industrial policy objectives of the EU.

TNCs AND REGIONAL INTEGRATION IN EUROPE The establishment of the EC and its later enlargement have generated many studies of how the Common Market would affect and has affected the strategies and performance of TNCs, which account for a significant and growing share of the EU’s output, employment and trade. The issue of how, if at all, the EU should attempt to influence their operation has also been much debated from time to time, as just noted. Outside Europe, the Canada—US Free Trade Agreement has also generated studies of the impact of TNCs in that context. One of the earliest aspects to attract empirical investigation was the impact of European integration on the magnitude of FDI inflows from outside the EU and, more specifically, from the US. The principal early studies have been reviewed by Yannopoulos (1990) and in a United Nations study (UNCTAD, 1992). Some of these studies provide strong evidence for the view that integration had a significant positive influence on inflows of FDI from the US during the 1960s. During the past fifteen years, and especially since the end of the last decade, FDI in the EU has continued to expand vigorously. The rise was accounted for primarily by inflows from the US and from Japan, which has become an increasingly important source. Studies of more recent FDI flows and of the impact of ‘1992’ and of the Canada—US Free Trade Agreement on those flows can be found in UNCTAD studies (1990, 1992) and in Balasubramanyam and Greenaway (1992). A rare recent statistical study by Aristotelous and Fountas (1996) provides evidence in favour of a significant single market effect on inflows of FDI from the US and Japan, though the results are not as strong for Japan. The results of the 119

TNCs AND REGIONAL INTEGRATION

statistical studies are often persuasive, but well-known problems of identification arise with all of them, and the issue cannot be regarded as settled. A second aspect to attract attention is the impact of the EU on intra-EU direct investment flows. This subject has been less well researched, partly because of lack of adequate data. One such study by Molle and Morsink (1991), however, found evidence of a positive association between trade intensity, which is taken as an indicator of the degree of economic integration, and intra-EC flows of FDI during the period 1975–83. Apart from statistical and econometric studies of FDI, there have been numerous industrial studies that focus on the impact of regional economic integration on various aspects of TNC activity and, notably, on integration at the production level and on the employment impact of intra-EU FDI. The variety of factors at work with respect to transnational production, the interaction between firm-specific and structural variables and the lack, until very recently, of a substantially integrated internal market or even of a Common Commercial Policy, suggest that a uniform pattern of response on the part of TNCs to European integration is unlikely to be observed. And, in any case, few strong generalizations emerge from the more realistic models of international business or market structure and foreign trade. To the extent that rationalized investment and reorganization investment are stimulated by market completion, however, two effects on industrial structure may be expected. First, the removal of tariffs and barriers to trade and investment may be expected to result in decreased horizontal integration. Firms producing substantially identical products in plants in different countries of the bloc should be stimulated to remove duplication if costs differ or if economies of scale are significant, and to concentrate their activities—by divestment—on fewer countries and plants. Second, the opportunity that integration affords for a better international division of labour may also be expected to stimulate increased vertical integration and perhaps component specialization by TNCs seeking lower costs and scale economies. One of the earliest systematic studies of the impact of integration on the activities of TNCs in the EC was undertaken by Franko (1976). Using data for 1958–71, he analysed the experience of more than eighty of the largest European TNCs. Paradoxically, he found little evidence of rationalization, or even of significant intra-bloc trade, even in the case of those industries such as glass, processed foods, 120

TNCs AND REGIONAL INTEGRATION

industrial chemicals and paints in which economies of scale were large. However, Franko’s analysis disregarded the experience of those TNCs with headquarters outside the original six EC countries, including the numerically important, and in some industries dominant, TNCs from the US. Many of these enterprises operated in industries where technology was an important source of competitive strength. In such firms, by contrast, there has been, since the inception of the EC, a well-documented tendency for European affiliates that had previously been merely truncated replicas of their parent companies to specialize in particular products and processes for all markets in the region and to trade these products across national boundaries (UNCTAD, 1992). The formation of Ford (Europe) in 1967 was one of many notable examples of that kind of investment response to the formation of the EC on the part of TNCs from the US. It was set up specifically to co-ordinate and integrate the company’s motor vehicle production on a European basis and represented a major step towards the globalization of its product and marketing strategies. More recent studies of TNC activity convey a complex picture, but certain points stand out. First, whatever may have been the case prior to 1970, in subsequent years European-headquartered TNCs have sought actively to rationalize their European investment and production, often stimulated by competition from US TNCs and Japanese competition. Their efforts have not always met with success. Rationalization has been particularly noticeable in the European-owned motor vehicle industry, in agricultural machinery and in domestic electronic goods and computers. On the other hand, in a number of industries, and in particular in services, where nontariff barriers operated by member states are significant, rationalization has gone less far. For instance, in pharmaceuticals, at the end of the last decade, it had hardly occurred at all, largely for that reason (Cantwell, 1988); the white-goods market has also remained fragmented (Stopford and Baden Fuller, 1991). In many other industries, also, European producers based in one country often continued to feel obliged to maintain and establish manufacturing presences in other countries of the EU by the acquisition of interests, by mergers and, less commonly, by the construction of green-field plants. Their behaviour was motivated partly by the wish to jump non-tariff barriers such as those identified in Franko’s early study, but partly by specifically strategic behaviour directed against rivals. The major steps that have been taken in recent years towards the completion of the Eurpean internal market have profoundly 121

TNCs AND REGIONAL INTEGRATION

altered the strategic context for TNCs, though somewhat differently for insiders and outsiders. In particular, the full implementation by member states of the measures that have been adopted at EU level will erode, perhaps to vanishing point in respect of European operations, the differences between European TNCs in most industries and their near relation, multi-plant domestic firms. As a result, intra-EU defensive investment can be expected to fall. On the other hand, offensive asset-seeking investment and investment motivated by rationalization can both be expected to increase. A second wave of inward FDI can be observed, this time emanating substantially from Japan, but with some resurgence of investment from the US. This appears to be partly defensive, to counter the perceived trade diversion effects of 1992 (of which in fact there is little evidence), but it is also partly a reaction to the anticipated growth effect of the internal market. Apart from mergers and acquisitions, an important element of the response of TNCs to integration was their resort to cross-border strategic alliances as affording a more flexible, less costly, riskreducing alternative. From the inception of that approach, many alliances were entered into by European TNCs, though often with enterprises outside the EU. How is the completion of the European market likely to affect resort to the strategy of alliances? The Emerson—Cecchini report contended that EU trans-frontier collaborative business activity—already important—would be further stimulated (CEC, 1988a) by market completion. On the other hand, Kay (1991) has argued that, since it was essentially market fragmentation that encouraged those ventures in the past, the direct effect of 1992 on joint ventures will be to discourage them. It is too early to reach a firm conclusion on this question.

122

9 FISCAL INTEGRATION

Fiscal integration reflects the role of community public finance and its budget in arrangements for international economic integration. The integration of fiscal functions at community level results in some degree of fiscal union, a situation in which the choice of policy targets and the administration of policy instruments in relation to expenditure and taxation are, within its domain, a matter for the community’s authority. Fiscal integration in this sense is to be distinguished both from fiscal harmonization, which refers to agreement on the manner in which each member state will utilize a particular fiscal instrument over which it retains control, and from fiscal co-ordination, which refers to essentially voluntary alignments of national fiscal measures which cannot be enforced by the community authority. This chapter deals mainly with fiscal integration. The central issues that arise in fiscal harmonization are discussed in Chapter 8. Factors that affect the rational choice among integration, harmonization and co-ordination in the fiscal field are alluded to throughout these two chapters.

ECONOMIC EFFICIENCY AND THE ASSIGNMENT OF FUNCTIONS In contrast to the analysis of customs unions, common markets and monetary unions, there is no single well-demarcated case to address in the case of fiscal integration. A spectrum of workable states of regional economic integration can be envisaged that ranges from simple customs unions by way of common markets up to more advanced forms that approach the status of a full economic union. In each state the role of 123

FISCAL INTEGRATION

the community budget will reflect and encapsulate the deliberate political choices that have been made about the transfer of authority from the member states to the community over a range of policies that involve expenditure and its financing. Although policy integration in relation to these matters ultimately rests on a political decision, the pursuit of efficiency gains is central to all dimensions of economic integration initiatives. It is therefore important to elucidate the efficiency considerations that bear on the assignment of jurisdiction over policies from the member states to the community. A body of economic analysis has developed over the last four decades in terms of which the optimal assignment of public functions in a multi-governmental framework can be analysed. Since many functions and policies require expenditure and finance, this analysis was initially brought to bear on the issue of fiscal integration. But more recently similar efficiency considerations have been applied to the assignment of regulatory functions to the community, although such functions need have no significant budgetary implications. A convincing analysis of policy integration in relation to regulatory or other functions would have to encompass not only the orthodox efficiency considerations but a number of other dimensions that have relevance to political decisions. Among these are the considerations introduced by public choice theory, where competition among jurisdictions is seen as reducing the potential for exploitation by jurisdictions and interest groups. In the end, the appropriate assignment of responsibilities between a bloc’s authority and its member states must reflect a trade-off between the benefits of integration and its costs. In order to address the issues that bear on the assignment of policies from member states to the community, and their budgetary implications in these terms, it is useful to recall the public functions that underlie national budgets. In both unitary states and federations the national budget is a crucial instrument of national economic management whose size and structure reflect the nation’s political, social and economic objectives. The budget affects an economy in a variety of ways that in practice cannot be separated, but for analytical purposes it is convenient to distinguish three distinct impacts that are bound up with three separable objectives of public policy. These concern: (1) the allocation of resources; (2) the redistribution of incomes; and (3) stabilization. 124

FISCAL INTEGRATION

The allocation function of a budget has to do with the provision or encouragement through the budget of goods and services. In market economies the presumption is that these should be limited to ‘public goods’. Public goods are goods or services not provided effectively by the market, either because of the characteristics of the good itself or because of imperfections in the market mechanism deriving from externalities and spillovers. The redistribution function is concerned with the use of tax and expenditure policies to alter the regional or personal distribution of incomes that is brought about by the market. The stabilization function of the budget is concerned with its use to achieve the macroeconomic objective of stability, that is, to minimize the deviation of actual output from potential output. In international economic groupings, the same three functions must either be left in the hands of the member states that were already performing them prior to integration, or they must be assigned wholly or partly to institutions of the community, together with any additional functions, including regulatory ones, that the creation of the community makes desirable. The principles put forward in the literature to determine the efficient assignment of responsibilities amongst different tiers of authority typically appeal to one or more of three primary criteria, namely: (1) cross-border spillovers; (2) economies of scale; (3) political homogeneity (Musgrave, 1969; Oates, 1972, 1977; Forte, 1977; CEC, 1993b). The general bearing of the three indicated criteria on the assignment of responsibilities for policies, and so on fiscal integration, can be stated very simply. The cross-border spillover criterion is relevant whenever public policies in one jurisdiction necessarily have effects that significantly affect others. If these are ignored, the outcome is likely to be suboptimal for the group. Such effects constitute the main prima facie efficiency ground for assigning the activity in question, or its regulation, to a higher-tier authority that can internalize the major costs and benefits of the public policy in question. If, in addition, significant technological or pecuniary economies of scale are generated when certain public policy activities are jointly conducted, there may be a prima facie case, on efficiency grounds, for assigning those activities to an authority or tier of government that can pursue them on a sufficiently large scale to secure the benefits of reduced costs for the group. Evidently the indicated assignments would be conclusive only 125

FISCAL INTEGRATION

if the gains generated from overcoming the effects of market failure by the centralization of functions were not offset by one or more of the following: increased administrative or compliance costs; an inferior quality of policy at the upper tier by comparison with what would be provided under decentralized provision; the possibility that centralization would encourage perverse behaviour at the level of member states. Moreover, even if the criteria were to point to gains from policy integration, centralization would strictly be necessary only if policy coordination among member states could not procure the gains. Apart from these supplementary considerations which may weaken or even countermand any case for centralization that might be made out on grounds of spillovers or scale economies, the political dimension itself may provide positive grounds for decentralization. Political homogeneity, in the sense of similarity of preferences or tastes as expressed in shared criteria of political, social and economic choice in relation to the provision of public goods, strengthens the case for assigning particular functions to the broadest geographical jurisdiction, if efficiency justifications based on the significance of cross-border spillovers or scale economies pointed to doing so in the first place. On the other hand, the existence of a diversity of preferences operates in the other direction and is at the core of the orthodox case for supposing that welfare will be maximized by decentralization in the provision of public goods and the performing of public functions. In their application to the functions and objectives of budgetary policy, the general intergovernmental assignment that is proposed for nation states by reference to these criteria as they find expression in the theory of economic or fiscal federalism (Oates, 1972, 1977) can be summed up in this way. The resource allocation function should be shared amongst upper and lower tiers of government according to the particular public goods characteristics of the services to be provided and the homogeneity or diversity of the preferences for them, whereas both the redistribution function and the stabilization function should be performed at the highest level. The theory of fiscal federalism also analyses the vertical assignment of revenue instruments. In relation to taxation, it puts forward a basic economic rationale for integrating tax jurisdictions at the highest level that similarly derives from considerations of efficiency in the presence of spillovers. This centralizing prescription is reinforced by distributional aims, to the extent that only upper126

FISCAL INTEGRATION

level authorities can operate efficiently in relation to distributional objectives, but it is offset by political factors, since the right to impose taxes is so closely linked with sovereignty. The economic efficiency arguments that point to fiscal integration in taxation are three in number. (1) If there are cross-border externalities, as for instance where the tax base is mobile across jurisdictions, integration may be justified to avoid tax competition among states. Otherwise this may lead both to a lower level of public goods provision than is optimal and to a distortion of the location decisions of business enterprise. (2) There is an argument for fiscal integration if there are taxes whose base cannot be reasonably defined at state level. This is the case with customs duties, whose attribution in a customs union is essentially arbitrary. (3) A third efficency argument that may point to integration is that the integrated collection of taxes may lead to reduced costs of collection due to economies of scale.

THE RELEVANCE OF FISCAL FEDERALISM TO REGIONAL ECONOMIC INTEGRATION There are a number of serious objections to employing the foregoing analysis to provide guidelines for the assignment of functions and so of fiscal integration in international economic groupings. Most obviously, although the EU has assumed, notably in the Single European Act of 1986 and the Treaty on European Union of 1992, a number of the characteristics of federation (and in the view of some has become a federation in all but name), its situation is so far unique. The authorities of other regional economic groupings cannot qualify to be regarded as governments. The limitation that this represents on the application of the literature of fiscal federalism to the issue of fiscal integration in regional groupings is particularly evident in relation to the analysis of tax assignment, since all contemporary groupings depend on contributions from member states for the finance of their activities rather than on the exercise of tax powers that are independently possessed by the authority. An additional limitation of assignment analysis in terms of its relevance to regional economic integration is that the theory is essentially framed in static terms, in the sense that the political regime and its criteria are both taken as given. 127

FISCAL INTEGRATION

Yet all present-day economic communities are in active evolution. To the extent that this is a widespread, if not an essential characteristic, regional integration has to be viewed as a process and not merely as a state of affairs. The orthodox analysis of fiscal federalism throws little light on the dynamics of integration, yet nowhere are those dynamics more important than in relation to a consideration of community budgetary issues which encapsulate and reflect the whole political economy of integration. The public choice perspective is more illuminating from that point of view (Vaubel, 1994). There is a further important more general limitation to the orthodox assignment analysis that has already been alluded to. Even if there should be significant net efficiency gains from the assignment of certain functions to the community, that is not the only possible policy response to the existence of cross-border spillovers. Policy integration or centralization in a narrow sense will strictly only be necessary if the alternatives of policy coordination or harmonization amongst member states would be unable to deal adequately with the problem. The subsidiarity principle laid down by the Maastricht Treaty specifically requires not only that there should be a convincing rationale for any further extension of the competence of the EU, but also that a co-operative solution could not resolve the common problems. Co-operative solutions to crucial cross-border problems of spillover are often feasible, and have already, in the case of the EU, been successfully implemented in a number of areas. What are the circumstances in which agreements on policy coordination or harmonization would be adequate to deal with the problem of cross-border spillovers? It is arguable that the main gains from assigning a policy area with large cross-border spillovers to a community arise from the greater credibility that centralized policy may have in certain circumstances. Where it is difficult to monitor compliance with an agreement to co-ordinate or harmonize policy because it requires more information than individual parties to the agreement are likely to possess, member states may not have confidence that other signatories will comply, and compliance is thus likely to become an unattractive strategy. On this view, the key issues in deciding on the assignment of functions to a community centre on information and monitoring. The functions which a community needs to exercise are those where two conditions are satisfied: there are significant crossborder spillovers; and compliance with harmonization agreements 128

FISCAL INTEGRATION

cannot easily be monitored by individual member states (Gatsios and Seabright, 1989). It is clear from this brief survey that the primary assignment criteria of fiscal federalism are static, only partly operational, and in any case need to be supplemented by other efficiency and political considerations such as those mentioned above. Nevertheless, they remain within their limits analytically instructive categories that provide an important ingredient of any specific policy analysis of policy integration. The following three sections consider in more detail their implications for fiscal integration in the specific context of regional economic groupings.

The allocation function In the member states themselves that belong to present-day economic communities, a large proportion of public expenditure is allocated to the provision of such public goods as defence and social services, notably education, health and in some cases social security. In the case of defence, both economies of scale and spillover effects are without question highly significant. So far, however, without exception, the operation of political considerations has excluded the conduct of defence on an integrated basis. In the provision of social services, significant economies of scale are fewer, although modest spillover effects are not uncommon. But an overriding consideration with respect to integrated provision in this sphere of activity is that in most economic communities there is a marked lack of homogeneity of preferences, traditions and attitudes. To that extent the case for the integration of social services and policies cannot be strong. It is also necessary to consider the substantial range of regulatory public functions—including some that will be called into existence by the establishment of a bloc—that concern such matters as the implementation of technical, environmental, safety and health standards, issues of tax harmonization, the operation of the common market and customs union, state aid to industry and external trade policy issues and negotiations. For all these categories a strong case can be made out by appeal to spillovers, scale economies, and the need to make the market effective, for the exercise of some degree of centralized community regulation. Most of the activities undertaken by the EU that affect resource allocation in fact take this form rather than that of exhaustive expenditure. In the last few 129

FISCAL INTEGRATION

years there has been a growing debate in the EU on the degree of centralization that is appropriate on efficiency and other grounds in respect of such regulatory functions (Begg, 1996). But, even if the arguments for centralization are accepted in particular spheres, it will not usually call for any substantial public expenditure or public ownership, and so no substantial fiscal integration wil be required. In short, in terms of the allocation function, the application of efficiency criteria suggests that the strictly necessary and appropriate degree of fiscal integration for economic communities composed of market economies would appear to be modest. This indeed is what is found in the EU and in other blocs, where levels of ‘community’ expenditure represent only a minute fraction of the expenditures of member states and a still smaller fraction of their national incomes. Such an outcome is not formally inconsistent with the rather general policy stance of fiscal federalism with respect to the provision of public goods. When it comes to the indicated assignment of fiscal federalism in relation to the other two governmental functions, however, the arguments for policy integration appear on the face of it to be much stronger. Are they, however, appropriate or relevant in the case of regional economic integration?

The redistribution function In unitary states, large-scale fiscal redistribution takes place from the centre to the regions primarily and automatically as a result of the operation of policies of personal taxation and public expenditure affecting individuals that are nation-wide in operation but have diverse regional impacts in the face of the varying geographical incidence of prosperity. In federations too, a substantial measure of redistribution takes place from the federal level to the states, but in that case mainly as a result of specific or general intergovernmental grants. In its spatial aspects, fiscal redistribution serves several important and interrelated purposes: • •

It contributes to a more equitable spatial distribution of the burden of macroeconomic shocks. It effectively compensates for the inability of states or regions 130

FISCAL INTEGRATION

• •

to conduct their own trade and exchange rate policies for the purpose of mitigating their regional problems. It helps to promote more uniform and perhaps convergent economic standards and performance in the various regions. It helps to discourage unacceptably large inter-state or interregional migration flows.

In relation to the redistribution function, the analysis of fiscal federalism points unequivocally to an assignment on efficiency grounds in which the primary responsibility for its performance rests with the central or highest tier of government. This assignment finds its rationale chiefly in the high degree of factor mobility that is taken to characterize most unions and federations. A high degree of factor mobility will imply significant leakages and fiscal spillovers that will establish narrow practical limits on the effective jurisdiction of lower-tier governments with respect to distributional matters. In particular, high factor mobility can be expected to limit severely the capacity of lower-level governments to influence the personal distribution of incomes, the economic development of their region, or their standards of social and economic provision, by adopting tax and expenditure policies that are significantly different from those of other member states or of other regions. In present-day economic groupings, this justification for assigning the redistributive function solely or primarily to the community level would carry much less weight, for the following reasons. First, even when, as in a common market, restrictions on factor mobility are formally absent, the effective mobility of persons (though not necessarily of enterprises) is usually much lower than in unions or federations, as a result of linguistic or cultural factors. The effective jurisdiction of the member states can thus be maintained in the area of personal taxation (though perhaps not in the area of corporate taxation), and it is likely to be eroded by tax competition only in the very long term. Another factor that reinforces the ability of member states in a community to influence incomes is that few economic communities have yet instituted a monetary union, so that their members retain their nominal monetary autonomy. Consequently it is open to them to seek to influence their relative real wage costs through exchange rate changes, for the purpose of counteracting any temporary stabilization problem. The absence of monetary union in an economic community is in any case likely to pose severe 131

FISCAL INTEGRATION

practical difficulties for the centralized imposition of taxes. To that extent, fiscal integration and monetary integration have to go hand in hand. Despite the questionable relevance to economic groupings of the principal arguments from fiscal federalism for assigning the redistributive function to the top tier of authority, distributional objectives clearly cannot be ignored in economic blocs, and their pursuit may point to the need for action at community level. The central reason for supposing this to be the case is that the cohesion of a bloc is unlikely to be sustained unless each member state perceives itself to have an equitable stake in its operations. Sustainability is thus bound up with success in containing distributional issues. The federalist analysis assumes that the top tier of authority actually possesses supranational powers that enable it to tax individual citizens or enterprises directly. This is not the case in any present-day economic community, where in effect community activities must be financed by submitting demands to the member states. To the extent that fiscal federalism can usefully be applied to international integration, and in particular to its redistributive aspects, it is thus the characteristics of a confederal rather than a federal regime that are primarily relevant. Such a regime virtually excludes the possibility of significant and deliberate redistribution other than by common consent. In purely fiscal terms, therefore, so long as this situation prevails, the redistribution function in economic communities can be exercised only through the impact of agreed expenditure policies, and ultimately through the agreed criteria for determining the budgetary contributions of member states towards their financing. There are two well-established alternative fiscal principles in the light of which such contributions to a community budget may be determined, namely the benefit principle and the principle of ability to pay. Their distributional implications are very different. The benefit principle amounts to the proposition that those who reap the benefits of public expenditure should contribute correspondingly to its costs. One interpretation of this would be that each country should as far as possible get back in the form of attributable expenditure what it contributes to finance the common policies of the bloc. In the context of the EU budget debate that is briefly discussed below, this is the notion of juste retour. It carries the implication that the budget itself should be 132

FISCAL INTEGRATION

neutral in its impact on inter-country income distribution. If it were adopted, budgetary transfers could not be used on a continuing basis to promote the convergence of living standards amongst the member states or for any long-run redistributive purposes whatsoever, although short-term spatial redistribution on the basis of neutral long-term effects would not necessarily be ruled out. The contributory principle that underlies most national fiscal systems is different, and rests instead on a concept of equity that has horizontal and vertical aspects. Horizontal equity is usually taken to mean that equally situated individuals should be taxed equally, while vertical equity requires that differently situated individuals should bear different tax burdens. These notions have been widely interpreted to justify the use of the criterion of ability to pay as a basis for distributing tax burdens, and this in turn is usually taken to imply the need for a progressive system of taxation. Applied to arrangements for international economic integration at the level of nations, the adoption of the principle would require that net budgetary contributions should broadly reflect differences in the ability of different member states to pay. Of these two principles, the first has in its favour two important advantages: it is unambiguous and it is substantially operational. If it were to be adopted as a criterion for determining inter-state financial contributions to a community budget, it would be illogical to confine its application to the budget if there should be other important transfers directly attributable to the community’s policies that arose outside the formal framework of its budget. A country might be a net contributor to a community budget but might nevertheless derive economic benefits—or incur costs through other channels, notably through the impact of intra-community free trade on its income, employment and so on its public revenues. Thus, even if a primary criterion of contribution were accepted to be that of juste retour, it would not be possible to determine the effective revenue contribution of a member state without considering what its public revenues would have been in the absence of the policies reflected in the budget. Nevertheless, concentration on the purely formal incidence of financial transfers or budgetary contributions is understandable because such transfers are transparent and they are for the most part capable of being attributed to member states, whereas the accurate quantification and attribution of other transfers are generally difficult and sometimes impossible. 133

FISCAL INTEGRATION

The alternative contributory principle, which emphasizes ability to pay in relation to contribution and need in relation to expenditure, would evidently be more appropriate for a community that accepts that balanced development or even economic convergence should be one of the objectives of its operations. The crux of the matter would then be the translation of the principle into operational terms. This can result only from a political process through which acceptable guidelines on the degree and pace of its attainment can be evolved.

The stabilization function Discussion of the assignment of responsibility for stabilization policy in an economic community must acknowledge that there is not full agreement on (1) whether in principle macroeconomic policies can have real effects on output and employment; (2) the practical feasibility of public intervention for stabilization purposes, having regard to problems of identification and timelags; and (3) the relative role of monetary and fiscal policy. The following discussion assumes that there is no monetary union but that member states have adopted a system of fixed but adjustable exchange rates in order to facilitate market integration. Monetary policy is thus subordinated to exchange rate policy and cannot be used, except to a very limited extent, as an instrument of domestic stabilization policy. Current macroeconomic analysis broadly accepts the natural rate of unemployment hypothesis that is discussed in Chapter 11. Against that background, the case for budgetary intervention within a single country for the purpose of stabilization rests on the failure of labour markets to clear in the face of shocks, either through wage flexibility or by migration. If there are such rigidities, in particular in the mechanisms of the labour market that hinder rapid adjustment, there will be a potentially useful role for fiscal policy intervention if an economy experiences an increase in unemployment over its natural rate as a result of a negative shock whose source and nature are correctly identified by the authorities and appropriately responded to. Essentially the role of policy in that context would be threefold. (1) In the context of permanent shocks, it could help to mitigate the effects of a fundamental adjustment in the real exchange rate and real wages that would be required and would eventually occur, but perhaps not before long periods of substantial unemployment had been experienced, 134

FISCAL INTEGRATION

with the welfare losses that would imply. (2) The natural rate of unemployment itself may be adversely affected over time if the work force experiences protracted unemployment. Policy intervention can then be justified to avoid prolonged periods of unemployment above the natural rate that would have adverse long-term consequences. (3) Where the shocks are temporary, fiscal intervention to smooth out fluctuations can lessen the need for painful labour market adjustments in the shape of real wage movements and migration that will not be needed in the long run. In practice it is difficult to determine whether a shock is permanent or temporary. Any intervention would have to be carefully designed so as not to hinder needed fundamental adjustments. This is a matter that has received much attention in the recent literature (Goodhart and Smith, 1993). Assuming that stabilization through fiscal policy is potentially useful and practically feasible, the key question in the present context is whether a case can be made out for undertaking it at the community level rather than at the level of the member states. Two reasons are commonly put forward in favour of performing the stabilization role at community level. These are bound up with the following factors: (1) the spillover effects of national policies; (2) the advantages in that context of sharing the risks of random adverse shocks. The main reason for arguing that stabilization policy should be conducted at the community level derives from the existence of significant cross-border spillover effects that affect other member states. To illustrate their implications, suppose that, in the face of a community-wide shock, one member state seeks to mitigate the effects by means of a debt-financed increase in public expenditure. Some part of the additional expenditure will be spent on imports and will thus leak. Leakage will reduce the impact on real income of any given expenditure, part of which will alleviate the shock for other states. At the same time, the state in question will have to bear the full burden of the debt in the shape of the higher future taxes required to service it. It is therefore argued that the more the benefits of stabilization spill over into other countries, the less likely it is that states will feel able to undertake as much action for stabilization as would be optimal if the externalities were internalized by policy centralization. Although it is true that, if the shocks were common, all countries would have a similar incentive to act, the problem is that no one country can rely with certainty on the positive intervention of the other member states that would, 135

FISCAL INTEGRATION

by generating reciprocal spillovers, largely eliminate the problem of insufficient stabilization. The practical importance of such policy externalities will depend on the degree of integration of the economies in question, which is partly a function of their size. Although small countries may cover too small an area and their economies may be too open to make it feasible for them to adopt stabilization policies effectively for spillover reasons, such would not necessarily be the case for larger states in a regional grouping. There is in any case likely to be considerable variation in the size of bilateral spillovers in a multi-country community as between core and peripheral countries. The second main reason for assigning stabilization policy to the community is the insurance principle (Goodhart and Smith, 1993). If adverse shocks affect countries in an economic community randomly, then if each country has to finance its own stabilization policies by borrowing on the market, the country that has bad luck will be in a worse position in the sense that it has to incur more public debt and will engage in less stabilization than if a community financial mechanism of a non-redistributive nature were instituted. In such a case, a suitable community stabilization scheme could advantageously provide a degree of mutual insurance against the occurrence of adverse country-specific shocks that were macroeconomically significant and help to move the amount of stabilization undertaken at the national level nearer to the optimum. If the benefits were related to changes in real income or employment over time in the country in question, or to the relative severity of such changes vis-à-vis the rest of the community, substantial long-run redistribution across countries could be avoided, which might help to avert political obstacles. At the same time, the possibility of such a policy providing an incentive to perverse behaviour on the part of potential beneficiaries would also be minimized. To return to the problem of common shocks in the face of spillovers from national policies, it should be noted that optimizing the amount of stabilization undertaken does not necessarily require the centralization of the stabilization function. In principle, the problems could be addressed through policy co-ordination aimed at influencing the magnitude and timing of the fiscal policy interventions of member states. Policy co-ordination faces other difficulties, however, not the least of which is that, even if it is based upon rules, and still more so if it is discretionary, it requires 136

FISCAL INTEGRATION

continual intra-community negotiation, and that process takes time. The time lags inherent in the decision-taking process at community level, coupled with others that are in any case inherent in the response of households and enterprises to experienced changes in policy variables, may severely limit the capacity of such an approach to smooth out fluctuations in output and employment. Even at the purely national level, a discretionary stabilization policy might, for such reasons, exacerbate rather than reduce instability. Within a bloc, the process of achieving co-ordination by perhaps protracted intra-community negotiation is likely to mean that some of the practical difficulties in the way of timely discretionary intervention may be even more intractable than they would be at the national level. Finally, quite apart from the problem presented by time lags, it cannot be concluded that either approach to stabilization, namely the centralization of the stabilization function or the co-ordination of the fiscal policies of member states would per se necessarily contribute to the overall stability of a regional bloc. Apart from the uncertainty that besets both forecasts of future economic conditions and estimates of key parameters of the system, there may be uncertainty about the policy model that is appropriate. This is an issue on which countries may take different views. This itself would make the task of agreeing on policy co-ordination more difficult. If the difficulty were overcome, but the jointly agreed policies were inappropriate, co-ordination could result in the exaggeration of any inherent instability. In conclusion, certain practical implications of undertaking direct fiscal action for stabilization purposes through the budget of an economic community may be briefly noted. There are three important related issues to consider: • • •

the scale of the budgetary operations that would be required to achieve adequate stabilization impact; the specific expenditure and revenue implications of the operations; the financing and management of any ensuing debt.

The first of these problems, the critical scale required for effective macroeconomic intervention at community level, would depend on the size of the shocks and their nature—symmetrical or asymmetrical—to which the community was exposed. If the shocks were large and symmetrical, then unless decisions on fiscal 137

FISCAL INTEGRATION

integration had already led to the adoption of community-level policies that would provide an automatic stabilizing effect, or could facilitate a substantial and readily rephasable programme of community revenue and expenditure changes, the scope for significant stabilization action would be limited. That consideration might give weight to arguments for assigning additional tax and expenditure functions to the community level in order to provide it with enough scope and flexibility for performing a community-wide stabilization function without incurring huge efficiency costs. Asymmetrical shocks on the other hand can be expected to diminish in relative importance if economies converge in their economic structures as a result of integration, but the analytical and empirical indications of integration from this point of view are somewhat ambiguous. But in any case, if direct fiscal intervention for pure stabilization at community level were to be limited to providing assistance to member states facing asymmetrical shocks, leaving the overall budgetary stance of the community to be influenced by the coordination of the fiscal policies of member states, the budgetary cost of stabilization would be substantially reduced. The operation of a community-wide fiscal stabilization policy might from time to time require debt to be incurred. In the absence of monetary integration, a community could be authorized to finance stabilization initiatives by issuing its own bonds on the separate national capital markets of the member states. However, since that would necessarily interfere with national control over monetary conditions, it is unlikely to be feasible except within narrowly prescribed limits. In the absence of monetary integration, a community’s borrowing capacity to finance major stabilization initiatives must ultimately depend on the goodwill and the policies of the national monetary authorities. Consequently, the borrowing and debt management policy needed to operate a community-level fiscal stabilization policy of the order required to deal with large symmetrical shocks could be conducted only in conjunction with arrangements for effective monetary co-operation with the member states or on the basis of monetary integration. An insurance scheme to deal with asymmetrical shocks might escape these problems if it could be financed through the institution of a budgetary reserve, as is claimed to be feasible in the case of the EU (CEC, 1993a). In the light of this discussion of the economics of fiscal integration we turn next to an outline of the EU budget and to a brief review 138

FISCAL INTEGRATION

of how some of the issues that have been identified in this section have made themselves felt.

THE BUDGET OF THE EUROPEAN UNION The budget of the EU differs in several key respects from the national budgets of its member states: • • • •

•

Its size is minuscule in relation to the EU’s economy. The EU has no power to levy taxes: the right to tax is the exclusive preserve of member states. Under the Treaty of Rome, revenue and expenditure must be in balance. The budget does not serve a significant planned redistributive function, although de facto it redistributes resources amongst its members in a somewhat haphazard way. The budget has no stabilization role.

At the present stage of the EU’s development, its budget is to be viewed simply as the instrument for financing a range of policies of varying importance that the EU has agreed to establish, notably the common agricultural policy (CAP), regional, social and cohesion policies that operate through the structural and cohesion funds, overseas aid policies, and certain activities in the fields of industry, technology and research. Many of the EU’s policies are implemented not through the provision of goods and services, or of grants to member states, but through its regulatory activities as, for instance, in the case of competition policy. Those regulatory activities are reflected only in administrative costs, which are relatively small. It should be noted that the general budget of the EU does not provide a complete picture of the financial operations of the EU, since substantial expenditure is undertaken by EU institutions on the basis of finance that falls outside the general budget. Thus, through the European Development Fund (EDF), which is partly financed by separate national contributions, aid is provided to those African, Caribbean and Pacific (ACP) countries that have signed the Lomé conventions. The European Coal and Steel Community (ECSC) has a separate budget. The European Investment Bank (EIB) raises finance on the capital markets to support a range of expenditures in Europe in the field of industry, energy and 139

FISCAL INTEGRATION

infrastructure, and priority investment in the regions, as well as development projects in Third World countries. The Commission itself borrows funds for the purpose of financing investment, for instance in nuclear power.

The development of the Community budget During the first 20 years or so of the EC’s existence, the evolution of the budgetary system was determined primarily by the Community’s need to procure funding for its common policies and by its attempts to endow itself with financial autonomy for that purpose. Another important factor was the struggle for the exercise of power over the budget between the Community’s institutions. The foundations of the EU’s main expenditure categories were laid during the first twenty years. The price guarantee and structural investment funds associated with the CAP were created in 1962, and by 1973 they accounted for more than 80 per cent of the budget. The Social Fund established under the Treaty of Rome was reformed in 1971 and given a strengthened role. In 1975, following the first enlargement of the EC, the European Regional Development Fund (ERDF) was established. From 1958 to 1970 the general budget was financed by a system of member state contributions. In 1970 the system of ‘own resources’ was introduced. Own resources may be defined as tax revenue allocated to the EC and accruing to it automatically. Initially they consisted of customs duties, agricultural levies and the revenue of a contribution from member states of up to 1 per cent of their VAT base. The latter soon became the Community’s principal source of revenue. Own resources are, however, collected largely by member states, which set an effective limit to this source of revenue by jointly fixing on the basis of unanimity the contribution of its most important components and by agreeing periodically to any increases that are sought. Under the provisions of the Treaty of Rome the adoption of the budget fell within the exclusive competence of the Council. Since 1970, however, the European Parliament’s budgetary powers and its influence over the budgetary process have been considerably strengthened, first as a result of the Amending Treaty of 1970, which provided that the EC’s operations should be financed by ‘own resources’, and second by the 1975 Brussels Treaty. Since then the Council and Parliament have formally constituted the two arms of 140

FISCAL INTEGRATION

the budgetary authority. The Parliament has the final say on some expenditure items, excluding ‘compulsory’ expenditure, such as agricultural price support payments, which is required to meet the EC’s legal obligations. It can also reject the budget altogether. Ultimately, however, the budgetary process continues to turn on negotiations carried on among the governments of member states at the level of the Council of Ministers or the European Council on the basis of proposals from the Commission. Parliament’s role remains a modest one. The decisions of the 1985 Intergovernmental Conference on the Treaty and the resulting Single European Act did nothing to meet Parliament’s demands for further budgetary powers. Towards the end of the 1970s the Community’s budgetary arrangements entered upon a prolonged period of conflict and difficulty. Clashes between the two arms of the budgetary authority occurred as Parliament sought to exploit its new powers. A serious conflict also arose over distributive equity in relation to the specific position of the UK in respect of its budgetary contributions. In addition, a growing imbalance emerged between revenue requirements and expenditure needs as a result of the combined effect of the erosion of revenues from own resources and built-in pressures producing strong rises in expenditures. Revenue from existing own resources failed to grow adequately owing to the declining yield of customs revenues and agricultural levies and because the VAT base did not expand as quickly as GNP. By contrast, EC expenditure went up steeply as the Social and Regional Funds were reinforced and new programmes were developed. However, the most important reason for the steep rise in expenditure was that the Community was unable to keep CAP expenditure down. In view of the requirement imposed by the Treaty of Rome that expenditures and revenues should balance, a new financial settlement became imperative.

The Fontainebleau settlement of 1984 The two problems of revenue shortfall and budgetary contributions were jointly tackled at the 1984 Fontainebleau summit. On the revenue side, the problem was addressed by raising the VAT resource ceiling to 1.4 per cent. At the same time it was decided that the root of the problem, namely the growth of agricultural price support spending, should be contained by limiting it in future 141

FISCAL INTEGRATION

to less than the rate of growth of the own resource base. This curtailment proved impossible to implement, largely because of a fall in the value of the dollar, which automatically raised the cost of the CAP. But in any case there was no specific agreement on how to impose the necessary budgetary discipline. With respect to the issue of budgetary contributions, which the UK claimed to be regressive and redistributive, it was agreed to redress the perceived imbalance. Between 1980 and 1983 the UK had already negotiated ad hoc refunds from the EC which resulted in its receiving back 75 per cent of its gross financial contribution to the budget. However, these refunds were agreed only after yearly negotiations that threatened to absorb the bulk of the energies of the EC institutions, if not to paralyse their operation. At Fontainebleau a more enduring solution was reached. Although the source of the problem was primarily on the expenditure side, where the UK benefited relatively little from agricultural price support expenditure, which accounted for the bulk of budgetary expenditure, the problem was dealt with by providing for a cut in the UK’s VAT contributions amounting to two-thirds of the difference between its share of VAT payments and its percentage share of EC expenditure. The VAT contributions of other members were correspondingly raised but Germany’s share of the increase was restricted in recognition of its disproportionately heavy contribution. The Fontainebleau settlement appeared to offer the prospect of a few years’ respite from budgetary disputes. By 1986, however, that prospect had receded, and the results of the Community’s failure to address basic issues, and in particular to grasp the nettle of reform of the CAP, had become all too evident. The additional resources from the newly raised VAT ceiling had been completely used up by the increased costs of the CAP, by the abatement of the UK’s contribution, by the refunds to Spain and Portugal that were required under the Treaty of Accession to enable them broadly to enjoy budgetary neutrality in terms of their net contributions during the transitional period, and by the cost of implementing earlier commitments to expenditures in other parts of the budget.

The reform of 1988 The inadequacy of the Fontainebleau settlement to meet pressures on revenue and the need to fund the commitments of the Single European Act in relation to cohesion led to the adoption of a new 142

FISCAL INTEGRATION

budgetary settlement in 1988, known as Delors I, which embodied significant reforms. To deal with the overall revenue constraint it was decided to express financial ceilings not in terms of the VAT take but in terms of Community GNP, so linking the size of the budget with the economic performance of the Community. Annual ceilings were laid down rising to 1.2 per cent of Community GNP by the end of the five-year period in 1992. At the same time, the Community decided to change the contributory basis of its income. First, the VAT contribution was capped at 1.4 per cent of the harmonized VAT base of each member state. Second, this base was taken into account only to the extent that it did not exceed 55 per cent of a country’s GNP. Third, a so-called fourth revenue resource was introduced, based on national product. This was to meet any revenue gap remaining after the changes in the VAT-based contributions. These three changes represent the Community’s first attempt to relate budgetary contributions to relative prosperity. On the expenditure side, the budgetary settlement of 1988 included a commitment to double the size of the structural funds by 1992, representing an increase from ECU 7.8 billion to ECU 13 billion in real terms. This followed the inclusion in the 1986 Single European Act of cohesion as a treaty objective. The goal of cohesion is to reduce disparities between the levels of the various regions. The pressure for this change came mainly from the poorer member states, which feared that the benefits of the single market programme might not be equally shared and that help was needed to enable poorer members to compete successfully. As part of this settlement, two other important changes were made. First, a legally binding agricultural guideline decreed that the annual increase in market support expenditure could not exceed three-quarters of the annual increase in Community GDP, so that the share of agricultural spending in the total budget must decline. Second, a medium-term financial perspective was established which sets expenditure ceilings on broad categories of expenditure reflecting the main political priorities. The ceilings are binding on all three parties to the so-called Interinstitutional Agreement, namely Council, Parliament and Commission, and will eliminate conflict between the two arms of the budgetary authority (since Parliament agrees to respect the annual ceilings when exercising its powers) while retaining needed flexibility. 143

FISCAL INTEGRATION

The reforms of 1993 The 1988 budgetary accord expired after five years, necessitating a new accord on the future of the budget. For this purpose the Commission proposed a new financial perspective, the so-called Delors II package. The outcome, in the shape of the decisions taken at the Edinburgh summit of 1992, provides the perspective for budgetary developments until the end of the century. The coming into effect of the Single European Market and the implementation of the Maastricht Treaty represent important elements of its context. In particular the settlement was conditioned by the reinforced emphasis placed in the Maastricht Treaty on economic and social cohesion and by the specific commitment to correct the regressive nature of the contributory system in the interests of the less prosperous members. In terms of budgetary ceilings the settlement was influenced by the reluctance of member states to concede resources for significant real growth in the face of the severe pressures on their own positions that were imposed by recession. The agreed budget ceiling was to rise to 1.27 per cent by 1999. The basis of funding was again modified. It was decided to cap the VAT rate base to a maximum of 50 per cent and to reduce the call-up rate by stages to 1 per cent. This will increase the relative importance of the fourth, GNP-based, resource and reduce the regressive nature of the contributory system. The new own-resource decision came into effect in 1995. Among the expenditure proposals, its most prominent features include a further substantial increase in EU structural funds and the decision to establish a new Cohesion Fund. Access to the new fund is limited to member states with a per capita GNP less than 90 per cent of the EU average. Currently this group includes Greece, Portugal, Spain and Ireland. Including the Cohesion Fund, the increase in the ceiling on structural expenditure will be of the order of 50 per cent by 1999 in comparison with 1992. The guideline for agricultural expenditure remains in force, but CAP expenditure will be affected independently by the significant decision taken later in May 1992 to move from price support to income support. Certain guaranteed prices are cut, coupled with offsetting conditional direct income payments in compensation. The controversial formula governing the UK rebate which was negotiated at Fontainebleau remained untouched. 144

FISCAL INTEGRATION

The present EU budget Tables 9.1 and 9.2 provide a synopsis of the present EU budget that has emerged from the developments just outlined. For 1996 the general budget adopted by the EU amounted to ECU 82 billion, a sum equivalent to 1.2 per cent of the combined GNPs of member states. A decade earlier, at ECU 36 billion, the budget was equivalent to 1.0 per cent of combined GDP. The EU’s revenue from own resources consists of agricultural duties,1 customs duties, VAT resources and the fourth resource. The agricultural duties are levied mainly to make the CAP effective. Customs duties are levied on products imported from the rest of the world at rates determined by the common external tariff. Customs revenues accounted in 1996 for 16 per cent of total revenue. Their relative importance has declined steadily since the establishment of the EC. VAT resources derive from the application of a uniform rate to each member state’s VAT base. Since systems and rates of VAT vary among member states this precept is applied to a notional common base. Following the 1988 reforms, a member state’s base may not exceed 55 per cent of its GNP. Since that reform, the uniform rate is found by applying a 1.4 per cent rate to the VAT base and deducting the gross compensation paid to the UK. Under the latest own-resources decision, which was ratified in 1995, the VAT rate will be gradually reduced to 1 per cent, and the capping rate to 50 per cent, by 1999. The fourth resource, or additional resource, was created in 1988. This is based on GNP and is derived from the application of a rate to the sum of the GNPs of all the member states. It is a variable, budget-balancing resource for which the call-in rate is calculated in such a way as to cover the shortfall over the amounts yielded by other budget revenues. For 1997 VAT and the fourth resource will together account for more than four-fifths of the EU’s revenue, contributing equally. So far as the revenue side of the budget is concerned, the EU’s aspiration to fund itself autonomously has evidently not so far been succcessful. Unlike customs revenue and agricultural levies, which currently provide about a quarter of total revenue, the VAT and GNP-based so-called own resources are effectively national contributions that balance revenues and expenditures in the general budget. The EU’s budgetary authority is not empowered to introduce any taxes on its own initiative. Any decision to expand 145

Notes a Preliminary draft budget, 1997.

Source: The Community Budget: the Facts in Figures, 1996 edition.

Table 9.1 EU budget revenue, 1988–97

Notes a Preliminary draft budget, 1997. b Excluded from allocated expenditure in calculations of net budgetary contribution.

Source: The Community Budget: the Facts in Figures, 1996 edition.

Table 9.2 EU expenditure, 1988–97

FISCAL INTEGRATION

or modify own resources requires the unanimous consent of member states. The EU’s expenditures are summarized in Table 9.2 for the same period, 1988–97. Over the period 1988–96, EU expenditure (including ECSC, Euratom and EDF) more than doubled in nominal terms, representing an increase in real terms of about 50 per cent. Apart from expenditure on administration and certain other items, including compensation to new member states in respect of budgetary contributions, the bulk of expenditure is on economic and social purposes within the member states. It is referred to as the allocated budget, that is to say, it is attributed to the member states. Except for expenditure on agricultural guarantees, allocated expenditure is mainly of a matching nature, that is, the EU finances a proportion of the cost of certain projects submitted to it by member states or by public or private bodies. The remainder of the budget consists of expenditure outside the EU on overseas aid. This is termed the unallocated budget. The main categories of expenditure for 1996 were as follows: •

•

Expenditure under the CAP in support of farm prices (EAGGF Guarantee Section) alone constituted the largest category of budgetary expenditure in 1996, absorbing nearly 50 per cent of the budget. Agricultural expenditure first entered the EC’s budget in 1965 when the CAP started to come into effect. Its share of the budget rose to more than 80 per cent in 1973 but was subsequently reduced, partly as a result of the development of new policies—such as regional policy—in other sectors and partly as a result of modest reform measures. The second largest group of expenditures consists of those made under the four structural funds, which are concerned largely with the promotion of structural adjustment and growth. The four structural funds are: the European Regional Development Fund (ERDF); the European Social Fund (ESF); the Guidance Section of the Agricultural Fund, EAGGF; and the Financial Instrument for Fisheries Guidance (FIFG). Since the 1988 budgetary settlement the four funds have been focused largely on regional policy objectives. Together with the Cohesion Fund they accounted in 1996 for approximately one-third of the EU budget.

The European Regional Development Fund is the EU’s original vehicle for regional policy. It was set up in 1975 to develop lagging 148

FISCAL INTEGRATION

regions and areas affected by the decline of traditional industry. It co-finances infrastructural development and productive investment in designated regions. In 1996 the ERDF absorbed 13 per cent of the total budget. The European Social Fund was set up under the Treaty of Rome. In 1996 it absorbed 7 per cent of the budget. The fund co-finances vocational training and retraining schemes and job creation projects. Finance is also provided for education, culture, the environment and consumer protection under this head. The Agriculture Fund was established in 1962 in the context of the CAP and is intended to promote structural adjustment in agriculture. In 1996 it acccounted for nearly 5 per cent of total budget expenditure. The Cohesion Fund was established under the Maastricht Treaty and came into operation in 1993. Its purpose is to help the less prosperous member states to prepare for EMU. It accounted for 2.5 per cent of the budget in 1996. •

•

A third category of expenditure is concerned with a range of internal policies. Research, energy, industry and transport are the most important, research alone accounting for the bulk of expenditure at nearly 4 per cent of the budget. Overseas co-operation absorbed nearly 6 per cent of the budget in 1996. It supported technical assistance to the developing countries of the Mediterranean, Asia and Latin America and to the countries of Central and Eastern Europe. Some food aid was provided under this head.

