Foreign Trade Reforms and Development Strategy

Foreign trade reforms and development strategy

In recent years developing countries have come under increasing international pressure to liberalize their trade regimes. In particular, many aid and adjustment programmes recommend that import controls should be relaxed. This advice is normally based on theoretical considerations of ‘optimal’ specialization and insertion in international trade or on empirical studies demonstrating the counter-productive effects of policies aimed at protecting domestic industry or promoting import-substitution. Foreign Trade Reforms and Development Strategy argues that trade liberalization is in fact inappropriate for many developing countries. The theoretical framework on which it is predicated takes no account of the instability of the international financial problems. Nor is the empirical evidence for liberalization all that robust. In other words, we know very little about the consequences of one of the major reforms advocated by Structural Adjustment Programmes. The object of this book is to clarify the objectives, constraints, and dangers of foreign trade reforms, and to contribute to the articulation of a positive case for flexible, targeted and reversible import controls. Its aim is to reintroduce long-term strategic considerations into a debate dominated by the management of shortterm constraints. Both theoretical and empirical issues are addressed. Questions such as the place of foreign trade reforms in SAPs, the impact of trade liberalization upon industrial efficiency, the evolution of protectionist tendencies in the industrialized countries and conflicts between policy objectives are examined, and bases for alternative policies are outlined. Analytical case studies illustrate the main theoretical findings. Jean-Marc Fontaine studied at the universities of Montpellier, Paris (Panthéon-Sorbonne) and Oxford. He has worked in a government research department in Quebec as well as teaching in the Universities of Lille and Amiens. Currently he is attached to the Institut d’Etude du Développement Economique et Social (IEDES) at the Université Panthéon-Sorbonne. Research activities include participation in CNRS (Centre National de la Recherche Scientifique) research programmes and consultancy work for the French Ministère de la Cooperation,UNCTAD and the OECD.

Foreign trade reforms and development strategy Edited by Jean-Marc Fontaine

London and New York

First published 1992 by Routledge 11 New Fetter Lane, London EC4P 4EE This edition published in the Taylor & Francis e-Library, 2005. “To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.” Simultaneously published in the USA and Canada by Routledge a division of Routledge, Chapman and Hall, Inc. 29 West 35th Street, New York, NY 10001 © 1992 Jean-Marc Fontaine 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 or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data Foreign trade reforms and development strategy. I.Fontaine, Jean-Marc II. [Controle des importations et strategies de developpement economique. English] 382.17 ISBN 0-203-97923-0 Master e-book ISBN

ISBN 0-415-07294-8 (Print Edition) Library of Congress Cataloging in Publication Data Contrôle des importations et strategies de développement économique. English Foreign trade reforms and development strategy/edited by Jean-Marc Fontaine. p. cm. Translation of: Contrôle des importations et strategies de développement économique. Includes bibliographical references and index. ISBN 0-415-07294-8 (Print Edition) 1. Developing countries—Commercial policy. 2. Free trade —Developing countries. 3. Economic stabilization—Developing countries. I. Fontaine, Jean-Marc, 1942- II. Title. HF1413.C61513 1992 382′.3′091724–dc20 91–28661 CIP

Contents

List of Figures

vi

List of Tables

vii

Contributors

ix

Acknowledgements

xi

Introduction Jean-Marc Fontaine

1

1

Structural adjustment: a general overview, 1980–9 Paul Mosley

18

2

An overview of trade policy reform, with implications for Sub-Saharan Africa John Nash

33

3

Import controls and the sequencing of trade policy reform, with special reference to Africa David Evans

57

4

The effect of trade liberalization on industrial-sector productivity performance in developing countries Colin Kirkpatrick and Jassodra Maharaj

66

5

Trade reform and growth resumption in Latin America José María Fanelli, Roberto Frenkel and Guillermo Rozenwurcel

80

6

Protectionist pressures in the 1990s and the coherence of North-South trade policies Juan A.de Castro

119

7

Inter-African integration and protection policies: unavoidable failure or missed opportunities? Jean Coussy

148

8

Import liberalization and the new industrial policy in Senegal Anne-Marie Geourjon

162

9

Impact of price-based and quantity-based import control measures in Nigeria T.Ademola Oyejide

174

Import liberalization in Kenya Jean-Marc Fontaine

182

10

v

11

Industrialization strategies, foreign trade regimes and structural change in Turkey, 1980–8 Nurhan Yentürk-Coban

199

12

Exchange rates and subsidies in an export promotion policy: the case of South Korea Mario Lanzarotti

212

Index

219

Figures

1.1 1.2 2.1 2.2 4.1 6.1

‘Structural adjustment’ in relation to other elements of development policy Relation between slippage and growth of GDP, with possible explanation of ‘outliers’ Real exchange rate indices, 1978–88 GDP and export growth for developing countries, 1965–88 Productive efficiency of the firm Non-tariff protectionist response surface against imports of manufactures from developing countries, as a factor of trade and non-trade related factors in 1984 6.2 Projected protectionist pressures in manufacturing sectors of developed market-economy countries as a result of the expansion of trade with developing countries 10.1 Kenya: imports of processed consumer goods, industrial inputs and capital goods, 1970–88 10.2 Kenya: year-to-year rate of growth of manufacturing output, 1969–87 10.3 Kenya: index of manufacturing activity and volume index of imports of inputs, 1971–88 10.4 Kenya: imported input-substitution and liberalization periods 10.5 Kenya: shares of exports to regional markets and of manufactured exports in total exports 12.1 Korea: nominal rate of exchange 12.2 Korea: real exports and real effective rate of exchange 12.3 Korea: real subsidies and real effective rate of exchange, 1963–75 12.4 Korea: real subsidies and real effective rate of exchange, 1962–78

19 29 39 40 71 125 127 185 186 188 189 190 213 213 216 217

Tables

1.1 Adjustment indicators for less-developed countries as a group, 1980-8 1.2 Types of policy measure requested and implemented by the World Bank in return for programme finance, 1980-8 1.3 Comparison of economic performance in adjustment lending and non-adjustment lending LDCs, 1982-6 1.4 Results of regression analysis: all SAL countries 1.5 Effectiveness of structural adjustment: summary of results 1.6 Case-study countries performing better and worse than predicted by policy change/growth regression line 2.1 Average tariffs and para-tariffs by geographical regions 2.2 Frequency of NTMs by geographical regions: percentage of tariff positions affected 2.3 Major trade policy reform proposals among countries receiving World Bank trade adjustment loans, 1979-87 2.4 Growth in manufacturing output, exports, and capacity utilization before and during reform period 2.5 Macroeconomic indicators before and after reform in trade adjustment loan countries, with implementation data 2.6 Exports of non-oil developing countries as a share of exports from developing countries to industrial countries, with selected country examples 4.1 Content of World Bank structural adjustment lending operations, 1980-7 4.2 Main components of trade policy in forty adjustment loan countries 4.3 Implementation of conditionality 5.1 Export real rates of exchange 5.2 Foreign debt as a proportion of GDP 5.3 External sector 5.4 Foreign debt as a proportion of exports of goods and services 5.5 Public foreign debt as a proportion of total debt 5.6 Colombia: public sector accounts 5.7 Public investment as a proportion of GDP 5.8 Argentina: public sector. Savings-investment accounts 5.9 Brazil: public sector deficit 5.10 Brazil: financial transfers from government to private sector, 1970–86 5.11 Brazil: taxes as a proportion of GDP 5.12 Mexico: public sector accounts 5.13 Gross Domestic Product and per capita GDP 5.14 Rate of growth of exports 5.15 Gross fixed investment as proportion of GDP at constant prices

19 22 25 27 28 30 35 37 38 42 44 48 66 68 69 88 89 91 93 94 96 98 99 100 101 102 102 110 111 112

viii

5.16 5.17 5.18 5.19 5.20 5.21 5.22 5.23 6.1 6.2

Savings, investment and resource transfers Chile: public sector deficit Chile: revenues and expenditures of central government Brazil: public sector expenditures in wages and investment Rates of inflation. Consumer prices Capital flight Argentina: monetary assets Brazil: monetary and financial assets Developing countries’ import penetration ratios of total manufactures in DMEC markets Changes in the structure of the value increment of DMECs’ imports of manufactures from developing countries 6.3 Concentration of DMECs’ imports of manufactures from developing countries classified by income levels, in sectors suffering significant ntm’s forms of trade restriction or under free trade, and prospects for protection in the 1990s 6.4 Concentration of DMECs’ imports of manufactures from developing countries classified by regions, in sectors suffering significant ntm’s forms of trade restriction or under free trade, and prospects for protection in the 1990s 6A.1 United States: projected protectionist pressures in manufacturing sectors as a result of the expansion of trade with developing countries 6A.2 European Economic Community: projected protectionist pressures in manufacturing sectors as a result of the expansion of trade with developing countries 8.1 Rate of effective protection by sector 9.1 Growth of Nigerian imports, 1960–88 9.2 Structure of Nigerian imports, 1960–88 9.3 An import market share model 9.4 Sign expectation 9.5 Estimated effects of import control 11.1 Composition of imports and proportion of imports to GNP 11.2 Ratio of exports to imports and GNP, and trade deficit 11.3 Immigrant workers’ remittances: value and percentage of trade deficit 11.4 Protective tariffs 11.5 Evolution of the exchange rate 11.6 Evolution of exports and share of exported manufactured goods in total exports 11.7 The ratio of total exports, manufactured exports and manufacturing value-added to GNP 11.8 Value of exports, imported capital goods and intermediate goods, and export/import ratio 11.9 Capital goods and intermediate imports as a proportion of total imports 12.1 Evolution of subsidies per export dollar 12.2 Estimate of exports’ elasticities

112 113 113 113 114 114 115 115 122 123 131 133 136 142 164 175 175 178 179 180 201 201 202 204 204 205 205 206 207 214 215

Contributors

JUAN A.DE CASTRO is an economist of the United Nations Conference on Trade and Development (UNCTAD) in Geneva. He previously worked for the Banco Hispano Americano, Economics Research and Strategic Planning Division (Madrid) as Deputy Director of the Economics Research Department. He has written and lectured widely on international trade and development and on inter-national finance issues. Among his publications is The Euromoney Guide to Capital Markets 1991 (The Spanish peseta), and various papers on economic journals such as El Trimestre Economico in Mexico. The Department of Economics of CEDES is a research and graduate teaching group devoted mainly to macroeconomic analysis of Argentine and Latin American economies. As a group and individually the authors have published books and articles in that field and acted as advisors of L.A.Governments and international institutions like UNCTAD, ECLA, SELA and IDB. JEAN COUSSY is a Senior Researcher at the Ecole des Hautes Etudes en Sciences Sociales, in Paris, and co-editor of the series ‘Relations Economiques Internationales’, published by the Review Economies et Société. His latest publication (in collaboration with PHILIPPE HUGON) is Integration africaine et ajustement structurel, Ministère de la Cooperation, Paris. DAVID EVANS is a Senior Fellow at the Institute of Development Studies, University of Sussex. He has done research work for various institutions, including UNCTAD, the OECD Development Centre and the Centre for Economic Policy Research. He contributed a chapter to the Handbook of Development Economics edited by H.B. Chenery and T.N.Srinivasan in 1989, and is the author of Comparative Advantage and Growth, published at Harvester-Wheatsheaf in 1989. JOSE MARIA FANELLI is now Senior Researcher at CEDES (Buenos Aires) and Professor of the University of Buenos Aires. JEAN-MARC FONTAINE is a Senior Lecturer in Economics and Director of the Centre de Recherche Economique at the IEDES (Institut d’Etude de Développement Economique et Social), PanthéonSorbonne University, Paris I. He has published in a number of reviews and journals in France, Britain and Germany. He has recently edited Réformes du Commerce Extérieur et Politiques de Développement at the Presses Universitaires de France, Paris. ROBERTO FRENKEL is now Director of the Department of Economics of CEDES and Professor of the University of Buenos Aires. ANNE-MARIE GEOURJON is Senior Lecturer and a researcher at the Centre d’Etudes et de Recherche sur le Développement Inter-national (CERDI), Clermont-Ferrand, France. She has done research work for UNDP, The World Bank, the French Centre National de la Recherche Scientifique (CNRS) and participated in many publications, both in French and English. Her latest contributions are in L’Ajustement Economique du Niger (P. & B.Guillaumont, eds.), L’harmattan, Paris.

x

COLIN KIRKPATRICK is Professor of Development Economics at the Development and Project Planning Centre, University of Bradford, UK. His research interests are in the area of trade and industrialization policy in developing countries. He is the author (with P.COOK) of Macroeconomics for Developing Countries, Harvester-Wheatsheaf, 1990. MARIO LANZAROTTI is a Senior Researcher at the Centre de Recherche Economique de l’IEDES (Institut d’Etude du Dével-oppement Economique et Social), Panthéon-Sorbonne University, Paris I, and presently an economic advisor to the Ministry of Finance in Santiago, Chile. He has published articles in a number of reviews in France and Latin America. His latest publication is La Corée du Sud, une sortie du sous-développement, Presses Universitaires de France, 1992. JASSODRA MAHARAJ is a graduate of the University of Birmingham and Research Assistant at the Development and Project Planning Centre, University of Bradford, UK. Her research interests are in the area of macroeconomic and trade policy in developing countries. PAUL MOSLEY is Professor of Development Policy and Director, IDPM. He has acted as economic advisor to the governments of UK and Kenya, and consultant to World Bank, FAO and UNICEF. He is the author of Aid and Power (2 vols, Routledge, 1991) and other books and articles on policy-making and development. JOHN NASH is a senior economist in the Trade Policy Division of the World Bank. His publications include contributions to several volumes and articles in journals, including the American Economic Review, the Journal of Economic Development and Cultural Change, Economic Inquiry, and The International Journal of Law and Economics. He is also a co-author of the book Best Practices in Trade Policy Reform. He holds a Ph.D. from the University of Chicago. T.ADEMOLA OYEJIDE is a Professor of Economics at the University of Ibadan, Nigeria, and Chairman of the Programme Committee of the Conference on African Economic Issues at the African Economic Research Consortium. He has published a number of articles in International Development Reviews and is a co-author of the latest African Development Report, published by the African Development Bank. GUILLERMO ROZENWURCEL is Senior Researcher at CEDES and Professor of the University of Buenos Aires. NURHAN YENTÜRK-COBAN is Associate Professor at Istanbul Technical Unit and Associate Researcher at the Centre de Recherche Economique de l’IEDES, Panthéon-Sorbonne University, Paris. She is the co-author of Technological and Structural Change in Turkish Clothing Industry (1988) and Technological Change and Industrial Policy: Three Cases from Automobile, Steel and Clothing Industries Using Industrial Electronics in Turkey (forthcoming).

Acknowledgements

First, thanks must go to Jean Masini for help in establishing contacts with scholars and researchers who participated in this book, and to Nadine Mettler for coordinating meetings where this material was discussed. Thanks also to Marianne Georgopoulos, who translated texts by Jean Coussy, Anne-Marie Geourjon, Mario Lanzarotti and Nurhan Yentürk from French into English. Grateful acknowledgements are also due to the General Direction for Development (DG.VIII) of the European Commission in Brussels, which financed the necessary translations and made it possible for the International Seminar, where the idea for this book originated, to be held in Paris, in January 1990.

Introduction Jean-Marc Fontaine

If dissemination is the mark of success, then liberal notions have won a victory in the battle of ideas. This victory seems so complete that one sometimes feels as if only one paradigm in development economics is left: the liberal version of neo-classical analysis (Fischer 1990). However, this victory is somewhat ambiguous. We are not sure whether it was won in a battle fought from beginning to end, or granted by default, following the withdrawal of one competitor. After the 1970s, when primary commodity booms succeeded one another, when growth fed on debt, and new manufactured goods exporters emerged, the 1980s opened under depressing omens. The world economy slowed down, financial problems rose to insoluble proportions, and economic instability settled in. This gloomy atmosphere witnessed the breakdown of development models premised on the dynamic exploitation of sectoral disequilibria, investment fetishism, and direct state intervention (Singer 1989), either because new economic conditions exterminated them, or simply because they revealed their intrinsic inadequacy. Strategies based on state intervention were discredited and replaced in the realm of accepted ideas by the notion that production should be extricated from the underbrush of regulations and interventions that choked down their latent economic potential. BACKGROUND TO THE DEBATE: THE STRENGTH OF THE LIBERAL PLEA The first and obvious explanation for the liberal victory is default. Acceptance and implementation of IMF and World Bank-designed programmes more often followed constraint than conviction and carried less enthusiasm than disenchantment.1 From this point of view, the agreement concluded between the IMF and Tanzania in October 1985 marked a turning-point. On that day, the death certificate of the Arusha Declaration was signed and the burial of ‘African Socialism’ celebrated (Fontaine 1988; Biermann and Campbell 1990). However, defeat of a model—and in the case in point, of a lighthouse model—does not in itself justify the doctrine that emerges from its ruins. Development models of the sixties largely collapsed under their own weight, and the liberal option filled in a vacuum. Still, the liberal option has many things to recommend it. First is obvious commonsense. The outdating of previous models begot such aberrations that any solution trimming down the costly relationships established between the public sector and overprotected, over-importing and over-privileged industries and public enterprises was welcome. In that respect, liberalization proposals raise an urgent alarm—perhaps with some overstatement and rhetorical emphasis.2 They also rely on a tradition going back more than a hundred years, which sees efficiency as resulting from simple rules, such as allocating resources in line with relative-price movements, rather than from behest, based on projects or dreams of the state.

2

INTRODUCTION

As an instrument of critique, liberal analyses have great potential. They stress blatant inconsistencies, for instance in price-systems, and the adverse effects of monopoly positions acquired in the process of industrialization. But, again, does this potential for critique mature into positive recommendation? This, of course, is the crucial question, and it receives no direct or simple answer. A few indeterminate points First, if the superiority of liberal solutions is a clear case in the classroom, empirical verification is quite another matter. Recent evaluations of structural adjustment programmes stumble over a few stones. To start with, results are ambiguous. At best, they do not contradict the optimistic view that adjustment improves performances. But often, blurred by methodological confusions, they do not indicate much,3 or they turn out to be frankly negative (Taylor 1988). Furthermore, when we can presume success, causality is so difficult to establish4 that we remain unsure whether it follows reforms, or some other external factor such as increased funding, for instance. The other source of ambiguity is of a theoretical nature and arises from a confusion regarding the proper content of neo-classical teachings. Although often presented as an integrated block, they really have two facets. One, which is the stone of the fruit, is an abstract theoretical construct of the optimal (or efficient) allocation of resources. The other is a practical recommendation in favour of resource allocation through the market. What the theoretical nucleus does establish is the superiority of one allocation scheme over others.5 But what it does not establish is the essential superiority of one institutional arrangement (the market) over all others.6 The market is a good arrangement only in so far as it approximates to the abstract allocation of resources singled out by theory.7 But the approximation holds under certain conditions, such as the absence of externalities, which are subject to a host of exceptions: decreasing cost industries, indivisibilities, infant industries, non-monetary exchanges and so on. The market is also an institution (Buchanan and Tullock 1967) that may not exist everywhere, in which case it has to be created, which is an important task (Polanyi 1947) and requires a strong involvement of the state, increasing with the degree of backwardness (Gerschenkron 1966). Yet, exceptions tend to be more frequent than rules in many developing countries where non-monetary transactions, non-market links and externalities abound, where industries are still in search of adequate markets, and large parts of populations survive on subsistence and informal activities. The fact that markets in developing countries do not, by a long shot, approximate to the efficient resource allocating mechanism postulated by theory does not, a priori, support any particular conclusion. We could argue that liberalization is pointless, since markets, in so far as they exist, are imperfect: institutions, habits, exchange networks oppose ‘spontaneous’ optimization, and corrective intervention is what is needed. Alternatively, we could argue (see World Bank 1987) that market imperfections are precisely the reason why the economy should be reformed along liberal lines. By removing imperfections, we would lift obstacles to spontaneous optimization, and the economy would move closer to efficiency. If markets were amendable, this last argument would carry much weight, although we would first have to know how they can be amended, in what time-span, and in which order various imperfections should be removed. None of these questions receives a general answer. Identifying the source of the market distortion can prove a delicate matter, and we might easily confuse the visible imperfection with the cause of the distortion. Many situations exhibit the natural monopoly paradox: a monopoly might be the source of the

INTRODUCTION

3

distortion (if borne by force, exclusion, or legal privilege) or its result (if the optimal scale of production does not allow for more than one firm in a market). In this case, suppressing the legal privilege (i.e the apparent imperfection) through privatization will maintain the monopoly (i.e. the distortion). Generally, the answer to the question ‘is the economy distorted because the state intervenes, or does the state intervene because the economy is distorted?’ is a difficult one. There is a great deal of naivety in thinking that removal of one imperfection will end the distortion. Removal might also call for another imperfection (as the account of Senegal’s trade reform by Anne-Marie Geourjon illustrates—see Chapter 8). One point, then, is that market imperfections might be extremely resilient, with correction proving a long-term task, which implies intervention, not liberalization. Another argument is that imperfections are many and pervasive. Even if markets are amendable, piecemeal reform might not be suitable, possible or optimal. We fall back here on the classical LipseyLancaster (1956) ‘second-best’ argument, recently reexpounded in a development perspective by Toye (1987:71–9): when many imperfections coexist, there is no presumption that removal of one of these will result in overall improvement. The result of any one reform is then indeterminate from the start, and sequencing of reforms will prove extremely hazardous, unless you are prepared to maintain the principle of corrective intervention. The last point of indetermination arises from the obsession with allocative efficiency upon which neoclassics thrive. The search for optimal combinations requires stability, not necessarily a static environment, but at least a predictable one. But world instability has been aggravated, world trade has slowed down, and domestic management has become more complex since the mid-1970s. World instability has been aggravated, hitting particularly primary products whose prices react more strongly to variations of activity in industrialized countries. This means that if production and specialization followed actual prices, they would be subjected to contradictory moves from year to year. The implication is that prices do not operate as efficient incentives, but on the contrary that coherent incentives have to be constructed. Even the simplest construction (replacing actual price by trend or moving average) requires public intervention to smooth out unruly variations and present producers with alternatives that allow specialization to proceed beyond short-term fluctuations (Guillaumont and Guillaumont 1990). World instability thus prevents the use of actual prices and requires intermediation by the state, which of course is inconsistent with liberalization. World trade has slowed down. The volume rate of growth of world imports fell from roughly 8 per cent per annum between 1965 and 1974 to 3.5 per cent p.a. between 1976 and 1987.8 This implies two things. First, redefinition of international specializations appears more difficult as competition between new producers hardens. Conditions that allowed the Asian Dragons to emerge no longer prevail (Cline, 1982). We might then wonder whether respecialization is within the reach of private enterprises, or whether states ought to intervene, following the lines of ‘New Trade Theory’ (Krugman 1987), to replace inherited economic structures by constructed comparative advantages. Second, the stimuli to which economies are supposed to react might no longer be there. While economies do seem to react to foreign trade liberalization in periods of sustained growth, amendments to trade regimes do not exert significant influences in periods of international dullness (Singer and Gray 1988). Finally, domestic management has become more complex, torn between the requirements of stabilization (which calls for demand compression) and those of adjustment (which implies supply-side action). Both exercises are contradictory (Krueger 1981; Sachs 1987; Fanelli, Frenkel and Rozenwurcel in Chapter 5 of this book). Reforms intended to promote liberalization generate tensions that aggravate domestic disequilibria or contradict stabilization objectives. This makes them dangerous or impossible. World instability, of course, aggravates matters, with exports receipts varying widely from year to year, and domestic

4

INTRODUCTION

policy priorities shifting anarchically from domestic absorption control to current account emergency management (Dick et al. 1983). FOREIGN TRADE-REFORM IN THE LIBERAL APPARATUS When moving into the area of foreign-trade reforms, the discussion gains clarity without losing its generality. The discussion is clarified because the number of interrelations is reduced, or at least ordered. It is also clarified because objectives, purposes and criteria are less elusive. The impact of reforms, for instance, is not judged upon such vague notions as ‘welfare’ or ‘optimality’ but according to more operational concepts such as rates of growth of production, industrial productivity, and stability of outlets (see Chapter 4 by Colin Kirkpatrick and Jassodra Maharaj). This makes it possible to move from polemics to controversy. We do not lose much generality because foreign trade reforms occupy a central position in liberalization proposals. By looking at things from the foreign-trade perspective, we shall unwind the whole coil of interactions between structural adjustment reforms (see Chapter 2 by John Nash). We shall first discuss some of the theoretical bases of the liberal perspective, before delineating an alternative argument to which this book hopes to contribute. A brief analysis of the contributions closes this introduction. The neo-liberal export-orientation paradigm The neo-liberal view is founded on a critical analysis of import-substituting industrialization (ISI). Although often attributed to neo-liberals (Little, Scitovsky and Scott 1970; Krueger 1978; Bhagwati 1978; Balassa and associates 1982), this critical analysis was first elaborated (CEPAL 1969) and further developed (SidAhmed 1981; Fajnzylber 1983; Ominami 1986) in the structuralist tradition. Only a few points will be made here, as an introduction to contributions. A more complete analysis is presented by Fanelli, Frenkel and Rozenwurcel in Chapter 5. Anti-export bias From the strict point of view of resource allocation, ISI is blamed for its depressing effect on exports.9 ISI induces a distortion in the structure of incentives, with productive factors being attracted to protected import-substituting sectors, and export activities directly and indirectly penalized. Attraction of investment and productive resources by protected import-substituting sectors results from higher profitability induced by protection. Limitation of imports weakens competition, protective tariffs guarantee high selling-prices on the domestic market, high profits, and rents. Relative prices shift in favour of import-substitutes, which then attract investment and productive resources. On the other side, exports are penalized by the comparative evolution of domestic and world prices. Protection of import-substitutes translates into domestic price increases. Through wages, in the first stage of import-substitution, which touches consumption goods, then through inputs, as ISI deepens into the productive structure. Since exporters use labour and home-produced inputs, this affects adversely their terms of trade. And since exporters are typically price takers on the world market, they cannot pass their costs on, profitability in export sectors falls and exports contract. Only the most profitable export-sectors survive, i.e. those where comparative advantages are firmly rooted, which in practice means traditional exports. Thus export-diversification is discouraged by import-substituting strategies.

INTRODUCTION

5

This anti-export bias is aggravated by a number of parasitic phenomena. One is the tariff-cascade, which aggravates the anti-export-bias as import-substitution deepens.10 Another is the imposition of export taxes, levied to finance industrial infrastructure. Finally, real rates of exchange tend to appreciate, either through a deliberate policy of keeping down the prices of imported industrial inputs, or following the increase in domestic absorption brought about by high investment in the industrial sector. The global result is that, comfortably wrapped in the blanket of protectionism, import-substitution sectors never grow out of infancy. They fail to reach their optimal scale of production on the domestic market, where high costs and prices arising from below-capacity operation discourage demand. Furthermore, they fail to break into the world market, because of anti-export bias. If, in addition, imports of inputs rise faster than imports of finished products fall, the economy turns globally more dependent on imports. And exportsectors are increasingly relied upon to provide the foreign-exchange needed to supply industrial sectors with spare-parts, inputs and machinery. The inward-orientation becomes a spiralling self-aggravating feature, industry becomes more and more dependent on export sectors and is threatened by collapse when export prices or volume fall. The same argument also applies to the case where imports are restricted to restore trade-balance equilibrium rather than for import substitution reasons. The effect is possibly worse, since import reductions will be imposed indiscriminately, contracting the supply of imported inputs as well as finished goods, thus disregarding the priorities of import-substitution strategies (see Chapter 9 by Ademola Oyejide). The central point here is more protectionist policies, and especially import-restrictions, than inputsubstitution as such. This accounts for the central importance attached by liberals to foreign-trade reforms. We should note, however, that the viciousness of protection does not so much arise from protection as such as from its perenniality. What happens is that, once granted, protection will prove hard to reverse. Protection thus graduates into anti-productive rent (see Chapter 11 by Nurhan Yentürk). If the protection mechanism could be made reversible, then quite a different picture would emerge (see Chapter 3 by David Evans and Chapter 12 by Mario Lanzarotti). Tariffs, quotas, rents and DUPes In abstracto, protection policies through tariff or physical restrictions are equivalent. If tariffs are properly calculated, they will produce the same protective effect as quotas. Equivalence is, in principle, easily calculated.11 Judging strictly from the point of view of resource allocation, quotas and ‘equivalent’ tariffs will induce the same distortions both on level of production and welfare. However, this principle of equivalence between tariffs and quotas has been questioned from the liberal point of view. One line of argument, initiated by Bhagwati (1965), implies that tariffs will generally be superior to quotas. The original argument by Bhagwati is that, if markets are not perfectly competitive—in the case, for instance, of monopoly in the supply of imported goods-quotas will induce higher distortions than tariffs. The reason is, quite simply, that quotas are more rigid. They reinforce monopoly power, and allow the supplier of importables to push prices up by restricting supply. However, in the case of tariff protection, supply remains flexible: once domestic prices have risen to import-parity level, imports will resume. Tariffs thus define an upper bound on the prices of importables, which will be below the domestic ‘quota plus monopoly’ price. The argument could be rephrased in terms of market-contestability (Baumol, Panzar and Willing 1982): while quotas strengthen monopoly, tariffs will allow ‘imperfect contestability’ (the ‘right of entry’ of the foreign competitor is equal to the import-tariff rate) thus limiting the possible distortion arising from domestic monopoly. Since restricted competition is a common characteristic of markets in developing

6

INTRODUCTION

countries (see Chapter 4 by Colin Kirkpatrick and Jassodra Maharaj), the argument carries some weight in the present context. Quota rigidity also prevents quantity adaptation to shifts in demand. This can oppose the modification of the structure of exchanges (i.e. oppose moves along the production possibility frontier), thus preventing positive international respecialization (Bhagwati and Srinivasan 1982). One further consideration against quantitative restrictions is that, since quotas yield important scarcityrents, prospecting and lobbying for the allocation of quotas will develop, distracting high-level labour resources from productive purposes to non-productive, but lucrative, canvassing. Labour-resources will then also be misallocated, aggravating the overall resource-misallocation resulting from import restriction. The notion of ‘rent-seeking’ activities was first developed by Anne Krueger (1974), and later generalized to a broader class of activities by Bhagwati (1982) under the appellation of DUPes (Directly Unproductive Profit-seeking activities). Tariffs, on the other hand, while containing distortions, will yield public revenue instead of private-rents, thus facilitating stabilization, which is a prerequisite for foreign-trade liberalization (see Chapter 2 by John Nash). This directly inspires the sequencing philosophy of reforms. The first step should be to abolish quantitative restrictions and replace them by tariff-equivalents. This would enhance government revenue and ease stabilization, and supply the government with a flexible instrument, making it possible to modify gradually the trade-regime (see Chapter 8 by Anne-Marie Geourjon, retracing the reform strategy in Senegal). Starting from an inward-orientation, tariff-protection will allow the trade-bias to be modified gradually by changing tariff rates. The eventual objective is to bring about a balanced trade-regime. Reducing the dispersion of tariffs between all classes of imports (inputs, machinery, processed goods and finished consumer goods) will abolish discrimination between sectors, and lowering the average tariff will induce neutrality between home-market, import-substitution and export activities. In such a situation, production opportunities on the domestic markets will reflect opportunities on the world market, and international specialization will reach optimality (Bhagwati 1987). Arguments in favour of export-orientation All this explains why import-controls are considered so negatively by liberals: they are held responsible for the existence and generalization of the anti-export bias. Reforms should begin by changing the trade regime, shifting it to neutrality or export-orientation. The critique of import-controls is supplemented by a positive case in favour of export-orientation, argued on two grounds. The first point is that an industry which ventures on the world market must avoid two pitfalls. One is the limitation of domestic demand, generally too small to allow optimal use of installed equipment. Assuming developing countries to be ‘small countries’, their industry will face a potentially unlimited demand, and the traditional import-substitution vicious circle (where domestic outlet inadequacy prevents the optimal range of production, thus keeping costs too high to allow expansion of domestic demand) will be broken. The other concerns technological evolution. The protective cocoon will be broken, industry will have to face foreign competitors on a price-taker market, and will be pushed into adoption of modern production techniques. With direct access to foreign exchange, they will stop relying on traditional exports and will have both the means and incentive to adopt new techniques and equipment. The second point is rather more subtle and has to do with automaticity of corrective measures in case of policy mistakes. While import-substitution sparks off a self-reinforcing call for protection, exportpromotion policies will either succeed, or be abandoned.