It is not easy to measure the shares of EU expenditure which correspond to allocational objectives as opposed to redistributive objectives. Although the CAP embodied allocational objectives in its initial mandate, spending has come chiefly to serve sectoral redistributive purposes. The CAP redistributes large amounts of income to farmers, primarily from consumers and secondarily from taxpayers. This transfer, by reducing consumption below, and increasing production above, what they would be if world market prices prevailed, involves a deadweight welfare loss. The mean estimate of this substantial loss that emerged from a range of empirical studies undertaken during the 1980s amounted to around 1 per cent of the Community’s GDP (Demakas et al. 1988). The inefficiency of the policy can be measured by the transfer ratio, which is defined as the cost to the economy (consumers and 149

FISCAL INTEGRATION

taxpayers) of increasing farmers’ income by one unit. This ratio was put by Buckwell et al. (1982) at 1.5 for all CAP commodities, but others have put the ratio at more than 3. Most countries were shown to be losers in the studies reviewed, but the distribution of the loss was not uniform between countries. The deadweight costs should be reduced as a result of the move towards income support in 1992. Much of the aid to backward and declining regions through the structural funds is designed to improve supply-side conditions, but its distribution nevertheless has a strong cross-country redistributive element. Leaving aside these two major categories of expenditure, the Commision has estimated (CEC, 1993a) that in 1993 only ECU 3 billion of budgetary expenditure was justified by the promotion of efficient allocation where spillover considerations are important. Of this amount, some two-thirds was accounted for by support for R&D. It should be borne in mind that the principal policy instrument that is available to affect resource allocation at EU level is regulation, rather than expenditure on the provision of goods and services, which the EU can undertake only if there is a clear cross-border spillover dimension. In respect of the pursuit of common allocational objectives, the EU provides essentially a regulatory form of governance. With respect to the second public function, namely redistribution, the EU budget’s general redistributive capacity is weak. It is essentially determined by three factors: a country’s contribution to the EU’s VAT receipts; the share of agriculture in the GDP; and by the country’s share of the interventions of the structural funds which exhibit strong geographical concentration. These three factors largely determine the net redistribution of resources that is brought about through the budget. The structural funds transfer quite significant amounts to some parts of the small and less developed member states. In 1992 the grants received by Ireland, Portugal and Greece each represented more than 2.5 per cent of GDP, and were equivalent to more than 75 per cent of gross fixed capital formation by the public sector (CEC, 1993a, p. 23). With regard to stabilization, the EU’s budget cannot yet play any role in the EU economy as a whole for three reasons: the budget is too small; it is prohibited from running a deficit; owing to the system of multi-annual programming that is now adopted, it in any case lacks the flexibility that would be required. Indeed, the budget 150

FISCAL INTEGRATION

may be said at present to display a procyclical bias, since spending ceilings are geared to GDP. The budget cannot serve an intracommunity stabilization purpose either because of the lack of automatic stabilizers or other regional shock absorbers operating at EU levels.

EU BUDGETARY ISSUES IN THE LONGER TERM The EU budget at present plays a negligible role in the allocation sphere except for its nefarious impact on the agricultural sector; likewise its distributional impact has been insignificant (though, if anything, overall it has been regressive and haphazard), while any stabilization role has been totally absent. Nevertheless, the possibility of assigning a more important role to the budget in the longer term in each of these spheres has been considered from time to time. The impetus to do so was provided initially by aspirations on the part of some of the EU’s leaders to move towards closer economic integration and to a monetary union, and by the simultaneously held belief that the establishment of an enlarged role for the budget would be crucial to this process. Closer economic integration in the shape of a single market will almost certainly intensify and accelerate economic change in ways that cannot be fully predicted. A built-in system of net transfers from whoever is doing well at the moment to whoever is doing badly as an insurance against economic misfortune may therefore be regarded as essential to induce participants to accept the risks. This is the primary relevance of the scale and pattern of the budget to closer integration. Monetary integration would further enhance the need for insurance by reducing the member state’s own ability to deal with its adjustment problems in the face of asymmetrical shocks. The development of the EU budget in the longer term will depend fundamentally on decisions on goals and on the pace at which movement towards those goals should be undertaken. There are, however, three important issues with possibly major budgetary implications which recent reforms and initiatives have left unresolved, and which promise to demand attention in the near future in the context of the current thrust of policy. These are: (1) budgetary equity; (2) the implications of economic and monetary 151

FISCAL INTEGRATION

union; (3) the implications of the enlargement of the EU to include the countries of Central and Eastern Europe. The issue of budgetary equity needs to be addressed in relation not only to the UK adjustment, but also in the context of the prospective accession of the Central European countries. The advent of a single currency may require the EU budget to assume new functions, particularly in relation to stabilization. The enlargement of the EU will have important budgetary implications if the CAP and the EU’s structural policies remain unchanged. To conclude this chapter, these three issues are briefly addressed in turn.

Budgetary equity Budgetary equity has been an issue since 1973, when it became clear that the application of previous budgetary arrangements and spending policies to new members, and in particular to the UK, would give rise to serious inequity. Even after the introduction of the ERDF, which was designed to ameliorate the situation, a situation emerged before long in which net contributions to the budget (that is, the difference between attributable gross contributions less attributable receipts from the allocated budget) did not correspond to either vertical or horizontal equity. In 1980, for example, Germany and the UK were the only net contributors, and to a roughly similar extent, whilst all other member countries were net beneficiaries. The UK rebate and the German adjustment negotiated at Fontainebleau represented an attempt to provide a durable and automatic solution to the problem. The settlement was renewed in the budgetary settlements of 1988 and 1992 and will remain in effect until 2000. Although the main source of the British imbalance was to be found on the expenditure side and in the dominant share of agricultural support expenditure in the EC budget, from which Britain benefited to a relatively small extent, the solution that was adopted, as already noted, dealt with it by cutting the UK’s contribution. The UK’s VAT-based contribution was reduced by twothirds of the difference between the UK’s share of VAT payments and its percentage share of EC expenditure. At the same time, Germany’s share in the cost of the UK correction was restricted, the additional cost of this adjustment falling on the other members. Since that time, the changes introduced in the 1988 and 1992 settlements have moved the EU towards a budget in which 152

FISCAL INTEGRATION

contributions take more account of relative national income, and expenditure through the structural funds is directed more towards the poorer member states. The budget will move still further in that direction with the projected reduced relative contribution from the VAT resource. Other changes, such as German reunification and the accession of Finland, Austria and Sweden, have also affected the outcome. The combined effect of all these factors has been to bring about a marked change in the position with respect to net contributions to the EU budget. As recently as a decade ago, the only substantial net contributors were Germany and the UK. Most other members were roughly in balance or received more than they paid in. Table 9.3, section A, summarizes the greatly altered position for 1995. Today the Netherlands, France and Italy have all become substantial net contributors, together with Sweden and Austria, which acceded in 1995. The only major net recipients are Ireland, Portugal, Greece and Spain. This change greatly affects the political dynamics of budgeting in the EU. Some anomalies appear to have been mitigated. But fairness in the sense of contributions proportional to prosperity has not yet been attained. Among the net contributors, the payments of Germany and the Netherlands appear prima facie high, while those of France and Italy appear disproportionately low. There are several qualifications to bear in mind in using such figures to measure the net transfer of resources through the EU budget from the taxpayers and consumers of the net contributory countries to the beneficiaries of EU policies in the net recipient countries. One of the most important is that the calculation assumes that customs duties and import levies collected at a member state’s borders represent a charge on the taxpayers or consumers of that state. This is not the case where a member state’s products enter or leave the EU through the ports of another member state. In consequence, Germany’s contribution is understated while those of Belgium and the Netherlands are overstated. Export refunds that are attributed to the country where the goods leave the EU pr esent a similar problem of overstatement of benefits. The Commission claims that 25 per cent of revenue and 40 per cent of expenditure are hard to apportion on a national basis (CEC, 1993a). It must be borne in mind, in any case, that the net contribution does not represent the full redistributive effects of the EU’s policies, but only the transfer of resources that takes place through 153

Notes a Figures for actual net contributions are taken from the report of the European Court of Auditors. The receipts of member states refer only to allocated expenditure, made inside the EU (i.e. disregarding external action). b Net contributions with a redistributive budget are illustrative only. The figures assume (1) that gross contributions to the budget are in proportion to relative GDP; (2) that notional receipts of member countries from the EU budget are distributed, taking into account relative income per head, as described in Annexe D to the Padoa—Schioppa report (1987). A redistributive factor of 0.15% is arbitrarily assumed. Franklin (1992) assumed a factor of 0.05. c Rank refers to the ordering in terms of per capita contributions.

Table 9.3 EU net budgetary transfers, 1995

FISCAL INTEGRATION

the budget. Important non-budgetary transfers are automatically produced as an integral consequence of the impact of the CAP and the EU’s common external tariff on the prices received by exporters in intra-EU trade. Their effect, which is significant only for agricultural products, is to transfer resources from net importing countries to net exporters through the higher prices that are paid directly by consumers to producers as a result of trade diversion. This effect may operate to reinforce or partly offset the results shown in the table. With these qualifications, whether or not the EU wishes to redistribute income towards poorer member states, the issue of horizontal budgetary equity must be borne in mind. It is not clear either that the present system is wholly acceptable or that the changes now in train as a result of the 1993 reforms can themselves be relied upon to produce an outcome that will be more acceptable in the longer term and will remain so and can accommodate enlargement. If not, regularly recurring budgetary controversies can be anticipated. What are the main options for reform? One possibility would be to change further the character of both revenue and expenditure policies systematically in order to bring about the desired results in terms of equity. This could be done, for example, on the income side by eliminating the third resource, and perhaps making the fourth resource progressive by varying the rate that is applied to countries according to their prosperity. Expenditure changes could likewise be made that would work in the same direction by, for example, extending the principle introduced by the Cohesion Fund of limiting access to poorer states. In the past the Community has placed great emphasis on the use of policies, rather than fiscal adjustments, to address emergent inequities, and the inception of the Regional Fund in 1975 was itself just such a response. There are clear efficiency dangers, however, in utilizing policies for redistributive purposes, from which the Regional Fund operations have not been free. If the purpose of certain EU policies is mainly redistributive, more direct means could be chosen which would probably be more efficient. Another idea that has been advanced from time to time since 1976 is that policies, and in particular expenditure policies, should be left unaffected, but that a corrective financial mechanism should be established that would be triggered automatically and applied uniformly in the event of an emergent equity problem in relation to net budgetary transfers. This proposal was elaborated in the Padoa—Schioppa report (1987) 155

FISCAL INTEGRATION

and has been supported more recently by Franklin (1992). To make such a scheme operational a formula would have to be formulated to give effect to guidelines that would have to be determined through a process of political negotiation. The object would be to produce a systematic and lasting outcome. Once settled, this approach could minimize budgetary conflict. It would also have the great merit of enabling political negotiation over the allocative policies of the EU to be concentrated on efficiency considerations and divorced from equity considerations, which is not the case at present. If a corrective mechanism existed, it would no longer be necessary to invent schemes of doubtful efficacy in order to achieve desired transfers towards poorer member states. Its disadvantage would be that, unless a wide range of outcomes were permitted before compensation came into play, the interest of member states in the conduct of specific policies would be likely to diminish, since the adjustments would be equivalent to unconditional grants. The Reichenbach Study Group has expressed well-founded doubts as to the political feasibility of introducing such a radical change (CEC, 1993a). Instead, it favours the continuation of recent reforms, with a further move in the longer term towards proportionality on the revenue side. If this is to be the goal, the simplest way to achieve it would be by abolishing the VAT resource and replacing it by increased—or even complete—reliance on national contributions on the basis of GDP. Redistribution could then be effected through the expenditure side by way of performance-related grants. Table 9.3 illustrates in section B the pattern of net contributions that would have resulted for 1995 if gross contributions to the budget had been entirely based on relative GDP while the country distribution of receipts reflecting current EU policies remained unchanged. The table also shows in section C for purposes of illustration the approximate pattern of net contributions that would be operative for that year if EU expenditures were designed systematically to favour the less welloff member states according to a particular redistributive factor. If policies were to be left unchanged, a comparison of such figures with actual net contributions could be used to trigger automatic abatements or additional contributions in the following year, on the lines of the safeguard mechanism envisaged in the Padoa— Schioppa report. If there is to be a move towards a more equitable budget, and GDP or GNP is to be used as a measure of relative prosperity, 156

FISCAL INTEGRATION

several implementation issues would have to be resolved. One concerns whether GDPs should be compared in money terms or in purchasing power terms. Income disparities between member states are typically smaller in purchasing power standard (PPS) terms. In principle, the PPS term would provide a superior yardstick in terms of equity, but there is no consensus on how to make the adjustments. For this reason the Commission argues for the use of unadjusted figures.

The implications of monetary union for the EU budget As the moves toward monetary union progress, it is necessary to consider whether there are any important longer-term implications for the evolution of the EU’s budget. There are two main issues. First, can any presumptive link be established between the operation of the monetary union and the development of disparities between member states and between regions of the EU? If so there may be additional implications for the budgetary efforts aimed at cohesion which so far have been justified in terms of the impact of the single market. In other words, are there additional effects to take into account over and above those of the single market, whose potentially disequalizing effects supposedly were adequately addressed in the policy changes adopted at the Edinburgh summit? The second issue turns on whether there is a need for the budget to exercise a stabilization function which, as has been seen, it does not yet perform, and what the implications would be for the size of the budget and budgetary ceilings.

The spatial effect of monetary union The issue turns on whether an adverse link can be established between monetary union and disparities amongst member states and regions. The spatial effects of market integration and its policy implications are discussed in general terms in Chapter 12. There is not much analysis that bears on the specific issue of the incremental effects of monetary union in a spatial context however. What little there is seems to point to the conclusion that, at the micro-level, small economies could be expected to benefit disproportionately from transaction costs savings and from the suppression of exchange 157

FISCAL INTEGRATION

rate variability (Santos, 1993). They would lose seignorage revenues, but that loss could be expected to be offset by reductions in interest rates and, through them, in the burden of their debt. On the basis of these considerations, there seems little reason to suppose that an expansion of the cohesion budget would be called for in the longer run by EMU since, if anything, the handicaps of small economies would be reduced.

Stabilization The remaining issue is whether the monetary union would require the EU to assume a stabilization function and, if so, what its budgetary implications would be. In the past, when this issue has been discussed, the budgetary implications for the EU of exercising a stabilization function have been thought to be immense and to imply an unacceptable degree of fiscal integration. In considering the implications, it is useful first to draw a distinction between (1) the exercise of a stabilization function for the EU as a whole in relation to Union-wide cyclical fluctuations or longer-term shocks, and (2) the need for the EU to assume a stabilization role in relation to shocks affecting specific member states only. The issue of an EC-wide countercyclical stabilization policy was first analysed by the MacDougall Committee (CEC, 1977) when a move to economic and monetary union was seriously under consideration by the Community for the first time. It was then concluded that even if the EC budget of the time were doubled or trebled, massive swings in expenditure and revenue would have been required to exercise any significant leverage. In an analysis undertaken a decade later, the Padoa—Schioppa report (1987) also concluded that, to be of significance, budgetary swings would have to be massive—equal to the total of the EC budget at that time (1 per cent of GNP). The possibility of engineering swings of such magnitude would obviously imply a much larger regular EU budget than at the present time and even then could hardly be implemented without incurring major economic and efficiency costs, so long as the EU budget did not dispose of taxes that directly impinged on personal incomes, since it would otherwise have to rely purely on costly variations in expenditures to exercise its leverage. It is arguable, however, that to attempt to deal with EU-wide stabilization problems in this way would be both inappropriate and unnecessary. It will clearly be necessary for the EU to ensure 158

FISCAL INTEGRATION

a budgetary stance for the Union as a whole that is consistent with the monetary policy that is being conducted in the interests of internal and external balance. The EU intends that this reconciliation shall be effected by national fiscal policy coordination. There are many difficulties that confront the coordination of national policies to which attention has been drawn earlier in this chapter, but it is only if it proves to be unfeasible that major changes in the EU budget on this account alone would need to be contemplated. There is, however, a second aspect of stabilization to consider, namely regional stabilization in circumstances where a member state experiences severe shocks which are specific to it. In the context of monetary union it has been argued that there would be a strong case for the creation of an EU budgetary instrument to assist member states to absorb and adjust to such shocks. The rationale is that the EMU implies the loss of the exchange rate mechanism and independent monetary policy as stabilization instruments. A recent Commission study group on the implications of EMU for EU public finances (CEC, 1993a) supports budgetary intervention to assist member states to adjust to country-specific shocks, but contends that it can be done inexpensively and does not therefore imply a large EU budget. The message of the report is that, contrary to the conventional wisdom, a small budget—of the order of 2 per cent of EU GNP— can sustain EMU, so long as the budget itself is not required to perform an EU-wide stabilization role. The regional stabilization support mechanism which is suggested, following a study by Goodhart and Smith (1993), would take the form of an automatic non-repayable block grant to the government concerned. It would be triggered by changes in an indicator of real activity such as unemployment. If this showed a fall below trend that was significantly worse than the performance of the weighted average for the other countries, the member state would be eligible for support. The degree of stabilization to be achieved and the precise mechanism would have to be settled by political negotiation. What is important is that the proposed mechanism would be highly efficient in the sense that if passed through to the incomes of individuals it would be capable of achieving, at relatively low budgetary cost, a degree of stabilization not dissimilar to that reported in the literature for certain federal states such as the US, in relation to transfers from the federal government to the individual states. For an estimated annual cost of about 0.2 per cent of EU 159

FISCAL INTEGRATION

GNP (or ECU 10–11 billion for the Union of twelve) a full stabilization mechanism that would operate automatically for asymmetrical shocks of all sizes could be set up. A limited stabilization scheme that would operate only in the case of severe shocks beyond a certain threshold would also provide a reasonable level of stabilization but would be even cheaper. The basic reason for the high level of efficiency is that, unlike the automatic regional stabilizers that operate in existing federations, this proposed mechanism is explicitly designed for regional stabilization purposes rather than being a by-product of programmes which have other purposes. If such an approach were to be followed, the degree of stabilization to be achieved and the precise mechanism would have to be settled by political negotiation, and the establishment of such a fund would call for an amendment to the Maastricht Treaty. Up to the present time, no action has been taken on this proposal. The EU appears to be taking the risky line of embarking on monetary union while lacking any instrument through which member states could be assisted, notwithstanding the constraints that are to be imposed on the stabilization that the member states might otherwise implement through their own budgets as a result of the Stability Pact negotiated at the Dublin summit of December 1996. Eichengreen has argued (1997) that most countries will enter EMU right up against the 3 per cent deficit which is permissible and that this will render fiscal policy strongly procyclical. He therefore argues that the provisions of the Maastricht Treaty and of the Stability Pact should be applied not to the actual deficits of member countries but to what is compatible with the nonaccelerating inflation rate of unemployment (NAIRU).2 Procedural and institutional reform might then reasonably be demanded of member states as a quid pro quo following the guidelines put forward by von Hagen and Harden (1994), in order to offset the present bias towards deficits that derives from the operation of national political systems. Further integration in the EU could either weaken or strengthen the case for a regional support mechanism. Increased fiscal harmonization and factor mobility could further limit the ability of member states to undertake stabilization measures. As for the need for such measures, the issue turns largely on whether integration encourages a greater measure of industrial diversification, so that country-specific shocks may be weaker, or whether, on the contrary, increased industrial specialization results. 160

FISCAL INTEGRATION

The budgetary implications of EU enlargement The most likely first new entrants to the EU are the Visegrad group, which includes the Czech Republic, the Slovak Republic, Hungary, Poland and Slovenia. Since the discussion began on enlargement, its budgetary cost has formed an important element of the debate. There is an extensive literature on the matter, which focuses on the potential budgetary contributions of the new entrants on the one hand compared with the cost of providing access to the structural funds and the CAP on the present basis on the other. In 1992 the two least developed member states, Greece and Portugal, received from the structural funds about ECU 200 per capita. This should double by 1999. If similar transfers were made to the Visegrad countries at the turn of the century, it would imply a rise in the structural funds of ECU 26 billion (CEC, 1993a). This estimate has been questioned on the ground that it implies unrealistically high levels of grant absorption, bearing in mind that matching funds have to be provided by recipient governments. Baldwin et al. (1997) suggest that spending of about half that amount might be a more realistic estimate. The cost of extending the CAP to the Visegrad group is more problematical. The purely budgetary costs are linked with the purchase by the EU of food surpluses at guaranteed prices, and their subsidized export. The difficulty is to determine how output in Eastern Europe would rise under the CAP. Estimates have ranged from ECU 4 billion to ECU 37 billion. Baldwin suggests a consensus estimate of ECU 10 billion. The upshot on these bases would be a gross cost of ECU 23 billion, but with budgetary contributions of ECU 4 billion assuming rapid growth (6 per cent) to the end of the century, the net cost would be about ECU 19 billion, which is about one-fifth of what the 1999 budget would be without enlargement. If these costs were borne in proportion to shares of gross budget contributions in 1994, three-quarters would be borne by Germany, France, the UK and Italy, with Germany alone bearing 30 per cent of the cost. If the costs are not affordable, the redistributive policies in question would have to be scaled down. According to Baldwin et al.’s estimates, however, there would be offsetting real income gains to the EU as a whole amounting to about half the net budgetary costs. 161

FISCAL INTEGRATION

NOTES 1

2

Agricultural duties have, in accordance with the GATT agreements, replaced the variable levies charged up to 1995 on imports of agricultural products to implement the CAP. Under this head are also included sugar and isoglucose levies. See Chapter 11 for a discussion of this concept.

162

10 FISCAL HARMONIZATION

International economic integration encompasses a spectrum of arrangements ranging from simple free trade areas to economic and monetary unions. For a full economic union to be feasible, member countries must share similar policy objectives in major fields and similar preferences for public goods. In such circumstances a high degr ee of fiscal integration, and substantially uniform fiscal systems and tax and expenditure policies, will be appropriate. The theory of fiscal harmonization takes a different starting point—namely, forms of integration short of economic union in which diverse public policy objectives and preferences for public goods exist amongst the member states, which also retain responsibility for key policy areas such as stabilization. Its concern is to analyse the significance of differences in jurisdictional principles, and in types and rates of taxation and of expenditures, and the implications of different ways of co-ordinating fiscal systems, in the light of the common or shared objectives of the union and the goals of member states. The bulk of the literature on the subject relates purely to tax harmonization and is not concerned with expenditure harmonization. This is a reflection of policy concerns in as much as expenditure harmonization is not a central issue in the EU1 or in any other contemporary grouping. This chapter reflects that reality and is solely concerned with tax harmonization. The analysis of tax harmonization may be conducted for any form of integration, but a basic distinction must be drawn between customs unions on the one hand, in which products are free to move without tariff barriers, and common markets on the other, where factors and products are both free to move without fiscal or other barriers. For each form of integration, har monization issues may be analysed for different monetary arrangements, the principal alternatives 163

FISCAL HARMONIZATION

being: (1) per fectly flexible exchange rates; (2) fixed exchange rates; (3) monetary union. The analysis of tax harmonization is conducted with the usual range of economic criteria in mind, and in particular, its effects on the goals of the union and of its member states with respect to allocative efficiency and distributional equity. One aspect of the latter, in the shape of the inter-country revenue implications of alternative tax harmonization arrangements, may be perceived to constitute a significant source of gain or loss that needs separate consideration. Ultimately, however, since tax adjustments are normally capable of replacing any revenues that may be lost in the process of tax harmonization, it is the welfare losses and gains associated with tax harmonization that are crucial. These cannot be straightforwardly identified with the extent of revenue gains and losses themselves. It remains to emphasize that, as with market integration itself, the potential efficiency gains to a union from fiscal harmonization do not guarantee that each member will gain, in the absence of compensation. It is also true that any income gains secured from tax harmonization may have a perceived cost to member states in the shape of the constraint that harmonization imposes on the use of indirect taxes to pursue objectives of national policy in the fields of distribution or social policy. However, alternative means of achieving such ends will normally be available, and some of those means may not be significantly inferior, on welfare grounds, to the use of taxation. To that extent, any costs to member states on grounds of distributional or social policy considerations may, after policy adjustments, be very limited. The efficiency criterion has always dominated tax harmonization analysis, reflecting the emphasis of orthodox integration theory from which it springs upon improving the allocation of resources. In that context, tax harmonization analysis has two principal foci, corresponding to the main types of internal taxes. For indirect taxes the problem is to determine the efficiency impact of tax-induced price effects on trade and specialization. In the case of direct taxes the problem is to determine the efficiency impact of tax-induced differences in factor earnings by country or nationality on productive efficiency. This chapter commences by analysing the implications for tax systems if tax-induced distortions in resource allocation are to be avoided. This is the ‘neutrality’ criterion. Its implications are taken up for indirect and direct taxation in turn. Indirect taxation is alone relevant to the customs union case whereas, for a common market, 164

FISCAL HARMONIZATION

tax harmonization issues arise for both forms of taxation. Broadly speaking, if the tax system is not neutral, production and trade will not be optimized, and welfare lossses will be produced. Neutrality may be affected by several different dimensions of taxation: (1) the treatment of cross-border transactions by competing jurisdictions, (2) the computation of the tax base, (3) tax rates and structures, and (4) the tax systems. Considerations other than efficiency— notably the implications of fiscal harmonization for member governments’ ability to make use of fiscal policy in pursuit of domestic stabilization, distributional or structural objectives—are only touched upon. The final section briefly comments on the progress of tax harmonization in the EU and current policy issues.

THE HARMONIZATION OF INDIRECT TAXES The issue of allocative efficiency Even when a customs union has eliminated the impact of all external duties within the union, different systems and rates of internal indirect taxation may influence intra-union trade flows. Such influences may become more significant, and are in any case likely to be more evident, when intra-union tariffs no longer exist. The main efficiency issues that arise in connection with the harmonization of indirect taxes are: •

• •

•

Does the use of different forms of indirect taxation in the various member countries produce a tax-induced misallocation of resources? Does the choice of jurisdictional principle affect the allocational effects of indirect taxes? Do differences among member countries in their relative reliance on direct as compared with indirect taxes give rise to distortions? If tax-induced distortions are to be avoided, is it necessary for countries’ rates of indirect taxation to be equalized?

To simplify the discussion it is assumed that, apart from traditional excise taxes on such commodities as alcohol, tobacco and petroleum products, indirect taxation is levied by means of a value-added tax 165

FISCAL HARMONIZATION

(VAT). This is increasingly being adopted as the preferred form of consumption taxation throughout the world. The first of the four issues mentioned above can therefore be largely neglected. It is further assumed that the tax base itself is harmonized, thus ruling out another possible source of distortions. We may begin the discussion with a consideration of the second issue. This is posed by the fact that, with indirect taxes, a choice exists between two alternative jurisdictional principles, which correspond to the imposition of the tax on a production or on a consumption basis. Under the origin principle, the tax is imposed on the domestic production of goods, whether they are exported or not, but not on imports. Under the destination principle, the same tax is imposed on imported goods as on domestically produced goods destined for consumption by domestic consumers, but domestically produced goods destined for consumption by foreigners are not subject to the tax. Are there any theoretical reasons, stemming from resource allocation considerations, why one principle should be preferred to the other when indirect taxes are applied to internationally traded commodities? The issues may be illuminated with the help of a simple comparative cost model, following the approach of the Tinbergen Committee (ECSC, 1953). Recent statements (Keen, 1993; Genser et al., 1995; Keen and Smith, 1996) are more general, but leave the conclusions of the basic analysis essentially unimpaired. For simplicity, a two-country framework is adopted. The analysis assumes that the initial rate of exchange is the equilibrium rate, that trade is in products only and that exchange rates are perfectly flexible. The model assumes two countries, H and P, that produce two commodities, namely clothes and motor vehicles. Country H has a comparative advantage in clothes and country P in vehicles, but neither country specializes completely because of increasing cost conditions. In the initial situation there are no tariffs or internal indirect taxes affecting trade between the two countries. Country H will thus export clothes and country P will export vehicles. Internal indirect taxes are now introduced, and two principal cases are considered: in the first case each country imposes VAT at a uniform rate on every product and applies the same principle to each product; in the second case a differentiated tax is assumed to be applied in at least one of the countries. It is further assumed that the proceeds of the tax are used to reduce the general income tax from which public revenues are derived, leaving aggregate public revenues and expenditures unchanged. 166

FISCAL HARMONIZATION

Uniform indirect taxes: different levels Assume that country H imposes a perfectly general VAT at a rate of 20 per cent, whereas country P imposes a similar tax at a lower rate of 10 per cent. How will the imposition of these taxes affect equilibrium? If the tax is imposed on the destination principle and factor prices are given, then, since all exports will be exempt from the tax, the relative costs of clothes and of vehicles produced for export will not be affected by the tax. Internally, both domestic production and imports will be subject to taxation at the domestic rate, leaving their relative prices unchanged. Comparative costs will therefore be unaffected. No trade effects will result, and there will be no effect on exchange rates. If the tax is imposed on the origin principle and factor prices are given, all exports will have to pay the tax; but, as it is levied on both products, the relative costs of clothes and vehicles will be unchanged. Comparative costs will again be unaffected. At the initial rate of exchange there will, however, be an increase in the price of country H’s exports relative to foreign goods, which will be unfavourable to its exports, and a reduction in the price of country H’s imports relative to domestic goods, which will be favourable to its imports. These effects will be automatically offset by a tendency for country H’s exchange rate to depreciate relative to that of country P. If the tax is imposed on the destination principle and factor prices are flexible, a fall in factor prices will accompany the increase in indirect taxation and comparative costs will again be unaffected; but for country H, at the initial rate of exchange, the price of exports relative to foreign goods will fall, and the price of imports relative to domestic goods will increase. This will give rise to a tendency for the exchange rate of country H to appreciate relative to that of country P, maintaining trade in balance. If the tax is imposed on the origin principle and factor prices fall with the increase in taxation, leaving prices unchanged in each country, there will be no trade effects and no exchange rate change. In general, the direction and extent of the price adjustment that comes about will depend on the nature of monetary policy and on the degree of price flexibility in factor and product markets. If factor prices are rigid downwards, however, the price effects that 167

FISCAL HARMONIZATION

accompany increases in both origin and destination taxes are likely to be upwards. Several important conclusions may be drawn from this analysis. First, given flexible exchange rates and a perfectly general indirect tax, the choice between the origin and destination principles has no effect on the composition or level of trade, irrespective of the direction of price change. Second, the existence of different levels of the uniform general indirect tax in the countries in question will not affect the composition or level of trade either. Third, a decision by a country to change from one principle to the other should have no effect on trade. It can readily be shown that these conclusions hold for more than two countries. They also hold if the members of a customs union use the origin system among themselves and the destination principle for trade with countries outside the customs union—the so-called restricted origin system (Johnson and Krauss, 1970; Meade, 1974). There may, nevertheless, be important fiscal differences between the two systems. Since the location of the tax imposition determines which government receives the revenue, taxes imposed on the origin and destination principle may differ with respect to their revenue impacts. This could present a particular problem if the restricted origin system were employed while at the same time internal fiscal frontiers were abolished in the customs union, since trade deflection could occur, resulting in losses of revenue for the high-tax country. One way of overcoming the problem would be to institute a common external tax rate that was applicable to all compensatory import taxes and export refunds (Shibata, 1967; Johnson and Krauss, 1970). In short, however, with flexible exchange rates, perfectly general indirect taxes imposed on the same principle on all commodities in any country at a uniform rate will not prevent the basic resource allocation objectives of a customs union from being achieved. The maximization of production in member countries and the optimum distribution of the commodities produced among its members will both be achieved. This is the equivalence theorem. There are several considerations that may qualify the practical significance of this conclusion for tax harmonization in economic blocs. To apply perfectly general indirect taxes accurately may be extremely difficult. For instance, in practice, trade in services is often important but may partially escape destination-principle adjustments. In addition, if the initial position is not one of equilibrium, it is not clear that the equivalence of the two taxes 168

FISCAL HARMONIZATION

will hold (Shoup, 1953). Moreover, the conclusion relates to longterm static equilibrium and ignores adjustment problems. In the short run, even under a flexible exchange rate regime, exchange rates or factor prices may be fixed. Consequently, the choice of one jurisdictional principle rather than another may have temporary implications for a country’s international competitive position and its balance of payments, for the tendencies noted above will not be immediately offset. But, above all, the equivalence theorem depends crucially on exchange rate flexibility. In a common market, exchange rate stability is likely to be an important goal and, in a monetary union, exchange rates amongst the member states will be immutably fixed. In these circumstances, the choice of jurisdictional principle may have important longer-term implications for trade and resource allocation in the union.

Differentiated indirect taxes If the assumption that the indirect tax is levied at a uniform rate on all commodities is removed, it will no longer be possible to achieve full optimization. In general, when any member of a customs union applies indirect taxes the rates of which are differentiated according to product, a choice must be made between the objectives of optimizing production and optimizing trade. The destination principle is compatible with the maximization of production but not with the optimization of trade, while the origin principle is compatible with the optimization of trade but not with the maximization of production. Assume that country H levies a 20 per cent VAT on vehicles only, whereas country P imposes a uniform tax of 10 per cent. If in this situation both countries apply the destination principle, the relative prices net of tax of producers supplying country H’s market will be unaffected, and country H’s exporters will be free from tax under the destination principle. Similar considerations will apply to country P. Thus the comparative costs of producers in countries H and P will be unaffected by the tax, and production will continue to be maximized. On the other hand, relative consumers’ prices will be distorted because vehicles will be more expensive relative to clothes in country H than they will be in country P. As a result, consumers in country H will buy more clothes and fewer vehicles than they would do otherwise, and the opposite tendency will prevail in country P. Consumers in both countries would gain from 169

FISCAL HARMONIZATION

a transfer of vehicles from country P to country H and of clothes from country H to country P. In other words, trade between the two countries will not be optimized. Alternatively, assume that both countries apply the origin system. Relative consumer prices will then be the same in both countries, and therefore trade will be optimized. However, the tax will interfere with the maximization of production, because producers’ prices net of tax will be reduced in a non-proportionate way. Producers in country H will therefore be encouraged to produce clothes rather than to manufacture vehicles, and the opposite tendency will prevail in country P. The important conclusion that emerges from the foregoing analysis is that if the rates of tax are not identical for all commodities in any one country, distortion will arise, no matter whether the origin system or the destination system is chosen. The question then to pose is whether it is more important to bring about production efficiency or exchange efficiency. If the destination principle is adopted (involving explicit border tax adjustments between member countries), the location of production in the low-cost country will not be interfered with and production costs are minimized. At the same time, consumption losses may arise because the prices faced by consumers will not be uniform. Which aspect should be given priority? Musgrave (1969) argues that there are two reasons for suggesting that fiscal harmonization in a customs union should be concerned primarily with production efficiency. In the first place, potential production losses are likely to outweigh the consumption burden. Second, consumption burdens resulting from discriminatory destination taxes are largely borne by the country that imposes them, whereas production inefficiencies are shared outside the taxing country. The substantial literature on this issue that has appeared during the last two decades has not upset the presumption established earlier in favour of a destination-based system of indirect taxation. This remains the professional consensus. Under conditions of imperfect competition, however, this presumption may call for some qualification (Keen and Lahiri, 1994) as may the pursuit of national self-interest through strategic tax behaviour.

Other issues in indirect tax harmonization In evaluating the rationale of tax harmonization in economic groupings, considerations other than union allocational efficiency 170

FISCAL HARMONIZATION

may have to be taken into account, although some are relevant only to forms of international integration that go beyond free trade areas and customs unions. This section briefly reviews some of these other considerations and indicates their interrelationships with the resource-use considerations discussed so far.

Transaction costs The origin principle was perceived for many years to have one major advantage over the destination principle, namely that its adoption would make it possible to dispense with border tax adjustments and fiscal frontiers within a union. Rebates and compensating taxes would not be required at intra-union frontiers, and goods might therefore pass without control, except for what might be required for other purposes (for instance, health or drug control). The resultant savings in costs both for the state and for private enterprises could be considerable. In the EC the procedures used and the formalities required in connection with VAT and excises at intra-Community frontiers were notoriously complicated, cumbersome and costly. Their cost prior to the introduction of the single market has been put at 5–7 per cent of the value of intra-EC trade (Cnossen and Shoup, 1987). Moreover, the retention of border controls constitutes a substantial psychological obstacle to intraunion trade and market integration. The retention of fiscal controls at frontiers makes it technically feasible to resort to non-tariff barriers in cases of economic difficulty. The mere possibility of such action may affect the credibility of integration and thereby constitute an important obstacle to the full integration of markets and the maximization of potential gains. Without doubt, if it is desired that a common market should have characteristics that are similar to those of an internal market, and if internal indirect taxes imposed on the destination principle can be collected only at the borders, the origin principle would have important advantages from this standpoint. The principal difficulty with origin taxes, as noted already, is that if they are imposed on a significantly differentiated basis they will distort the location of production within the union. From this point of view, an origin tax could be compatible with allocational efficiency only if it were imposed at a fairly uniform rate within each country, although it would not be necessary for it to be imposed at the same initial rate in all member countries. If exchange 171

FISCAL HARMONIZATION

rates are fixed, however, then although initial tax differences reflected in the initial equilibrium exchange rate may continue, changes in origin tax rates may not be compatible with continued equilibrium in the absence of offsetting changes in factor and product prices. The argument that the employment of the destination principle requires border controls is, however, no longer accepted. In principle, border controls for trade between firms can be fully eliminated by shifting border tax adjustments to the books of account of taxable enterprises in other member states. This approach would make it possible to maintain the destination principle as the basis of intra-Union trade and the allocation of revenue, but would substantially eliminate the transaction costs involved in border controls, whilst still, in principle, leaving member states free to set their own tax rates. There are two main variants of this approach (Cnossen and Shoup, 1987). 1

2

Exports would be zero-rated but the collection of VAT on importation would be shifted to the first taxable enterprise in the importing member state under what is termed a deferred payment scheme. A scheme of this kind has been in use in the Netherlands since the introduction of VAT. Exports would not be zero-rated, but the first taxable person in the importing member state would be entitled to a credit in respect of the tax that had been invoiced by the exporter of another member state. These union tax credits would be offset through a clearing house, leaving net exporters to make payments of the net balances to net importers.

Either of these variants would remove border controls, but would imply the retention of fiscal frontiers in the sense of destinationbased border-tax adjustments. The clearing house system would, however, be much more demanding in terms of its requirements for co-operation amongst the internal tax administrations of the member countries and for similar standards of performance. To enable border controls to be fully removed it would also be necessary to make special provision for excisable products such as alcohol, tobacco and petroleum. Except for petroleum products, which can be controlled by other means, the removal of border controls on excisable products would point to the establishment of a linked bonded warehouse system, possibly supplemented by physical marking to prevent goods that have paid duty in one 172

FISCAL HARMONIZATION

country from being sold in another. Furthermore, if border controls are to be fully removed, either approach would have to be supplemented by provision for dealing with cross-border purchases by final consumers. The difficulty of policing this category of trade even with frontiers would point to the use of the origin system in its case. However, its adoption would raise the problem of crossborder shopping and of tax competition amongst member states in pursuit of their own national interests. The first problem, of course, is that where the markets are highly integrated, substantial initial divergences in tax-paid prices amongst the different countries of the union will give rise to opportunities for profitable arbitrage by individuals. This cross-border shopping may result in substantial revenue losses for some countries. What is perhaps even more important is that open borders may create incentives for certain countries to fix their indirect tax rates with an eye on strategic considerations, that is, for the purpose of capturing revenue gains from cross-border shopping. The outcome of such behaviour could be a reduction in indirect tax rates as a result of the tax competition so generated. This prospect has been a major concern in discussions of tax harmonization issues. On one view such tax competition is to be welcomed, to the extent that it can be expected to bring about spontaneous tax harmonization through a downward convergence of effective tax rates and a beneficial limitation of the public sector. The alternative view is that disregard of the fiscal spillovers that give rise to tax competition leads to arbitrage activities which are essentially unproductive from the standpoint of the union and to a welfare loss for the union as a whole as a result of the setting of tax rates that are too low. On this view, concerted rate harmonization is required to avoid such losses. Arguments for rate harmonization derive further support from consumption efficiency considerations.

The harmonization of rates It has already been pointed out that, even if border tax adjustments are employed to give effect to the destination principle, rate differences amongst member states will give rise to consumption distortions. This may suggest that, within a union, tax rates (or subsidies) should be harmonized on efficiency grounds, unless they are necessary to offset some other imperfection in the market mechanism. It is easy to show, for a two-country model with a 173

FISCAL HARMONIZATION

single product in which demand curves are identical, that averaging tax rates will necessarily produce welfare gains for the union as a whole. The case for harmonization is, in fact, a very general one, which applies irrespective of demand conditions and of the number of products, and it remains valid even when firms behave oligopolistically (Keen, 1993). There are, however, several objections to rate harmonization and, in particular, to complete uniformity. The first is that, to the extent that differences in rate structures reflect the exploitation by national fiscal authorities of differences in demand elasticities across countries, harmonization may have a significant impact on the tax revenues of some member states. For this and other reasons, compensation (or adjustment assistance) may be required to ensure that all countries share in the potential gain, although not all revenue changes provide grounds for compensation. The second reason is that indirect taxes are used not merely to raise revenue but also, particularly through the imposition of excise taxes, to influence patterns of consumption and production in the interests of health, social policy and income distribution. Analyses of fiscal harmonization typically proceed on the basis that countries should be permitted to retain differences in taxation that reflect important differences in social philosophy. In the case of the EU, the need to do so has been repeatedly emphasized and is accepted. Rate differences can, of course, be used for the purpose of providing protection for a national product. And while the most obvious cases of discrimination can be fairly easily identified (as, for instance, with the high tax rate imposed on Scotch whisky in France before the intervention of the European Court), it is sometimes difficult in practice to separate tax differences aimed at justifiable interference with the market based on compensation for externalities or on the weight attached to social values from differentiation that is imposed purely for the purpose of providing protection against the product of a partner country. If substantial differences in social values exist in a union, and are permitted to influence fiscal systems, it is difficult to see how the retention of a system of excises can be avoided even if there is substantial progress in the harmonization of rates of VAT. If substantial excise rate differences should be retained, any attempt to shift the required adjustments away from borders in the way described in the previous section for VAT would, in the case of these normally very high rates, almost certainly call for the development of both a system of in-bond transportation and a 174

FISCAL HARMONIZATION

connected system of bonded warehouses if fraud is to be contained to acceptable levels.

Adjustment and stabilization There may be a further conflict between the desirability on the one hand of a high degree of uniformity of structure and rates on allocational and administrative grounds, and to avoid harmful tax competition, and the merits on the other hand of retaining flexibility in the interests of facilitating national structural adjustment and stabilization policies. Any such conflict will be more pronounced in a situation of monetary union where the national use of monetary and credit policy for facilitating domestic adjustments is excluded. The conflict is likely to be more important for some taxes than for others. On allocational grounds, rate variations for the purpose of facilitating adjustments should be confined to those that will least affect the flows of factors and commodities, or those that can be offset by neutralizing arrangements in the form of border tax adjustments. This suggests, for instance, that variations in corporation tax rates should not be employed for stabilization purposes because small differences are likely to affect net profit margins significantly and to have long-term distorting effects, since capital is a highly mobile factor. On the other hand, in so far as labour is not highly mobile, variations in personal income tax rates for stabilization purposes need not be severely detrimental to efficiency considerations. As to rate variations in indirect taxation, the implications may be considered for two alternative situations—namely, fixed exchange rates and monetary union. It will be assumed that indirect taxation takes the form of a uniform VAT. Suppose that an external imbalance arises under a fixed exchange rate regime, rendering some longrun adjustments necessary. If the VAT were levied on the destination principle (that is, with border tax adjustments), variations in its rate would be neutral with respect to trade and would therefore have no effect on the balance of payments, assuming that such variations were accompanied by product price changes rather than by factor price changes. If, instead, the VAT were levied on the origin principle, without border tax adjustments among the members of the union, then, for a situation in which it was again assumed that variations in VAT 175

FISCAL HARMONIZATION

were accompanied by product price changes rather than factor price changes, the outcome would be different. In that case a reduction in VAT would increase the cost of country H’s imports from its partners relative to that of domestic goods and reduce the cost of its exports by comparison with those of country P. The effect of the change would be akin to a depreciation of the currency of country H, and there would be a tendency for the external imbalance to be corrected in the case of a common market with fixed rates of exchange. If the situation were one of monetary union and the problem were that of a structural imbalance between countries H and P, with surplus labour and capital in country H, a downward variation of VAT in country H would also result in a tendency for the imbalance to be corrected. If the problem in country H were a cyclical or conjunctural one, again a downward variation in VAT should contribute to the restoration of equilibrium. The above point may be put in a slightly different way. The merit of the destination system is that, given the above-mentioned assumptions on tax-shifting, it enables countries to pursue their own indirect tax policies without regard to balance of payments considerations. This is not possible if the origin system is employed. In short, from the standpoint of stabilization and adjustment objectives in a common market, there would be an argument for adopting the origin system of indirect taxation for intra-union trade (so dispensing with border tax adjustments among members) while retaining the destination system for external trade. This is the socalled restricted origin principle, which was recommended in the Neumark report (CEEC, 1963). In a common market with a monetary union but an undeveloped union fiscal policy the argument for such a course might be even stronger. But any advantages secured in terms of stabilization or adjustment would be obtained at the cost of sacrificing some of the resource allocation advantages of a distortion-free system. Administrative considerations may also favour the use of the destination system.

THE HARMONIZATION OF DIRECT TAXES In a common market, tax harmonization issues also arise in relation to direct taxes. To the extent that such taxes exert their influence on factor incomes, rather than on the absolute and relative prices of goods and services (except for the possible forward shifting of 176

FISCAL HARMONIZATION

the corporation income tax), the impacts that give rise to a case for harmonization on efficiency grounds, and the wider implications of harmonization, are different from those that arise in connection with indirect taxes. The principal efficiency issue concerns the impact of differences in net tax burdens between countries on the intra-bloc flow of factors and its implications for resource allocation and for the fiscal autonomy of members. Other important issues have to do with: (1) the problem of double taxation arising from the overlapping of tax jurisdictions on a single economic activity, which raises issues of productive efficiency and of equity between taxpayers; and (2) the problem of the international distribution of tax revenues, which raises questions of equity between member states. As in the case of indirect taxes, two alternative jurisdictional principles may be used for direct or income taxes. The first is the residence principle, under which taxation is applied to the total income of each resident—personal or corporate—regardless of the place where the income is earned or the capital is located. The second principle is the source principle, under which tax is applied to all income earned within the taxing jurisdiction regardless of the owner’s residence. If there is no cross-border investment and no labour mobility the two principles will be identical. These principles are normally modified by jurisdictional tax adjustments that attempt to accommodate foreign investment and factor mobility. The adjustments may be motivated by considerations of equity, revenue or efficiency. On the tax credit approach, taxes imposed abroad are fully credited by the country of residence against its own tax assessment. This effectively treats taxpayers in the country of residence equally, irrespective of whether their income arises from domestic or foreign sources. A country employing this rule has to adjust to income taxes paid abroad, reducing the domestic tax taken fully by the amount of the foreign tax where that is less than the domestic tax and, in principle, granting a refund if the foreign tax is more than the domestic tax. A second approach permits only the deduction of foreign income taxes for the purpose of arriving at net income for tax purposes. On this approach, taxpayers with equal incomes pay different amounts of tax, depending on the proportions in which their incomes are derived from domestic or foreign sources. Each country is then able to collect whatever income tax it chooses, without any adjustment to income tax paid by its residents to other tax jurisdictions. 177

FISCAL HARMONIZATION

The efficiency issue What are the efficiency implications of direct tax differentials and alternative jurisdictional principles within an economic bloc? They will depend primarily on the degree of factor mobility between the countries, on efficiency differences between countries and the enterprises of different member states, and on the actual incidence of the tax. If factors of production are immobile, as they are assumed to be in a simple customs union, and direct taxes are not shifted, they will play only a minor allocative role. If intra-bloc factor mobility is assumed, as in a common market, direct tax differentials may have considerable efficiency significance if there are locationspecific advantages attached to producing in one country rather than another or cross-border firm-specific competitive differences. Direct tax differentials can affect the intra-union location of capital or labour, depending on whether the incidence of the tax is on income from capital or on income from labour. If the direct tax on profits is shifted to leave returns to capital unaffected in the short run, profits tax differentials among the members of a common market may also make themselves felt in trade distortions among the member countries. In this chapter, attention is concentrated on the taxation of income from capital on the grounds of both relevance and realism. Capital is the most mobile of factors in economic blocs, and the need for tax harmonization measures in respect of corporate taxation is consequently most felt in this area. By contrast, the harmonization of taxes on labour is not a felt need in any economic bloc today, though it may become so in future should labour mobility increase to an extent sufficient to undermine effective national tax autonomy in that field. Differences in corporate taxation can affect the efficiency of economic activity in an economic bloc in a variety of ways. For instance, cross-border rationalization of production may be impeded unless trans-frontier mergers leave tax liabilities unaffected. The main argument for corporate tax harmonization, however, derives from the potential distortionary effects of unharmonized corporate taxes on the location of investment and production, and on the relative competititive ability of enterprises from different member states. The following discussion assumes that the tax is not shifted. In the absence of taxation, an enterprise will invest in the member state where production can be carried out at lowest cost. 178

FISCAL HARMONIZATION

Where taxes are imposed, the firm is drawn to the location that generates the highest after-tax profit. If corporate taxes are imposed and are higher in the country where production is cheapest, the enterprise may have an incentive to locate in another member state where the production cost is higher but the tax bill is lower enough to more than offset the higher cost. A higher return is thus produced for the firm but economic resources in the bloc are wasted. For this loss to be avoided, a firm must face the same effective tax rate wherever it invests in the bloc. This is termed capital export neutrality. If taxation throughout the bloc is harmonized so that it is purely residence-based, capital export neutrality will be attained. Differences in corporate taxation may also affect the relative competitive advantage of enterprises from different member states, and in doing so have an adverse effect on production efficiency. If taxes discriminate between companies, a less efficient producer may be competitive in a particular activity simply because a more efficient competitor is subject to higher taxation. For instance, in the absence of taxation, in a system of competitive tendering, the firm with lowest production costs—let us suppose an enterprise in a partner state—will be able to submit the lowest bid for a particular project, and the production cost will be minimized if that firm undertakes it. If corporate taxes are imposed, and the enterprise from the partner state has to bear the tax of the home country where the project takes place, as well as the tax of its country of residence, the bid price of the more efficient enterprise from the partner state will have to reflect its higher tax burden relative to that of the domestic enterprise in order to earn an adequate rate of return for its shareholders. This may tip the outcome in favour of the less efficient domestic enterprise. If so, the production cost of the project will not be minimized and resources in the bloc will not be used as efficiently as possible. For this kind of loss to be avoided, all enterprises capable of producing a particular good must face the same effective tax rate throughout the bloc, irrespective of their nationality. This is termed capital import neutrality. If taxation throughout the bloc is harmonized so that it is based purely on the source principle, it will be neutral in that sense. Import and export neutrality with respect to capital cannot both be achieved, except by the adoption of uniform tax rates in all countries. Otherwise, one or other type of neutrality must be sacrificed. Musgrave (1969) has argued that the two requirements are not of equal importance and that, from the point of view of 179

FISCAL HARMONIZATION

efficient resource allocation under competitive conditions, it is export neutrality that should be aimed at. Giovannini (1989) also concludes that violating export neutrality is likely to be much more costly than violating import neutrality, mainly because investment decisions are supposed to be highly responsive to rates of return whereas savings decisions are less so. However, depending on the precise nature of the tax-shifting assumptions made, a wide variety of different adjustments would be called for to produce neutrality (Musgrave, 1967). On these and on other grounds analysed by Keen (1993), it seems justifiable to conclude that some doubts remain with respect to the usual presumption in favour of the residence principle.

TAX HARMONIZATION IN THE EU At the time of the establishment of the EC the member countries exhibited substantial differences of structure and rates with respect to the five main types of tax—namely, sales taxation, excises, corporation taxation, personal taxation and social security levies. In the field of indirect taxation the general sales tax took the form of a VAT in France, a cumulative or cascade sales tax in West Germany and a single-stage tax in Italy. Excises were even more varied and included fiscal charges of all kinds, often exercised through state monopolies of manufacture and sale. In the field of direct taxation similar diversity existed in corporation and personal taxation. Three standard forms of corporation taxation were found: the so-called separate or classical system; the split rate system; and the imputation (tax credit) system. One country (Italy) had no corporation tax in the modern sense. Personal income taxes were also extremely diverse in their bases, rates, allowances and degrees of progressivity. In the field of social security finance there were large differences in coverage and methods of finance. Over and above formal statutory differences in structures and bases, there were differences in the efficacy of enforcement, which were reflected in the differences between member countries in the percentage of the due tax that was actually collected. Of the five types of tax distinguished above, fiscal harmonization has so far been sought by the EU only for sales taxation, excises and corporation tax, all of which are significant for the operation 180

FISCAL HARMONIZATION

and allocational efficiency of the common market. There are no plans for the harmonization of personal income taxes, which are regarded as falling exclusively within the domain of national economic policy, or for the harmonization of social security levies. The harmonization of indirect taxation was explicitly demanded by Article 99 of the Treaty of Rome, and the harmonization of corporation tax was implicitly required by Article 100. The treaty itself, however, provided little specific guidance on policy formulation in those areas, and proposals for tax harmonization in such areas had to be worked out subsequently with the help of a number of important reports produced for the Community by independent experts. Of these, the Tinbergen report (ECSC, 1953), the Neumark report (CEEC, 1963), the van den Tempel report (1969) and the Ruding report (CEC, 1992b) are among the most notable. In terms of academic studies, much light was initially shed on the issues by the important work initiated by Carl Shoup of Columbia University (Shoup, 1967), but the subject was then largely neglected until the mid 1980s, when the need to examine the implications of the single market initiative stimulated a tidal wave of studies which have greatly advanced an understanding of the issues. Of these newer studies the notable conference volume edited by Cnossen (1987) is basic. To date, progress in concerted institutional tax harmonization has been slow, and in the case of corporate tax harmonization, it has not gone very far, although in that field significant piecemeal adjustments have been made or are in prospect and a certain degree of spontaneous harmonization has occurred. Effectively the pace of progress is limited by the fact that every decision relating to taxation requires the unanimous agreement of the Council of Ministers. Qualified majority voting does not apply to tax issues. Since 1992 the emphasis in the Treaty of Maastricht on subsidiarity as a guiding principle has coloured the context of the debate on further tax harmonization, as have the implications of prospective monetary union for fiscal policy. The prospect of monetary union is perceived by member states to enhance the need to retain as much flexibility as possible in their revenue collection for stabilization purposes to compensate for the loss of the monetary policy instrument. The principal achievement in tax harmonization has been the adoption of a unified form of general sales taxation. Following the Neumark report (CEEC, 1963), which recommended that VAT should be adopted as the EC’s sales 181

FISCAL HARMONIZATION

tax, directives were eventually adopted establishing a common structure for VAT in all member states in the early 1970s based on the destination principle. Other important recommendations of the Neumark report that envisaged rate harmonization and an eventual shift to the restricted origin principle were not implemented. It was not until the late 1980s that indirect taxation became once again a central issue of Community policy. Following the single market initiative, and the decision to abolish border formalities, minimum rates of VAT have been prescribed. Border tax adjustments have also been abolished, as described below, but the Community did not accept the Commission’s proposed mechanism to accompany their abolition and, strictly speaking, fiscal frontiers still remain in the current ‘transitional’ system. A shift to some kind of origin system but with tax attributed on the basis of destination appears still to be on the agenda of the Commission. In relation to excise duties, little progress has been made towards harmonization of structures and rates. Although minimum rates are now in force that are subject to review every two years, and target rates towards which convergence is expected have been laid down, the minimum rates still permit wide differences—rightly so, in the view of some analysts. With respect to corporation tax, the van den Tempel report (1969) recommended that the separate system (the so-called ‘classical’ system) should be adopted as the EC’s harmonized tax. Harmonization proposals were not made until 1975, when the Commission submitted to the Council of Ministers a proposal for a directive on the harmonization of systems of corporation taxation that aimed at tax neutrality. The Commission’s proposals departed from van den Tempel’s recommendations by suggesting a common imputation (tax credit) system, which would have partly relieved the double taxation of dividends and would have involved also some narrowing of differences both in the rates of tax on profits (the band suggested was 45–55 per cent) and in those of the tax credit. No proposals were initially put forward on the appropriate corporate tax base for the calculation of profits. Six members, including the largest member states, now employ the imputation system, making agreement on any other system unlikely. In 1990 the Commission withdrew its 1975 proposals and established a committee of experts to advise on future policy on corporation tax. Some aspects of the resulting report (CEC, 1992b) are discussed below. 182

FISCAL HARMONIZATION

The harmonization of VAT and the transitional regime As a result of the initiative which began in 1985 to complete the internal market, important changes have been made in the way in which VAT is operated within the EU. (1) Consumers may now cross internal borders with unlimited amounts of tax-paid goods purchased for personal use in the course of cross-border shopping. Final purchases are therefore taxed at the rate specified by the member state in which the final purchase is made, which represents a significant modification of the destination principle. (2) To limit tax competition and the revenue losses associated with cross-border shopping, a significant move towards rate harmonization, or approximation, as it is termed, has been accepted. In particular, from 1 January 1993, a standard rate of VAT of at least 15 per cent must be applied, and no more than two reduced rates may be set. Certain special arrangements were also introduced—for instance, for motor vehicles and mail order selling—to limit the scope for tax arbitrage. (3) New arrangements were introduced for operating VAT in respect of transactions between firms, since border checks can no longer be used to verify exportation for zero-rating, or to impose VAT on imports. The new system did not essentially modify the destination basis on which transactions between firms were subjected to tax prior to 1993 but shifts its administration away from frontiers. Exports continue to be zero-rated as before but the exporter must obtain and report the VAT identification number of the customers in other member states and the value of sales to them and hold commercial documentary evidence of exportation. The collection of VAT on imports, on the other hand, is shifted to the first taxable entity in the importing country, thus effectively integrating the collection of VAT into the domestic collection system. Since large amounts of revenue are involved, controls are required to provide member states with assurances that their revenues are safeguarded. To limit error and fraud, tax administrations are to exchange information on VAT numbers and aggregate intra-EU sales through a computerized network and to intensify administrative co-operation. The system introduced in 1993 is characterized as transitional, and it should have been replaced in January 1997 by the definitive regime. Although the present system has overcome border controls and has succeeded in avoiding major distortions—in particular, 183

FISCAL HARMONIZATION

revenue losses from cross-border shopping—it has not resulted in the elimination of fiscal frontiers, since transactions between member states are not treated for tax purposes in the same way as transactions within a member state. The differences entail both psychical and real costs. Any replacement of the present system in the interests of overcoming these costs will have to strike a difficult balance between the conflicting objectives of completely removing fiscal frontiers while preserving the tax autonomy of member states. It will have to do so, moreover, while retaining the perceived advantages of the present system in terms of revenue allocation, neutrality and the limitation of fraud. The neutrality requirement itself appears to exclude any recourse to an origin-based system, which would otherwise have some merit. The most widely discussed alternative is some variant on the system originally proposed by the Commission in 1985. Under that system, it was proposed that VAT should be levied as if member countries belonged to the same fiscal territory. That is, cross-frontier sales would be subjected to VAT by the taxable vendor in the exporting country and there would be a full credit for those taxes to the taxable importer. From the standpoint of an enterprise there would then be no difference in the VAT treatment of goods whether they were for home consumption or for export to the rest of the EU. At the same time, the principle was still accepted that VAT should continue to accrue to the country of final consumption in respect of such transactions. Since intra-EU trade flows are not in balance, this would require the institution of a clearing house system, or some alternative arrangement of equivalent effect, to ensure that, in broad terms, VAT collected on export sales in one member state was ultimately reimbursed to the member states in which VAT on the imported input had been credited, and in which final consumption took place. On the basis of 1986 figures, for example, it would have been necessary for Germany and the Netherlands to have made net payments into such a clearing system of ECU 3.5 billion and 1.5 billion respectively, while France and the UK would have received ECU 2.4 billion and 1.8 billion. Such a direct transfer mechanism would be self-financing. There is no fundamental difference, in economic terms, between the clearing house system and the transitional system. There are, however, important practical objections to set against the gain from abolishing fiscal frontiers that the clearing house 184

FISCAL HARMONIZATION

system would allow. First, its adoption may involve an increase in the bureaucracy of the EU, partly or wholly offsetting any administrative savings to enterprises. In addition, in terms of the originally proposed system, there would be an adverse effect on the incentive for member states to devote resources to identifying fraudulent or dubious claims for credit if they could automatically recover the cost by billing the clearing house. The identified problems of a clearing house system could no doubt be substantially overcome by appropriate reshaping of the scheme. It is far from clear, however, that any improvements currently discussed could offer any substantial advantages over the transitional regime. The single new proposal for an alternative strategy for VAT to emerge in the current debate involves the adoption of a fully harmonized rate of VAT for transactions between enterprises combined with flexibility in rates on final sales (Keen and Smith, 1996). Such an arrangement, which can be thought of as combining a variable retail sales tax with a completely harmonized VAT, harks back to a proposal of the Neumark Committee. It would in principle have a number of advantages by comparison with the transitional scheme and could facilitate the introduction of a clearing house system. The suggestion remains to be fully evaluated. In the meantime, in the absence of any solid grounds for change, and the requirement of unanimity before any change can be made, the present transitional system may turn out to be remarkably enduring.

The harmonization of corporate taxation Many of the complex issues raised by corporate tax harmonization in the EU received extensive examination in the report of the Ruding Committee (1992). The committee was asked to address three issues: Do differences in corporate taxation cause major distortions in the functioning of the internal market? If so, will they be eliminated spontaneously by market forces and tax competition? If not, what measures are needed to remove or mitigate them? In relation to policy measures the committee limited its recommendations to ones designed to mitigate the main distortions in cross-border investment within the Community. The committee assumed that member states would wish to maintain the structure of their present corporation taxes and therefore did not consider radical proposals for moving 185

FISCAL HARMONIZATION

the EU to novel types of corporate taxation that have been advocated recently and which are reviewed by Cnossen (1996). The committee found strong evidence of distortions in the EU in the sense of departures from capital export neutrality and capital import neutrality. Differences in taxation also result in non-neutrality with respect to portfolio investment and financial structures. Some of the evidence is reviewed in the Ruding report (CEC, 1992b). Three features emerged: there are clear differences in the tax burden on domestic companies in different member states; there is discrimination against foreign companies; an EU investor faces very different tax burdens in his investments in other member states. To illustrate the last point, to obtain the same after-tax return, the before-tax return that an Irish company must generate is twice as high on an investment in Spain as on one made in Germany. Do such differences affect business decisions? To answer this question the committee sponsored a survey. The responses indicated that taxation differences were an important factor in decisions on the location of real productive activity and even more important with respect to financial activities. The committee considered the argument that the distortions would be eliminated by a process of tax competition among member states. In this connection the business survey disclosed a large majority in favour of Community action to produce convergence. The committee itself concluded that reliance on market forces and on national tax competition to deal with distortions was an inadequate response, and it therefore made a number of phased recommendations for specific measures at Community level for remedying the situation. Phase 1 should have been complete by 1994, while the second and third phases were linked with the completion of EMU. Its recommendations affect four areas: (1) cross-border discrimination, (2) tax rates, (3) the tax base, (4) tax systems. Many of the Phase 1 recommendations deal with distortions affecting cross-border investment within the EU. The evidence presented by the Ruding Committee suggests that the two most important sources of distortion are the existence of withholding taxes on dividend payments between countries and the way in which the country of residence taxes foreign source income at the corporate level. The anti-discrimination proposals concern the abolition of withholding taxes, transfer pricing, and offsetting losses in one country against profits in another. A number of the proposals 186

FISCAL HARMONIZATION

are relatively uncontroversial and have been put into effect by directives or are the subject of draft directives. Other Phase 1 proposals include a suggested minimum tax rate of 30 per cent and various proposals for harmonizing the tax base. The latter do not seem to be called for in the light of the committee’s evidence that variations in the way in which profits are computed for tax purposes constitute relatively minor sources of non-neutrality and moreover, as Devereux (1992) has pointed out, they have inherent weaknesses. There is nevertheless a case for the proposed minimum rate to counter the possibility that member states may engage in tax competition for the purpose of encouraging inward profit-shifting. There is at present, however, very little support for a minimum rate or for harmonizing the tax base (CEC, 1996b). The committee felt unable to make substantive recommendations on the tax system. Whether proposals for harmonizing tax bases and rates can be justified if the tax systems which govern the link between the taxation of the corporation and its shareholders themselves remain unharmonized is questionable. Cnossen (1996) argues that domestic and crossborder investment decisions in the EU will remain distorted by widely divergent effective tax rates on company earnings and other capital income, even if the Ruding proposals are implemented, and that more radical reform is required. His preferred solution is the adoption of a separate, purely source-based, tax on capital income, uniform for persons and corporations, on the lines of what has been introduced in Denmark and Sweden. The rate would not need to be the same across the EU in order to be neutral, and the requirements of efficiency and sovereignty would thus be reconciled. An ingenious double matrix which considers seven main options and ranks them in terms of economic efficiency against considerations of sovereignty, administrative and compliance costs and other factors has been prepared by Devereux and Pearson (1989). This suggests that only a single EU corporation tax would be ideal in terms of economic efficiency, but it is evidently far from being politically acceptable. All other reforms would progressively trade off economic efficiency against fiscal sovereignty and administrative simplicity. The preceding outline of the tax harmonization initiatives of the EU makes it clear that they have been undertaken primarily from the standpoint of resource allocation considerations. In relation to indirect taxation, the strategy has been to seek to unify structures 187

FISCAL HARMONIZATION

and to equalize rates in the first place in order to prevent distortions in the functioning of the market with respect to trade. Rate equalization (and initially the adoption of the restricted origin principle) has also been sought for administrative and other reasons, in order to permit the abolition of fiscal frontiers within the EU. Similarly, unification of the structure of the corporation tax and harmonization of rates have been sought in order to avoid distortions in the location of investment and of competitive advantage. The equalization or unification approach traditionally emphasized by the Commission may be criticized on two grounds. First, unlike the position with the tariff, the equalization of tax rates among member countries cannot be undertaken at zero but requires a positive rate. A nominal equalization of rates need not equalize the effective rates of taxation facing enterprises, because of different tax-shifting possibilities in different sectors and markets. Consequently, even when VAT and excise rates are equalized, indirect taxation may not necessarily be neutral or distortion-free. In the case of corporation tax, the absence of a uniform base and differences in inflation rates may mean that similar rates would imply very different effective rates. More recently, however, the Commission seems explicitly to recognize the need to approach the problem in effective terms (CEC, 1996b). Second, the equalization approach gives no weight to the adjustment or regional policy objectives of member states. If the levels of major revenue-producing taxes are harmonized, their use to promote such objectives will necessarily be largely ruled out, even though tax differentials and tax adjustments may be appropriate for the purpose. For instance, from the standpoint of the maintenance of external equilibrium among the members of a common market, it might be advantageous, if the restricted origin principle were in force, for a country to be able to raise its VAT rate above a uniform rate or band if it experienced a balance of payments surplus, and vice versa for a deficit, thus reducing or eliminating the need for exchange rate changes. In the case of a monetary union such flexibility could be even more important, since variations in VAT (if origin taxation were in force) would be one of the few remaining national measures open to a country for alleviating its adjustment problems. The full harmonization of VAT and of corporation tax rates in the EU is unlikely to be acceptable until the fiscal role of the EU itself 188

FISCAL HARMONIZATION

becomes much greater than it is today. This would almost certainly imply an EU budget that was: (1) substantially larger in relation to national budgets than is now the case; (2) perhaps endowed with the capacity to levy taxes of its own; and (3) able to utilize its revenues for an expanded regional policy and for adjustment and stabilization transfers among member states. These considerations underline once again the close links between proposals for monetary integration, fiscal harmonization and regional policy.

CONCLUSION This brief survey of the issues and problems involved in fiscal harmonization in economic groupings indicates their complexity. In order to reduce the analysis to manageable dimensions, the approach commonly followed is to consider a single tax in isolation from other taxes, expenditure patterns and other areas of government policy, such as exchange rate and monetary policy, and on the basis of alternative assumptions about incidence to deduce a suitable pattern or system of tax harmonization to promote neutrality or other relevant objectives. Despite its highly restrictive nature, even this approach is not free from analytical difficulties. In the first place, the criteria for tax harmonization are several, and some of the requirements conflict. Second, the impact of taxes is likely to be affected by other variables that are subject to government control, including monetary conditions, public expenditure patterns and competition policy. In particular, different monetary conditions in the member countries of a common market may produce differing tax-shifting patterns and, notably, very different patterns of effective corporate tax. Consequently, where a monetary union does not exist—or, at least, a high degree of monetary policy co-ordination— it would be an impossible task to elaborate an appropriate regime to secure effective, as opposed to nominal, rate harmonization. The analysis becomes still more complex if the simplifying assumptions of the orthodox comparative cost model are relaxed.

NOTE 1

In the context of EMU, however, social security expenditure may soon become a matter of common interest. See Kopits (1997).

189

11 MONETARY INTEGRATION

Monetary integration has two essential characteristics: (1) exchange rates in the integrated area must bear a permanently fixed relationship to each other, although they may jointly vary with respect to other currencies; and (2) there must be full convertibility in the sense that there are no exchange controls on either current or capital transactions within the area. (Convertibility for traderelated transactions is indispensable for the effective functioning of a customs union; convertibility for capital transactions is a principal ingredient of capital market integration, which is itself an essential characteristic of a common market.) If these two characteristics are to be adequately guaranteed and to be truly immutable, two other requirements would also be essential. First—since immutability of fixed exchange rates would depend on monetary policies within the area being consistent with that object—the use of the standard instruments of monetary policy must be assigned to the community and exercised solely by its monetary authority, leaving no autonomy for member states in this field. Inter alia this would imply that, beyond any agreed amount of monetary credit to which a member might be given access, any member state’s budget deficit would have to be financed in the capital market. Second, since any change in the rate of exchange between an external currency and those of the area must be uniform, responsibility for exchange rate policy with other currencies and for the balance of payments of the entire community with the rest of the world must also be assigned to the community, and its monetary authority must control the pool of exchange reserves. Under such a system it would be difficult if not impossible for a member state to calculate its balance of payments with its partners and with the rest of the world. 190

MONETARY INTEGRATION

When all these requirements are met, the resulting arrangement can be described as one of complete monetary integration, or of monetary union. In effect there would then be a single currency, although nominally differentiated currencies might continue to coexist, as is the case in Belgium and Luxembourg and in some monetary unions in Africa. It would be possible for the member states of a customs union or a common market to seek to establish the two essential characteristics of monetary integration merely by agreeing amongst themselves to fix their exchange rates permanently, to maintain full convertibility and to back the agreement up with promises of economic policy co-ordination, but without going so far as to integrate their monetary policies or to establish a common pool of foreign exchange reserves and a single central bank. Such an arrangement might help to make market integration more effective, but it could not guarantee the immutability of the relation between the currencies of a regional bloc that is essential to the concept of complete monetary integration or monetary union. The term pseudounion (Corden, 1972a) is often applied to this degree of monetary integration. Monetary union involves an additional limitation of national autonomy in economic policy beyond that required by the membership of a customs union or common market. Although monetary integration may offer potential benefits in terms of making market integration more effective through reductions in transaction costs and resource allocation gains, it would, at the same time, entail a definitive loss of control over a supposedly important instrument of macroeconomic policy that is particularly bound up with the emotive concept of national sovereignty. What are the considerations that have to be taken into account in evaluating the case for taking such an important step? The traditional approach to determining whether a country should participate in a common currency or should instead maintain its separate currency, tried to identify crucial economic characteristics that, having regard to macro-economic objectives, would serve as criteria for determining the optimal domain of a currency area. Optimality requires that the existence of a common currency should not reduce the ability of member states to adjust to shocks or increase their vulnerability to such shocks. The leading contributors to this optimum currency area debate have in turn emphasized different attributes whose existence in the area would enable the effects of country-specific shocks on output and employment to be 191

MONETARY INTEGRATION

mitigated or overcome, when movements in intra-bloc exchange rates are no longer available for that purpose. The seminal contributor to this debate was Mundell (1961), who identified factor mobility as the strategic attribute of an optimum currency area. His argument was that, when factors move freely within the area, adjustments to real disturbances can take place without the need for large and damaging price and income changes amongst its constituent members. McKinnon (1963) emphasized instead the importance of a high degree of openness (in the sense of the ratio of trade to ouput) of the individual economies, which is crucial for determining the extent to which, in any case, the prices of domestic outputs must remain in line with foreign prices. In a later contribution, Kenen (1969) stressed the primacy of a high degree of diversification, on the grounds that, the higher the degree of diversification, the more likely it would be that real disturbances would average out. MacDougall (1975) and Allen (1982, 1983) for their part emphasized the crucial significance of a high degree of fiscal integration, which would maximize the feasibility and effectiveness of intra-union fiscal transfers in facilitating the process of stabilization. The traditional optimum currency area approach provides a number of insights into the issues of monetary union but is open to several crucial objections, apart from its exclusive focus on costs. In the first place, several of the relevant characteristics are difficult to measure unambiguously. Second, it is evident that no single country is likely to possess all the attributes required to make it an ideal member of a monetary union, yet the criteria proposed cannot formally be weighed against each other. Vaubel (1978) was thus led to propose as the crucial characteristic a measurable attribute, namely the smallness of countries’ revealed need for real exchange rate variability which, he claimed, comprised the relevant implications of all other criteria. Viewed as a criterion for judging whether to enter a monetary union, this attribute is inadequate, since past experience can provide no assurance that a need for significant real exchange rate variability may not arise in the future, while the formation of a union may itself change the behaviour which generated the extent of variability in the past. The third criticism is even more fundamental, namely that the traditional theory pays no attention to problems of credibility. The application to the problem of monetary integration in the past decade of newer macroeconomic theories that stress problems of time consistency and credibility (Barro and 192

MONETARY INTEGRATION

Gordon, 1983; Giavazzi and Pagano, 1988) has introduced a new dimension into the analysis. This has not only undermined the implicit assumption of optimum currency area theory that exchange rate policy is a tool that is available for discretionary use but, in the process, has identified a major new source of benefit from monetary integration. Newer appraisals of monetary integration predominantly follow a broader approach than that of the traditional optimum currency area theories. Instead of appealing to any single criterion or goal and in that context focusing on costs only, emphasis is placed on the range of costs and benefits that may be anticipated from monetary integration. Attempts are then made to evaluate their weight and their significance for the attainment of the objectives of the individual member states and of the group as a whole, in the light of the trade-offs amongst the objectives that are involved. In this respect, the analysis of monetary integration has evolved in a similar manner to that of other aspects of the economics of international integration. Some of the questions which are still not adequately addressed by the current theory of the optimum currency area are indicated by Mélitz (1995). This chapter selectively considers some of the leading aspects of the recent analytical discussion of monetary integration. It also considers the EU context and the prospects of monetary union in Europe.

THE BENEFITS OF MONETARY INTEGRATION The fundamental economic case for superimposing monetary integration upon product and factor market integration is that it will make the integration of those markets more effective, with beneficial effects on resource costs, economic efficiency and growth. The argument has been particularly well put by Scitovsky (1958). If the creation of a customs union or common market is to have its full beneficial effects on output, it must modify the nature, scale and geographical distribution of economic activity. In market economies this will result primarily from the influence exercised by the existence of the union on the investment decisions of private enterprises. Since business investment decisions are long-run in nature, that influence will not make itself fully felt unless 193

MONETARY INTEGRATION

entrepreneurs can assume with confidence that the free market will be maintained and that intra-union economic relations will not be disrupted by trade restrictions, exchange controls or exchange rate adjustments that may have similar effects on trade as tariffs. The complete fulfilment of these conditions would in effect demand the establishment of a common currency. This basic argument has been questioned in its application to trade, on the grounds that enterprises can hedge the risk of exchange rate fluctuations, and that in any case empirical studies have not found a significant relationship between exchange rate variability and trade. However, the potentially adverse impact of exchange rate variability on long-term investment through its effects on risk cannot so easily be dismissed. If the elimination of exchange rate uncertainty reduces risk and causes a decline in the risk-adjusted cost of capital, that would yield, in terms of neoclassical growth theory, a temporary increase in the rate of growth through its positive effects on investment (Baldwin, 1990), or even a sustained increase if there are dynamic external economies of the kind that are formalized in some new growth theories. It is possible, however, that the elimination of exchange rate variability may also reduce the expected value of profits and hence the rate of return. If so, from a purely theoretical point of view, the reduction of risk will have an ambiguous effect on investment and growth (Emerson et al. 1992, chapter 3). The question is therefore an empirical one. Some of the limited empirical evidence for Europe is reported in Molle and Morsink (1990). A second important source of potential allocational and resource-saving benefits derives from the high degree of integration of financial markets that would necessarily be implied in a situation of monetary integration. First, the removal of any controls over FDI should reinforce the allocational benefits from more rational investment decisions. Second, the increase in the effective size of the market may enable operational economies of scale to be exploited so that the resources employed in the processes of financial intermediation and in transforming savings into investment are reduced. Third, financial market integration may improve the allocational efficiency of the financing process itself to the extent that it provides both borrowers and lenders with a broader spectrum of financial instruments, so enabling more efficient choices to be made in terms of duration and risk. Discontinuities in the range of available finance may also be reduced, so facilitating the financing of operations of optimal scale 194

MONETARY INTEGRATION

by enterprises. Resource allocation gains may also be obtained from increased intra-union trade in securities as a result of a tendency for returns on different kinds of capital to come closer together in the member countries. The creation of an effective union-wide market for assets that could be expected to maximize such effects would evidently require the concerted adoption of positive measures to promote financial integration through the harmonization of national financial regulations and the structures of financial and capital market institutions of member countries. Some segmentation must be expected to remain, in the presence of information imperfections across countries, but government bonds of the same degree of maturity and risk should become close if not perfect substitutes and interest rate differentials should disappear. A third significant potential source of benefit from monetary integration results from the permanent resource savings to be expected from the elimination of the direct transaction costs involved in converting one currency into another. There would also be savings in transaction costs in trade with third countries. The latter gains would be larger if payments outside the union formerly denominated in other currencies came to be denominated in the union currency. In addition, internal administrative costs imposed on enterprises where sales within a community involve several currencies would be eliminated. Fourth, when a common currency is established and a common pool of foreign exchange is created, once-and-for-all economies in holdings of foreign exchange reserves should be possible for two reasons: (1) if members are structurally diverse, any payments imbalances may be offsetting, and the pooling of reserves should enable the minimum level to be reduced; and (2) foreign exchange will no longer be needed to finance intra-union trade. The reserves set free could be redirected to other uses. They could partially offset the costs associated with the introduction of a common currency, which would be considerable. Reduced costs of financial management may also be possible and yield continuing benefits since a common currency should make it possible to spread the overhead costs of financial transactions more widely. In addition some part of the activities of institutions dealing in foreign exchange could be dispensed with, thus generating further resource use savings. In addition to the microeconomic and cost-reducing effects of 195

MONETARY INTEGRATION

monetary integration, just outlined, more recent theorizing emphasizes a further major benefit of a macroeconomic nature that derives from the problem of time inconsistency and policy credibility. In essence, it is argued that a monetary union may improve the anti-inflationary credibility of monetary policy. A government which seeks on its own to maintain anti-inflationary policies to avoid the distributional, efficiency and growth costs of inflation may find it difficult to convince the private sector and the financial markets of its commitment to low inflation. Any such lack of credibility will result in higher equilibrium inflation and/or lower output and employment. By contrast, if a monetary union is set up in such a way that its central bank is independent of any national government and has a clear mandate to operate a monetary policy that will keep prices stable, output losses should be reduced because private expectations of inflation will be favourably affected, as will nominal interest rates and wage rates. The potential gains under this heading are larger for countries where the monetary authorities (the central bank and, where the latter is not independent, the ministry of finance) have a poor reputation as a result of a past record of high inflation. For countries with a better record and greater credibility, the formation of a monetary union may hold out little gain from this point of view, and may even involve a potential cost. The magnitude of gains from most of the allocational and costsaving sources is virtually impossible to quantify accurately, though the European Commission has attempted to do so for EMU for several of the direct benefits (CEC, 1990b), notably the savings in transaction and other efficiency costs. But even modest gains would be worth seeking if monetary integration were to involve no offsetting costs with respect to other goals of policy. Are there such costs to be set against the favourable effects that monetary integration can certainly be expected to have through its resource allocation and cost-reducing benefits and impact on policy credibility? If so, what are they, and what are the factors that influence their magnitude? More generally, what are the conditions that have to be satisfied if monetary integration is to be feasible, sustainable and successful?

196

MONETARY INTEGRATION

THE COSTS OF MONETARY INTEGRATION Historically, debate on the costs of monetary integration has focused principally on the macroeconomic implications of the constraints that such a regime would place on the conduct of the monetary, fiscal and, in particular, exchange rate policies of the participants. The crucial issue is the extent to which acceptance of those constraints would significantly impair the effective pursuit of national macroeconomic policy objectives, in respect of economic stability, growth and external balance. But, given the uncertainties that exist regarding the way in which economies behave and should be modelled, this issue is not easy to resolve. The following discussion distinguishes between a largely Keynesian ‘traditional’ approach and a ‘modern’ approach that reflects the mainstream of modern macroeconomics. This mainstream draws on both sides of the Keynesian—monetarist debates of the 1960s and early 1970s. It has also absorbed a number of insights from new classical macroeconomics, notably the assumption of rational expectations, but it does not accept the key new classical assumption of perfect market-clearing. It also draws in some cases on the insights which new Keynesian economics derives from notions of asymmetrical information and transaction costs. The way in which exchange rates are determined is clearly of central importance to the analysis. Here contemporary analyses will be followed which view exchange rate determination as predominantly a function of financial asset market variables in the short run and (in some but not all treatments) of current account variables in the long run. In this framework, considerations of competitiveness that are the only important factors when capital is immobile or exogenous retain their fundamental relevance under capital mobility by determining the long-run real exchange rate of one currency for another (indicating the price of tradable goods relative to home goods). Capital flows themselves are viewed as being determined not merely by nominal differences in yield but also by expectations about the exchange rate. Changes in expectations, perhaps related to changes in monetary policy, can in these interactive analyses be an independent cause of exchange rate changes. Typically these models predict that, in response to a change in monetary policy, the exchange rate will overshoot the new long-run equilibrium, that is, it will depreciate (or appreciate) more in the short run than is required in the long run. 197

MONETARY INTEGRATION

Competitiveness and aggregate output may thus be adversely affected in the short run. The range of specific models in this vein is considerable (MacDonald and Taylor, 1992). Differences in their conclusions hinge critically on the assumptions that are made about expectations and wage—price flexibility. In particular, such assumptions crucially affect the relative time paths of adjustment towards the long-run real level or current account-determined path.

The traditional approach The traditional case for retaining exchange rate flexibility within a customs union supposes that the member states wish to maintain internal balance in the sense of full employment with price stability and external balance in the sense of balance of payments equilibrium. Tacitly the case rests on three propositions: (1) that the economies of the member states are ‘insular’, in McKinnon’s (1981) sense that prices are determined domestically because of limited financial and commodity arbitrage with the outside world; (2) that flexible exchange rates can therefore provide significant national autonomy in matters of macroeconomic policy; (3) that national governments require such autonomy because of differences in national objectives. To achieve simultaneously a given number of independent economic targets, at least a similar number of policy instruments is required (Tinbergen, 1952). The simultaneous attainment of internal and external balance therefore requires the use of two policy instruments. The orthodox Keynesian approach to the role of exchange rate flexibility assumes these to be: (1) demand management policy, which (prices being given) influences the level of employment and is assigned to the internal balance objective; and (2) exchange rate policy, which is designed to influence the composition of expenditure in favour of, or away from, domestically produced goods through its influence on international competitiveness, and is assigned to the external balance target. The basic case for resisting the fixity of exchange rates that is a necessary part of monetary integration is simply that, if the exchange rate instrument is forgone, conflicts between the objectives of internal and external balance are likely to arise for the members of a customs union or common market. The underlying model (Swan, 1955; Johnson, 1961) postulates that the level of domestic employment and the balance of payments 198

MONETARY INTEGRATION

both depend on the level of domestic expenditure and the cost of a country’s exports relative to those of its competitors. If the ratio of foreign prices to domestic prices is high (which means that a country’s competitive position is favourable), a given level of domestic employment can be supported with a relatively low level of domestic demand. If the ratio is unfavourable, employment generated by exports will be relatively low, and a high level of domestic expenditure will be necessary to maintain a given level of employment. Devaluation is the instrument by which the international cost ratio is affected, so as to switch expenditure from foreign goods to domestic goods. Consequently, if countries deprive themselves of the possibility of making needed exchange rate adjustments by participation in some form of monetary integration such as a pseudo-union, they will inflict costs upon themselves as a result of enforced departure from internal balance unless their competitiveness happens to be just right and remains so. That conclusion assumes that no alternative means are available of altering competitiveness. Any expenditure-switching measures that take the form of discriminatory restrictions would, as already noted, be ruled out in a common market and, in any case, would generally be undesirable on resource allocation grounds. As to more general measures, it is conceivable that incomes might be additionally taxed and the proceeds used to subsidize the prices of domestic production—a fiscal operation that would have similar effects to devaluation. But that too would almost certainly be excluded by international and customs union obligations. The only possible general alternative way of bringing about a required adjustment in competitiveness that would not be inconsistent with the obligations of customs union membership would be to change the level of domestic wages and prices directly. But, as Friedman (1953) has tellingly observed, even if it could be done, in just the same way as it is more efficient to change the clock by one hour than to reschedule the time of every activity if the benefits of daylight saving are sought, so single changes in nominal exchange rates are likely to be a more efficient way of producing any required adjustment than the revision of thousands or millions of individual wages and prices in terms of national currencies that would otherwise be needed to produce the same effect. But, in any event, exchange rate flexibility can be successful in maintaining balance in this model only if real wages can be reduced by devaluation, so permitting the real rate of exchange to adjust. 199

MONETARY INTEGRATION

This condition will not necessarily be fulfilled if wages are determined either within an expectations-augmented Phillips curve framework or within a ‘target real wage’ framework of a more Keynesian kind. In that event, devaluation will typically lead to a full ‘pass-through’ from the exchange rate to prices and wages so that in the medium run (perhaps of two to five years), real wages and international competitiveness will be unaffected (Artis and Miller, 1979). If real wages cannot be reduced by devaluation, then, in the absence of other solutions, perhaps of a structural kind, external equilibrium will be incompatible with internal balance, no matter whether exchange rates are fixed or flexible. The loss of the option of making exchange rate adjustments that is necessarily involved in monetary integration would then not of itself be of any significance. No ‘cost’ would be involved. If monetary integration takes the form of a monetary union with a single currency, rather than a pseudo-union, essentially similar considerations arise but they would present themselves in a different form. In a monetary union, the need to settle deficits and surpluses in foreign currency would not arise for individual members. But intra-union current deficits and surpluses would still have to be regulated. If a member state were in deficit, correcting it would require the same kind of downward real wage adjustment if employment is to be maintained in that state as would occur in the absence of a monetary union with an effective currency devaluation. But inside a monetary union such an adjustment could be brought about only by a relative fall in the deficit country’s wage and price levels. The need for such adjustments may, of course, be postponed by a redistribution of private financial assets within the union. If the member states of the monetary union had attained a significant level of fiscal integration as well, it might also be substantially reduced by automatic or discretionary fiscal transfers, such as are discussed in Chapter 9. A later approach to the analysis of the implications of exchange rate unification for internal and external balance in a similar analytical tradition focused attention on a different aspect of the problem, namely the cost of unifying policy targets in a monetary union in which the objectives of national economic policies differ. This approach, initiated by Fleming (1971) and Corden (1972a), crucially assumed an inverse relationship between the rates of change of money wages and prices and the level of unemployment of a kind first postulated by Phillips (1958). A stable Phillips curve relationship of this kind would imply that a government confronts a trade-off 200

MONETARY INTEGRATION

between unemployment and inflation, and it is assumed to have a particular preference ordering between alternative achievable levels of unemployment and wage rate changes. A particular welfaremaximizing combination of unemployment and wage rate changes could thus be chosen that represented the optimal point of internal balance. However, since countries with permanently fixed exchange rates or a common currency cannot maintain different inflation rates (at least for traded goods), some might be compelled to depart from their optimal positions if a uniform rate of change of costs and prices were to ensue. Some might be compelled to accept on a permanent basis more inflation than they would choose while others would have to accept more unemployment. For those countries that were obliged to suffer additional unemployment, its excess (valued by the corresponding loss of output) would be one significant measure of the cost of monetary integration. Figure 11.1, for example, shows the Phillips curve locus of the different possible combinations of unemployment and price inflation (assumed here to be the same for both countries). Suppose points a and β represent the positions which the home country and the partner country would each choose in a situation of floating exchange rates: the home country prefers lower

Figure 11.1 The original Phillips curve trade-off between inflation and unemployment

201

MONETARY INTEGRATION

unemployment even at the cost of higher inflation, while the partner country prefers lower inflation even at the cost of higher unemployment. If these two countries formed a monetary union on the basis of the preferred position of the partner country, then the home country would have to accept a move from a to b, since in the union the inflation rate of the two countries would have to be the same, and this would mean accepting that unemployment would be continuously higher, at u1 instead of u0.

The modern approach A very different view of the significance of monetary integration is implied by the incorporation of expectations into the Phillips curve framework following the work of Phelps (1967) and Friedman (1968). Monetarists such as Friedman saw the role of monetary policy in terms of the control that it affords, in principle, over nominal targets—the price level, the exchange rate or the rate of inflation. Whether and, if so, to what extent changes in such targets have effects on real variables such as output, employment and the real rate of exchange and competitiveness is seen to depend largely on the behaviour of suppliers of labour and goods. In this connection, the original Phillips curve relationship failed to distinguish between changes in real and changes in nominal wage rates, implicitly assuming that labour suppliers and demanders expected the price level or its rate of change to be constant and did not change their price expectations or their willingness to supply and demand labour at different nominal wage rates even when, as an inevitable result of monetary policy changes, those price expectations turned out to be false. But if, as Friedman and Phelps argued, labour suppliers and demanders should be regarded as forming expectations of future price inflation which feed into their labour supply and demand decisions and hence into the determination of nominal wages, there ceases to be any long-run trade-off between inflation and unemployment. Figure 11.2 shows the effect of introducing expectations into the Phillips curve analysis of Figure 11.1. Here the vertical curve at un shows the combinations of inflation and unemployment rates which are compatible with long-run equilibrium: any inflation rate can be attained, but unemployment in the long run must always be at the rate given here as un, the so-called ‘natural rate’ of unemployment. The downward-sloping lines indicate the trade-offs that exist in the 202

MONETARY INTEGRATION

Figure 11.2 The expectations-augmented Phillips curve trade-off

short run, but in the short run only, between inflation and unemployment; each is drawn for a different level of expected .e inflation ( p ), and actual and expected inflation are always equal on the vertical long-run Phillips curve at un. On the expectations-augmented Phillips curve view, the natural rate is the rate of unemployment that is consistent with the structure of real wage rates. Other things given, at that level of unemployment, the rise of real wage rates in the long term is governed by capital formation and technological change. Thus the natural rate is essentially the rate that is consistent with general equilibrium, defined as the absence of excess demand in all markets or as a state in which all expectations are realized. It is determined by the structural characteristics of the labour (and product) markets, including market frictions, minimum wage legislation, unemployment benefit rates and labour mobility. Many of the characteristics that determine it are thus, as Friedman has put it, ‘man-made and policy-made’ (1968, p. 9) and it is not immutable. Natural rate theorists accept that the rate may not be optimal and that it may be desirable to aim to reduce it. Other economists have emphasized this by using the term ‘non-

203

MONETARY INTEGRATION

accelerating inflation rate of unemployment’ (NAIRU) instead of ‘natural rate’. But both groups agree that the NAIRU or natural rate—un in Figure 11.2—is largely determined by structural and microeconomic factors, that it can be reduced only by appropriate structural policies, and that it is largely invariant with respect to macroeconomic policies. On this view, raising the rate of inflation once and for all lowers the rate of unemployment only temporarily. A corollary of this is that the only way to keep the unemployment rate permanently below the natural rate is by continuously increasing the rate of inflation. Each time the authorities raise the inflation rate they achieve a temporary increase in employment. Ultimately, however, the process must be self-defeating. A further corollary is that a single exchange rate change will not have lasting effects on employment, because it will be neutralized by induced inflation, so that the ability of a country to achieve relatively low unemployment by depreciating its exchange rate is also a short-term matter. The implications of the natural rate hypothesis for the evaluation of monetary integration are quite profound for, in contrast to the traditional approach, the policy stance it suggests is that such a regime could be adopted without incurring any lasting cost in terms of unemployment. For instance, if entering a monetary union requires a country to lower its inflation rate this will require a movement such as that . . from p 2 to p 0 down the curve marked d in Figure 11.2, where the economy first moves down to the right along the short-run Phillips . curve for expected inflation of p 2, and then expectations begin to . fall and inflation falls more rapidly towards p 0. Thus the country has to suffer only a short-run transitional cost in terms of the rise in unemployment required during the reduction of inflation to the new union level. The above analysis has been couched implicitly in terms of adaptive expectations, that is, it assumes that expectations of inflation adjust gradually towards actual inflation. This was the hypothesis associated with Friedman’s critique of the original Phillips curve relationship, but later new classical macroeconomists introduced the idea of rational expectations. According to that hypothesis, economic agents do not merely respond to actual inflation but make forecasts of inflation based on all the information available to them, including their knowledge of the way in which the economy works and the way in which the government or monetary authority operates. If expectations are 204

MONETARY INTEGRATION

adaptive, although there is no long-run trade-off between inflation and unemployment, a short-run trade-off still exists. But if expectations are rational there is no trade-off in the short run either for the systematic element of macroeconomic policy, where ‘systematic’ means that policy responds in a particular way to deviations of output and employment from the natural rate (e.g. policy is tightened (loosened) a certain amount for every 1 per cent fall (rise) of unemployment below (above) the natural rate). The private sector can then predict what the monetary authority is going to do and, as a result, prices adjust rather than output or unemployment. The economy moves up and down the vertical long-run Phillips curve without any movement to the left or right. Thus the short-run transitional unemployment cost, for a country entering a monetary union that requires it to adopt a lower inflation rate, disappears altogether. This point is illustrated in Figure 11.3, where the left-hand diagram refers to the home country, which is initially at point a, and the right hand to the partner country, which is initially at point b. If the two countries decide to form a monetary union they must have the same rate of inflation, but provided that the implications of the move to monetary union are known and understood, neither

Figure 11.3 Rational expectations, credibility and time consistency

205

MONETARY INTEGRATION

country has to accept any long-run change in unemployment in . order to move to the union’s chosen inflation rate of p 0 at point b (g): expectations adjust and the home country simply moves down its long-run vertical Phillips curve without any change in unemployment. Figure 11.3 can also be used to illustrate time inconsistency and the possible benefit of monetary union in the form of lower inflation. Time inconsistency means that a policy which is optimal in the long-term is not consistent with short-term incentives. The problem of time inconsistency arises as follows. Suppose that the government would like to reduce unemployment below its natural rate, in order to increase employment and to improve its electoral prospects. In a world of rational expectations the private sector understands this desire. The result is that zero actual and expected inflation cannot be reached: the private sector believes (correctly) that if actual and expected inflation fell to zero the government would then succumb to the temptation to expand the economy so increasing inflation, so that the private sector’s expectations of inflation never fall to zero. In Figure 11.3 (left-hand diagram), for example, uH* represents the point the home government would prefer to reach and the curves labelled IH1 are its indifference curves. Their concavity implies that as the inflation rate declines, the government attaches greater weight to reducing unemployment. Indifference curves closer to the origin represent a higher level of welfare than those further . away. If inflation fell to p0 at point g and expected inflation remained . equal to p0, the government would be able to reach a higher level of welfare (on indifference curve IH1 which is nearer to uH*) by relaxing monetary policy to move the economy to point d But the private sector understands this temptation, and therefore never . expects inflation of p0. Consequently, the economy ends up at a (higher) equilibrium rate of inflation at which the government has no incentive to increase (or decrease) the rate of inflation and the private sector perceives this to be so. Such an undesirable equilibrium can be avoided only by improving the anti-inflationary credibility of monetary policy. One solution to the problem is for the government to ‘precommit’ itself to low inflation, for example by announcing a target for monetary growth or inflation such that it will suffer a damaging loss of reputation if it misses the target; if the projected loss is large enough the private sector will be convinced that the government will fulfil the target. An alternative strategy would be for the government to transfer responsibility for monetary policy to an independent 206

MONETARY INTEGRATION

authority which has no need to seek re-election and no desire to increase employment beyond the sustainable natural rate, and therefore has, and is perceived to have, no incentive to inflate. Within this perspective joining a monetary union constitutes a way of transferring monetary responsibility to a non-domestic institution with different objectives from those of any member state. In Figure 11.3 the curves labelled IH and I P represent the indifference curves of the two governments. Assume initially that both countries have floating exchange rate regimes, and that the preferences of the home country’s government are such that its long-run time-consistent equilibrium is at point a, while the partner country’s preferences (either because its government is more averse to inflation or because it has given monetary responsibility to a central bank which is inflation-averse) are such that its longrun equilibrium is at b. The problem for the home country, which suffers from a marked lack of credibility (i.e. it has a high longrun inflation equilibrium), is to persuade the private sector that if inflation comes down the government will not then stimulate inflation in order to get on to an indifference curve nearer to its preferred point of uH*. If the home country pegs its exchange rate irrevocably and credibly to the currency of the partner country, or if the two countries set up a monetary union with arrangements which ensure that monetary policy is operated as in the partner country, the home country will be able to reach point g without the private sector expecting it to inflate towards point d. In short, if the monetary union is set up in such a way that the new union-level central bank is independent of national governments and has a clear mandate to keep prices stable, the benefits in terms of lower inflation for countries whose alternative domestic arrangements had previously been ineffective in keeping inflation down could be substantial. In the more general case in which non-traded goods exist it would be possible for national inflation rates to diverge in such a union because it is only identical price changes of traded goods that are implied. If rates of growth of productivity in the traded goods sectors differed in the two countries while there were equal growth rates of productivity in the non-traded goods sector, the measured rates of inflation of the two countries would necessarily differ. In that event, price stability in a dominant partner could even require a fall in the price level of the other country. This consideration, which may have some practical importance in the context of the European 207

MONETARY INTEGRATION

Monetary System (EMS), complicates the model but does not alter its essential implications and will therefore be disregarded.