INTRODUCTION

7

In the event where strict ‘neutrality’ in the Bhagwati sense12 cannot be attained, implementation of a ‘proexport’ bias provides a good second-best policy, where unavoidable policy mistakes will be self-corrective. Export-promotion policies require direct and costly incentive manipulation (e.g. export-subsidies, tax rebates, tax-free imports, tax allowances and remittances, etc.). Since price-manipulation on international markets is beyond government’s control, costs cannot be passed on to the consumer and directly bear on the government budget. With costs directly identifiable and bearing on the decision-maker’s budget, detection and correction of policy mistakes will be almost automatic (Krueger 1980). Conversely, protection of import-substituting industries is based upon the possibility of price manipulations on the domestic market. The costs of protection are then diffuse, bear on many private budgets, and benefit a small group, that is, the protected entrepreneurs. The resulting lobbying structure will favour protection, while no natural spokesman will rise to articulate the case for the vague collective identity of consumers. This explains why, although inefficient and costly, protectionist policies could prevail for so long. QUALIFICATIONS TO THE LIBERAL ARGUMENT AND OUTLINE OF A POSITIVE CASE FOR IMPORT CONTROLS By focusing on distortions, the neo-liberal argument follows the rhetoric of double-negation, where suppression of a negative feature is equated with positive reform. This overlooks the possibility that importcontrols have positive influences. The remaining part of this introduction will delineate the outlines of a positive case for import controls. We shall start by tuning-down the tone of neo-liberal critiques, bringing in a few points missed in the discussion. Then the potential of direct import controls in a long-term and strategic perspective is considered. Tuning-down the tone When considered in its own setting, the argument against import controls appears intellectually sound. But adaptation to reality reveals a number of weaknesses or over-simplifications. The tariff vs quota argument is crucial in defining the sequencing and timing of reforms. As such, it immediately bears upon the shaping of policy. It is built upon three considerations: generation of fiscal revenue, allocative efficiency, and welfare, which will be examined in turn below. The first argument, which pertains to fiscal revenue generation of tariffs, is not a very strong one. Nothing prevents the auctionning of quotas (a possibility considered by David Evans in Chapter 3), in which case the rent would be shifted to the government budget, and quotas would yield fiscal revenue. In so far as private appropriation of scarcity rent is concerned, quotas can operate as tariffs, and the fiscal revenue argument can be disposed of. The argument based upon greater flexibility of tariffs—which commands both allocative and welfare effects—is apparently a stronger one. But it rests upon a heavy assumption, that of stability of prices of importables. Quite surprisingly, developments of the tariff vs quota argument taking uncertainty into account (Fishelson and Flatters 1975; Pelcovits 1976; Driscoll and Ford 1983) have passed unnoticed in the discussion of policy options. And yet, since industrial goods display increased price-instability, implications could be important.13 With unstable prices of importables on the world markets, shifts in domestic opportunity costs might prove quite brutal and disruptive. In such a situation, tariffs will reflect and accentuate world-price instability,14 and physical containment of imports might be preferred as yielding greater stability. This depends upon the source of uncertainty. If it originates from shifts on the domestic

8

INTRODUCTION

market, then the usual argument favouring tariffs against quotas applies: adaptation to domestic shifts (e.g. shifts in the structure of demand) requires flexibility. And since tariffs allow flexibility of imports, they should be preferred. Conversely, if uncertainty originates on the world market (e.g. in the case where prices of imports fluctuate widely from year to year), then quotas might prove superior to tariffs. This will be especially true if domestic industry has some strategic objectives beyond static efficiency (e.g. maintaining a given level of activity in import-substituting sectors). In this case, predetermined domestic objectives can only be disrupted by world price instability, which twists the option in favour of quotas. Since international instability is now a common feature, the uncertainty argument carries a lot of weight, and discussions on the abolition of quotas need reconsideration. One other important point concerns the reversibility of policies. Protection through import-controls, either through discriminating tariffs or quantitative restrictions, is seen as self-perpetuating and aggravating. But the notion that export-orientation policies are self-corrective, although intellectually appealing, might be just that: intellectually appealing. The case of Turkey (analysed by Nurhan Yentürk, in Chapter 11) shows that export-oriented lobbies can be as powerful, effective and baneful as protectionist inward-looking ones. Conversely, the analysis of the Korean import-strategy (see Chapter 12) shows import-controls to be flexible, reversible, and adaptable to strategy. Neo-liberal analysis rests on a somewhat simple, negative and naïve view of the interrelation between public power and private interests. The possibility of building positive coalitions between the state and the business community, centred around joint management of import controls (as advocated by David Evans) should not be dismissed. If properly explored and relied upon, it might yield promising policy alternatives, which is perhaps what is happening in Senegal (see Chapter 8). Finally, empirical verification does not follow the abstract demonstration of the benefits of foreign trade liberalization. Although Papageorgiou, Choski and Michaely (1990), in a study prepared for the World Bank, presented evidence of positive links between foreign trade liberalization and growth, re-examination of the methodology and data by Evans, Goldin and van der Mensbrugghe (1991) casts serious doubts on the reliability of their results. As noted earlier in a more general context (see p.3 in ‘Background to the debate’), attribution of causality is uncertain. Macroeconomic reforms, upturns and downturns in the general economic outlook, as well as availability of fundings, also contributed to performance. Attribution of success to trade-reforms is somewhat arbitrary. Short of solid and unquestionable empirical verification, we are driven back to theory, which was seen to yield indeterminate results. Outline of a positive case for import-controls The neo-liberal argument against import-controls is based upon price distortions. Discriminating tariffs directly modify relative prices, and quantitative restrictions repress demand, pushing prices up. But, depending on the class of imports that are in short supply, shortage can also exert direct physical effects. This will be the case when import-shortages touch inputs. One class of effects, disregarded in the neoliberal discussion, has to be reintroduced: that which follows the change in the structure of imports that import-liberalization brings about. The outline of the argument is as follows: if import-restrictions are lifted, imports of goods that were previously banned or restricted will catch up with formerly repressed demand (Ocampo 1987). Importliberalization induces a shift in the structure of imports in favour of processed consumer goods (see Chapters 9 and 10). Unless import capacity has risen to allow adequate domestic supply of both classes of imports, domestic input availability will contract, thus strangling industrial sectors. Domestic industry will then be taken in pincers, between competition with imported processed goods from the demand side—

INTRODUCTION

9

which is supposed to exert positive rationalizing influences—and input shortages—which will constrain and disrupt the supply-side. In other words, depending upon the ‘spontaneous’ or ‘uncontrolled’ structure of imports, import strangulation might bear heavily on industrial inputs, disrupting production from the supplyside. In such a situation, physical control of imports will be required to limit input-scarcity. Imported-input strangulation obviously threatens sectors that manufacture goods for the domestic market. Since import-compression has also been shown to constrain exports (Khan and Knight 1986), inadequacy in the domestic supply of export-oriented imported inputs will also threaten exports. In other words, facing the neo-liberal argument that import restrictions generate a price bias against exports, one other argument is that import liberalization will induce a quantity-based bias against all activities requiring imported inputs, including exports. In some ways then, in import compressed economies, import protection is part of an export promotion policy (Krugman 1984). This, as is well known, was actually the case in South Korea. Consider now the case of countries that need to diversify both domestic production and exports, for instance countries with a regressive export base, locked into ‘fallacy of composition’ as most African countries are (Godfrey 1985; Karunasekera 1984). Will trade liberalization ‘spontaneously’ ensure exportdiversification? Will incentives resulting from devaluation plus foreign-trade liberalization prove strong enough to stimulate the creation of new exporting sectors? The answer can very well be ‘No’ (Kulatunga 1990; Fontaine 1992). Initial costs are high, and non-intervention will typically favour existing export activities, which means traditional exports only. Strategic considerations ought then to enter into the discussion. Although the ‘strategic trade’ argument might not apply widely to developing countries (F.Stewart forthcoming), some important parts of it, those that have the strongest affiliation with ‘infant industry’ arguments, do have relevance here (Stewart 1984). The modification of international insertion, especially when it requires breaking away from existing specialization and developing sectors with a strong potential for economies of scale and learning effects, will justify interferences in the foreign trade sector. This is so not only in terms of export-subsidies or classical infant industry protection, but also in terms of monitoring adequate quantitative supply of inputs. The choice of the relevant form of import-control stems from a straight-forward logical argument: if the aim is adequate quantitative supply, then a quantitative instrument will be fittest. It is selective, directly influences both volume and composition of imports, and allows the by-pass of the price transmission mechanism which, in periods of world instability, exerts disruptive influences (the structuralist arguments in favour of direct import controls are listed in Bird, 1984, pp.119 ff). Finally, we should put import controls in proper perspective. These are measures that respond to current account disequilibria and result from excess of domestic absorption, or from insufficient import capacity. In other words, import controls are, and should be considered, as disequilibrium management instruments. As such, they have to be weighed against other disequilibrium management instruments, which all have their drawbacks. Reduction of domestic absorption carries the danger of contraction and ‘overkill’ (DiazAlejandro 1981). Devaluation can initiate downward spirals in rates of exchange, spark off inflation (Assidon and Jacquemot 1988), cause economic contraction (Krugman and Taylor 1978). An increase in import-capacity is a long-term job, which requires strategic planning of export diversification. Perhaps this puts neo-liberal recommendations in perspective: most economists will agree that, except for strategic reasons that are in essence transitory, imports should be liberalized at some point in time. Determination of this ‘point in time’ gets one obvious answer: when equilibrium has been restored. Short of this ‘sub specie aeternitatis’ answer, there is scope for choice among many instruments, and scope for intervention in the management of imports. Where does this bring us to? Perhaps simply to the recognition that import-controls are, and should remain, transitory measures. This is not an unexpected discovery. This is what they were originally designed

10

INTRODUCTION

to be: transitory. Crises, failures of import-substituting sectors to mature, and attempts to deepen importsubstitution have constantly pushed the time of resolution ahead. Policy-relations, lobbying and coalitions have built transitory responses into quasi-permanent features. But the cause of departure from sound practice might have been mislocated by neo-liberals. We saw that the blame bore not on import-substitution as such, but on accompanying protectionist policies. Then, not protection as such, but irreversibility was spotted as the source of failure. The neo-liberal argument shrinks as analysis proceeds. Which brings us back to the question originally asked at the beginning of this introduction: does the vogue for neo-liberal concepts result from its positive content, or from the failure of alternatives? If such is the case, has the element of failure of the alternatives been properly identified? Since neither equilibria nor world stability have been restored, the immanent signs for overall liberalization have not manifested them-selves yet. In the meanwhile, we could convincingly advocate a return to the original design of import-control: strategic, transitory, reversible. PRESENTATION OF CONTRIBUTIONS This book is organized in two parts: the first one examines the general perspectives and circumstances of foreign-trade reforms, the second part is devoted to analysis of country-experiences. In Chapter 1, Paul Mosley presents an overall empirical evaluation of structural adjustment programmes (SAPs). Three points ought to be emphasized. In regard to degree of implementation, foreign-trade reforms rank quite low. Second, impact of SAPs is mixed: if SAPs exert positive influence on exports, and perhaps on GNP growth, they depress investment, including foreign investment, thus endangering the shift from structural adjustment to self-sustained growth. Finally, analysis of a sample of countries illustrates the importance of the point of departure: depending on the initial stage of the economy, a completely different policy package, centred upon rehabilitation and public investment in a Keynesian fashion, could be necessary. Which qualifies structural adjustment as a ‘somewhat, and on balance a middle-income, item of consumption’. Chapter 2 by John Nash focuses more specifically on foreign-trade reforms and presents a very complete panorama of the World Bank’s expectations regarding these reforms, of their degree of implementation and of obstacles opposing implementation or success. Implementation remained below expectations, which is accounted for by the existence of lobbies. Analysis of sequencing problems shows that, if in principle foreign-trade liberalization is a prerequisite for other reforms, it might have to be delayed. This is the case when obstacles to foreign-trade liberalization do not result from foreign-trade regulation, but from domestic features (e.g. domestic distortions such as administered prices) or when other policy tasks, such as domestic stabilization, have to be conducted before reform is possible. One of the major conflicts between policy objectives will concern the rate of exchange, which must respond both to the requisites of international trade (which imply devaluation) and to those of anti-inflationary action (which is made easier by a real appreciation). Replacement of quantitative restrictions by tariffs will go some way towards lessening the conflict, since it will enhance public revenue, thus facilitating stabilization. Chapter 3 by David Evans presents a general argument in favour of import-controls. By taking into account rent-seeking attitudes and potential externalities in a strictly neo-classical framework, he explores the possibilities of introducing performance-linked incentives. The nature and philosophy of state intervention consistent with the administration of such non-conventional incentive schemes is then analysed, and joint-management schemes of import-controls by the state, intermediate organizations and industrialists are recommended. The chapter ends with a comparison between two sequences of foreign-trade reform

INTRODUCTION

11

(based upon neo-liberal and interventionist principles) and concludes with the possibility of building positive, coherent and productive protectionist coalitions, thus overturning the usual negative image of coalitions conveyed by the DUPes literature. In Chapter 4, Colin Kirkpatrick and Jassodra Maharaj look into the determinants of industrial productivity and assess in particular the impact of foreign-trade reforms. Surveying empirical literature on the subject, they note the absence of significance of the link between trade reform and total factor productivity. When such a link does appear, as in South Korea, the direction of causality is ambiguous and we cannot decide whether productivity increased because of export-expansion, or whether exports followed from increased productivity. The influence of market structure is equally examined. One possible sequence links trade reform, increased industrial concentration and rationalization of production, as in the case of Chile, but entails a phase of de-industrialization. The general tone on the question of the impact of liberalization upon industrial efficiency is one of scepticism. We chose to place Chapter 5, by José Maria Fanelli, Roberto Frenkel and Guillermo Rozenwurcel, with the general contributions rather than with the country studies, although they analyse the experience of five Latin-American countries. The reason is that, Latin America being the mother continent of importsubstitution, analyses originating from there bear somewhat naturally general relevance to all importsubstituting and foreign-trade reform experiences. The other reason is that their analysis examines countries that display many simultaneous symptoms: anti-export bias, domestic distortions, heavy government intervention, internal deficits, high domestic inflation and excessive foreign debt. This accumulation of handicaps makes them both atypical, as regards situations analysed by theory, and representative of practical difficulties and conflicts between policy-objectives that countries attempting foreign-trade reform have to face. Starting from the description of the ‘Washington Consensus’ policy package, they develop a critical analysis based upon two sets of factors: the possibility that economic dynamics might lead to explosive paths, and the permanent self-restoring tendency of internal disequilibria. In the face of both these tendencies, foreign-trade liberalization proves impossible, or suicidal. The interaction between foreign shocks, especially financial shocks, and requisites of domestic stabilization results in a situation where public deficits constantly reconstitute themselves, feeding inflation and domestic instability. Contradictory pressures then bear upon policy instruments, especially on the rate of exchange which has to exert three incompatible functions: promote exports, stabilize the domestic situation, and anchor expectations. When all available policy instruments have to be used, government control upon export receipts proves an invaluable asset which explains part of the relative (Colombia or Chile) or transitory (Mexico) successes. Regarding foreign-trade liberalization proper, one necessary prerequisite, a stable and balanced rate of exchange, is out of reach, unless one is prepared to use multiple (i.e. administered) rates of exchange. In any case, foreign-trade liberalization can only be partial: articulation between liberalization of trade balance and capital account is extremely delicate and dangerous, and trade account liberalization itself should be restricted to intermediate and equipment goods only, lest import structure massively shifts in favour of consumer goods. The last two chapters of this first part examine actual external conditions of trade liberalization. Chapter 6 by Juan de Castro identifies factors commanding the emergence of protectionist tendencies in industrialized countries, and assesses their impact on future trade. Starting from a framework analysing past protectionist tendencies, a model is constructed that links reactions of lobbies to the evolution of structure of trade and domestic evolution of industrial sectors. Sectors in industrial countries, where adjustment costs to changes in international trade are highest, where the general outlook is depressing, or where imports for developing countries are most threatening, will exhibit strongest protectionist tendencies. These are also the

12

INTRODUCTION

sectors where developing countries can most easily break into the world market. Protectionism on the part of industrialized countries will then prove a major obstacle to the expansion and diversification of exports from developing countries, although new forms of international trade, especially intra-firm trade, will oppose this adverse tendency. In Chapter 7, Jean Coussy reviews problems of regional integration in Africa. Considering things from a historical perspective, he expounds three successive notions that inspired attempts at regional integration and spots factors responsible for failure. The first attempts, based upon the notion of educative protectionism, failed because import-substituting industries never grew out of infancy. Second attempts, where existing activities were to be integrated regionally, failed because protection against the rest of the world prevented adaptation to new competitors, especially from Asia. Present-day integration attempts through liberalization of goods and factors of production are hampered by the existence and resilience of informal exchanges and strategies as well as by the divergences between various domestic policies and expectations. One feature common to all these stories is policy conflict between objectives: the multiplicity of objectives assigned to integration complicated matters to the point where policy instruments were used in a self-contradictory manner, ending in paralysis. Country studies begin in Chapter 8, by Anne-Marie Geourjon, who examines the relation between trade liberalization and the new Industrial Policy in Senegal. Although measures appeared as ‘intell ectually sound’, they were quite promptly reversed. The notion that foreign-trade reform should be implemented prior to others stumbled against the difficulty to reform segments of the domestic economy, especially industrial policy and labour legislation. Too rapid liberalization thus placed the economy in overhang. Trade-liberalization had a negative impact on industrial activity: not only did it accelerate industrial recession, but it favoured traders over industrialists and, by increasing the number of participants in importexport activities, weakened the state’s control capacity on fraud. Restoration of import-controls shows how effective these can be in limiting disequilibria. The positive outcome was that, in front of failure to implement reform as planned, all actors, including the World Bank, realized that new attitudes were required, involving greater cooperation. This perhaps announces the new coalitions recommended by David Evans. Chapter 9, by Ademola Oyejide, compares the impact of price-based and quantity-based import controls in Nigeria. One essential point concerns the evolution of the structure of imports between consumer goods, intermediate products and equipment. The impact is measured through a model which breaks down the growth of various imports between market expansion and modification of import structure. Quantitative controls are shown to reinforce the substitution in favour of intermediate goods and equipment—which is consistent with import-substituting strategies—while relaxation (i.e. price-based controls) reverses this trend. Crisis management, on the other hand, which was resorted to in periods of balance of payment difficulties, induces global compression rather than structure monitoring and induces inconsistency. Which illustrates again the importance of conflict between policy objectives. In Chapter 10, Jean-Marc Fontaine analyses the impact of trade liberalization upon both volume and composition of imports in Kenya. Two tendencies established by Oyejide in the case of Nigeria are also found here: a long-term substitution of imports in favour of industrial inputs and equipment, and a liberalization bias in favour of consumer goods. The hypothesis of an import-compressed manufacturing sector does not seem verified, because domestic industry has substituted domestic for imported inputs. However, trade-liberalization slows down this input-substitution, giving strength to infant industry arguments in favour of input-substituting sectors. Owing to the importance of regional markets for manufactured exports, the liberal sequence linking export-diversification to productivity increases resulting from liberalization does not apply here. Finally, increased reliance on rate-of-exchange policy to control

INTRODUCTION

13

volume and structure of imports might stumble against ‘devaluation fatigue’, whereby the impact of nominal devaluations upon the volume of imports weakens when devaluation becomes a usual policy instrument. The last two chapters concern two countries often quoted as successful examples: Turkey and South Korea. Chapter 11, by Nurhan Yentürk-Coban, analyses the case of Turkey, where the trade regime was shifted to export-promotion, breaking away from previous inward-looking and protectionist policies. When judged strictly in terms of volume of industrial exports, the policy is a success. If, however, we look into the structure of Turkish industry, we find that no proper adjustment has taken place: foreign markets were substituted for domestic ones, the share of manufacturing in GNP has remained stable, and no technological improvements have occurred. Investment had come to a halt, while costs remained high. In fact, exportpromotion was brought about by a mixture of compression of domestic demand, which crowded manufacturing protection out of internal markets, and export protection, with high export-subsidies paid to commercial exporters rather than to industrialists themselves. This contradicts two neo-liberal tenets, namely that export-promotion policies are self-correcting, and that technological improvements are a quasiautomatic by-product of export-promotion. Finally, the case of Korea is examined by Mario Lanzarotti in Chapter 12. One well-known feature is illustrated here: the importance of state-intervention in foreign-trade management. Precise analysis of the means utilized to influence exports shows how subsidies have been used to avoid frequent devaluations, with subsidies increasing when the rate of exchange appreciated. Three important theoretical intuitions are substantiated here. One is that the use of non-conventional incentives tied to past performance can prove very effective. The second is that such specific instruments, as targeted subsidies, prove more flexible and adaptable than global rate-of-exchange management. If properly administered and monitored in cooperation with industrialists, subsidy schemes prove reversible, not only in the sense of not growing into a permanent rent, but also in the sense that they allow, and can promote, changes in export strategy, that is, reverse previous export priorities. The third one is that, in incompletely industrialized countries, mobilization of resources, especially as regards reallocations from non-tradable to tradable sectors, reallocations within trad ables, from importables to exportables, and, within exportables, from old, or traditional, to new, or non-traditional sectors, might display higher elasticity in relation to subsidy than to the rate-of-exchange (Schydlowsky 1982). The general lesson we might draw from this last chapter is that some research could be fruitfully devoted to the exploration of ways of introducing flexibility in import controls, rather than in theoretical elaborations within the fairly idealistic, formalized and restricted set of neo-liberal assumptions. As they stand, contributions to this book cover, we hope, a wide enough range to allow the reader to build an opinion of the importance, the necessity and also the potential dangers of import-liberalization. While perhaps adapted to circumstances that prevailed in the 1970s, systematic import-substitution strategies have displayed their weaknesses. Bets on rapid industrialization based upon domestic markets alone have been lost—and dramatically so in Africa. International specializations have to be reconsidered. But in a world where minus times minus does not systematically yield plus, the systematic back-scent followed by international organizations has probably not uncovered the new philosophy that the 1990s require. Although it cannot claim its exclusive paternity, an important potential for critique is embodied in the neo-liberal corpus. Resting on a rigorous and coherent analytical apparatus, it has fused a number of critical observations into one coherent set of doctrine. The argument flows naturally as in a demonstration of pure logics, delineating adaptation strategies that have a touch of perfection. But this is perhaps where the weak point lies: in formal perfection.

14

INTRODUCTION

Because concrete situations do not respect postulates, even less so when postulates are more static. Because emergencies do not follow the logical order of appearance, but crowd one another in as they come. And, finally, because evolutions are discontinuous, unpredictable and brutal. And because resilience, rather than optimality, is the master-word for survival today. The search for resilience requires greater attention to dynamic adaptation mechanisms, to conditions under which governments must control domestic imbalances, and to means for attaining long-term objectives—even if the price to be paid is short-term distortions. NOTES 1 See, for instance, the interventions of African high-ranking civil servants in G.Helleiner (1987). 2 See Killick (1989). 3 For instance, the assessment by Zulu and Nsouli (1985) shows a good success record on a group of indicators. However, close examination of the success record reveals that success is higher on indicators that are really instruments (devaluation, domestic credit expansion or inflation) than on indicators of objectives such as GDP growth or Balance of Payments (Fontaine 1989). Similarly, evaluations presented by the World Bank (World Bank 1987; World Bank and UNDP 1989) have been severely criticized for methodological inconsistency (ECA 1989). 4 See Chapter 1 in this book, by Paul Mosley, or Mosley, Harrigan and Toye (1991). 5 And even so, quite restrictively, under the assumption that, either prices do have a clear meaning and interpretation, e.g. as indicators of social preferences, or that quantities to be produced are unambiguously and totally ordered according to some criterion. 6 The market was introduced into the theory by Walras only as an illustration. It was then a tale, with peasants exchanging surpluses of apples and pears like in children’s books. 7 This was in fact the starting point of a famous controversy, which opposed, inter alia, L.von Mises and F.von Hayek to Oskar Lange in the 1930s, i.e. central planners and liberals. See von Hayek (1935) and Lange (1936). 8 Rates of growth based on figures from IMF, International Financial Statistics. 9 Balassa (1980) is most often quoted on this point. A very complete, clear and compact exposition of the theoretical argument is given in Schydlowsky (1982). 10 When import-substitution extends to production of input B, input B will have to be protected, and its price will rise above world-price. Since B is an input in formerly protected activity A, production costs in A will rise accordingly, requiring increased protection. Practically, this means that the protective tariff for sector A will rise to include the effects of increased tariffs on input B. Tariffs ‘cascade’ from one sector to the next. As importsubstitution deepens, cascades multiply, and anti-export bias strengthens. 11 Quantitative restriction, e.g. through imposition of quota, increases domestic scarcity of goods, pushing marketprice up. If the resulting price can be calculated, imposition of an import tax equivalent to the difference between the world-price and the domestic market price resulting from the imposition of a quota would induce a reduction in domestic demand, so that the quota would be ‘spontaneously’ respected (see for instance Winters 1985, Ch.7). 12 That is, a situation where incentives would not favour either home or export-markets. 13 This paragraph does not follow the original arguments by Fishelson and Flatters, Pelcovits and Driscoll and Ford. This is a heuristic elaboration upon their results, which were established in terms of welfare comparisons. 14 If the domestic price of an imported good is determined by world price multiplied by the rate of exchange, and increased by a given proportion through imposition of a classical ‘ad valorem’ tax, then the ‘ad valorem’ tax accentuates an upward shift (or absorbs part of the downward shift) of the international price of the imported goods.

INTRODUCTION

15

REFERENCES Assidon, E. and Jacquemot, P. (eds) (1988) Politiques de Change et Ajustement en Afrique, Paris: Etudes et Documents, Ministère de la Cooperation. Balassa, B. (1980) ‘The Newly Industrializing Developing Countries after the Oil Crisis’, World Bank Staff Working Paper No. 437, Washington DC: World Bank. Balassa, B. and associates (1982) Development Strategies in Semi-Industrial Countries, Baltimore, Md: Johns Hopkins University Press. Baumol, W.J., Panzar, J.C. and Willing, R.D. (1982) Contestable Markets and the Theory of Industry Structure, New York: Harcourt Brace Jovanovitch. Bhagwati, J.D. (1965) ‘On the equivalence between tariffs and quotas’, in R.E.Baldwin (ed.) Trade, Growth and the Balance of Payments: Essays in honor of G.Haberler, Chicago: Rand McNally. (1978) Foreign Trade and Economic Development: Anatomy and Consequences of Exchange Control Regimes, New York: National Bureau of Economic Research. (1982) ‘Directly Unproductive, Profit-seeking (DUP) activities’, Journal of Political Economy, 90 (5) October: 988–1002. (1987) ‘Outward orientation: trade issues’ in V.Corbo, M.Goldstein and M.Khan, Growth-Oriented Adjustment Programs, Washington DC: IMF and World Bank. Bhagwati, J.D. and Srinivasan, T.N. (1982) ‘The welfare consequences of directly-unproductive profit-seeking (DUP) activities: Price versus quantity distortions’, Journal of International Economics, 13 (2): 33–44. Biermann, W. and Campbell, J. (1989) ‘The chronology of crisis in Tanzania: 1974–86’, in B.Onimode (ed.) The IMF, the World Bank and the African Debt, Vol. I, London: Zed Books. Bird, G. (1984) ‘Balance of payments policy’, in T. Killick (ed.), The Quest for Economic Stabilisation, Vol. I, London: Heinemann. Buchanan, J.M. and Tullock, G. (1967) The Calculus of Consent, Ann Arbor, Mich: University of Michigan Press. CEPAL (1969) El Pensamiento de la Cepal, Santiago de Chile: Editorial Universitaria. Cline, W. (1982) ‘Can the East Asian model of development be generalized?’ World Development, 10, February. Diaz-Alejandro, C. (1981) ‘Southern cone stabilization plans’, in C. Cline and S.Weintraub (eds), Economic Stabilization in Developing Countries, Washington DC: The Brookings Institution. Dick, H., Gupta, S., Mayer, T. and Vincent, D. (1983) ‘The short-run impact of fluctuating primary commodity prices on three developing countries: Colombia, Ivory Coast and Kenya’, World Development, 11:5. Driscoll, M.J. and Ford, J.L. (1983) ‘Protection and optimal trade-restricting policies under uncertainty’, The Manchester School for Economic and Social Studies, March, 21–33. Economic Commission for Africa (ECA) (1989) Statistics and Policies, Addis Ababa: United Nations ECA. Evans, D., Goldin, I. and van der Mensbrugghe, D. (1991) ‘Trade reform and the small country assumption’. Paper presented to the Conference on International Dimensions to Structural Adjustment, CEPR (London) and OECD Development Centre, Paris. Fajnzylber, F. 1983, La Industrializacion trunca de America Latina, Mexico City: Nueva Imagen. Fischer, S. (1990) ‘Comment’ to Williamson, in J. Williamson, Latin American Adjustment: How Much Has Happened? Washington DC: Institute for International Economics, pp. 25–9. Fishelson, G. and Flatters, F. (1975) ‘The (non) equivalence of optimal tariffs and quotas under uncertainty’, Journal of International Economics, 5 (4):385–95. Fontaine, J.M. (1988) ‘L’Economie Tanzanienne: involution ou survie?’ in F. Constantin and D.C.Martin (eds) Arusha (Tanzanie), vingt ans après, Pau: CREPAO. (1989) ‘Diagnostics et remèdes proposes par le Fonds Monétaire International pour l’Afrique: quelques points critiques’, Tiers Monde, XXX (117), Paris. (1992) ‘Bias overkill? Removal of anti export-bias and manufacturing investment: Ghana, 1983–9’, in C.Kirkpatrick, R.Adhikari and J. Weiss, Industrial Trade and Policy Reforms in Developing Countries, Manchester: Manchester University Press.

16

INTRODUCTION

Gerschenkron, Alexander (1966) Economic Backwardness in Historical Perspective, Cambridge, Mass.: Harvard University Press. Godfrey, M. (1985) ‘Politique commerciale et politique de taux de change: une nouvelle contribution au débat’, in T. Rose (ed.) Afrique Sub-sharaienne: de la crise au redressement économique, Paris: OECD Development Centre. Guillaumont, P. and Guillaumont, S. (1990) ‘Why and how to Stabilize Producer Prices for Export Crops in Developing Countries, UNDP/World Bank Trade Expansion Programme occasional Paper, No. 6, Washington DC: Trade Policy Division, The World Bank. Helleiner, G. (ed.) (1987) The IMF and Africa, Washington DC. IMF. Karunasekera, M. (1984) Export Taxes on Primary Products: A Policy Instrument in International Development, Commonwealth Economic Papers No. 19, London: Commonwealth Secretariat. Khan, M. and Knight, M.D. (1986) Import Compression and Export Performance in Developing Countries, Development Research Department Discussion Papers, Report DRD 197, Washington DC: World Bank. Killick, T. (1989) A Reaction Too Far: Economic Theory and the Role of the State in Developing Countries, London: ODI. Kirkpatrick, C, Adhikari, R. and Weiss, J. (eds) (1991) Industrial Trade and Policy Reforms in Developing Countries, Manchester: Manchester University Press. Krueger, A.O. (1974) ‘The political economy of rent-seeking society’, American Economic Review, June: 291–303. (1978) Foreign Trade Regimes and Economic Development: Liberalization Attempts and Consequences. Cambridge, Mass.: NBER Ballinger Press. (1980) ‘Trade policy as an input to development’, American Economic Review, May. Krueger, A.O. (1981) ‘Interactions between inflation and trade objectives in stabilization programmes’, in W.Cline and S.Weintraub, Economic Stabilization in Developing Countries, Washington DC: The Brookings Institution. Krugman, P. (1984) ‘Import-protection as export-promotion: international competition in the presence of oligopoly and economies of scale’, in H. Kierskowski (ed.), Monopoly competition and International Trade, Oxford: Oxford University Press. Krugman, P.R. (ed.) (1987) Strategic Trade Policy and the New International Economics, Cambridge, Mass.: MIT Press. Krugman, P. and Taylor, L. (1978) ‘Contradictory effects of devaluation’, Journal of International Economics 8. Kulatunga, S. (1990) Export Development Prospects and Problems, UNDP/ ITC project Gha/97/004, June, Geneva: ITC. Lange, Oskar (1936) ‘On the economic theory of socialism’, reprinted in E. Lippincott (ed.) On the Economic Theory of Socialism, New York: McGraw Hill, 1964. Lipsey, R.G. and Lancaster, K. (1956) The general theory of the second best’, Review of Economic Studies, 24:1–32. Little, I.M.D., Scitovsky, T. and Scott, M. (1970) Industry and Trade in Some Developing Countries: A Comparative Study. New York: Oxford University Press. Ocampo, J.A. (1987) ‘The macroeconomic effects of import controls, a Keynesian analysis’, Journal of Development Economics, 27:307–38. Ominami, C. (1986) Le Tiers Monde dans la Crise, Paris: La Découverte. Papageorgiou, D., Choski, A. and Michaely, M. (1990) Liberalising Foreign Trade in Developing Countries: The Lessons of Experience, Washington DC: World Bank. Pelcovits, M.D. (1976) ‘Quotas versus tariffs’, Journal of International Economics, 6 (4): 363–71. Polanyi, Karl (1947) La Grande Transformation, French translation, Paris: Gallimard, 1983. Sachs, J. (1987) ‘Trade and exchange rate policies in growth oriented adjustment programs’, in Corbo, V. (ed.) Growth Oriented Adjustment Programs, Washington DC: IMF-World Bank. Schydlowsky, D. (1982) ‘Alternative approaches to short-term economic management’, in T.Killick (ed.) Adjustment and Financing in the Developing World, Washington DC: ODI/IMF. Sid-Ahmed, A. (1981) Croissance et Développement: l’Expérience des Economies du Tiers-Monde depuis1945, Algiers: OPU.

INTRODUCTION

17

Singer, H., 1989, ‘Lessons of post-war experiences’, Revue Africaine de Développement, Banque Africaine de Développement, Abidjan, Décembre. Singer, H. and Gray, H. (1988) ‘Trade policy and growth of developing countries: some new data’, World Development, 16, March. Stewart, F. (1984) ‘Recent theories of international trade: some implications for the South’, in H.Kierkowski (ed.) Monopoly Competition and International Trade, Oxford: Oxford University Press. Stewart, F. (forthcoming) ‘A note on “strategic” trade theory and the South’, Journal of International Development. Taylor, L. (1988) Varieties of Stabilization Experiences, Oxford: Clarendon Press. Toye, J. (1987) Dilemmas of Development, Oxford: Blackwell. Von Hayek, F. (1935) Collectivist Economic Planning, London: Routledge. Winters, A. (1985) International Economics, London: Allen & Unwin. World Bank (1987) World Development Report, Washington DC: World Bank. World Bank and UNDP (1989) Africa’s Adjustment and Growth in the 1980s, Washington DC: World Bank. Zulu, J.B. and Nsouli, S.M. (1985) Adjustment Programs in Africa: the Recent Experience, Washington: IMF.

1 Structural adjustment: a general overview, 1980–9 Paul Mosley

BACKGROUND ‘Structural adjustment’, for the purposes of this essay, is that part of development policy devoted to achieving a boost to the supply side of an economy by the removal of market imperfections; it is therefore to be contrasted with stabilization, which seeks to control the demand side, and also with long-term supplyside policies, such as research and sectoral investment policy. In the 1980s the phrase was used by many as a synonym for appropriate development policy and treated, like motherhood, as a good and necessary thing in itself; but even among those using our own strict definitions, there remain multiple differences between those who, like the Latin-American structuralists of the 1980s, favour the removal of imperfections through state intervention (for example, in land and credit markets) and those who, like the World Bank in the 1980s, favour their removal by state withdrawal; and within the latter group, disputes persist concerning which markets should be liberalized and in what order, which are further clouded by enormous inter-country differences in what is politically feasible. Figure 1.1 provides a map of the various pathways into which development policy, and debate about it, ramifies. The focus of this paper is on structural adjustment, in the sense defined above, and on the extent to which it has been able to bring about a lasting improvement in output and living standards in less-developed countries (LDCs). The World Bank, in the 1980s, has staked its reputation on the claim that ‘outwardlooking’ economies show higher levels of production efficiency than ‘inward-looking economies’ (e.g. World Development Report 1983, Chapter 4; World Bank 1989, Table 20) and that governments of all developing countries can increase living standards by making their economies more ‘outward-looking’, that is, by implementing a structural adjustment programme of the state-intervention-reducing variety. We shall be concerned to test this claim: by discussing the extent of implementation of structural adjustment programmes, and their aggregate effectiveness, and, then, attempting to discover why effectiveness varied between countries. The general conclusions are two. First, whereas there is indeed a statistically significant payoff to ‘structural adjustment’ considered as a package, certain elements of this package appear to be vital —notably exchange-rate policy and the rationalization of government investment—and other things irrelevant—notably privatization and foreign trade liberalization. Second, that such effectiveness as can be observed is acutely sensitive to a cluster of intermediate variables, of which some relate to the behaviour of the nation’s political economy (the extent of real wage compression and political tolerance of it, the nature of the private sector response to liberalization, the extent of political cohesion among exporters) and others to the relationship between the structural adjustment effort, and the success of the government’s stabilization and long-term development policies. In general, it will be argued, ‘appropriate structural

STRUCTURAL ADJUSTMENT: 1980–9

19

Figure 1.1 ‘Structural adjustment’ in relation to other elements of development policy

adjustment’ means an entirely different thing in a resource-poor economy with a shattered infrastructure than in a newly industrializing country (NIC). THE IMPLEMENTATION OF ‘STRUCTURAL ADJUSTMENT’ POLICIES We begin from the truism that no country can choose whether to adjust to external shocks, only how to do so; from this point of view, the distinction often made (for example, World Bank 1988, 1989) between ‘adjusting’ and ‘non-adjusting’ countries is unfortunate. A country that experiences an increase in its balance of payments deficit, that is, an increase in the net value of claims by nationals of overseas countries on its own domestic economy, must either meet these claims directly, persuade creditors to accept deferred payment or, if it chooses to repudiate the claims, work out strategies to mitigate the consequences of any retaliatory behaviour by creditors. All of these behaviours have a right to be called an adjustment policy. If we examine the behaviour of less-developed countries as a group to the shock imposed by the oil crisis and world depression of 1979–83, what is striking is that all less-developed countries, not only those who have been happy to accept economic support packages from the World Bank and IMF, have chosen to adopt policies that have sharpened incentives to producers to increase supply, in other words, policies of structural adjustment, as well as simple stabilization. As Table 1.1 demonstrates, even in those LDCs Table 1.1 Adjustment indicators for less-developed countries as a group, 1980–8

Real effective exchange rate (index 1978=100) Public sector deficit as share of GDP Real interest rates

1980

1988

All LCDs

All LCDs

‘Strong reform programmes’**

‘Weak or no reform programmes’**

104

78

68

82

10.2

7.2

6.1

8.0

−7.9

0.2*

8.6

−6.5

20

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

1980

1988

All LCDs

All LCDs

‘Strong reform programmes’**

‘Weak or no reform programmes’**

Agricultural 96 125* 146 108 export prices (index 1980– 82=100) (Africa only) Sources: Interest rates, World Bank (1989b) World Development Report 1989, Figure 4.2; exchange rates, ibid., Figure 1.7; agricultural prices, World Bank (1989b), Table 19, p. 29; public sector deficit, World Bank 1988, Figure 3.3, p. 61. Notes: *Data for 1986. **For the first three indicators a ‘strong reform programme’ is defined by the existence of a World Bank policy-based lending programme in that country.

characterized, in the World Bank’s view, by ‘weak or no reform programmes’, the real effective exchange rate has declined since 1980, on-farm prices for agricultural exports have risen and real interest rates have risen. As a broad generalization, policy adjustment has gone furthest in respect of those policy instruments that can be manipulated within the central bank or the ministry of finance without any need for legislative sanction or complex institutional planning, and to which political barriers are therefore smallest—in other words, public expenditure and the exchange rate. All nations, therefore, have adjusted; but each nation has adjusted in its own way, reflecting its own political imperatives and social priorities as well as the variations in the shocks to which they have been subjected. At this point a major difference between stabilization and structural adjustment becomes apparent: whereas it is common ground between all parties that aggregate demand needs to be reduced (whether through cutting expenditure or raising taxes) if an economy’s balance of payments or inflation problem is to be stabilized, there is no common ground concerning the right way to increase aggregate supply for those wishing to embark on a structural adjustment programme; there is no analogue, either in the World Bank or anywhere else to the IMF’s ‘Polak model’, which prescribes the amount by which domestic, absorption must be reduced if the balance of payments is to be improved by $X. There is a presumption, backed by some empirical analysis in the World Development Report 1983, that few market distortions are better than many, but there is nothing in economic theory to tell us which distortions should be removed in which order, or even that it will necessarily help if any one distortion is removed. This territory is still ruled, after thirty years, by the Theory of the Second Best (Lipsey and Lancaster 1956), which warns that in an economy characterized by many market imperfections there is no presumption that the removal of any one such imperfection will necessarily ‘make things any better’, that is, take the economy closer to its production possibility frontier. Any ‘structural adjustment programme’, that is, programme to remove a cluster of such imperfections, is therefore not an application of economic principles, but rather an improvization, a gamble based on the premiss that if past micro-economic policies have yielded unsatisfactory results, an alteration of those policies may help. Perhaps for this reason, the structural adjustment programmes actually implemented around the Third World have shown extra-ordinary diversity, and the World Bank, which has been responsible for designing a good number of them in conjunction with the economic ministries of less-developed countries, certainly escapes the censure that has often been levelled at the Fund’s stabilization programmes, of imposing identikit packages on different countries without consideration of differences in economic structure or capacity to

STRUCTURAL ADJUSTMENT: 1980–9

21

adjust. It is necessary, of course, to make a distinction between programmes as designed and programmes as implemented, and in Table 1.2 we set out, in the left-hand part of the table, the frequency with which different types of structural adjustment reform were advocated by the Bank and, in the right-hand part, the nature of the reforms recommended and implemented in a sample of nine countries recently studied by a team led by the author (Mosley, Harrigan and Toye 1991: Vol. 2, case studies). Looking down the left-hand column of Table 1.2, the alert reader will immediately notice something missing: where is the exchange rate, long cited by the World Bank as the key price that has to be set right in the micro-economy of developing countries? Answer: the IMF is supposed to have sorted it out beforehand. By a division of labour between the two Bretton Woods institutions reconfirmed in 1966 (Feinberg 1986), the IMF takes primary responsibility for advice on the exchange rate, as a key instrument of stabilization; the difficulty, as we have moved into the age of structural adjustment, is that it is also a key instrument of resource allocation, which must be at a competitive level if the Bank’s other measures are to be expected to work. In principle, this difficulty has been sorted out by requiring any developing country to have an IMF standby agreement in position before it is able to negotiate with the World Bank for a structural adjustment loan (SAL), and Table 1.1 testifies to the Fund’s success in bringing this element in the adjustment process into being, but there have been a good number of cases (Jamaica and the Philippines between 1981 and 1985, for example) where for various reasons the Fund, having concluded a standby, allowed the real exchange rate to appreciate in the middle of a Bank structural adjustment programme, thereby provoking the need for a second stabilization episode, which temporarily terminated all adjustment on the supply side. The second thing to be noted is that not all structural adjustment measures which the Bank asked for have actually been implemented - in its own internal evaluation (World Bank 1988, Table 4.1) the Bank estimates the percentage at 60 per cent, and our own country case studies, which are intended to be a balanced sample, range from 25 to 90 per cent and average at less than half. This draws attention to two fundamentally important features of structural adjustment.