Balanced and unbalanced growth Two further aspects of monetary integration can be underlined at this point. In the first place, the institution of a monetary union that would entail the acceptance of a common rate of inflation would not necessarily imply the balanced growth of the member countries in the sense of equivalent real wage growth. Suppose that productivity growth rates differ between the member countries. This means that, with the same inflation rate, the rates of growth of nominal wages (and therefore of real wages) will also differ, since in long-run equilibrium the rate of change of nominal wages must be equal to the inflation rate plus the rate of productivity growth. Each country would have the same inflation rate and each would be at its own long-run equilibrium. Provided the divergence in rates of growth of real wages corresponded to the difference in rates of productivity growth, that difference would not in itself generate problems of internal balance. A lowproductivity-growth member state could be competitive with a high-productivity-growth member and could remain so if it accepted a lower increase in real wages. However, it is not inconceivable that the unbalanced growth in this sense of the prospective partners might be unacceptable. Political considerations may make it essential that the high degree of economic integration that is involved in monetary union should also be accompanied by the balanced or convergent growth of its member countries. One possible justification for such a requirement in a common market would be that, otherwise, divergent growth rates of real wages would ultimately lead to large-scale movements of labour towards the high-productivity-growth countries. The political and economic costs for member states of such intra-group migration might be perceived to be unacceptably high and to threaten the cohesion of the union. If for any such reasons it were required that a monetary union should also be accompanied by balanced growth, it would be necessary to bring the rate of change of (nominal and real) wages between countries together. One means of doing so would be to introduce fiscal transfers at the union level from one country to another. This could be effected 208

MONETARY INTEGRATION

by a wage tax in the country with the higher productivity growth and a wage subsidy in the lower-productivity-growth country. Equilibrium would be attained when the fiscal transfer was such that the net growth of wages was the same in each country. A second possibility would be to introduce structural policies at union level designed to equalize the growth rates of productivity. Some of the policies that might be employed to promote the convergence or equalization of productivity growth rates are discussed in Chapter 10, which deals with regional problems and policies in a common market. The limited success of regional policies within nation states suggests, however, that within an economic community the equalization of national productivity growth rates by union policies might be very difficult to achieve. A still more difficult task would be presented if it were a goal of policy that monetary union should be accompanied by convergence of the level of real wages, which would imply a relatively rapid rate of growth of productivity in the lagging areas. If it cannot be assumed that ‘regional’ policies at the union level can be made to produce the required convergence, monetary union among countries with significantly different productivity growth rates may not be politically feasible unless transfer payments are made from the high- to the low-productivity growth members. At the same time, if equalization of the growth rates of productivity among the member states implies a reduction of the growth rate of net incomes in high-productivity-growth countries, the latter will be less likely to see monetary union in itself as an attractive proposition. A decision on their part to join a union with such implications would rationally turn on whether other benefits that may be associated with a monetary union, such as trade gains or gains of a political nature, might be expected to compensate for their loss in real income growth. It should also be noted that, even if a monetary union enjoys balanced income growth, this does not imply that unemployment rates are equalized throughout the union. If the convergence of unemployment rates were regarded as a further essential part of a programme of monetary integration, policies aimed at harmonizing conditions in the national labour markets would also be called for in order to move the long-run Phillips curves into identical positions. Thus, even if monetary integration is capable of generating significant resource allocation gains, they may not be distributed amongst the member states in an acceptable way. 209

MONETARY INTEGRATION

Adjustment costs and shocks The second aspect of monetary integration that needs to be taken into account is the issue of the differences between countries in terms of their exposure to demand shocks and supply-side disturbances such as raw material shocks or productivity changes. Since monetary and exchange rate policy in a union is set at the union level, it can be used to help the union to adjust to symmetrical shocks which affect all member states equally, but it cannot be used to help individual countries (or regions) to adjust to asymmetrical shocks which strike them in different ways. In a new classical macroeconomic model such shocks pose less of a problem, since all markets are supposed to adjust fully and rapidly to shocks, and stabilization policy is therefore largely unnecessary as well as ineffective. However, very strong assumptions are required for these results—namely, perfect market clearing and a high level of information as well as rational expectations—which in combination lack plausibility. Recognizing this, the mainstream of economic thinking does not regard stabilization policy as entirely unnecessary or ineffective: within a single country, policy intervention may be able to influence employment and output favourably in the short run if the economy experiences an upward departure from its natural rate of unemployment as a result of shocks whose source and nature are correctly identified by the authorities and appropriately responded to. Moreover, the natural rate of unemployment may itself be adversely affected over time by the work force’s experience of protracted unemployment (Cross, 1987). This hysteresis effect may provide a further ground for policy intervention, to avoid large and/or prolonged periods of unemployment above the existing NAIRU, which would lead to an upward drift in the NAIRU itself, with important long-term consequences for unemployment. But, in a monetary union of several countries, policy cannot intervene in this way to deal with asymmetrical shocks. For these reasons, there may be significant costs for a country that would be attached both to the fixing of exchange rates within a pseudo-union (since monetary policy is then subordinated to the pursuit of an exchange rate target) and to the establishment of a common currency in a complete monetary union. These costs may be seen as falling into two main categories: (1) the adjustment costs of the transition to monetary integration involved in the achievement 210

MONETARY INTEGRATION

of a common rate of inflation; and (2) the adjustment costs associated with later shocks and disturbances. With respect to the transitional costs of monetary adjustment, if there are differences between the initial inflation rates of the prospective member countries, adjustment costs would necessarily be imposed on some members as a result of the adoption of a common target inflation rate, even if the members were initially on their long-run Phillips curves. The magnitude of these costs will be determined by a number of factors, including: the time horizon adopted for the achievement of the common rate of inflation; the speed at which expectations adjust to actual inflation; the form of the short-run Phillips curve; and the initial discrepancies between the initial rates of inflation of each member country and the union-adopted target rate. If, for instance, a member country initially had a higher rate of inflation than the union-adopted rate, the downward adjustment required on its part could be brought about only by maintaining its unemployment rate above its natural level for a period. Moreover, the extent of the required rise in unemployment would increase as the time period for bringing inflation rates into line was shortened. The firm adoption of the target of convergence of partner countries’ inflation rates might itself facilitate the adjustment of expectations and thus reduce the transitional real cost. There might also be a long-term cost arising from the process of transition if this were to bring about a permanent rise in the NAIRU. Whether it would do so would be a matter of how long, and how much, the actual unemployment rate was held above the NAIRU in order to reduce inflation. The cost would obviously be smaller when the short-run transitional cost itself was smaller. However, if the initial inflation differential was large and expectations did not adjust rapidly towards the union inflation rate, a hysteresis effect could in principle lead to a significant rise in the NAIRU (Cobham and Williams, 1997). After monetary integration has become effective and inflation rates are harmonized, costs may continue to be incurred by individual members as a result of the constraints imposed on monetary policy responses in the face of adjustments required by later localized shocks and disturbances. Their magnitude will depend primarily on the size of the asymmetrical shocks which individual economies experience, on the speed with which expectations adjust, and on institutional and contractual rigidities, which will affect the degree and duration of disequilibria. The extent of asymmetrical 211

MONETARY INTEGRATION

demand shocks should be small, since overall aggregate demand is governed largely by the (symmetrical) union-wide monetary policy. Nevertheless, important asymmetrical shocks could still result from shifts of demand between products where these are produced largely in separate regions. There may also be significant asymmetrical supply-side shocks of various kinds arising from country-specific nominal wage changes or productivity improvements. For both demand and supply shocks, a crucial determinant of the costs they impose will be the flexibility of markets in general and the labour market in particular. If an adverse shock can be absorbed via a reduction in real wages its effect on output and employment will be much smaller. Wage flexibility depends on the nature of labour market bargaining and institutions. Empirical evidence suggests that flexibility appears to be greatest both in markets with very low levels of trade union organization and in those with high levels of centralized collective bargaining. (See Calmfors and Driffill, 1988, for further discussion of labour market issues.) Another factor that will influence costs will be the scope for national stabilization measures through national fiscal policy, which is considered next.

FISCAL POLICY IN A MONETARY UNION When a country joins a monetary union it loses the ability to use monetary and exchange rate policy to stabilize its economy. At the same time the union’s monetary authority cannot address asymmetrical shocks by operating a regionally differentiated interest rate or credit policy. In the absence of fiscal integration and a community stabilization policy operated through the community budget, the only instrument available for dealing with countryspecific shocks will be national fiscal policies. Some of the constraints on the effective use of national fiscal policies that derive from spillovers have already been discussed in Chapter 7. In the context of a monetary union, the orthodox Mundell—Fleming framework of analysis suggests that the domestic effectiveness of fiscal policy would be enhanced. But given that fiscal policy intervention is perceived to be desirable and feasible in national terms, how autonomous can national fiscal policies be allowed to be in a monetary union? The basic issue is whether market forces be relied upon in such a union to effect the necessary consistency

212

MONETARY INTEGRATION

between national fiscal policies and the common monetary policy, or whether some concerted regulation of national fiscal policies by the community is also called for. The argument that in a monetary union national fiscal policies should be flexible and autonomous, except for any requirements imposed by fiscal harmonization of the kind discussed in Chapter 10, is a powerful one. If countries are hit by adverse country-specific shocks that the union’s monetary policy cannot address, and if the lack of fiscal integration rules out any automatic or discretionary transfers through the community budget for stabilization purposes, member states must themselves be able to respond by incurring a budget deficit if the union is not to be subjected to intolerable stresses. Such deficits would automatically arise as tax revenues declined and social outlays increased, and their magnitude could be reinforced by discretionary fiscal changes. In a monetary union the deficit would be financed not by monetary means but by borrowing on the financial market. The issue is whether the extent and timing of such borrowing should be a matter solely for the member state or whether there is also a community interest in the debt position and fiscal stance of members that may justify the imposition of restrictions on their fiscal autonomy. The argument that there is such a community interest centres on considerations of (1) fiscal discipline and (2) fiscal policy co-ordination. Both are related to spillovers, but they are of different kinds. The root of the distinction is that fiscal discipline has to do with the avoidance of a longer-term risk of monetary instability, whereas fiscal policy coordination has to do with a shorter-term focus on stabilization and the appropriate fiscal policy stance of the community in the light of demand and interest rate externalities.

Fiscal discipline Fiscal discipline has to do with the avoidance of unsustainable public debt. If a member state runs a deficit and incurs public debt over a period and at a rate that come to threaten its capacity to service that debt, the debt is ultimately unsustainable. The incentives to engage in deficit financing will certainly be modified in various ways by the institution of a monetary union (Masson and Taylor, 1993). In their light, is is conceivable that on balance the incentives to fiscal discipline may be reduced. One important reason for supposing this to be so is that the sanction of exchange rate 213

MONETARY INTEGRATION

depreciation against imprudent national policies would be lacking in a monetary union. Such policies could pose a threat to the common monetary policy, since the pressure on the monetary authority to bail the member out by monetizing the debt or by other means may be difficult to resist. It may be argued that this danger implicitly assumes that capital markets do not work efficiently. If the market assesses risk accurately, long before that time, any member state would find it progressively more costly to borrow. However, even if the market exerts a disciplinary effect in this way through its assessment of country risk, it cannot be taken for granted that the effect would be sufficient to deal with the problem of credibility if either the market believes that ultimately a community bail-out would occur or, alternatively, if member states themselves fail to respond adequately to the increased cost of borrowing. If, for such reasons, the view is accepted that the market cannot be wholly relied upon to deal with this problem, a decision has to be made on the means to be adopted for protecting the monetary union and the financial system from the effects of excessive deficits by member states. The choice seems to lie between the imposition of rigid ex ante limits on deficits and debt, which have the merit of simplicity but which may hinder legitimate and sustainable deficits, on the one hand, and some more flexible form of fiscal surveillance that would be more clearly defensible in terms of national conjunctural situations and community policy objectives on the other.

Policy co-ordination Co-ordination of policy becomes an issue whenever spillover effects of domestic fiscal policies on partner countries are significant and are not taken into account by member countries. Intra-union spillover effects may pose problems for the union even when governments behave in a prudent fashion. In that context, it is arguable (Emerson et al., 1992; Allsopp et al., 1995) that the danger for a monetary community may be fiscal conservatism, in the sense that too little fiscal stabilization of any kind may be undertaken. The institution of monetary union certainly creates a major new channel through which spillovers operate. As the Emerson report (Emerson et al., 1992) put it in relation to European monetary integration:

214

MONETARY INTEGRATION

the common exchange rate vis-à-vis the rest of the world will be turned into a major spill-over channel, since budgetary policy in any country will affect this EC-wide variable. In the absence of co-ordination, the current account of the Community becomes for Member States a kind of ‘public good’ whose determination is beyond the reach of any single public body. (1992, p. 114) The implications of this point can be seen by supposing that the movement of the common external account demanded a deflationary fiscal policy. In the absence of co-ordination, it is unlikely that any single member would be willing to deflate for the benefit of the group as a whole. Even if a number of states had an incentive to act, others might be tempted to ‘free ride’, and to reap the benefits of others’ action without incurring the costs of taking part in it. In such a situation there would be a strong prima facie case for co-ordination to determine the common budgetary stance of the community as a whole in the light of the common exchange rate and the common external account. If this were not done there would be no means of ensuring that the overall monetary—fiscal policy mix of the community would be appropriate. Co-ordination could still, in principle, leave scope for differentiating among member states according to their individual need at any particular time to engage in stabilization. The weight of these sometimes conflicting considerations in a monetary union will be an empirical matter, bound up with economic structures, the degree of market integration, and the institutional arrangements that are in force. In general, however, if spillovers are significant, their existence will provide grounds for a community role to influence the way in which national fiscal policies are exercised in a monetary union.

STRATEGIES FOR MONETARY INTEGRATION If the net benefits of complete monetary integration in an economic community are perceived to be positive, what strategies might be adopted to bring it about? Vaubel (1978) has distinguished two main types of strategy—co-ordination and centralization—and suggested

215

MONETARY INTEGRATION

that they should be appraised principally in terms of the criteria of gradualism and automaticity. For currency unification to be feasible, progress towards it must be gradual, since otherwise the associated transitional adjustment costs may be prohibitive. At the same time there is a need for automaticity to limit the discretion of participants once they have opted for a particular strategy. Since the crucial objective of monetary integration is to eliminate uncertainty about future exchange rates, automaticity is required to prevent governments from renegotiating or violating agreements or opting out of them. Co-ordination strategies include (1) agreements to reduce the variability of exchange rates by ex post intervention in foreign exchange markets; (2) agreements to harmonize monetary policies ex ante so as to produce exchange rate stability without the need for intervention in the foreign exchange markets; and (3) agreements involving a combination of (1) and (2). Major technical difficulties confront each of these approaches. That consideration apart, although all are capable of conforming to the criterion of gradualism, they all lack the attribute of automaticity and therefore cannot guarantee progress towards monetary integration. Centralization strategies include: (1) the ‘big leap’ solution involving the replacement of national currencies by a community currency at once and completely; (2) free competition amongst the currencies of all member states; and (3) the introduction of a parallel currency—such as a newly created community currency—that would compete with purely national currencies. The ‘big leap’ would clearly conflict with the requirement of gradualism unless it were to follow on from a phase of successful co-ordination. The other two variants of the centralization strategy envisage a gradual but automatic centralization of monetary policy whose pace would be dictated by market forces as the most useful currency displaced purely national currencies. A choice between these options cannot be made on the basis of purely technical criteria. Indeed, strategies of co-ordination and centralization are essentially looking at different issues. On the one hand, co-ordination strategies seek to establish the preconditions for political agreement on monetary integration. Centralization strategies, on the other hand, are essentially to be viewed as relevant alternatives for achieving monetary union once a political agreement on establishing it has been arrived at. Arriving at that agreement, with its implied transfer of political authority, remains the nub of the problem. Ultimately this is the leap that has to be made. 216

MONETARY INTEGRATION

Within the centralization category, attention in recent years has turned more towards what is now sometimes called the ‘institutional approach’ (Gros and Thygesen, 1990) such as is currently being followed by the European Union. It involves a ‘big leap’ as the culmination of a long process of preparation involving convergence between member countries and the creation of the institutions which will be needed to operate the centralized union-level monetary policy. Some economists still argue for different forms of currency competition, but others have emphasized the difficulties for such a strategy posed by the ‘gravitational pull’ towards existing currencies, which results from the fact that a monetary system constitutes a public good rather than a private good (Cobham, 1989).

MONETARY INTEGRATION AND THE EU L’Europe se fera par la monnaie ou ne se fera pas. (Jacques Rueff)

The Treaty of Rome was concerned primarily with the liberalization of trade and factor movements and with establishing a common commercial policy and a common agricultural policy. It contained no specific provisions for the adoption of a common monetary policy. Apart from general statements on the need for co-ordinating economic policies, the treaty confined itself to asserting that countercyclical policy and changes in exchange rates (which at the time and until 1973 were fixed to the dollar within the Bretton Woods international monetary system) should be treated ‘as a matter of common concern’ (Articles 103 and 107). No prior commitment or understanding existed from which negotiations on monetary integration could start.

Early initiatives on monetary union The initiative for economic and monetary union (EMU) in Europe that subsequently developed in the late 1960s was predominantly motivated not so much by any obviously pressing economic necessity for the conduct of the common market itself but mainly by a political desire to provide a focus and a driving force for European integration in the next decade, when the initial task of 217

MONETARY INTEGRATION

establishing the customs union and the CAP would have been completed. However, it was already clear that the Bretton Woods system was likely to disintegrate, and EC member states were worried about the strains which floating exchange rates might pose for the CAP and other EC policies—indeed, special provision in the shape of monetary compensatory amounts (MCAs) had to be made to deal with the problems arising from the French franc devaluation and German mark revaluation of 1969. The monetary integration initiative of the EC has been interpreted by some in terms of a ‘neofunctionalist’ strategy through which spillover effects would force governments to extend their co-operation to new fields of economic activity, culminating ultimately in political union. Whatever its rationale, its proponents were forced to admit that many practical problems would arise during the process. In the debate over strategy and timing which began in the late 1960s, two main viewpoints emerged among the member states, viewpoints which emerged again in the discussions of the late 1980s and early 1990s. On the one hand, a group that included France, Belgium and Luxembourg (termed ‘monetarist’) argued that positive steps towards monetary integration (in the shape of an arrangement to fix exchange rates irrevocably) would themselves strengthen and accelerate the process of economic integration. Member nations would then be compelled to co-ordinate their economic and financial policies, thus reducing disparities in wage and price trends and making full monetary union easier to achieve. The other group of countries, which included West Germany and the Netherlands (termed ‘economist’), took the view that policy harmonization and economic convergence must come first, and that major steps towards monetary integration should not be taken—and indeed would not be feasible—until wage and price changes had converged and structural adaptations to intraCommunity free trade had been completed. West Germany and the Netherlands also argued that premature attempts to fix exchange rates irrevocably might impede rather than stimulate the further liberalization of trade and of capital markets, because some member states would be obliged to retain controls in order to maintain those rates. Their opponents countered by contending that, if capital liberalization were introduced while exchange rates were merely pegged, massive speculative flows could develop with which official reserves could not cope, forcing exchange rates to move in the anticipated directions. The 218

MONETARY INTEGRATION

solution, they suggested, lay in taking a decision to fix exchange rates irrevocably rather than in attempting to approach the objective by a succession of small steps. The compromise initially worked out between these two views was reflected in the Werner report (Werner, 1970), which was broadly accepted by the Council of the European Communities early in 1971. The Werner report proposed the establishment by stages of an economic and monetary union by 1980. There was by then to be a single EC currency, or at least a rigid fixing of exchange rates and the irreversible inter-convertibility of all such currencies. In addition, all capital movements were to be completely liberalized, and a common central banking system was to be established. Until the last moment, however, the separate monetary authorities were to stay in existence, and would retain the right to follow their own monetary policies, to withdraw from the arrangement and to refuse to finance partners in deficit. The first stage of the plan aimed to narrow the margins of exchange rate fluctuations. The concrete outcome of the Werner report was the creation of the ‘Snake’ arrangement, under which EC member states’ currencies floated within margins of 2.25 per cent against each other (which contrasted with the margins of 1 per cent against the dollar under Bretton Woods up to 1971 and those of 2.25 per cent against the dollar from late 1971 to the breakdown of Bretton Woods in early 1973). However, the Snake rapidly became narrowed down to a Deutschmark zone, as the other major currencies—sterling, the Italian lira and the French franc—chose to float separately on their own, leaving only the Benelux countries, Denmark and some nonEC countries linked with Germany, and no further progress was made. The principal reason for the failure of the Snake was that countries (which were far from wholly committed to monetary union) chose to react in different ways to the oil price shock of late 1973, thereby allowing a divergence of inflation rates which made fixed exchange rates untenable. The Community at that time seemed to be moving away from, rather than towards, monetary union, and this was recognized in the Tindemans report (Tindemans, 1976), which tried to maintain some momentum towards monetary integration by advocating that countries within the Snake could move forward even without the others. As a strategy for monetary integration the main weakness of the Snake, or any similar system, is simply that, until an irrevocable commitment is made to staying within it (as 219

MONETARY INTEGRATION

evidenced, for instance, by the pooling of reserves and the establishment of a common central bank), there can be no certainty that national monetary developments will produce uniform inflation rates and stable exchange rates. This would be the case even if the system were to be operated in conjunction with an ex ante harmonization of monetary policies involving the adoption of supposedly consistent monetary growth rate targets. In operational terms there is no means of ensuring that even the full implementation of ex ante monetary targets would be consistent with exchange rate objectives, since the relationship between such intermediate targets and the policy objectives is variable and difficult to predict. Moreover, in the last resort, participants can opt out.

The European Monetary System For some years after the appearance of the Tindemans report, little attention was paid by European policy makers to the issue of monetary union. Roy Jenkins, as President of the European Commission, attempted in his 1977 Monnet lecture to relaunch public debate on the subject and to stimulate action (Jenkins, 1978). But the new European Monetary System (EMS) established in March 1979 was presented primarily as a means of mitigating the increasingly disruptive effects on intra-EC exchange rates that had resulted from pressure on the dollar rather than as a way of proceeding to monetary union. Although it was clear that the EMS might facilitate a later move towards monetary union if it should succeed in promoting a close convergence of economic policies and performance in the EC, its principal declared objective was to establish a greater measure of monetary stability—i.e. lower inflation and more stable exchange rates—in the Community. The EMS was essentially an enlarged system of fixed but adjustable exchange rates. The Exchange Rate Mechanism (ERM) which was its main element differed from the Snake in having a wider membership—including France and Italy, and later Spain, Portugal and (for a short period) the UK. It also involved more consultation on economic policies in general and exchange rate changes in particular, and incorporated larger short-term credit facilities than those available in the Snake. Each participating currency had a central rate expressed in terms of the European Currency Unit (ECU), the value of which was calculated as a 220

MONETARY INTEGRATION

‘basket’ of specified amounts of each member state’s currency. These central rates determined a grid of bilateral central rates around which permissible fluctuation margins of ±2.25 per cent were established (6 per cent for the weaker Irish punt and the Italian lira). At these margins, intervention by all participating central banks was obligatory and unlimited. Adjustments of central rates were subject to mutual agreement by a common procedure involving all countries participating in the ERM and the Commission itself. The ECU served as a numéraire for the ERM, as a means of settlement and a reserve asset for EMS central banks, and as a reference point for the divergence indicator. The latter was a device to ensure that balance of payments disequilibria would be corrected through changes in policies in both surplus and deficit countries, but it was never really used because countries chose to use the system as a means of reducing their own inflation more easily by pegging to the low-inflation Deutschmark. The EMS was also intended to include the creation of a European Monetary Fund, but this proved impossible because of basic differences of opinion amongst the member states with respect to its structure, tasks and powers and the extent to which it should be independent of governments. The introduction of the EMS was linked with a separate arrangement to provide subsidized loans through the European Investment Bank under the so-called New Community Instrument to two less prosperous member states (Italy and Ireland) for the period 1979–83 to finance infrastructure projects in the countries concerned. This was designed to strengthen their economic base and to encourage their acceptance of the disciplines that the EMS would impose. The establishment of the EMS was attended by considerable scepticism. Some of its critics argued that the system would exercise a deflationary influence, while others insisted that it would encourage inflation by serving as a vehicle for the expansion of liquidity and would thus help to make economic divergences more persistent. It was widely predicted that divergent national economic developments would encourage speculative capital movements that would rapidly disrupt the system. That did not happen, at least for the first thirteen years of its existence; the system not only survived, but is widely regarded as having achieved some of its objectives and as having contributed to the decision to proceed further to EMU itself. 221

MONETARY INTEGRATION

The performance of the EMS Any attempt to assess the performance of the EMS since 1979 confronts the difficulty that it is impossible to know how events would have developed in its absence. Enthusiasts are inclined to base their judgements on favourable movements of certain key indicators of policy and performance in the EMS since its establishment, which are imputed to the system. More critical observers have sought to judge success partly by comparing EMS experience with that of the principal non-participating countries during the same period, namely Japan, the US and the UK, where, for example, there were also large falls in inflation. The judgements of some critics on the role of the EMS are also coloured by their mistrust of the argument that exchange rate volatility has had harmful effects upon trade. Five phases in the development of the EMS can be distinguished. In the first phase, from March 1979 to March 1983, which included the second oil price shock, there were seven exchange rate realignments—typically involving devaluations of the French and Belgian francs, the Danish krone and the Italian lira against the Deutschmark—so that exchange rates were not entirely stable, and inflation was relatively high. The second phase, from April 1983 to January 1987, followed from the major change in French macroeconomic policy when the Mitterrand government abandoned its attempt to reflate the French economy on its own in favour of closer integration with other continental countries. During this period there were only four realignments, and exchange rates within the ERM (though not necessarily those between ERM and non-ERM currencies) were more stable. At the same time inflation was lower and different countries’ inflation rates showed signs of converging, reflecting, in part, an increase in the co-ordination of macroeconomic policy between countries. This was the period in which the much debated and much tested idea of ‘German leadership’ and ‘asymmetry’ (in the sense of policy in other countries responding to German policy but not vice versa) was most convincing: other countries were consciously using the peg to the Deutschmark to assist their own disinflations. In the third phase, from February 1987 to August 1992, there were no realignments at all (except the ‘technical’ realignment of the Italian central rate when the lira moved from 6 per cent margins to 2.25 per cent margins in January 1990 and altered its central parity in line with the current market rate), and three more countries 222

MONETARY INTEGRATION

joined the ERM: Spain in 1989, the UK in 1990 and Portugal in 1992, each with a 6 per cent margin. Inflation was on average lower, although there was a rise at the end of the 1980s, and was clearly converging (though part of the convergence was due to the rise in German inflation following the 1990 unification). Capital controls were abolished in most member countries in 1989–90 and, partly for that reason, the ERM seems to have operated in a less asymmetrical way in what Giavazzi and Spaventa (1990) termed the ‘new’ EMS. This phase was also the period during which, following the report (CEC, 1989) of the Delors Committee (made up mainly of EC central bank governors), the decision was reached in the Maastricht Treaty of December 1991 to establish an Economic and Monetary Union (EMU). Although there had long been a widespread (though not universal) aspiration towards monetary union, it was now argued (e.g. by Padoa-Schioppa, 1988) that the abolition of capital controls, necessitated by the desire to extend the single market to financial and other services as well as goods, made fixed exchange rates no longer compatible with national monetary sovereignty and that the latter should be given up in favour of Union-level monetary policymaking. Indeed, there is no question that unfettered capital mobility allows large (possibly speculative) capital flows which can pose serious problems for fixed exchange rate systems, problems that disappear in a monetary union. In addition, the experience of the last few years, with declining use of realignments and growing inflation convergence, had led to a consensus that the EMS had been successful and made EMU appear more viable and less costly.

The Maastricht process The Maastricht Treaty set out the details for a three-stage move to EMU, involving the setting up of a European System of Central Banks (ESCB) which would include existing national central banks and the new European Central Bank (ECB), and the eventual introduction of a new single currency (later named the ‘euro’). The ECB was to be independent of national governments and of the European Commission, and its primary objective was to be that of price stability. Thus the framework for monetary policy under EMU was in accordance with that widely identified in the academic literature as the best guarantee of permanently low inflation; it was also in part designed to replicate the arrangements under which 223

MONETARY INTEGRATION

German monetary policy, with its enviable record for low inflation and stability, had long been operated. The treaty also set out the conditions for admission to EMU. Countries would have to fulfil four conditions. •

•

• •

They must have inflation rates no more than 1.5 per cent higher than the average of the three lowest-inflation countries in the EU. They must have long-term interest rates no more than 2 per cent higher than the average for the three lowest-inflation countries. They must have maintained their currencies within the ‘normal’ ERM margins for at least two years. They must have fiscal deficits which were not ‘excessive’, where that would be judged by the Council of Ministers in relation to the two ‘reference values’ of 3 per cent for the ratio of the deficit to GDP and 60 per cent for the ratio of general government debt to GDP.

The first three criteria are requirements for nominal convergence before entry to EMU. However, critics have argued that, since EMU is going to involve a new common currency and therefore a ‘monetary reform’ as well as monetary union, prior nominal convergence is not really necessary on economic grounds, though it may be necessary on political grounds, as part of the process of making EMU acceptable to the German government and the German electorate (De Grauwe, 1994). The fourth criterion is presumably designed to avoid problems of fiscal discipline and will have to be observed by countries after, as well as before, entry to EMU. However, it is not possible to provide a convincing rationale for the precise numbers of the fiscal reference values, even if—which some critics also dispute—a broad-brush justification for limits of some sort can be established by appealing to the need to avoid an inappropriate fiscal—monetary policy mix for the EU as a whole and to contain undesirable spillover effects in aggregate demand from one country to another, as discussed in the section on fiscal policy above. The fourth phase in the development of the EMS was one of repeated foreign exchange crises, from September 1992 to July 1993, during which period both sterling and the lira left the ERM, a number of other countries devalued (some several times), and it was finally decided to widen the permitted margins of fluctuation. An important 224

MONETARY INTEGRATION

factor in these crises was the high level of German interest rates, itself due to the inappropriate ‘policy mix’—loose fiscal policy and tight monetary policy—in Germany following unification in the summer of 1990. Other factors included the doubts cast on the EMU project by the Danish referendum in June 1992, which rejected the Maastricht Treaty, and the polls leading up to the French referendum of September 1992, which suggested that there might be a majority vote against in France also. Explanations of the crises have also focused on speculative pressures arising from expectations of changes in governments’ policies and on the monetary policies operated in France and the UK in particular (see Mélitz, 1994; Cobham, 1996). The fifth phase that can be distinguished is the period from August 1993, when the EU countries agreed to widen the bands of permissible fluctuation in the ERM from 2.25 per cent to 15 per cent, until the present. At the beginning of this period it seemed to some observers as though the EMS had collapsed, and the prospects for EMU with it, but it soon became clear that countries were not in fact allowing their currencies to fluctate within the full extent of the widened band, that convergence was still an aim of governments’ policies, and that the political will to introduce EMU remained strong. By the second half of the 1990s countries’ attempts to fulfil the convergence criteria agreed in the Maastricht Treaty were exercising a dominant influence on policy. In particular, many countries experienced difficulty in reducing their fiscal imbalances to the levels required. Although the treaty allowed some room for discretion in the decisions on which countries qualified for entry to EMU (decisions due to be taken during 1998), notably on the debt to income ratio, critics of the Maastricht process argued that undesirably restrictive policies in many countries were having harmful effects on unemployment and economic growth. The inflation and interest rate criteria, on the other hand, had by 1997 been met by all those countries that were anxious to enter EMU, while the exchange rate criterion had lost much of its force, for it had become accepted that ‘normal’ now referred to the 15 per cent bands in existence from August 1993 rather than the 2.25 per cent narrow bands at the time of the Maastricht Treaty. At the time of writing (mid 1997) it seemed probable that EMU would commence in January 1999, although whether the initial membership would include all the ‘southern’ countries (Italy, Spain and Portugal) was still unclear. If EMU does go ahead, the EMS will 225

MONETARY INTEGRATION

be seen to have played a fundamental role in bringing about monetary union, despite the criticisms levied against it as a coordination strategy incapable of ensuring continued progress towards the goal. But EMU itself will also have required a final ‘big leap’ in the form of the irrevocable fixing of exchange rates and the subsequent introduction of the euro.

Unresolved issues Although the Maastricht Treaty itself covered a range of detailed issues regarding the functioning of EMU, and the constraints on fiscal policy under EMU were formalized in the Stability Pact agreed at the Dublin summit in December 1996, there are several important issues yet to be resolved. The first of these concerns the exchange rate regime for countries which are seeking entry but are not initially admitted to EMU and for those, like Denmark and the UK, which may wish to remain aloof. Partly because the aspirant countries will naturally wish to peg to the euro, and partly because maintaining unchanged parity within the ERM is one of the Maastricht criteria, the aspirant countries will be expected to participate in an ERM Mark II, but the details of that arrangement have yet to be agreed. All outsiders will be required to treat their exchange rate policies as a matter of common concern and, in that connection, to take account of EMS experience. Fears that the integrity of the single market will be compromised by large exchange rate fluctuations are not groundless, and are certain to expose those countries which choose to remain outside EMU to strong pressure to join ERM II.1 At the time of writing it appears that the initial members of EMU will include Germany, France, the Benelux countries, Austria, Ireland and Finland. Spain, Portugal and Italy may also be admitted. On that basis, the main candidates (or targets) for ERM Mark II membership, apart from the latter three countries, would be Sweden, Denmark, the UK and (perhaps at a later date) Greece. The enlargement of the EU to the Central and Eastern European countries will pose the issue of ERM membership for the new entrants. The position of the UK in relation to this issue is of considerable importance because of the weight of its economy in the EU and the role of London as the EU’s leading financial centre. The UK is not at present a member of the ERM. Following the rise in German interest rates in 1992 that followed German reunification, the ERM 226

MONETARY INTEGRATION

was placed under severe pressure, since the rise required the UK and other countries to accept interest rates higher than were justified by their domestic circumstances. The UK refused to raise interest rates and, in the face of this fundamental inconsistency, speculation developed against the pound, forcing sterling out of the ERM in September. In terms of the Maastricht deficit and debt criteria, the UK would qualify for admission to EMU. The 1997 legislation that gives the Bank of England operational responsibilty for determining interest rates to meet inflation targets that are set by the government also moves the UK nearer to the institutional requirements for entry to EMU in terms of central bank independence and would make it easier for the UK to enter in the first wave if the government wished to do so. This has not been ruled out but seems unlikely, despite the election in 1997 of a government more sympathetic to European integration and the emergence of considerable support for UK participation on the part of business and, to a lesser extent, banking circles. A prior referendum will be necessary in any case. Purely in terms of its structural characteristics, research studies suggest that the UK is one of several EU countries that do not fall definitely within the optimum currency area core of the bloc but, equally, would not be definitely inappropriate members of EMU (e.g. Bayoumi and Eichengreen, 1993; De Grauwe and Vanhaverbeke, 1993). The second issue is that of the exchange rates which should be fixed irrevocably at the start of EMU and which will form the basis for the conversion into euros a few years later. The main choice here is between specific agreed rates which are announced well advance, e.g. at the time when the decisions on membership of EMU are announced, and rates which are an average of market rates over some pre-specified period up to January 1999. The former involves specifying parities which may then invite speculative pressure, but the latter allows drift which could lead to minor misalignments being incorporated into EMU and may also, according to De Grauwe (1996), generate unpredictable volatility at the time when the rule to be applied is announced. The third issue which has yet to be resolved is that of the monetary policy to be operated by the new European Central Bank. Here the main choice is between monetary targeting as practised and advocated by the Bundesbank and some other continental central banks (but abandoned by the UK authorities in the mid1980s) and inflation targeting as introduced by a number of countries, including Finland, Sweden, the UK and New Zealand, in 227

MONETARY INTEGRATION

the early 1990s. The issue has been considered by the European Monetary Institute, set up in Frankfurt in 1994 as a precursor of the ECB, but left for the ECB to decide when it comes into existence later (EMI, 1997). The fourth issue is the relationship between the euro and other major currencies (the dollar and the yen). In principle, this relationship will emerge as a result of the monetary policies pursued in the three main currency blocs. However, the Maastricht Treaty leaves the possibility open that the European Council could enter into a binding agreement like that of Bretton Woods which would fix the external value of the ECU. This seems unlikely to occur except in the context of a world-wide initiative to stabilize exchange rates, of which there is as yet no sign. The EU cannot, however, conclude agreements outside such a system unless those agreements are compatible with price stability. Whatever relationship may emerge between the exchange rates of the three major currencies, the establishment of the EMS is likely to have significant implications for the international monetary system and the world economy in a number of respects. The euro will at the outset become the second most important reserve currency, and if it proves to be strong it could well overtake the dollar. The substitution of the euro for EU national currencies will reduce the number of currencies traded and could in due course reduce the various roles of the dollar in international trade. The establishment of EMU will also have implications for international co-ordination of policy through such forums as the G7 and G10, which will presumably require Europe to speak with one voice in matters of monetary policy. Relations of member states with the International Monetary Fund (IMF) will also have to be resolved. The directions that will be taken remain to be seen, but if the euro establishes itself as a strong currency it will clearly become a central pillar of the international monetary system.

CONCLUSION Any purely economic assessment of the merits of monetary integration and of monetary union must be related primarily to the magnitude of three principal elements of a cost—benefit appraisal and the weight that is attached to them by its prospective participants. 228

MONETARY INTEGRATION

The first element is made up principally of the lasting beneficial impact on the efficiency of industry and finance—and so on output—that can be expected to result from improved capital mobility and investment decisions in the area when exchange risks are eliminated in complete union, or reduced in a pseudo-union. The second element is made up of anticipated real adjustment costs. These take the form of (1) the loss of output and employment that some members may incur in the transition to monetary integration, during which inflation rates have to be aligned, and (2) any differential short-term costs of adjustment to local or unionwide disturbances that would arise even in the longer run after the monetary union has come into effect. Other costs may be associated with the choice of the target inflation rate, not only during the transitional period but in the longer run. These costs and their intraunion distribution would depend on the monetary policies adopted by the monetary authority or authorities with respect to inflation and short-term stabilization problems. The third element in the appraisal consists of the potential costs entailed for certain members by the need for intra-Union transfers of resources in order to ease adjustment costs that may otherwise be perceived to be unacceptably high for certain other members. In the transitional period such transfers may be a necessary political condition for a commitment to monetary integration. If the Union is to endure, such transfers may even be needed indefinitely to facilitate structural adjustments to asymmetrical real disturbances or to promote the convergence of economic performance in the community. Their costs would inevitably fall mainly on its econmically stronger members and would from their standpoint reduce any potential economic attractions that monetary integration might otherwise possess in terms of its efficiency-promoting effects. Evaluations of the merits of monetary integration and of monetary unions in the light of these conflicting considerations cannot usefully be made in general and in the abstract. Judgement must be related first to the structural behaviour and characteristics of particular economies for which the magnitude of the effects and the weight to be attached to them may be concretely evaluated from the standpoint of the Union and its constituent member states; and, second, to the policies that could be expected to be pursued by the monetary authority or authorities of an integrated area. Those policies could to some extent be constrained and predicted by the treaty or monetary agreement that would necessarily have to be negotiated for any form of monetary integration that was 229

MONETARY INTEGRATION

not merely a de facto arrangement resting on the hegemony of a pivotal currency. The European project of monetary union, which in mid-1997 seems likely to commence on 1 January 1999, was the culmination of a long process of debate and of real and nominal convergence between a group of countries which had already undertaken considerable policy integration in a range of other areas. European debates and experience both drew on and contributed to the substantial amount of academic research on monetary integration undertaken in the 1980s and 1990s and earlier. The Maastricht process itself attracted considerable, though by no means universal or uncritical, support from academic research. In the end, it is clear that the political will to create EMU and make it work is of fundamental importance to its realization.

NOTE 1

Following the UK’s departure from the ERM, several cases occurred of TNCs in France and the Netherlands switching production to Britain, which inspired charges of ’exchange-rate dumping’.

230

12 EQUITY, COHESION AND CONVERGENCE: REGIONAL ISSUES

The benefits of regional integration are potential benefits for the common market or economic community as a whole. The question of how those benefits will be distributed is a vitally important issue that will affect the feasibility of integration initiatives and the sustainability and progress of those that have been established. Distributional questions arise at several levels. • • • • •

•

To what extent will different countries and regions benefit? How will the distribution of the gains between capital and labour be affected? How will factor prices be affected? How will different economic sectors be affected? Since industries tend to concentrate in particular regions of particular countries, how will the sectoral impacts of integration affect the regions within member states? In a growth context, can integration be expected to produce convergence or divergence at the level of regions and countries?

In the analysis of the distributional aspects of regional integration, attention has focused principally on the two spatial or regional dimensions, namely the distribution of benefits between countries and its fairness or acceptability, and second on the corresponding implications for regions within countries. Both aspects will be bound up with the market-induced impact of integration on regional disparities in levels and rates of growth of incomes, output and employment between and within member countries. It is these disparities that encapsulate the so-called regional problem in its broadest sense. Regional policies to limit disparities or promote convergence by improving the economic performance of designated regions are 231

EQUITY, COHESION, CONVERGENCE

a means of addressing the problem. The formulation of appropriate regional policies requires an understanding of the causes of spatial impacts and disparities and how particular strategies may affect them. The impact of regional economic integration on the distribution of its benefits among member countries and within them, and more specifically on regional disparities, will depend, apart from policy measures, on the types of market adjustment mechanism that operate. Integration impacts operate primarily through changes in relative prices and costs in member states. The changes in relative prices and costs brought about by integration will affect trade flows (whether governed by comparative costs or by economies of scale and product differentiation), production and consumption, intra-bloc direct investment and labour migration. Accumulation may also be affected. The adjustment mechanisms that operate in these respects at the level of member states vis-à-vis each other and those that operate at the level of regions within member states are not identical; but many of the fundamental factors at work are similar, and to that extent the two problems may be approached by similar analytical methods. There are four main differences between regions within countries on the one hand and member states within economic communities on the other that may be expected to be relevant: • •

•

•

Factors are more mobile within regions of member countries than between them. National economies are much less open than regional economies in the sense that the ratio of their external transactions to their incomes is smaller. Members of economic groups often retain control over their monetary and exchange policies, whereas regions within states do not. National budgets produce a substantial redistribution of income between the regions of member states, but redistribution among member states of economic groupings is much less important.

As integration progresses along the spectrum that ranges from common markets to economic union, these differences between the national and regional contexts of the spatial aspects of integration can be expected to diminish.

232

EQUITY, COHESION, CONVERGENCE

THE REGIONAL DIMENSIONS OF ECONOMIC INTEGRATION The analysis of the distributional and regional dimensions of integration draws on a number of theoretical approaches. A broad distinction can first be made between those approaches which follow a comparative static approach and those which focus on growth. Within each category, three main approaches can usefully be identified. Some of these approaches are entirely lacking in any specific regional or spatial content. It is only in the course of the past decade that attempts have been made to develop models that encompass both the spatial aspects of production as considered in the traditional approaches to location theory and regional economics and the microeconomic foundations of integration theory.

Comparative static analyses The inference from market integration theory The first approach simply brings out the country-wide or spatial implications for costs and benefits of the traditional or new allocational analyses of market integration. The traditional approach considers the implications for the inter-country distribution of costs and benefits of the orthodox partial equilibrium analyses conducted in Chapters 2 and 3. There is no specific consideration of the determinants of comparative advantage. General equilibrium repercussions are disregarded. Transport costs are excluded, as are all other spatial determinants of production. In terms of such analyses there is no presumption that all countries must gain unambiguously from integration. In a simple customs union the outcome will depend on the specific character of the common external tariff and the conditions of production (which govern comparative advantage) and of demand in each country for each product. Where some trade diversion occurs, it is possible for its costs to fall exclusively or disproportionately on certain members. Some members may then suffer an overall welfare loss that will be reflected in their forgone tariff revenues. In the context of the new economics of market integration, where trade is determined by scale economies and imperfect competition, the shifting of production and of rents may also give rise to a situation in which, although 233

EQUITY, COHESION, CONVERGENCE

integration produces gains for the market as a whole, some countries may suffer. Some form of compensation may then be needed if all countries are to benefit. Compensation payments will normally entail allocational costs, since they have to be financed, and the receipts may also have adverse effects on the behaviour of the recipients.

Trade theory and disparities The second mainstream approach that is drawn on to elucidate the regional effects of economic integration is the neoclassical general equilibrium theory of international trade. That theory views income disparities as arising from initial differences in factor endowments which are given. According to this comparative static approach, free trade amongst the members of a customs union on the basis of comparative advantage should promote convergence in factor returns and living standards. In a common market, the effect of factor flows would reinforce this effect. Capital and labour would be redistributed amongst countries and regions, in the manner described in Chapter 6, to the benefit of all countries and regions. This orthodox model depends on a number of highly restrictive assumptions. If they are not satisfied, a reduction in disparities is not necessarily assured. In the first place, the model assumes homogeneous factors of production. That would imply, for instance, that a given quantity of capital can produce the same amount of product in any location. That need not be the case. A better educated or more highly skilled labour force in one location may be able to produce more output from a given quantity of capital than a less well educated or less skilled labour force in another location. The pattern of productivity differentials may thus be more important in determining competitive or comparative advantage than differences in nominal factor endowments. If productivity differentials favour high-income countries, or regions within countries, the effects of integration on the distribution of output may disproportionately benefit those countries or regions. The orthodox theory only implies a reduction of disparities in efficiencyadjusted factor rewards. If workers differ in their productivity levels, and wages are adjusted to those levels, disparities in real wages and incomes need not necessarily be reduced. If wage rates are not kept in line with productivity levels, disparities in unemployment rates will also tend to arise. Other limitations of the orthodox model must be mentioned. First, it assumes perfect competition. There are no internal 234

EQUITY, COHESION, CONVERGENCE

economies of scale which would be incompatible with a competitive structure, nor are there external economies associated with the process of technological advance. Strategic behaviour is absent. Prima facie, therefore, on these counts, the orthodox model is not well suited to illuminate the impact of integration on those sectors of the economy where such characteristics are present and to which the new economics of market integration is more relevant. Its primary limitation from the standpoint of an analysis of regional issues, however, is that the theory provides no explicit treatment of the spatial determinants of inter-regional competitiveness and so of the location of production.

The new economic geography The third comparative static approach is represented by the socalled ‘new economic geography’. This approach, which is particularly associated with the writings of Krugman and Venables, has attempted to deal with the above-mentioned limitations first by adopting the perspective of the new economics of integration, and second by explicitly incorporating a spatial element in the shape of transport costs and other barriers to trade. Later variants encompass the external effects of the concentration of industry and of population upon costs and profits. The core model of Krugman and Venables (1990) demonstrates the possibility that, depending on the extent of barrier reduction, regional integration may harm some participants by widening disparities in production structures and incomes. The basic Krugman—Venables model (1990) focuses on the effect on production in a peripheral country and in a central country of the progressive reduction of trade barriers when economies of scale, imperfect competition and product differentiation are allowed for in a context in which location matters because of transport costs and the importance of market access. The general point that emerges is that the regional effect of economic integration is ambiguous. The reason for the ambiguity can be seen in the following very simplified example. Assume that there is a ‘footloose’ manufactured product that could be produced either in the centre or at the periphery. The demand for the product is completely inelastic, so that total shipments can be taken as given. Suppose that production costs are higher in the centre than at the periphery but that, because there are economies of scale, it would be more expensive to 235

EQUITY, COHESION, CONVERGENCE

produce part of the given output in both locations. However, in addition to production costs there are transport costs. If the good is produced only at the centre, some of the output must be shipped to the periphery. If the good is produced only at the low-cost periphery, a large part of the output will have to be shipped to the centre, where the main market is located. Assume finally that production will be located where the total costs of production and distribution are minimized. Table 12.1, following Krugman and Venables, illustrates a situation that corresponds to these assumptions. Although it is cheapest to produce the good in the peripheral country (Spain), and most expensive to produce in both locations because of scale economies, the costs of transport also affect the outcome. If shipping costs are ‘high’, the cheapest locational structure requires production in both locations, since the savings in transport costs outweigh the extra production cost. However, a reduction in shipping costs does not necessarily cause production to move to the low-cost periphery. A shift from the high to the medium shipping cost case, where costs are reduced by half, will cause production to be located in the centre (Belgium). If, however, shipping costs are completely eliminated, the periphery will represent the least-cost location. The general point is that, while high barriers to trade encourage local production, moderate barriers interacting with economies of scale may encourage a concentration of production in high-cost locations at the centre which enjoy good market access rather than in low-cost locations at the periphery which are remote from the market. In their 1990 analysis Krugman and Venables explore the effect of a reduction in trade costs in a less ad hoc way on the basis of several variants of a highly abstract model. Trade costs are interpreted as a synthetic measure of transport costs and a wide range of other barriers to trade. The model allows for

Table 12.1 Effects of lowering trade barriers on the location of production

236

EQUITY, COHESION, CONVERGENCE

economies of scale, product differentiation and the effects of oligopolistic firms seeking to exploit the markets of partner countries. General equilibrium effects on factor prices and costs are also taken into account and these together with movements in trade costs are the long-run driving force for adjustment. The results of their particular numerical examples in that broader framework are still suggestive of a tendency for the initial process of concentration to reverse itself when barriers to trade fall sufficiently. The fundamental ambiguity thus remains in this broader framework. Lower barriers to trade could make it more attractive to move production out towards the lower-wage periphery, and thereby allow a rise in peripheral wages; or, on the other hand, they could make it more attractive to concentrate production in the centre, requiring a fall in peripheral wages at least relative to those of the centre. Anything that impedes the necessary change in relative wages will reinforce the tendency to concentrate production in the centre. (Krugman and Venables, 1990, p. 74) Esssentially the model suggests that a reduction of trade barriers has a U-shaped impact on the concentration of industry. The outcome to be expected at a particular time depends on where on the U-shaped curve we find ourselves. Economies of scale operate to produce divergence only temporarily, because the economies of scale are internal and can be attained in any location and are not restricted to core economies. Ultimately therefore, the basic Krugman—Venables model, like the traditional neoclassical model, predicts convergence, but the path by which it is reached is different. In other developments of his models Krugman has drawn heavily on the main traditional currents of location theory, including the market potential approach and the idea of circular and cumulative causation. An integrated treatment of his analysis is presented in the appendix to Krugman (1995). A most important point is that the inclusion of agglomerative forces, for instance in the shape of traditional Marshallian cost linkages such that an increase in the number of firms in a location raises the return to other firms, and demand linkages, can reinforce the effects making for concentration, and even exclude convergence in the long run (Venables, 1995). The extension of the analysis from one to several poles also opens 237

EQUITY, COHESION, CONVERGENCE

up the possibility that not only the main pole or poles may benefit but also intermediate locations, notably those in border regions. In many respects the broad picture which emerges from these analyses of the new economic geography is not dissimilar from earlier writings on location that are somewhat outside the mainstream of economics. Krugman’s contribution is to have integrated the ‘outcast’ traditions into a single formal model having a firm microeconomic base and, in the present context, to provide a link with the analysis of economic integration. In so doing, he claims to have validated those contributions. In his own words, however, the jury is still out on the value of his approach.

Growth and spatial effects The second category of analyses of the effects of economic integration on regional disparities looks at the determinants of growth in macroeconomic terms and seeks to explain why the growth rates of different countries and regions may differ, whether growth rates and levels of per capita GDP can be expected to converge, and how integration may affect the process. There are a number of competing approaches which offer different predictions.

The neoclassical model The orthodox mainstream neoclassical supply-constrained approach to growth is characterized by Solow’s model (1956) and later variants. The model employs an aggregate production function that relates output to factor inputs in the shape of physical capital and labour. The function displays diminishing returns with respect to each factor and constant returns to scale. In this model, an economy’s long-run growth path is determined by the growth of the labour force and by technical progress. One basic prediction from the orthodox theory is that, given the technology, which is supposed to be universally available, and the inverse relationship between income per head and the productivity of capital, the level of per capita income across countries should show tendencies towards convergence because, for the same amount of investment, low-income countries should grow faster than high-income countries. The basic model assumes a closed economy, but relaxing this assumption strengthens its convergence properties. The theory 238

EQUITY, COHESION, CONVERGENCE

has nothing to say about the crucial factor of technical progress, which is treated as exogenous. The theory also assumes full employment and lacks any treatment of spatial factors. On all these counts its usefulness in the analysis of regional problems is limited.

The cumulative causation model In marked contrast to the various orthodox neoclassical models there is the cumulative causation paradigm. This approach originated in the work of Myrdal (1957), Hirschman (1958) and Kaldor (1970, 1971) and was in part a response to the failure of orthodox theory to provide answers to basic issues that arise from the economic history of growth and development, including the persistence of divergence. Their analysis clearly foreshadows aspects of the new economic geography that are discussed in the previous section, and of the new growth theory that is discussed below. To explain the persistence of regional divergences Myrdal invoked ‘the principle of circular and cumulative causation’, which is closely bound up with increasing returns. In Kaldor’s words: These are not just the economies of large-scale production commonly considered, but the cumulative advantages accruing from the growth of industry itself—the development of skill and know-how; the opportunity of ever-increasing differentiation of processes and specialisation in human activity. (1970, p. 340) On this view, initial differences become perpetuated by a cumulative movement that is reinforced rather than offset by factor movements and trade. Factor movements themselves can generate increasing returns; thus flows of labour, capital, goods and services from poor to rich regions may occur, which serve not to ameliorate disparities in per capita incomes and growth rates, as neoclassical theory may seem to suggest, but rather to enhance them. In the presence of increasing returns, trade between regions may also operate so as to widen, not narrow, differences in comparative costs and may not necessarily be beneficial to all. The solution to regional disparities is therefore not to be found in the indiscriminate promotion of factor mobility. The process by which the initial advantage of the 239

EQUITY, COHESION, CONVERGENCE

growing region is sustained and reinforced has been termed polarization by Hirschman and backwash by Myrdal. For the lagging areas an important influence on their rates of growth is the induced effects of growth in the prosperous centre. These effects are of two kinds: (1) the unfavourable effects that result from polarization; and (2) the favourable effects, termed spread by Myrdal and trickle-down effects by Hirschman. The latter include such repercussions as the increased demand for imports generated by growth at the centre, and the effects of the diffusion of technology from the more advanced centres. If spread effects were sufficiently strong, all countries or regions could benefit from growth at the centre, and the problem of inequalities would become one of differential rates of progress. Hirschman argued that the spread effects would eventually dominate the backwash effects, so that growth is initially divergent owing to polarization but eventually becomes convergent as external diseconomies build up at the core. Myrdal took the more pessimistic view that spread effects are likely to be weak, so that, failing government intervention, the success of fast-growing areas will actually inhibit the development of the others. Myrdal has seen this process as operating for long periods of time, but he too envisaged the possibility that, ultimately, natural checks would be imposed on the growing areas as a result of diseconomies associated with high rates of industrial growth, including congestion and other environmental problems. A more developed version of the cumulative causation theory has been presented by Kaldor (1970, 1971) in specific application to the regional issue. Its essential features may be briefly summarized. Resource increases are primarily self-generated, being determined by the rate at which technical progress is embodied in capital equipment, which in turn is primarily determined by the rate of growth of output. In contrast to the orthodox neoclassical approach, the long-run growth of output is constrained by the rate of growth of autonomous demand and not by factor supplies. In a regional context the main autonomous demand factor is the demand for exports, which in turn induces new investment in the region. The behaviour of exports, and of a region’s production, depends on two factors: (1) the rate of growth of world demand for the region’s products; and (2) the movement of ‘efficiency wages’ in the region relative to other producing regions. The movement of efficiency wages depends on the relative movement of money wages 240

EQUITY, COHESION, CONVERGENCE

and of labour productivity. If this relationship (the index of wages divided by the index of productivity) moves in favour of the area, it will gain in competitiveness, and vice versa. There is empirical evidence to suggest that money wages can be expected to move in a similar fashion in each region, even though regional growth rates of employment vary widely. However, because of increasing returns, higher growth rates of labour productivity will be experienced by regions with faster rates of output growth and particularly of export growth. This is the relationship known as Verdoorn’s law (Verdoorn, 1949) which states that:

. . where p = growth of productivity, Q = growth in output, . a = autonomous productivity growth and λ = the Verdoorn coefficient. Thus efficiency wages tend to fall in areas where output and so productivity are increasing relatively rapidly, allowing those areas to acquire a cumulative competitive advantage over relatively slowgrowing areas. This is the mechanism through which, according to Kaldor, the process of cumulative causation works. A formal model of regional growth differences on Kaldorian lines has been developed by Dixon and Thirlwall (1975). This clarifies the role of the Verdoorn effect in contributing to regional growth differences, helps to elucidate the role of structural factors in explaining growth rate differences, and throws light on the question of whether regional growth rate differences will tend to narrow or diverge through time. The model shows a region’s equilibrium growth rate as depending on four main factors: (1) the income elasticity of demand for exports; (2) the price elasticity of demand for exports; (3) the rate of inflation; and (4) the Verdoorn effect, which itself operates through the price elasticity of demand for exports. Regional differences persist because of the sustaining influence of the Verdoorn effect, which provides the link between exports and growth via prices and completes the circle of circular and cumulative causation. It is essentially the Verdoorn relationship that makes the model circular and cumulative and gives rise to the possibility that, once a region obtains a growth advantage, it will, other things being given, keep it. Suppose, for instance, that a region obtains an advantage in the production of goods with a high income elasticity of demand, which 241

EQUITY, COHESION, CONVERGENCE

causes its growth rate to rise above that of another region. As a result of the Verdoorn effect, productivity growth will be higher, the rate of change of prices will be lower (assuming the rate of growth of money wages to be the same in both regions), the rate of growth of exports (and hence the rate of growth of output) will be higher, and so on. Moreover, the fact that the region with the initial advantage will obtain a competitive advantage in the production of goods with a high income elasticity of demand will mean that it will be difficult for other regions to establish similar activities. The formal model just described considers each region in isolation from the others, and takes interregional relations into account only through the Verdoorn effect. There is no explicit analysis of spatial and locational aspects. From this point of view it is narrower than the theories of Myrdal, Hirschman and Kaldor, who all explicitly recognized the importance of interregional relationships that operate through the spread effect. The less formal approach of those earlier writers also enabled them to encompass more of the complexity of the development process than is possible in a formal model, and their analyses contain many useful insights. Kaldor, for instance, has argued, by analogy with oligopoly theory, that cumulative causation is likely to lead to the concentration of industrial development in several successful regions rather than in one. It is then possible that those regions may hold each other in balance through increasing specialization between them, some becoming more prominent in some industries and some in others. Kaldor’s approach was not widely followed, however, in part because his emphasis on increasing returns was perceived to be incompatible with competitive markets and stable equilibrium.

New growth theory The third and latest theoretical approach that has some bearing on regional development and policy is new growth theory. The mainstream neoclassical growth theory had very little to say about technical progress, which is the mainspring of economic growth. Since it also failed to throw any light on a number of crucial questions concerning the historical process of growth, it was widely regarded as sterile. From the mid 1980s, however, the so-called new theory of endogenous growth that emerged from the writings of Romer (1986), Lucas (1988) and others has led to a revival of 242

EQUITY, COHESION, CONVERGENCE

interest in growth theory and to attempts to elucidate the determinants of the rate of technical progress. Endogenous growth models fall into one of two main types. Both hypothesise an aggregate production function that exhibits increasing returns to scale, but the increasing returns derive from externalities. Although individual enterprises face constant returns to scale, a positive production externality is postulated to give rise to increasing returns at the aggregate level. In one type of model the elasticity of output with respect to aggregate capital is unity, implying increasing returns to capital and labour together. Although individual firms still face a constant returns production function, their overall efficiency level is a function of the aggregate capital stock in the economy. In a second type of model, the existence of a specific growth factor which raises the total productivity of other factors is hypothesized. The growth factor may consist of human capital (Lucas, 1988; Romer, 1989), or knowledge from R&D. From the specific standpoint of regional policy there are two aspects of the new growth models which are of particular interest. First, most of these endogenous growth models imply that an optimal allocation of resources is not achieved by the market. For instance, in the first class of model, efficiency would be greater with more investment but, since part of the benefit cannot be appropriated by the private investor, he ignores it in his calculations. The second important point is that it can be shown that in terms of most models there is no presumption that per capita incomes will converge across countries because the marginal product of capital does not decline as countries get richer, so that they do not grow more slowly.

THE RATIONALE OF REGIONAL POLICY IN AN ECONOMIC COMMUNITY The broad thrust of modern theorizing that has been reviewed in the previous section suggests that: (1) it is not necessarily the case that all countries gain from a process of customs union or market integration; (2) there are specific reasons why various regional disparities in a common market may persist and even widen for long periods of time; (3) there is no presumption that convergence can be expected. But why should an attempt be made to influence these disparities through regional development initiatives? Not all

243

EQUITY, COHESION, CONVERGENCE

spatial disparities are undesirable. There are advantages from the concentration of industries in particular locations that derive from externalities in the shape of technological spillovers and other local external linkages such as are documented for Silicon Valley and other locations. Although some member states and regions within them may suffer from a widening of disparities or polarization as markets and spillovers interact, there is no implication in any of the theories reviewed that the community as a whole does not benefit on balance in terms of income and growth from the economic geography that results. To that extent, it may appear that, although compensation may be necessary if all are to gain, a regional policy involving intervention to reduce disparities by encouraging dispersion and more balanced development will necessarily conflict with the basic rationale of economic integration, namely, income gains from improved resource allocation. This is not necessarily implied if negative as well as positive locational spillovers are also present and constitute significant resource costs. But even if that is not the case, other economic goals will have to be taken into account in considering whether there are economic arguments for regional policies or whether the adoption of regional policies simply implies a trade-off between the pursuit of economic goals and political considerations. At a national level, elements of an economic case for the pursuit of regional policies may be founded on three principal considerations, subject to the usual public choice qualifications. The first is a resource allocation argument, which asserts that market forces alone may fail to operate satisfactorily in relation to location decisions. The hypothesis is that the externalities and spillovers of location decisions justify policy intervention in order to ensure an efficient economic framework within which private locational decisions can operate optimally. Congestion in the centre is a traditional example of the possible need for policy deterrents. To the extent that the diseconomies of congestion are external to the individual producer and are not fully reflected in selling costs or prices, concentration in growing areas may be carried much further than is desirable on efficiency grounds. On the other hand, beneficial externalities in the shape of technological spillovers and labour market linkages of the kind much stressed by Marshall, Kaldor, Krugman and some new growth theorists might suggest that there may be too little polarization for maximum efficiency rather than too much. The second consideration is a resource 244

EQUITY, COHESION, CONVERGENCE

utilization argument, which orthodox integration theory abstracts from by its assumption of full employment. Output and growthmaximization considerations require that an economy’s resources should be fully utilized. If labour is immobile, and its price fails to reflect its opportunity cost, that may justify intervention through regional policy to influence the level of utilization. The third consideration involves stabilization. The avoidance of inflation and the achievement of steady growth may be mutually compatible only if large differences in regional unemployment rates can be avoided. The existence of large regional disparities in unemployment rates is likely to affect a country’s ability to control inflation by making it difficult to implement general measures that avoid excessive inflationary pressures in some areas where employment is high, since they may imply unacceptably high rates of unemployment in others. Regional policy measures may from this standpoint be regarded in part as an attempt to manage the level of aggregate demand on a more selective basis than is possible by the use of orthodox fiscal and monetary measures. The case for regional policies at a national level rests, however, not on purely economic arguments such as these but, perhaps mainly, on political and social considerations. Wide economic disparities generate political tensions and are felt to be inconsistent with the equality of economic opportunity that a common citizenship implies. Labour mobility will not necessarily alleviate regional disparities. Moreover, even if increased labour mobility can reduce regional income and unemployment disparities, the political and social costs of major geographical movements of the population may be unacceptable. It is a combination of such political and social considerations with the previously outlined economic arguments that has historically led to the adoption of regional policies in so many countries. If there is a case for regional policies on such grounds at the level of member states, it will not be diminished by membership of an economic grouping. It may indeed be reinforced if integration encourages polarization or if community policies have an adverse sectoral effect. The establishment of an economic community does, however, raise the issue of the assignment of responsibility for dealing with regional disparities. Should that responsibility be left with the national governments or assigned to the community or should both levels have a role? In the latter event, it would be necessary to co-ordinate the policies of community and member states. 245

EQUITY, COHESION, CONVERGENCE

Where the object of policy is to influence inter-country disparities, the issue is clear. The fundamental case for the adoption of a regional policy at the level of an economic community stems essentially from considerations of cohesion. However satisfactory the initial integration arrangements may be for all member states, the dynamic effects of integration cannot be foreseen. Although they should be favourable for the group, they may not benefit each member. If the operation of an economic grouping enhances the forces of polarization between countries, or if for any other reason disparities grow between them, the necessary political basis of integration may be eroded or disappear. Only the community can address this issue of cohesion, since whatever specific strategies are adopted its implementation will ultimately require a transfer of resources between member states. Also, community sectoral policies may exacerbate regional problems. Only at the community level can this problem be addressed. Where the object of regional policy is instead to influence intracountry disparities, the appropriate assignment of responsibility is not so obvious. The fact that disparities between the regions of each member state constitute an important element of the disparities between the regions of a community and, as such, are of concern to the community as a whole (through its common interest in output and growth maximization and economic stabilization) does not provide an unambiguous conclusion. There are on the one hand two principal arguments that may tell against the assignment of regional policy to the level of the community. First, there are likely to be differences in preferences among member countries with respect to reducing regional disparities of different kinds. A centralized policy at community level would not be able to take these into account. The second argument relates to policy efficiency. The direct operation of regional policy at the central level may be feasible, but implementation at the level of national governments that are closer to the problems and have more information may be more efficient. There are, on the other hand, several reasons for not leaving regional policies exclusively in the hands of member states. First, some kinds of national regional aids may be incompatible with the working of the common market because they distort cross-border competition. Second, national funding ability may not correspond to the severity of the problem. Exclusive reliance on member states’ own policies may therefore fail to contribute to the reduction of interregional disparities. If the poorer countries are left entirely to 246

EQUITY, COHESION, CONVERGENCE

finance their own regional policies, while the poorer regions of richer countries receive generous assistance although they may be as rich or richer than the poorer countries, the outcome may be increased rather than decreased inter-regional disparities. Third, there may be gains from co-ordination where spillover effects from regional development are present. On the basis of such considerations, a legitimate common interest in inter-regional disparities in income, employment and growth can be identified in economic communities that implies a specific role for community-level policies aimed at containing and reducing such disparities. At the same time, since no a priori argument can be made out in terms of assignment considerations for the centralization of national regional policies, the co-ordination of the two levels of intervention will be needed. There will be a specific role for a community policy to co-ordinate national policies (1) to prevent regional aids from causing distortions of competition, (2) to ensure that any community sectoral policies do not exacerbate regional problems, and (3) to enable regional policy to contribute more effectively to a reduction in disparities over the community as a whole. The relative importance attached to the regional policy of a community vis-à-vis national regional policy, and also its specific character, must clearly depend on the degree of integration aimed at or achieved. In a simple customs union a policy limited to avoiding distortion by measures of incentive harmonization may suffice. In a common market, where large-scale labour and capital flows may occur, there may be a greater need to introduce community resource transfers to influence the scale and character of those flows or to deal with their spatial and financial repercussions. In an economic and monetary union there would be grounds for assigning a still greater role to community regional policy, since national policy instruments for dealing with regional problems would then be greatly reduced or constrained and, in addition, the sustainability of a non-inflationary monetary policy might otherwise be jeopardized.

THE IMPLEMENTATION OF A COMMUNITY REGIONAL POLICY In order to implement a concrete regional policy at community level, decisions have to be taken on a number of issues. Objectives 247

EQUITY, COHESION, CONVERGENCE

must be specified and a strategy chosen to achieve the objectives. If responsibility for carrying out policy is decentralized, grant mechanisms will have to be devised and the criteria specified for determining eligibility for benefit. Lastly the financial resources needed for carrying out the policy have to be estimated and provided. The design of policy also needs to address the issue of additionality (to ensure that that community funding does not simply replace nationally funded expenditure that would have been made anyway) and the choice of strategy should reflect the need to minimize the possibility of adverse incentive effects (moral hazard). The broad objective of a regional policy is to improve economic performance in designated regions. More specifically, a convergence policy aims to generate extra growth in lagging regions so as to reduce some part of their real income difference vis-à-vis the community average over a certain period of time. The primary difficulty in determining a regional strategy for this purpose is that the determinants of regional income and growth disparities are complex and not fully understood. The discussion in earlier sections of this chapter has outlined several approaches that throw some light on the issues. The broad policy implications that derive from each approach are different. Orthodox neoclassical models generally point simply to the need to ensure the efficient functioning of markets if convergence is to be encouraged, but policy can have no long-run effect on growth (unless it raises the rate of growth of the labour force). Cumulative causation theories, on the other hand, such as those of Kaldor, Dixon and Thirlwall, that are rooted in macroeconomic models do allow a potential role for specific regional policies to reduce disparities. So do the recent microeconomic analyses presented by Krugman. Endogenous growth theories, in their turn, envisage the possibility of a failure of convergence and point to factors such as infrastructure through which growth might be influenced. Specific proposals arising from any of these approaches that could be directly translated into policy are, so far, few. In general terms the factors identified as significant for growth point to the desirability of relating regional policy incentives to activities that have favourable structural characteristics rather than to particular factors of production, whether labour or capital. In terms of cumulative causation, for instance, it was argued that greater concentration on activities with high income elasticities of demand should be encouraged in peripheral regions or countries. Krugman (1989) has also emphasized the role of this factor. Fundamentally, 248

EQUITY, COHESION, CONVERGENCE

a community strategy for raising the growth rate of a region or a country and its level of income must aim at encouraging a relatively rapid rate of growth of productivity in the regions in question. Ultimately, in the absence of ever increasing budgetary transfers between regions and countries, it is only by such means that divergences in real incomes can be reduced. The problem is to identify the enabling factors and to assess their significance in an operationally useful manner. Little serviceable guidance can be derived from export-led growth models in themselves or, so far, from the models presented by the new economic geographers. Recent analyses using cross-country regressions (Thirlwall and Sanna, 1996) throw some light on the significance of the various factors, notably investment, that are claimed to affect growth in the competing theoretical approaches. Pugno (1995), for instance, finds the factors stressed in the Kaldor—Thirlwall approach to be significant. That, however, is a far cry from the provision of any useful policy guidelines. At first sight, more concrete indications of appropriate policy could be derived from some of the new growth models. Thus, if specific factors such as R&D, human capital or public infrastructure can be shown to be bound up with higher productivity growth rates, then, since it is implied that there is divergence between private and social costs due to externalities, a case could be made out for regional subsidies or other forms of intervention to raise investment in those areas. However, the theories themselves provide little guidance on the relative importance of the different growth factors and of their effects, such as would be necessary if specific subsidy policies were to be justified and framed. The new growth theory is largely at the stage of showing how various factors could influence growth rather than of demonstrating how certain factors do have such an influence and to what extent (Boltho and Holtham, 1992). Whatever the specific character of the strategy adopted by an economic community to limit disparities, if the administration of policy is decentralized for reasons of efficiency, its implementation will require the provision of grants from the community to the member states. The object of such grants is to promote additional spending on specified purposes—for instance, public infrastructure—over and above what it would have been without the grant. Two problems arise in this connection. The resource allocation effect is normally aimed at by the provision of conditional matching grants. The first difficulty is that in operational terms 249

EQUITY, COHESION, CONVERGENCE

additionality is difficult to ensure, since strictly it refers to a purely hypothetical alternative. In practice, therefore, grant aid may be partly fungible and may be used to support general expenditure that would have been made anyway, rather than for expanding the targeted categories of expenditure. Its resource allocation effect will to that extent be reduced. The second difficulty has to do with the possibly adverse incentive effects of regional strategies. The transfer of resources to peripheral, low-income or lagging areas may in certain respects slow down rather than accelerate a reduction of disparities if they turn out to have a negative impact on the incentive to migrate, work or innovate. This danger is likely to vary with the type of strategy adopted. It would be a factor that could hardly be ignored in a strategy of regional wage subsidization, for instance, but in relation to subsidies for the provision of public infrastructure its significance may be more doubtful.

REGIONAL POLICY IN THE EUROPEAN UNION

The evolution of policy From its inception the EU has had a concern with regional development. The preamble to the Treaty of Rome expresses a common desire to ensure harmonious development by reducing the differences between the various regions and the backwardness of the less favoured regions—in a word, what is now termed convergence. These allusions, however (as Robert Marjolin has remarked of many similar Community declarations), represented an act of piety rather than a commitment to specific action, and they were not translated into the adoption of a common regional policy. Instead, regional policy was initially left entirely to the member states. Under the treaty itself, the Commission was merely assigned the ‘negative’ function of overseeing state regional aid. From the time of the Community’s establishment in 1957 until its first enlargement in 1973, the principal achievement of the Commission in the area of regional policy was the establishment and implementation of principles for the co-ordination of national regional aid. Here the Commission’s primary concern has been to prevent national aid from causing undue distortions of competition 250

EQUITY, COHESION, CONVERGENCE

and to render regional policies more consistent with Community objectives. The principles first laid down on regional aid harmonization in 1971 reflect these and other concerns, and have been gradually redefined over the years. The subsequent attempts to develop a more positive, coherent and autonomous regional policy have been shaped by two main factors: (1) the successive enlargements of the Community, which have increased the scale of regional disparities; and (2) the deepening of EU policies as reflected in the Single Market programme and the initiative on EMU. Both these initiatives have been perceived to have potentially adverse implications for convergence and for the distribution of the benefits of closer integration. The first enlargement of the Community took place in 1973 with the accession of the UK, Denmark and Ireland. It was followed by the southern enlargements of the 1980s, with the accession of Greece in 1981 and of Portugal and Spain in 1986. More recently, in 1995, Austria, Sweden and Finland have acceded. Widening, and in particular the southern enlargements, has greatly increased the number of relatively underdeveloped regions in the EU and the size of key disparities. With the southern enlargement, moreover, a geographically distinct pattern of disparities has emerged, with the poorest countries and regions predominantly located on the southern and western periphery. If the initiatives to widen the EU further, to include some of the countries of East and Central Europe, are successful, regional disparities will become even wider. The first enlargement was followed by the creation in 1975 of the European Regional Development Fund. The ERDF was the first Community financial instrument specifically designed as a vehicle for regional policy by providing grant support for regional policy expenditures of member states on infrastructure and industrial projects. The main limitations of the original ERDF as an instrument for reducing disparities were apparent from the outset. The fund was small in size both in relation to the scale of regional problems and in relation to the level of regional expenditure by member states. The quota system, which gave all countries an allocation, dissipated its modest impact. The provision of conditional specificpurpose grants was supposed to influence the level and character of regional expenditures but additionality was difficult to ensure. In fact the fund largely amounted to the provision of disguised block grants (CEC, 1993a), the distribution of which was to some extent bound up with the need to address the issue of equity in relation to the net contribution of certain member states to the 251

EQUITY, COHESION, CONVERGENCE

Community budget. Finally, the fund could only respond to national initiatives and decisions in the field of regional policy. It was not empowered to support initiatives of the EC itself. The fund was thus merely an appendage to the regional policies of the member states. In an attempt to overcome these constraints, and to provide a basis for a more active and comprehensive Community regional policy, the Commission put forward a number of proposals for improvement in 1977, 1981 and 1984. Some were adopted, but others were resisted. Major reform and funding improvement had to wait until the late 1980s. Since the late 1980s, notwithstanding the greater disparities that have accompanied the southern enlargement, the EU has pressed forward with initiatives to deepen economic integration, first through the Single Market programme and, second, through the proposals for economic and monetary union. These have had important repercussions on regional policy. Concern that a single market may increase regional disparities has led to the provisions on economic and social cohesion in the Single European Act. For the first time the Act explicitly incorporated the aim of economic and social cohesion, defined as promoting ‘the harmonious development of the Community and a reduction in disparities between the various regions and the backwardness of the least-favoured regions’ as a treaty objective and provided the political basis for the major reform of the EC’s structural policies in 1988. The 1988 reform of the structural funds, which largely governs the present operation of regional policy, represented a concerted attempt to deal with some of the more important limitations of the ERDF and to reshape regional policy in the context of the perceived implications of the Single Market programme. These reforms, which were paralleled by a substantial increase in financial provision from 1989, also reflect a renewed attempt on the part of the EU to assume a more autonomous role in relation to regional policy. The centrality of regional policy was further emphasized following the 1992 Treaty on Economic and Monetary Union, which made the strengthening of economic and social cohesion alongside the creation of the single market and economic and monetary union one of the central pillars of the Union. The treaty also created a Cohesion Fund to help the four less prosperous member states to adopt the budgetary disciplines required in preparation for EMU while enabling them to maintain their expenditure on disparity-reducing measures. A further reform of the structural funds occurred in 1993 when new regions (mainly in Germany) were added and a new objective was established to 252

EQUITY, COHESION, CONVERGENCE

facilitate the adaptation of the labour force to industrial change and restructuring. Provision was also made for a further substantial increase in spending on the structural funds up to 1999.

Inter-country and inter-regional disparities in the EU What is the present pattern and character of regional disparities in the EU, and how has it evolved? The enlargement of the EC from six countries with a population of 175 million to fifteen countries with a population of 370 million has been accompanied by an increase in its economic diversity. Major differences now exist in living standards between the member states and regions in terms of income per head, employment and productivity (Tables 12.2 and 3) and in growth performance. These disparities, moreover, exhibit a pronounced core—periphery pattern in regional terms. The high-income regions are clustered within Germany, Denmark, the Benelux countries, France, northern Italy and parts of the UK. Around this core lies a periphery of regions with markedly lower per capita incomes, notably in the Mediterranean area, Ireland, the northern and western parts of the UK and eastern Germany.

Disparities between member states At the member state level income per head (as measured by per capita GDP in PPS terms)1 is, in 1995 terms, significantly—10 per cent or more—above the EU average in Belgium, Denmark, Luxembourg and Austria. It is also above average in Germany, France, Italy and the Netherlands, around the average in the UK and slightly below the average in Sweden and Finland. The remaining four member states have income per head between 64 per cent of the EU average (Greece) and 90 per cent (Ireland). The average income per head in the two poorest member states, Greece and Portugal, is about 40 per cent below that of the four most prosperous member states considered together. This is a considerably wider gap than existed before the first enlargement of the Community in 1973, when GDP per head in Germany at one extreme was only 25 per cent above that in Italy at the other. Although enlargement obviously increased absolute disparities at the time, a key issue is whether relative growth rate differences 253

EQUITY, COHESION, CONVERGENCE

have produced a narrowing of absolute disparities in levels of income over time. Growth rates in poorer countries have typically been higher, but the differences have to be substantial for convergence in levels of income to occur. Over the decade to 1993, significant differences in growth rates have been experienced in the EU. Four member states, namely Spain, Ireland, Luxembourg and Portugal, have grown on average significantly faster than the

Table 12.2 Regional disparities in income, productivity and unemployment in the EU, 1983 and 1993

Source: CEC, First Cohesion Report, 1996, p. 131. Notes a For unemployment, highest unemployment rate/lowest unemployment rate. GDP is measured in purchasing power standard.

254

EQUITY, COHESION, CONVERGENCE

rest. In the cases of Spain and Portugal, their growth accelerated after accession. For the rest, apart from Greece, Finland and Sweden, growth was close to the average. Finland and Sweden experienced a marked deterioration in their relative positions (CEC, 1996a). The outcome of the relatively rapid growth of three of the four poorest members (the so-called ‘Cohesion Four’) is that they improved their relative position from an average of 66 per cent of the Union GDP per head in 1983 to 74 per cent in 1993. For the EU as a whole, the standard deviation, reflecting the dispersion around the average of incomes per head between countries, fell from 17.2 in 1983 to 12.8 in 1993 (Table 12.2). In addition to disparities in levels of GDP per capita and in growth rates, there are also considerable disparities in unemployment rates, both at the national and at the regional level. Table 12.3 shows that unemployment rates varied between countries in 1993 from less than 5 per cent in Luxembourg and Austria to 16 per cent in Ireland, 18 per cent in Finland and as much as 22 per cent in Spain. In general, economic growth has been less employment-intensive in the EU than in, for example, the US, where employment growth has been 1.5 per cent a year over the last decade, whereas in the EU productivity growth of around 2 per cent a year has generated an increase in employment of only 0.5 of a percentage point a year. The 1995 enlargement added three member states where throughout the 1970s and 1980s unemployment was very low, but the fall in GDP in Finland (associated with the collapse of trade with the USSR) and in Sweden produced large rises in unemployment from which some recovery has subsequently been made. The highest unemployment at country level is that of Spain, where between one-sixth and one-fifth of the labour force has been affected since the beginning of the 1980s. Of the other members of the ‘Cohesion Four’, Ireland has continued to experience a high rate of unemployment despite a GDP growth rate of nearly 5 per cent during the decade to 1993. Portugal also experienced rapid economic growth but was able to reduce its already low rate of unemployment despite little employment expansion because the labour force increased only slowly. Its rate of unemployment has consequently remained below the EU average. Greece has enjoyed below-average growth, but low labour force growth has helped to keep nominal unemployment below the average of the EU though higher than in Portugal. 255

Source: CEC, First Cohesion Report, 1996, p. 131.

Table 12.3 Regional disparities in income and unemployment in the EU, by member state, 1983 and 1993

EQUITY, COHESION, CONVERGENCE

Disparities between regions Although there are considerable differences between member states in levels of income, in unemployment and in growth rates, such economic disparities are much more prominent at the regional level, and are particularly evident between the EU’s centre and its periphery. Income per head is below or well below average in all the southern peripheral Mediterranean regions, including southern Italy, as well as in those on the eastern and northern periphery (in eastern Germany and northern and eastern Finland) and on the north-western periphery. They are well above average in a cluster of regions in northern Italy, southern Germany and Austria, and above average again in the Benelux countries and northern Germany. Disparities can be measured in a number of ways. A simple comparison between regions with the highest and lowest levels of income per head (in PPS terms) shows that the average level in Hamburg, the most prosperous region of the EU, was four times that of the Epirus in Greece (CEC, 1996a). Taking more representative groups, Table 12.2 shows that a comparison of the ten richest and the ten poorest regions indicates that in 1993 average GDP per head in the former was three times higher than that of the latter. The southern enlargement was responsible for more than doubling, to some 50 million, the total number of persons living in low-income regions (defined as those with a per capita GDP of less than 75 per cent of the EU average). At the regional level, growth in GDP has varied markedly between regions over the decade. The outcome is that the absolute difference between the ten richest regions and the EU average has widened, but overall, as indicated by the difference in the standard deviation computed on a regional basis in Table 12.2, little net change has occurred. With regard to unemployment, disparities in unemployment are most acute at the regional level. Thus Table 12.3 shows that, in 1993, the ten worst affected regions had an unemployment rate averaging 26.4 per cent, nearly seven times the average for the twenty-five least affected regions (4.6 per cent). The changes over time in these groups of regions are revealing. For the least affected regions average unemployment was virtually the same in 1983 as in 1993, but for the worst affected regions the picture is different. For the ten worst affected regions the average unemployment rate increased significantly, from 19.4 per cent in 1983 to 26.4 per cent in 1993, an increase of seven percentage points. The tendency for 257

EQUITY, COHESION, CONVERGENCE

disparities in unemployment to widen over time is confirmed by statistics for all regions (CEC, 1996a). The long-term trend for regional differences to increase, which dates back to the mid 1970s, was interrupted by the economic upturn between 1987 and 1990. The subsequent recession led to unemployment increasing throughout Europe, which was associated with a substantial widening of disparities that continued until 1995. Regional dependence on particular sectors has an important bearing on unemployment performance. Consequently the regions worst affected by unemployment are not always the same as ten years ago. In particular, the regions particularly dependent on the primary sector, notably in Spain and southern Italy, have become much more prominent. Some insight into the underlying causes of regional disparities in income per head can be obtained by decomposing it into two components: productivity in the sense of GDP per person employed, and the proportion of the population in employment. GDP per head is the product of these two factors. There are significant disparities in average productivity and employment between member states and between the lagging regions (CEC, 1996a; Dignan, 1995). This makes it unlikely that any single policy prescription can provide a remedy for such disparities. For instance, average labour productivity in Spain exceeds that in both the UK and Denmark and is quite close to the EU average. The problem of Spain centres on its high unemployment rate. By contrast, in both Greece and Portugal, low productivity is the main factor underlying the large gap between the GDP per capita in those countries and the EU average, since the employment rate in Portugal is substantially above the UK average, while that in Greece is not far short. It must therefore be the case that considerable underemployment exists in both Greece and Portugal. Underlying the productivity gaps (Table 12.2) are marked disparities between different parts of the EU in the conditions that underlie production. In particular, the more prosperous parts of the EU generally are better provided with a communications infrastructure in terms of transport and telecommunications, a more skilled labour force, and capacity in R&D. There is also a correlation between income per capita and distance from the EU’s core, suggesting that transport costs are an inhibiting factor in the peripheral regions, but the fact that Finland and Sweden face problems of distance from the core yet are relatively wealthy

258

EQUITY, COHESION, CONVERGENCE

suggests that the impediment of distance can be offset by other factors—as is indeed the case globally. To sum up, over the past decade, the EU has experienced some lessening of disparities in levels of income per head between the member states, but little convergence in the disparities between the regions of different member states. Within member states, several countries have experienced a widening of internal disparities. In terms of unemployment, some widening of disparities has been experienced between countries and regions. Historical experience in the EU and elsewhere suggests that, overall, cross-country disparities tend to narrow in periods of rapid growth, while in recession divergence gains ground. Since cyclical and adjustment factors are likely to have affected member states and regions differently during the short period discussed above, the trend over a longer period of time is of more interest. At the level of regions, long-term trends are particularly difficult to establish on account of data problems, but several analysts have suggested that, although for much of the post-war period there was a slow but steady reduction of differences in GDP across countries and regions, there are some signs of a reversal of this trend (Dunford, 1993), and even that we may now be in the presence of a Europe of different speeds (Fagerberg and Verspagen, 1996). Two recent studies of convergence at the regional level that are outlined in the following section do, however, suggest that, over longer periods of time, the data point to convergence rather than divergence in the experience of regional growth in the EU.

Convergence or divergence among the regions: some long-run evidence The literature distinguishes two concepts of convergence. The first, termed beta convergence, measures the rate at which GDP per capita converges, which reflects the extent to which the growth rates of poorer countries grow faster than those of the richer ones. The second, termed sigma convergence, involves a decline over time in the cross-country or cross-regional dispersion of GDP per capita. The first of these recent empirical studies of regional growth in Europe (Barro and Sala-i-Martin, 1991) presents evidence which suggests that convergence in regional GDP per capita values within the twelve-country EU (beta convergence) has occurred at a steady 259

EQUITY, COHESION, CONVERGENCE

rate of 2 per cent per annum over the period 1950–85, a rate which is remarkably similar to that found in the US states and elsewhere. A second, more recent, study by Armstrong (1996) looks at regional growth in the EU for a longer period, 1950–92, and for more regions but still for the EU twelve, on the basis of the same kind of cross-sectional regression models developed by Barro and Sala-i-Martin. Armstrong also concludes that the evidence overwhelmingly points to conditional convergence but at a slower rate of some 1 per cent. Both studies reach their conclusion on the basis of a negative relationship between a region’s growth rate and its initial GDP per capita, which means that the initially poorer regions grow at a faster rate than the initially richer regions. The dispersion of per capita incomes across countries and regions at the end of the period could nevertheless increase because the growth of the richer countries starts from a larger absolute base. Such an outcome has been termed ‘weak’ convergence by Chatterji (1992). Armstrong’s results suggest, however, that EU regional growth rates in each of the periods 1950–70, 1970–90 and 1975–92 have exhibited ‘strong’ or sigma convergence, that is, the differences between levels of per capita incomes have narrowed over the periods studied. The results of these studies are not inconsistent with a neoclassical convergence view of regional growth within the EU but, as Cheshire and Carbonaro (1995) emphasize, they are also consistent with a range of other possible explanations. These might include ‘catch-up’ through technological diffusion, and the possibility that structural change and the shift of resources from low productivity to high productivity sectors in poor regions may be going on at a much faster pace than in rich regions. In any case, it is not possible to conclude unequivocally that the underlying growth that has been experienced by the EU is convergent rather than divergent. The convergence that is observed among the EU regions may have been influenced in part by national regional policies and, after 1975, by the EC’s regional policy also, and to that extent it cannot be attributed to purely market mechanisms. Equally, the results cannot throw light on the issue of whether the deepening of EU integration that is under way through the implementation of the Single Market legislation and the moves towards monetary integration will be accompanied by convergent or divergent growth.

260

EQUITY, COHESION, CONVERGENCE

Regional policies since 1987 Current EU regional policies aim to reduce disparities between the regions and the backwardness of the less favoured regions, and so to promote cohesion, though the extent of the reduction and the pace at which progress towards it should be made have not been clarified. Intervention for this purpose takes place through the EU’s structural policies. These are implemented through the structural funds—of which the ERDF (accounting for 45 per cent of expenditures during 1989–99), the Social Fund (30 per cent) and the Guidance Section of the EAGGF are most important—and through the Cohesion Fund and loans from the EIB. The structural funds have certain EU-wide functions, but during 1989–99, 85 per cent of their funding was allocated to regional policy objectives. The funds currently operate within a common framework of reforms mainly laid down in 1987. The Cohesion Fund and the EIB also have an important role in promoting the reduction of regional disparities. Their approach is based on project financing and they are governed by their own specific rules. The following discussion relates mainly to the structural funds. The reform of the structural funds which was implemented in 1989 has involved three main dimensions: • • •

Concentrating funding on a limited number of objectives and on regions in greatest difficulty; An attempt to introduce more integrated planning into regional policy; A substantial increase in funding.

In terms of concentration, three of the six regional policy objectives are to account for 85 per cent of funding in 1989–99. Objective 1 is to develop regions which are lagging (i.e. those with GDP per capita less than 75 per cent of the EU average). About a quarter of the EU’s population live in such regions and more than twothirds of total funding is earmarked for such regions. Objective 2 is aimed at assisting areas affected by the decline of traditional industries. About 11 per cent of funding is allocated to this objective, which covers 16 per cent of the population. Objective 5b is to assist the development of rural areas undergoing structural decline. Five per cent of total allocations is earmarked for this objective, which covers 9 per cent of the EU’s population. Assistance is provided by matching grants for three broad 261

EQUITY, COHESION, CONVERGENCE

purposes: infrastructure, human resources and productive investment. About 30 per cent of the funds is spent on infrastructure investment in such areas as transport, telecommunications, energy, water supply and other environmental infrastructure. A further 30 per cent is devoted to strengthening education and training systems and supporting labour market policies. The remaining 40 per cent of funding goes mainly on what is termed productive investment. Much of this is aimed at helping to build a more dynamic and innovative milieu for business and supporting investment aid schemes for small and medium-size enterprises. The emphasis is increasingly on the last two purposes. The second dimension of reform involves an attempt to improve the operational efficiency of the structural funds by introducing more integrated planning into regional policy. This involves the replacement of ad hoc support for particular projects by support for monitored programmes budgeted on a multiannual basis and geared to specific targets and objectives. On the basis of regional plans submitted to the Commission, Community support frameworks (CSFs) are agreed after consultation between the member states and the Commission. These frameworks aim at a ‘partnership’ approach that incorporates EU, national and regional interests. There is now a more transparent system of policy making but doubts remain (CEC, 1993a; Begg et al., 1995) as to whether the acknowledged operational deficiencies of the funds have been overcome by the reforms. The third dimension of reform has involved a substantial increase in the provision of funds. This has been put into effect by doubling in real terms the allocations for structural fund activities for the first programming period following the reforms. There followed a further increase for the period 1994–9. As a result, for the period 1994–9, expenditure on structural policies is expected to amount to about one-third of the total EU budget. The Cohesion Fund is the other budget-financed instrument of EU structural policy. It was established by the Maastricht Treaty and came into operation in 1993. Although it is not a structural fund, it does have important regional policy objectives. Its purpose is to help the less prosperous member states to tighten their fiscal policy in preparation for EMU while maintaining expenditure directed towards reducing their backwardness. At present four member states benefit from the Cohesion Fund, namely Greece, 262

EQUITY, COHESION, CONVERGENCE

Spain, Ireland and Portugal. Its allocation for the period 1994–9 is ECU14.5 billion.

The impact and efficacy of EU regional policy The expenditure undertaken through structural policies can have two main effects on disparities in the EU: • •

It can increase output and employment in the assisted countries and regions simply by increasing expenditure. It can improve supply-side efficiency by improving the labour force, and by strengthening production structures through improved public and business-related infrastructure.

The direct relationship of fiscal flows to GDP in the EU can be readily ascertained. During the 1989–93 framework programme, structural intervention, mainly through the ERDF, would have increased GDP in Greece, Portugal and Ireland by 3 per cent, and by 1 per cent in Spain, assuming full additionality. During 1994–9 the corresponding increases in GDP are estimated at 4 per cent for Greece and Portugal, 3 per cent for Ireland and 2 per cent for Spain, including the Cohesion Fund for that period (CEC, 1996a). The main concern of EU structural policy, however, is to improve supply-side efficiency. An increase in the level of GDP of lagging countries or regions can be sustained without continuing subsidies from the EU only if structural policy increases the productivity of those areas by shifting their production functions upwards and by improving their capacity for autonomous selfsustaining growth. A once-and-for-all shift, for instance, by increasing capital formation, will have some short-term effect on growth, but the effect will be sustained only if the factors affecting productivity growth are favourably affected. It is much more difficult to assess the impact of EU regional policy from either of these points of view. Rates of return are difficult to calculate where the investment is directed towards public goods. Although the short-term growth effects have been modelled by the EU (CEC, 1996a) for 1989–93, and suggest a positive impact, the results are illustrative rather than demonstrative. In looking at the record, the impact of EU regional policy is also difficult to assess, since it is impossible to isolate the impact of policy measures from that of other effects, including those of the 263

EQUITY, COHESION, CONVERGENCE

member states’ own policies. But whatever the influence of policy— and it is only recently that it has been sharply focused—the enduring extent of the disparities seems clear. Even the member states’ own policies have had only limited effect in the many years since they have been in operation. Partly for that reason, in many of the more advanced countries there has been a marked decline in expenditure on regional policy incentives since the 1980s. In the UK the decline was as much as 40 per cent during 1985–92, and in France the corresponding figure was 32 per cent (Dignan, 1995).

Emergent regional policy issues in the EU Since the reforms of 1988 there has continued to be debate on whether EU policy towards regional disparities may need further modification. Three aspects in particular merit attention: • • •

Whether the 1988 reforms went far enough; Whether present policies can be maintained in the face of the prospective accession of Central and Eastern European states; Whether the implications of monetary union may not point to a need for the development of new instruments.

As already pointed out, the structural funds in their post-1988 shape mainly intervene by way of matching grants. The interventions continue to be subject to ex ante national quotas by objective with indicative allocations. All countries, including the most wealthy, receive a quota. The quota arrangement creates a presumption of automatic entitlement and makes it hard to secure project-related or programme-related additionality. Moreover, if the funds are largely exhausted by national quota allocations, incentives are also lacking for potential beneficiaries to compete on the basis of the merits of the programmes. As the Commission’s Reichenbach Group pointedly put it (CEC, 1993a, p. 68), ‘Evaluation of programmes is rather pointless if it is inconsequential.’ One change advocated by the group would be to subject the interventions to micro or macro economic conditionality, including, for instance, conditions relating to measures to promote adjustment in, for instance, the labour market, taxation reform, or fiscal policy (there is such a link in the framework of the Cohesion Fund). In other words, the funds should become performance-related rather than expenditure-related. Since the contribution that structural interventions can make is intimately 264

EQUITY, COHESION, CONVERGENCE

bound up with the quality of overall economic policy, it is a not unpersuasive point. A more fundamental question is whether the EU should continue to involve itself at all with regional policy when it involves transfers to the more advanced member states that are in effect indirectly financed by their own net budget contributions. One possible alternative would be to implement the principle of concentration fully in its geographical dimension and for the EU to limit its interventions entirely to the poorest states, leaving regional policy elsewhere to be conducted by national governments using their own resources (subject to common rules on regional aids, which would continue to be necessary to avoid market distortions) with perhaps some co-ordination where there are significant spillovers. Such a reform would find support in the principle of subsidiarity and should help to minimize ineffective bureaucratic intervention (Begg et al., 1995). Regional policy will be confronted with major new problems when the EU is enlarged to include some of the countries of Central and Eastern Europe. Under current rules, the countries of Eastern Europe would be eligible for substantial net transfers. If the EU were to maintain its practice of directing some three-quarters of its structural funds to regions where GDP per capita is less than 75 per cent of the EU average, the accession of the Czech Republic, Hungary, Poland and Slovakia would require a doubling of the funds if those countries were merely to benefit on the present scale despite their larger income gaps. If the Baltic countries and Bulgaria and Romania were to accede also, a tripling of the funds would be required. (Major transfers to those countries would also be required under the CAP.) A recent study (Baldwin et al., 1997) argues, however, that the net budgetary cost of both regional aid and CAP transfers would be roughly balanced by the gains from trade to the existing members. Though there are only small gains to the EU from expansion, the budgetary costs are similarly small. A more difficult potential problem is that the accession of the Eastern European countries is likely to place an uneven burden of adjustment on the existing member states in terms of its trade impact. This could have significant adverse implications for particular regions and for regional disparities that could not easily be resolved within the present framework. A third question is whether monetary union itself has any potential spatial effects that have implications for regional development and policy. The benefits of monetary union in terms 265

EQUITY, COHESION, CONVERGENCE

of the reduction in transaction costs and uncertainty, and interest rate reductions, have been discussed in the last chapter. In general, it is not clear that these effects could be expected to impinge in a differential way on lagging countries and regions. There is, however, some reason to expect that monetary union may favourably affect inward investment in lagging countries by removing major macroeconomic uncertainties. If so, this could reinforce the relative gains in investment inflows that the lagging countries apparently recorded during the period up to 1991, so contributing to convergence. The main direct cost of monetary union is represented by the loss of the exchange rate instrument. All countries are affected by the loss, but it is possible that the lagging countries may be particularly affected. In the first place, common shocks may impact more severely on those countries because of their narrower economic base. Also, if the adoption of a common currency increases pressure for the equalization of wage levels throughout the EU, it will operate to reduce the pace and extent of catching up. For both reasons monetary union may have some adverse effects on lagging regions and countries. The overall regional effects of monetary integration from these and other sources of benefit and cost are, however, essentially ambiguous, and empirical studies that might throw light on the situation are lacking. The outcome can be expected to depend a great deal on the domestic economic structures of each country and especially on the soundness of the economic policies that they pursue. The preparatory phase of monetary union may in any case have regional implications. Some steps have already been taken through the Cohesion Fund to provide additional financial support to assist the poorer states to keep up regional expenditures while they pursue the stringent budgetary policies that are needed to enable them to meet the macroeconomic criteria for admission to the monetary union. It is arguable, however, that, in the longer term, further measures may be needed to deal with acute regional problems that may later make themselves felt. Such measures might take the form of a further strengthening of intervention through the structural funds. In addition, the introduction of a contingency fund designed to provide general budgetary support to a member state whose economy is hit by a country-specific exogenous shock has also been advocated (CEC, 1993b). The case for such a fund has not yet been accepted by the EU. 266

EQUITY, COHESION, CONVERGENCE

The general case in favour of promoting a greater measure of regional convergence in order to promote the smooth functioning of the EU remains as strong as it ever was, and it appears to command widespread assent. The task of overcoming regional disparities has, however, so far proved to be quite intractable. Further enlargement will certainly increase the extent of disparities. It is by no means evident that EU initiatives so far taken have had a sustainable impact or that those whose effects have been positive have been cost effective. Transfers can evidently reduce disparities for so long as they continue, but effective policies that can be relied upon to accelerate the autonomous growth of the less successful regions remain elusive.

NOTE 1

GDP expressed in terms of PPS, or purchasing power standards, allows for exchange-rate differences and differences in the price level of one country relative to that of others.