Table 1.2 Types of policy measure requested and implemented by the World Bank in return for programme finance, 1980–8

22 FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

Source: For left-hand column, Mosley (1987), table 2; for right-hand columns, case studies in Mosley, Harrigan and Toye (1991). Vol. 2.

STRUCTURAL ADJUSTMENT: 1980–9 23

24

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

The first is that, for a variety of reasons (less precise performance criteria; less power held by the Executive Board; greater pressure to disburse), the World Bank has weaker leverage over the structural adjustment process than the IMF has over the stabilization process. If an IMF performance criterion is not met, disbursement of the loan is suspended; if a World Bank performance criterion is not met, nearly always disbursement of the loan proceeds. The second is that the political odds are stacked much more heavily against the implementation of structural adjustment than the implementation of stabilization. Most stabilization instruments (exchange rate, public investment programme, money supply) can be changed in an ‘afternoon by one official in the central bank or ministry of finance’, whereas ‘rationalize the structure of protection’ or ‘make arrangements to privatize the Maize Marketing Board’ (both typical conditions of structural adjustment programmes) takes a lot longer and involves many more people; in addition, the political threat posed by structural adjustment measures is much more direct than that posed by stabilization measures, since the costs of the latter are scattered across the entire population whereas the costs of the former are borne by specific holders of privileges (‘rent-holders’ in the World Bank jargon: local industrialists who lose from the abolition of import quotas, rich maize farmers who lose from the abolition of maize movement restrictions) who often have considerable political power. If these two considerations are set side by side, policymakers in developing countries have a strong temptation to take advantage of the World Bank’s bargaining weakness in order to avoid arousing political opposition at home, which they can do by negotiating a structural adjustment loan and then only complying with selected ‘harmless’ conditions. The financing of structural adjustment, then, is a game in which the recipient has considerable bargaining power, and in the circumstances what is remarkable is not how much slippage there has been, but how much policy adjustment there has been (Table 1.1) in spite of the formidable political barriers. What is clear after ten years of experience, however, is that there has been massive variation in implementation experience both between countries and between types of policy. As consideration of bargaining theory might lead us to expect, implementation levels tend to be higher in those countries where economic crisis was most acute, the recipient’s financial dependence on the donor was greatest and the donor’s aid programme was relatively small—in short, where the recipient’s bargaining programme was weakest (Mosley, Harrigan and Toye 1991, Chapter 5). But considerations of domestic politics were also vital: in particular, the successful carrying through of any structural adjustment programme required the technocracy in the central bank and the ministry of finance to build an alliance in support of the programme, first of all with their own line agencies (ministries of agriculture, commerce and industry, for example), who were most vulnerable to lobbying by aggrieved rent-holders, and beyond that with the industrialists and farmers’ representatives directly affected. So powerful are the political barriers to this process, as previously described, that the number of countries where we can truly speak of a ‘coalition in support of structural adjustment’ can be counted on the fingers of one hand: Turkey, Mauritius, Philippines under Aquino for a couple of years. Elsewhere, the technocrats, assuming that they were unable to impose their will by force (as in Chile), had to fight a somewhat lonely battle, adjusting those instruments that they could (classically the exchange rate and interest rates) but often being forced to accept defeat in respect of those instruments that did lie under their direct control: import quotas, privatization, composition of the public investment programme. (This ‘politically biased’ pattern of structural adjustment appears clearly in the final column of Table 1.2.) For reasons that we shall develop in a later section (Inter-country variations in experience; see p. 39) this mattered less than it might have done.

STRUCTURAL ADJUSTMENT: 1980–9

25

EFFECTIVENESS In assessing the effectiveness of structural adjustment programmes we confine ourselves to those sixty or so countries that have conducted those programmes with the World Bank, ignoring that significant number (for example, India) that have implemented major supply-related reform programmes without involving the Bank. In evaluating structural adjustment efforts, we are in essence comparing what actually happened in a given country with our best guess of what would have happened in the absence of structural adjustment policies. Since the latter is hypothetical, there is always room for honest disagreement regarding what such policies achieved. In principle, there are three possible ways of comparing the with-policy and withoutpolicy situation: comparisons of ‘structural adjustment lending’ countries with a control group of countries which are similar in all respects to the first group except the specific supply-side policies which they implemented; cross-section regressions in which the structural adjustment effort features as one among several independent variables influencing growth; and simulation exercises on econometric models of individual countries. For reasons of space, our discussion of the third approach here will be minimal, but there is a report on results from the modelling approach in Chapter 8 of Mosley, Harrigan and Toye (1991). We shall also confine our discussion to those performance indicators stressed by the World Bank—growth and trade performance—and ignore the important distributional effects of structural adjustment. Table 1.3 compares the levels of GNP growth, export growth, balance of payments, investment, consumption and foreign private capital inflow—five of the main targets of World Bank structural adjustment lending programmes—as between twenty countries that received World Bank structural adjustment programmes and a ‘control group’ of developing countries that did not receive adjustment loans but which in other respects (income level, prior growth rate, trend in terms of trade) were selected to be as similar as possible to the adjustment lending group. We observe (without any comment at this stage) that the balance of payments performance of adjustment lending countries is substantially better than the performance of the control group, the performance of GNP and export growth trivially better (average export growth being negative in both groups), but the performance of investment is a good deal worse. Table 1.3 Comparison of economic performance in adjustment lending and non-adjustment lending LCDs, 1982–6 Performance criteria

Average value in 20 ‘adjustment lending’ countries, 1982–6

Real GNP growth (% p.a.) 1.7 [2.95] Consumption as percentage of GNP 86.7 [5.03] Investment as percentage of GNP 17.0 [4.0] Balance of payments deficit on current 4.0 [4.6] account (% of GNP) Real export growth (% p.a.) −0.4 [6.2] Figures in brackets are standard deviations Source: Mosley, Harrigan and Toye (1991), Chapter 6.

Average value in 20 ‘non-adjustment lending countries’ 1982–6 2.0 [2.12] 81.4 [11.4] 20.5 [7.9] 6.8 [5.6] −1.3 [7.5]

We now move to the regression approach. Table 1.4 shows the results of an OLS regression conducted across a sample of nineteen countries receiving World Bank SALs for the seven years 1980–6 (133 observations), in which the dependent variables are the targets that the Bank has declared to be the objectives of structural adjustment (GDP growth, export growth, investment and the inflow of private foreign finance) and the independent variables are the finance provided by and the implementation of the policy conditions attached to World Bank adjustment loans, plus a cluster of additional independent

26

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

variables which may also be expected to exert an influence on growth (weather, terms of trade, and IMF finance). Once again, we discover that the influence of structural adjustment on exports and GNP growth appears to be mildly positive, and on investment negative. But from this exercise we also discover that the influence of SAL money on its own appears to be negative, whereas the implementation of SAL conditions has a positive (but lagged) effect. The influence of structural adjustment programmes on inflows of private foreign investment, which it was intended to stimulate, is statistically insignificant. Finally, let us bring together the results of these two exercises with some results from our own modelling work on Malawi and from the World Bank’s own internal review of adjustment lending (World Bank 1988). The results show a remarkable consistency (see Table 1.5). We would draw attention in particular to two implications of the results given in Table 1.5. First, the immediate negative effect of World Bank loan finance, coupled with the positive effects of implementation of loan conditions, suggest that such finance, far from acting as a leverage mechanism to encourage structural adjustment reforms, may in a large number of countries have simply reduced the immediate pressure to adjust by delaying the moment when radical measures have to be undertaken (see also Mosley, Hudson and Horrell 1987); but when they were undertaken they seem on balance to have helped. Second, however, there is the worrying evidence, at a time when attention is being switched from concerns of ‘adjustment’ to those of ‘sustainable growth’, that structural adjustment programmes have acted to reduce investment, and hence long-term growth prospects. Hypotheses concerning the causal mechanisms at work are various (adjustment lending is not project-tied, and hence can more freely be diverted to consumption; adjustment lending is limited to stabilization programmes, of which the most faithfully carried out part has been cuts in public investment;

Source: Mosley, Harrigan and Toye (1991), Chapter 7, Table 10. Note: ** denotes significance of a coefficient at the 5 percent level and * at the 5 percent level. Key: IMF=drawings of IMF finance as a percentage of GDP SAL=SAL and SECAL finance as a percentage of GDP CI=percentage implementation of policy conditions set by World Bank on SALS and SECALS W=weather index (rainfall in capital city for years stated as percentage of 25 year index 1961–86) TOT=terms of trade index (1980=100) EPI=export price index (1980=100) INV=investment as a share of GDP gEX=growth rate of exports.

Table 1.4 Results of regression analysis: all SAL countries

STRUCTURAL ADJUSTMENT: 1980–9 27

28

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

Table 1.5 Effectiveness of structural adjustment: summary of results [Indicators] Methods

Real GDP growth Real export growth Investment Balance of payments Foreign finance

1 Tabular comparisons with ‘control group: (a) Mosley, Harrigan & neutral/weak +ve +ve Toye 1990 (b) World Bank 1988 neutral +ve 2 Multiple regressions weak +ve +ve 3 Single country N/A neutral/weak +ve simulations (Malawi, Morocco) Sources: World Bank (1988); Mosley, Harrigan and Toye (1991).

−ve

+ve

N/A

−ve −ve N/A

+ve +ve −ve

N/A neutral N/A

private investment, as revealed by Table 1.4, has failed to respond to the intended stimulus of SALs, possibly because the private sector expects that any changes implemented under a structural adjustment programme will be soon reversed). Whatever the reasons for it, the fact of declining investment levels is worrying, more particularly because we shall proceed to argue that in the poorest countries a boost to public investment is the key element in a correctly designed structural adjustment programme. INTER-COUNTRY VARIATIONS IN EXPERIENCE The findings reported in the previous section have been aggregative; the high standard deviations of Table 1.3 provide a warning of how much adjustment experiences varied between countries. Just as Lance Taylor’s recent study (1988) for the World Institute for Development Economics Research (WIDER) has reminded us that there are Varieties of Stabilisation Experience, so even the most casual review of what happened in different developing countries during the 1980s makes us aware that the stimuli listed in Table 1.2 produced enormously diverse responses. Again, we confront the point that structural adjustment is not a standard and reliable cure for a clearly identified disease, but an improvisation, a collection of measures thrown together in the hope of inducing greater supply-side efficiency in a wide variety of economies. In this section we speculate in summary manner, drawing on the case studies in Volume 2 of Mosley, Harrigan and Toye (1991), on some of the factors, over and above the political variables discussed earlier (see ‘The implementation of “structural adjustment” policies’), which may have differentiated the effective from the ineffective programmes. Three such factors, which stand out as potential influences on the effectiveness of structural adjustment, are: (a) the response of domestic and overseas investment to the structural adjustment programme; (b) the behaviour of real wages and, as a consequence, the size of the competitive advantage conveyed by devaluation; (c) whether or not the structural adjustment exercise was interrupted mid-term by a stabilization episode.

STRUCTURAL ADJUSTMENT: 1980–9

29

Figure 1.2 Relation between slippage and growth of GDP, with possible explanation of ‘outliers’ Source: Mosley, Harrigan and Toye (1991), Vol. 2, Chapter 10.

Figure 1.2 graphs these factors onto the relationship between implementation of structural adjustment measures examined in the previous section. In simple two-variable form this becomes:

Let us now see whether the variables mentioned above explain any tendency for countries to perform better or worse than this regression line might predict. Malawi, Ghana and Jamaica, of our case-study countries, lie very close to the regression line, and so can be disregarded for this purpose. The status of the others in relation to our chosen indicators is as set out in Table 1.6. The indications set out in Table 1.6 should not be mistaken for econometric evidence (which cannot be provided with a sample of only nine countries); nonetheless, it does appear at least as though the less-thanexpected performance of foreign investment inflows and the unexpected stabilization episodes may have something to do with the lower-than-predicted growth rates observed in Guyana and the Philippines. But Kenya’s structural adjustment process was inter KEY Variables in regression: Levels of implementation of World Bank SAL/SECAL measures, from case studies in Mosley, Harrigan and Toye (1991). (N.B. ‘Slippage’ is 100% less ‘level of implementation’ as measured here.) Growth rate of GDP: annual rate of growth of real GDP between first year of World Bank adjustment lending (three-year moving average) and fifth year after the date (three-year moving average), data from

30

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

World Bank, World Development Reports as summarized in Mosley, Harrigan and Toye (1991), Appendix to Chapter 6 Possible explanations of ‘outliers’: negative real growth in private foreign investment during five years after first World Bank adjustment loan: data from World Bank World Development Reports as summarized in Mosley, Harrigan and Toye (1991) positive growth in real manufacturing wages during five-year period after first World Bank adjustment loan as measured in World Bank, World Development Report 1989, Appendix, Table 7 stabilization episode involving IMF during five years after first World Bank adjustment loan Table 1.6 Case-study countries performing better and worse than predicted by policy change/growth regression line Unfavourable indication in relation to: Private foreign investment Real wages Stabilization episode during structural adjustment period Countries performing ‘better than predicted’ Turkey Thailand Yes Kenya Yes Countries performing ‘worse than predicted’ Guyana Yes Yes Philippines Yes Yes Source: Mosley, Harrigan and Toye (1991). Note: a rise in real wages during the adjustment period is treated as an ‘unfavourable indication’.

rupted by a stabilization episode, which it managed to shrug off; nor did a rise in real wages in Thailand manage to prevent its becoming one of the most successful of all developing economies, with double-figure growth rates of both GDP and exports, during the second half of the 1980s. At the very least, therefore, falling real wages do not constitute a necessary condition for the success of a structural adjustment programme. We would also draw attention to the case of Guyana, by some measure the poorest performer within our sample. Jane Harrigan’s case study of this country (Mosley, Harrigan and Toye 1990, Chapter 18) makes it clear that this fiasco is by no means solely due to its failure to implement the policy advice of the World Bank, or even to the collapse of public expenditure and foreign investment that occurred during the stabilization episode. Rather, in a country where infrastructure and the export sector’s capital stock had decayed, ‘structural adjustment’ should have taken the form of rehabilitation of the productive structure, rather than privatization and liberalization. As one of the World Bank’s most percipient senior economists has acknowledged: Conventional text-book economics is not written for economies in decline, but for static or growing economies. One can structurally adjust an economy which is growing but growing inefficiently, or one that is static. But an economy in cumulative decline forces one to confront a systemic problem. Such an economy requires transformation of a Keynesian type, in which the emphasis is on the

STRUCTURAL ADJUSTMENT: 1980–9

31

quantity and quality of investment, particularly the stimulative role of public sector investment. The Bank’s structural adjustment programmes however are rooted in the marginality theories of neoclassical text-book economics with their emphasis placed on price incentives, exchange rate adjustments, and trade liberalisation. (cited in Mosley, Harrigan and Toye 1991, Chapter 18, p. 28) The importance of this observation is that it is not, of course, the economy of Guyana alone that is in ‘cumulative decline’, but of many of the world’s poorest countries, including much of sub-Saharan Africa. Mozambique, Sudan, Ethiopia, Somalia, Sierra Leone, Guinea, Zaire, even Nigeria: all of these, as much as Guyana, require ‘transformation of a Keynesian type’, plus state intervention to make land and smallholder credit available to those who need them, rather than stimuli for a ‘private capitalist sector’, which in many of the poorest countries scarcely exists. Our reading of the structural adjustment experience of the poorer developing countries, then, leads us to take inspiration from the Latin-American structuralists of the 1950s, who advocated that the state should promote development by itself removing bottlenecks in the economy— the left-hand branch of the two structural adjustment options in Figure 1.1—rather than from the ‘new structuralist’ experiments of the 1980s. SYNTHESIS AND HISTORICAL CONTEXT The episode of reforms known as ‘structural adjustment’, then, have been implemented initially across the entire developed world during the 1980s, but in a patchy manner, which reflects the fact that exchange rate and public expenditure reforms, being administratively simpler and politically less visible, are easier to implement than liberalization and privatization. Where implemented, they appear to have raised income and exports, but to an extent that varied across countries according to the response of private foreign investment and the incidence of emergency stabilization measures. The liberalizing brand of structural adjustment fashionable in the 1980s appears to have more relevance to richer than to poorer LDCs and least of all to those that are in absolute decline. It remains to ask how this episode fits into its historical context. Gerschenkron (1959), comparing the industrialization patterns followed by successive countries before 1914, observed that each such process appeared to be more heavy industry-oriented, more monopolistic and more heavily state-protected than the last; but, as has frequently been noted (for example, Hirschman 1968; Ranis 1990), this trend towards progressively increasing state intervention has been bucked by the ‘late late industrializers’ of the twentieth century, many of whom (classically Hong Kong and Singapore) have been remarkably sparing in their use of industrial protection. In an attempt to make sense of the complex reality of post-colonial industrialization, Fei and Ranis (1988, pp. 15–23) have argued that the general pattern is for an initial inward-looking (import-substitution) phase (artificially prolonged in Latin America) to be succeeded by a trend towards progressively increasing liberalization and export orientation, as the economy becomes more complex and less manipulable by a few powerful rent-holders. This long-term trend will be interrupted by exogenous shocks provoked by the world business cycle, primary product prices, etc. Where does the structural adjustment episode of the 1980s fit into this long-term pattern? What immediately becomes clear is that the appropriateness of packages such as those set out in Table 1.2 depends entirely on the stage of development that the recipient country had reached. In countries such as Turkey and the Philippines, the World Bank was able to achieve spectacular success by catching the point of transition from the inward-oriented to the export-oriented phase, at a point when an internal political change had made it unnecessary, and indeed undesirable, for the state to propitiate most of the rent-

32

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

holders from the previous ancien regime. In countries such as Kenya and Malawi, it tilted at windmills by asking for liberalization at a point when the country was still, and rightly, in its inward-oriented phase, and needed protection for its infant activities both industrial and agricultural before a switch to export-led growth could make any policy sense. In countries such as Guyana and Ghana in the early 1980s, caught in a process of systemic decline, ‘structural adjustment’ in the liberalizing sense can only be described as a historical irrelevance, but many elements of the Bank’s approach to Ghana in 1983–5 showed that the Bank had learned to appreciate the need for a phase of capital stock rehabilitation in advance of any changes in the structure of industrial protection or ownership. In the early 1980s one of the architects of structural adjustment policy at the World Bank described the structural adjustment process as a necessary item of consumption for all less-developed countries, designed to confront ‘the central problem of development at present’ (Stern, 1983, p. 107). The experience of the 1980s has taught us to see structural adjustment in the Bank’s sense, rather, as a somewhat specialized, and on balance a middle-income, item of consumption. REFERENCES Fei, J.C.H. and Ranis, G. (1988) ‘The political economy of development policy change: a comparative study of Thailand and the Philippines.’ Unpublished paper. New Haven, Conn.: Economic Growth Center, Yale University. Feinberg, R. (1986) The Changing Relationship between the World Bank and IMF. Washington DC: Overseas Development Council. Unpublished paper. Gerschenkron, A. (1959) Economic Backwardness in Historical Perspective. Cambridge, Mass.: Harvard University Press. Hirschman, A. (1968) ‘The political economy of import-substituting industrialization in Latin America’, Quarterly Journal of Economics, 82:1– 32. Lipsey, R.G. and Lancaster, K. (1956), ‘The general theory of second best’, Review of Economic Studies, 24:11–32. Mosley, P. (1987) Conditionality as Bargaining Process: Structural Adjustment Lending 1980–6. Princeton NJ: Princeton Essays in International Finance, no. 168. Mosley, P., Harrigan, J. and Toye, J. (1991) Aid and Power: The World Bank and Policy-Based Lending in the 1980s, London: Routledge. Mosley, P., Hudson, J. and Horrell, S. (1987) ‘Aid, the public sector and the market in less developed countries’, Economic Journal, 97:616–42. Ranis, G. (1990) ‘Asian and Latin American experience: lessons for Africa’, Journal of International Development, 2 (April). Stern, E. (1983) ‘World Bank financing of structural adjustment’, in J. Williamson (ed.) IMF Conditionality, Washington DC: Institute for International Economics/MIT Press. Taylor, L. (1988) Varieties of Stabilisation Experience: Towards Sensible Macroeconomics in the Third World, Oxford: Clarendon Press. World Bank (1983) World Development Report 1983, Washington DC: World Bank. (1988) Report on adjustment lending. Report R88–199. Washington DC: World Bank, Country Economics Department. (1989a) Africa’s Adjustment and Growth in the 1980s. Washington DC: World Bank with UNDP. (1989b) World Development Report 1989, Washington DC: World Bank.

2 An overview of trade policy reform, with implications for Sub-Saharan Africa John Nash

Trade policy reform, because of its economy-wide effects on relative prices and efficiency, has often been a key component of the first phase of adjustment. Such reforms, when implemented well, have contributed to improved economic performance in developing countries. Nonetheless, both the implementation and outcomes of reforms have been less successful than might have been hoped for, given their prominence in adjustment policies and World Bank lending. There remains substantial room for improvement. Through the insights provided by economic theory and the experience of many developing countries that have attempted to implement reforms, a number of lessons can be derived to improve implementation and performance. EXPECTED GAINS FROM TRADE POLICY REFORMS Most of the discussions of the gains from reforms have focused on the once-and-for-all increases in GDP that come from reducing the static costs produced by resource misallocation associated with the restrictions (see, for example, Corden 1974). In a static and partial equilibrium framework, direct costs coming from the misalignment of domestic and international prices are generally estimated to be a few per cent of gross domestic product (GDP) a year or less. The costs are much greater, however, when the likely effects on market structure are also considered (Condon and de Melo 1986; Bergsman 1974). Costs are further increased when protection is provided by non-tariff barriers, which inevitably encourage rent-seeking activities (Grais, de Melo and Urata 1986; Krueger 1974; Mohammad and Whalley 1984). The dynamic effects of reforming the trade regime (e.g. increasing the long-run rate of growth of technical change, productivity, and incomes) are more difficult to demonstrate than the static costs, but are of great practical importance (Edwards 1989a; Chenery, Robinson, and Syrquin 1986; Grossman and Helpman, 1989a, 1989b; Romer 1989). Trade and domestic restrictions also have macroeconomic implications, which usually become most visible when a country faces severe external shocks, such as higher oil prices or commodity booms and busts. Delinking the domestic economy from world prices contributed to poor adjustment performances in Nigeria, Côte d’Ivoire, Senegal and Kenya in response to commodity price cycles (Bevan, Collier, and Gunning 1987; Devarajan and de Melo 1987; Pinto 1987). Concerns regarding trade policy reforms Arrayed against arguments in favour of initiating reforms are the arguments that protecting high-cost ‘infant industries’, usually import-substituting industries, are justified by the economic benefits that eventually flow from these industries once they become efficient. However, experience has shown that protected infant industries often fail to grow up. The World Bank has generally found that industries and firms that are

34

TRADE POLICY REFORM: AN OVERVIEW

inefficient received high protection for relatively long periods, while those that are efficient (notably exporting industries) received relatively low protection and incentives in earlier periods. Industrial performance sometimes deteriorates over prolonged periods of insulation from world markets as a result of protection policies (for example, the steel, glass, and many engineering industries in India). Furthermore, infant industry protection creates a bias against exports and agriculture, both of which have the characteristics used to justify protection for infant industry, such as the ability to reduce costs by ‘learning while doing’ and imperfection of capital markets. It is not clear then, why policies should be followed that protect import-substitutes while dis-protecting export infants. One variant of the infant industry argument concedes the superiority of neutral policies with no antiexport bias in the long run. But it posits that in early stages of development, import substitutes should be protected so that the economy can develop the infrastructure, management skills, human capital and efficient production techniques that can eventually help export industries compete in world markets. There are, however, reasons to doubt that export success is made easier by the earlier import-substituting (and therefore anti-export) policy bias. For one thing, the techniques of production chosen, as well as the industries that develop, under an import-substituting regime are likely to be inappropriate for exports. Overvalued exchange rates, escalated tariff structures and subsidized interest rates create incentives for production that is intensive in imported capital goods and intermediates, rather than in the domestic factors on which competitive exports must be based (Bruton 1970). Linkages with the rest of the economy are poor and become weaker over time, as Killick showed in pre-reform Ghana (Killick 1989). One reason why so many industries in Latin America and Africa have required extensive restructuring as part of adjustment packages is inappropriate production techniques induced by protectionist policies (Sood and Kohli 1989). Second, the physical infrastructure needed in an open economy -where much of the transportation service is to and from the borders— is very different from that needed in a more closed economy. To take one example, the limited and run-down port facilities in Mexico were not much of a problem until the new outward orientation greatly increased the volume of traffic. Third, the skills developed in a protectionist regime are not those needed in a more open environment. In the former, fortunes are made and lost because of lobbying skills and political contacts. A former Under Secretary of the Economy in Argentina is quoted as noting that ‘it is more profitable to spend time in these corridors [of the Ministry of the Economy and the Central Bank] than in the manufacturing plant’ (Nogués 1989b). Such skills may not be readily transferable to the practices that make for an efficient exporter. Finally, import-substituting policies are likely to create inappropriate demographic patterns. Protected industries—even agroindustries that would otherwise locate near their sources of supplies—are led to locate in urban areas, and the labour supply follows. When protection (and the accompanying anti-agricultural bias) is reduced, this rural-urban migration must be reversed, at great human cost. Evaluating this proposition based on countries’ experiences is difficult. There are certainly countries that have successfully made the transition from import-substitution to export-orientation. However, these experiences cannot be compared with those of many countries that have followed export-oriented policies from early stages of development, since there are few if any such countries outside east Asia (e.g. Hong Kong, Singapore). Thus, it is not clear whether the eventual export success of other countries was achieved because of, or in spite of, the early protectionist policies. But the specific arguments laid out above, and the large number of countries that have not yet successfully made the switch, argue that countries may be better off if they begin to follow neutral, outward-oriented policies and develop export industries as soon as possible, instead of trying to change in this direction at a later stage of development. Some recent literature has developed theoretical support for strategic trade policy, that is, providing protection rather than reducing it for certain key industries, based on the excess profits they are likely to

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

35

earn in oligopolistic global markets (Spencer and Brander 1983; Krugman 1987b) or the external economies they create for other firms (Kemp and Negishi 1970; Panagariya 1980). However, it is hard to target these interventions successfully in practice. For this and other reasons, even the theorists that developed this literature acknowledge that the conditions under which the benefits of strategic trade policy exceed its costs are unlikely to be met in practice (Krugman 1987a). (For a detailed account of an unsuccessful attempted application in the Brazilian aircraft industry, see Baldwin, forthcoming). Observers have also raised a variety of other concerns about trade policy reform (see, for example, Sachs 1987; Taylor 1988). First, some assert that trade liberalization may aggravate the balance of payments and fiscal problems that have afflicted many countries in the 1980s. Second, the benefits of liberalization and greater openness are disputed. In addition to the putative benefits of protection to infant industries, some doubt that world trade conditions will allow reforming countries to increase their exports. Third, it is feared that liberalization produces transitional unemployment and that devaluation increases inflation rates. These concerns are discussed in some detail below. At the outset, however, it should be recognized that there is a possibility of conflicts between trade and other reforms. But, the evidence suggests that sequencing of reforms can help avoid these conflicts. The evidence also shows that well-implemented trade reforms improve economic performance (exports and income growth), and that complementary actions will further augment the supply response. While short-term transitional costs are usually expected from resource reallocation, the findings of the Bank’s previous research indicate no clear relationship between trade liberalization and unemployment. REFORM EXPERIENCE AND IMPLICATIONS Overview of developing country trade regimes Trade regimes throughout the developing world are heavily distorted by both tariff and non-tariff barriers. Compared to other regions, Africa tends to have fairly high legal tariffs, but to impose few other surcharges (Table 2.1). Of course, actual tariff collections are much lower than the level of legal tariffs would suggest. This is true because many firms (especially parastatals) have exemptions from tariffs for their inputs, highly taxed imports enter clandestinely or under the guise of lower-duty items, and legal imports tend to be in lowduty capital and intermediate goods. But tariffs do not tell the full story, since in Africa, many tariffs are redundant; a better measure of protection is the coverage of non-tariff barriers (NTBs). By almost any measure of NTBs, sub-Saharan Africa is far more protective than any other region (Table 2.1). While studies in Africa showing the net impact of NTBs and tariffs on effective protection rates (EPRs) are scarce, it appears that protection is high and highly dispersed across sectors. In Zambia, for example, 24 per cent of consumer goods had EPRs over 500 per cent in 1975, while 17 per cent had negative EPRs; for intermediates, only 5 per cent had EPRs over 500 per cent, while 30 per cent had negative EPRs. Table 2.1 Average tariffs and para-tariffs by geographical regions (in percentages)

Tariffs Countries unweighte da

Caribbea n

Central America

South America

North Africa

Other Africa

West Asia Other Asia

All Regions

16

23

34

29

32

7

26

36

36

TRADE POLICY REFORM: AN OVERVIEW

Caribbea n

Central America

South America

North Africa

Other Africa

West Asia Other Asia

All Regions

Countries 17 24 38 30 35 4 22 24 importweightedb Tariffs plus para-tariffs Countries 18 65 46 36 34 9 42 34 unweighte da Countries 17 66 51 39 36 5 25 30 importweightedb Source: UNCTAD computer files based on published official national sources. From: Erzan, R., Kuwahara, H., Marchese, S. and Vosennar, R. undated, The Profile of Protection in Developing Countries, UNCTAD Discussion Paper No. 21, Geneva: United Nations Conference on Trade and Development. Notes: a Simple average across products and countries. b Simple averages across products; across countries averages weighted by total imports.

Policy changes Since 1979 the World Bank, usually in conjunction with the IMF, has supported trade policy reforms through adjustment loans. During 1979–87, an estimated 81 adjustment loans to 41 countries contained substantial proposals for trade policy reform (Table 1.3).1 With nine ‘Intensity of Proposed Reform’ indicators (not mutually exclusive), the proposed reform measures could be considered ‘strong’ in 31 per cent of the cases, ‘moderate’ in 45 per cent, and ‘mild’ in the remaining 24 per cent. Policy proposals for exports and quantitative restrictions were generally stronger than those related to the level and dispersion of tariffs. In Africa, many of the initial structural adjustment programmes were concerned more with issues not directly related to import protection, but dealt with agricultural pricing, duty drawbacks for exporters, fiscal reform, mining sectoral reforms, etc. Later, programmes in some countries began to liberalize import restrictions, though proposals were on average modest compared with initial conditions and to countries in other regions (Halevi 1989).

Source: UNCTAD computer files based on published official national sources. From: Erzan, R., Kuwahara, H., Marchese, S. and Vosennar, R. undated, The Profile of Protection in Developing Countries, UNCTAD Discussion Paper No. 21, New York: United Nations Conference on Trade and Development. Notes: a In calculating the stack total, different NTMs affecting the same product are counted cumulatively; in non-stack total, even if more than one NTM affects a product, it is counted only once. b Excluding NTMs which are across the board.

Table 2.2 Frequency of NTWs by geographical regions: percentage of tariff positions affected (countries in each group import weighted)

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY 37

38

TRADE POLICY REFORM: AN OVERVIEW

Reform implementation by trade loan recipients varied substantially across countries and policy areas.2 In some countries, there was little progress in reforms—or reforms were reversed—during 1979– 87 (Guyana, Yugoslavia, Zambia and Zimbabwe); there was modest progress in others (Bangladesh and Madagascar); and there was considerable reform in others (Chile, Mexico and Turkey). A few countries, notably Bolivia and Haiti, also substantially reformed their trade policies during 1979–87 without World Bank support. Substantial progress has been made in correcting misaligned exchange rates (Figure 2.1). In Africa, however, the progress has been slower. An index of the real effective exchange rate in sub-Saharan Africa with the years 1971–2 as a base of 100 appreciated to 125 in the mid-1980s before beginning to depreciate. By 1987, it had finally declined to a level below 100. In the meantime, Asian and Western hemisphere developing countries’ rates had depreciated greatly since 1971–2 and by 1987 were at levels 50–60 per cent of the base year. Average parallel exchange rates in Africa reached almost 250 per cent of the official rates around 1983, and by 1985 were still one-third higher. Recently, devaluations have made some African countries more internationally competitive (Tanzania, Ghana, Guinea), although the CFA franc countries’ exchange rates remain overvalued. In addition to correcting misaligned exchange rates, progress in many countries has been made in reducing impediments to exports, including restrictions on imports needed by exporters. On the import Table 2.3 Major trade policy reform proposals among countries receiving World Bank trade adjustment loans, 1979–87 Area of reform proposal

Present

Not present

Strong

Moderate

Mild or absent

Exchange rate 38 2 Export promotion 33 7 Protection studies 28 12 Intensity of reform proposals Overall export policy 15 15 10 Imported inputs used in exports 17 15 8 Overall import policy 14 15 11 Protective quantitative restrictions 14 15 11 a Non-protective quantitative restrictions 14 16 10 Tariff level 7 21 12 Tariff dispersion 7 24 9 Schedule of future action 6 29 5 Overall reduction in anti-export bias 17 12 11 Percentage of total 31 45 24 Note: a Non protective quantitative restrictions refer to controls on items that are not domestically produced.

side, several countries have substituted tariffs for quantitative restrictions, but seldom with protection reduced by much. Reductions of both quantitative restrictions and tariff levels have been more modest, although some countries have made good progress in this regard, among them Bolivia, Costa Rica, Chile, Ghana, Korea, Mexico and, until recently, Turkey. In nearly all these countries, the bias against trade, and exportables in particular, declined in part as a result of the reduction in import protection. For most other countries, however, the data to evaluate with any precision reductions in anti-export bias are lacking. For the recipients of trade adjustment loans, the measured ratios of imports to GDP (in current and constant

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

39

Figure 2.1 Real exchange rate indices, 1978–88 (unweighted averages for country groupings; 1980=100) Source: Based on IMF data 21 INDs=twenty-one industrial countries 40 TALS=forty trade adjustment loan recipient countries 47 NTALs=forty-eight non-recipients of trade adjustment loans Note: Increase in index indicates a real appreciation a. Multilateral index of the real exchange rate measured against a basket of currencies of trading partners

prices) have risen relative to those for non-recipient countries, indicating the influence of both increased financing and some import liberalization. Policy changes and outcomes have been more extensive in trade policy than in most other areas of adjustment lending, although reduction in protection levels has been slow. Nonetheless, four sets of domestic factors constrain stronger and more sustained reforms. First, vested interests against reform and inadequate convictions concerning its benefits have combined to weaken commitment to reform (Kenya, Peru, Zimbabwe, Yugoslavia). Second, inadequate implementation capacity resulting from administrative and institutional bottlenecks have contributed to implementation setbacks (Bangladesh, Côte d’Ivoire, Malawi). Reforms to strengthen public sector institutions have received inadequate attention. Third, weak macroeconomic performance and conflicts between policy reform and stabilization goals have sometimes slowed (Morocco, the Philippines, Malawi, Kenya) or reversed (Argentina, Zambia) trade liberalization. Fourth, lags in supply response to policy reform have, by reducing the apparent benefits, limited the enthusiasm for reform, especially in the low-income countries of sub-Saharan Africa.