267

13 ECONOMIC INTEGRATION IN A GLOBAL PERSPECTIVE: EXPERIENCE AND INITIATIVES

Although Europe provides the leading example of international economic integration, most of the numerous arrangements for regional integration that are found in the world today concern countries outside Europe. Of the seventeen instances of integration mentioned in Chapter 1, a far from inclusive list, fourteen solely concern developing countries. Of those, five are in Latin America and the Caribbean, eight are in Africa and two are in Asia. Two other of the arrangements mentioned concern the advanced countries of North America and Australasia. A new dimension in regional integration has been the emergence of arrangements and initiatives that involve both advanced and developing countries. These concern both the western hemisphere and Asia and the Pacific. Other cases which have some similarities with this development are represented by the Europe Agreements of the EU with countries of Eastern Europe and with Turkey. In its various forms, the phenomenon of regional integration in a broad sense is now globally pervasive. But although virtually all members of WTO are involved in regional arrangements and, in particular, in free trade arrangements, many of these are limited in scope and coverage and have the character of trade agreements rather than of arrangements for regional integration. Of those which qualify to be considered as examples of regional integration many take the form of a free trade area1 rather than of a customs union, common market or economic union. This chapter provides a brief review of the global scene. It considers in turn: (1) the rationale and experience of old-style economic integration among developing countries; (2) the global and policy context and characteristics of the so-called ‘new’ or ‘second wave’ regionalism 268

GLOBAL EXPERIENCE AND INITIATIVES

that dates from the late 1980s; and (3) the new dimension in international integration that is presented by the emergence of arrangements such as NAFTA that involve both advanced and developing countries. If all the initiatives now under way were fully implemented, the world could find itself at the beginning of the twenty-first century largely arranged into three separate discriminatory trading blocs that would dominate the world trading system. The EU and NAFTA are already economic and political realities, although of very different kinds, while the market-led open regionalization of trade and investment that has been a prominent feature of recent Asian development now shows some signs of being extended by discrimatory trading initiatives. These developments could have major implications for the multilateral system and for WTO. Some analysts believe that regionalism provides building blocks for more effective multilateral systems and promotes deeper economic co-operation than can be achieved at a global level. Others speculate that major blocs may be tempted to turn inwards or to seek to extract strategic trade concessions from competitors (de Melo and Panagariya, 1993) with potentially damaging effects on world welfare. The possibility of such dangers has almost certainly been greatly reduced by the successful completion of the Uruguay Round in December 1993. The further enlargement of the EU to include new members in Eastern Europe and the Mediterranean, and the likelihood that NAFTA will become embedded in a Free Trade Area of the Americas are developments that further reduce the dangers of increased protectionism and of trade and investment diversion. Existing regional arrangements have so far coexisted with the world multilateral system on a reasonably satisfactory basis and are generally thought to have been mutually supportive. There is no evidence in the dominant case of the EU that its operation has been detrimental to world trade (except in agriculture). In the longer term, however, reforms in the WTO provisions that regulate discriminatory regional arrangements may be desirable if the global extension of regionalism is to continue to be consistent with a satisfactory world trading order. These systemic issues—many highly speculative—fall outside the scope of this book.

269

GLOBAL EXPERIENCE AND INITIATIVES

THE ‘OLD’ REGIONALISM AMONG DEVELOPING COUNTRIES The orthodox theory of integration was evolved with the developed countries of Europe in mind and expressly for the purpose of throwing light on the problems of integration in Western Europe. This poses the question of whether the theory is also relevant to the circumstances of developing countries. To answer the question requires a consideration of two principal issues: (1) to what extent the characteristics of less developed countries favour trade creation in regional integration; and (2) whether trade creation is the significant criterion for evaluating customs unions and free trade areas among such countries. It is difficult to generalize usefully about the circumstances that favour trade creation, but the most widely accepted proposition suggests that it is more likely to arise when the existing external trade of prospective members is small relative to their domestic production and where a large share of external trade is already undertaken with prospective partners. This guideline is, on the face of it, persuasive.2 Where external trade is unimportant relative to domestic production, market integration offers more scope for creating intra-union trade by the displacement of high-cost domestic products. Similarly, where external trade with non-union members is relatively unimportant, the union can have little effect in diverting imports to higher-cost sources within the union. The conditions that favour trade creation are thus precisely the opposite of those typically found in many developing countries when regional integration first took off. At that time, the external trade of those countries was usually large relative to their domestic production, while intra-group trade was often a relatively small component of their total trade. Many such developing countries rely heavily on the exportation of primary products, which are freely traded on world markets. Integration is unlikely to affect significantly the volume of resources allocated to the production of such commodities. On the other hand, the imports of developing countries consist chiefly of intermediate products and final manufactures, which many such countries either did not produce at all or produced only to a limited extent behind high protective barriers. In terms of the criterion of gain of orthodox customs union theory, therefore, integration among developing countries may appear at best to be irrelevant and at worst to be positively harmful, except for the more developed of such countries. 270

GLOBAL EXPERIENCE AND INITIATIVES

To accept that conclusion, however, is to suppose that the rationale of ‘old’ market integration initiatives among developing countries lay in the gains to be derived from changes in their existing pattern of trade, which was necessarily based on their existing pattern of production. The traditional rationale for regional integration among developing countries was grounded instead on the putative gains from rationalizing the emergent structure of production on a regional basis in the context of an importsubstitution industrialization strategy supported by investment inflows. In evaluating the merits of integration in that context, what is relevant is not only its impact on the existing patterns of production and trade, but more particularly its impact on those patterns that would otherwise emerge in the absence of the formation of regional groupings and on the rate of economic growth. That traditional case for economic integration among developing countries thus relied on the validity of one or more of the several arguments justifying protection in the partner countries, and in particular on the case for import substitution. If a case for protecting industry can be made out, then, as explained in Chapter 3, integration may offer the prospect of reducing its costs, so freeing resources for further investment and offering the prospect of increased output and a faster rate of economic growth. In this perspective the conventional static criteria of trade diversion and trade creation may not dominate the policy argument, and a policy of regional integration among developing countries may be warranted even if the conditions for static trade creation do not exist. The merits of regional integration in that perspective would still need to be demonstrated for any proposed grouping of developing countries. No general case can be made out for integration among developing countries from the traditional perspective, any more than it can for integration among advanced countries in terms of the classical analysis of customs unions. The strength of the traditional economic case for regional integration in the context of import-substituting industrialization strategies would then depend on the empirical significance of a variety of factors, including: (1) the weight attached to the case for industrialization in economic development; (2) the possibilities, if any, of exporting manufactures to world markets rather than to protected regional markets; (3) the magnitude of scale economies in prospective regional industries; (4) the differences in the costs of producing industrial products in 271

GLOBAL EXPERIENCE AND INITIATIVES

the different member countries; (5) the location of markets in member countries; and (6) the costs of transporting raw materials and finished products within the region.

Approaches to market integration among developing countries During the earlier, import substitution phase of regional integration, four principal approaches to international economic integration have been employed in groupings of developing countries, sometimes in combination. The classical approach has involved across-the-board trade liberalization for all or most products, through the formation of customs unions or free trade areas. A second approach is the so-called ‘complementarity agreement’ approach which involves trade liberalization for certain existing industries or product groups in the context of a deliberate, planned rationalization of production. A third approach, also limited to specified products or industries, takes the form of measures to promote and regulate investment in new regionally based industries that enjoy economies of scale, so as to meet the combined demand of member countries more efficiently on the basis of agreed specialization. A fourth approach has involved the provision at the discretion of particular member states of margins of preference for specified regional products or industries rather than complete free trade. In each approach, the objective was to encourage profitable regional specialization but subject to distributional constraints except in the classical mode. In the case of classical, across-theboard trade liberalization, the procedural emphasis was on attaining the objective through the operation of market forces by the negotiation of a suitable tariff structure that was expected to work broadly in the desired directions. In the case of the other three approaches, the emphasis was on first determining the appropriate scope for specialization in existing or new industries, usually on the basis of feasibility and cost—benefit studies, and then utilizing a variety of policy instruments, such as external tariffs, fiscal incentives (including intra-bloc tariff preferences) and administrative controls to bring about the desired changes in the pattern of production. The classical mode of integration was the basis of some of the integration arrangements that were initially established among a 272

GLOBAL EXPERIENCE AND INITIATIVES

number of developing countries in the various regions of Africa, and in Central America and the Caribbean. Complementarity agreements, which were widely used in the Latin American Free Trade Association (LAFTA), represent a second, narrower, approach to regional integration. Their object is to promote specialization among existing plants or processes in order to exploit scale economies and to utilize excess capacities. In principle this approach has some attraction, particularly where structural or policy distortions may have rendered the price system an unreliable guide to regional specialization. Its practical disadvantage is that it is administratively and operationally complex, it demands much time-consuming inter-firm and inter-industry negotiation and it is exposed to protectionist pressures. LAFTA’s experience suggests that this approach is unlikely to be successful except where the sectors involved are fairly narrow and well defined, and where a reallocation of production can take place without adversely affecting the interests of particular enterprises. A third approach to integration among developing countries during the import-substituting phase rested on inter-governmental agreement on the establishment of designated new plants or industries, coupled with the adoption of tariff or other policy measures to ensure implementation. Agreed specialization was attempted in several groupings during the import substitution phase, including the East African Community, the Central American Common Market and the Andean Pact. The measures proposed for sectoral industrial co-operation among the countries of the Association of South East Asian Nations also rested on this basis. A fourth approach to integration involved the provision on the initiative of a member state of tariff preferences for regional industrial products or industries rather than complete free trade. This was the approach followed in both the West African Economic Community (CEAO) and in the Equatorial African Union (UDEAC). This arrangement leaves each member state with substantial discretion in respect of the degree of market integration it accepts with its partners in relation to industrial products. A striking feature of integration among developing countries in its import-substitution phase was the gradual but marked change of approach that occurred. Although earlier integration schemes relied largely or entirely on trade liberalization, subsequently, in almost every instance, attempts were made to plan or regulate specialization or to influence the location of integration-induced industrial activity. In groupings such as the Andean Pact the 273

GLOBAL EXPERIENCE AND INITIATIVES

regulation of industrial development constituted an integral part of the arrangements from the outset. The widespread resort to a regulated approach to market integration reflected in the first place the problems of equity that confront market integration that is based on import substitution fostered by extremely high rates of protection. These problems were subsequently compounded in both Africa and Latin America by the further market distortions and non-tariff barriers affecting regional and extra-regional trade that were introduced over the years by the policy responses to foreign exchange and debt crises, which were themselves partly bound up with the anti-export bias of those early arrangements. These often made the price system an unreliable guide to socially profitable trade and investment.

The failure of old regionalism The general experience Detailed studies of the operation of integration initiatives during the first wave of regional integration reveal a picture of failure. Only a small number of regional initiatives fully met their commitments to trade liberalization and fiscal and tariff harmonization. Of those that did succeed in implementing their programmes, most proved to be incapable of generating significant benefits in terms of intra-bloc trade expansion, regional specialization and faster growth. Their failure is attributable to a variety of common factors. The highly protectionist and inward-looking bias of most arrangements meant that gains from actual or potential trade creation were often small and sometimes outweighed by trade diversion. Import-substituting industrial development during this period often resulted in no net saving of foreign exchange. National policy responses first to the oil shocks and later to the debt crises gave rise to major distortions and severe interferences with markets, the effect of which was to render market-driven integration inappropriate and ineffective. Attempts to regulate regional industrial development in the interests of balanced development were also often ineffective and—as notably exemplified in the operations of the Andean Group—severely discouraged FDI. Where policy was successful in encouraging balance, it was often achieved only at a high fiscal cost and at the price of sacrificing the regional 274

GLOBAL EXPERIENCE AND INITIATIVES

specialization which was the putative source of gain and the rationale for integration in the first place. The pattern of industrial development that was brought about in the CEAO and UDEAC exemplifies that outcome (Robson, 1988). A formal statistical study of the experience of trade blocs based on cross-country comparisons and regressions (de Melo, Panagariya and Rodrik, 1993) lends further support to the conclusion that firstphase regional integration among developing countries did not have a positive effect on their income and growth. The one exception noted by de Melo, Panagariya and Rodrik is the case of the Southern African Customs Union (SACU), where favourable growth effects were found for Botswana, Lesotho and Swaziland. SACU has an unbroken and quite successful history which goes back eighty years, and it is virtually the only instance of a customs union amongst developing countries (as opposed to a free trade area) that is fully operational. Its leading aspects merit note as a case from which lessons can be learned.

The Southern African Customs Union SACU encompasses the relatively advanced Republic of South Africa together with the much smaller and less developed economies of Botswana, Lesotho, Namibia and Swaziland (BLNS). With the exception of Botswana, all the countries also form part of a common monetary area. The existing SACU agreement was signed in 1969 and amended in 1976 and is currently being renegotiated. The main features of the 1969 agreement were: (1) a revenue-sharing formula that was designed to compensate BLNS for the trade diversion and polarization of industrial development within the Republic of South Africa that accompanied the republic’s import substitution policies and for the loss of fiscal sovereignty which made it difficult for BLNS to use fiscal policy to support their social policies; and (2) provision to enable BLNS to develop industries and diversify their economies. The revenue formula of the 1968 agreement provided the smaller countries with a basic rate of duty approximately equivalent to the product of an iso-price tariff (that is, tariff revenue equivalent to what they could have raised by imposing their own tariffs while leaving prices unchanged), so that the cost of trade diversion was in effect offset by this element. This basic rate was enhanced by a factor of 42 per cent, which can be regarded as compensation for 275

GLOBAL EXPERIENCE AND INITIATIVES

the cost to those countries of industrial polarization in South Africa itself, and for their loss of fiscal discretion in relation to the main sources of their public revenues. Little or no use was made of the provisions intended to enable BLNS to promote their industrial development by protecting their infant industries and such limited industrial development as has taken place in BLNS since 1969 owes little to the customs agreement. Rather it has been stimulated by the relatively low wage rates in those countries; the effect of sanctions on South Africa; and by the access to EU markets provided by the Lomé. Convention. Experience suggests that few if any significant new industries set up in BLNS could rely on unimpeded access to what is nominally a single market in the absence of safeguards. Such safeguards will undoubtedly be sought by the smaller countries in any revised agreement. The price demanded by South Africa must be expected to be a less generous revenue formula. That could be justifiable in any case as a result of trade liberalization in the customs area resulting from commitments entered into in the Uruguay Round, which will reduce the costs of trade diversion for the smaller members, which import a large proportion of their requirements of manufactured products from South Africa. SACU is clearly a special case, (1) because it is dominated by one of its members with a strong economic and political interest in making it work; (2) because the permeability of frontiers places a premium on customs union for all members, including the smaller members, which might otherwise find difficulty in collecting independent tariff revenues; (3) because a centralized system of administering customs and sales tax or VAT revenues made it possible to eliminate fiscal frontiers and thus to operate the area as a single market. The benefits of a single market area have moreover not been dissipated in uneconomic industry-sharing arrangements. The enhancement factor which represents a significant transfer of income to BLNS has been a major source of benefit to the smaller partners. No doubt it was this that largely contributed to the positive growth effect that was identified in the study by de Melo, Panagariya and Rodrik. In marked contrast to the poor performance of initiatives for market integration, the arrangements for monetary integration among developing countries that existed during the first phase of market integration appear to have been rather successful. During the 1960s and 1970s, for instance, there is evidence from cross-section studies that African monetary unions contributed to economic integration 276

GLOBAL EXPERIENCE AND INITIATIVES

in the form of expanded trade and cross-border investment flows during that period and that they were accompanied by better macroeconomic discipline, low inflation and a generally better growth performance than that of other similar countries lacking monetary unions (Guillaumont et al., 1988). Relative performance more recently is, however, less easy to evaluate and remains a controversial issue (Devarajan and de Melo, 1991). Currency boards—a form of monetary integration—also operated successfully in the Eastern Caribbean Common Market. These exceptions do little to temper the picture of a deeply flawed pattern of regional integration among developing countries during the first or import substitution phase. By the 1980s, integration arrangements among developing countries were almost everywhere in dire straits. There was a lack of political commitment on the part of their participants, and thinly veiled hostility on the part of the international institutions. Interest in regional integration appeared to be withering away as structural adjustment programmes took effect and emphasis on export promotion policies became the order of the day, based in many cases, particularly in Latin America, on unilateral trade liberalization initiatives.

THE ‘NEW’ REGIONALISM Context and character During the past decade there has been a resurgence of interest in regional integration in developing countries throughout the world and in advanced countries. Old arrangements are in the process of revival and new ones have been initiated. International institutions have become less hostile to integration initiatives. How is this to be explained? A major factor that has transformed the integration situation in the western hemisphere and beyond is the conversion of the US. That country was traditionally hostile to regional integration initiatives (except in the case of the EC) but in recent years it has become a proponent of, and participant in, several initiatives for regionalism, some of which are certainly significant. Its change of heart has been attributed to a variety of factors, such as frustration with the lack of progress in the Uruguay Round of GATT and the threat that the deepening of European integration and the prospective enlargement of the EU to include

277

GLOBAL EXPERIENCE AND INITIATIVES

much of Eastern Europe and several Mediterranean countries was perceived to offer to its own markets. Other countries have embarked on regional initiatives to avoid being left out of the process. The new economics of market integration has also played a part in pointing to newly identified sources of potential gains from market integration that are much greater than those previously identified. New regional initiatives include the Common Market of the Southern Cone (MERCOSUR) set up in 1991 between Argentina, Brazil, Paraguay and Uruguay; the ASEAN Free Trade Area (AFTA), set up in 1992; NAFTA, which includes Canada, the US and Mexico, which was set up in 1993; UMEOA, the Economic and Monetary Union of West Africa that was set up in 1994. Attempts were also initiated during this period to restructure established schemes such as CACM, CARICOM and the Andean Pact. These varied initiatives are taking place in an entirely different national and global policy context from that of the earlier ones. They typically involve countries that have already committed themselves to lowering external trade barriers and which are trying to move from the promotion of industrial development on the basis of import substitution behind high tariff barriers to outward-looking policies based on export promotion. Those policies significantly alter the context and the rationale of the new phase of regionalism. First and foremost, liberalization reduces the scope of trade diversion costs and of trade creation benefits from regional integration. To that extent, the direct intra-bloc trade effects of integration will be lessened, perhaps greatly. This poses the question of whether significant benefits can be expected from integration in the context of the adoption of outward-looking policies. Part of the answer is that, in the new regionalism, static trade liberalization gains are not the dominant consideration, any more than they were in the old-style schemes. Any contribution that they make to export promotion must stem from different sources. The new economics of regionalism stresses the potential gains from reduced administrative and transaction costs and other barriers to trade. These show up for an economy in increased inter-firm competition that promises a reduction of production costs and monopoly rents. Production integration and crossborder investment creation and diversion should also be stimulated. These gains can be expected to result in increased competitiveness in a global context and thus are perfectly consistent with and can contribute to the objective of export 278

GLOBAL EXPERIENCE AND INITIATIVES

promotion. There can be benefit even if integration does not lead to substantial increases in intra-bloc intra-industry trade. To achieve these gains, however, much more than a simple free trade arrangement is called for if transactions and administrative costs are to be significantly reduced and market segmentation is to be overcome. It is not yet clear to what extent the new arrangements will encourage the policy reforms that could enable these initiatives to make a significant contribution to enhancing internal and external sources of competitiveness and growth, though there are some promising signs. There is a second important consideration that is prominent in a number of new initiatives. The new regionalism among developing countries is taking place in the context of increased globalization of production and business. This process has been stimulated by technological advances in transport and communications that have reduced natural barriers to trade. World-wide financial liberalization has also facilitated the pursuit of new strategies of investment and production organization on the part of TNCs, and this has been reflected in a spectacular growth of FDI flows during the past decade. This, and the associated flows of technology, training and trade, have been the primary means by which a growing number of developing countries have been integrated into the world economy. The pattern of FDI is, however, extremely polarized. In that context, regionalism is seen by some economists (Oman, 1994) as one means by which developing countries may be able to avoid exclusion from the growth benefits of the globalization of production. Many recent studies suggest that the prospects that particular developing countries or blocs have of attracting FDI are primarily dependent on their overall economic performance and growth. That said, however, the investment strategies of TNCs have been shown to turn largely on the relative importance of economies of scale and of distance costs. The latter include all those penalties that arise if the producer is far removed from the final market for his products. A variety of factors appear to be at work to enhance the relative role of distance costs, including synergy with suppliers and customers, and the need for flexibility and rapid adaptation. The outcome can be seen in the growing importance of market-based FDI and of regional, as opposed to global, production and sourcing. In that context, regional integration may offer benefits in terms of attracting FDI by way of its impact on the relative importance of scale economies and 279

GLOBAL EXPERIENCE AND INITIATIVES

distance costs (Robson, 1993). There is certainly evidence, discussed elsewhere in this book, that FDI inflows can be induced by regional arrangements, particularly in cases where they promise access to significant markets, and the markets of industrialized countries. Smaller countries that lack such links may therefore find themselves increasingly marginalized in the quest for investment inflows. During the first round of integration among developing countries, investment inflows were also sought, but regional blocs rarely succeeded in attracting them on a significant scale. This was partly because of self-imposed obstacles against foreign investment, such as the notorious Decision 24 of the Andean Group. But it was partly because the inflows originated largely in investment diversion, so that their products wer e uncompetitive in export markets. In addition, the blocs lacked long-term credibility, which increased the risk premium on investment. The lack of credibility also accentuated problems of regional balance, since it gave investors an additional incentive to locate in or near the larger markets of the bloc so as to minimize the risk of losses in the event of break-up. Enhanced credibility may thus be doubly critical for the future success of the new wave of integration initiatives. One approach to creating it is to build-in external constraints and incentives. Credibility was given to the African francophone monetary unions by guarantees from the French Treasury. A hegemonic power, as in the case of the Republic of South Africa, provided assurance and predictability in the case of SACU. Current Latin American initiatives for free trade areas may similarly be able to enhance their credibility by locking their policy reforms in by treaties with a dominant partner if North—South links such as that provided by NAFTA for Mexico are extended by the admission of blocs to a hemispheric system of free trade that provides predictable and assured access to the market of the US. The following section briefly considers some of the principal new initiatives in Africa, Latin America and Asia, including those that embrace both advanced and developing countries. Scope for traditional gains from trade creation is certainly not lacking in relation to the current initiatives. But in terms of assessing their contribution to outward-looking policies and to avoiding marginalization in a global economy—which turns on their ability to attract foreign investment on a competitive basis—what is mainly important is their potential impact on the interrelated aspects of 280

GLOBAL EXPERIENCE AND INITIATIVES

competitiveness and credibility. These are the aspects to bear primarily in mind in the following brief discussion.

The West African Economic and Monetary Union Since independence most of the francophone West African states have been linked in a monetary union (Union Monétaire OuestAfricaine, UMOA) as well as in a separate economic community (CEAO) with a largely similar membership. It is generally agreed that the monetary union operated remarkably successfully, although the machinery for macroeconomic co-ordination and for the supervision of financial instruments left much to be desired. Also, over the years the currency (the CFA franc), which was tied to the French franc and convertible, became progressively overvalued. It was eventually devalued by 50 per cent early in 1994. The CEAO itself was characterized by a higher degree of regional integration in trade than any other bloc in Africa apart from SACU, but it lacked a common external tariff, its trade liberalization programme was ad hoc, and effectively each member country could determine the degree of protection it accorded to the industries of its partners. Trade liberalization was coupled with a system of fiscal compensation that was linked to the degree of preference that countries offered their partners and received from them, and to net intra-CEAO trade balances. Ivory Coast and Senegal, the principal net exporters of manufactured products in intra-CEAO trade, were the contributors to the compensation funds. In no sense, however, did the region constitute a single market, and it is clear that the transnational enterprises which largely accounted for the region’s industrial production did not treat the area as such. There was a failure to exploit the economies of scale which showed itself in the replication of plants. In other fields of shared economic interest, such as transport and agriculture, co-operation was limited. But the Community did contribute to bringing about and maintaining what was by African standards a relatively high level of intra-community trade. By the end of the last decade, however, financial and other pressures had combined to produce the abandonment of the compensation arrangements and with them the trade preferences, and the virtual demise of the CEAO soon followed. In 1990 the seven heads of state of UMOA decided to transform the monetary union into an economic and monetary union (Union 281

GLOBAL EXPERIENCE AND INITIATIVES

Economique et Monétaire Ouest-African, UEMOA). This was a response in part to the deficiencies of the CEAO and in part to certain operational weaknesses of the monetary union. To deal with the weaknesses of the monetary union, the initiative proposed measures for the multilateral surveillance of national macroeconomic policies but, apart from that, the existing monetary union and its central bank remained intact and subject to the existing treaty, which had been in force since 1974. Most of the treaty of January 1994 is therefore concerned with the requirements for economic union. Here the treaty provides for the institution of a customs union and the establishment of a single market, together with a variety of measures to give effect to these objectives. The approach to the objectives builds on the practices of previous communities in Africa, but fiscal compensation is to be purely transitional. In the longer term the treaty provides for the establishment of structural funds, the object of which would not be to compensate for experienced disparities in costs and benefits but to encourage the balanced development of UEMOA in order to ensure that it is sustainable. The UEMOA has given itself three major tasks in relation to the implementation of the customs union and the institution of a sustainable single market. The first is to put in place a common external tariff and fiscal harmonization measures that would enable a single market to operate without fiscal frontiers. The second is to put in place a transitional system for compensating for certain revenue losses associated with the introduction of a common external tariff. The third is to elaborate longer-term instruments and measures that could moderate the structural forces that have already produced a polarization of development upon Ivory Coast and Senegal and which may be strengthened by an effective single market, and to do so without losing the benefits of production specialization. A basic task will be to elaborate fiscal harmonization measures, including a common external tariff and internal indirect tax structures that will reduce to a minimum the need for border tax adjustments and inspections. The establishment of a common external tariff will involve a shift from the CEAO system of a free trade area for agricultural products combined with a preferential trading arrangement for approved manufactures. A simplified tariff of the kind under discussion would permit a modest reduction in overall tariff rates while maintaining or increasing revenues for the principal countries, but would involve important changes in the 282

GLOBAL EXPERIENCE AND INITIATIVES

competitiveness of industries in certain countries, and for low-tariff countries such as Benin and Togo there could be serious revenue problems. A major benefit of fiscal harmonization would be a decline in rent-seeking trade and arbitrage, which is a principal motive for frontier controls inside the region. Experience in Africa strongly suggests that, where border tax adjustments are maintained, the operation of regional markets is likely to be seriously impeded. But, even with fiscal harmonization, the prospects of removing frontier controls in UEMOA without at the same time instituting a unified fiscal administration for customs and VAT of the kind found in SACU may not be promising. A second task for UEMOA is to elaborate a system of compensation for revenue losses arising from the institution of the customs union. The introduction of common tariff rates accompanied by some modest rate reductions would not involve revenue losses except for Benin and Togo but, if it were also accompanied by free trade, Benin, Togo, Burkina Faso, Niger and Mali would suffer considerable revenue losses in the absence of compensation if they continued to be net importers from Senegal and Ivory Coast. One basis for such compensation could be the static real income losses that would be associated with the institution of tariff-free trade. These losses would equate to the additional revenue that each partner could raise by an iso-price tariff. This would not be so very different from the basis of compensation adopted in the former CEAO, which ultimately proved unacceptable to the net contributory countries. However, unlike the CEAO arrangement, the UEMOA arrangement is to be transitional. Furthermore, a fully effective single market could be expected to yield more substantial benefits to the more advanced countries, which would be the prospective net contributors, than the previous ad hoc arrangements. To that extent, such arrangements may be more acceptable. After a transitional period, compensation for revenue losses is to be replaced by the institution of structural funds. These, in conjunction with the eventual system to be adopted for financing the UEMOA’s budget (a share of national VAT proceeds), could contribute to convergence and more balanced development as well as to ameliorating any real income losses that might result from disequalizing effects from implementing a customs union and common market. The importance of credibility for securing the favourable effects of integration, particularly in relation to the restructuring of investment, is a commonly stressed factor. A recent 283

GLOBAL EXPERIENCE AND INITIATIVES

survey of the intentions of TNCs in UEMOA concerning the alternatives of concentrating production in one country or continuing to operate several plants, each mainly serving the individual national markets, makes it clear that credibility would be the paramount factor in that choice. Up to the present time the UEMOA does not appear to have determined the policy objectives that are to be pursued through its funds. The operations are apparently to be led by distributional considerations, and so the size, use and distribution of the funds must be expected to be defined by political negotiation that will reflect the degree of political and social solidarity within UEMOA. At the same time, if convergence is to be encouraged, a crucial role must be assigned to efficiency considerations. There appears to be scope for doing so through the provision of improved cross-border transport links and investment in improving human capital in the less advanced partner states. UEMOA is at present in a formative stage. In favour of an optimistic prognosis is the commitment to tariff liberalization and the evident political commitment to the maintenance of established economic links, which have an unbroken history since independence, coupled with the new awareness of the many benefits of reducing transaction costs through the institution of a single market. It also builds on a solid monetary union whose operation excludes problems of macroeconomic divergence that have impeded the operation of other blocs in Africa and Latin America. At the same time, a number of the sources of conflicts of interest that have constituted obstacles to earlier, less ambitious initiatives among the countries concerned are unlikely to disappear. There is no painless way of financing either compensation or regional policies that depend on expenditure. Without major steps towards tariff liberalization which could reduce the static welfare costs of integration for the less developed partners, or the adoption of other measures that might generate non-traditional or dynamic gains, the single market aspects of this project, which is more ambitious and demanding in terms of harmonization, administrative requirements and political commitment than its predecessor, may prove difficult to implement. The direct intra-bloc gains from reductions in transaction costs could be relatively large, but UEMOA, like most other African blocs, is not well placed through outward-looking regionalism to stimulate FDI inflows because of its underdeveloped infrastructure and lack of a highly skilled labour force. 284

GLOBAL EXPERIENCE AND INITIATIVES

The Common Market of the Southern Cone A prominent example of the new regionalism in Latin America is represented by MERCOSUR, which emerged from an attempted process of sectoral integration between Argentina and Brazil in 1985. Its four member countries comprise Argentina, Brazil, Paraguay and Uruguay. Together they have (1996) a population of more than 200 million and an aggregate GDP of some US $650 billion in terms of 1990 prices and they account for about 45 per cent of the population of Latin America and more than 50 per cent of its GDP. It is a heterogeneous group. Brazil accounts for two-thirds of regional output and four-fifths of the population and possesses an extensive industrial base, while Argentina accounts for most of the rest of output. Per capita GDP differs markedly among the four partners, ranging from US $5,480 in Argentina and US $3,704 in Uruguay to US $2,716 in Brazil and US $1,471 in Paraguay. MERCOSUR’s objective under the 1991 Asunción Treaty was to constitute a customs union by 1995. The two main instruments were a four-year trade liberalization programme and a commitment to implement a common external tariff by January 1995. The treaty also established the objective of co-ordinating sectoral and macroeconomic policies, but no specific procedures or time-tables were set out. The liberalization programme had two main components: (1) a linear and automatic import tariff reduction programme, and (2) a commitment to reduce all non-tariff barriers to intra-regional trade. Implementation of the liberalization programme was to lead to the establishment of a free trade area by 1995, while the common external tariff was to be negotiated during the transition period and enforced from January 1995. Free trade in services was also formally provided for, but no specific calendar was laid down for bringing it about. Issues relating to the free movement of factors of production have yet to be adequately addressed, although the Colonia Protocol of 1994 in principle established national or most favoured nation treatment for investment from member states. MERCOSUR’s ambitious target date for the inception of the customs union was formally met, although it is subject to important limitations. First, the trade liberalization programme had authorized countries to except a limited number of sensitive products from the schedule. Products included in those lists will not enjoy full tarifffree access for a further five to seven years. Second, the common external tariff came into effect for some 88 per cent of the tariff 285

GLOBAL EXPERIENCE AND INITIATIVES

schedule only. The remainder, which accounts for some 25 per cent of intra-bloc trade, consists mainly of capital goods, computer and telecommunications products and automobiles. The national rates on these products are to converge to the negotiated common external tariff rates by the year 2000, except in the case of Paraguay. Meanwhile, such categories of intra-regional trade continue to be subject to general and specific rules of origin. Progress in removing other border and non-border non-tariff barriers has been slow and not uniform. Significant progress on many of the latter will require the convergence of standards and practices which, if EU experience is any guide, will be a lengthy process. Although MERCOSUR cannot yet qualify as a full customs union, let alone a common market, its formation has been accompanied by a considerable increase in economic interdependence among its members. Intra-regional exports rose during the transition period from US $4 billion in 1990 to nearly US $12 billion in 1994. Although this rapid growth cannot be attributed solely to MERCOSUR, the much faster rate of growth of intra-regional imports vis-à-vis total imports that occurred during this period suggests that the preferential system has played an important role. By 1994 Brazil had become the leading destination for Argentine exports (absorbing about a quarter) and Argentina had become the second most important market for Brazil, accounting for about one-tenth of Brazil’s exports. Much of the expansion in trade between Argentina and Brazil has taken the form of intra-industry trade in manufactures. Investment flows within the region have displayed some signs of response to the liberalization initiative, but they remain marginal. Other activities associated with regional integration at the level of production have also become more prominent. This includes ‘reorganization investment’ carried out by existing foreign investors, which has taken place in automobiles, food and beverages, petrochemicals and textiles (Bouzas, 1997). MERCOSUR’s performance to the end of the transition period is generally admitted to have been outstanding. Several factors have contributed: the policy context was one of significant unilateral trade liberalization; the method of proceeding by automatic tariff preferences with limited protective exceptions reduced the need for negotiation; macroeconomic convergence helped to minimize unsustainable trade flows. In the future it may prove more difficult to maintain the momentum. The costs of adjustment for sensitive sectors will become more evident as the customs union is more fully implemented. In addition, to secure the cost-reducing benefits 286

GLOBAL EXPERIENCE AND INITIATIVES

of market integration that are stressed by the new economics of integration, and which are crucial to supporting export promotion, a considerable degree of policy deepening would have to be brought about to overcome non-tariff barriers and market segmentation in the region and to ward off the threat of disruption and the lack of predictability that would accompany the emergence of macroeconomic divergence. Whether MERCOSUR will constitute a form of regional integration that can generate the gains of new regionalism, or will merely turn out to be a limited customs union whose main benefits are found in more traditional gains, remains to be seen.

The North American Free Trade Agreement: a new dimension? The background Regionalism in North America began with the 1988 Canada—US Free Trade Agreement (CUFTA), which took effect in 1989, although a sectoral pact between the two countries for automobiles and parts had existed since 1965. Regionalism has been further extended with the tripartite North American Free Trade Agreement of 1993, which came into effect in 1994 between the US, Canada and Mexico following the Bush initiative of 1990, which is discussed in the following section. NAFTA is clearly a special case of regional integration. Its membership includes one dominant economic partner, the US, and two countries with much smaller economies. A further distinguishing feature is that two of its members are advanced, whilst Mexico is a developing country with much lower per capita incomes. The new arrangements have been driven by several factors. One factor has been the frustration felt by the US with the slow progress of the Uruguay Round of multilateral tariff negotiations, coupled with concern for the possible impact of the EU’s single market measures on its trade. This produced a volte-face on the part of the US, which hitherto had been staunchly opposed to regional initiatives (except for Europe). An equally important factor has been the desire on the part of the smaller countries for secure and predictable access to the markets of their larger trading partners. In the case of NAFTA both Canada and Mexico rely mainly on the US market for their exports (70 to 85 per cent in each case). In the 287

GLOBAL EXPERIENCE AND INITIATIVES

case of Canada, anticipated cost-reducing effects have been a factor. In the case of Mexico, the desire to attract inward FDI and to use the agreement as a way of establishing the credibility of domestic policy reforms was a major consideration. The NAFTA agreement sets a four-category timetable for implementing tariff reductions towards an eventual zero tariff after fifteen years. The basis initially favours Mexico in terms of pace of access. By the year 1999, 60 per cent of intra-NAFTA trade should be free of tariffs, and by 2005 some 95 per cent should have a tariff level of zero. However, many sensitive sectors receive special treatment, notably the automotive sector, agriculture, energy, textiles and apparel and petrochemicals. Although NAFTA involves a less deep form of integration than the EU in several respects, including the lack of a CET and provision for resource transfer from stronger to weaker members, the agreement may turn out to be more like a structured free trade area capable of yielding broad integration gains rather than merely a simple free trade agreement. It contains provisions relating to the reduction of technical barriers to trade by the harmonization of industrial and sanitary standards, the facilitation of customs clearances and the opening up of government procurement. Working groups are established to deal with many of the issues. Other issues concerning the environment and labour standards have also been coupled with the negotiations. There is a comprehensive framework to facilitate intra-NAFTA investment, which provides for national and most favoured nation treatment amongst the members, and freedom for financial transfers. The agreement also has chapters on liberalizing trade in services (including a special chapter on financial services, where some liberalization has already taken place) and provisions on competition policy. Since the agreement constitutes a free trade area and not a customs union, an important part of the agreement is concerned with rules of origin, ostensibly to avoid trade deflection (see Chapter 2). Eligibility normally involves a change of tariff classification and a value-added rule of 50 per cent or 60 per cent for regional components. Protection is inherent in these criteria. The rules for certain sensitive goods such as automotive products and textiles and clothing are more complex and even more restrictive and seem certain to hinder trade. Apart from restrictive rules of origin, another limitation of the agreement (and of CUFTA) is that it did not eliminate the use of national anti-dumping or countervailing duty actions in the free trade area. Much use has 288

GLOBAL EXPERIENCE AND INITIATIVES

been made of such actions by domestic interests to restrict market access for competing imports from the rest of the area. Although dispute resolution mechanisms exist, it is claimed that they are not yet capable of ensuring the predictable commercial environment that is needed for the maximization of trade gains (Dearden, 1997). A potentially important aspect of the agreement for the future of western hemisphere trade is its accession clause. This requires the consent of all member countries to enlargement, but a country can agree to the accession of another country without allowing the agreement to apply between itself and the acceding country, so opening up the possibility of hub-and-spoke arrangements. There are, however, no explicit accession requirements or rules. Chile’s application for accession is under consideration. But, beyond that, recent developments suggest that NAFTA is unlikely directly to play a greatly expanded role in the structure and process of western hemisphere trade integration in the near future.

Assessing the impact of NAFTA There are a number of reasons for expecting NAFTA to generate economic benefits for all three countries although, for the US, noneconomic benefits, including the containment of migration and the extension of political influence, have clearly been perceived as major merits. First, its members are large and are already natural trading partners. Canada and Mexico are respectively the first and third largest trading partners of the US, and the US is the largest trading partner of both Canada and Mexico, accounting for more than twothirds of their trade. Second, the tariffs of the US, Canada and Mexico are on average relatively low, amounting in unweighted terms to some 6.0 per cent, 8.5 per cent and 11.5 per cent respectively in 1994. These factors should help to ensure that trade diversion is limited. There are, however, a number of other factors that suggest that, even if the agreement is fully implemented, conventional economic welfare benefits from NAFTA will be limited by comparison, for instance, with those that were generated by the establishment of the EEC. The first is that, since the tariff barriers are already low (and Mexico already benefited from the Generalized System of Preferences, for developing countries, in US markets), the possibility of a large beneficial 289

GLOBAL EXPERIENCE AND INITIATIVES

expansion of trade due to tariff elimination alone is correspondingly limited. Tariffs and other trade restrictions were much higher in Europe in 1958, giving more scope there for beneficial trade expansion. The second factor limiting the scope for beneficial trade expansion is the lower level of regional integration in NAFTA by comparison with the EU as measured by the share of intra-bloc trade in total trade. In 1994, intra-EU trade was 67 per cent of total trade, as compared with 32 per cent for NAFTA. Admittedly the EC has been in existence for forty years, but regionalization measured by the share of intraEC trade in 1958 was already higher than it is in NAFTA now, even though the share of world product traded then was much lower than it is today. A third factor that may limit the benefit from trade expansion in NAFTA by comparison with the EU is that the US, Canada and Mexico are more complementary than competitive, unlike the members of the EC which in 1958 were more homogeneous in terms of factor endowments. Consequently, trade expansion among NAFTA partners can be expected to a considerable extent to take the form of interindustry trade based on differences in factor endowments rather than of intra-industry trade. That suggests that the offsetting costs of industrial adjustment and of income redistribution may be more acute within NAFTA, particularly for Mexico, than they were in the EC. The conventional welfare benefits are, however, not the only ones to consider. The role of liberalization in generating economies of scale and increased competition in imperfectly competitive product and resource markets that should produce lower prices for consumers was a motive for Canada in the case of CUFTA. In the case of NAFTA, predictable tariff-free access to Mexico will also encourage the US to import labourintensive components from Mexico and help to retain other stages of production in the US rather than losing production entirely to low-wage economies outside NAFTA. This should help to increase the ability of the US economy to compete in the world market. Mexico may gain in employment that requires low skills, but it can be expected to be partly at the expense of the newly industrialized economies or of other Latin American countries rather than at the expense of the US. The US is expected to lose unskilled jobs in the future, but many of them might have been lost anyway as a result of globalization. The establishment of the free trade area can also be expected to 290

GLOBAL EXPERIENCE AND INITIATIVES

benefit Mexico, first reducing the risk of investment in Mexico by helping to lock-in structural reforms, so contributing to an increase in the inflow of FDI, and second by leading to greater export-led growth resulting from increased access to the large American market. The incremental benefits to Canada over and above those ensuing from CUFTA are less obvious, and it appears that Canada entered the tripartite negotiations with some reluctance. The effects of North American trade liberalization have been studied extensively with computable general equilibrium models and other techniques (Brown et al., 1992; Klein and Salvatore, 1995). Table 13.1 refers to one such estimate of the impact of NAFTA on Mexico and the US for selected macroeconomic variables. The table shows that the annual growth of Mexican real GDP over 1994– 2003 is estimated to be 1.4 per cent higher with NAFTA, resulting in a real GDP level 14 per cent higher than without NAFTA by the year 2003. The annual rate of increase of the consumer price index would also be substantially lower with NAFTA. NAFTA would also reduce the risk of investing in Mexico, as reflected in the lower rate of interest on Treasury bills. The higher growth rate of GDP

Table 13.1 NAFTA’s macroeconomic impact on Mexico and the US, 1994–2003

Source: Adapted from Klein and Salvatore (1995), tables 2 and 3.

291

GLOBAL EXPERIENCE AND INITIATIVES

and the lower inflation rate with NAFTA are due primarily to the larger FDI inflow that NAFTA would encourage. For the US, the table shows that GDP in 1987 prices is expected to be $42.8 billion, or 0.64 per cent higher with NAFTA by 2003. This results from higher US exports to Mexico induced by the more rapid growth of income in Mexico under NAFTA. The increase in real GDP is significant in absolute terms but small as a percentage of US GDP because the US economy is so much larger (about twenty times larger) than the Mexican economy. The table also shows that US industrial production is expected to be 2 per cent higher with NAFTA than without it, the unemployment rate 0.3 percentage point lower, and a net total of half a million additional jobs is expected to be created. These aggregate results are broadly in line with those estimated in other empirical studies reviewed by Brown et al. (1992) and by the US International Trade Commission (USITC, 1993) and others (Hufbauer and Schott, 1993). They say nothing about the distributional impacts of the arrangement. Those effects have been a major source of apprehension, particularly in the US. In its case, unskilled labour may be expected to suffer, and certain regions of the US are also likely to be particularly affected. The USITC (1993) study throws additional light on these issues.

Western hemisphere regionalism: towards a Free Trade Area of the Americas The initiative for a broader regionalism in the western hemisphere, embracing the countries of North and South America, stems originally from the 1990 Enterprise for the Americas initiative of President Bush, which held out the prospect of hemispherical free trade. Free trade agreements were to be considered with Latin American countries that could meet certain conditions: notably they should have stable market-oriented economies and have made demonstrated progress towards open trading regimes. A preference was expressed for negotiating with blocs rather than with individual countries. It was Mexico’s response to this initiative that led to NAFTA. The direct trade benefits for most other Latin American countries of free trade with the US would be small (Erzan and Yeats, 1992). However, the benefits for certain countries in terms of changes in investment and production and in binding policy reforms in place 292

GLOBAL EXPERIENCE AND INITIATIVES

could be not dissimilar to those expected for Mexico in NAFTA. NAFTA membership has consequently been seen as a goal by some countries such as Chile, with which negotiations have already begun. For most other countries, particularly the smaller ones, this option has never been a real prospect, and others, like Brazil, do not desire it. Meanwhile, MERCOSUR has been set up and steps have been taken to reactivate the CACM and the Andean Pact. In further initiatives, several members of the three major blocs have simultaneously entered into commitments involving yet more preferential links. Some amount to pieces of variable geometry designed to facilitate closer integration among the members of a bloc, as has occurred in the CACM and the Andean Group, but others involve preferential links with outsiders. Chile has reached an agreement on associate membership of MERCOSUR (as has Bolivia) while still seeking membership of NAFTA. The compatibility of some of these arrangements with the engagements of the primary blocs remain unclear, as do their trade implications. But their number and the complexity of the different rules of origin that now coexist seem unlikely to make it easy to reduce border obstacles to free trade. What has finally emerged from the initiative for western hemisphere integration is a decision to start formal negotiations on a Free Trade Area of the Americas (FTAA) in March 1998. One issue that is already settled is that the free trade area should coexist with sub-regional groups such as NAFTA and MERCOSUR. That decision appears to limit the scope for a full multilateralization of free trade in the Americas.

New initiatives for regionalism in Asia This brief survey has left Asia until last because, for almost three decades, regionalism has scarcely been a significant issue in that region. While countries in all other regions embarked on experiments in free trade areas, customs unions and common markets, those in Asia remained largely aloof. An increased degree of regionalization of trade and investment has occurred under the stimulus of market-led FDI undertaken by Japan and the newly industrializing countries of the region as certain sectors lost their competitiveness at home, but it has been brought about largely through the internal operations of TNCs and without the support 293

GLOBAL EXPERIENCE AND INITIATIVES

of institutionalized regionalism resting on discriminatory trade arrangements. The preoccupations of the ASEAN regional bloc (established in 1967 by Indonesia, Malaysia, the Philippines, Singapore and Thailand) were initially political and strategic. It was some time before economic co-operation became a concern, and it was not until 1977, ten years after its foundation, that the members first entered into an agreement on a Preferential Trading Arrangement. Little progress was made in implementing the agreement, and as late as the end of the last decade a mere 5 per cent of intra-ASEAN trade was covered by it. Not surprisingly, all empirical studies of its operation (Langhammer, 1991; Imada, 1993) conclude that it had no significant impact on intra-ASEAN trade. An alternative approach to regional integration was simultaneously envisaged in the ASEAN Industrial Projects Programme that was agreed in 1977. This programme is an example of the third approach to integration that was distinguished above. It involved the agreed assignment to each member state of certain industrial projects offering large economies of scale that would be capable of meeting the regional requirements of a number of products, including, for instance, fertilizers. This alternative approach also proved to be largely ineffective as countries showed themselves unwilling to accept the costs of trade diversion that were involved. Since the end of the 1980s, both ASEAN and a number of East Asian and other regional countries have become involved in new initiatives for regionalism. These have been stimulated by the same external factors that have played a role elsewhere, notably concern about the impact of regional initiatives in other regions. Two approaches may be noted. In the first place, the ASEAN member states3 agreed in 1992 to establish gradually an ASEAN free trade area (AFTA). Its objective is to bring about a reduction in intra-ASEAN tariffs on manufactures to 0–5 per cent within fifteen years from January 1993. It was also proposed to eliminate nontariff barriers, but the procedure for doing so was not specified. Studies of the impact of AFTA (Imada, 1993) suggest that the outcome of its full implementation would be greater specialization according to comparative advantage and increased intra-ASEAN trade that would reflect both trade creation and trade diversion. The expected changes in trade flows are not, however, large. The credibility of the process has been in question almost from the outset as failures to implement the tariff reduction programmes and the extensive use of exclusions—echoing previous experience— have manifested themselves. 294

GLOBAL EXPERIENCE AND INITIATIVES

The second noteworthy initiative for regionalism in Asia has centred on the formation of much wider Asian groupings. In 1989 Australia, which had also become concerned at the implications of regionalization for those countries that were excluded, put forward a proposal that was eventually implemented for a very wide regional grouping, known as Asia-Pacific Economic Cooperation (APEC), that would include not only the East Asian countries, and Australasia, but also the US and Canada. Its initial minimalist programmatic outcome is heralded by some as a new phenomenon that has been termed ‘open regionalism’ (Garnaut and Drysdale, 1994), characterized by openness in terms of membership, voluntary understandings and the absence of discrimination. At its 1994 summit, however, this forum appears to have abandoned non-discrimination by adopting an agreement to institute a free trade area by the year 2020, with the industrialized countries agreeing to reach that goal by 2010. The agreement is not binding on its members, however, and it may turn out to have little impact.

CONCLUSION The first phase of regional integration initiatives outside Europe disappointed their proponents. Most initiatives failed to generate significant benefits in terms of welfare-increasing intra-bloc trade expansion or a rationalization of the emergent industrial structure or in stimulating growth. In part this outcome stemmed from failure to implement programmes that in principle embodied many constructive elements. The new regional integration schemes that have been initiated outside Europe during the past decade are rightly viewed as differing in important respects from those of old-style regionalism. The differences should not be exaggerated. So far, they reside chiefly in the general policy context in which they have been established, and in their aspirations, rather than in any substantial difference of embodied approach. The most obvious difference is that they have reverted to the classical trade liberalization approach discussed above and have abandoned attempts at promoting agreed specialization. In addition the agreements usually contain provision designed to foster intra-bloc investment. Trade creation through the cost reduction effect from economies of scale is likely to remain

295

GLOBAL EXPERIENCE AND INITIATIVES

a source of benefit, but against the background of deep unilateral tariff reduction its potential will be diminished in the absence of other measures. Other objectives have coloured the new initiatives, but they have yet to be matched in most cases by specific policies, instruments and programmes to attain them. The ‘new regionalism’ initiatives have emerged against the background of the pursuit of outward-looking policies to which those initiatives are expected to contribute. If they are to do so, integration must result in a favourable effect on the global competitiveness of participants in terms of foreign trade and FDI. The two main channels through which such an effect can operate are: (1) through the impact of integration on the credibility of the initiatives and the predictability of market access that they promise; (2) through policy-deepening measures that can effect a reduction in the cost of real trade barriers and the costs of market segmentation in production. Where the credibility of policy reforms and macro-stability are enhanced by integration, a reduction in risk premiums can be anticipated which can contribute to this objective. Credibility and predictability should be fostered to the extent that integration provides a framework for discipline. That is significant for Mexico in relation to NAFTA, where it was expected that export competitiveness would be significantly improved primarily through these channels. What have been termed ‘participatory supranational agencies of constraint’ have provided the discipline and credibility in monetary integration in West Africa, and it was hoped that this effect would spill over into the new Economic and Monetary Union. Credibility has been an important consideration also in relation to the Europe Agreements between the EU and the much less developed countries of Eastern Europe. These agreements have been justified in part as a means of first promoting and then lockingin economic (and political) reform. But groupings among developing countries alone do not obviously provide much advantage from that point of view. Few such agreements (UEMOA is an exception) involve supranational elements, in any case. The second channel through which integration may contribute to an increase in the relative competitiveness of integrating economies vis-à-vis the outside world is through the potential gains that are stressed by the new economics of market integration. To secure these gains evidently calls for measures that go well beyond the implementation of tariff reductions. Deeper measures of policy co-operation will be required to reduce non-tariff barriers (including 296

GLOBAL EXPERIENCE AND INITIATIVES

non-tariff fiscal barriers) that otherwise necessitate the retention of border controls and contribute to market segmentation. So far the new generation of schemes outside Europe promise little in this respect, since they lack specific policies, instruments or programmes for reducing real trade costs or market segmentation except in relation to the provisions relating to the treatment of investment amongst partners. The slender institutional base underlying several of those initiatives also suggests that an internal impetus towards policy deepening in those directions is unlikely. The new regionalism has so far taken the form mainly of initiatives on free trade areas rather than customs unions (or common markets). This in itself reflects lack of agreement over tariff policy among the partners. The overlapping and often inconsistent nature of some of the current supplementary bilateral and multilateral initiatives is a further symptom of a fundamental lack of common policy goals on the part of the countries concerned. This must limit the prospects of overcoming not merely those basic border barriers that at present impede market integration but also other obstacles whose removal hinges on deeper measures of policy integration. In the past, one of the principal problems associated with regional integration among countries with disparate income levels has been the distribution of costs and benefits. It can be argued that there is less need for concern about this matter in the new regionalism, since tariffs are lower and in most cases integration does not involve a common external tariff. But even MERCOSUR, which is working towards a common external tariff, has no provision for addressing distributional issues. The longer-term progress of new customs unions and common markets may ultimately be hindered if polarization should be encouraged by freer trade, and the problem is not addressed by transfers or regional policy initiatives. It is too early for a proper empirical assessment of the performance of the new regionalism in terms of its conventional trade gains or its ability to contribute to export promotion. As to the former, growth in intra-bloc trade is already apparent in several cases. MERCOSUR has shown remarkable dynamism in this respect, and intra-bloc trade has become much greater than it ever was during earlier phases of integration. In the case of CACM, where reforms are in progress, there had also been a significant recovery in intra-bloc trade by 1994 (to 22 per cent of total trade) although that level is still below what was recorded in 1970. In the new 297

GLOBAL EXPERIENCE AND INITIATIVES

policy context, intra-bloc trade is not such an important criterion of gain. It has yet to be seen how foreign trade and investment inflows develop under new regionalism and what part can be ascribed to the regional dimension itself in that process.

NOTES 1 2

3

A report by WTO (1995) lists more than 100 notified preferential trading arrangements. Many are bilateral and some are non-reciprocal. Krugman (1991b) suggested a 50 per cent share as a rule of thumb for the threshold level in determining a ‘natural’ grouping. This would exclude all blocs except the EU. Kreinin and Plummer (1992) look instead at the pattern of intra-bloc trade in terms of revealed comparative advantage. Brunei joined the original five founding members in 1984 and Vietnam in 1995. Laos and Myanmar (Burma) joined in 1997. The accession of Cambodia has been agreed but postponed following a coup. A tenmember ASEAN would be larger in population than either the EU or NAFTA.

298

REFERENCES AND SELECT BIBLIOGRAPHY

Acocella, N. (1991), ‘Stragetic foreign direct investment in the EC’, Economic Notes, vol. 20, pp. 279–302. Aitken, N.D. (1973), ‘The effects of the EEC and EFTA on European trade: a temporal cross-section analysis’, American Economic Review, vol. 63, pp. 881–92. Aliber, R. (1985), ‘Transfer pricing: a taxonomy of impacts on economic welfare’, in A.M.Rugman and L.Eden (eds), Multinationals and Transfer Pricing (London: Croom Helm), pp. 82–97. Allen, P.R. (1975), Organization and Administration of a Monetary Union, Princeton Studies in International Finance (Princeton, N.J.: International Finance Section, Department of Economics, Princeton University). Allen, P.R. (1982), ‘Increased wage or productivity differentials in a monetary union’, in M.T.Sumner and G.Zis (eds), European Monetary Union: Progress and Prospects (London: Macmillan), pp. 195–215. Allen, P.R. (1983), ‘Cyclical imbalance in a monetary union’, Journal of Common Market Studies, vol. 21, pp. 313–27. Allen, P.R. and Kenen, P. (1980), Asset Markets, Exchange Rates and Economic Integration (Cambridge: Cambridge University Press). Allsopp, C. and Vines, D. (1996), ‘Fiscal policy and EMU’, National Institute Economic Review, Ocober, pp. 91–107. Allsopp, C.J., Davies, G. and Vines, D. (1995), ‘Regional macroeconomic policy, fiscal federalism, and European integration’, Oxford Review of Economic Policy, vol. 11, pp. 126–44. Aquino, T. (1978), ‘Intra-industry trade and inter-industry specialisation as concurrent sources of international trade in manufactures’, Weltwirtschaftliches Archiv, vol. 114, pp. 275–95. Aristotelous, K. and Fountas, S. (1996), ‘An empirical analysis of inward foreign direct investment flows in the EU with emphasis on the market enlargement hypothesis’, Journal of Common Market Studies, vol. 34, pp. 571–83. Armstrong, H.W. (1996), ‘An appraisal of the evidence from cross-sectional analysis of the regional growth process within the European Union’, in H.W.Armstrong and R.W.Vickerman (eds), Convergence and Divergence among European Regions (London: Pion), pp. 40–65.

299

SELECT BIBLIOGRAPHY Arndt, S.W. (1968), ‘On discriminatory versus non-preferential tariff policies’, Economic Journal, vol. 78, pp. 971–9. Artis, M. and Miller, M.H. (1979), ‘Inflation, real wages and the terms of trade’, in J.K.Bowers (ed.), Inflation, Development and Integration (Leeds: Leeds University Press). Artis, M. and Ostry, S. (1986), Inter national Economic Policy Coordination, Chatham House Papers, Fifth Series, no. 30 (London: RIIA). Balassa, B. (1962), The Theory of Economic Integration (London: Allen & Unwin). Balassa, B. (1965), ‘Trade liberalization and “revealed” comparative advantage’, The Manchester School, vol. 33, pp. 99–123. Balassa, B. (1966), ‘Tariff reductions and trade in manufactures among industrial countries’, American Economic Review, vol. 56, pp. 466–73. Balassa, B. (1967), ‘Trade creation and trade diversion in the European Common Market’, Economic Journal, vol. 77, pp. 1–21. Balassa, B. (1975), ‘Trade creation and diversion in the European Common Market: an appraisal of the evidence’, in B.Balassa (ed.), European Economic Integration (Amsterdam: North-Holland; New York: Elsevier), pp. 79–118. Balasubramanyam, V. and Greenaway, D. (1992), ‘Economic integration and foreign direct investment: Japanese investment in the EC’, Journal of Common Market Studies, vol. 30, pp. 175–93. Baldwin, R.E, (1989), ‘On the growth effects of 1992’, Economic Policy, no. 9, pp. 248–81. Baldwin, R.E. (1990), ‘On the microeconomics of the European Monetary Union’, European Economy, Special Edition, 1. Baldwin, R.E. (1992), ‘Measurable dynamic gains from trade’, Journal of Political Economy, vol. 100, pp. 162–74. Baldwin, R.E. and Venables, A.J. (1995), ‘Regional economics integration’, in G.M.Grossman and K.Rogoff (eds), Handbook of International Economics (Amsterdam: Elsevier), vol. 3, pp. 1597–644. Baldwin, R., Francois, J. and Portes, R. (1997), ‘The costs and benefits of eastern enlargement: the impact on the EU and Central Europe’, Economic Policy, no. 24, pp. 127–76. Barrell, R. and Whitley, J. (1992), Macroeconomic Policy Coordination in Europe: the ERM and Monetary Union (London: Sage Publications). Barro, R. and Gordon, D. (1983), ‘Rules, discretion and reputation in a model of monetary policy’, Journal of Monetary Economics, vol. 12, pp. 101–21. Barro, R.J. and Sala-i-Martin, X. (1991), ‘Convergence across states and regions’, Brookings Papers on Economic Activity, vol. 1, pp. 107–82. Bayoumi, T. and Eichengreen, B. (1993), ‘Shocking aspects of European monetary integration’, in F.Torres and F.Giavazzi (eds), Adjustment and

300

SELECT BIBLIOGRAPHY Growth in the European Monetary Union (Cambridge: Cambridge University Press), pp. 193–229. de Beers, J.S. (1941), ‘Tariff aspects of a federal union’, Quarterly Journal of Economics, vol. 56, pp. 49–92. Begg, I. (1996), ‘Regulation in the European Union’, Journal of European Public Policy, vol. 3, pp. 525–35. Begg, I., Gudgin, G. and Morris, D. (1995), ‘The assessment: regional policy in the European Union’, Oxford Review of Economic Policy, no. 11, pp. 1–17. Berglas, E. (1979), ‘Preferential trading theory: the n commodity case’, Journal of Political Economy, vol. 87, pp. 315–31. Berglas, E. (1981), ‘Harmonisation of commodity taxes, destination, origin and restricted origin principles’, Journal of Public Economics, vol. 16, no. 3, pp. 377–87. Berglas, E. (1983), ‘The case for unilateral tariff reductions: foreign tariffs rediscovered’, American Economic Review, vol. 73, pp. 1141–2. Bliss, Christopher (1994), Economic Theory and Policy for Trading Blocks (Manchester: Manchester University Press). Boltho, A. and Holtham, G. (1992), ‘The assessment: new approaches to economic growth’, Oxford Review of Economic Policy, vol. 8, pp. 1–14. Bouzas, R. (1997), ‘Mercosur and preferential trade liberalisation in South America: record, issues and prospects’, in R.G.Lipsey and P.Meller (eds), Western Hemisphere Trade Integration (Basingstoke: Macmillan), pp. 58–89. van Brabant, J. (1980), Socialist Economic Integration (Cambridge: Cambridge University Press). Brander, J.A. and Krugman, P.R. (1983), ‘A reciprocal dumping model of international trade’, Journal of International Economics, vol. 15, pp. 313–21. Brander, J. and Spencer, B.J. (1985), ‘Export subsidies and international market share rivalry’, Journal of International Economics, vol. 24, pp. 83–100. Brown, D.K., Deardoff, A.V. and Stern, R.M. (1992), ‘North American integration’, Economic Journal, vol. 102, pp. 1507–18. Buckley, P.J. and Artisien, P. (1987), ‘Policy issues of intra-EC direct investment: British, French and German multinationals in Greece, Portugal and Spain, with special reference to employment effects’, in J.H.Dunning and P.Robson (eds), Multinationals and the European Community (Oxford: Blackwell), pp. 105–28. Buckley, P.J. and Casson, M. (1976), The Future of the Multinational Enterprise (London: Macmillan). Buckwell, A.E., Harvey, D.R., Thomson, K.J. and Parton, K.A. (1982), The Costs of the Common Agricultural Policy (London: Croom Helm).

301

SELECT BIBLIOGRAPHY Buiter, W.H. and Marston, R.C. (eds) (1985), International Economic Policy Co-ordination (Cambridge: Cambridge University Press). van den Bulcke, D. et al. (1979), Investment and Divestment Policies of Multinational Corporations in Europe (Farnborough: Saxon House/ ECSIM). Byé, M. (1950), ‘Unions douanières et données nationales’, Economie Appliquée, vol. 3, pp. 121–57; reprinted in translation as ‘Customs unions and national interests’, International Economic Papers, no. 3 (1953). Calmfors, L. (ed.) (1983), Long-run Effects of Short-run Stabilisation Policy (London: Macmillan). Calmfors, L. and Driffill, J. (1988), ‘Bargaining structure, corporatism and macroeconomic performance’, Economic Policy, no. 6, pp. 13–61. Cantwell, J. (1988), The reorganisation of European industries after integration: selected evidence on the role of transnational enterprise activities’, in J.Dunning and P.Robson (eds), Multinationals and the European Community (Oxford: Blackwell), pp. 25–49. Cantwell, J. (1992), ‘The effects of integration on the structure of multinational corporation activity in the EC’, in M.W.Klein and P.J.J.Welfens (eds), Multinationals in the New Europe and Global Trade (Berlin: Springer). Carnoy, M. (1972), Industrialization in a Latin American Common Market (Washington, D.C.: Brookings Institution). Caves, R. (1980), ‘Investment and location policies of multinational companies’, Schweizerische Zeitschrift für Volkswirtschaft und Statistik, vol. 116, pp. 321–38. Caves, R. (1982), Multinational Enterprise and Economic Analysis (Cambridge: Cambridge University Press). Cawley, R. and Davenport, M. (1988), ‘Partial Equilibrium Calculations of the Impact of Internal Market Barriers in the European Community’, in CEC, Research on the Cost of Non-Europe, vol. 2, Studies on the Economics of Integration, chapter 8 (Luxembourg: European Communities). Cecchini, P. (1988), The European Challenge, 1992 (Aldershot: Wildwood House). Chatterji, M. (1992), ‘Convergence clubs and autonomous growth’, Oxford Review of Economic Policy, vol. 8, pp. 57–69. Cheshire, P. and Carbonaro, G. (1995), ‘Convergence—divergence in regional growth rates: an empty black box?’ in H.W.Armstrong and R.W.Vickerman (eds), Convergence and Divergence among European Regions, European Research in Regional Science 5, pp. 89–111 (London: Pion). Cnossen, S. (ed.) (1987), Tax Co-ordination in the EC (Deventer: Kluwer).

302

SELECT BIBLIOGRAPHY Cnossen, S. (1996), ‘Company taxes in the European Union: criteria and options for reform’, Fiscal Studies, vol. 17, pp. 67–97. Cnossen, S. and Shoup, C.S. (1987), ‘Co-ordination of value-added taxes’, in S.Cnossen (ed.), Tax Co-ordination in the EC, pp. 59–84(Deventer: Kluwer). Cobham, D. (1989), ‘Strategies for monetary integration revisited’, Journal of Common Market Studies, vol. 27, pp. 203–18. Cobham, D. (1996), ‘Causes and effects of the European monetary crises of 1992–93’, Journal of Common Market Studies, vol. 34, pp. 585–604. Cobham, D. and Williams, S. (1997), ‘Hysteresis, the Phillips curve and monetary union’, Applied Economics Letters, forthcoming. Collier, P. (1979), ‘The welfare effects of customs unions: an anatomy’, Economic Journal, vol. 89, pp. 84–95. Commission of the European Communities (monthly) Bulletin of the European Communities (Luxembourg: European Communities). Commission of the Eurpean Communities (1973), Report on the Regional Problems in the Enlarged Community (Thomson report) (Brussels: the Commission). Commission of the European Communities (1977), Report of the Study Group on the Role of Public Finance in European Integration, vols 1 and 2(Brussels: the Commission). Commission of the European Communities (1985), Completing the Internal Market (Luxembourg: European Communities). Commission of the European Communities (1987), Report by the Commission to the Council and Parliament on the Financing of the Community Budget, COM (87) 101 (Brussels: the Commission). Commission of the European Communities (1988a), ‘The Economics of 1992’, European Economy, no. 35. (Republished as M.Emerson et al., The Economics of 1992: The EC Commission’s Assessment of the Economic Effects of Completing the Internal Market, Oxford: Oxford University Press.) Commission of the European Communities (1988b), Research on the Cost of non-Europe, vol. 2, Studies on the Economics of Integration, chapter 8 (Luxembourg: European Communities). Commission of the European Communities (1989), Report on Economic and Monetary Union in the European Community, (Luxembourg: European Communities). Commission of the European Communities (1990a), ‘One market, one money: an evaluation of the potential benefits and costs of forming an economic and monetary union’, European Economy, no. 44. (Republished as M.Emerson et al. (1992), One Market, One Money, Oxford: Oxford University Press.) Commission of the European Communities (1990b), ‘The Economics

303

SELECT BIBLIOGRAPHY ofEMU: Background Studies for One Market, One Money’, European Economy, Special Edition, 1. Commission of the European Communities (1992a), Treaty on European Union (Luxembourg: European Communities). Commission of the European Communities (1992b), Report of the Committee of Independent Experts on Company Taxation (Brussels: the Commission). Commission of the European Communities (1993a), ‘Stable money—sound finances: community public finance in the perspective of EMU’, European Economy, no. 53. Commission of the European Communities (1993b), ‘The economics of community public finance’, European Economy no. 5, Reports and Studies. Commission of the European Communities (1996a), First Cohesion Report, COM (96) 542 (Brussels: the Commission). Commision of the European Communities (1996b), Taxation in the European Union: Report on the Development of Tax Systems, COM (96) 546 (Brussels: the Commission). Commission of the European Communities (1996c), The Impact and Effectiveness of the Single Market, COM (96) 520 (Brussels: the Commission). Commission of the European Communities (1996d), The Single MarketReview, Commission Staff Working Paper, SEC (96) 2378, (Brussels: the Commission). Commission of the European Communities (1996e), The Community Budget: the Facts in Figures (Luxembourg: the Commission). Commission of the European Communities (1997), Single Market Review, thirty-nine volumes published on behalf of the Commission (London: Kogan Page). Commission of the European Economic Community (1963), Report of the Fiscal and Financial Committee on Tax Harmonization in the Common Market (Neumark report) (Brussels: the Commission). Reprinted in translation as Tax Harmonisation in the Common Market (Chicago: Commerce Clearing House). Commission of the European Economic Community (1966), The Development of a European Capital Market (Segré report) (Brussels: the Commission). Cooper, C.A. and Massell, B.F. (1965), ‘Towards a general theory of customs unions for developing countries’, Journal of Political Economy, vol. 73, pp. 461–76. Cooper, R.N. (1985), ‘Economic interdependence and co-ordination of economic policies’, in R.W.Jones and P.B.Kenen (eds), Handbook of International Economics, vol. 2 (Amsterdam: North-Holland), pp. 195–234.

304

SELECT BIBLIOGRAPHY Corden, W.M. (1970), ‘The efficiency effects of trade and protection’, in I.A.McDougall and R.H.Snape (eds), Studies in International Economics (Amsterdam: North-Holland), pp. 1–17. Corden, W.M. (1971), The Theory of Protection (Oxford: Clarendon). Corden, W.M. (1972a), Monetary Integration, Essays in International Finance, no. 93 (Princeton, N.J.: International Finance Section, Princeton University). Corden, W.M. (1972b), ‘Economies of scale and customs union theory’, Journal of Political Economy, vol. 80, pp. 465–75. Corden, W.M. (1974), Trade Policy and Economic Welfare (Oxford: Clarendon). Corden, W.M. (1975), ‘The costs and consequences of protection: a survey of empirical work’, in P.B.Kenen (ed.), International Trade and Finance: Frontiers for Research (Cambridge: Cambridge University Press), pp. 51–91. Corden, W.M. (1976), ‘Customs union theory and the non-uniformity of tariffs’, Journal of International Economics, vol. 6, pp. 99–106. Corden, W.M. (1985), Inflation, Exchange Rates and the World Economy, 3rd edn (Oxford: Clarendon). Cossetti, G. and Roubini, N. (1993), ‘The design of optimal fiscal rules for Europe after 1992’, in F.Torres and F.Giavazzi (eds), Adjustment and Growth in the European Monetary Union (Cambridge: Cambridge University Press), pp. 46–82. Council for Mutual Economic Assistance (1980), The Evening out of Levels of National Economic Development and the International Economic Relations (Moscow: CMEA). Council for Mutual Economic Assistance (n.d.), Basic Documents of the Council for Mutual Economic Assistance (Moscow: CMEA). Cox, D. and Harris, R. (1985), ‘Trade liberalization and industrial organization: some estimates for Canada’, Journal of Political Economy, vol. 93, pp. 115–45. Cross, R. (ed.) (1987), Unemployment, Hysteresis and the Natural Rate Hypothesis (Oxford: Blackwell). Curzon, V. (1974), The Essentials of Economic Integration (London: Macmillan). Cushman. J. (1983), ‘The effects of real exchange rate risk on international trade’, Journal of International Economics, vol. 18, pp. 45–63. Davenport, M. and Page, S. (1991), Europe 1992 and the Developing Countries (London: Overseas Development Institute). Dearden, R.G. (1997), ‘Trade disputes and settlement mechanisms under the Canada—US Free Trade Agreement’, in R.G.Lipsey and P.Meller (eds), Western Hemisphere Trade Integration: a Canadian—Latin American Dialogue (Basingstoke: Macmillan), pp. 207–31.

305

SELECT BIBLIOGRAPHY De Grauwe, P. (1975), ‘Conditions for monetary integration: a geometric interpretation’, Weltwirtschaftliches Archiv, vol. 111, pp. 634–44. De Grauwe, P. (1994), ‘The political economy of monetary union in Europe’, World Economy, vol. 16, pp. 653–61. De Grauwe, P. (1996), ‘How to fix conversion rates at the start of EMU’, EIB Papers, vol. 1, pp. 97–118. De Grauwe, P. and Vanhaverbeke, W. (1993), ‘Is Europe an optimum currency area? Evidence from regional data’, in P.Masson and M.P. Taylor (eds), Policy Issues in the Operation of Currency Unions (Cambridge: Cambridge University Press), pp. 111–29. Delors Committee report (1989), see Commission of the European Communities (1989). Demakas, D.G., Kasper, B., Gupta, S., Lipschitz, L., and Mayer, T. (1988), ‘The effects of the Common Agricultural Policy of the European Community: a survey of the literature’, Journal of Common Market Studies, vol. 27, pp. 113–46. Devarajan, S., and de Melo, J. (1991), ‘Membership in the CFA zone: Odyssean journey or Trojan horse?’, in A.Chibber and S.Fischer (eds), Economic Reform in Sub-Saharan Africa (Washington, D.C.: World Bank). Devereux, M. (1992), ‘The Ruding Committee: an economic assessment’, Fiscal Studies, vol. 13, pp. 96–107. Devereux, M. and Pearson, M. (1989), Corporate Tax Harmonisation and Economic Efficiency (London: Institute for Fiscal Studies). Devereux M. and Pearson, M. (1990), ‘Harmonising corporate taxes in Europe’, Fiscal Studies, vol. 11, pp. 21–35. Dignan, T. (1995), ‘Regional disparities and regional policy in the European Union’, Oxford Review of Economic Policy, vol. 11, no. 2, pp. 64–95. Dixit, A.K. and Kyle, A.S. (1985), ‘The use of protection and subsidies for entry promotion and deterrence’, American Economic Review, vol. 75, pp. 139–52. Dixit, A.K. and Norman, V. (1980), Theory of International Trade (Cambridge: Cambridge University Press). Dixon, R.J. and Thirlwall, A.P. (1975), Regional Growth and Unemployment in the United Kingdom (London: Macmillan). Dosser, D. (1967), ‘Economic analysis of tax harmonization’, in C.S. Shoup (ed.), Fiscal Harmonization in Common Markets (New York and London: Columbia University Press), vol. 1, pp. 1–144. Dosser, D. (1972), ‘Customs unions, tax unions, and development unions’, in R.M.Bird and J.G.Head (eds), Modern Fiscal Issues: Essays in Honor of Carl S.Shoup (Toronto: Toronto University Press), pp. 86–103. Dunford, M. (1993), ‘Regional disparities in the European Community: evidence from the REGIO databank’, Regional Studies, vol. 27, pp. 727–43.

306

SELECT BIBLIOGRAPHY Dunning, J.H. (1977), ‘Trade, location of economic activity and the MNE: a search for an eclectic approach’, in B.Ohlin et al. (eds), The International Allocation of Economic Activity (London: Macmillan), pp. 395–418. Dunning, J.H. (1981), International Production and the Multinational Enterprise (London: Allen & Unwin). Dunning, J.H. (1989), ‘The theory of international production’, in K. Fatemi (ed.), International Trade (London: Taylor & Francis). Dunning, J.H. (1997), ‘The European internal market programme and inbound foreign direct investment’, Journal of Common Market Studies, vol. 35, pp. 1–30. Dunning, J.H. and Robson, P. (eds) (1988), Multinationals and the European Community (Oxford: Blackwell). Eden, L. (1985), ‘The microeconomics of transfer pricing’, in A.M. Rugman and L.Eden (eds), Multinationals and Transfer Pricing (London: Croom Helm), pp. 13–46. Eichengreen, B. (1997), ‘Saving Europe’s automatic stabilizers’, National Institute Economic Review, February, pp. 92–8. El-Agraa, A. (ed.) (1997), Economic Integration Worldwide (Basingstoke: Macmillan). El-Agraa, A. and Jones, A.J. (1980), The Theory of Customs Unions (Oxford: Philip Allan). Emerson, M., Aujean, M., Catinat, M., Goybet, P., and Jacquemin, A. (1988), The Economics of 1992: the E.C.Commission’s Assessment of the Economic Effects of Completing the Internal Market (Oxford: Oxford University Press). Emerson, M., Gros, D., Pisani-Ferry, J., Italianer, A., and Reichenbach, H. (1992), One Market, One Money (Oxford: Oxford University Press). Erzan, R., and Yeats, A. (1992), ‘Free trade agreements with the United States: what’s in it for Latin America?’, World Bank Working Papers no. 872. Ethier, W. and Horn, H. (1984), ‘A new look at economic integration’, in H.Kierszkowski (ed.), Monopolistic Competition and InternationalTrade (Oxford: Oxford University Press). European Coal and Steel Community (1953), Report on the Problems Raised by the Different Turnover Tax Systems Applied within the Common Market (Tinbergen report) (Luxembourg: ECSC). European Free Trade Association (1969), The Effects of EFTA on the Member Countries (Geneva: EFTA). European Monetary Institute (1997), ‘Executive summary of the single monetary policy in Stage 3’, reproduced in Bank of England Quarterly Bulletin, vol. 37, pp. 208–9. Fagerberg, J. and Verspagen, B. (1996), ‘Heading for divergence? Regional growth in Europe reconsidered’, Journal of Common Market Studies, vol. 34, pp. 431–48.

307

SELECT BIBLIOGRAPHY Fleming, M. (1962), ‘Domestic financial policies under fixed and under floating exchange rates’, International Monetary Fund Staff Papers, vol. 9, pp. 369–79. Fleming, M. (1971), ‘On exchange rate unification’, Economic Journal, vol. 81, pp. 467–88. Forte, F. (1977), ‘Principles for the assignment of public economic functions in a setting of multilayer government’, in Commission of the European Communities (1977), vol. 2, pp. 321–58. Franklin, M. (1992), ‘The EC Budget: Realism, Redistribution and Radical Reform’, RIIA Discussion Paper no. 42 (London: Chatham House). Franko, L.G. (1976), The European Multinationals: a Renewed Challenge to American and British Big Business (London: Harper & Row). Frenkel, J. and Goldstein, M. (1990), ‘Monetary policy in an emerging European economic and monetary union’, IMF Working Paper (Washington, D.C.: IMF). Friedman, M. (1953), ‘The case for flexible exchange rates’, in M.Friedman, Essays in Positive Economics (Chicago: Chicago University Press), pp. 157–203. Friedman, M. (1968), ‘The role of monetary policy’, American Economic Review, vol. 58, pp. 1–17. Friedman, M. (1975), Unemployment versus Inflation? An Evaluation of the Phillips Curve (London: Institute of Economic Affairs). Garnaut, R. and Drysdale, P. (eds) (1994), Asia Pacific Regionalism (Sydney: Harper Collins). Gasiorek, M., Smith, A. and Venables, A.J. (1991), ‘Completing the internal market in the EC: factor demands and comparative advantage’, in L.A.Winters and A.J.Venables (eds), European Integration: Trade and Industry (Cambridge: Cambridge University Press) pp. 9–33. Gatsios, K. and Seabright, P. (1989), ‘Regulation in the European Community’, Oxford Review of Economic Policy, vol. 5, pp. 37–60. Genser, B., Haufler, A. and Sorensen, P.B. (1995), ‘Indirect taxation in an integrated market: is there a way of avoiding trade distortions without sacrificing national autonomy?’, Journal of Economic Integration, vol. 10, pp. 178–205. Geroski, P. (1989), ‘The choice between diversity and scale’, in J.Kay (ed), 1992: Myths and Realities (London: London Business School). Giavazzi, F. and Pagano, M. (1988), ‘The advantage of tying one’s hands: EMS discipline and central bank credibility’, European Economic Review, vol. 32, pp. 1055–82. Giavazzi, F. and Spaventa, L. (1990), ‘The “new” EMS’, in P.DeGrauwe and L.Papademos (eds), The European Monetary System in the 1990s (Harlow: Longman), pp. 65–85. Giavazzi, F. and Torres, F. (eds) (1993), Adjustment and Growth in the European Monetary Union (Cambridge: Cambridge University Press).

308

SELECT BIBLIOGRAPHY Giersch, H. (1950), ‘Economic union between nations and the location of industry’, Review of Economic Studies, vol. 17, pp. 87–97. Giovannini, A. (1989), ‘National tax systems versus the European capital market’, Economic Policy, no. 9, pp. 346–74 and 381–6. Goodhart, C.A.E and Smith, S. (1993), ‘Stabilisation’, European Economy, no. 5, pp. 417–55. Gotur, P. (1985), ‘Effects of exchange rate volatility on trade: some further evidence’, International Monetary Fund Staff Papers, vol. 32, pp. 475– 512. Greenaway, D. and Milner, C. (1986), The Economics of Intra-industry Trade (Oxford: Blackwell). Gregory, T.E. (1921), Tariffs: a Study in Method (London: Griffin). Grinols, E.L. (1984), ‘A thorn in the lion’s paw: has Britain paid too much for Common Market membership?’, Jour nal of International Economics, vol. 16, pp. 271–93. Gros, D. and Thygesen, N. (1990), ‘The institutional approach to monetary integration in Europe’, Economic Journal, vol. 100, pp. 925–35. Grubel, H. (1967), ‘Intra-industry specialization and the pattern of trade’, Canadian Journal of Economics and Political Science, vol. 22, pp. 374– 88. Grubel, H. (1982), ‘The theory of international capital movements’, in J. Black and J.H.Dunning (eds), International Capital Movements (London: Macmillan), pp. 1–21. Guillaumont, P., Guillaumont, S., and Plane, P. (1988), ‘Participating in African monetary unions: an alter native evaluation’, World Development, vol. 16, pp. 569–76. Gunter, F.R. (1989), ‘Customs union theory: retrospect and prospect’, in D.Greenaway, T.Hyclak and R.J.Thornton (eds), Economic Aspects of Regional Trading Arrangements (New York: New York University Press). Haaland, J. (1990), ‘Assessing the effects of EC integration on EFTA countries: the position of Norway and Sweden’, Journal of Common Market Studies, vol. 28, pp. 379–400. Haaland, J. and Wooton, I. (1992), ‘Market integration, competition and welfare’, in L.A.Winters (ed.) Trade Flows and Trade Policy after 1992 (Cambridge: Cambridge University Press). Haberler, G. (1936), The Theory of International Trade, trans. A.W. Stomer and F.Benham (London: Hodge). Haberler, G. (1964), ‘Integration and growth in the world economy in historical perspective’, American Economic Review, vol. 54, pp. 1–22. von Hagen, J. (1993), ‘Monetary union and fiscal union: a perspective from fiscal federalism’, in P.Masson and M.P.Taylor (eds), Policy Issues in the Operation of Currency Unions (Cambridge: Cambridge University Press), pp. 264–96.

309

SELECT BIBLIOGRAPHY von Hagen, J. and Harden, I.J. (1994), ‘National budget processes and fiscal performance’, European Economy no. 3 (Reports and Studies), pp. 311–408. Hamada, K. (1985), ‘The political economy of monetary integration: a public economics approach’, in The Political Economy of International Monetary Interdependence (Cambridge, MA: MIT Press), pp. 27–44. Hamilton. B. and Whalley, J. (1985), ‘Geographically discriminatory trade arrangements’, Review of Economics and Statistics, vol. 67, pp. 446–55. Han, S.S. and Liesner, H.H. (1971), Britain and the Common Market: the Effect of Entry on the Pattern of Manufacturing Production, University of Cambridge Department of Applied Economics, Occasional Paper 27 (Cambridge: Cambridge University Press). Harris, R. and Cox, D. (1984), Trade, Industrial Policy and Canadian Manufacturing (Toronto: Ontario Economic Council). Heiss, H.W. (1970), ‘The Council of Mutual Economic Assistance: developments since the mid-1960s’, in Economic Development in Countries of Eastern Europe, Joint Economic Committee Print (Washington D.C.: Government Printing Office). Helliwell, J.F. and Padmore, T. (1985), ‘Empirical studies of macroeconomic interdependence’, in R.W.Jones and P.B.Kenen (eds), Handbook of International Economics, vol. 2 (Amsterdam: NorthHolland), pp. 1107–51. Helpman, E. (1984), ‘A simple theory of international trade with multinational corporations’, Journal of Political Economy, vol. 79, pp. 1059–72. Helpman, E. and Krugman, P.R. (1985), Market Structure and Foreign Trade (Brighton: Wheatsheaf Press). Helpman, E. and Krugman, P.R. (1989), Trade Policy and Market Structure (Cambridge, MA: MIT Press). Hewett, E.A. (1974), Foreign Trade Prices in the Council for Mutual Economic Assistance (London and New York: Cambridge University Press). Hewett, E.A. (1977). ‘Prices and resource allocation in intra-CMEA trade’, in A.Abouchar (ed.), The Socialist Price Mechanism (Durham, N.C.: Duke University Press), pp. 95–128. Hewett, E.A. (1984), ‘Soviet economic relations with the CMEA countries’, in P.Joseph (ed.), The Soviet Economy after Brezhnev (Brussels: NATO), pp. 241–53. Hirschman, A.O. (1958), The Strategy of Economic Development (New Haven, CO: Yale University Press). Holland, S. (1980), Uncommon Market (London: Macmillan). Holzman, F.D. (1974), Foreign Trade under Central Planning (Cambridge, MA: Harvard University Press).

310

SELECT BIBLIOGRAPHY Holzman, F.D. (1976), International Trade under Communism: Politics and Economics (London: Macmillan). Holzman, F.D. (1985), ‘Comecon: a trade-destroying customs union?’, Journal of Comparative Economics, vol. 9, pp. 410–23. Horst, T. (1973), ‘The simple analytics of multinational firm behaviour’, in M.B.Connolly and A.K.Swoboda (eds), International Trade and Money (London: Allen & Unwin), pp. 72–84. Horstman, I. and Markusen, J. (1992), ‘Endogenous market structures in international trade’, Journal of International Economics, vol. 32, pp. 109–29. Hufbauer, G.C. (1974), ‘Multinational corporations and the international adjustment process’, American Economic Review, vol. 64, pp. 271–5. Hufbauer, G.C. (1990), Europe 1992: an American Perspective (Washington, D.C.: Brookings Institution). Hufbauer, G.C. and Schott, J.J. (1993), NAFTA: an Assessment (Washington, D.C.: Institute for International Economics). Hufbauer, G.C. and Schott, J.J. (1995), Western Hemisphere Economic Integration (Washington, D.C.: Institute for International Economics). Hymer, S.H. (1970), ‘The efficiency (contradictions) of multinational corporations’, American Economic Review, vol. 60, pp. 441–8. Imada, P. (1993), ‘Production and trade effects of the ASEAN Free Trade Area’, The Developing Economics, vol. 31. Ingram, J.C. (1973), The Case for European Monetary Integration, Essays in International Finance no. 98 (Princeton, N.J.: International Finance Section, Princeton University). Institute for Latin American Integration (1984), The Latin American Integration Process in 1984 (Buenos Aires: Inter-American Development Bank). International Monetary Fund (1984), Exchange Rate Volatility and World Trade, Occasional Paper no. 28 (Washington, D.C.: IMF). IMF (1986), The European Monetary System: Recent Developments, Occasional Paper no. 48 (Washington, D.C.: IMF). Ishiyama, Y. (1975), ‘The theory of optimum currency areas: a survey’, International Monetary Fund Staff Papers, vol. 22, pp. 344–83. Jacquemin, A. (ed.) (1984), European Industry: Public Policy and Corporate Strategy (Oxford: Oxford University Press). Jenkins, R. (1978), ‘European monetary union’, Lloyds Bank Review, no. 127, pp. 1–14. Jensen-Butler, C. (1987), ‘The regional economic effects of European integration’, Geoforum, vol. 18, pp. 213–27. Johnson, H.G. (1960), ‘The cost of protection and the scientific tariff’, Journal of Political Economy, vol. 68, pp. 327–45. Johnson, H.G. (1961), ‘Towards a general theory of the balance of payments’, in International Trade and Economic Growth: Studies in Pure Theory (Cambridge, MA: Harvard University Press), pp. 153–8.

311

SELECT BIBLIOGRAPHY Johnson, H.G. (1962), ‘The economic theory of customs unions’, Pakistan Economic Journal, vol. 10, pp. 14–32; reprinted in H.G.Johnson, Money, Trade and Economic Growth (London: Allen & Unwin). Johnson, H.G. (1965), ‘An economic theory of protectionism, tariff bargaining and the formation of customs unions’, Journal of Political Economy, vol. 73, pp. 256–83. Johnson, H.G. and Krauss, M. (1970), ‘Border taxes, border tax adjustments, comparative advantage, and the balance of payments’, Canadian Journal of Economics, vol. 3, pp. 595–602. Kaldor, N. (1970), ‘The case for regional policies’, Scottish Journal of Political Economy, vol. 17, pp. 337–48. Kaldor, N. (1971), ‘The dynamic effects of the Common Market’, in D. Evans (ed.), Destiny or Delusion: Britain and the Common Market (London: Gollancz), pp. 59–91. Kaser, M. (1967), Comecon: Integration Problems of the Planned Economies, 2nd edn (London: Oxford University Press). Kay, J. (1991), ‘Industrial collaborative activity and the completion of the internal market’, Journal of Common Market Studies, vol. 29, pp. 347–62. Keen, M. (1993), ‘The welfare economics of tax co-ordination in theEuropean Community; a survey’, Fiscal Studies, vol. 14, no. 2, pp. 15–36. Keen, M. and Lahiri, S. (1994), ‘The comparison between destination and origin taxation under imperfect competition’, Working Paper Series (London: Institute of Fiscal Studies). Keen, M. and Smith, S. (1996), ‘The future of VAT in the European Community’, Economic Policy, no. 23, pp. 375–420 (including discussion). Kemp, M.C. (1969), A Contribution to the General Equilibrium Theory of Preferential Trading (Amsterdam: North-Holland). Kemp, M. and Wan, H. (1976), ‘An elementary proposition concerning the formation of customs unions’, in M.Kemp (ed.), Three Topics in the Theory of International Trade: Distribution, Welfare and Uncertainty (Amsterdam: North-Holland). Kenen, P. (1969), ‘The theory of optimum currency areas: an eclectic view’, in R.A.Mundell and A.K.Swoboda (eds), Monetary Problems of the International Economy (Chicago, Ill.: University of Chicago Press), pp. 41–60. Kenen, P. (1976), Capital Mobility and Financial Integration: a Survey, Princeton Studies in International Finance (Princeton, N.J.: International Finance Section, Department of Economics, Princeton University). Kenen, P. (1995), Economic and Monetary Union in Europe (Cambridge: Cambridge University Press). Kindleberger, C.P. (1966), ‘European integration and the international corporation’, Columbia Journal of World Business, vol. 1, pp. 65–73. Klein, L.R. and Salvatore, D. (1995), ‘Welfare effects of the North American Free Trade Agreement’, Journal of Policy Modelling, vol. 17, pp. 163–176.

312

SELECT BIBLIOGRAPHY Klepper, G. (1990), ‘Entry into the market for large transport aircraft’, European Economic Review, vol. 34, pp. 775–803. Knickerbocker, F.T. (1973), ‘Oligopolistic Reaction and Multinational Enterprise (Boston, MA: Harvard Business School). Kohn, M.J. (1976), ‘Developments in Soviet—Eastern European terms of trade, 1971–75’, in Joint Economic Committee, Congress of the US, Soviet Economy in a New Perspective (Washington, D.C.: Government Printing Office). Kojima, K. (1971), Japan and a Pacific Free Trade Area (London: Macmillan). Kopits, G. (1997), Are Europe’s Social Security Finances Compatible with EMU?, IMF Paper on Policy Analysis and Assessment no. 97/3 (Washington, D.C.: IMF). Kowalczyk, C. (1990), Welfare and Customs Unions, National Bureau of Economic Research Working Paper no. 4983 (Cambridge, MA: NBER). Krauss, M.B. (1972), ‘Recent developments in customs union theory: an interpretative survey’, Journal of Economic Literature, vol. 10, pp. 413–36. Kreinin, M.E. (1969), ‘Trade creation and diversion by the EEC and EFTA’, Economia Internazionale, vol. 22, pp. 273–80. Kreinin, M.E. (1972), ‘Effects of the EEC on imports of manufactures’, Economic Journal, vol. 82, pp. 897–920. Kreinin, M.E. (1973), ‘The static effects of EEC enlargement on trade flows’, Southern Economic Journal, vol. 39, pp. 559–68. Kreinin, M. and Plummer, M.G. (1992), ‘“Natural” Economic Blocs: an Alternative Formulation’ (mimeo). Krueger, A. (1983), Exchange Rate Determination (Cambridge: Cambridge University Press). Krugman, P. (1979), ‘Increasing returns, monopolistic competition andinternational trade’, Journal of International Economics, vol. 9, pp. 469–79. Krugman, P. (1983), ‘The “new” theories of international trade and the multinational enterprise’, in C.P.Kindleberger and D.B.Andretsch (eds), The Multinational Corporation in the 1980s (Cambridge, MA: MIT Press), pp. 57–73. Krugman, P. (1989), ‘Differences in income elasticities and trends in real exchange rates’, European Economic Review, vol. 33, pp. 1031–47. Krugman, P. (1991a), Geography and Trade (Leuven: Leuven University Press). Krugman, P. (1991b), ‘Is bilateralism bad?’, in E.Helpman and A. Razin (eds), International Trade and Policy (Cambridge, MA: MIT Press), pp. 9–23. Krugman, P. (1995), Development, Geography and Economic Theory (Cambridge, MA: MIT Press). Krugman, P. and Venables, A.J. (1990), ‘Integration and the competitiveness of peripheral industry’, in C.J.Bliss and J.Braga de

313

SELECT BIBLIOGRAPHY Macedo (eds), Unity with Diversity within the European Community (Cambridge: Cambridge University Press), pp. 56–77. Krugman, P. and Venables, A.J. (1993), Integration, Specialisation and Adjustment, Discussion Paper no. 4559 (Cambridge, MA: NBER). Laidler, D. (1986), ‘The new classical contribution to economics’, Banca Nazionale del Lavoro Quarterly Review, no. 156, pp. 27–55. Lamfalussy, A. (1963), ‘Intra-European trade and the competitive position of the EEC’, Transactions of the Manchester Statistical Society, session 1962–63, pp. 1–19. Langhammer, R.J. (1991), ‘ASEAN economic co-operation: a stock-taking from a political economy point of view’, ASEAN Economic Bulletin, vol. 8, pp. 137–50. Langhammer, R.J. (1995), ‘Regional integration in East Asia: from marketdriven regionalization to institutionalized regionalism?’, Weltwirtschaftliches Archiv, vol. 131, pp. 167–201. Lavigne, M. (1983), ‘The Soviet Union inside Comecon’, Soviet Studies, vol. 35, no. 2, pp. 135–53. Leibenstein, H. (1966), ‘Allocative efficiency versus “X-efficiency”’, American Economic Review, vol. 56, pp. 392–415. Levine, R. and Renelt, D. (1992), ‘A sensitivity analysis of cross-country growth regressions’, American Economic Review, vol. 82, pp. 942–63. Lipsey, R.G. (1960), ‘The theory of customs unions: a general survey’, Economic Journal, vol. 70, pp. 496–513. Lipsey, R.G. (1970), The Theory of Customs Unions: a General Equilibrium Analysis (London: Weidenfeld & Nicolson). Lipsey, R.G. and Smith, M.G. (1985), Taking the Initiative: Canada’s Trade Options in a Turbulent World (Toronto: C.D.Howe Institute). Lipsey, R.G. and Meller, P. (eds) (1997), Western Hemisphere Trade Integration (Basingstoke: Macmillan). List, F. (1885), The National System of Political Economy (London: Longman). Lloyd, P.J. (1982), ‘A 3×3 theory of customs unions’, Journal of International Economics, vol. 12, pp. 41–63. Lucas, R.E. (1988), ‘On the mechanics of economic development’, Journal of Monetary Economics, vol. 22, pp. 3–42. Lunn, J. (1980), ‘Determinants of US direct investment in the EEC: further evidence’, European Economic Review, vol. 13, pp. 93–101. McCulloch, J.R. (1832), A Dictionary of Commerce (London: Longman). MacDonald, R. (1988), Fluctuating Exchange Rates: Theories and Evidence (London: Allen & Unwin). MacDonald, R. and Taylor, M.P. (1992), ‘Exchange rate economics’, International Monetary Fund Staff Papers, vol. 39, pp. 1–57. MacDougall, G.D.A. (1960), ‘The benefits and costs of private investment from abroad: a theoretical approach’, Economic Record, vol. 36, pp. 13–35.

314

SELECT BIBLIOGRAPHY MacDougall, G.D.A. (1975), ‘Discussion Paper’, in Report of the Study Group on Economic and Monetary Union in 1980, Annex II (Brussels: the Commission), pp. 97–103. MacDougall, G.D.A. (1992), ‘Economic and Monetary Union and the European Community budget’, National Institute Economic Review, May, pp. 64–8. MacDougall report (1977), see CEC (1977). McKinnon, R.I. (1963), ‘Optimum currency areas’, American Economic Review, vol. 53, pp. 717–24. McKinnon, R.I. (1981), ‘The exchange rate and macroeconomic policy: changing post-war perceptions’, Journal of Economic Literature, vol. 19, pp. 531–57. McMillan, J. and McCann, E. (1981), ‘Welfare effects in customs unions’, Economic Journal, vol. 91, pp. 697–703. Machlup, F. (ed.) (1976), Economic Integration: Worldwide, Regional, Sectoral (London: Macmillan). Machlup, F. (1977), A History of Thought on Economic Integration (London: Macmillan). Makower, H. and Morton, G. (1953), ‘A contribution towards a theory of customs unions’, Economic Journal, vol. 63, pp. 33–49. Marer, P. (1985), Dollar GNPs of the USSR and Easter n Europe (Washington, D.C.: World Bank). Markusen, J.R. (1985), ‘Canadian gains from trade in the presence of economies of scale and imperfect competition’ in J.Whalley and R. Hill (eds), Canada—United States Free Trade (Toronto: University of Toronto Press). Markusen, J.R. (1995), ‘The boundaries of multinational enterprises and the theory of international trade’, Journal of Economic Perspctives, vol. 9, pp. 169–89. Markusen, J.R. and Venables, A. (1995), ‘Multinational firms and the new trade theory’, NBER Working Paper no. 5036 (Cambridge, MA: NBER). Marrese, M. and Vanous, J. (1983), Soviet Subsidization of Trade with Eastern Europe: a Soviet Perspective (Berkeley, CA: Institute of International Studies). Masson, P.R. (1996), ‘Fiscal dimensions of EMU’, Economic Journal, vol. 106, pp 996–1005. Masson, P.R. and Taylor, M.P. (eds) (1993), Policy Issues in the Operation of Currency Unions (Cambridge: Cambridge University Press). Mayes, D. (1978), ‘The effects of economic integration on trade’, Journal of Common Market Studies, vol. 17, pp. 1–25. Meade, J.E. (1953), Problems of Economic Union (London: Allen & Unwin). Meade, J.E. (1955a), The Theory of Customs Unions (Amsterdam: NorthHolland).

315

SELECT BIBLIOGRAPHY Meade, J.E. (1955b), Trade and Welfare (London: Oxford University Press). Meade, J.E. (1957), ‘The balance of payments problems of a European free trade area’, Economic Journal, vol. 67, pp. 379–96. Meade, J.E. (1974), ‘A note on border tax adjustments’, Journal of Political Economy, vol. 82, pp. 1013–15. Mélitz, J. (1994), ‘French monetary policy and recent speculative attacks on the franc’, in D.Cobham (ed.), European Monetary Upheavals (Manchester: Manchester University Press), pp. 61–77. Mélitz, J. (1995), ‘The current impasse in research on optimum currency areas’, European Economic Review, vol. 39, pp. 492–500. de Melo, J. and Panagariya, A. (eds) (1993), New Dimensions in Regional Integration (Cambridge: Cambridge University Press). de Melo, J., Panagariya, A., and Rodrik, D. (1993), ‘The new regionalism: a country perspective’, in J.de Melo and A.Panagariya (eds), New Dimensions in Regional Integration (Cambridge: Cambridge University Press), pp. 159–93. Mendershausen, H. (1959), ‘Terms of trade between the Soviet Union and smaller Communist countries, 1955–57’, Review of Economics and Statistics, vol. 41, pp. 106–18. Mendershausen, H. (1960), ‘The terms of Soviet—satellite trade: a broadened analysis’, Review of Economics and Statistics, vol. 42, pp. 152–63. Mennes, L.B.M. (1973), Planning Economic Integration among Developing Countries (Rotterdam: Rotterdam University Press). Michaely, M. (1976), ‘The assumptions of Jacob Viner’s theory of customs unions’, Journal of International Economics, vol. 6, pp. 75–93. Miller, M.H. (1971), ‘Estimates of the static balance of payments and welfare costs of United Kingdom entry into the Common Market’, National Institute Economic Review, no. 57, pp. 69–83. Miller, M.H. and Spencer, J.E. (1977), ‘The static economic effects of the UK joining the EEC: a general equilibrium approach’, Review of Economic Studies, vol. 44, pp. 71–93. Molle, W. and Morsink, R. (1990), ‘Direct investments and monetary integration’, in European Economy, Special Edition, pp. 81–101. Molle, W. and Morsink, R. (1991), ‘Intra-European direct investment’, in B.Burgenmaier and J.Mucchielli (eds), Multinationals and Europe 1992 (London: Routledge), pp. 81–101. Monti, M. (1996), The Single Market and Tomorrow’s Europe: a Progress Report from the European Commission (London: Kogan Page). Mundell, R.A. (1961), ‘A theory of optimum currency areas’, American Economic Review, vol. 53, pp. 657–64. Mundell, R.A. (1962), ‘The appropriate use of monetary and fiscal policy for internal and external stability’, International Monetary Fund Staff Papers, vol. 9, pp. 70–7.