40

TRADE POLICY REFORM: AN OVERVIEW

Figure 2.2GDP and export growth for developing countries, 1965–88 Source: Based on World Bank data GDP=average GDP growth rate during 1965–88 X=average growth rate of exports of goods and nonfactor services during 1965–88 LL=fitted line based on least square regression

Performance outcomes If countries are to compete in a rapidly modernizing international economy, they need to increase their efficiency, become flexible in their production structures, and adapt to changing global demands. Experience in the 1980s corroborates the positive association between exports and economic growth found in previous work. Since exports are usually relatively labour-intensive, this is also expected to be associated with employment generation. Without implying causality, there is a positive correlation between export and output growth (Figure 2.2). But tracing the influence of specific policies behind superior export and GDP growth is complex, however, because of the simultaneous presence of other contributing factors. Policy reforms can also take time to produce the expected improvements in resource allocation, efficiency and growth. Bearing these caveats in mind, the adjustment episodes of the 1980s and longer-term experience suggest that trade policy and structural reforms have contributed positively to the growth in output and exports. Real exchange rate depreciation and commercial policy reforms are linked to such improved performance. Both additional financing and policy reforms connected with adjustment lending have contributed to relative improvements in the 1980s. After controlling for some other factors, trade adjustment lending is associated with a mild improvement in GDP, exports, and other variables. This improvement is stronger and statistically more significant among early and intensive loan recipients (those receiving three or more

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

41

adjustment loans with a substantial trade component). Accordingly, the results measured for all recipients are likely to understate the benefits because they include outcomes of early stages of a continuing process for some. The results are also stronger and statistically more significant when the comparison is between trade policy reformers and non-reformers rather than simply between trade loan recipients and nonrecipients. The evidence suggests that developing countries have benefited economically from a combination of exchange rate depreciation and import and export policy reform. The stimulus to exports and output from a real devaluation, combined with reductions in export restrictions and measures for export development, are likely to be more immediate than those from a real devaluation and import liberalization (Lopez 1989). At the same time, efficient longer-term development of exports and output depends not only on export policies but also on import liberalization. Thomas (1989) and Heitger (1987) each found a negative effect of import restrictions on growth. In Africa, the response to trade policy and other reforms has varied across countries and across sectors. The response in agriculture, which was severely depressed by overvaluation and export taxation, has been quite good. Countries with favourable policy environments have seen agricultural value added increase by 18 per cent between the periods 1980–2 and 1985–8, while those with unfavourable policies saw value added fall by 2 per cent (Jaeger 1989). Since agricultural exports in many countries was the sector most depressed by past policies, this has also shown the greatest response, but food crop production has also grown (by 12 per cent) in countries with favourable policies. Redressing the historical pro urban bias has increased incentives to return to the farm, with pronounced effects in some countries, such as Nigeria and Ghana. Survey data from Ghana show a 2 per cent increase in the share of the population earning a living in agriculture between mid-1984 and 1988, compared with previous rates of migration to the city of 1 per cent per year. The effect on the manufacturing sector has been more mixed. Trade reforms in general have helped export industries and those that faced constraints because of a scarcity of imported inputs. Highly protected industries—often parastatals—tended to suffer. But the overall effect has been positive (Table 2.4). In some countries, reforms revived the economy so much that even the relatively disprotected industries were not hurt in absolute terms. A survey in Ghana in December 1987 showed that 56 per cent of the firms increased their output following the reforms; only 15 per cent lowered output (Steel 1989). Much of the increase came from recovery of export markets and expansion of existing operations, e.g. in textiles from Ghana and Nigeria to neighbouring countries and cocoa products from Côte d’Ivoire and Nigeria to Europe. Some expansion came from new industries—garments in Madagascar and glycerol from Ghana. The reforms tended to shift the industrial structure from assembly industries, highly dependent on imported inputs, to resource-based industries. The supply response Country studies identify several factors that may contribute to or constrain the supply response. First, growing protectionism in international markets in the 1980s has discouraged trade policy reforms, particularly those relating to agriculture. For manufactured exports, industrial countries represent promising markets because of low tariffs, but non-tariff barriers in some important product categories (for example, textiles, clothing, steel) have hurt developing country export growth. This has not been a serious problem for African producers, who so far represent little threat to domestic industries in industrial countries. Second, the supply response is influenced by perceptions concerning the sustainability of reforms, which depends on the country’s previous track record in policy reform, the effectiveness of the initial steps,

42

TRADE POLICY REFORM: AN OVERVIEW

macroeconomic stability, and consistency with other reforms (financial sector reform and agricultural pricing, for example). Third, insufficient attention to the institutional and infrastructural needs of exporters has often been a problem (weak Table 2.4 Growth in manufacturing output, exports, and capacity utilization before and during reform period Output Before

Exports During

Before

Capacity utilization During

Côte d’Ivoire −1.8(81–83) 5.8(84–86) -6.9 12.3 Ghana −17.1 (80–83) 15.0 (84–87) −10.4 51.3 Nigeria −7.8(82–85) 0.2(86–87) −15.4 18.1 Zambia −3.1 (82–84) 4.0 (84–86) 5.4 7.2 Source: World Bank, World Tables 1988–9, Washington DC, 1989. From: ‘The Saharan Africa: a strategy for recovery and growth’, Box 5.4.

Before

During

– – 19 32 30 57 38 54 long-term perspectives for Sub-

systems for providing duty-free and restriction-free access to imported inputs, inadequate port and transport and telecommunications facilities, poor information and market services for exporters). Fourth, domestic regulatory policies influence the supply response by determining whether incentives actually change in response to reform (price controls) and by affecting the mobility of factors of production in response to changes in incentives (labour regulation, market entry and exit regulations, foreign investment controls). Fifth, some public sector policies are not supportive of rapid adjustment to a changed incentive structure and thus inhibit the supply response (allocation mechanisms in centrally planned economies, parastatal monopolies or domination of agricultural markets, as with many export crops in Africa). Finally, when entrepreneurial and managerial capacities are relatively underdeveloped, the supply response is inevitably slower. Shortages of trained labour and poorly developed input supply lines have also been serious problems in many cases. SEQUENCING AND TIMING OF REFORMS More often than not, countries that are in great need of trade policy reforms also have a number of other policy-induced distortions and rigidities that demand resolution as well. Often, the most pressing of these are macroeconomic imbalances, but they include regulatory policies, price controls on selected products, investment licensing, and other domestic policies. In designing the adjustment programme, the multifaceted nature of the problem raises the issue of how reforms should be sequenced and paced. Should trade reforms be delayed until inflation is brought under control? Once begun, how quickly should the reform package be carried out? Should policies to directly encourage exports precede those to reduce import barriers? These are important questions. But there are no universally applicable answers; the prescription for each country will depend on its initial conditions. There are, however, some general principles. First, in designing the package, one should ask whether some of the needed policy changes would interfere with the trade policy reforms. In practice, the answer will usually be ‘no’, and it is easy to overestimate the conflicts. Nonetheless, in some conditions (discussed in more detail later), conflicts can arise, and some reforms may need to be postponed. The corollary to this is that, in some cases, trade reforms may be highly complementary with certain other reforms, implying that a special effort should be made to implement them simultaneously. Second, when not all reforms can be carried out together (whether because of conflicts,

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

43

political considerations, or shortage of administrative capacity), priorities should be set by locating the areas of greatest distortions. Often, this will be the trade regime, which is why this has been the starting point for many adjustment programmes. But in some countries in Africa, the restrictions on trade from tariffs or nontariff barriers are clearly redundant; the real distortions come in the first instance from exchange restrictions or from direct state ownership or control of trading mechanisms. Issues of sequencing reforms to avoid conflicts and maximize their salutary impact are taken up on this section. Trade policy reforms and macroeconomic stabilization Some analysts have argued that the fiscal deficit and inflation should be reduced before trade policy reform is introduced. One issue is possible inconsistency with stabilization efforts: tariff reforms may reduce revenue, devaluation may fuel inflation, and liberalization may exacerbate balance of payments deficit. In practice, the significant trade policy reformers have also generally managed to reduce the fiscal deficit, inflation and the balance of payments deficit more than have the weaker reformers in the sample (Table 2.5). The substitution of quantitative restrictions by tariffs has increased revenues. Where the fiscal deficit has been sufficiently reduced, the current account deficit has also declined under import liberalization. Devaluation raises domestic prices of tradables and can fuel inflation; but, an adequate reduction in the fiscal deficit has lowered inflation under a devaluation, as in Nigeria and Ghana (Steel 1989). And import liberalization can also dampen inflationary expectations by providing much needed competition in the domestic markets, as in Tanzania and Uganda. A second issue is whether, under macroeconomic instability, trade policy reforms will be productive. Under most situations, they have been effective. The experience of the sample countries in the 1980s corroborates Krueger’s (1981) conclusion that, in general, trade policy reform and stabilization can proceed in parallel successfully. When the real exchange rate is overvalued, reform requires a real depreciation, which can usually be achieved by using a crawling peg, while correcting the underlying fiscal imbalance (Edwards 1989b). Where inflation has been very high and variable, however, Fischer (1984) points out that leads and lags in the movement of individual prices have made the resulting relative prices a poor guide for economic decisions. In addition, if the authorities use the exchange

Source: World Bank data. Note: Extent of reform (1980–7) is based on a combination of changes in policies (high, moderate, or low) with respect to exchange rate depreciation and commercial policy reforms. Countries in each group are as follows: significant (high in both categories or high in one and moderate in the other): Chile, Colombia, Ghana, Jamaica, Korea, Mauritius, Mexico and Turkey; moderate (moderate and moderate, or high and low): Bangladesh, Madagascar, Morocco, Pakistan, Panama, Philippines and Thailand; and mild (others): Côte d’Ivoire, Guyana, Kenya, Malawi, Senegal, Togo, Yugoslavia, Zambia and Zimbabwe. (Mild includes countries that reversed reforms.) a Excludes Mexico, for which changes in operational deficit is a more meaningful measure of fiscal effort.

Table 2.5Macroeconomic indicators before and after reform in trade adjustment loan countries, with implementation data (unweighted average for each group in percentages)

44 TRADE POLICY REFORM: AN OVERVIEW

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

45

rate (instead of adequate macroeconomic policies) as a ‘brake on inflation’, the real exchange is likely to appreciate and thus reduce the effectiveness of trade reform. This has occurred in a number of Latin American countries (Corbo and de Melo 1987; Kiguel and Liviatan 1988). Under these conditions, aspects of trade reform whose effectiveness depends on relative price changes are unlikely to be successful, and they should wait until very high rates of inflation are brought down. Trade policy and fiscal reform Eliminating non-tariff barriers—especially converting them to equivalent tariffs—and eliminating tariff exemptions are revenue-enhancing reforms (Jamaica, Kenya, Mauritius). Tariff reform can be revenueneutral or revenue-enhancing insofar as lower tariff rates are offset by a more depreciated exchange rate. Reducing very high tariff rates can increase trade tax revenue if tariff evasion rates fall or if import demand is price elastic. But, such increase in tariff revenue cannot automatically be relied on. In a sample of countries that primarily reformed non-tariff barriers, tariff revenue increased from 2.7 per cent of GDP to 3. 4 per cent. But, in a sample of tariff reformers, revenue fell on average from 2.8 per cent of GDP to 2.3 per cent. The fiscal effects of a devaluation in general depend on whether the government is a net buyer (Ghana, Sierra Leone, Somalia, Uganda, Zaire) or seller (Nigeria) of foreign exchange. In countries highly dependent on tariff revenues, therefore, revenue effects of specific trade reforms involving tariff reductions should be evaluated before implementation. Measures to reduce expenditure or enhance revenue from other sources (such as consumption taxes) may need to be implemented. Mexico generated additional revenue through tax reform when trade taxes fell; Morocco did not generate new revenues, leading to a partial reversal. Ghana offset losses from devaluation and successfully implemented reforms. Zambia did not offset lost revenue and also increased spending; the fiscal situation got out of control, and reforms were reversed. Trade policy and domestic reforms The benefits of trade policy reforms are usually greater when accompanied by domestic economic reforms. But successful trade reformers have not waited until all complementary domestic reforms and infrastructural investments were in place. In fact, the initiation of trade reforms often exposes the need for domestic reforms and investments (Madagascar, Mexico, Tanzania, Zaire).3 In Tanzania, for example, increased cotton production highlighted inadequacies in transportation that created bottlenecks for cotton exports. With more openness, unforeseen infrastructural demands have arisen as industries developed based almost entirely on foreign demand (fruits, measuring instruments, salmon, and translation services in Chile; cut flowers in Colombia; wigs, cutlery, furs, and colour television sets in Korea.) Sometimes domestic reforms should be deferred until it is made clear to the business and financial communities that import protection will be reduced. Otherwise, for example, if investment or price controls are removed in highly protected sectors, increased investment and production might be encouraged in the wrong sectors. Conversely, trade policy reform in some cases may need to wait until a domestic control is relaxed. For example, a processed product (such as textiles) may become rapidly disprotected if its tariffs are reduced while the price of its basic input (cotton), set by a monopolistic parastatal, remains high. In general, external and domestic reforms are best carried out simultaneously because of their complementary effects. But there are also some circumstances, as noted above, when it is better to introduce them in sequence.

46

TRADE POLICY REFORM: AN OVERVIEW

Sequencing trade policy reforms In case of a substantially overvalued exchange rate, priority ought to be given to measures to achieve and maintain a real devaluation.4 In many countries, a real devaluation has made many quantitative import restrictions redundant and facilitated their removal. The rapid removal of a large number of quantitative restrictions has often been preceded or accompanied by large devaluations (Bolivia, Chile, Ghana, Laos, Mexico, Nigeria, Sri Lanka, Zaire). The shift from commercial policy protection to ‘exchange rate protection’ is a major step toward neutrality in incentives between and among exportables and import substitutes. Introducing other export policy reform shortly before, or at least at the same time as, import reforms permits an earlier export supply response and allows unification of the tariff structure to proceed without burdening exporters. Import policy reform often starts by replacing non-tariff barriers with tariffs providing roughly the same protection. This step, as well as elimination of tariff exemptions, improves resource allocation, increases transparency, reduces rent-seeking, and helps to improve the fiscal situation. These steps ought to be followed by a rationalization of the tariff structure to provide more uniform incentives. This should usually be accompanied or followed by a reduction in tariff rates in order to reduce protection, while any significant adverse effects on the fiscal deficit are offset. Depending on revenue needs, however, non-tariff barriers could be phased out without equivalent tariffs. Pace of reforms One argument for expeditious reform is that a long-drawn-out reform process allows opponents time to organize and lobby for a reversal. Furthermore, difficulties in importing intermediate inputs and capital goods may persist, which would interfere with the restructuring that is expected to occur in the initial phase of liberalization. Finally, the sooner the supply response begins, the better the prospects for sustainability. An argument against rapid reform, on the other hand, is that it concentrates transitional costs, including unemployment, in a short period. Some successful reforms have been comprehensive, intensive and fast, as in Bolivia, where extensive reforms were carried out virtually overnight. Chile’s phasing out of quantitative restrictions was rapid, while tariff reductions took place over five years. As with Chile’s tariff reforms, it is desirable to announce the trade reforms in advance, even though implementation is spread over a period of years to give affected activities time to adjust. Mexico quickly reduced the coverage of quantitative restrictions and reduced tariffs in about two years. Korea has carried out its comprehensive reforms over twenty years, with substantial import liberalization occurring since 1980. While the decision on the pace of import reform depends on country circumstances, experience reviewed in this study suggests that no longer than five to seven years—including decisive actions in the first year—is appropriate in most cases for a substantial liberalization. This should allow time for quantitative restrictions to be phased out, with tariffs reduced to, say, 15–30 per cent. Later stages may reduce rates further. In Africa, the pace of reform has often been very slow, but there are exceptions. In two years, Mozambique went from a centrally planned, price-controlled economy to a relatively liberalized market environment with a competitively valued currency. While there is still much to do, the reforms appear to be paying off; for example, a Hong Kong firm has recently established a joint venture for exporting garments. Nigeria’s programme, begun in 1986 after a public debate, was both intensive and rapid. It liberalized the exchange rate system, eliminated import licensing and loosened export regulations, removed price controls and abolished agricultural marketing boards. While an economy as distorted as Nigeria’s (following the overspending of the oil boom and bust) cannot be expected to recover quickly, the signs in 1987 and 1988 were encouraging. Non-oil exports—especially cocoa —rose 40 per cent per year. After 5 years of contraction at 4 per cent per year, the manufacturing sector’s output also began growing in 1987. The new

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

47

manufacturing growth is based on local inputs, rather than the highly protected assembly operations of the past. Lax fiscal policy has recently created more problems, and investment still has not begun to recover, partly because of regulatory constraints that were slow to be relaxed and partly because of uncertainty caused by the slow pace of setting tariff targets. Nonetheless, the outlook overall seems positive. TRADE POLICY MEASURES Neutral trade policy A more outwardly oriented economy requires a policy regime that does not discriminate against exports. This can be achieved by a relatively hands-off approach (Chile, Hong Kong, Mexico) or through selective government assistance (South Korea, Taiwan, China). Relying on noninterventionist and neutral policies supported by an adequate exchange rate and an overall stable macroeconomic environment has several merits. It avoids the susceptibility to misjudgement and abuse to which targeted investment policies are prone. It also avoids the practical problems that arise because direct export support may become subject to countervailing duties under the General Agreement on Tariffs and Trade (GATT). Studies of large samples of countries have indicated that, on average, less interventionist regimes have been more effective in promoting exports and growth (Agarwala 1983; Balassa 1988; Easterly and Wetzel 1989; Edwards, 1989a; Marsden 1983; Landau 1983; Scully 1988). Exchange rate policy The adequacy of the real exchange rate should be judged with respect to attaining equilibrium in the balance of payments and domestic markets and to its compatibility with growth in tradables and output over the longer term. Overvaluation of the domestic currency indirectly taxes exportables and lightly protected importables, relative to non-tradables and importables protected by binding restrictions. Uncertainty in the real exchange rate also hurts exports (Caballero and Corbo 1989). A real devaluation, with supportive macroeconomic policies and accompanied by exchange rate unification where relevant, improves the incentives for exports and efficient import substitutes. A crawling peg system helped many countries to prevent domestic inflation from offsetting the nominal exchange rate adjustment (Brazil, Chile, Colombia, Turkey) (Edwards 1989b). The importance of the exchange rate to trade policy reform is emphasized by the experience of Côte d’Ivoire. Côte d’Ivoire’s programme has floundered largely because its currency is pegged at a fixed rate to the French franc, and a simulated devaluation using import tariffs and export subsidies was not fully implemented and proved to be a poor substitute for devaluation. This is a major issue also in other countries in Francophone Africa. Export policy Despite their increasing use of non-tariff barriers, industrial countries remain attractive markets for developing countries’ manufactured exports. About one-third of the countries that have received trade adjustment loans increased their share of exports to industrial countries (Table 2.6). Macroeconomic stability, with low fiscal deficits and inflation rates, and stable and adequate exchange rates, has been the hallmark of East Asia’s economic success. Relatively low protection for import substitutes helps to sustain a more depreciated real exchange rate than would otherwise be the case. It also makes it easier to administer schemes that exempt exporters from restrictions and tariffs on their imported

48

TRADE POLICY REFORM: AN OVERVIEW

inputs. A few East Asian economies have been successful with protective import policies (South Korea and Taiwan, China) by avoiding exchange rate overvaluation and by using export measures to offset the antiexport bias from import protection. Korea’s approach during the 1960s and 1970s, however, would be difficult for other countries to replicate. Rent-seeking activities viewed as incompatible with export growth were vigorously suppressed in Korea, which would be difficult with less authoritarian governments. The approach included export subsidies and special incentives, which other countries would countervail today (Nam 1986; Balassa 1982) and which have had pernicious effects in other countries, when tried. (See the discussions on Yugoslavia in Havrylyshyn 1988; Argentina in Nogués 1989b; China in World Bank 1988a; and seven countries in Fitzgerald and Monson 1989.) Even in Korea, when the policy was Table 2.6 Exports of non-oil developing countries as a share of exports from developing countries to industrial countries, with selected country examples (in percentages) Category

Number of countries

1981

1988

Export share (%)

Non-oil developing 74 100.0 100.0 countries Trade adjustment loan 37a 55.9 59.5 countries With rising share 12 30.4 42.4 Korea 9.8 18.5 Turkey 1.7 2.9 Ten others 18.9 21.0 With declining share 25 25.1 16.9 Zambia 1.4 0.7 Côte d’Ivoire 1.6 0.7 Non-trade adjustment loan 37b 44.1 40.5 countries With rising share 9 12.3 17.5 China 8.4 11.8 Portugal 2.7 4.2 Seven others 1.2 1.5 With declining share 28 31.8 23.0 Peru 1.9 0.9 Sudan 0.3 0.1 Source: IMF, Direction of Trade Statistics. Note: Exports of the 75 non-oil developing countries in the sample of 88 countries considered in this report as a percentage of exports from all developing countries (IMF definition) to industrial countries. a The forty trade adjustment loan countries excluding Indonesia, Mexico, Nigeria which are oil exporters. b The forty-seven non-trade adjustment countries excluding ten oil exporters.

shifted from one of providing incentives to all exporters to one of targeting selected industries (the ‘Heavy and Chemical Industries’ drive), the consequences were on balance negative. This was recognized by policymakers, who in the 1980s accelerated the liberalization of imports and reduction of offsetting export

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

49

subsidies. (A former deputy Prime Minister describes how the government recognized and corrected its mistake; see Kim 1987.) Korea now is among those countries with very substantial reduction in protection (currently, non-tariff barriers cover only about 5 per cent of imports and the average tariff is only 20 per cent), with a positive effect on GDP and exports. Import protection translates into a bias against exports by contributing to a more appreciated exchange rate in addition to making import substitution more profitable. While exporters face world prices for their output, import protection increases the costs and reduces the availability of the inputs used in exports. This bias can be partially offset by schemes that reduce input costs, such as giving direct and indirect exporters restriction-free access to inputs at duty and tax-free international prices.5 One way is to have an essentially free trade regime with no tariffs or restrictions on imports (Hong Kong and Singapore). In regimes with import protection, one approach that can be used for larger firms that import inputs for export and domestic production is to provide duty waivers (and exemptions from other import restrictions) or temporary admissions of imported inputs (India, Indonesia, Mexico, Morocco and Turkey). But these systems must be designed to be user-friendly. Madagascar also has implemented a temporary admission system, but it is so cumbersome (51 documents need to be stamped and verified three times on average) that exporters must employ special firms or a great deal of staff-time to process each request. Second, for small or irregular exporters for which waivers or temporary admission schemes are not practical, a quick reliable system uses drawbacks or rebates of duties and indirect taxes actually paid (India, Korea, Taiwan, Thailand). But collecting and refunding duties is inefficient compared to waiving duties, and drawbacks do not offset nontariff barriers. Third, a duty-free scheme can be provided that allows in-bond manufacturing exporters to locate almost anywhere (Mauritius, Mexico). Manufactured exports in Mauritius increased from almost nothing to 39 per cent of current exports. A fourth alternative, used in more than thirty developing countries, is to establish physically separate export processing zones. Many of these zones have proved to be poor investments as a result of unwise location, high investment costs, mediocre management, or uncooperative customs officials, but the best ones—some public, some private—have done well. Of course, duty-free schemes for exporters’ inputs also involve costs. These may include drawing resources from more efficient activities, temporarily increasing a fiscal deficit,6 disprotecting domestic suppliers of importable inputs, and creating new opportunities for rent-seeking. These costs need to be balanced against the likely economic gains. Manufactured exports must meet exacting and frequently changing requirements and must be delivered reliably and on time. Valuable factors in that respect are: efficient infrastructure and telecommunication, readily available export credit, technology development, quality control, production planning, attention to trade logistics and facilitation, and organization. The East Asian experience suggests the value of adjusting regulations for all firms on layoffs, fringe benefits, minimum wages, and collective action, to reduce labour costs and increase flexibility at the enterprise level. Industrial location and regional development policy may also have to be changed, since exports on a large scale cannot be expected from backward areas with poor infrastructure. It may help to assist exporters in securing technical assistance services from consultants and information suppliers of their choice. Much of the Bank’s technical assistance to exporters has been channeled through official export promotion organizations, with disappointing results (Keesing and Singer 1989). Foreign direct investment can be a valuable source of technology, capital, and connections to world markets. Policies attractive to foreign investors include macroeconomic stability, protection of property rights (including intellectual property rights), a stable and transparent regulatory environment, and liberal access to foreign exchange for profit remittances and imported inputs and services (World Bank 1989c). This kind of good climate for investment is likely to be superior to special incentives, such as tax holidays, which may attract footloose industries that leave when the holiday is over.

50

TRADE POLICY REFORM: AN OVERVIEW

Exports for primary sectors are frequently discouraged by overvaluation, taxes, low administered prices, and inefficient government marketing monopolies, as well as by industrial protection and restrictions on foreign investment. For some key agricultural products that compete with temperate-zone agriculture, export growth is also constrained by industrial countries’ protection. In other cases the constraint arises from the slow growth in the market and the price inelasticity of demand. Malaysia in the 1960s and Chile in the later 1970s created a positive policy environment for primary exports, with good results. Chile’s reforms helped to reinvigorate the export-oriented mining sector, based to a large extent on foreign private initiative. They also led to a spectacular increase in agricultural and wood product exports, which grew from US$ 44 million in 1972 to US$ 1,102 million in 1986. In the 1980s, some countries (Bolivia, Guinea) opened mining to foreign investment. Producer prices for leading primary product exports have been improved by eliminating direct or implicit taxes (Bangladesh, Côte d’Ivoire, Ghana, Malawi, Philippines, Turkey). Argentina and Uruguay reduced export taxes, but the action was quickly reversed in Argentina. Regulatory controls on exports have been reduced in Colombia, Mexico, Morocco and Tanzania. Malaysia and Thailand achieved sustained and strong growth in primary exports by avoiding major currency overvaluation, heavy taxation of the sector, and high protection of manufacturing industries. Some countries have eliminated public sector marketing boards or stripped them of their monopoly procurement powers (Ghana, Madagascar, Mali, Morocco, Nigeria, Senegal, Tanzania). Such reforms have significantly increased primary exports and agricultural growth in general (Balassa 1986, 1988; Jaeger 1989). Import policy Import licensing, prohibitions, exceptions, quotas, official reference prices and foreign exchange allocation schemes are common non-tariff barriers. Because they depend on discretionary decisions by the authorities, they make the system less transparent and predictable and encourage lobbying, rent-seeking and corruption. Even with little or no decrease in protection, reduction of nontariff barriers can have major salutary effects. One simple reform is to switch from a positive import list (that allows unlicensed imports only of listed items) to a negative list (that allows unlicensed imports of all items not listed). This was the first major step in Korea’s liberalization programme in 1967. Auction systems can be substituted for administrative rationing. In quantitative quota systems, quotas can be auctioned, with the quota amount increased until its protective value falls to zero, at which time it can be abolished. Alternatively, tariffs providing approximately equivalent protection can be imposed on product categories as non-tariff controls are eliminated. This tariffication reestablishes the link between domestic and international prices, ensuring that they move in the same direction and do not diverge by more than the amount of the tariff. Because tariffs on finished products are usually higher than on intermediates and raw materials, and because tariff exemptions are common, effective protection varies greatly across industries. In most countries, production efficiency requires that effective protection be reduced, and that protection among imports be made more uniform, taking into account the protective effect of the domestic tax system.7 By coordinating tariff reform with domestic tax reform to offset revenue losses, deeper reductions in tariffs are possible than otherwise. This policy should eventually result in a low, equal rate of tax on consumption of imports and domestically produced products. Raising low tariff rates (usually on inputs) also increases revenue, allowing high rates to be reduced further, and makes effective protection more uniform between inputs and finished goods. However, raising low rates encourages domestic production of products whose effective tariff rates are raised, perhaps drawing resources from exports. Exporters would thus need to be insulated from paying prices above world levels for these protected inputs. In general, a low and relatively uniform tariff structure (across sectors and between inputs and outputs) not only encourages productive

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

51

efficiency, but also has the advantage of being less subject to lobbying and the use of political clout. Some of the more successful reformers (Bolivia, Chile, Mexico) have converged their tariff structures towards 15 per cent. Even with relatively low tariff rates, producers in most regions of Africa would remain protected by the natural barrier of the high cost of transporting goods to the area. Among the paths toward a relatively low and uniform tariff structure, the most difficult to implement seems to be an industry-by-industry approach, since changes in one industry may have repercussions in others. What seems to be better is a ‘concertina’ approach, whereby at each stage all the top rates are collapsed to the next highest level. Even better would be a ‘radial’ reduction, in which all rates at each step are cut to an equal fraction of their previous level, thereby reducing protection as well as tariff rates at each step (Harberger 1974). Radial tariff reduction promises faster production efficiency gains, but is more likely to be revenue-reducing in the first stages. The concertina approach concentrates initial reductions on very high rates, at least some of which may be so high that reducing them will increase import volume and tariff revenues. In many cases, greater uniformity might best be achieved by a combination approach: a collapse of the very high rates and a radial reduction of all other rates. POLICIES TOWARDS TRADING PARTNERS Newly industrialized countries have much to gain from reduced external barriers to their exports. They also have some leverage in the Uruguay Round negotiations since the industrial countries are seeking an opening of their markets (Finger and Olechowski 1987). In contrast, countries interested in agricultural issues may need to negotiate as a group if they are to have leverage. One issue raised by those countries with some leverage or interest in getting concessions is whether to reduce barriers unilaterally or to delay reforms in order to gain concessions in multilateral negotiations. An unambiguous statement by all negotiating parties in the Uruguay Round that credit would be granted for unilateral reduction of barriers would eliminate this incentive to delay reform pending negotiations. Another alternative is to bind tariffs at a higher level than the actual level, and offer to reduce the bound level at negotiations. This approach, however, requires a credible threat to raise tariffs to the bound level if negotiations fail, as well as a credible commitment not to raise the actual tariffs—by resorting to article XVIII of GATT—if negotiations succeed. Aside from the credit issue, the costs a country imposes on itself by its own trade policies are likely to be higher than the costs imposed on it by other countries’ barriers, which argues for unilateral reforms (Nogués 1989a). This is especially true of African countries, which are subject to relatively few barriers. During the last three decades, industrial country tariffs have been reduced from an average of approximately 40 per cent in the late 1940s to less than 5.5 per cent today. But non-tariff barriers have increased, with the proportion of developed country imports affected by non-tariff barriers nearly doubling in the period 1966–86 (Laird and Yeats 1988; Kelly et al. 1988). Non-tariff barriers, coupled with the surpluses resulting from high OECD agricultural support prices, has caused major distortions and destabilization of agricultural markets. Despite ‘voluntary’ quotas (especially textiles and steel), industrial country markets remain attractive for a wide range of manufactured exports. Nonetheless, the non-tariff barriers and agricultural policies in the industrial countries reduce the potential benefits from trade reforms by developing countries and make reforms politically more difficult. They also seriously depress exports of the highly indebted countries and exacerbate the debt crisis (Laird and Nogués 1988), while reducing the welfare of consumers in the industrial countries and threatening the world trading system. Industrial countries must act to reform these policies, but the future of efforts to do so in the Uruguay Round are clouded with uncertainty.

52

TRADE POLICY REFORM: AN OVERVIEW

The potential for expanding trade with neighbours has led many countries to form regionally integrated groups. These groups were expected to allow members to realize gains from increased trade, to take advantage of economies of scale by production for a regional market, and to provide initial exporting experience under protection. But intraregional trade expanded little (Economic Community of West African States) or fell (UDEAC, West African Economic Community, East African Common Market). Industries established as a result of integration usually had high production costs, and experience gained in marketing to neighbours proved not very useful in exporting to wider markets. And most regional groups raised barriers against extraregional trade, discouraging integration into the world economy, where the gains from trade are likely to be greater. Low benefits, high costs and practical implementation problems have led to the breakdown of many schemes. Despite the shortcomings, integration efforts continue. Integration has been more successful among countries with generally outward-oriented economies (ASEAN or the European countries.) Two lessons from the experience with integration efforts are noteworthy. First, integration schemes might focus more on improving infrastructure and factor mobility than on trade policy measures. Increased trade should follow naturally. Second, any trade policy measures might focus on reducing barriers to all trade, not just trade among members. In the African context, this would generally require making all members’ currencies convertible and removing existing artificial barriers to trade among members. (There is considerable scope for this kind of action; some African countries have easier access to EC markets than to their neighbours’. See World Bank 1989b.) Above all, integration should accelerate, or at least not interfere with, reduction of barriers to trade with the outside world. The Central American Common Market, for example, has explicitly recognized the need not to impede member countries’ progress in overall trade policy reform. CREDIBILITY AND SUSTAINABILITY If the private sector is to invest in new sectors, it must believe that the changed incentives will be sustained. For credibility, the first steps should be decisive: reforms begun with tentative steps have often been reversed (Michaely et al. 1986). A strong endorsement by the head of state—as by Flt Lt Rawlings of Ghana or President Babangida of Nigeria—can help to establish credibility. A general interest agency (central bank, finance ministry) rather than one with specific protectionist interests (ministry of trade or industry) is usually more reliable in executing reforms. Credible announcement of a timetable - although it carries the risk of giving time for opposition to be mobilized against the reform—can usually strengthen the reform process if subsequently adhered to (Korea). Reforms are generally easier to introduce after a crisis that discredits old policies and accomplishes the necessary but painful reduction of absorption. In Nigeria, reforms followed a crisis that discredited the use of quantitative controls. In Zambia, implementation of a foreign exchange auction coincided with a period of shrinking exchange availability. The public (and some international non-governmental organizations!) viewed reforms as the cause of the crisis, and they were subsequently reversed. Since the crisis is often related to the balance of payments, a strong devaluation is probably necessary and useful in sending a signal to producers, whose supply response will help to justify the programme. Sustainability may be enhanced by an external commitment to maintain reforms, such as accession to the GATT. However, Article XVIII decreases the value of this commitment by allowing developing countries to impose barriers for purposes of correcting balance of payments disequilibria or protecting infant industries (Hudec 1987). Despite its net benefits, trade policy reform can have short-term transitional costs for some segments of the population. Steps to partially compensate losers may increase the odds that the reforms will be sustained

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

53

and have other benefits. Some measures, such as worker retraining, can improve factor mobility and speed the inter-sectoral adjustment process. Other measures, such as antidumping procedures (set up in Chile, Mexico), may reduce resistance to the reduction of more costly forms of protection. Other steps, such as targeted food assistance programmes and employment programmes, may mitigate effects on real wages and employment. Often, however, workers displaced from protected industries are not among the poorest groups in the society. Moreover, new programmes run the risk of creating new distortions (food programmes only available in cities exacerbate the anti-rural bias and consequent migration) and exacerbating fiscal deficits. But the compensation issue needs to be addressed, and pragmatic means must be found to target the programmes well. NOTES 1 During June 1989, there were 98 loans to 44 countries that contained significant trade policy components. 2 Among the forty trade loan recipients, implementation data were available for twenty-four. Detailed findings on implementation refer to the twenty-four, while aggregate data were also considered for the forty as well as for fortyeight non-recipients. 3 To take one example, deficiencies in trucking and port regulations in Mexico became more evident when expanding exports increased the demand for transport services. 4 Countries with multiple exchange rates in the goods market can gain from unifying them at this time. 5 This avoids wide variations in effective incentives between different exports that use inputs subject to different import controls and tariffs. It is not subject to the countervailing measures that importing countries are increasingly applying to direct or indirect export subsidies. 6 Short-run negative revenue effects may occur when protection of inputs is principally provided by tariffs. Tariff revenue should increase, however, as expanding exports increase the supply of foreign exchange for imports. 7 The domestic tax structure can be quite important. Shalizi and Squire (1986) found that domestic tax rates in Ghana varied so much across sectors that a uniform tariff would result in effective protection rates of 0 to 50 per cent. This underscores the importance of coordinating tariff and domestic tax reform.

REFERENCES Agarwala, R. (1983) Price Distortions and Growth in Developing Countries, World Bank Staff Working Paper No. 575, Washington DC: World Bank. Balassa, B. (1982) ‘Development strategies and economic performance’, in B.Balassa and associates, Development Strategies in Semi-Industrialized Countries, Baltimore, Md: Johns Hopkins University Press. (1986) ‘Economic incentives and agricultural exports in developing countries.’ Paper presented at the Eighth Congress of the International Economic Association, New Delhi, India. (1988) Incentive Policies and Agricultural Performance in Sub-Saharan Africa, World Bank PPR Working Paper No. 77, Washington DC: World Bank. Baldwin, R.E. (forthcoming) ‘High technology exports and strategic trade policy in developing countries: the case of Brazilian aircraft’, in G.K.Helleiner, (ed.) New Trade Theory and Industrialization in Developing Countries, New York and London: Oxford University Press. Bergsman, J. (1974) ‘Commercial policy, allocative efficiency, and X-efficiency’, Quarterly Journal of Economies 88 (August): 409–33. Bevan, D.L., Collier, P. and Gunning, J.W. (1987) ‘Consequences of a commodity boom in a controlled economy: accumulation and redistribution in Kenya 1975–83’, World Bank Economic Review1 (May): 489– 513. Bruton, H.J. (1970) ‘The import substitution strategy of economic development: a survey’, Pakistan Development Review 10 (2):123–46.

54

TRADE POLICY REFORM: AN OVERVIEW

Caballero, R.J. and Corbo, V. (1989) How Does Uncertainty about the Real Exchange Rate Affect Exports? World Bank PPR Working Paper No. 221, Washington DC: World Bank. Chenery, H., Robinson, S. and Syrquin, M. (1986) Industrialization and Growth: A Comparative Study, New York: Oxford University Press. Condon, T. and de Melo, J. (1986) ‘Industrial organization implication of QR trade regimes: evidence and welfare costs.’ Paper prepared for meetings of Applied Econometric Association, Istanbul (10–12 December). Washington, DC: World Bank. Corbo, V. and de Melo, J. (1987) ‘Lessons from Southern Cone policy reforms’, World Bank Research Observer 2 (2). Corden, W.M. (1974) Trade Policy and Economic Welfare, Oxford: Clarendon Press. Devarajan, S. and de Melo, J. (1987) ‘Adjustment with a fixed exchange rate: Cameroon, Côte d’Ivoire, and Senegal’, World Bank Economic Review (1 May): 447–88. Easterly, W.R. and Wetzel, D.L. (1989) Determinants of Growth: Survey of Theory and Evidence ,Work Bank PPR Working Paper No. 343 (December), Washington DC: World Bank. Edwards, S. (1989a) Openness, outward orientation, trade liberalization and economic performance In developing countries, World Bank PPR Working Paper No. 197, Washington, DC: World Bank. (1989b) Real Exchange Rates, Devaluation and Adjustment: Exchange Rate Policy in Developing Countries. Cambridge, Mass.: MIT Press. Finger, J.M. and Olechowski, A. (eds.) (1987) The Uruguay Round: A Handbook for the Multilateral Trade Negotiations, World Bank, Washington DC. Fischer, S. 1984. Real Balances, the Exchange Rate, and Indexation: Real Variables in Disinflation, NBER Working Paper No. 1497, Cambridge, Mass.: NBER. Fitzgerald, B. and Monson, T. (1989) ‘Preferential credit and insurance as means to promote exports’, World Bank Research Observer 4:89–114. Grais, W., de Melo, J. and Urata, S. (1986) ‘A general equilibrium estimation of the effects of reductions in tariffs and quantitative restrictions in Turkey in 1978’, in T.N.Strinivasan and J.Whalley, (eds) General Equilibrium Trade Policy Modeling, Boston: MIT Press. Grossman, G.M. and Helpman, E. (1989a) Comparative Advantage and Long-run Growth, National Bureau of Economic Research Working Paper 2809. Cambridge, Mass.: NBER. (1989b) Growth and Welfare in a Small Open Economy, National Bureau of Economic Research Working Paper 2970, Cambridge, Mass.: NBER. Halevi, N. (1989) ‘Trade liberalization in adjustment lending’. Background paper for Strengthening Trade Policy Reform, See M89–1454/1 CECTP. Washington, DC: World Bank. Harberger, A.C. (1974) ‘Notes on the dynamics of trade liberalization.’ Prepared for a Conference on Trade Liberalization, Santiago, Chile (October). Havrylyshyn, O. (1988) ‘Yugoslavia: the experience of trade policy reform, 1965–1975, in Trade Liberalization: The Lessons of Experience, World Bank LAC Regional Series Report No. IDP14, Washington, DC: World Bank. Heitger, B. (1987) ‘Import protection and export performance: their impact on economic growth’, Weltwirtschaftliches Archiv, 123:249–59. Hudec, R. (1987) Developing Countries in the GATT Legal System, London: Gower Press, for the Trade Policy Research Centre. Jaeger, W. (1989) ‘The impact of policy on African agriculture: an empirical investigation’, (December 14), mimeo, Washington, DC: World Bank AFTTF. Keesing, D. and Singer, A. (1989) How to Provide High-Impact Assistance to Manufactured Exports from Developing Countries, Washington, DC: CECTP, World Bank. Kelly, M., Kirmani, N., Xafa, M., Boonekamp, C. and Wingle, P. (1988) Issues and Developments in International Trade Policy. International Monetary Fund Occasional Paper No. 63, Washington DC:IMF. Kemp, M.C. and Negishi, T. (1970) ‘Variable returns to scale, commodity taxes, factor market distortions and their implications for trade gains’, Swedish Journal of Economics 73(1): 499–526.

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

55

Kiguel, M. and Liviatan, N. (1988) ‘Inflationary rigidities and orthodox stabilization policies: lessons from Latin America’, World Bank Economic Review 3 (September): 273–98. Killick, T. (1989) ‘Industrialization Policies in Ghana’, in G.M.Meier and W.F.Steel (eds) Industrial Adjustment in SubSaharan Africa, London: Oxford University Press, pp. 71–6. Kim, Mahnje (1987), ‘Korea’s adjustment policies and their implications for other countries’, in V.Corbo, M.Goldstein and M.Khan (eds) Growth Oriented Adjustment Programs, Washington, DC: IMF and World Bank. Krueger, A.O. (1974) ‘The political economy of the rent-seeking society’, American Economic Review 64 (June): 291–303. (1981) ‘Interactions between inflation and trade objectives in stabilization programs’, in W.Cline and S.Weintraub, (eds) Economic Stabilization in Developing Countries, Washington, DC: Brookings Institution. Krugman, P.R. (1987a) ‘Is free trade passé?’ Journal of Economic Perspectives 1(Fall):131–44. Krugman, P.R., (ed.) (1987b) Strategic Trade Policy and the New International Economics, Cambridge, Mass.: MIT Press. Laird, S. and Nogués, J. (1988). Trade Policies and the Debt Crisis. World Bank PPR Working Paper No. 99, Washington, DC: World Bank. Laird, S. and Yeats, A. (1988). ‘Nontariff barriers of developed countries, 1966–1986’, Finance and Development25 (March): 12–13. Landau, D. (1983) ‘Government expenditure and economic growth: a cross-section study’, Southern Economic Journal49 (January): 783–92. Lopez, R. (1989) ‘Trade Policy, Growth and Investment.’ Background paper for Strengthening Trade Policy Reform. Sec M89–1454/1, CECTP, Washington, DC: World Bank. Marsden, K. (1983) Links Between Taxes and Economic Growth: Some Empirical Evidence, World Bank Staff Working Paper No. 605, Washington DC: World Bank. Michaely, M., Choksi, A. and Papageorgiou, D. (1986) The Phasing of a Trade Liberalization Policy: Preliminary Evidence. World Bank CPD Discussion Paper No. 1986–42, Washington, DC. Mohammad, S. and Whalley, J. (1984) ‘Rent-seeking in India: its costs and policy significance’, Kyklos 37:387–413. Nam, C.-H. (1986) Export Promoting Policies under Countervailing Threats: GATT Rules and Practices. World Bank Discussion Paper No. VPER59, Development Policy Issues Series, Washington DC: World Bank. Nogués, J. (1989a) The Choice Between Unilateral and Multilateral Trade Liberalization Strategies, World Bank PPR Working Paper No. 239, Washington DC: World Bank. 1989b. Latin America’s Experience with Export Subsidies, World Bank PPR Working Paper No. 182, Washington, DC: World Bank. Panagariya, A. (1980) ‘Variable returns to scale in general equilibrium theory once again’, Journal of International Economies 10 (November): 499–526. Pinto, B. (1987) ‘Nigeria during and after the oil boom: a policy comparison with Indonesia’, World Bank Economic Review 1(May): 419–46. Romer, P.M. (1989) What Determines the Rate of Growth and Technological Change?World Bank PPR Working Paper Series 279, Washington DC: World Bank. Sachs, J. (1987) Trade and Exchange Rate Policies in Growth-Oriented Adjustment Programs, Cambridge, Mass.: Department of Economics, Harvard University. Scully, G.W. (1988) ‘The Political Economy of Free Trade and Protectionism.’ Paper prepared for Conference on the Political Economy of Neomercantilism and Free Trade, Big Sky, Montana, June 9–11. Shalizi, Z. and Squire, L. (1986) Tax Policy for Sub-Saharan Africa, Washington DC: Country Policy Department, Resource Mobilization Division, World Bank. Sood, A. and Kohli, H. (1989) ‘Strategies for Restructuring.’ In Meier and Steel, p. 98–103. Spencer, B.J. and Brander, J.A. (1983) ‘International R&D rivalry and industrial strategy’, Review of Economic Studies 50:707–22. Steel, W.F. (1989) ‘Recent policy reform and industrial adjustment in Zambia and Ghana’, in G.M.Meier and W.F.Steel (eds) Industrial Adjustment in Sub-Saharan Africa, London: Oxford University Press, pp. 152–9.

56

TRADE POLICY REFORM: AN OVERVIEW

Taylor, L. (1988) Economic Openness: Problems to the Century’s End, Helsinki: World Institute for Development Economics Research. Thomas, V. (1989) Developing Country Experience in Trade Reform, World Bank PPR Working Paper No. 295, Washington, DC: World Bank. World Bank (1988a) Adjustment Lending: An Evaluation of Ten Years of Experience, Policy and Research Series No. 1, Washington, DC: World Bank. World Bank (1989b) The Long-Term Perspective for Sub-Saharan Africa: A Strategy for Recovery and Growth, Washington, DC: World Bank. World Bank (1989c) ‘The role of foreign direct investment in financing developing countries’, Board Memorandum (July 11), Washington DC: World Bank.

3 Import controls and the sequencing of trade policy reform with special reference to Africa David Evans

INTRODUCTION AND RATIONALE Theory of rent-seeking behaviour, and its generalization to encompass a wider set of DUPe or Directly Unproductive Profit-seeking-activity (Krueger 1974; Bhagwati 1982; Bhagwati and Srinivasan 1982), has been widely regarded as one of the seminal contributions of the neo-liberal tradition.1 This acclaim has not, however, been without a strong reservation about the narrow context within which the neo-liberal theory of rent-seeking behaviour has been set. The economic model within which rent-seeking and DUPe behaviour is analysed is the standard competitive one, save for the modification that allows for the expenditure of real resources by private agents in order to acquire access to rents created by policy-induced market distortions. Typically, there is no role for the state within the simple abstract model other than the identifier of market distortions and the designer of optimal policies to overcome those distortions. In the area of trade and industrial policy, once the scope for state intervention has been limited by the argument that, whatever the residual case might be for intervention to overcome market distortions, the welfare loss through imperfect state intervention is likely to greatly exceed the cost of continued market distortions, then the role of the rent-seeking and DUPe models is to establish whether or not competition for rents, which may undermine the effects of sub-optimal policy instruments, is welfare improving or not. The argument that direct quotas on trade are the worst form of policy-induced distortion frequently encountered in developing countries, either in the pursuit of importsubstituting industrialization (IS) or in balance of payments policy, quickly follows. The greater rigidity of a quantity-fix distortion, compared with policy instruments that only introduce price differentials, leaving the quantities to adjust to the new equilibrium, leads to this result. In this chapter, I have started from the position that unproductive economic activity can be induced within both the private sector and the state, and that the state or other non-private sector institutions can carry out productive activity. However, the private sector is given the benefit of the doubt, and no unproductive behaviour is directly considered.2 As with the standard rent-seeking and DUPe model, the state is the sole site of potential unproductive behaviour. To set the ball rolling, two key assumptions are made. First, a set of strategic imperfections affecting trade and industrial policies are identified in the markets for information and technology. Second, the set of direct controls considered is extended to encompass direct controls tied to administered incentives. It is argued that both assumptions capture important elements of the East Asian experience in designing trade and industrial policies. Consideration is then given to possible forms of state intervention and coordination to overcome those market distortions, so that private producers, the state, or some mixed private-state institution, can act to increase the supply of the strategic inputs. Once

58

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

these modifications to the standard model are allowed, then a modified set of quantitative controls tied to administered incentives may be superior to both fixed controls or tariffs. THE SEQUENCING OF TRADE POLICY REFORM: A REVIEW OF THE STANDARD ARGUMENT Consider the case of the introduction of trade policies designed to foster infant import competing industry.3 The case for this may arise for a number of reasons, such as scale economies, external economies such as labour training, or where important producer goods required for the infant industry such as knowledge and technology have mixed public and private goods attributes. Or it might be the case that, because of imperfections in the capital market, future benefits of setting up the infant industry either are not foreseen or cannot be privately appropriated. In all of these cases, the argument for infant industry protection involves time. However, the standard case for the sequencing of trade policy instruments need not consider the above distortions explicitly; it is enough to consider the relative merits of each trade policy instrument at a given point in time. In the standard two-commodity two-factor competitive trade model without rent-seeking, the introduction of a quantitative import control aimed at protecting the infant industry has a deadweight cost arising from the short-run consumption and production distortion. In addition, the rents for the import licence accrue to the import licence holders, rather than to the issuing authority, the government. An equivalent tariff has the same deadweight production and consumption losses, but has the beneficial effect (from most perspectives) of providing the government, rather than import licence holders, with the revenue benefits of restricted access to the domestic market. On redistributive and revenue grounds, the import licence is inferior to the tariff as a means of protecting the infant industry. Suppose that the same protective effect was to be achieved by a direct subsidy to the infant industry. In this case, there is no byproduct consumption distortion. Leaving revenue and redistributive arguments aside, the direct subsidy is a more efficient means of providing infant industry protection. The advantage of the direct subsidy over tariff protection is lessened when the government tax base is weak, so that there is a revenue constraint. In the presence of a revenue constraint, it may be possible for the consumption distortion arising for raising tariff revenue to be less than alternative means of raising government revenue to finance infant industry protection, or for the additional costs of the consumption distortion caused by the tariff to be offset by future infant industry benefits. In either case, the presence of a government revenue constraint may make tariff protection superior to a direct production subsidy. However, the presence of a revenue constraint reinforces the inferiority of import quotas to either tariff protection. The case against import quotas is further strengthened when rent-seeking or DUPe behaviour is allowed for. In the standard case, when rent-seeking behaviour designed to seek access to import quotas leads to Pareto-sub-optimal production (that is, where the production of both the importable and exportable commodity is lowered) the welfare cost of import quotas is increased. Where tariff protection itself attracts rent-seeking behaviour, the argument for the superiority of the tariff over the import control is a little more subtle. Suppose that the rent-seeking behaviour around the allocation of import quotas, and around the allocation of tariff revenue benefits, favours the producers of the exportable commodity. Suppose also that the cost of the rent-seeking behaviour is the same in both cases. Then the presence of rent-seeking behaviour will lead to production inside the production possibilities’ frontier, but in comparison to the situation without rentseeking, there will be more production of the exportable and less production of the importable. With import quotas, the relative scarcity of the importable in consumption will increase and there will be a welfare loss

IMPORT CONTROLS AND TRADE POLICY REFORM

59

in comparison with the case without rent-seeking. However, with tariff protection, there is no absolute restriction on the quantity of imports, and the relative scarcity of the importable in consumption can be lowered through trade and the tariff with rent-seeking situation may be an improvement on the tariff or import control case without rent-seeking. It is certainly true that the tariff with rent-seeking leads to an improvement in welfare compared with import quotas with rent-seeking. Since rent-seeking is generally regarded as more prevalent around import quotas than around tariffs, the case against import quotas is strengthened by the introduction of rent-seeking behaviour. THE STANDARD THEORY AND THE IMPLICIT VIEW OF THE STATE The above arguments on the sequencing of trade policy reform have an implicit model of the state whose economic functions are confined to minimalist intervention to secure a basic infrastructure, plus policies designed to ensure the provision of basic education, health and security.4 Wherever possible, state policy seeks to find private and market solutions to the problems of economic development. This perspective is in sharp contrast to recent work on the Developmental State, which tends to stress the importance of state intervention in the formation and execution of a development strategy. It focuses upon the importance of the class base and the relative autonomy of the state, its administrative capacity, its ability to make long-run strategic decisions, and the role of intermediate levels of state intervention in the formulation and execution of a selective industrial policy.5 Some of this is recognized by analysts, who characterize interventionist East Asian states as clearing houses for information and technology in the face of market failure.6 Yet even this narrow economic interpretation of the role of the East Asian developmental state is resisted by neo-liberals, who remain highly sceptical of any non-market institutions or non-arms-length policy instruments.7 There are two aspects of the experience of the developmental state in East Asia that are relevant to the present discussion. First, a wider productive role for the state is recognized than in the neo-liberal minimalist state case, focusing on strategic market distortions in the markets for information and technology, over and above the standard minimalist state functions. In addressing market failure in these areas, East Asian states have typically taken on a direct or an enabling role in developing institutions to provide needed information and technology, and to undertake a strong coordinating role in the development process. Second, the East Asian states have modified and adapted direct controls as pervasive policy instruments to facilitate their trade and industrial policies.8 THE ROLE OF THE STATE IN TRADE AND INDUSTRIAL POLICY The economic rationale for state intervention through indicative planning rests on the notion that, in an uncertain world, the allocation of resources can be improved upon by providing individuals and firms with information about government and overall economic activity. However, in practice, government policy interventions centre around the government development budget and projects, very often in the context of crisis management: there is no role for an on-going process of two-way coordination and information flows between government policymakers and economic agents required for medium and longer run strategic policy intervention. Generally, state intervention in development policy is very much a top-down process, with a very limited bottom-up component. The standard formulation for trade and industrial policy is no exception to this general rule.9 The case for an interventionist and bottom-up approach to trade and industrial policy has gained wider respectability in recent years.10 This approach was the basis for Japanese industrial planning and the key to

60

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

the success of Japanese industrial growth in the postwar period, and of the Taiwanese and South Korean cases, which adapted the Japanese model. Information required by the policymaker includes where the industrial sector’s competitive edge lies, what innovations from other sectors can be introduced, who the leading producers are, and where growth potential exists. All of this is in strong contrast to narrow neoclassical or neo-liberal perspectives on microeconomic policy, which argue that the most important policy problem is to remove government interference in the price mechanism, to ‘get the prices right’, and that sector detail is not the concern of government policymaking. The above ideas are forcefully elaborated by Pack and Westphal (1986). They argue that the key role of an industrial strategy is to achieve dynamically efficient industrialization through the management of technical change. They stress (a) the imperfect tradeability of industrial products and technology, (b) the cost and the lack of markets for establishing technological capability, and (c) the interaction between the organization of economic activity, markets, relative prices and technology, all of which undergo major changes as industrialization proceeds. It would seem, then, that there is a set of generic microeconomic issues behind the advocacy of a sectoral industrial strategy, which go far beyond the neo-liberal ‘get the prices right’ view. The institutional context within which successful industrialization has taken place in other countries or regions has varied considerably, a point developed in the historical and in the present context by Best (1990). One possibility is the Japanese model with big vertically integrated MNCs around which large numbers of small firms cluster, combined with state institutions such as the Japanese Ministry of Trade and Industry and the Anti-Trust laws (see Best 1990, ch. 6). More relevant for small developing countries is the experience of small-scale export-oriented industry in the so-called Third Italy, where local government has played a decisive role in fostering interfirm associations to facilitate cooperation in the provision of services with substantial economies of scale. The local state has also assisted in generating technological externalities through the direct role of coordinator and information clearing house for regional industrial districts and through encouraging networking between firms.11 Best argues further that the industrial districts of the Third Italy have the institutional capacity for collective learning (see Best 1990, chs 7 and 8). Just as there is a wide variety of institutional forms for trade and industrial policy, so too is there a variety of policy instruments. In some cases, where markets are generally highly developed, the creation of enabling institutions that can intervene at the point of imperfection in the strategic markets for information and technology may be appropriate; many of Best’s examples from the Third Italy and Japan would fall into this category. In other contexts, such as the south Korean and Taiwanese cases, much more over-arching state involvement using stronger policy instruments has emerged. In particular, both of the latter cases have adapted direct quantity controls on imports and many other inputs into highly flexible instruments of trade and industrial policy. The key to the adaptation of quantity controls in both Taiwan and South Korea is their use in conjunction with performance indicators and, in some cases, the threat of removal of the control if the desired performance is not attained. Thus, in South Korea, direct allocation of import quotas, trading licences, foreign exchange and investment funds to firms depended on export performance (see LeuddeNeurath 1988, particularly pp. 71–88) and Amsden (1989, especially pp. 64–76). In Taiwan, quantitative restrictions on the import side were governed primarily by the government’s aims with respect to technological and supply capacity and exporters were given an apparently free-trade regime for current and capital inputs. However, exporters were in fact subject to a tightly controlled monitoring, so that imported inputs were only freely permitted if no suitable domestic substitute were available (see Wade 1988, especially pp. 48–53). In both South Korea and Taiwan, the threat of removal of an import restriction meant that domestic prices could be kept close to world prices if desired. The adaptation of direct controls to performance criteria in

IMPORT CONTROLS AND TRADE POLICY REFORM

61

both South Korea and Taiwan also provided an important nexus for the two-way exchange of a wide range of information between firms, governments and sectoral institutions, particularly with regard to markets and technology, as well as the nexus around which government coordination of the development process could proceed. In conjunction with a strong developmental ideology, direct controls with performance criteria also provides a framework within which rent-seeking behaviour can be minimized.12 THE SEQUENCING OF TRADE POLICY REFORM WHEN ORGANIZATIONAL VARIABLES ARE TAKEN INTO ACCOUNT Another way to put the above is to argue that, quite independently of the institutional form, a successful industry strategy requires the presence of an agent or agents who have the capacity to take on the coordinating and information clearing-house role, providing a nontraded intermediate input into the production process, which yields strong external economies. Typically, such institutions will be industryspecific; in the East Asian context, they have had an important role in administering and monitoring the effectiveness of tariffs, subsidies and other direct controls that affect the sector. For the purposes of the argument, it is assumed that there are no sectorwide scale economies, and that the size of sectoral institutions is chosen so that any scale economies that cannot be captured by firms within the sector can be captured by the sectoral institution. Again, in the East Asian context, these institutions have also played an important role coordinating the allocation of non-competing imports and the building of infrastructure, and coordinating sectoral policies with the central policymaking process. The coordination function requires some form of direct relationship between the sectoral institution and the firms. It might be organized around common inputs and through informal networks, such as in the industrial districts described by Best (1990). In the East Asian context, the administration of direct controls themselves and the monitoring of performance standards has provided the central nexus around which the co-ordination function has been organized. In this case, the chief sources of external economies derive from the coordination functions and the two-way information flows established within the sector, and between the sector and the central policymaking process, built around instruments of direct control and the associated performance criteria. Abstracting from the specific institutional context, the standard competitive trade model can be modified to include a highly stylized version of the above story. Thus, suppose that there are constant returns to scale and an importable and exportable commodity each producing with a homogeneous labour input and a composite and sector-specific non-traded strategic input. In the absence of technological externalities, the strategic input will reduce to sectorspecific capital and infrastructure and all of the standard sequencing arguments follow. With strong technological externalities arising from coordination and two-way information flows, but where the nexus around which coordination and two-way information flows is not specified, the strategic inputs will require the expenditure of labour in their production as well as sectorspecific capital and infrastructure. When the nexus for coordination and two-way information flows is built around the monitoring and administration of direct controls and the allocation of non-competing imports, the composite strategic inputs will require in addition non-competing imports in their production and will take on an ‘East Asian’ institutional character. The standard trade model extended along these lines provides a framework for examining the sequencing of trade policy reform. It is critical to all the arguments that follow that the presence of technological externalities in the production of the tradeables arising from the strategic input introduces a distortion into the competitive pricing process. This can be looked at from two aspects. From the point of view of the producers of the strategic inputs, the technological externalities cannot be captured in a competitive price and there will be

62

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

under-investment in the provision of the strategic inputs. From the point of view of the users of the strategic inputs, there will be benefits not included in the price of the inputs. If, in the initial situation, the resource base for the exportable is poor (the country has a low agricultural and/or mineral resource to labour ratio), a long-run development strategy will inevitably need for focus first on import-substituting industrialization and, later, on the export of industrial commodities. On the supply side, the under-provisioning of the strategic input into importables due to market failure will impede import-substituting industrialization. If the main technological externalities are to be reaped in the importable rather than the exportable sector, then the relative price of exportables to importables will be higher than the marginal rate of transformation between the two commodities. Thus, on the demand side, for any given level of provision of strategic inputs, the process of import substitution will be enhanced by the market failure. In a fully dynamic context, the underprovision of the strategic inputs acting as a brake on the industrialization process is likely to be dominant. (i) Initial situation Typically, the arguments for the sequencing of trade policy reform begin with an initial situation in which there has been a period of import-substitution characterized by a relative lack of coordination of the industrialization process and extensive use of direct import quotas. For simplicity, suppose that there are quota allocations of competing imports and non-competing imports, and that there is a net protective effect to the importables sector as a result of the rents created by the two types of quotas. The quota premium on noncompetitive inputs to the exportable sector acts as a subsidy on that sector, and the rent on the competing import quota complements the premium on non-competing inputs. The processes by which the quotas are allocated may be characterized by extensive rent-seeking. (ii) Trade policy reform on neo-liberal lines Given the above set of initial conditions, it is not difficult to design a set of trade policy reforms that are welfare-enhancing in the long run, discounting transition costs. By initially transforming the quota premiums to tariff equivalents, both rent-seeking behaviour may be reduced and government revenue enhanced as traditional and non-traditional exports expand. Further reform by lowering tariffs and moving towards freer trade will lower the consumption costs of protection, and any residual production costs of protection of the importables sector may be made more commensurate with longer-run infant industry protection. Revenue constraints permitting, this implies targeted subsides to deal with ‘viable’ infant industry protection. If the neo-liberal argument, that imperfect markets are better than imperfect states, is taken seriously, the faster is the move towards complete liberalization of the tariff structure with no targeted subsidies for ‘viable’ infant industry protection, and the greater is the welfare improvement. There is no scope seen for reformed interventionist policies to enhance the provision of strategic inputs into either the importable or the exportable sector; private agents are left to deal with the imperfect markets for coordination and two-way information flows.

IMPORT CONTROLS AND TRADE POLICY REFORM

63

(iii) Reforming the trade policy regime with enhanced strategic input provision The obvious counter-example to the neo-liberal sequencing argument is to reform the quota allocation system by tying the provision of strategic input allocation to performance criteria, along the lines of the East Asian experience. If the allocation of input quotas can be transformed into the nexus for sectoral coordination and two-way information flows, there may be increased allocation of labour to the coordination of production in the importable or industrial sector. However, given the enhanced production possibilities in the production of importables (at the expense of traditional exportables, so that the modified production possibilities frontier crosses that for the initial situation), it is perfectly possible for the reformed quota allocation system to be superior both to the initial with-controls situation and the ‘reformed’ freer trade situation. This outcome is more likely when parts of the industrial sector can quickly expand, using available capacity, into export production, and where non-traditional exports can be encouraged. Moreover, the allocation of quotas, according to export performance or genuine ‘ex ante’ foreign exchange savings through efficient import substitution, will lead to a set of domestic prices that exactly mirrors world prices in the ideal case. Thus, if from the initial situation, quotas for non-competing imports are only allocated to exportable production, a balance of payments surplus will quickly emerge. As administrative arrangements adjust the allocation of foreign exchange to this situation, an initial allocation of some of the non-competing inputs to importable industrial sector production will permit some domestic production of the importable. If there are no quota allocations of competing imports to the importables sector, there will be a large domestic price premium over world prices for importables. The administered foreign exchange regime will exactly match world prices if competing import quotas are relaxed to remove the premium over world prices, the quota premiums on non-competing imports, and any incentives for rent-seeking behaviour. The closer the initial import-substituting process was to the point where some importable industrial products could become exportable, then the change to the allocation of non-competing import quotas by export performance criteria may be associated with a change in the pattern of trade. THE EAST ASIAN EXPERIENCE: SCOPE FOR REPLICATION IN AFRICA? The above theoretical argument is meant to highlight what is increasingly widely accepted: that the East Asian case is a special one in which there was a significant role both for government and sectoral institutions to create significant technological externalities through the provision of strategic inputs. The immediate question arises for replication in Africa: is it reform package (ii) which is most likely to be welfareenhancing, or is there some scope for adapted institutional reform to generate an outcome more like (iii). There is no easy answer to this question, since it depends so much on the particular conditions pertaining in any particular country. However, there are two general considerations that may encourage the search for individual country solutions, which have more of the shape of (iii) than of (ii): (a) Because of the difficulties in estimating the tariff equivalents of import controls, it is in fact very difficult to know what the reformed tariff rates should be. This will in turn make it very hard to estimate or predict what an appropriate new equilibrium exchange rate should be, whether or not it is to be administered or completely market-determined. (b) In the light of (a), it is likely to be very difficult to prevent a very rapid draining of foreign exchange reserves in the first stage of the reform process. This is particularly likely where it may be possible to use the removal of import restrictions to facilitate capital movements through transfer pricing when capital market restrictions remain intact.

64

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

The combined force of these three considerations may mean that there is a degree of unity in government opposition to trade policy reform when posed in terms of the standard sequencing arguments. Responsible and capable officers of finance ministries and central banks may find themselves in alliance with ‘rentseeking’ elements of other ministries and their political masters against trade policy reform that removes administered quota allocations, particularly on the grounds of consideration (b) above. Then an alternative sequencing of the reform process, by allocating the ‘rents’ from quota allocations according to export or ex ante foreign exchange saving performance, accompanied by an administrative and political reform coalition within government which excludes rent-seeking interests, may be able to sufficiently improve the allocation of strategic inputs to obtain welfare improvements more on the lines of outcome (iii) than (ii). NOTES 1 There is widespread recognition given to this literature outside the neo-liberal tradition. See, for example, Bardhan (1988), Evans (1989a) and Colclough and Manor (1991, forthcoming). 2 There is by now a considerable literature which identifies unproductive or socially sub-optimal behaviour by profit-seeking firms arising from their choice of technology, considered in the widest sense, and from their profitmaximizing strategies at the point of production. For a recent survey of this literature, see Evans (1989a, section 4). 3 The arguments in this section are based on Evans (1989b, chapter 8). See also Michaely (1977, 1986). 4 See, for example, the view of a minimalist state in relation to trade and industrial policy set out in the World Bank World Development Report, 1987, Ch. 4. 5 See for example Bardhan (1984) and White and Wade (1988). My own attempt to summarize this literature is in Evans (1989a, section 4.3, and 1989b, ch 8.3). See also Amsden (1989). 6 This view is reflected in World Bank (1987, Box 4.4, p. 71) where some of the debates on industrial policy are surveyed. For an elaboration of the role of the East Asian developmental state in developing technological capability through a selective industrial policy, see Pack and Westphal (1986). 7 See, for example, the discussion in the World Bank (1987, Box 4.4 and p. 71), where the text discussion appears to side with the neo-liberal view of East Asian industrial policy against the interventionist arguments summarized in Box 4.4. 8 See for example the discussion in Leudde-Neurath (1986, 1988), Wade (1988) and Amsden (1989). 9 For a review of the current state of trade and industrial policies, see Helleiner (1991, forthcoming). 10 There is, of course, an enormous literature on this. But recent contributions by the MIT Commission on Industrial Productivity (see Dertouzoz, Lester and Solow (1989)) has increased the respectability of a more microeconomic bottom-up approach. 11 For an application of some of these ideas to a small developing country, see Murray et al. (1987). 12 One of the features of the literature on the Developmental State discussed above is the role of ideology in curtailing and limiting rent-seeking behaviour. See also Gerschenkron (1966), North (1981) and White (1984).

REFERENCES Amsden, A. (1989) Asia’s Next Giant: South Korea and Late Industrialisation, Oxford: Oxford University Press. Bardhan, P.K. (1984) The Political Economy of Development in India, Oxford: Blackwell. Bardhan, P.K. (1988) ‘Alternative approaches to development economics’, in H.Chenery and T.N.Srinivasan (eds) Handbook of Development Economics, Vol. I, Amsterdam: North-Holland. Best, M. (1990) The New Competition: Institutions of Industrial Restructuring, Cambridge: Polity Press. Bhagwati, J.N. (1982) ‘Directly unproductive profit-seeking (DUP) activities’, Journal of Political Economy 90 (5) (October): 988–1002.

IMPORT CONTROLS AND TRADE POLICY REFORM

65

Bhagwati, J. and Srinivasan, T.N. (1982) ‘The welfare consequences of directly-unproductive profit-seeking (DUP) activities: price vs quantity distortions’, Journal of International Economics13:33–44. Chenery, H. and Srinivasan, T.N. (eds) (1988 and 1989) Handbook of Devel-opment Economics, Vols I and II, Amsterdam: North-Holland. Colclough, C. and Manor, J. (eds) (1991, forthcoming) States or Markets? Neo-Liberalism and the Development Policy Debate, Oxford: Oxford University Press. Dertouzos, M.L., Lester, R.K. and Solow, R.M. (1989) Made in America: Regaining the Productive Edge, Cambridge, Mass. and London: MIT Press. Evans, H.D. (1989a), ‘Alternative perspectives on trade and development’, in H.Chenery and T.N.Srinivasan (eds) (1989) Handbook of Development Economics, Vol. II, Amsterdam: North-Holland Evans, H.D. (1989b) Comparative Advantage and Growth: Trade and Development in Theory and Practice, Hemel Hempstead: Harvester-Wheatsheaf Press. Gerschenkron, A. (1966) Economic Backwardness in Historical Perspective: A Book of Essays, Cambridge, Mass.: The Belknap Press of Harvard University Press. Helleiner, G.K. (ed.) (1991, forthcoming) Trade Policy, Industrialization and Development: New Perspectives, Oxford: Clarendon Press. Krueger, A.O. (1974) ‘The political economy of rent seeking society’, American Economic Review, 64 (June): 291–303. Leudde-Neurath, R. (1986) Import Controls and Export Oriented Development: A Re-assessment of the South Korean Case, Boulder, Colo., and London: Westview Press. Leudde-Neurath, R. (1988) ‘State intervention and export orientated development in South Korea’, in G.White(ed.) Developmental States in East Asia, London: Macmillan. Michaely, M. (1977) Theory of Commercial Policy: Trade and Protection, Oxford: Philip Allan. Michaely, M. (1986) ‘The timing and sequencing of trade policy reform’, in A.M.Chokin and D.Papageorgiou (eds), Economic Liberalization in Developing Countries, Oxford: Blackwell. Murray, R. (1987) Main Report, Cyprus Industrial Strategy Mission, Brighton: Institute of Development Studies, University of Sussex, prepared for UNDP/UNIDO. North, D. (1981) Structure and Change in Economic History, New York: Norton. Pack, H. and Westphal, L.E. (1986) ‘Industrial strategy and technological change: theory versus reality’, Journal of Development Economics 22:87– 128. Wade, R. (1988) ‘State intervention in “outward-looking” development: neo-classical theory and Taiwanese practice’, in White ed (1988). White, G.M. (1984) ‘Developmental states and socialist industrialisation in the Third World’, Journal of Development Studies21(1) October: 97–120. White, G. (ed.) (1988) Developmental States in East Asia, London: Macmillan. White, G. and Wade, R. (1988) ‘Developmental states and markets in East Asia: an introduction’, in G.White (ed.) Developmental States in East Asia, London: Macmillan. World Bank (1987) World Development Report, New York: Oxford University Press.