316

SELECT BIBLIOGRAPHY Mundell, R.A. (1964), ‘Tariff preferences and the terms of trade’, The Manchester School, vol. 32, pp. 1–13. Mundell, R.A. (1973), ‘Uncommon arguments for common currencies’, in H.G.Johnson and A.K.Swoboda (eds), The Economics of Common Currencies (London: Allen & Unwin), pp. 114–32. Musgrave, P. (1967), ‘Harmonization of direct business taxes: a case study’, in C.S.Shoup (ed.) Fiscal Harmonization in Common Markets (New York and London: Columbia University Press), vol. 2, pp. 207–343. Musgrave, R.A. (1969), Fiscal Systems (New Haven, Conn., and London: Yale University Press). Myrdal, G. (1957), Economic Theory and Underdeveloped Regions (London: Duckworth; reprinted London: Methuen, 1963). Neary, P. (1988), ‘Tariffs, quotas and voluntary export restraints with and without internationally mobile capital’, Canadian Journal of Economics, vol. 21, pp. 714–35. Negishi, T. (1969), ‘The customs union and the theory of second best’, International Economic Review, vol. 10, pp. 391–7. Neumark report (1963), see CEEC (1963). Neven, D. (1990), ‘EC integration towards 1992: some distributional aspects’, Economic Policy, no. 10, pp. 13–62. Neven, D. and Siotis G. (1993), ‘Foreign direct investment in theEuropean Community: some policy issues’, Oxford Review of Economic Policy, vol. 9, no. 6, pp. 72–93. Norman, G. and Dunning, J.H. (1984), ‘Intra-industry foreign direct investment: its rationale and trade effects’, Weltwirtschaftliches Archiv, vol. 120, pp. 522–40. Norman, V.D. (1989), ‘EFTA and the internal European market’, Economic Policy, vol. 9, pp. 424–61. North Atlantic Treaty Organization (1977), Comecon: Progress and Prospects (Brussels: NATO). NATO (1983), External Economic Relations of CMEA Countries: their Significance and Impact in a Global Perspective (Brussels: NATO). NATO (1987), The Economies of Eastern Europe and their Foreign Economic Relations (Brussels: NATO). Oates, W. (1972), Fiscal Federalism (New York: Harcourt Brace Jovanovich). Oates, W. (1977), ‘Fiscal federalism in theory and practice: applications to the European Community’, in CEC (1977), vol. 2, pp. 279–320. O’Brien, D. (1976), ‘Customs unions: trade creation and trade diversion in historical perspective’, History of Political Economy, vol. 8, pp. 540–63. Ohlin, B., Hesselborn, P. and Wijkman, P.M. (eds) (1977), The International Allocation of Economic Activity (London: Macmillan). Oman, C. (1994), Globalisation and Regionalisation: the Challenge for Developing Countries (Paris: OECD). Oudiz, G. (1985), ‘European policy co-ordination: an evaluation’, Centre

317

SELECT BIBLIOGRAPHY for Economic Policy Research, Discussion Paper no. 79 (London: CEPR). Oudiz, G. and Sachs, J. (1984), ‘Macroeconomic policy co-ordination among the industrial economies’, Brookings Papers on Economic Activity, vol. 1, pp. 1–75. Owen, N. (1983), Economies of Scale, Competitiveness, and Trade Patterns within the European Community (Oxford: Clarendon Press). Padoa-Schioppa, T. (1985), ‘Policy co-ordination and the EMS: experience in economic policy co-operation’, in W.H.Buiter, and R.C. Marston (eds), International Economic Policy Co-ordination (Cambridge: Cambridge University Press), pp. 331–59. Padoa-Schioppa, T. (1987), Efficiency, Stability and Equity: a Strategy forthe Evolution of the Economic System of the European Community (Oxford: Oxford University Press). Padoa-Schioppa, T. (1988), ‘The European Monetary System: a longterm view’, in F.Giavazzi, S.Micossi and M.Miller (eds), The European Monetary System, (Cambridge: Cambridge University Press), pp. 369–84. Pagno, M. (1995), ‘On competing theories of economic growth: crosscountry evidence’, International Review of Applied Economics, vol. 9, pp. 249–74. Panic, M. (1991), ‘The impact of multinationals on national economic policy’, in B.Burgenmeier and J.L.Mucchielli (eds), Multinationals and Europe 1992 (London: Routledge), pp. 204–22. Pelkmans, J. and Gremmen, H. (1983), ‘The empirical measurement of static customs union effects’, Rivista Internazionale di Scienze Economiche e Commerciali, vol. 30, no. 7. Pelkmans, J. (1984), Market Integration in the European Community (The Hague: Nijhoff). Pelkmans, J. (1987), ‘The new approach to technical harmonisation and standardisation’, Journal of Common Market Studies, vol. 25, pp. 249–69. Pelkmans, J. and Robson, P. (1987), ‘The aspirations of the White Paper’, Journal of Common Market Studies, vol. 25, pp. 181–92. Pelkmans, J. and Winters, L.A. (1988), Europe’s Domestic Market (London: Routledge). Pelzman, J. (1977), ‘Trade creation and trade diversion in the Council of Mutual Economic Assistance, 1954–70’, American Economic Review, vol. 67, pp. 713–22. Petith, H. (1977), ‘European integration and the terms of trade’, Economic Journal, vol. 87, pp. 262–72. Phelps, E. (1967), ‘Phillips curves, expectations of inflation and optimal unemployment over time’, Economica, vol. 34, pp. 254–81. Phillips, A.W. (1958), ‘The relation between unemployment and the rate of change of money wages in the United Kingdom 1862–1957’, Economica, vol. 25, pp. 283–99.

318

SELECT BIBLIOGRAPHY Prewo, W.E. (1974), ‘Integration effects in the EEC: an attempt at quantification in a general equilibrium framework’, European Economic Review, vol. 5, pp. 379–405. Pugno, M. (1995), ‘On competing theories of economic growth: crosscountry evidence’, International Review of Applied Economics, vol. 9, pp. 249–74. Ricardo, D. (1817), The Principles of Political Economy and Taxation, ed. P. Sraffa (Cambridge: Cambridge University Press, 1951). Richardson, M. (1995), ‘On the interpretation of the Kemp—Wan theorem’, Oxford Economic papers, vol. 47, pp. 696–703. Riezman, R. (1979), ‘A 3×3 model of customs unions’, Journal of International Economics, vol. 9, pp. 341–54. Robson, P. (1968), Economic Integration in Africa (London: Allen & Unwin; Evanston, Ill: Northwestern University Press). Robson, P. (1970), ‘The distribution of gains in customs unions between developing countries’, Kyklos, vol. 23, pp. 117–19. Robson, P. (1971), Fiscal Compensation and the Distribution of Benefits in Economic Groupings of Developing Countries (New York: United Nations). Robson, P. (ed.) (1972), Inter national Economic Integration (Harmondsworth: Penguin). Robson, P. (1983), Integration, Development and Equity: Economic Integration in West Africa (London: Allen & Unwin). Robson, P. (1988), ‘The West African Economic Community’, in A.M ElAgraa (ed.), International Economic Integration, 2nd edn (London: Macmillan), pp. 162–88. Robson, P. (1993), Transnational Corporations and Regional Economic Integration (London: Routledge). Robson, P. (1997), ‘Integration in Sub-Saharan Africa’, in Ali El-Agraa (ed.), Economic Integration Worldwide (London: Macmillan), pp. 348–67. Robson, P. and Wooton, I. (1993), ‘The transnational enterprise and regional economic integration’, Journal of Common Market Studies, vol. 31, pp. 71–90. Rollo, J.M.C. and Warwick, K.S. (1979), The CAP and Resource Flows among EEC Member States, Government Economic Service Working Paper no. 27 (London: Ministry of Agriculture, Fisheries and Food). Romer, P. (1986), ‘Increasing returns and long-run growth’, Journal of Political Economy, vol. 94, pp. 1002–37. Romer, P. (1989), Human Capital and Growth, NBER Working Paper No. 3173. Rossi, V. et al. (1986), ‘Exchange rates, productivity, and international competitiveness’, Oxford Review of Economic Policy, vol. 2, no. 3, pp. 56–73. Ruding Committee report: symposium (1992), Fiscal Studies, vol. 13, no. 2, pp. 84–121.

319

SELECT BIBLIOGRAPHY Rugman, A. and Verbeke, A. (1988), ‘Strategic responses to free trade’, Hitotsubashi Journal of Commerce and Management, vol. 23, pp. 69–79. Rugman, A. and Verbeke, A. (1990), ‘Multinational corporate strategy and the Canada—US Free Trade Agreement’, Management International Review, vol. 30, pp. 253–66. Sala-i-Martin, X. (1994), ‘Regional Cohesion: Evidence and Theories of Regional Growth and Convergence’, Discussion Paper no. 1075 (London: Centre for Economic Policy Research). Santos, P. (1993), ‘The spatial implications of economic and monetary union’, in ‘The Economics of Community Public Finance’, European Economy, no. 5, pp. 353–68. Sapir, A. (1996), ‘The effects of Europe’s internal market programme on production and trade: a first assessment’, Weltwirtschaftliches Archiv, vol. 132, pp. 457–75. Sato, M. and Bird, R.M. (1975), ‘International aspects of the taxation of corporations and shareholders’, International Monetary Fund Staff Papers, vol. 22, pp. 384–455. Sawyer, W.C. (1984), ‘The effects of the second enlargement of the EC on US exports to Europe’, Weltwirtschaftliches Archiv, vol. 120, pp. 572–9. Scaperlanda, A. and Balough, R. (1983), ‘Determinants of US direct investment in the EEC—revisited’, European Economic Review, vol. 21, pp. 381–90. Scherer, F.M., Beckenstein, A., Kaufer, E., and Murphy, R.D. with BougeonMaassen, F. (1975), The Economics of Multi-plant Operations: an International Comparisons Study (Cambridge, MA: Harvard University Press). Scitovsky, T. (1958), Economic Theory and Western European Integration (London: Allen & Unwin). Scully, G.W. and Yu, E.S.H. (1974), ‘International investment, trade diversion and trade creation’, Economic Record, vol. 50, pp. 600–4. Segré report (1966), see CEEC (1966). Sellekaerts, W. (1973), ‘How meaningful are empirical studies on trade creation and trade diversion?’, Weltwirtschaftliches Archiv, vol. 109, pp. 519–51. Sheffrin, S.M. (1983), Rational Expectations (Cambridge: Cambridge University Press). Shibata, H. (1967), ‘The theory of economic unions: a comparative analysis of customs unions, free trade areas, and tax unions’, in C.S. Shoup (ed.) Fiscal Harmonization in Common Markets (New York: Columbia University Press) vol. 1, pp. 145–264. Shoup, C.S. (1953), ‘Taxation aspects of economic integration’, Travaux de l’Institut International de Finances Publiques, ninth session (The Hague: Stockum); reprinted in abridgement in P.Robson (ed.), International Economic Integration (Harmondsworth: Penguin), pp. 197–218.

320

SELECT BIBLIOGRAPHY Shoup, C.S. (ed.) (1967), Fiscal Harmonization in Common Markets, vol. 1, Theory; vol. 2, Practice (New York and London: Columbia University Press). Smith, A. (1776), An Inquiry into the Nature and Causes of the Wealth of Nations, ed. R.H.Campbell, A.S.Skinner and W.B.Todd (Oxford: Clarendon Press, 1976). Smith, A. (1987), ‘Strategic investment, multinational corporations and trade policy’, European Economic Review, vol. 31, pp. 89–96. Smith, A. and Venables, A.J. (1988a), ‘Completing the internal market in the European Community: some industry simulations’, European Economic Review, vol. 32, pp. 1501–25. Smith, A. and Venables, A.J. (1988b), ‘The costs of non-Europe: an assessment based on a formal model of imperfect competition and economies of scale’, in CEC, Studies on the Economics of Integration (Luxembourg: CEC), vol. 2. Smith, A., Venables, A. and Gasoriek, M. (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). Smith, S. (1992), ‘Financing the European Community: a review of options for the future’, Fiscal Studies, vol. 13, no. 4, pp. 98–127. Sobell, V. (1984), The Red Market: Industrial Co-operation and Specialisation in Comecon (Aldershot: Gower). Solow, R.M. (1956), ‘A contribution to the theory of economic growth’, Quarterly Journal of Economics, vol. 70, pp. 65–94. Spaventa, L., Koopmans, L., Salmon, P., Spahn, B. and Smith, S. (1986), The Future of Community Finance, CEPS Papers no. 30 (Brussels: Centre for European Policy Studies). Spraos, J. (1964), ‘The conditions for a trade-creating customs union’, Economic Journal, vol. 74, pp. 101–8. Statistical Office of the European Communities (1986), Eurostat Review 1975–84 (Luxembourg: European Communities). Stopford, J. and Baden Fuller, C. (1991), ‘Globalisation frustrated: the case of white goods’, Strategic Management Journal, vol. 12, pp. 473–507. Stopford, J. and Turner, L. (1985), Britain and the Multinationals (Chichester: Wiley). Sumner, M. (1976), ‘European monetary union and the control of Europe’s inflation rate’, in M.Parkin and G.Zis (eds), Inflation in the World Economy (Manchester: Manchester University Press; Toronto: Toronto University Press), pp. 97–112. Swan, T.W. (1955), ‘Longer-run problems of the balance of payments’, reprinted in R.E.Caves and H.G.Johnson (eds), Readings in International Economics (Homewood, Ill: Irwin), pp. 455–64. van den Tempel, A.J. (1969), Corporation Tax and Individual Income Tax in the European Communities (Brussels: Taxation Directorate, Commission of the European Communities).

321

SELECT BIBLIOGRAPHY Thirlwall, A.P. (1974), ‘Regional economic disparities and regional policy in the Common Market’, Urban Studies, vol. 11, pp. 1–12. Thirlwall, A.P. and Sanna, G. (1996), ‘The macro determinants of growth and new growth theory: an evaluation and further evidence’, in P.Arestis (ed.), Employment, Economic Growth and the Tyranny of the Market: Essays in Honour of Paul Davidson, vol. 2 (Aldershot: Edward Elgar). Thomson report (1973), see CEC (1973). Tinbergen, J. (1952), On the Theory of Economic Policy (Amsterdam: North-Holland). Tinbergen, J. (1965), International Economic Integration, 2nd edn (Amsterdam: Elsevier). Tinbergen report (1953), see European Coal and Steel Community (1953). Tindemans, L. (1976), ‘European union: report by Mr Leo Tindemans to the European Council’, Bulletin of the European Communities, vol. 9supplement no. 1. Tironi, E. (1982), ‘Customs union theory in the presence of foreign firms’, Oxford Economic Papers, vol. 34, pp. 15–71. Torres, F. and Giavazzi, F. (eds) (1993), Adjustment and Growth in the European Monetary Union (Cambridge: Cambridge University Press). Truman, E.M. (1969), ‘The European Economic Community: trade creation and trade diversion’, Yale Economic Essays, vol. 9, pp. 201–57. Truman, E.M. (1972), ‘The production and trade of manufactured products in the EEC and EFTA: a comparison’, European Economic Review, vol. 3, pp. 271–90. Truman, E.M. (1975), ‘The effects of European economic integration on the production and trade of manufactured products’, in B.Balassa (ed.), European Economic Integration (Amsterdam: North-Holland; New York: Elsevier), pp. 3–40. United Nations Conference on Trade and Development (1979), Transnational Corporations in World Development (New York: United Nations). UNCTAD (1990), Regional Economic Integration and Transnational Corporations in the 1990s: Europe 1992, North America and Developing Countries (New York: United Nations). UNCTAD (1992), World Investment Report 1992: Transnational Corporations as Engines of Growth (New York: United Nations). UNCTAD (1993), Handbook of International Trade and Development Statistics, supplement (New York: United Nations). United States International Trade Commission (1993), Potential Impact on the US Economy aand Selected Industries of the North American Free Trade Agreement (Washington, D.C.: USITC). Vaitsos, C. (1982), The Role of Transnational Enterprises in Latin American Economic Integration Efforts: Who Integrates with Whom, How, and for Whose Benefit (New York: United Nations).

322

SELECT BIBLIOGRAPHY Vanek, J. (1965), General Equilibrium of International Discrimination: the Case of Customs Unions (Cambridge, MA: Harvard University Press). Vaubel, R. (1978), Strategies for Currency Unification (Tübingen: Mohr). Vaubel, R. (1990), ‘Currency competition and European monetary integration’, Economic Journal, vol. 100, pp. 936–46. Vaubel, R. (1994), ‘The public choice analysis of European integration’, European Journal of Political Economy, vol. 10, pp. 227–49. Venables, A.J. (1993), Equilibrium Locations and Vertically Linked Industries, Discussion Paper no. 802 (London: Centre for Economic Policy Research). Venables, A.J. (1995), ‘Economic integration and the location of firms’, American Economic Review, vol. 85, no. 2, pp. 296–300. Venables, A.J. and Smith, A. (1986), ‘Trade and industrial policy under imperfect competition’, Economic Policy, no. 2, pp. 622–72. Verdoorn, P.J. (1949), ‘Fattori che regolano lo sviluppo della produttivita del lavoro’, L’lndustria; trans. G. and A.P.Thirlwall in Research in Population and Economics (autumn 1978) and circulated privately. Verdoorn, P.J. (1952), Welke Zijn de Achtergronden en Vooruitzichten van de Economische Integratie in Europe en Welke Gevolgen zou deze Integratie Hebben, met Name voor de Welvaart in Nederland? Overdrukken no. 22 (The Hague: Centraal Planbureau). Verdoorn, P.J. (1960), ‘The intra-bloc trade of Benelux’, in E.A.G. Robinson (ed.), The Economic Consequences of the Size of Nations (London: Macmillan). Verdoorn, P.J. and Meyer zu Schlochtern, F.J.M. (1964), ‘Trade creation and trade diversion in the Common Market’, in H.Brugmans (ed.), Intégration européene et réalité économique (Bruges: de Tempel), pp. 95–137. Verdoorn, P.J. and Schwartz, A.N.R. (1972), ‘Two alternative estimates of the effects of EEC and EFTA on the pattern of trade’, European Economic Review, vol. 3, pp. 291–335. Viner, J. (1950), The Customs Union Issue (New York: Carnegie Endowment for International Peace; London: Stevens & Sons). Viner, J. (1965), ‘Letter to W.M.Corden dated 13 March 1965’, Journal of International Economics, vol. 6, pp. 107–8. de Vries, B. (1951), ‘Price elasticities of demand for individual commodities imported into the US’, International Monetary Fund Staff Papers, vol. 1, pp. 397–419. Wallace, H. and Wallace, W. (eds) (1996), Policy Making in the European Union, (Oxford: Oxford University Press). Werner, P. (1970), ‘Report to the Council and the Commission on the realisation by stages of economic and monetary union in the Community’, Bulletin of the European Communities, Supplement, vol. 3 (Werner report). West Indian Commission (1992), Overview of the Report of the Commission (Barbados: the Commission).

323

SELECT BIBLIOGRAPHY Whalley, J. (1979), ‘Uniform domestic tax rates, trade distortions and economic integration’, Journal of Public Economics, vol. 11, pp. 213–21. Whalley, J. (1981), ‘Border adjustments and tax harmonization: comment on Berglas’, Journal of Public Economics, vol. 16, pp. 389–90. Whalley, J. (1993), ‘Regional trade arrangements in North America: CUSTA and NAFTA’, in J.de Melo and A.Panagariya (eds), New Dimensions in Regional Integration (Cambridge: Cambridge University Press), pp. 352–82. Wilczynski, J. (1965), ‘The theory of comparative cost and centrallyplanned economies’, Economic Journal, vol. 75, pp. 63–80. Wiles, P.J.D. (1968), Communist International Economics (Oxford: Blackwell). Williamson, J. (1965), ‘Regional inequality and the process of national development: a description of the patterns’, Economic Development and Cultural Change, vol. 13, pp. 3–45. Williamson, J. and Bottrill, A. (1971), ‘The impact of customs unions on trade in manufactures’, Oxford Economic Papers, vol. 23, pp. 323–51. Winters, L.A. (1984), ‘British imports of manufactures and the Common Market’, Oxford Economic Papers, vol. 36, pp. 103–18. Winters, L.A. (1987), ‘Britain in Europe: a survey of quantitative trade studies’, Journal of Common Market Studies, vol. 25, pp. 315–35. Winters, L.A. (ed.) (1992), Trade Flows and Trade Policy after ‘1992’ (Cambridge: Cambridge University Press). Winters, L.A. and Venables, A. (1991), European Integration: Trade and Industry (Cambridge: Cambridge University Press). Wonnacott, P. and Wonnacott, R. (1967), Free Trade between the United States and Canada: the Potential Economic Effects (Cambridge, MA: Harvard University Press). Wonnacott, P. and Wonnacott, R. (1981), ‘Is unilateral tariff reduction preferable to a customs union? The curious case of the missing foreign tariffs’, American Economic Review, vol. 71, pp. 704–14. Wonnacott, P. and Wonnacott, R. (1984), ‘How general is the case for unilateral tariff reduction?’, American Economic Review, vol. 74, p. 491. Wonnacott, P. and Wonnacott, R. (1992), ‘The customs union issue reopened’, World Economy, vol. 60, no. 2, pp. 119–135. Wooton, I. (1986), ‘Preferential trading agreements: an investigation’, Journal of International Economics, vol. 21, pp. 81–97. Wooton, I. (1988), ‘Towards a common market: factor mobility in a customs union’, Canadian Journal of Economics, vol. 21, pp. 525–38. World Trade Organization (1995), Regionalism and the World Trading System (Geneva: WTO). Yannopoulos, G. (1990), ‘Foreign direct investment and European integration: the evidence from the formative years of the European Community’, Journal of Common Market Studies, vol. 28, pp. 235–59.

324

INDEX

ability to pay principle 132–4 accumulation effects 6, 39, 106, 232 Acocella, N. 115 adaptive expectations 204–5 additionality 248, 250–1, 264 adjustment costs 97;fiscal integration 151–2;monetary integration 192, 209–12, 215, 228, 229, 287; in NAFTA 290; policy 188–9; stabilization 175–6 Africa, regional integration 1, 3, 12, 268, 273, 284, 296 Agriculture Fund (EU) 149 aid, regional 246, 248–51, 265 Aliber, R. 111 Allen, P.R. 192 allocation: effects 9, 18–19, 106; neutrality 164–5, 179–80, 184, 186–7, 189 allocation function 129–30, 148–51, 156 Allsopp, J. et al. 214 Andean Pact 13, 273–4, 278, 280, 293 Aristotelous, K. 119 Armstrong, H.W. 260 Arndt, S.W. 40 Artis, M. 200 ASEAN Free Trade Area (AFTA) 14, 278, 294 ASEAN Industrial Projects Programme (1977) 294 Asia, regional integration 1, 13–14, 268–9, 293–5 Asian-Pacific Economic Cooperation (APEC) 14, 295

assignment of policy functions 126, 128, 130, 132, 135–7; regional policy 246–9 Association of South East Asian Nations (ASEAN) 273, 293–4 Asuncíon Treaty (1991) 285 Australia and New Zealand Closer Economic Relations Agreement (ANZCERTA) 11 backwash 240 Baden Fuller, C. 121 balance of payments 169, 175–6, 188, 190–1, 198, 220 Balassa, B. 48, 101 Balasubramanyam, V. 119 Baldwin, R. 39, 95, 98–9, 161, 194, 265 barriers: non-tariff 14, 63–5, 69, 100, 114–15, 121, 171, 274, 285–7, 294, 296; technical 63–4, 288; trade 2, 61, 84–7, 92–5, 99, 101–3, 109, 120, 235–7, 278–9, 296 Barro, R.J. 193, 259–60 Bayoumi, T. 227 de Beers, J.S. 7, 9 Begg, I. 130, 262, 265 benefit principle and budget 132–3 benefits of integration: allocation 4–7, 18–35, 41–8, 64, 67, 72–3, 81, 194; competition 5, 83–90; European Union 93–101; growth 6, 98–9; vs. unilateral reduction 52–5, 57–9; see also policy credibility; spillovers; trade costs; transaction costs

325

INDEX Berglas, E. 25, 51, 52 beta convergence 259 Boltho, A. 249 border: controls 182, 297; tax adjustments 171–6, 182–5 Botswana, Lesotho, Namibia, Swaziland (BLNS) and SACU 275–6 Bouzas, R. 286 Brander, J.A. 5, 82, 87, 95, 118 Bretton Woods Agreement 217, 219, 227 Brown, D.K. et al. 291, 292 Brussels Treaty (1975) 140–1 Buckley, P.J. 112 Byé, M. 7, 9 Calmfors, L. 212 Canada and NAFTA 287–92 Canada—US Free Trade Agreement (CUFTA) 3, 11, 91, 109–10, 119, 287, 289–91 Cantwell, J. 121 capital: export and import neutrality 179–80, 186; foreign 76–9; market 213–14, 218; mobility 10, 64, 72–80, 92, 104–5, 175, 178–9, 197, 223, 228, 234, 238–9, 247 Carbonaro, G. 260 Caribbean Community (CARICOM) 13, 278 Caribbean Free Trade Area (CARIFTA) 13 Caribbean, regional integration 1, 3, 13, 268, 273 Cartagena Agreement (1969) 13 Casson, M. 112 Caves, R. 112 Cawley, R. 94 Cecchini Report (1988) 98, 108, 122 Central American Common Market (CACM) 13, 273, 278, 293, 297 Chatterji, M. 260 Cheshire, P. 260 clearing house system (VAT) 184–5 Cnossen, S. 171, 172, 181, 186, 187 co-ordination: definition 65; monetary 215–16; policy 2, 65–8, 123, 126, 128, 136–8, 159, 189, 191, 213–15, 218, 222, 225, 281,

285; regional policy 246–7, 265; TNCs 112 Cobham, D. 211, 216, 224 cohesion 64, 67–8, 132, 142–4, 157–8, 208, 246, 251–2, 261 Cohesion Fund (EU) 139, 144, 148–9, 155, 252, 261–4, 266 Collier, P. 25 Colonia Protocol (1994) 285 COMECON 4, 12 Common Agricultural Policy (CAP) 10, 139–45, 148–50, 152, 155, 161, 217, 265 common external tariff (CET): efficient 57–8; ineffective or redundant 22, 33; tariffaveraging 22–3 common market: European Union as 10, 92 Common Market of the Southern Cone (MERCOSUR) 13, 278, 285–7, 293, 297 community budget 4, 68, 123–6, 128, 132–3, 137 community expenditure 130–2, 134–5, 137–8, 163, 198–9; European Union 139–44, 146, 148–50, 152–3, 155–6, 158 comparative advantage 4–5, 47, 49, 78, 82–4, 107, 166, 189, 233, 234, 294; relevance for integration 82–4 compensation 6, 26, 53, 58–9, 144, 148, 156, 164, 174, 233–4, 244, 275, 281–4 competition: jurisdictional 124; policy and TNCs 117; tax-rate 127, 173, 186–7 complementarity agreement 272–3 consumption effect 17, 19, 22, 24, 30–1, 33–5, 54–5, 57 convergence: customs union 64, 67–8, 80–1; European Union 10, 186, 216, 218, 222, 224–5, 229, 250–1, 254, 259–60, 266; fiscal integration 131, 134, 138; in MERCOSUR 286; monetary integration 208–9, 211, 229; regional 232, 234, 237–8, 240, 243, 248; in UEMOA 284 Cooper, C.A. 8, 51, 54–5, 58

326

INDEX Corden, W.M. 25, 42, 44, 46, 60, 191, 200 corporate taxation 131–2, 178–82, 185–9 cost reduction effect 44–5, 47 Council for Mutual Economic Assistance (CMEA) 11 Cross, R. 210 cross-border: integration of TNCs 109, 111, 116–18, 122; investment 177, 185–7, 277, 278; spillover 6, 61–2, 64–5, 125, 127–8, 135–6; taxation 282–3 cumulative causation 237, 239–42, 248 currency: boards 277; parallel 216; single 3, 67, 191, 194–5, 200–1, 210, 215–16, 229; single in EU (euro) 10, 102, 152, 218–19, 223–8, 266 customs duties 102, 145–6, 153 Davenport, M. 94, 100 De Grauwe, P. 224 Dearden, R.G. 289 defensive investment 114–15 deflation 214–15 Delors Report 143–4, 222 Demakas, D.G. et al. 149 destination principle 166–73, 175–6, 182–3 devaluation 199–200, 222, 224 Devarejan, S. 277 developing countries 12–14, 58, 60, 268, 270–7, 277–95 Devereux, M. 187 Dignan, T. 258, 264 discrimination 2, 67, 72, 99, 104, 117, 175, 179, 186, 199 disparities 61, 67, 73, 231–2, 234–5, 239, 243–52; European Union 10, 143, 157, 261, 263–6 distance costs 279 distribution function 131–2 distributional effects 17–18, 73, 110, 231–3 Dixit, A.K. 82, 118 Dixon, R.J. 241, 248 double taxation 106, 177, 182 Driffill, J. 212 Drysdale, P. 295

Dunford, M. 259 Dunning, J.H. 109, 112, 113, 117, 121 dynamic effects 38–9, 47–8, 80, 98–9 East African Community 273 East Europe, integration into EU 268 Eastern Caribbean Common Market (ECCM) 13, 277 Economic and Monetary Union (EMU) 10, 149, 158–60, 196, 217, 221, 223–7, 229, 251–2, 262 economic and monetary union, European Union as 10, 252, 260, 265–6 economic and monetary unions 2–4, 123, 163, 247, 268, 281 economies of scale 5, 17, 20, 37–9, 41–9, 63, 73, 80–1; in European Union 94–6, 98–9; fiscal integration 125–7, 129–30; market integration 82–3, 100; monetary integration 194; in other blocs 271–3, 279, 281, 290, 294–5; regional 233, 235–8, 243; TNCs 114, 121; unilateral reduction and customs unions 52–3, 56–7 Economic Community of West African States (ECOWAS) 12 Eden, L. 111 efficiency: allocational 67–8, 70, 73, 109, 111; assignment of functions 123–31, 138; common market 76, 81; costs 109–10; European Union 104, 156, 160, 187, 263; gains 47–9, 63–5, 106, 111–12, 116–17; polarization 244; redistribution 155; regional policy 246, 249; taxharmonization 164–6, 170–1, 173, 177–80; technical 37–8 efficient transfer price 111 Eichengreen, B. 160, 227 El-Agraa, A. 25 Emerson, M. 194 Emerson Report (1988) 91, 93, 97, 98, 106, 122, 214 employment: European Union 253,

327

INDEX 258, 263;fiscal policy 134, 136–7; monetary integration 192, 198–200, 202, 205–6, 210, 212, 228;in NAFTA 290; regional 231, 239, 241, 245, 247 Equatorial African Union (UDEAC) 273, 275 equilibrium analysis 97; generality of 24–5 equity 133, 141, 151–7, 164, 177, 252, 274 equivalence theorem 168–9 Erzan, R. 292 European Central Bank 223, 227 European Coal and Steel Community (ECSC) 139, 147–8 European Community (EC) 1, 3–5, 38, 61, 64, 82–3, 91, 114, 116, 119–21, 180, 243–7 European Court of Justice 64 European Currency Unit (ECU) 220, 227 European Development Fund (EDF) 139, 147–8 European Economic Community (EEC) 1, 3, 9, 11 European Economic Interest Grouping Regulation 106 European Economic Space 11 European Free Trade Association (EFTA) 3, 10–11, 83 European Investment Bank (EIB) 139, 221, 261 European Monetary Fund (EMF) 221 European Monetary Institute 227 European Monetary System (EMS) 207, 220–5, 228 European Regional Development Fund (ERDF) 140, 147–9, 152, 155, 251–2, 261, 263 European Social Fund (ESF) 140–1, 147–9 European System of Central Banks (ESCB) 223 European Union (EU): budget 132, 138–61; comparison with NAFTA 287–90; as economic and monetary union 3–4; enlargement 268–9, 277–8, 296; federalism 127–8, 130; fiscal

policy 214; impact on third countries 99–100, 107; intercountry and inter-regional disparities 253–60; intra-bloc trade 67; as largest economic bloc 9–10, 12; market integration 91–101; monetary integration 217–29; policy integration 70; regional policy 250–3, 260–7; as single market 101–8; tax harmonization 175, 180–9; TNCs and 109, 118– 22;VAT 171; see also convergence Exchange Rate Mechanism (ERM) 220, 222–6 exchange rates 58, 66, 131–2, 134, 164, 166–9, 171–2, 175–6, 188, 190–4, 197–202, 210, 212–13, 215, 228; European Union 157, 159, 217–20, 222–8, 266 excise taxation 165, 172, 174, 180, 182, 188 expectations 202–5, 210–11, 297 externalities 54, 65, 125, 127, 136, 244 factor: endowment 49, 83–4, 100, 234, 238, 290; influence on growth 248–9; mobility 5, 18, 64, 66, 72–4, 76, 80–2, 131, 160, 163, 177–8, 192, 217, 232, 239, 285 Fagerberg, J. 259 Financial Instrument for Fisheries Guidance (FIFG) 147–8 fiscal federalism 124, 126–32, 160 fiscal frontiers 171–2, 182, 184, 188, 282–3 fiscal harmonization 112, 123, 160, 163–89, 195, 212, 274, 282–3 fiscal integration 4, 102, 123–61, 192 fiscal transfers 58, 68, 153–6, 200, 208–9 Fleming, M. 200, 212 Fontainebleau settlement (1984) 141–2, 144, 152 foreign direct investment 77, 79–80, 107, 113–15, 117, 119–20, 122, 194; regional integration 274, 279–80, 284, 288, 291–3, 296

328

INDEX foreign profit creation and diversion effects 77–9 Forte, F. 125 Fountas, S. 119 Franklin, M. 156 Franko, L.G. 120–1 free (non-discriminatory) trade 7, 48, 51, 53, 59, 111 Free Trade Area of the Americas (FTAA) 13, 292–3 free trade areas 1–4, 10–11, 13–14, 28–30, 30–5, 37–50, 51–3, 60, 69, 72, 163, 268–9, 272–3, 280, 282, 285, 288–90, 292–3, 297 Friedman, M. 199, 202, 203, 204 Garnaut, R. 295 Gatsios, K. 129 GATT 60, 269, 276, 277, 287 Genser, B. et al. 166 Geroski, P. 98 Giavazzi, F. 193, 222 Giovannini, A. 180 globalization and international integration 268–98 Goodhart, C.A.E. 135, 136, 159 Gordon, D. 193 grants 130, 150, 156, 248–51, 261, 264 Greenaway, D. 119 Gros, D. 216 growth: economic 17, 98–9; European Union 97–9, 106, 144, 253–7, 259–60, 263, 266; global 271, 274–6, 279, 282–3, 286, 295; monetary integration 194, 208–9; regional 231, 244–5, 247–9; spatial effects and regional integration 238–43; TNCs and 116–18 Grubel, H. 47, 48, 76 Guidance Section (EAGGF) 147–8, 261 Guillaumont, P. et al. 277 Haaland, J. 89, 100 Haberler, G. 1–2 Harden, I.J. 160 harmonization: definition 65 Helpman, E. 41, 48, 49, 83 Hirschman, A.O. 239, 240, 242

Holtham, G. 249 homogeneity, political 125–6, 129 Horstman, I. 109 Hufbauer, G.C. 100, 116, 292 Hymer, S.H. 116 hysteresis effect 210–11 Imada, P. 294 import substitution 14, 19, 22, 114–15, 271–5, 277, 278 inflation 107, 188, 196, 200–8, 210–11, 219–23, 225–6, 228–9, 241, 245, 277, 292 insurance principle 136, 151 integration: definition 65 interdependence 61, 64, 66–7 Interinstitutional Agreement (EC) 143 internalization 110, 113, 118 International Monetary Fund (IMF) 228 intra-bloc capital flows 74–6 intra-bloc trade, in EU 48, 64, 67, 82, 94–5, 100, 103, 120, 155;in other blocs 12–13, 274, 278–9, 281–2, 285–6, 289–91, 294, 297 intra-industry trade 5, 18, 38, 48–9, 82, 87, 89, 107, 110 investment: creation 114;diversion 114, 269; fiscal harmonization 178–80; global 271, 279–80, 283, 285–6, 288, 290, 292–3, 297–8; intra-bloc 120; market integration 84, 98–9; monetary integration 193–4; regional 240; TNCs 113–17, 121–2 Jenkins, Roy 220 Johnson, H.G. 8, 51, 53–4, 168, 198 Jones, A.J. 25 juste retour principle 132–3 Kaldor, N. 239, 240, 241, 242, 244, 248, 249 Kay, J. 122 Keen, M. 166, 170, 174, 180, 185 Kemp, M. 25, 26, 53 Kenen, P. 192 Kindleberger, C.P. 114 Klein, L.R. 291 Klepper, G. 118

329

INDEX Knickerbocker, F.T. 113, 115 Kojima, K. 47 Krauss, M.B. 60, 168 Krugman, P.R. 5, 48, 49, 82, 87, 95, 118, 224, 235–7, 238, 248, 249 Kyle, A.S. 118 labour: mobility 10, 64, 72, 92, 105–6, 175, 177–8, 208, 234, 238–9, 245, 247 Lahiri, S. 170 Langhammer, R.J. 294 Latin America, regional integration in 1, 13, 268, 273, 277, 280, 284–5, 292 Latin American Free Trade Association (LAFTA) 13, 273 Latin American Integration Association (LAIA) 13 Leibenstein, H. 38 liberalization: financial 279; tariff 284; trade 2, 14, 46–8, 64, 72, 83, 217–18, 272–4, 276–8, 281, 285–6, 288–91, 295 Lipsey, R.G. 51 List, F. 8 Lloyd, P.J. 25 location effects 6, 106, 112–14, 127, 170–1, 178–9, 186, 188, 233–8, 242, 244, 272–3, 280 Lomé Convention 139, 276 Lucas, R.E. 242, 243 Maastricht Treaty (1992) 3–4, 10, 128, 144, 149, 160, 181, 222–6, 229, 262 McCulloch, J.R. 8 MacDonald, R. 197 MacDougall Committee 158 MacDougall, G.D.A. 76, 79, 192 Machlup, F. 9 McKinnon, R.I. 192, 198 Macmillan, Harold 38 Managua Treaty (1960) 13 Marjolin, Robert 250 market: access 104, 235–6; enlargement effects 37–9; failure, market 65–6, 109, 126; predictability 296; share 85, 87, 95 Markusen, J. 109

Marrese, M. 4 Marshall, Alfred 244 Massell, B.F. 8, 51, 54–5, 58 Masson, P.R. 213 Meade, J.E. 9, 25, 47, 51, 60, 73, 168 de Melo, J. 25, 51, 269, 275, 276, 277 mergers 105, 107, 115, 118, 122, 178 Mexico, and NAFTA 287–93, 296 migration 105, 131, 134–5, 208, 232 Miller, M.H. 200 Molle, W. 120, 194 monetary compensatory amounts (MCAs) 217 monetary integration/union 2–3, 6, 70, 131–2, 191–3; in Africa 275–7, 281–4; benefits and costs 193– 207; community fiscal policy 138, 151, 157–60, 208–9; and employment 198–207; in EU 223–8; fiscal harmonization 164, 169, 175–6, 188–9; national fiscal policy 212–15 Monti Report (1996) 92, 101–2, 103 Morsink, R. 120, 194 multilateralism 269, 287 Mundell, R.A. 39, 192, 212 Musgrave, P. 180 Musgrave, R.A. 125, 170, 179 mutual recognition principle 64, 102–4 Myrdal, G. 239, 240, 242 natural rate of unemployment 134–5, 202–6, 211 Neary, P. 77 negative integration 2 Neumark Report (1963) 176, 181–2, 185 Neven, D. 100, 118 new classical macroeconomics 197, 204, 210 new economic geography 235–9, 249 new growth theory 239, 242–3, 244, 248–9 new regionalism 14, 269, 277–80, 297 non-accelerating inflation rate of

330

INDEX unemployment (NAIRU) 160, 203–4, 210–11 Norman, V. 82 North America, regional integration 268 North American Free Trade Agreement (NAFTA) 3, 13, 269, 278, 280, 287–93, 296 Oates, W. 125, 126 O’Brien, D. 8 offensive investment 114–15 oligopoly 46, 48–9, 84, 90, 111, 113, 117, 174, 237, 242 Oman, C. 279 open regionalism 14, 192, 295 optimality 6, 191–3 origin: principle 166–73, 175–6, 179, 182, 184–5, 187; rules of 28–30, 102, 286, 288–9, 293 own resources system 140–1, 145, 148 Padoa-Schioppa Report (1987) 156, 158 Padoa-Schioppa, T. 223 Pagano, M. 193 Page, S. 100 Panagaryia, A. 269, 275, 276 Panic, M. 116 Pearson, M. 187 Pelkmans, J. 38, 98 personal taxation 130–2, 158, 180–1 Phelps, E. 202 Phillips, A.W. 200 polarization 80, 240, 244–6, 275, 282, 297 policy: autonomy 5, 73, 140, 177, 184, 190–1, 198, 213; constraints 4–5; credibility 84, 128, 171, 192–3, 196, 206–7, 214, 280–1, 283–4, 288, 294, 296; discipline 213–14, 224, 277, 296; interdependence 61, 64, 66–7 positive integration 2 power, market 79, 82, 85, 87–9, 110, 115 price: —cost margins 85, 99, 107; discrimination 87–92 pro-competitive effect 84, 90 product: differentiation 18, 38, 48–9,

82–5, 91, 100, 235, 237; mobility 102–3, 163 production effect 17, 19, 22–4, 31, 33–41, 44–9;multiplant 111–12; tax harmonization 165, 168–71; theory of international 113–14 protectionism 14, 24, 27, 30, 38, 45, 54–5, 58–60, 64, 69, 102, 175, 269–71, 273–4, 276, 281, 288 pseudo-(monetary) unions 191, 199–200, 210, 228 public goods 54–62, 65, 68, 125–7, 129, 163 public procurement 103 Pugno, M. 249 purchasing power standard (PPS) 154, 157, 254, 256 rational expectations 204–6, 210–11 rationalization 82, 116, 178, 271–2, 295 rationalization investment 114–15, 120–2 reciprocal dumping 87, 95 redistribution 58, 77, 80–1, 100, 130–4, 136, 139, 150–1, 153–6, 161 regional integration: and competition 82–4, 100– 1;dynamics of 68–71; and fiscal federalism 127–9;and globalization 279; and transnational corporations 109–22 regional policy: assignment 245–7; European Union 139, 148–50, 157, 159–60, 188–9, 250–67; rationale for 243–5 regulatory functions 129–30 Reichenbach Group 156, 264 reorganization investment 114–15, 120, 286 residence principle 177, 179–80, 186 resource transfers 247, 249–50, 260; see also redistribution restricted origin system 168, 176, 182, 188 revenue: community 127, 133, 137–8; European Union 99, 139–41, 143– 6, 153, 155–6, 158

331

INDEX Ricardo, D. 8 Richardson, M. 26 Robson, P. 109, 112, 275 Romer, P. 99, 242, 243 Ruding Report (1992) 181, 185–6 Rueff, Jacques 217 Rugman, A. 116 Sala-i-Martin, X. 259–60 sales taxation 180–1, 185 Salvatore, D. 291 Sanna, G. 249 Santos, P. 157 Sapir, A. 107 Schott, J.J. 292 Scitovsky, T. 193 Seabright, P. 129 segmentation, market 5, 64, 83–4, 87–90, 95–6, 111, 122, 195, 279, 287, 296–7 separation theory 112 services, free movement 92, 104, 107, 121, 285, 288 Shibata, H. 168 Shoup, C.S. 9, 169, 171, 172, 181 sigma convergence 259–60 Single European Act (1986) 92, 127, 141–4, 252 Single European Market (1993) 3, 11 single market: EC 61, 91; EU 64, 92–9, 101–8, 143, 151, 157, 171, 181–2, 226, 251–2, 260; SACU 276; UEMOA 281–4; Siotos, G. 118 Smith, A. 49, 90, 95–6, 97, 115, 118 Smith, Adam 8 Smith, S. 135, 136, 159, 166, 185 Snake arrangement 219–20 Social Fund (EU) 261 social security 105, 180–1 Solow, R.M. 238 source principle 177 South Africa and SACU 275–6, 280 Southern African Customs Union (SACU) 4, 12, 58, 275, 280–1, 283 Spaak Report (1956) 72 spatial effects of integration 69, 157–8, 232–5, 238–44, 265 Spaventa, L. 222 specialization 4–5, 17–18, 46–9, 56–7,

61, 82–4, 106–7, 114, 120–1, 160, 164, 166, 242, 272–5, 282, 294–5 Spencer, B.J. 118 spillover 6, 61–2, 64–7, 125–31, 135–6, 150, 173, 212–15, 217, 224, 244, 247, 265 spread effect 240, 242 Stability Pact (1996) 160, 225–6 stabilization 66–7, 70, 116, 125, 127, 132, 150–2, 157–60, 163, 175–6, 181, 189, 192, 198, 210, 212–15, 227, 229, 245, 266 stabilization function 134–9 standards, industrial 63–4, 100, 102–3, 129, 288 Stern, R.M. 291 Stopford, J. 121 strategic intervention theory 118 strategies, monetary integration 215–16 structural funds: European Union 139, 143–4, 146, 148, 150, 153, 161, 252–6, 261–3; UEMOA 282–4 subsidiarity principle 79, 128, 181, 265 subsidies 59–60, 112, 173, 199, 249–50, 263 Swan, T.W. 198 tariffs: TNCs 111, 120; unilateral vs. discriminatory reduction 51–62 taxation: common market 5, 79, 81; competition 66; European Union 105, 139–40, 145, 158; fiscal integration 127–8, 129, 131–3, 135, 138; harmonization 70, 164–89; TNCs and 111–12 Taylor, M.P. 197, 213 van den Tempel Report (1969) 181, 182 terms of trade 5, 17, 20, 25, 37, 39–42, 51, 100 Thirlwall, A.P. 241, 248, 249 Thygesen, N. 216 time consistency 193, 196, 206 Tinbergen, J. 2, 6, 9, 198 Tinbergen Report (1953) 166, 181 Tindemans Report (1976) 219 Tironi, E. 77, 78 trade: costs 85, 87, 89, 94, 100, 236–7; deflection 28, 30, 34, 40, 168, 288; developing countries

332

INDEX 270–1;fiscal harmonization 164–5, 168–73, 175, 178; intensity 120; measurement of effects 48, 107; monetary integration 194; suppression 45–7; theory 83, 234–5; see also barriers, trade; intra-bloc trade; intra-industry trade; liberalization, trade trade creation and diversion: classical economics 8– 9;common markets 74, 77–8; customs unions 19–20, 22–8; customs unions and free trade areas 31, 33–5, 38, 40, 44–9; European Union 107, 155; free trade areas 28, 30–1; in ‘new’ economies 82, 86, 90, 95–6, 99–100, 101, 107; in other blocs 269–71, 274–6, 278, 280, 289, 294–5; regional 233; TNCs 114–15, 122; unilateral reduction and customs unions 52–3, 59 transaction costs 61–2, 70, 82, 100, 109–10, 112, 117, 157, 171–3, 195–6, 266, 278–9, 284 transnational corporations (TNCs) 79, 81, 84, 90–1, 109–22, 279, 281, 284, 293 transport costs 18, 34, 47, 53, 233, 235–6, 258, 272 Treaty of Montevideo (1980) 13 Treaty of Rome (1956) 1, 9, 72, 92, 104–5, 139–41, 149, 181, 217, 250 trickle-down effect 240 Truman, E.M. 107 UK and EU 142, 144, 152–3, 226–7 unemployment: European Union 159, 225, 254–9; fiscal integration 134–5;monetary integration 200–6, 209–11; regional 234, 245, 292 Union Economique et Monétaire Ouest-Africaine (UEMOA) 12, 278, 282–4, 296 Union Monétaire Ouest-Africaine (UMOA) 281 US: and integration 1, 277, 280; and NAFTA 287–91 value-added tax (VAT) 102, 141–6,

150, 152–3, 156, 166–7, 169, 171–2, 174–6, 276, 283;European Union 181–5, 188 Vanek, J. 25, 26 Vanhaverbeke, W. 227 Vanous, J. 4 Vaubel, R. 70, 128, 192, 215 Venables, A.J. 49, 95–6, 235–7 Verbeke, A. 116 Verdoorn, P.J. 48, 241 Verspagen, B. 259 Viner, J. 7, 8, 9, 27–8, 40, 41, 45, 51 Visegrad Group 161 Von Hagen, J. 160 wages: European Union 105, 266; fiscal integration 131, 134–5; monetary integration 196, 199–201, 203, 208–9, 212, 217–8; regional 234, 237, 240–2 Wan, H. 26, 53 welfare effects of integration: common market 72–3, 78–9; customs unions 17–18, 20, 22–4, 26, 68, 70; customs unions and free trade areas 39–40; European Union 94–6, 99, 149; fiscal harmonization 164–5, 173–4; fiscal integration 126, 135; global 269; market integration 82, 84, 86–7, 89–90; monetary integration 201, 206; NAFTA 289– 90, 295; regional 233;TNCs 111– 12; UEMOA 284; vs. unilateral tariff reduction 53–5, 62 Werner Report (1970) 218–19 West African Economic Community (CEAO) 12, 273, 275, 281–3 Williams, S. 211 Winters, L.A. 83, 98 Wonnacott, P. and R. 51, 52, 53 Wooton, I. 25, 40, 74, 75, 89, 112 World Bank 51 World Trade Organization (WTO) 60–1, 268–9 Yannopoulous, G. 114, 119 Yeats, A. 292 Zollverein 3, 8

333