4 The effect of trade liberalization on industrial-sector productivity performance in developing countries Colin Kirkpatrick and Jassodra Maharaj

INTRODUCTION The role of trade policy in the process of industrial growth and economic development continues to be widely debated. In the first two postwar decades, the production of import substitutes was seen as the main avenue for rapid industrial growth, and trade policy became the major instrument of industrial planning. Most developing countries created high protective barriers to international competition, using tariffs, quantitative import restrictions and exchange rate controls. The protective measures were often reinforced by domestic market policy interventions in the form of selective credit policy, licensing and price controls. With the re-emergence in the late 1960s of neo-classical orthodoxy as the dominant paradigm, policymakers were encouraged to shift from import-substituting to export-oriented industrial strategies. An outward looking, export-based policy stance came to be seen as a necessary catalyst for improved industrial performance and rapid growth in the developing countries. The emphasis on adopting a more open and less regulatory trade regime continued in the 1980s, and trade policy has been a key component of the adjustment programmes adopted by many developing countries in recent years. Trade reform has featured prominently in the conditionality attached to World Bank lending for structural adjustment: during the period 1979–89 the Bank, often in conjunction with the International Monetary Fund, supported trade policy reforms in forty-four countries through ninety-eight adjustment loans containing substantial proposals for trade policy reform (Thomas and Nash, forthcoming). Table 4.1 shows the proportion of World Bank structural adjustment loans (SALs) containing one or more conditions in various policy areas. For the total sample of 51 Table 4.1 Content of World Bank structural adjustment lending operations, 1980–7 Percentage of total number of loans with conditions in various policy areas

1 Exchange rate 2 Trade policies 3 Fiscal policy 4 Budget/public expenditures 5 Public enterprises

SubSaharan Africa Highly indebted (13) middleincome countries (22)

Other developing countries (16)

All countriesa (51)

30.8 76.9 61.5 69.2

18.2 90.0 72.7 50.0

0.0 62.5 56.3 37.5

15.7 78.4 64.7 51.0

61.5

54.5

43.8

52.9

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

67

Percentage of total number of loans with conditions in various policy areas SubSaharan Africa Highly indebted (13) middleincome countries (22) 6 Financial sector 38.5 36.4 7 Industrial policy 53.8 9.1 8 Energy policy 7.7 13.6 9 Agricultural policy 76.9 40.9 10 Other 23.1 9.1 Source: World Bank (1988), Table 4.2. Note: a Numbers in parentheses are total numbers of loans.

Other developing countries (16)

All countriesa (51)

43.8 25.0 50.0 37.5 12.5

39.2 25.5 23.5 49.0 13.7

countries, trade policy reform was required in almost 80 per cent of programmes; in the highly indebted middle-income countries, the share increases to 90 per cent. Table 4.2 shows the specific areas of trade policy reform included in structural adjustment loans received by 40 countries. Substitution of quantitative restrictions by tariffs, liberalizing imported inputs for exports, and reduction in general protection levels have all been widely used components. The record of implementation, based on 51 adjustment loans (both SALs and Sector SALs) in 15 developing countries is shown in Table 4.3. The table is based on a judgemental grading, derived from country reports and verbal information supplied by persons with knowledge of the countries. A noteworthy feature of Table 4.3 is the relatively low degree of full implementation in the area of trade policies. Of the nine areas for policy reform identified in the table, the percentage grade for trade reform is among the three lowest figures, both during and after the loan period. This suggests that the implementation of policy reforms in the area of trade may be more difficult

Source: World Bank (1988), Table 3.1. Note: The assessments refer to proposals supported by the World Bank. They do not necessarily refer to policy implementation. a Judgement on the significance of the overall reform proposals. b Often these were not explicit conditions, but constituted understandings, frequently made under the programme. c Includes such schemes as export credits, insurance, guarantees, and institutional development.

Table 4.2 Main components of trade policy in forty adjustment loan countries

68 FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

69

Table 4.3 Implementation of conditionality Percentage of conditions fully implemented 1 Exchange rate 2 Trade policies Import QRs Import duties Import/export finance Export incentives Other trade policies 3 Fiscal policy Tax policy 4 Budget/public expenditure 5 Public enterprise reforms (incl. restructuring) 6 Financial sector 7 Industrial policy (excl. restructuring) 8 Energy policy Energy pricing 9 Agriculture policy Agricultural pricing All conditions, Total All conditions, SALs All conditions, SECALs All conditions, SSA countries All conditions, HICs All conditions, other developing countries Source: World Bank (1988), Table 1.3.

During the loan period

Current situation

70.0 54.9 62.8 61.5 20.0 60.6 33.3 53.2 46.2 68.0 61.3 71.4 53.3 79.2 84.6 57.1 64.3 60.3 68.3 50.9 52.4 66.9 52.8

62.5 63.4 69.0 72.7 42.9 62.5 41.1 69.8 86.7 71.7 70.0 73.5 42.9 83.3 100.0 58.1 61.5 67.5 73.5 60.0 62.2 73.2 56.0

than in the other areas of the economy. Thomas and Nash (forthcoming) suggest four sets of domestic factors as constraints on more sustained reform in trade policy. First, a weak commitment to reform, stemming from inadequate conviction concerning its benefits; second, inadequate implementation capacity, resulting from administrative and institutional bottlenecks; third, weak macroeconomic perform ance and conflicts between trade policy reform and stabilization goals; fourth, lags in supply response to policy reform. Despite the enthusiasm of its advocates, the results of the trade policy reforms carried out under structural adjustment programmes have been less successful than hoped for. The World Bank’s internal report on the impact of its trade policy conditionality lending calculated the change in nine performance indicators three years after the loan compared with three years before for twenty-six pre-1986 trade loan recipients and forty-seven non-recipients comparators. The results showed that, on average, the change in performance was better for the trade loan recipient, but in the majority of cases the difference in mean

70

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

performance indicators was not statistically significant (World Bank 1989, Table 3.2). Based on this crosscountry comparative analysis, regression analysis and selected country studies, the Report concluded that ‘trade adjustment loans are associated with a mild improvement in performance compared to their absence’ (P. 39). The failure to uncover strong empirical evidence of a link between improvements in performance and the implementation of trade policy reform measures does suggest that many of the claims made in favour of trade liberalization have been exaggerated, and that the efficacy of the reform measures currently being recommended to, and adopted by, many developing countries can be questioned.1 Our understanding of the interrelationships between policy changes and industrial performance is, at best, partial, a point acknowledged in the literature, but largely ignored in the context of policy formulation. The study by Chenery et al. (1986) of the pattern of industrialization in thirty economies (18 developed and 12 developing countries) concluded that: in sum, our analysis stresses that industrialisation must be examined in several dimensions. It may well be that all the associated elements follow naturally from a policy regime of sensible incentives in an environment of free markets. However, such a conclusion is certainly not obvious…[our analysis] …does suggest that a certain modesty in claims for particular policy choices is called for (Chenery et al. 1986:358–9) More recently, a leading World Bank advocate of liberalization has noted that: The analytical framework…provides a good basis for the design of adjustment policies in most countries. But it falls short of providing a quantitative structure for linking inputs (policy actions) and outputs (macroeconomic performance)…Indeed, the framework has little to say about the quantitative impact of key components of policy reform (e.g. trade reform, improvements in pricing, public investment reviews, and the like) on medium term economic performance (on which the Bank usually focuses). (Michalopoulos 1987:46–7) Our objective here is to review the state of knowledge concerning the relationship between trade policy reforms and the possible gains in industrial performance, as measured by efficiency changes and productivity growth. The remainder of the chapter consists of three sections: a review of the empirical evidence on trade policy, static and dynamic efficiency; a discussion of the relationship between domestic market competition and productivity performance, drawing attention to the indeterminacy of the performance response to trade liberalization when the domestic industrial sector is uncompetitive in nature; a summary and concluding comments. PRODUCTIVITY GROWTH AND TRADE POLICY The analytical framework used in measuring productivity performance is based on the economic theory of production and cost. A distinction can be made between static and dynamic productivity measures, and is illustrated by referring to the standard textbook treatment of the firm’s production and cost relations shown in Figure 4.1, where PP is the efficient production frontier and is determined by the existing state of technology. If technology changes, the production frontier will shift to the left.

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

71

Figure 4.1 Productive efficiency of the firm

Productivity improvements may be due to changes in (static) efficiency or changes in (dynamic) technical progress or productivity. The concept of static efficiency gains is shown in Figure 4.1 by comparing points A, B and C. The firm producing at A is inefficient, since it incurs higher costs, TC3, than the firm at C with costs TC3, The level of inefficiency can be decomposed into (i) the cost of technical inefficiency, TC3−TC2, which is due to low factor productivity as compared to firms at B on the efficient isoquant employing the same capital-output ratio, and (ii) the cost of allocative inefficiency, TC2 −TC1, due to the choice of the wrong technique at existing relative factor prices. Static efficiency gains will be measured, therefore, by a move from A towards C. A gain in dynamic productivity or what is usually referred to as technical progress, will occur when the isoquant shifts towards the origin. In practice, an observed change in total factor productivity will consist of changes in both technological progress and technical and allocative efficiency.2 Static efficiency gains and trade policy Much of the discussion of the gains from trade policy reform has focused on the once-and-for-all benefits that would result from the reduction in static resource inefficiencies associated with the restrictions on trade. A considerable body of evidence has been accumulated on the technical and allocative inefficiencies resulting from tariffs and other protectionist measures used to encourage import-substituting industrialization, with domestic resource cost and effective protection rate measures used to establish the magnitude of these costs. Early studies used calculations of the effective rate of protection to calculate the total inefficiency costs for the economy as a whole.3

72

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

Typically, these estimates suggest that the deadweight ‘welfare triangle’ cost of protection, measured in terms of its impact on GNP, is small, although as Diaz-Alejandro (1975) observed: The standard model can be made to generate hypothetical situations in which the costs of protection and self-sufficiency loom large… Faced with the alternatives of strict adherence to the pure, model yielding small quantitative effects or adding to it epicycles that make free trade look quantitatively better, most authors have chosen the latter.4 (Diaz-Alejandro 1975:98) Few of these studies of the total efficiency costs of protection have distinguished between the technical and allocative inefficiency components, assuming instead that the estimates indicate allocative inefficiency only. Page (1980), however, has shown how the domestic resource cost measure can be expressed in terms of technical and allocative efficiency, and provides estimates for each for Ghana (Page 1980) and India (Page 1984). The earlier study by Bergsman (1974) for Malaysia, Mexico, Brazil, Philippines and Pakistan separated allocative inefficiency costs from X-inefficiency, which can be interpreted as technical inefficiency (i.e. where production costs exceed the ‘best practice’ level of costs). Will these (relatively small) static inefficiency costs be reduced or eliminated by liberalizing the trade regime? As noted by Bhagwati and Srinivasan (1979:11–12) the repercussions of a change in the trade policy on different activities cannot be forecast from a simple examination of the relative DRCs or ERPs in an initial, protected situation. Estimates of ERP and DRC can only indicate the potential gains from resource reallocation: ‘even if effective rates give a reasonable indication of the direction of resource pulls they cannot on their own tell one what the responses to the pulls will be’ (Corden 1975:66). The actual gains in improved allocative and technical efficiency will depend upon the behavioural reaction of firms to the lowering of protection. The response of domestic producers to a liberalization of the trade regime is discussed in detail later (see ‘Domestic competition and productivity performance’, p. 103). Dynamic efficiency and trade policy Confronted by the evidence of static inefficiencies, proponents of protectionism would argue that the efficiency losses are offset by the dynamic productivity gains generated by import-substituting industrialization. The argument is commonly advanced in terms of infant industry protection, which is based on the proposition that rates of productivity change will be higher in the protected infant industries, thereby permitting these industries to catch up with their international competitors. The competing hypothesis is that exposure to international competition raises the rate of productivity growth, forcing domestic producers to achieve international performance standards:5 trade liberalisation results in the contraction of inefficient sectors and the expansion of new, efficient ones. Over time, a new and more efficient production structure develops that will be better suited to the international environment’. (Michalopoulos 1987:24) Nishimizu and Robinson (1986) identify three hypotheses linking growth and trade policies. The first is a positive link between export expansion (or increased import-substitution) and total factor productivity (TFP) growth arising from Verdoorn’s law and the growth of market size. Second, there is a positive link with exports arising from competitive cost-reducing pressures. A third positive link arises from the

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

73

relaxation of the foreign exchange constraint and increased availability of non-substitutable imports of intermediate and capital good inputs. There is a substantial body of literature which shows that TFP growth has been a significant contributory factor in the growth of developing countries. The Chenery et al. (1986) analysis of 30 countries (18 developed, 12 developing) classified economies into three groups distinguished by the rate of TFP growth and by the relative contribution of TFP change to growth of value added. The ‘typical developing economies’ showed faster growth rates of TFP than the developed countries, but TFP growth accounted for less than 2 per cent of output growth. Page (1990) and Nishimizu and Page (forthcoming) have assembled data on TFP and output growth for the industrial sector in eighteen countries, covering the 1956 to 1982 period. For the industrial sector as a whole, productivity change accounted for between 10 and 30 per cent of total output growth, with more rapid rates of output growth being associated with a lower TFP contribution to output growth. Total factor productivity growth and variations in productivity performance at the two digit industry level were both found to vary inversely with income per capita level. Attempts to establish empirical confirmation of a causal relationship between trade policy and productivity growth have been less successful. The major National Bureau for Economic Research project on liberalization in developing countries, conducted by Krueger and Bhagwati in the 1970s, attempted to relate total factor productivity growth to different trade policy regimes, but failed to find any clear evidence to support the hypothesis that liberalization will stimulate productivity growth (Bhagwati 1978, ch. 5). Reviewing the evidence accumulated by this and other studies during the 1970s, Bhagwati and Srinivasan (1979), concluded that: contrary to the enthusiasm of many proponents of liberalised regimes, there is no systematic evidence on their side either of dynamic efficiencies…and no general conclusions seem warranted. (Bhagwati and Srinivasan 1979:14) Individual country studies have also failed to find systematic evidence of an association between trade policy orientation and productivity growth.6 Nishimizu and Robinson (1986) found that TFP growth was more rapid in the export-oriented Korean economy than in the internally oriented economies of Turkey and Yugoslavia, but, as the authors acknowledge, the expansion of exports may have been caused by productivity growth, rather than the other way round. Tsao’s study of Singapore, probably the most open of all developing countries, found that productivity growth at the two digit level in the manufacturing sector during the 1970s was negligible, with many branches showing negative productivity growth (Tsao 1985). A study by Handoussa, Nishimizu and Page (1986) found that the majority of Egyptian industries recorded considerable improvements in total factor productivity after trade liberalization in 1973, but the gains were of the static, once-and-for-all type, rather than technical progress. The study by Nishimizu and Page (forthcoming) attempts to relate cross-country variations in TFP growth for the period 1956–83, to differences in the extent to which domestic producers were protected from import competition by quantitative restrictions on industrial imports. The results show that, prior to 1973, TFP growth in economies using quantitative trade restrictions was similar to that in more open market economies, but, after 1973, the more closed economies exhibited lower TFP growth rates.7 Pack (1988) provides a succinct summary of the empirical evidence on the relationship between trade policy orientation and manufacturing productivity growth in developing countries: Comparisons of total factor productivity growth among countries pursuing different international trade orientations do not reveal systematic differences in productivity growth in manufacturing, nor do

74

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

the time-series studies of individual countries that have experienced alternating trade regimes allow strong conclusions in this dimension. (Pack 1988:327) DOMESTIC COMPETITION AND PRODUCTIVITY PERFORMANCE The positive relationship between productivity growth and trade policy, as hypothesized by advocates of liberalization, rests on an implicit ‘challenge-response’ mechanism whereby international trade forces domestic industries to improve their efficiency performance and to adopt new technologies. Early literature argued that trade liberalization would challenge the monopoly rents created under protectionism and force entrepreneurs to abandon the ‘quiet life’ reflected in X-inefficiency (Leibenstein 1966).8 It is only in recent years however, that attention has begun to focus on the way in which the domestic producers’ supply response will be affected by the structure of the industrial sector market and by the domestic regulatory framework. The ‘new theory’ of international trade marries the insights of industrial economics to those of international trade, and provides a new perspective to the efficacy of trade policy reform in promoting cost reductions and productivity changes. By concentrating on the typically uncompetitive nature of the domestic industrial sector, this literature draws attention to the indeterminacy of the performance outcome under trade liberalization. The ‘new theory’ offers few clear-cut predictions: ‘it all depends on how the economy is expected to adjust, which in turn depends on the frustrating ambiguities of oligopoly theory’ (Rodrik 1988:110). The available evidence of market structure in LDCs suggests that the industrial sector can be characterized by conditions of imperfect competition. Aggregate market concentration measures show that the level of concentration in LDCs is generally higher than in industrialized countries.9 Kirkpatrick et al. (1984) provide a summary of country studies of market concentration at the industry level.10 There are the usual problems of comparability and data reliability, but a number of broad features of seller concentration can be identified (Lee 1988). In general, the levels of establishment and firm concentration measures tend to be higher in LDCs than in developed countries.11 This suggests that LDCs’ markets are typically more oligopolistic in structure than in developed countries. Second, LDCs with small domestic markets tend to have higher seller concentration levels than LDCs with large domestic markets. Third, the ranking of markets according to level of seller concentration tends to be broadly similar in different countries. The contestable markets’ literature argue that oligopolistic or monopoly producers may be driven to adopt competitive price and output levels if there is an effective threat of entry from other firms. The main issue, therefore, is not market concentration but the degree to which there are natural or policy related barriers to entry. Economies of scale, product differentiation, capacity licensing and other market reservation policies, can all act as barriers to entry. Direct evidence on the importance of scale economies in specific industrial sectors in developing countries is virtually non-existent, but the small internal markets of many LDCs, combined with the domestic orientation of most industries, would suggest that economies of scale often are not fully exploited. Licensing and protection policies typically combine to encourage firms to operate below full capacity. If, then, the industrial sector in developing countries is characterized by imperfectly competitive conditions, in what ways will this affect the technical efficiency and productivity performance of domestic producers following trade liberalization? A commonly made argument for trade liberalization maintains that protection tends to lead to too many firms producing at low output levels.12 Liberalization, therefore, will encourage rationalization, by forcing inefficient firms out of business. As Rodrik (1988b) notes, this industry-rationalization argument is based on

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

75

the assumptions of economies of scale and freedom of entry and exit. With the lowering of protection, the market price is reduced, some firms are forced out, and the remaining firms have to move down their average cost curve to produce at a sufficient scale of output for the reduced level of average costs to match the lower domestic price. With freedom of entry and exit, and assuming firms can adjust rapidly to the removal of protection, domestic firms can expand their output to meet the increase in demand following the reduction in domestic price. With restricted entry and exit the case for trade liberalization takes a different form, and relies on the ‘import discipline’ hypothesis. Exposure to increased imports reduces the market power of domestic producers and affects their pricing and production decisions accordingly. Efficiency changes take the form of technical efficiency improvements, with domestic firms moving to a lower cost curve. Firms that are unable to compete by lowering their costs will be displaced by imports. Collusion among domestic producers may block the efficiency gains from trade liberalization. If, for example, importers are also major sellers, the increase in imports may increase seller concentration. If sellers are able to maintain domestic prices at their existing levels, the lower cost of imported supplies will increase profitability. A similar outcome will emerge if there is collusion between domestic producers and importers. Domestic producers may also be able to influence the pattern and degree of implementation of trade liberalization to leave their domestic rents largely unaffected.13 Published research on the effect of trade reform on industrial market structure and firms’ efficiency performance is limited. The study by de Melo and Urata (1986) examined the effect of trade liberalization in Chile in the mid-1970s, and shows that the reforms increased concentration but reduced profitability, suggesting both that economies of scale were realized and that oligopolistic interaction diminished. The evidence also suggested that the exploitation of economies of scale was greatest for sectors that experienced the largest reduction in effective protection.14 The study also found evidence in support of the importdiscipline hypothesis, namely that sectors that had the highest import-penetration ratios also showed the largest decline in price-cost margins. Current research by the World Bank seeks to uncover new evidence on the protection-efficiency issue by taking a more disaggregated approach than previous studies. Firm level data have been used to examine the impact of trade liberalization on efficiency performance in the context of imperfectly competitive markets. Tybout, de Melo and Corbo (1989) consider two aspects of efficiency—scale and technical—using firm level data for pre- and post-trade liberalization periods in Chile. Measuring changes in exposure to foreign competition by changes in effective protection, it was found that increased foreign competition led to industrial rationalization, with the surviving firms producing at output levels closer to minimum efficient scale. With respect to technical efficiency, the results suggest that firms moved closer to the best-practice isocost frontier. The authors are careful to point out, however, that both the scale and technical efficiency results are open to alternative plausible explanations, with, for example, the apparent exploitation of scale economies reflecting a reduction in monopoly power of large producers, or the improvement in technical efficiency reflecting a reduction in product diversity. Harrison (1989) uses a panel of firms in the Ivory Coast to examine the impact of trade liberalization on total factor productivity growth. It is shown that the standard measures of TFP will be biased if the competitive market environment shifts during the estimation period. The results show that, whereas there appears to be a strong relationship between trade reform and productivity when perfect competition is assumed in the product markets, this relationship virtually disappears when the variations in price-cost margins resulting from trade liberalization are allowed for.

76

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

CONCLUSION Our objective has been to subject the arguments commonly made concerning the trade policy reform— industrial productivity nexus to critical evaluation. Despite the frequency and fervour with which the proliberalization case is advanced, it has been seen that the theory yields few clear and unambiguous predictions regarding the link between trade policy and productivity growth. Empirical evidence is equally difficult to interpret, and gives little firm support to the hypothesis that trade liberalization is generally conducive to improvements in efficiency and productivity performance. The indeterminacy of the effect that trade liberalization has on productivity performance can be traced to uncertainty about the way in which industrial enterprises respond to the new set of incentives established by trade policy changes. The reaction of firms is conditioned by the non-competitive structure of the industrial sector market. The ‘new’ trade theory, which incorporates the industrial organization literature on economies of scale and imperfect competition, provides an array of theoretical outcomes, depending on the particular set of assumptions made. It appears, therefore, that the focus for further research should be at the country level, using disaggregated industrial sector data to identify the linkages between trade liberalization, industrial market structure and firms’ productivity performance. Comparative case study material of this kind ‘may be able to shed more light on controversies that at present feed mainly on theory and selected anecdotes’ (Helleiner 1990:892).15 If successful, such studies will assist policymakers to identify the conditions under which trade reforms are likely to result in significant improvements in performance. Conversely, they should also assist analysts in recognizing the circumstances where trade liberalization will leave efficiency performance largely unchanged. But until such information becomes available, there is little basis for claiming that trade liberalization will generally enhance industrial sector efficiency and productivity growth in developing countries. NOTES 1 Bhagwati, for example, has claimed that ‘the question of the wisdom of an outward-oriented (“exportpromoting”) strategy may be considered to be settled’ (Bhagwati 1987, quoted in Helleiner 1990). 2 Nishimizu and Page (1986) show how total factor productivity change can be decomposed into technical efficiency and changes in allocative efficiency. The procedure is applied to industrial sector data for Thailand. 3 For examples of this approach, see Harberger (1959), Krueger (1966), Balassa et al. (1971), Little et al. (1970). 4 Balassa et al. (1971) widened the cost estimates to include consumption costs and the terms of trade losses due to the reduction in export prices. 5 For an early presentation of the argument, see Keesing (1967). 6 This section draws on Pack (1988). 7 TFP growth is also found to be positively correlated with export growth, but negatively correlated with import penetration. 8 A formal presentation of the argument is given in Martin (1978). 9 Frischtak et al. (1989) reports on the following sample of market concentration estimates: Country

Year

Average of four-firm concentration ratios

Number of industries

Argentina Brazil Chile

1984 1980 1979

43 51 50

172 119 41

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

Country

10 11

12 13

14

15

Year

Average of four-firm concentration ratios

77

Number of industries

India 1984 46 n/a Indonesia 1985 56 119 Mexico 1980 48 186 Pakistan 1985 69 n/a Turkey 1976 67 125 United States 1972 40 323 This list of country studies of industrial concentration is updated in Lee (1988). Rodrik (19 88a) argues that concentration ratios proba ably understate the extent of market power enjoyed by major oligopolistic firms. First, in many LDCs there is no effective monopoly regulation policy. Second, the reliance on quantitative restrictions allows higher price-cost margins than would tariffs that generate the same level of imports. Third, the measures do not allow for the existence of conglomerates that cover a number of industries. The argument that protection leads to excessive entry is developed in Eastman and Stykolt (1962). This condition sits uncomfortably with the notion that there are major barriers to entry in the industrial sector. See Haggard and Kaufman (1989), World Bank (1989), Chapter 4. Kirkpatrick and Onis(1991) discuss the case of Turkey, where trade liberalization policy was formulated to shield large-scale enterprises from the potentially adverse effects of sudden exposure to foreign competition. However, as Rodrik (1988a) notes, the de Melo and Urata (1986) study excludes almost half of the existing firms on the grounds that they did not cover average costs. This leaves unanswered therefore, the ultimate position of these enterprises. ‘until more evidence becomes available then, a healthy scepticism is in order. In the meantime, if truth-inadvertising were to apply to policy advice, each prescription for trade liberalisation would be accompanied with a disclaimer: “Warning! Trade liberalisation cannot be shown to enhance technical efficiency: nor has it been empirically demonstrated to do so”.’ (Rodrik 1988a:28).

REFERENCES Balassa, B. and Associates (1971) The Structure of Protection in Developing Countries, Baltimore, Md: Johns Hopkins University Press. Bergsman (1974) ‘The costs of protection and X-efficiency’, Quarterly Journal of Economics 88:909–33. Bhagwati, J. (1978) Foreign Trade Regimes and Economic Development: Anatomy and Consequences of Exchange Control Regimes, Cambridge, Mass.: Ballinger. Bhagwati, J. (1987)‘Outward orientation: trade issues’ in V.Corbo, M.Goldstein and M.Khan (eds) Growth-oriented Adjustment Programs, Washington DC: IMF-World Bank. Bhagwati, J. and Srinivasan T.N. (1979)‘Trade policy and development’, in R.Dornbusch and J.A.Frenkel (eds) International Economic Policy: Theory and Evidence, Baltimore, Md: Johns Hopkins University Press. Chenery, H., Robinson, S. and Syrquin M. (eds) (1986) Industrialisation and Growth: A Comparative Study, Oxford: Oxford University Press. Chenery H., and Srinivasan, T.N. (eds) (1988, 1989) Handbook of Development Economics, vols 1 and 2, Amsterdam: North-Holland. Corbo, V., Goldstein, M. and Khan, M. (eds) (1987) Growth-Oriented Adjustment Programs, Washington DC: IMFWorld Bank. Corden, 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. de Jong, H.W. and Shepherd, W.T. (eds) (1986) Mainstreams in Industrial Organisation, Dordrecht: Kluwer.

78

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

de Melo, J. and Urata, S. (1986) ‘The influence of increased foreign competition on industrial concentration and profitability’, International Journal of Industrial Organisation 4:287–304. Diaz-Alejandro, C.F. (1975) ‘Trade policies and economic development’, in Kenen, P.B. (ed.) International Trade and Finance: Frontiers for Research, Cambridge: Cambridge University Press. Eastman, H. and Stykolt, S. (1962) ‘A model for the study of protected oligopolies’, Economic Journal7:336–47. Frischtak, C., with B.Hadjimichael and O.Zachan (1989) ‘Competition poli cies for industrializing countries’, Policy and Research Series No. 7, Washington DC: World Bank. Haggard, S. and Kaufman, R. (1989) ‘Notes on the politics of stabilisation and trade reform’ (mimeo), Paper presented at the Conference on Structural Adjustments and Policy Reform, Brighton: Institute of Development Studies, University of Sussex. Handoussa, H., Nishimizu, M. and Page, J. (1986) ‘Productivity change in Egyptian public sector industry after ‘the opening’, 1973–9’. Journal of Development Economics, 21:1. Harberger, A. (1959) ‘Using the resources at hand more effectively’, American Economic Review, Papers and Proceedings XLIX: 134–46. Harrison, A. (1989) ‘Productivity, imperfect competition and trade liberalisation in the Côte d’Ivoire’ (mimeo), Washington DC: World Bank. Helleiner, G.K. (1990) ‘Trade strategy in medium-term adjustment’, World Development 18, 6:879–98. Keesing, D. (1967) ‘Outward-looking policies and economic development’ Economic Journal LXXVII, 306:303–20. Kenen, P.B. (ed.) (1975) International Trade and Finance: Frontiers for Research, Cambridge: Cambridge University Press. Kirkpatrick, C., Lee, N. and Nixson, F. (1984) Industrial Structure and Policy in Less Developed Countries, London: Allen and Unwin. Kirkpatrick, C. and Onis, Z. (1991) ‘Structural adjustment lending and policy reform in Turkey, 1980–6’, in Mosley P., Harrigan, J. and Toye, J. (eds), Aid and Power: The World Bank and Policy-Based Lending in the 1980s, London: Routledge. Krueger, A. (1966) ‘Some economic costs of exchange control: the Turkish case’, Journal of Political Economy LXXIV: 466–80. Krueger, A. (1978) Liberalization Attempts and Consequences, Cambridge, Mass.: Ballinger. Lee, N. (1988) ‘Market structure and trade in developing countries’ (mimeo), Helsinki: WIDER. Leibenstein, H. (1966) ‘Allocative vs. X-efficiency’, American Economic Review 56:382–415. Little, I., Scitovsky, T. and Scott, M. (1970) Industry and Trade in Some Developing Countries, London: Oxford University Press. Martin, J. (1978) ‘X-efficiency, managerial effort and protection’, Economica 45:273–86. Michalopoulos, C. (1987) ‘World Bank programs for adjustment and growth’ in V.Corbo, M.Goldstein and M.Khan (eds) Growth-oriented Adjustment Programs, Washington DC: IMF-World Bank. Mosley, P., Harrigan J. and Toye J. (1991) Aid and Power: The World Bank and Policy-Based Lending, London: Routledge. Nishimizu, M. and Page, J. (1986) ‘Productivity change and dynamic comparative advantage’, Review of Economics and Statistics 68:241–7. Nishimizu, M. and Page, J. (forthcoming) ‘Trade policy, market orientation and productivity in industry’, in J.de Melo (ed.) Trade Theory and Economic Reform: North, South and East. Essays in Honour of Bela Balassa. Nishimizu, M. and Robinson, S. (1986) ‘Productivity growth in manufacturing’ in H.Chenery, S.Robinson and M.Syrquin (eds) Industrialisation and Growth: A Comparative Study, Oxford: Oxford University Press. Pack, H. (1988) ‘Industrialisation and trade’, in H.Chenery and T.N.Srinivasan (eds) Handbook of Development Economics, Vol. 1, Amsterdam: North-Holland. Page, J. (1980) ‘Technical efficiency and economic performance: some evidence from Ghana’ OxfordEconomic Papers 32:319–39. Page, J. (1984) ‘Firm size and technical efficiency: applications of production frontiers to Indian survey data’, Journal of Development Economics 16:129–52.

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

79

Page, J. (1990) ‘The pursuit of industrial growth: policy initiatives and economic consequences’, in M.Scott and D.Lal (eds) Public Policy and Economic Development: Essays in Honour of Ian Little, Oxford: Clarendon Press. Roberts, M. and Tybout, J. (1990) ‘Rationalisation and trade exposure in developing countries’ (mimeo), Washington DC: World Bank. Rodrik, D. (1988a) ‘Imperfect competition, scale economies, and trade policy in developing countries’, in R.E.Baldwin (ed.) Trade Policy Issues and Empirical Analysis Chicago: University of Chicago Press. Rodrik, D. (1988b) ‘Closing the technology gap: does trade liberalisation help?’ Helsinki: WIDER. Thomas, V. and Nash, J. (forthcoming) ‘Trade policy reform: recent evidence from theory and practice’, in R.Adhikari, C.Kirkpatrick and J.Weiss (eds) Industrial and Trade Policy Reform in Developing Countries, Manchester: Manchester University Press. Tsao, Y. (1985) ‘Growth without productivity: Singapore manufacturing in the 1970s’, Journal of Development Economics 19:25–38. Tybout, J., de Melo, J. and Corbo, V. (1989) ‘The effects of trade reforms on scale and technical efficiency: new evidence from Chile’ (mimeo), Washington DC: World Bank. World Bank (1988) Adjustment Lending: An Evaluation of Ten Years of Experience, Policy and Research Series No. 1, Washington DC: World Bank. World Bank (1989) Strengthening Trade Policy Reform, volume II (restricted distribution), Washington DC: World Bank.

5 Trade reform and growth resumption in Latin America José María Fanelli, Roberto Frenkel and Guillermo Rozenwurcel

INTRODUCTION In this chapter, we examine the issues related to the problem of restoring growth in Latin America from the standpoint of economic policy. Consequently, one of our main purposes is to extract some conclusions from the experience of adjustment-with-stagnation experienced by most Latin-American countries throughout the 1980s. To fulfil this objective, it is almost as important to examine the policy reform proposals that are being made by the World Bank, the IMF and the literature produced by the Washington think-tanks, which gave rise to the ‘Washington Consensus’ about policy reform (Williamson 1990), as it is to analyse the structural disequilibria that arose during the adjustment period of the 1980s. These proposals are strongly determining the direction of the actual reforms that the most important countries in the region are putting into practice. Therefore, part of this chapter is devoted to assessing the suitability of the theoretical framework that supports the ‘Washington’ diagnosis and its concrete policy reform proposals as well. Besides the analysis of the actual economic disequilibria and the evaluation of the ‘Washington Consensus’, a third purpose of the chapter is to advance and evaluate some policy reform proposals aimed at restoring growth. According to the authors who share the Washington Consensus, the roots of the present Latin-American structural imbalances lie in the postwar development model, which assigned a leading role to the important substitution strategy of industrialization and to the state as an ‘engine of growth’. Consequently, the policy proposals are primarily oriented both to removing the structural and institutional features of the LatinAmerican economies that comply with that development strategy and to developing a new market-andoutward oriented strategy of growth. The first section discusses this diagnosis and the policy proposals based on the market-oriented strategy and evaluates some of the weaknesses that are present in this approach as well. We have decided to discuss the liberalization paradigm in detail here, not only because, nowadays, it is the most influential development paradigm, but also because some of the criticism that has been made of the postwar development process in terms of resource misallocation is correct, independently of our judgement on the theoretical underpinnings of the whole liberalization paradigm. From our point of view, there are two flaws in the market-oriented analysis and proposals, which should be highlighted. First, they are too abstract and general. Second, their diagnosis does not take into account some new structural features developed by the Latin-Amer-ican economies in the 1980s. These features did not originate in the weaknesses of the import-substitution strategy, but rather in the dynamics of the adjustment to the external shock that took place at the beginning of the decade. In fact, we consider that the principal constraints to growth today originated in the long-lasting features of the external and fiscal imbalances induced by the debt crisis that have still not reversed after ten years of adjustment. Therefore, the second

TRADE REFORM IN LATIN AMERICA

81

section addresses the analysis of the evolution of the external and fiscal gap during the 1980s. The analysis takes into account the empirical data of five representative Latin-American countries (Argentina, Chile, Colombia, Brazil and Mexico). On the basis of this analysis, we examine the transmission and propagation mechanisms operating at a macroeconomic level that contribute to the spillover of the destabilizing trends originating in the fiscal and external disequilibria. The last section presents and discusses some policy reforms for achieving adjustment with growth that our analysis suggests as suitable, given the constraints to growth that Latin-American countries are now facing. Policy reforms, concerning the trade structure, and the external sector are especially emphasized. STABILIZATION AND GROWTH IN THE ‘WASHINGTON CONSENSUS’ APPROACH The imbalances that most Latin-American economies are experiencing are not of a short-run character. The disequilibria in the current account, the government budget as well as the propagation mechanisms (in the form of high inflation and/or increased financial fragility) have led to a situation that cannot be reverted in the short run since it calls for a complete change in the present economic regime. The present context of disequilibrium is not the traditional one. It does not stem from a short-run disadjustment between absorption and domestic income that might be reversed implementing a traditional IMF-inspired policy package. This new characteristic of the situation is broadly perceived by both economic policymakers and scholars. There is a wide consensus that not only short-run stabilization measures but also others oriented to introducing structural1 changes are needed to ensure the closing of the gaps. One of the most important and influential responses to the challenges that the need for stabilization has posed is the policy package based on what Williamson (1990) termed the ‘Washington Consensus’2 (WC) approach to stabilization-with-growth. The diagnosis of the WC is based on both empirical and theoretical grounds. In the WC approach, the deepest roots of present Latin-American instability and lack of growth lie in the development strategy adopted in the postwar period in most countries of the region, that is, the import-substitution strategy of industrialization (ISI). According to the WC approach, this strategy entails an inward-oriented model of growth and a serious misallocation of resources, especially because of the central role played by the public sector as an ‘engine of growth’. From the theoretical point of view, the benchmark to judge the suitability of the ISI model in the WC view is the neoclassical-oriented model of development evolved, among others, by McKinnon, Shaw, Krueger and the staff of the World Bank.3 The intellectual influence of this view is so strong that Fischer (1990) has said that there are no longer two major competing economic development paradigms. The only paradigm is the market-oriented one and, consequently, participants in the development debate now speak the same language. However, given that this development paradigm has little to say about how to stabilize an economy in the short run, the basic paradigm is in fact appended with the traditional Fund approach stemming from the work of Polak and the staff of the IMF.4 This marriage between the neoclassical approaches to stabilization and development provides the basis for the WC proposals for structural reform, enabling recovery of both growth and macroeconomic stability in Latin America. The set of objectives pursued by the policy proposals based on this paradigm has been summarized by Fischer (1990) as the search for establishing (a) a sound macroeconomic framework; (b) an efficient and smaller government; (c) an efficient and expanding private sector; and (d) policies for poverty reduction (such as targeted food subsidies as well as medical and educational programmes).

82

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

We think that this approach has been successful in highlighting some important weaknesses of the development strategy adopted by Latin America in the postwar period. However, we also think that it has certain flaws, which render it an inappropriate tool for overcoming the disequilibria that will be analysed in the following sections. The empirical interpretation of the ISI model and of the role of the state, as well as the theoretical foundations of the liberalization approach are too general and abstract to represent a firm base on which to diagnose, design and apply sound policies oriented to stabilization-with-growth. It also shows weak theoretical foundations with regard to the relationship between short-term and long-term policies designed respectively or stabilizing the economy and recovering growth. The neoclassical approach is basically a static comparative exercise and, therefore, lacks the dynamic elements that have an important bearing on the relationship between stabilization and growth. Given the extreme influence that the WC has nowadays on policy-making in Latin America, it seems worth while discussing it further in this section. Then, we shall summarize the main elements of the diagnosis that the WC approach sustains with regard to the ISI model and the role of the state, its theoretical underpinnings and the policy proposals presented by Williamson (1990) as the core of the WC approach. The neoclassical interpretation of the import-substitution strategy of industrialization and the role of the state5 The main empirical findings that support the WC approach regarding the determinants of growth stem from the comparison between the growth records of the Far Eastern NICs (particularly Korea) and Latin America in the postwar period. As is well known, ex post, the growth strategy pursued in the Far East resulted in both impressive rates of economic growth and a high degree of macroeconomic stability, but such results were not achieved by Latin American countries. According to the supporters of the liberalization approach, such as Krueger and Balassa, the better performance of the Far Eastern countries is explained basically by the outward-oriented character of the industrialization process in those countries. Others, like McKinnon, tend to emphasize the role of the liberalization of the financial markets.6 There are other interpretations with regard to the main determinants of the economic success in Asia, and it is not our objective here to pursue the issue any further. For our purpose, it is unnecessary to discuss whether the neoclassical interpretation of the development in Asia is correct or not from an empirical point of view.7 The important issue is that a new paradigm has emerged from the neoclassical interpretation of such a process and that the stylized facts regarding the external and financial sectors that shape the neoclassical model of East Asian development are used as a pattern to assess the flaws of the growth process in Latin America. Thus, we shall concentrate here on the valuation of the Latin-American import substitution strategy made by the WC on the basis of those stylized facts. According to Balassa et al. (1986), it was only after the Second World War that import substitution became a doctrine that would guide policymaking in much of Latin America in the following years. In the WC view, this new doctrine reflected the rise of nationalism, the fear of another depression and the desire to provide employment. The ECLA is known to have been the main intellectual source of the ISI approach to development. The import substitution strategy based on protection of native industries was thought of as a response to the slackening of demand from the industrial countries for Latin-Amer-ican primary exports, the tendency for the terms of trade to deteriorate, and the inability to develop domestic manufactures. The ECLA recognized the importance of manufactured exports, but because of the protectionist policies in developed countries, it tended to recommend economic integration in Latin America as the way to foster exports.

TRADE REFORM IN LATIN AMERICA

83

Balassa’s criticism does not focus on the first ‘easy’ stage of import substitution. Since this stage of the manufacturing process is relatively labour-intensive, it does not entail much protection. Continued importsubstitution, however, requires the replacement of imports of intermediate goods and producer and consumer durables by domestic production. The production of intermediates, such as petrochemicals, paper and steel, and producer and consumer durables, such as machine tools and automobiles, is capital-intensive and shows economies of scale. As a consequence, the second stage needs a much higher level of protection. This second stage of the ISI is the one that has received the bulk of the liberalization-approach criticism. They have termed this stage ‘import-substitution at any cost’. According to neoclassical authors, the attempt to substitute intermediates and durable goods introduced an anti-trade bias in the development process. The main features of this stage were the following. In the first place, there was the loss of economies of scale. The large capital requirements and the degree of technological sophistication raised production costs within the confines of national markets and, therefore, import-substitution industries needed high protection to survive. In the second place, the limited size of domestic markets led to monopoly positions in some industries, whereas subsidiaries of foreign companies were attracted by high protection levels. The absence of domestic and foreign competition gave little incentive to improving technology. In the third place, a slackening technology and easy profits in sheltered domestic markets impeded the development of manufactured exports. Finally, the capital-intensive character of these industries raised the capital requirements of additional output increments, whereas there was no concomitant increment in the savings/ output ratio. A side-effect of the development strategy chosen by Latin-American countries was that primary production suffered discrimination, as protectionism and export taxes on primary exports (imposed to finance accumulation in the industrial sector) turned the terms of trade against primary activities. As a consequence, the lack of dynamics of manufactured exports, coupled with the bias against primary exports, impeded Latin-American countries from earning enough foreign exchange to sustain their economic growth. The above-mentioned characteristics of the import-substitution process may be thought of as inherent to such strategy. The main criticism with regard to the ‘engine of growth’ role played by most Latin-American states in the development process was directed toward the rapid expansion of the size of the government. This expansion was evident from the growing role of the state as investor, the increasing quantity and size of public enterprises and the mounting number of market regulations. The WC viewpoint sees this role of the state as excessive and suffocating. It led to a concomitant weakening of the private sector. While public investment tended to crowd out more profitable private projects, public enterprises undertook economic activities that private agents would have performed better. In this way, according to the neoclassical position, the crowding-out effect reduced the availability of funds for the private sector, while at the same time inefficient public enterprises worsened the distortion in the resource allocation provoked by the import-substitution strategy of development. According to liberalization literature, the Latin-American countries are considered to be among the world’s most regulated market economies. Among the more often criticized economic regulatory mechanisms, we find price controls, import barriers, discriminatory credit allocation, limits on the dismissals of employees, restrictions on inflows of foreign investment and outflows of profit remittance, control of capital movements and control of new investment as well as the establishment of firms in certain sectors. In the view of authors favouring liberalization the costs of overregulation stem from the fact that regulations and controls bias the market signals and also from the fact that there is a greater potential for

84

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

corruption, because the productive activity can be regulated by legislation, government decrees and case-bycase decision-making. However, there are some ambiguities and differences among the authors who share the basic guidelines of the liberalization approach with regard to regulations and government intervention. Some accept that state intervention is necessary before all markets have been liberalized. Complete release of market forces in some markets, while others are still distorted, might result in more rather than less misallocation of resources (Bhagwati 1987). Fischer (1990) went beyond this and distinguished two strategies of development. The first is the Chilean or Thatcher approach. This strategy requires that the government sets the right policies and incentives, behaves consistently and credibly and waits: growth will return. The second is the strategy followed by East-Asian countries, which entails a more active government role, especially with regard to industrial policy. In such a context, the government policies are oriented to maintaining an undervalued currency, to promoting exports and to keeping strictly time-limited import protection through tariffs. Fischer considers that other requisites of the strategy are an educated and disciplined labour force, a small government and the availability of entrepreneurial skills. Nonetheless, he does not advise most developing countries to attempt to run an industrial policy that picks the winners as opposed to setting a generally favourable economic environment for industry and for investment. Many governments do not have the ability to run a pick-the-winner industrial policy. Regardless of whether they think there is some room for efficient government intervention or that it is inherently bad, there is wide consensus among the authors of the liberalization approach on one point: the lack of correct economic policy management in Latin America has been one of the most important impediments of growth. In addition to the flaws corresponding to the development strategy that has been commented on before, it is considered that the management of the economy in the short run has been handled just as badly. Large government deficits are seen as the primary cause of excessive absorption and inflation. In this way, the lack of coordination and the loose control over fiscal, monetary and debt management policies led to both price and foreign exchange rate instability. Based on this diagnosis, the WC approach shows broad agreement with regard to the primary reasons for the present state of high indebtedness in most Latin-American countries: it originated in the poor policy management by the public sector. The inward-oriented development strategy led to a structural scarcity of foreign exchange while the loose control over public sector borrowing needs gave rise to an overaccumulation of public foreign liabilities. The Washington Consensus’s policy recommendation The WC diagnosis of the Latin-American development experience in the postwar period led them to conclude that: (a) the macroeconomic setting is characterized by severe instability; (b) the economic structure is highly distorted with regard to the allocation of resources; (c) the past performance of the government economic policy management precludes any serious hope of attaining a high growth rate in the future were a pick-the-winner development strategy chosen. What then should be done in such a context? Or, to put it in Williamson’s words, what does Washington mean by policy reform? Williamson (1990) identifies ten policy instruments with which Washington can master a reasonable degree of consensus. They can be summarized as: 1 Fiscal deficit. Fiscal deficit is a primary source of macroeconomic dislocation and it results, in the view of the WC, from a lack of the political courage or honesty to match public expenditures and public

TRADE REFORM IN LATIN AMERICA

85

revenues. Likewise, fiscal policy should aim to reduce expenditures rather than increase taxes in order to lower the borrowing needs of the public sector. 2 Public expenditures. Subsidies, especially indiscriminate ones, should be eliminated. Government investment should be almost exclusively directed toward public infrastructure. 3 Tax reform. The principle that should guide policymaking in this area is that the tax base should be broad and the marginal tax rate moderate. 4 Interest rate. Negative interest rates should be eliminated by letting the market do the job of determining a positive and (presumably) moderate real interest rate. The beneficial conse quences of this would be: (a) the discouraging of capital flight; (b) the rise in private savings; (c) the elimination of the corruption opportunities generated by segmented markets; (d) a better allocation of the resources available for investment. Some fears regarding a possible explosion of public domestic debt as a consequence of higher interest rates are expressed by some supporters of this policy measure. 5 Exchange rate. The exchange rate policy should not only assure a competitive level for the exchange rate but also assure that such a level will be maintained to support the private sector confidence. Growth of exports (particularly nontraditional ones) is the main purpose of this policy. Although, like interest rates, the exchange rate might be market-determined, the dominant view is that achieving a competitive rate is more important than how the rate is determined. There is no firm consensus about how and when to liberalize the balance of payment capital account. 6 Trade policy. The objective of the trade policy should be to attain a situation in which the cost of generating or saving a unit of foreign exchange is equalized between and among export and import competing industries. To move toward this objective, the government policy should eliminate non-tariff barriers such as import licensing, eliminate the taxation of exports and guarantee access to imports of intermediate inputs at competitive international prices. Protection to infant industries should be strictly temporal and a moderate general tariff (10 to 20 per cent) might be accepted as a mechanism for diversifying industry. There remain, however, some differences about the speed with which the import liberalization process should proceed. 7 Foreign Direct Investment. The countries should establish a legislation favouring foreign direct investment as a way to attract capital and technology. 8 Privatization. Policy reform should favour the privatization of state enterprises since they would be managed more efficiently by private agents. 9 Deregulation. The main rationale for deregulation is that a higher degree of competition would foster growth. Policies aimed at deregulating labour markets are especially emphasized. 10 Property rights. There is a conviction that property rights are highly insecure in Latin America. So, the legislation and institutions should be adapted to secure property during the process of structural reform. The assumption that lies behind the policy recommendations that make up the Washington Consensus is that there is a fundamental complement between adjustment and economic growth (Guitian 1988). And this is so, because they believe that most of the output growth during the first stage of the reform period should come from a better reallocation of existing resources and a better utilization of existing capacity. Longerrun, self-sustained growth can only be achieved when nationals prefer to invest domestically instead of abroad, and such a thing can only occur once the structural reforms and macroeconomic stability are soundly rooted (Selowsky 1990). On the basis of this approach to the relationship between stabilization and growth, the liberalization literature specifies that there are at least two stages that an economy should go through from stabilization to growth. During stage I, economic policy should almost completely focus on establishing a sound

86

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

macroeconomic setting. Given that instability is conceived as the result of excessive and volatile borrowing from domestic sources to finance the public sector deficit, achieving a primary fiscal surplus is the priority at this stage. The needed major fiscal reform to fulfil this policy objective should follow the abovementioned guidelines of the WC with regard to the public sector. However, the specified policy package lacks the essential elements to structure a truly short-run stabilization package aimed at attacking the most important short-run disequilibria, such as a high inflation rate or the existence of a misleading structure of relative prices (typically: overvalued exchange rates and lagging public prices). Owing to this, the policy package of the WC is normally complemented, with regard to the proposals for stage I, with the standard IMF stabilization procedures. In explaining the relationship between the Fund and the Bank policies Guitian (1987), said that a key function of economic management at this stage is that of keeping the level and the rate of growth of aggregate demand in a sound relationship with the level and growth prospects of the economy’s production capacity. As is well known, the Fund understands this to mean a tight control over the aggregate demand, which in turn entails not only eliminating fiscal imbalances but also keeping domestic credit expansion in an appropriate balance with the prospective path of desired money holdings in the economy.8 More often than not, this results in a deep recession. Stage II comprises the implementation of the set of policies aimed at bringing private incentives more in line with ‘true’ economic scarcities. During this stage, the core of the liberalization package should be applied, that is, the policy reforms recommended by the WC for the financial sector, the trade structure, the deregulation of the labour market, etc., that we have listed before. Only after going successfully through stages I and II could a country expect to recover sound growth and, perhaps, its credit worthiness. More optimistically, a country that has shown a firm political conviction in implementing structural reforms could expect, while applying them, some support from the official multilateral creditors. Yet, the compromise is vague. According to Fischer and Husain (1990), it is necessary to achieve some success in reforming the economy prior to being eligible. In fact, countries such as Argentina, who has shown willingness to reform but has suffered from some lack of conviction during the process, are not in a comfortable situation. Fischer and Husain are not very optimistic about the countries that have yet to put in practice a ‘realistic and sustainable program of economic reform’. They said that ‘as things now stand they will be left out of the strengthened debt strategy and continue to carry a heavy burden of external debt until they achieve the political will to adopt the necessary adjustment measures. For such countries, the main thrust should be to persuade them to undertake credible and lasting economic policy changes that can make them eligible for debt and debt-service reduction’ (p. 27). Some unresolved issues of the WC approach The conviction that the liberalization approach is the appropriate framework to tackle the issue of restoring growth and that the Fund model is the correct one to address stabilization seems to be held strongly by the participants in the WC. This, in turn, entails that the sequencing of policy reforms should broadly follow the above-mentioned stages. Nonetheless, we could ask whether there are sound bases for this consensus about the necessity and sequencing of the stages of the adjustment-with-growth process, and also for the Fischer contention about the uniqueness of the development paradigm. In principle, there is an ambiguous response to this question. On the one hand, even the authors that adopted the analytical framework provided by the liberalization approach have said that ‘more research is needed’ with regard to several key points of that framework. This is particularly so in relation to the issues that have a bearing on the dynamics of the economy when it is going from ‘representation’ to liberalization. On the other hand, the empirical evidence

TRADE REFORM IN LATIN AMERICA

87

from countries that have succeeded in obtaining stabilization-with-growth seems to contradict the idea that the economy should fulfil the policy sequence entailed by stages I and II. In the following, we shall specify more of these weaknesses because they have a determining influence on the correct design and implementation of the policy reforms. In the second place, as Michalopoulos (1987) clearly recognizes in his paper supporting the World Bank programmes for adjustment and growth, uncertainties are prominent at the implementation stage because implementation involves the dynamics of reform, about which less is known. Therefore, Michalopoulos concludes that, even though the framework provided a good basis for the design of adjustment in most countries, one can be more confident about the long-term outcome of the policy reform than about the precise dynamic profile of results following the implementation of the programme. Two conflicting issues are raised by Michalopoulos: (a) the sequencing of policy measures aimed at stabilization and those that focus on structural adjustment; and (b) the optimal sequence of reforms to remove distortions when many markets are initially regulated. Again, in another form, the problem of the relationship between short-run stabilization and structural reform appeared unresolved. If, as Michalopoulos recognizes, economic theory (it should be read: the liberalization approach) offers little guidance about an optimal sequence for removing market distortions, why should we assume a priori that the dynamics of the process would never be explosive? And, how can we evaluate the costs and benefits of policy reform? With regard to the dynamics issue, we shall show in the next section that there is enough empirical evidence from many Latin-American countries to at least not disregard a priori the possible generation of explosive paths in the dynamics of the adjustment process in the domains of the public accounts, inflation, the financial system and the evolution of the stock of public external debt. In the third place, the WC takes for granted the contention that a strong complement exists between domestic policies and external financing. Sound policies will be rewarded with the almost automatic return of debtor countries to normal and voluntary access to the international capital markets. Latin-American experience does not confirm this contention on empirical grounds. Countries that undertook deep reforms, such as Uruguay, Bolivia or even Mexico, did not receive any significant amount of additional external financing. Or if they did, it was only after a long delay and the amount was insufficient to recover growth. It should be taken into account that the total amount that the IMF and the World Bank are expected to provide within the Brady initiative will be no more than $20–5 billion, whereas the Latin-American debt amounts to $500 billion. On the other hand, the most successful adjustment experience with growth in Latin America is Chile. But we shall see that this country received preferential treatment with regard to external financing. The amount of external credit received was significant and was available from the beginning of the adjustment process. A fourth point, for which there is at least mixed evidence on both empirical and theoretical grounds, has to do with the sources of growth. The core of the liberalization approach contribution to development theory calls attention to the potential of improved resource reallocation for fostering growth. For many reasons, this strong contention about the sources of growth raised some doubts on empirical rounds. First, the research on the determinants of growth usually found that resource allocation does not explain too much, while the bulk of additions to the growth rate is explained by gains stemming from advances in knowledge and education. Second, successful liberalized countries such as Korea, Turkey and Chile showed both improvements in resource allocation and a high level of disposable savings either from domestic sources or from abroad.9 Third, countries such as Uruguay and Bolivia, which have undertaken structural reforms to liberalize the market forces but show very low levels of investment, have not resumed growth.

88

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

THE LATIN-AMERICAN EXPERIENCE IN THE 1980s: ADJUSTMENT AND STAGNATION Although the WC approach has rightfully pointed out some important weaknesses of the development strategy followed by Latin-American countries in the postwar period, we think that it shows certain flaws that render it an inadequate tool for analysing the functioning of these economies after the debt crisis and for proposing appropriate solutions to it. In our view, since the beginning of the crisis, most Latin-American economies suffer from two basic disequilibria: in balance of payments and fiscal accounts. To permanently close both disequilibria is a necessary condition for a successful stabilization-cum-growth solution. However, to accomplish this task with the standard tools of conventional stabilization policies seems to be quite difficult, as seen by the appalling performance of most Latin-American economies. Moreover, certain features of the financial, labour and goods markets contributed to amplifying the consequences of the two basic disequilibria in some of these economies. This was the case, in particular, in countries with prolonged experiences of high inflation and with small and very fragile financial systems. In order to overcome what we consider one of the main flaws of the WC approach, then, we present in this section our own diagnosis of the Latin-American economies in the late 1980s. To do so, we discuss in some detail the performance of five Latin-American countries—Argentina, Brazil, Colombia, Chile and Mexico—in relation to the two basic disequilibria mentioned above. The external gap It is generally acknowledged that the Latin-American region suffered the impact of a huge negative external shock between the late 1970s and the early 1980s. It originated in the combination of a significant fall in the terms of trade and an abrupt increment in the world interest rate. What was thought of as a temporary worsening of the external situation for the region resulted, in fact, in a permanent change in the international economic context. The 1982 Mexican moratorium gave rise to a new situation with a tight rationing of external financing for the region as a whole. The main and immediate consequence of the shock was that the services of the outstanding external debts experienced a huge increment. Between 1980 and 1982–3, external services rose by 7 percentage points of GDP in Argentina, 5 points in Chile, 4 points in Mexico and around 2 points in Brazil and Colombia. The fall in the prices of the main export goods was, during the same period, 34 per cent in Chile, 30 per cent in Argentina, 21 per cent in Brazil, 18 per cent in Mexico and 10 per cent in Colombia. It is worth while too reviewing the initial responses to this shock in terms of the adjustments implemented at a national level. In the case of Colombia, there was no initial adjustment. It was not until 1984 that an adjustment package was implemented, resulting in a positive trade surplus in 1986. The increment in the real exchange rate played a primary role in the attainment of this result (Table 5.1).10 However, the effects of the adjustment significantly differed from what was observed in other Latin-American adjustment processes. Neither the external shock nor the subsequent adjustment entailed dramatic macroeconomic consequences. There are many macroeconomic features in the Colombian case that explain this and they will be examined below. We will just mention here some specific characteris Table 5.1 Export real rates of exchange, (1980=100) 1970–9

Argentina

Brazil

Colombia

Chile

Mexico

143.7

78.5

111.8

98.1

109.4

TRADE REFORM IN LATIN AMERICA

Argentina 1980 100.0 1981 136.1 1982 165.8 1983 155.3 1984 148.4 1985 167.6 1986 186.0 1987 213.0 1988 192.4 Source: CEPAL(1989).

Brazil

Colombia

Chile

Mexico

100.0 84.0 82.0 98.2 91.9 92.6 94.1 96.5 87.4

100.0 92.6 86.2 86.1 91.4 103.2 128.2 140.9 142.8

100.0 82.7 95.2 112.6 116.3 144.3 165.0 173.3 180.7

100.0 91.0 134.3 137.4 114.3 113.6 147.1 153.3 126.8

89

ties that have to do with the external sector: (a) the measures aimed at balancing the current account were moderate and gradually applied; (b) there were quantitative controls on imports; (c) the degree of external financial exposure of Colombia was relatively low compared with most countries of the region (Table 5.2). This last feature was very important, because the smaller magnitude of the needed adjustment entailed smoother disequilibria at the domestic level. After decelerating between 1981 and 1983, the activity level recovered in 1984. Between 1984 and 1988, the average per annum growth rate was 4.7 per cent. Unlike Colombia, the other four countries that we are analysing here carried out an initial adjustment of the external sector that was far from gradual. Between 1980 and 1983, there was a huge drop in the amount of imports in these countries. They fell by 1 percentage point of GDP in Brazil, and by 5 points in Argentina, Chile and Mexico. Even though the cut in imports played a primary role, it was not the only factor explaining the increase in the trade surplus that followed the implementation of the adjustment policies. The quantity of goods exported also increased heavily during this period: 30 per cent in Argentina, 40 per cent in Brazil, 70 per cent in Mexico and 15 per cent in Chile. One of the most important consequences of this adjustment was the strong fall in domestic absorption which resulted, in turn, in a huge recession. In the first years of the adjustment, the accumulated fall in the activity level reached 14 per cent in Chile, 13 per cent in Argentina, 7 per cent in Brazil and 5 per cent in Mexico. Table 5.2 Foreign debt as a proportion of GDP (%) Argentina Brazil Colombia Chile Mexico Source: CEPAL(1989).

1980–1

1982–3

1984–6

1987–8

25 29 18 47 25

68 39 26 83 58

67 44 36 119 63

68 37 39 98 65

The devaluation of the domestic currency was a key instrument in achieving the initial adjustment in the current account and the later maintenance of a permanent surplus in the trade account. In 1983, real exchange rates had already reached very high levels compared with those prevailing in 1980, and also compared with the average level of such variables during the 1970s. This marked upward correction of

90

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

exchange rates did not revert afterwards. In 1987–8, the real exchange rates in Argentina, Chile and Mexico were respectively 100 per cent, 77 per cent and 40 per cent higher than those prevailing in 1980. Beyond the common characteristics assumed by the initial external adjustment in these four countries, the subsequent evolution of the external gap followed different paths. Four factors played a significant role in explaining the divergent courses followed by these economies: (a) the degree of external indebtedness at the beginning of the period; (b) the degree of openness of the economy when the shock took place; (c) the later evolution of the terms of trade for each country; (d) the relative amount of external financing made available by foreign creditors at the beginning and during the adjustment process. The interplay of these factors heavily contributed in determining the degree of ‘tightness’ of the external constraint and, therefore, in determining the need for additional measures to deepen the initial adjustment. At the beginning of the adjustment, Brazil was a very closed economy but, at the same time, its degree of external indebtedness was lower. Likewise, for a long period before the debt crisis, this country had been pursuing consistent import-substitution and export promoting policies. As a consequence, once the domestic currency was devalued in real terms, there was a sustained expansion of manufactured exports while imports remained low. Thus, from 1983 onwards, the trade balance permanently showed a surplus and the current account tended to balance and even to show a surplus (with the only exception being 1986, when the Cruzado Plan was implemented). It was unnecessary to put into practice further devaluations of the domestic currency in real terms. At the beginning of the 1980s, among the countries considered here, Chile showed the greatest degree of openness to foreign trade and obtained a very elastic response of exports to real devaluation. The exports/ GDP ratio rose from 20 per cent at the beginning of the decade to 40 per cent in 1988. Nonetheless, it should be taken into account that half of the expansion of exports originated in the increment of copper exports. This improved evolution of the copper revenues was due, in turn, both to the increase of copper prices in the international markets and to the government decision to invest heavily in the copper-mining enterprises, which consequently gave rise to a significant increment in the production level. However, in the case of Chile, in spite of the improved evolution of exports after the initial adjustment, the current account showed an important deficit over the whole adjustment period. This behaviour of the current account was due both to the high degree of external indebtedness of the country and to the upward trend of imports. Nonetheless, the country was able to finance the deficit in the current account because of the considerable amount of financial resources stemming from multilateral official creditors. In Table 5.3 it can be seen that, in 1984–5, while, Brazil, Argentina and Mexico had minimized their current account deficits, Chile showed an average deficit of about 9 percentage points of GDP without incurring arrears. In the 1983–7 period, the external financing received by Chile from bilateral and multilateral official creditors amounted to 40 per cent of its external financial payments.11 At the beginning of the 1980s, Mexico and Argentina represented an intermediate case from the point of view both of the degree of external indebtedness and of the openness of the economy. Still, in Argentina as well as in Mexico, the real devaluations and the fall in the activity level resulted in a rapid reversion of the external imbalance. These countries showed significant trade surpluses immediately after applying measures aimed at improving the external performance. However, this outcome would tend to revert by the mid-1980s. Nonetheless, there are some differences between Mexico and Argentina. Mexico’s debt service/output ratio was lower than Argentina’s and, consequently, the trade surpluses generated were

TRADE REFORM IN LATIN AMERICA

91

Table 5.3 External sector 1980

1981

Argentina EXP/GDP 11.46 15.67 IMP/GDP 15.23 16.77 Trade Bal./GDP −3.76 −1.09 Fac. Servs./GDP −1.74 −5.71 Curr. Acct./GDP −5.50 −6.80 Brazil EXP/GDP 8.40 8.70 IMP/GDP 10.68 9.27 Trade Bal./GDP −2.28 −0.57 Fac. Servs./GDP −2.65 −3.50 Curr. Acct./GDP −4.93 −4.07 Colombia EXP/GDP 15.7 11.7 IMP/GDP 15.6 16.3 Trade Bal./GDP 0.1 −4.6 Fac. Servs./GDP −0.8 −1.5 Curr. Acct./GDP 0.4 −5.9 Chile EXP/GDP 21.63 15.26 IMP/GDP 25.82 25.16 Trade Bal./GDP −4.18 −9.89 Fac. Servs./GDP −3.15 −4.75 Curr. Acct./GDP −7.33 −14.64 Mexico EXP/GDP 13.07 12.87 IMP/GDP 14.24 14.85 Trade Bal./GDP −1.17 −1.98 Fac. Servs./GDP −3.09 −3.83 Curr. Acct./GDP −4.26 −5.81 Sources: CEPAL (1989) and Villar (1989).

1982

1983

1984

1985

1986

1987

1988

16.39 11.63 4.76 −8.96 −4.20

16.01 10.03 5.98 −10.18 −4.20

15.92 9.87 6.05 −10.18 −4.13

14.73 7.76 6.98 −8.38 −1.40

13.58 10.43 3.16 −7.76 −4.60

11.34 10.63 0.71 −6.61 −5.90

14.75 9.64 5.11 −7.25 −2.15

7.60 8.57 −0.97 −4.67 −5.64

10.70 8.85 1.85 −4.99 −3.15

12.80 7.78 5.02 −5.07 −0.06

11.60 7.09 4.51 −4.69 −0.18

8.71 6.47 2.24 −3.87 −1.64

9.18 6.31 2.87 −3.38 −0.51

11.02 6.07 4.94 −3.48 1.47

11.3 17.3 −6.0 −2.7 −8.9

10.6 15.0 −4.6 −3.1 −8.5

11.9 13.2 −1.3 −4.1 −6.9

11.9 11.6 0.3 −4.4 −5.0

14.3 9.0 5.3 −4.1 1.5

12.3 8.9 3.4 −4.0 0.3

11.2 9.5 1.7 −3.2 -0.8

19.09 20.65 −1.56 −8.20 −9.76

23.38 20.46 2.93 −8.78 −5.86

23.37 23.89 −0.53 −10.48 −11.01

27.87 24.46 3.41 −11.79 −8.37

30.42 26.23 4.18 −11.20 −7.01

33.69 29.32 4.37 −9.01 −4.64

39.32 31.59 7.74 −9.07 −1.34

16.24 12.34 3.90 −7.80 −3.90

19.29 9.01 10.28 −6.46 3.82

17.26 9.21 8.05 −5.67 2.38

15.19 10.10 5.08 −4.77 0.31

16.17 11.95 4.21 −5.53 −1.32

19.39 12.05 7.33 −4.80 2.54

16.51 14.21 2.30 −4.10 −1.79

large enough to close the current account gap. The result of the current account became positive in 1983– 4. Argentina’s best external performance was achieved in 1985, but the country was unable to completely close the imbalance in the current account. In spite of the adjustment done in these countries, from 1985 in Mexico and 1986 in Argentina the results achieved in the external sector tended to revert. The primary reason was that there was a tendency for the terms of trade to deteriorate. Both countries consequently implemented additional measures aimed at restoring the current account equilibrium, then causing disturbing effects on the domestic side of these

92

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

economies. The main destabilizing effects were the acceleration of inflation and the fall in the activity level. We can now draw some conclusions about the relationship between the external sector and macroeconomic stability in the countries we are now analysing. Colombia was able to soften the disruptive consequences of the external shock because the country had shown a lower degree of financial exposure to the international capital market. This was an exceptional case in the Latin-American context at the time of the debt crisis. Consequently, in the other countries the initial adjustment process entailed a higher degree of macroeconomic instability. However, after the early stages of the external adjustment, in each case the economy evolved differently. Brazil and Chile achieved an initial closing of the external gap that would last throughout the 1980s. The reasons for these results, however, are different. Brazil obtained a long-lasting equilibrium in the current account because its trade surplus was large enough to cover the deficit in the financial services account. In the case of Chile, on the contrary, the closing of the external gap was made possible by the significant amount of external financing obtained by the country in the first years of the adjustment period. This made the initial adjustment smoother until the increase in the prices of copper would permit the country to balance the current account on the basis of its own resources. A certain degree of external equilibrium was also achieved by Argentina and Mexico in the years following the initial adjustment, when both countries implemented stabilization policies agreed upon with the IMF. Nonetheless, the closing of the gap was not long-lasting and the external gap once again became a destabilizing force. The achievement of a long-lasting closing of the external gap is not a sufficient condition for attaining macroeconomic stability. This fact is patiently exemplified by the case of Brazil. The closing of the external gap is, however, a necessary condition. We prefer the concept of ‘closing the external gap’ to that of ‘achieving equilibrium of the external sector’ in order to refer to the situation of credit rationing that Latin-American countries are now facing in the international capital markets. A ‘long-lasting’ closing is not merely a situation where ex post a country matches external outlays and revenues from different sources (including external financing). To be ‘longlasting’, the closing of the external gap must be perceived as lasting not only by domestic but also by foreign agents. For this to occur, the agents must be convinced that there is a very low probability that the government will face problems in raising the funds needed to finance the external gap. This specifically implies firmly sustained expectations on the future evolution of the exchange rate by the private sector. In turn, this has a bearing on the agents’ decisions regarding the allocation of resources on both the real and the financial side of the economy. One of the most important obstacles for achieving a long-lasting closing of the external gap is its magnitude and its long-run character. The external imbalance observed at present stems from an inconsistency between the stock of accumulated debt and the flow of external payments that these economies can afford without inducing severe macroeconomic imbalances in terms of the activity level and the rate of growth of the economy. The present disequilibria did not originate in the trade but in the financial services account of the balance of payments. The stock-flow character of the disequilibria explains its temporal persistence and the difficulties in managing them with conventional policy tools. Among the cases that we are analysing, this stock-flow inconsistency is particularly strong in Mexico, Argentina and Chile, given the current level of world interest rates and of the external indebtedness of these countries. Chile seems presently to be in a better position than Argentina and Mexico. However, the present equilibrium in the Chilean current account is heavily sustained by high international copper prices, which are not believed to be sustainable in the future. If copper prices fall, will the Chilean economy be in a

TRADE REFORM IN LATIN AMERICA

93

position to obtain external financing to offset the subsequent fall in its external revenues, thereby avoiding the disturbing macroeconomic consequences of implementing an additional adjustment? Empirical evidence does not seem to indicate that the external gap will soften in the immediate future. During the 1980s, the solvency indicators did not show any significant improvement and, in some cases, they worsened. In spite of the impressive magnitude of the real external transfers carried out by the LatinAmerican countries, the debt/GDP and debt/exports ratios suffered an increment, as can be seen in Tables 5.2 and 5.4. This evolution of the solvency indicators suggests that the size of the external gap does not show a declining trend. In the absence of offsetting mechanisms in a rationed credit market, the fragility of the present situation will likely become a permanent feature of the external accounts of these countries in the near future. Another negative feature of the present situation, from a long-run point of view, is that the generation of trade surpluses entailed a huge fall in the investment/GDP ratio. As a consequence, even if we were eager to accept optimistic assumptions about the productivity of the capital stock, the present investment rate is not high enough to ensure Table 5.4 Foreign debt as a proportion of exports of goods and services (%) Argentina Brazil Colombia Chile Mexico

1980–1

1982–3

1984–6

1987–8

302 319 165 249 237

480 415 267 400 340

527 415 261 433 378

590 374 247 279 359

Source: CEPAL (1989).

a rate of growth of domestic income consistent with the future stream of interest payments that the outstanding external debt will generate. The fiscal gap The impact effect of the external shock of the early 1980s on the fiscal accounts was large and long-lasting. It was so because, in Latin America, the public sector held much of the external debt. In some of the countries the public sector external commitments grew heavily because of aggressive investment policies implemented in the previous years. Brazil is a good example of this. In many others, in addition, the increase in the government liabilities stemmed from the nationalization of the stock of debt held by the private sector. This was the case in Argentina and Chile. The debt was ‘socialized’ by a variety of means, the typical ones being those operating through exchange rate guarantees, through the pressure of external creditors for obtaining public sector collaterals, or through the direct takeover of private liabilities by the state. Given that much of the external debt was held by the public sector, the rise in the world interest rates resulted in a sharp increase in government expenditures. To avoid the destabilizing macroeconomic consequences that a higher deficit would induce, either fiscal non-interest expenditures had to be cut or fiscal revenues had to be increased. Since these measures would also have strong repercussions on the macroeconomic equilibrium, there was no way out of the adjustment. This is what has been called the domestic transfer problem.12 The way in which each Latin-American country managed to face this domestic counterpart of

94

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

the external transfer problem (that has to do with the current account imbalance) considerably contributed to determining the degree of macroeconomic stability observed during the 1980s. In principle, in order to avoid the destabilizing macroeconomic consequences of the higher burden of external payments, the government had to generate a primary surplus pari passu with the increment in the trade surplus originating in the adjustment of the external sector. There are two factors that indisputably determined the degree of success in attaining such an objective during the 1980s. First, the situation of the public sector accounts when the external shock took place and, second, the repercussion of the external adjustment on the government budget. Colombia presents some particularities with regard to both elements, which differentiate it from the other four countries in the sample. In the first place, in the period prior to the external shock of the early 1980s, the Colombian authorities had shown a stricter discipline in managing the budget. Besides, at that time, the Colombian economy showed a lower inflation rate and a higher degree of monetization than the other four Latin-American economies. In such a context, the fiscal and monetary adjustment policies put into practice showed a higher degree of effectiveness. In the second place, in spite of the fact that the state held the major part of the Colombian external debt (see Table 5.5) the fiscal impact of the Table 5.5 Public foreign debt* as a proportion of total debt (%) Argentina

Brazil

Colombia

Chile

Mexico

1980 53.2 56.8 63.2 47.0 nd 1981 56.1 57.8 65.9 36.0 nd 1982 65.5 60.7 66.0 39.0 67.2 1983 70.4 71.2 68.7 58.0 66.7 1984 77.1 71.9 71.0 67.0 71.8 1985 80.8 76.3 75.7 74.0 73.7 1986 85.6 78.4 80.0 83.0 75.1 1987 nd 79.7 79.6 87.0 79.5 1988 nd nd 79.4 84.0 80.7 Sources: Argentina: Central Bank. Chile: Bande and Ffrench-Davis (1989). Brazil: Giambiaggi (1989). Mexico: CEPAL (1989). Colombia: CEPAL (1989). Note: * Including guarantees.

external shock was weaker, because the amount of this debt was lower than in the other countries. Between 1980 and 1984, the central government deficit rose from 1.2 per cent to 4.2 per cent of GDP (see Table 5.6). In 1984, the consolidated fiscal deficit (which includes the decentralized public sector) reached its highest level: 6.2 per cent of GDP. Consequently, the domestic borrowing needs of the public sector were 4.2 per cent of GDP. These figures reflected a budget deficit that was high by Colombian standards but not so high by Latin-American ones. Nonetheless, in the following years, the fiscal gap was brought under control. In the 1986–8 period, the average consolidated public sector deficit was around 2 percentage points of GDP, which entailed a practically zero level for the domestic borrowing needs. The improvement in the fiscal imbalance between 1984 and 1988 can be attributed to many reasons, the most relevant being both the improvement in the application of the tax policy and the higher level of economic activity. With regard to tax policy, it should be mentioned that, first, in 1983–4 there was a tax reform (comprising the introduction of a value added tax, an increment in import taxes, the implementation

TRADE REFORM IN LATIN AMERICA

95

of a ‘mandatory saving’ mechanism, etc.) and, second, from 1986 onwards, a policy was carried out aimed at improving and simplifying tax administration. The result was an increment in government revenues of about 2.5 per cent of GDP. This characteristic of the Colombian tax structure stems from the fact that the financing of the public sector heavily relies on non-tax revenues. Between 1984 and 1988, the government received annual average non-tax revenues of 7 percentage points of GDP. Two natural resources were the main sources of these receipts: coffee (because of the surplus generated by the Fondo Compensador del Café), and petroleum (because of Ecopetrol’s surplus). The balancing of the public sector accounts is also partly explained by a fall in public investment. After reaching a maximum of 9.6 per cent of GDP in 1984 (when the great electric infrastructure projects were terminated), the public investment/GDP ratio fell to 7.5 per cent in the 1986–8 period. This level of the public investment, however, is not very different from that corresponding to the precrisis period. Only Chile showed a better public investment performance. However, Chile’s strong increment in public investment was implemented starting from an initially low public investment/ GDP ratio. The improved performance of the tax receipts, coupled with the cuts in investment and minor restrictions in current government

Source: CEPAL (1989).

Table 5.6Colombia: public sector accounts (percentage of GDP)

96 FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

TRADE REFORM IN LATIN AMERICA

97

expenditures, resulted in a primary surplus that rose from –3.8 per cent of GDP in 1984 to 1.9 per cent in 1986–8. This explains why the consolidated fiscal deficit fell by 4 percentage points of GDP in the abovementioned period, in spite of the 2 points increment in the foreign interest payments. This sound performance of the government budget, in turn, resulted in a much better situation for the domestic financial system, since the public sector borrowing needs were covered by the available external financing: the financial needs of the public sector amounted to 2 per cent of GDP, but this sector received just as much in external credit. Consequently, Colombia, unlike other Latin-American countries, did not undergo the strong pressure exerted by the public sector borrowing needs on the domestic financial system that induced an increased financial fragility. In conclusion, the disturbing effects of the increment in the real external transfer on the domestic side of the economy were reduced to a minimum in the case of Colombia, and the fiscal disequilibria did not lead to significant macroeconomic instability. The widening fiscal gap was brought under control by means of gradual adjustments in the fiscal instruments. The impact effect of the external shock was stronger in Chile than in Colombia. However, its destabilizing consequences on both the government budget and the financial system were avoided to a great extent. Two factors played a significant role in explaining this result: Chile showed a sound fiscal situation at the moment of the shock and received much more financial support from the multilateral official lenders than the other countries in the region. In the period prior to the debt crisis (1980–1), the public sector in Chile showed a surplus of about 2.5 per cent of GDP. The fiscal policies implemented since the mid-1970s greatly contributed to attaining such a result. On the revenue side, a broad tax reform was carried out, introducing the value added tax and changes in the structure of indirect taxation. On the expenditures side, there was a sharp contraction of outlays in social services as well as a strong fall in public wages. Nonetheless, in spite of the sound initial fiscal situation, the increment in the foreign interest payments resulted in a worsening of the public sector budget from 1982 onwards. In 1985, the fiscal deficit reached 6.5 per cent of GDP. As was mentioned before, the flow of external financing from abroad, basically from multilateral organizations, contributed heavily to the financing of both the current account and the public sector deficits, because the bulk of external funds was received by the public sector. In this way, the increasing size of the budget imbalance in the aftermath of the external shock was not transformed into a further issuing of domestic liabilities by the public sector. Therefore, the destabilization of domestic capital markets was avoided. In any case, from 1985 onwards, the public sector borrowing needs systematically fell. As in the case of Colombia, another factor that played a central role in balancing the government budget was the exploitation of natural resources owned by the public sector. Copper constitutes the major item in the Chilean exports and is owned by the government. Thus, the authorities used the rents from copper exploitation to balance the budget. The share of copper in total exports rose from 5.2 per cent of GDP in 1981 to 14.1 per cent in the 1987–8 period. Roughly one-half of the additional value of exports stemmed from increments in the quantity produced (both because of higher productivity and of higher public investment in mining) and from the increase in copper prices, which rose significantly from 1987 onwards. The other half is basically explained by the real devaluation that took place during this period. In fact, the importance of copper is underestimated by the central government accounting. For example, the accounting system does not record the copper enterprises’ income utilized in the financing of investment undertaken by these enterprises. Nonetheless, the amount of public resources stemming from copper exploitation can be estimated on the basis of the financial surplus (net from depreciation allowances)

98

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

obtained by the Chilean state from copper exports. The amount of this surplus can be estimated as reaching on average 3.5 percentage points of GDP in the 1983–6 period and 6.5 points in the 1987–8 period.13 In sharp contrast to the other countries, who received a strong negative shock on the public sector budget, the government non-interest expenditures as a proportion of GDP showed no fall during the adjustment period in Chile. Likewise, while public investment in other countries was one of the expenditure items that suffered the harshest cuts, Chile’s public investment was considerably raised between 1982–4 and 1985–7 (see Table 5.7). The pattern of evolution of the fiscal sector in Chile and Colombia that we have highlighted here strongly differs from what happened in the other three countries. In Argentina, Brazil and Mexico, there was a recurrent tendency for the fiscal gap to induce severe macro-economic imbalances. This occurred because these countries did not succeed in obtaining a lasting and sustainable balancing of the government budget. Table 5.7 Public investment as a proportion of GDP (%) 1975–81

1982–4

1985–7

Argentina 10.1 7.9 6.5 Brazil 7.5* 5.9 5.7 Colombia 8.2* 7.6 7.7*** Chile 5.5** 5.1 9.2 Mexico 9.6 7.6 5.8 Sources: Chile and Mexico, Eyzaguirre (1980). Argentina, Escude and Guerberoff (1990). Brazil, Giambiaggi (1989). Note: * 1980–1 ** 1978–81 *** 1985–8

The fiscal accounts were imbalanced in the period prior to the external shock in Argentina, Brazil and Mexico. This imbalance was financed in great measure by external creditors. At the beginning of the 1980s the consolidated public sector borrowing needs were 7.5 percentage points of GDP in Argentina (1980), 6.5 points in Brazil (1980–1), and 10 points in Mexico (1980–1). Given the rationing of international credit markets after the debt crisis, these countries not only faced a rise in the fiscal deficit stemming from the increment in the world interest rates but also had to search for alternative sources for financing the existing budget gap. In Argentina, the fiscal deficit had fluctuated between 3 per cent and 7 per cent of GDP during the postwar period. After the debt crisis, in the 1981–5 period, it reached on average 12.5 per cent of GDP. This rise in the fiscal gap was a direct consequence of the state takeover of private external liabilities, coupled with the increment of the world interest rates. The authorities were unable either to raise the tax burden or to reduce the government expenditures in accordance with the higher government indebtedness. When significant adjustments in the public sector accounts were achieved, such a situation did not last long. To a certain extent, that was due to the inability of the successive stabilization attempts to completely revert the inertial behaviour showed by the main items in the fiscal budget. Nonetheless, on the expenditures side, a significant restriction was put into practice during the 1980s. In the 1986–8 period, government outlays were 5 percentage points of GDP lower than in the 1981–3 period. However, these cuts were not guided by sound efficiency principles and, therefore, the curtailing of government expenditures had negative consequences on the resource allocation of the economy. The bulk

TRADE REFORM IN LATIN AMERICA

99

of the adjustment fell on public wages and investment, while at the same time the government was unable to establish a tight control on the deficits stemming from social security, public firms and federal states. These adjustment traits not only worsened resource allocation but also heavily depressed the growth rate in a direct way. The public sector investment GDP ratio fell from 12 per cent in the late 1970s to 7 per cent in 1988. After the hyperinflationary episodes of 1989–90, this rate dropped even further. The government was also unable to increase tax revenues in order to meet the growing trend of external payments. As can be seen in Table 5.8 total revenues fell during the first stage of the adjustment period, increased to a maximum of 40 per cent of GDP during the stabilization attempt of 1985–6, and finally diminished again from 1987 onwards. Not only do the flaws of the tax structure but also macroeconomic instability explain the negative evolution of the public sector revenues. With regard to the tax system, the major factor contributing to the above-mentioned result was the erosion of the tax base. With regard Table 5.8 Argentina: public sector. Savings-investment accounts. (percentage of GDP) Resources Current Taxes Non-taxes Capital Economic Emerg. Financing1 Other Resources Expenditures Current Capital PSBR Source: Carciofi (1990). Note: 1 Compulsory savings, 2 Estimated

1980

1981

1982

1983

1984

1985

1986

1987

19882

36.43 35.53 23.26 12.27 0.31 – 0.59 43.90 34.39 9.51 7.47

35.78 34.66 20.35 14.32 0.25 – 0.87 49.03 39.39 9.65 13.26

33.07 32.25 18.71 13.54 0.49 – 0.33 48.18 39.62 8.56 15.11

34.59 34.12 18.40 15.71 0.23 – 0.24 49.75 40.06 9.68 15.15

33.46 33.03 18.19 14.84 0.24 – 0.19 45.38 37.56 7.82 11.92

41.52 40.53 22.05 18.48 0.25 0.66 0.08 47.53 40.47 7.06 6.01

39.36 38.24 22.34 15.90 0.21 0.58 0.33 44.09 36.61 7.48 4.73

36.62 35.98 21.14 14.84 0.20 0.11 0.33 44.51 36.11 8.41 7.89

35.74 34.39 19.18 15.21 0.51 0.75 0.09 43.85 36.43 8.42 8.11

to direct taxes, the tax base eroded both because of the existence of loopholes in the profits tax legislation and the denationalization of the country’s wealth caused by the capital flight phenomenon. The poor performance of direct taxes made the indirect ones the major source of government revenues. Equity considerations aside, there were also many reasons that impeded a significant increase in indirect tax collection. First, the base of the value added tax was eroded by the exemptions originating in the Fiscal Program for promoting the localization of industries in the less developed regions of the country. Second, the revenues stemming from the taxation of primary exports followed a changing pattern both because of the highly unstable evolution of the prices of primary exports during the 1980s and the conflictive political interaction with primary exporters that determined large and recurrent changes in the export tax rate (derechos de exportación). Lastly, the tax on fuel, which represented a mounting share of total tax collection throughout the period, greatly contributed to fostering inflation and then to eroding the real value of tax revenues via the Olivera-Tanzi effect.

100

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

There were two major channels through which macroeconomic instability affected tax collection: the Olivera-Tanzi effect and the deepening of fiscal evasion. The former negatively affected tax revenues because the inflation rate was permanently higher during the adjustment period than in the pre-adjustment one. As is well known, the Olivera-Tanzi effect states that there is a negative relationship between inflation and total real tax revenues because of the existence of lags in tax collection. Tax evasion, on the other hand, deepened because of the dollarization process of the economy and the worsening in the Central Government administrative efficiency provoked by the distortive cuts in government expenditures. It was estimated that the tax base erosion, the Olivera-Tanzi effect and fiscal evasion together represented a loss of real tax revenues in the order of 10 per cent of GDP per annum. It is worth while mentioning that this amount is similar to that of the average fiscal deficit observed during the 1980s. The evolution of the fiscal accounts throughout the entire adjustment period was characterized by the persistence of an average deficit/GDP ratio well above the one registered in the pre-crisis period. This was so in spite of the fact that real expenditures significantly diminished. In fact, after each stabilization policy briefly resulted in a certain balance between public receipts and outlays, the fiscal deficit tended to return to its previous levels. This particular feature of the Argentine experience mainly resulted from the dynamic properties shown by the tax system in a context characterized by sharp accelerations and decelerations in inflation. In fact, a kind of hysteresis phenomenon was observed, which was associated with the temporal pattern followed by the tax burden. On theoretical grounds, the Olivera-Tanzi effect assumes that the variations in the real value of tax collection is a function of changes in the inflation rates, and this relationship is supposed to by symmetric. Like the external gap, the fiscal gap in Argentina tended to become a self-perpetuating (rather than a selfcorrecting) disequilibrium because of the operation of the above-mentioned mechanisms originating in the macroeconomic instability. These features of the fiscal and external gaps, with other distortions of the domestic financial markets (that we shall analyse later), tended to put both the inflation rate and the evolution of domestic financial assets on explosive paths. These kinds of self-perpetuating disequilibria that tend to become explosive (for example, in the form of hyperinflationary episodes) characterized the macroeconomic setting in both Argentina and Brazil, in spite of the fact that Brazil showed a lower degree of external indebtedness and was less affected by the negative external shock. In the case of Brazil, there are different stages in the evolution of the fiscal deficit (see Table 5.9). Between 1980–2 and 1984, the fiscal deficit showed a significant decrease. This result was achieved despite the higher burden of financial payments, because of an important primary surplus. However, from 1984 onwards, the primary surplus Table 5.9 Brazil: public sector deficit (percentage of GDP) 1980 1981 1982 1983 1984 1985 1986

Deficit

External and domestic interests

Primary deficit

6.8 6.3 7.3 4.4 2.7 4.3 3.6

3.6 4.0 5.4 6.0 6.8 6.8 4.6

3.2 2.3 1.9 −1.6 −4.1 −2.5 −1.0

TRADE REFORM IN LATIN AMERICA

Deficit

External and domestic interests

1987 5.9 4.3 1988 4.3 4.5 Source: Giambiaggi (1989).

101

Primary deficit 1.6 −0.2

followed a declining path. Consequently, the fiscal deficit began to grow, even though the amount of interest payments on the external debt tended to diminish following the decreasing trend of world interest rates. The major explanation for the reduction registered in the deficit between 1980 and 1984 lies in the declining path followed by government expenditures, since the tax burden also tended to decline throughout this period. In the 1980–4 period, the sum of current expenditures, public investment, subsidies and transfers (including social security payments and excluding transfers to public enterprises) fell from 23 per cent of GDP to 19 per cent. The falling trend in public expenditures reverted from 1985 onwards, among other factors because there was an increase in public wages: between 1984 and 1987 wages rose by about 3 percentage points of GDP. However, beyond the fluctuations shown by the different expenditures items, as in the case of Argentina, the public investment rate was the item that suffered the greatest adjustment. While it averaged 7.5 per cent of GDP in 1980–1, it fell to 5.7 per cent in the 1985–7 period (Table 5.7). In Brazil, the destabilizing fiscal trends originated primarily in the growing burden of interest payments, associated with the increasing stock of domestic public debt in a context of high inflation and interest rates. Table 5.10 shows the high velocity at which the financial burden grew during the 1980s. Table 5.10 Brazil: financial transfers from government to private sector, 1970–86 (percentage of GDP) Payments of domestic interests Monetary correction 1970–4 0.72 1975–9 1.34 1980 1.15 1981 1.19 1982 2.19 1983 2.23 1984 3.50 1985 7.12 1986 7.28 Source: Bodin de Moraes (1989).

Real 0.50 0.57 0.74 1.07 1.23 1.93 2.71 3.72 2.96

The behaviour of tax revenues showed some similarities with Argentina (Table 5.11). The declining trend of tax collection in the early 1980s partially reversed in 1982 because of the increment in the amount of direct taxes levied. However, total tax revenues tended to decline again from 1983 onwards. The improvements introduced in the tax system and the increase in the nominal tax burden were insufficient to offset the joint effect of the recession and the acceleration of inflation. There was a temporary interruption of the declining trend during the Curzado Plan, but it returned after 1987.

102

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

The Mexican fiscal experience was similar to that of Argentina and Brazil. From Table 5.12, it follows that the 1981–2 external shock doubled the public sector borrowing needs, which reached 16.9 per cent of GDP in that year. Throughout the following three years, the government gained some control over the evolution of the fiscal accounts. The borrowing needs were reduced to levels fluctuating between 8.5 per cent and 9.5 per cent of GDP. In spite of the huge increment in the burden of both external and domestic interest payments, the above-mentioned reduction in the fiscal deficit was made possible by the generation of primary surpluses in the order of 4 to 5 per cent of GDP. This performance can be explained by the significant increase in the rents stemming from petroleum exploitation. The revenues originating in this activity rose from 7.8 per cent of GDP in 1980–1 to 15 per cent in 1983–5. The sharp reduction operating in the public investment/ GDP ratio (which fell from 7 per cent in 1980–2 to about 4 per cent in 1983–5) also helped to reduce the deficit. Table 5.11 Brazil: taxes as a proportion of GDP (%) 1970–74 1975–79 1980 1981 1982 1983 1984 1985 1986 1987 1988 Source: Bodin de Moraes (1989). Rezende et al. (1989).

25.9 26.4 24.2 24.6 26.2 24.7 21.6 22.0 24.3 21.6 19.9

Table 5.12 Mexico: public sector accounts (percentage of GDP) Revenues Oil revenues Parastatal sector Federal government Taxes Others Non-financial expenditures Current Capital Extra budget Primary surplus Interest payments Financial intermediation PSBR

1980

1981

1982

1983

1984

1985

1986

1987

1988

1989

25.5 7.8 6.5 11.2 10.4 0.8 28.8 21.0 6.9 0.9 −3.3 3.2 1.0 7.5

25.2 7.7 6.5 11.0 10.2 0.9 33.5 24.6 7.8 1.1 −8.3 4.7 1.1 14.1

27 .8 11 .1 6 .9 9 .7 8 .7 1 .1 31 .6 23 .3 6 .7 1 .6 −3 .8 11.8 1 .3 16 .9

31.7 16.1 6.2 9.4 8.4 1.0 27.1 21.8 4.5 0.8 4.6 12.7 0.5 8.6

31.3 15.1 7.2 9.0 8.3 0.8 26.0 21.1 4.2 0.7 5.3 12.4 1.4 8.5

30.4 13.3 7.8 9.3 8.4 0.9 26.7 22.0 3.8 0.9 3.7 11.7 1.5 9.6

29.3 11.3 8.1 9.9 8.9 1.0 26.9 22.1 3.8 1.0 2.4 16.8 1.1 15.5

29.5 11.9 7.9 9.7 8.6 1.1 24.0 20.1 3.7 0.2 5.5 19.8 1.0 15.3

29.3 10.0 7.9 11.4 9.6 1.8 21.6 18.4 3.1 0.1 7.7 16.9 1.6 10.8

28.8 9.1 6.9 12.8 10.5 2.3 20.8 18.1 2.6 0.1 8.0 13.4 0.6 6.0

TRADE REFORM IN LATIN AMERICA

1980

1981

1982

1983

1984

1985

1986

1987

1988

103

1989

Source: Ortiz and Noriega (1990).

Tax collection, on the contrary, moved in the opposite direction, adding to the destabilizing forces originating in the external shock. The tax revenues/GDP ratio fell from around 10 per cent of GDP to a lower level of about 8.5 per cent in the 1982–5 period. In the 1986–7 period, the fiscal gap worsened again. Two factors played a significant role. In the first place, inflation tended to accelerate, thereby inducing a strong increment in the nominal interest rates. In the second place, there was a fall in the petroleum prices, which provoked a severe reduction in government revenues. In spite of the maintenance of a depressed level of public investment rates, the absence of considerable improvements in either non-interest current expenditures or tax collection induced an increment of the public sector borrowing needs which returned to a level higher than 15 per cent of GDP. From 1988, there was a new reversion in the tendency for the fiscal deficit to increase. This was a direct consequence of the new stabilization plan that is still in effect. The public sector borrowing needs, 10.8 per cent of GDP in 1988, fell to 6 per cent in 1989. This was achieved in spite of the fact that the revenues from petroleum showed an unfavourable evolution. The main reasons behind this result were: the maintenance of depressed levels of the public investment rate, the reduction in the interest payments because of the decline in the inflation rate induced by the stabilization programme, the reduction in non-interest current expenditures, and an incipient but significative improvement in the amount of tax collection. It follows, then, that the new stabilization programme seems to have reverted some of the distortive trends that were present in the prior adjustment attempts. This is especially true concerning the tendency for the tax collection to deteriorate. Nonetheless, it has been too short a time since the implementation of the programme to say whether this new situation will be long-lasting. There are some conclusions that can be extracted from our analysis of the evolution of the fiscal gap. Mainly, the debt crisis made the fiscal structure developed by most Latin-American countries during the postwar development process inviable. The joint action of the fiscal gap and of the denationalization of savings added a new dimension to the problem of the external transfer, which had highly disturbing macroeconomic effects. However, our analysis suggests that the differences were correlated with the role played by: (a) the amount of external financing received by the public sector; (b) the degree of command over the rents of natural resources by the public sector; (c) the size of the initial impact of the external shock on public sector accounts; (d) the efficiency with which public expenditures were reduced, especially with regard to whether the bulk of the adjustment fell on public investment; (e) the efficacy and/or efficiency shown in obtaining the objective of at least maintaining the tax burden. We think that Colombia’s and Chile’s better performance has to do precisely with those factors. In the case of Colombia, the impact of the external shock was weaker. Therefore, the authorities were able to carry out the needed adjustment by implementing conventional fiscal policies, which operated on the margin. There were no major structural fiscal reforms. In addition, the fiscal restraint was facilitated by a less restrictive access to external financing compared with the other countries (excluding Chile) and by state appropriation of a part of the rents generated by natural resources (coffee and petroleum). This, in turn, prevented public investment from suffering a dramatic cut. In Chile, the impact of the external shock was far greater than in Colombia. However, there was a good performance of the fiscal accounts because, first, the government extracted and appropriated the rents of the principal natural resource of the country (copper), which is at the same time the principal export. Second,

104

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

the country managed to carry out a successful negotiation with foreign creditors. Third, during the adjustment period the level of public investment was increased, which in turn partially explains Chile’s better growth performance. Fourth, the tax policies implemented, beyond equity considerations, succeeded in attaining a tax collection in accordance with the higher burden of foreign financial payments. A common feature behind the worse fiscal performances of Argentina, Brazil and Mexico was the weakness of the state in facing the domestic side of the external transfer. In all three countries, this resulted in sharp decreases in the public investment/GDP ratio, which in turn depressed the overall growth rate. The three countries showed a marked inability to induce the needed permanent increases in tax collection to offset the higher level of foreign interest payments. In fact, Argentina’s and Brazil’s highly inflationary environment led to reductions in the tax revenues/GDP ratio. This, associated with the asymmetric behaviour of tax revenues when inflation rates changed, gave rise to a kind of hysteresis phenomenon that, in turn, led to the persistency of high fiscal deficits in spite of the declining trend of public expenditures. The disappointing evolution of the tax burden, in addition to the external credit rationing and the denationalization of savings, led to extremely high domestic interest rates, persistently high rates of inflation and explosive trends in the public domestic debt. STRUCTURAL REFORMS AND STABILIZATION WITH GROWTH According to the analysis developed in the last section, it is now possible to understand better why we chose to integrate Argentina, Brazil, Colombia, Chile and Mexico in our sample of Latin-American countries. Besides accounting for almost 80 per cent of the consolidated regional product, their contrasting growth performances during the 1980s accurately illustrate how different the adjustment to the debt crisis was, according to the structural features, initial conditions and availability of foreign finance for each LatinAmerican economy. They also suggest the limited validity of general policy prescriptions formulated for the region as a whole. In effect, as was said before, the growth rates achieved by Colombia and Chile in the late 1980s were satisfactory, while those of Argentina, Brazil and Mexico were truly disappointing—as was also the case in almost all other countries of the area. Consider first the two countries that did well. For several reasons— good luck and prudent pre-crisis economic policies among the most important ones—Colombia was less affected by the debt crisis. This, in turn, allowed it to successfully adjust to the new international environment in a relatively smooth way, without major changes in its previous growth strategy. From a long-run perspective, then, the main issue for Colombia is how to deal with the ‘neoclassical problem’ of improving resource allocation. Chile, unlike Colombia, was severely affected by the crisis and its adjustment implied a deep change of regime, involving both trade liberalization and a completely different role for the state. However, the Chilean success in restoring growth is still very fragile. On the one hand, the balance of payments and the fiscal accounts remain highly dependent on the behaviour of foreign finance and the terms of trade. On the other hand, because of its high debt/GDP ratio, the economy would continue to face an uncomfortable tradeoff between domestic consumption and investment were it to fulfil all of its external commitments. In the other three countries of our sample, the external shock of the early 1980s was equally severe but, worse than that, its initial impact was amplified through different propagating mechanisms. This resulted in a highly uncertain economic environment and the recurrent emergence of destabilizing tendencies that were reflected in the extreme volatility of the main macroeconomic indicators: the level of activity, the rate of inflation, the key relative prices, the public internal debt, the degree of monetization, etc. All this

TRADE REFORM IN LATIN AMERICA

105

considered, the sharp falls experienced by the investment and growth rates in these countries are hardly surprising. Besides having to resume growth, Argentina and Brazil have yet to overcome the consequences of hyperinflation and stabilize their economies. Mexico, at least, recently managed to stabilize. But stabilization came without growth (the so-called ‘bottom-of-the-well equilibrium’) and entailed the widening of the external gap. Both facts suggest that stabilization is still precarious and raise serious doubts about its continuity. Taking advantage of the comparative analysis developed in the previous sections, regarding the diverse adjustment experiences of these five Latin-American economies, in what follows we try to summarize the main lessons that can be drawn from the standpoint of growth recovery about (a) stabilization and (b) external sector and trade. Stabilization There seems to be a great deal of consensus on the priority of stabilization as a precondition for the recovery of growth. This idea is strongly supported by the experiences of the countries surveyed in the previous section. The worst performances of production, employment and investment were coupled with accelerations of inflation. In contrast, some stabilizations were followed by increases in the level of activity. Nonetheless, beyond this general idea, the consensus is not so broad. As we have seen in the first section, the dominant conventional view establishes a neat separation between stabilization and growth. Stabilization must come first, and it has to be sustained long enough to consolidate expectations and credibility. Only then is the economy ready to begin to grow. First of all, it must be stressed that the conventional view of stabilization heavily relies on pre-1980s diagnoses and prescriptions, as expressed, for example, by the IMF definition: A Financial Program is a set of coordinated policy measures mainly in the monetary, fiscal and balance of payments fields intended to achieve certain economic targets in a relatively short period of time. The task of setting the economic targets, choosing the policy instruments, and quantifying the appropriate magnitudes of the instruments required to reach the targets is described as financial programming. (International Monetary Fund 1981) Stabilization policies were conceived to resolve short-run disequilibria in the balance of payments, fiscal accounts or monetary aggregates. The disequilibria might result from shocks or policy mistakes disturbing the normal function of the economy. The stabilization policy was a set of measures intended to push the economy into a stable equilibrium by a quick correction of deviations. In each concrete episode, stabilization programmes combined short-term and long-term measures. Even more, in many cases stabilization was presented as a precondition of reforms aimed at improving the allocation of resources and increasing the long-term rate of growth. But structural reforms themselves had little to do with the success of the short-run stabilization effort. These notions of stabilization are no longer relevant in the present circumstances of Latin-American economies. First, as we saw in the second section, the disequilibria cannot be managed with the instruments traditionally used in stabilization policies. Second, the closing of the external and fiscal gaps requires structural reforms that, as a consequence, condition the possibility of stabilization.

106

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

There is a new character of stabilization that is derived from its new target. According to our diagnosis, many Latin-American economies follow explosive paths that are equivalent to processes of steady deterioration of their functioning and productive capacities. Under these circumstances, stabilization must be conceived of as a set of measures and policies capable of stopping these processes. A stabilization policy has to obtain: (a) sustainable closures of the fiscal and external gaps and (b) a quick reduction of the rate of inflation to a (relatively) low level. Stability of inflation and key relative prices are necessary conditions for stopping the flight from physical and financial domestic assets, but only necessary. Even in the case of a successful impact on the gaps and inflation, it cannot be expected to completely reverse previous tendencies. The stickiness of the pre-stabilization financial behaviour can be attributed to coordination failures and/or credibility problems. In any case, the post-stabilization financial situation will be fragile and very sensitive to signals from the fiscal or external fronts. The issue of the sustainability of the fiscal gap closure has two aspects. In the first place, because it requires deep changes in the amount and structure of fiscal expenses and receipts, a permanent closure of the fiscal gap involves a state reform, more than marginal adjustments. It will take time and, meanwhile, it has to be externally financed, as it was in Chile and other successful adjustment cases. In the second place, given the budgetary importance of the external debt, a sustainable closure of the fiscal gap requires a firm agreement on debt payments. This means that the fiscal and external gaps have to be closed simultaneously. In all cases, state reforms involve tough actions. Governments have to accumulate enough consensus to take them and to make them last. It seems impossible to reach these targets without external agreement and support. For the same reasons of reducing uncertainty and signalizing stability, the closure of the external gap should take the form of an extended agreement with foreign governments, multilateral lenders and commercial banks. The time structure and amounts of the agreed flow of funds should be compatible with the fiscal programme and the financing of some investment recovery. The agreement should also contain contingent clauses to cushion unfavourable shocks, given that voluntary lending will be inexistent in the foreseeable future. The durability of a low rate of inflation will require a persistent anti-inflationary activity instrumented by a consensual incomes policy. Stability of key relative prices is both a target and a necessary condition of this policy. In this respect, there exists a trade-off between the anti-inflationary role of the nominal rate of exchange and the required stability of the real exchange rate. The success of the anti-inflationary policy demands the utilization of all possible tools to perform a very difficult task. The set of measures includes the use of the nominal exchange rate as an anchor of price and wage decisions and expectations. Because the residual inflation is always higher than the nominal rate of devaluation there is a post-stabilization tendency for the domestic currency to appreciate. The lower the residual inflation the weaker the tendency, but this problem has no neat solution. In each case the stabilization programme has to combine in a difficult mix the two conflicting targets of the exchange rate policy. If all the necessary conditions for stabilization are fulfilled, it is possible to reach a financial and monetary equilibrium with a low rate of inflation. The economy stands in a state where the explosive tendencies have been stopped. Is it possible to consolidate this state without the recovery of some positive rate of growth? The main issue of the answer to this question lies in the fragility of the equilibria and rates of inflation reached by the stabilization effort. It seems to be very difficult, if not impossible, to sustain the poststabilization state without some positive signals from the real side of the economy. In a stagnant situation,

TRADE REFORM IN LATIN AMERICA

107

the probability of falling back to destabilization because of exogenous shocks or distributive conflicts increases with time. Beyond quantitative considerations, the stabilization-cum-growth recovery implies a reconstruction of the ability of the state to perform coordinating and promoting functions. It requires a global political effort to adopt and perform social rules and discipline. With the above-mentioned external agreement, a politically sustainable fiscal pact and a social agreement seem essential. This heroic character of stabilization policy in many Latin-American cases also comes from the need to revert pessimistic expectations, after about ten years of destabilizing economic tendencies and many policy failures. We have just said that stabilization-without-growth recovery is not impossible, but is difficult to sustain. At present, Bolivia and Mexico fall into the stabilized-stagnant category. Aside from fragility considerations, their economies suffer from high real rates of interest, a tendency for their domestic currencies to appreciate and no tendency to revert demonetization and increase domestic financial intermediation. In both cases, but more emphatically in Bolivia, the wait-and-see strategy dominates growth policy without yet exhibiting any positive growth result. Trade reform, debt and balance of payments Aside from some rather vague remarks on the ways to attract direct foreign investments and reverse capital movements, the Washington Consensus remains completely silent on the critical question of how to close the external gap in a way that is consistent with the objective of growth resumption. However, as was extensively discussed in the previous sections, in order to accomplish this goal a significant reduction of real transfers abroad (compatible with both the fiscal programme and investment recovery) seems unavoidable. Moreover, the amount and time profile of the flows of funds involved should be known by debtor countries with much more certainty than is the case nowadays. As Tanzi (1989) puts it: it is clearly better for economic policy when policymakers know for sure that the servicing of the debt will require, say, a future constant payment of 3 percent of GDP, so that they can adjust their policies accordingly, rather than knowing that there is a 50 percent probability that the payment will be 1 percent and a 50 percent that it will be 5 percent of GDP. Perhaps one of the greatest burdens associated with the debt crisis has been the creation of this uncertainty. Another has been the large amount of time that the policymakers had to allocate to this problem. (Tanzi 1989:25) Both questions stress the importance of achieving a firm and extended agreement between debtor countries on the one hand and the governments of creditor countries, multilateral agencies and commercial banks on the other. No matter what specific ways this agreement may take in each case, its targets should not be established in terms of the current account performance, but rather in terms of the evolution of real foreign indebtedness.14 And, it should also include contingent provisions to deal with unexpected shocks in the terms of trade and other variables out of the control of debtor countries. On the other hand, the Washington Consensus does emphatically stand for an outward-oriented growth strategy because it entails, among other things, a more efficient allocation of new investments and, therefore, higher growth.

108

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

The desirability of a more outward-oriented growth strategy raises no controversy at present. However, the critical question remains about how to accomplish it. Although there is no easy answer to this question, it seems worth while discussing it more closely. In the first place, an outward-oriented growth strategy should not be considered just a strategy to increase exports at any rate. Without a matching increase in imports, given the present context of most LatinAmerican countries, any improvement in export performance would only mean larger real transfers abroad, with negative repercussions on investment and growth. The case of Brazil in the 1980s, where the impressive performance of exports did not have any favourable effects on growth, constituted a good example of this point. In the second place, but no less important, the reforms taken with a long-term perspective in order to make the economy more open must not impair stabilization in the short run. This poses a complex problem to policymakers, especially regarding fiscal and exchange rate policies, which cannot be solved on a priori grounds. Let us consider the latter first. According to long-term considerations, there is no doubt that economic policy should ensure that the real exchange rate be high and stable in order to promote export growth and provide the right signals for new investments. However, stabilization requires that the nominal exchange rate be used as an anchor for price and wage decisions and expectations. This means, as we have seen above, that for a certain period the domestic currency is bound to appreciate in real terms. At least, this seems to be the initial outcome of most stabilization attempts, both successful and unsuccessful. Unfortunately, there are no clear-cut answers to this conflict and it makes no sense to try to find general solutions to it, applicable at any time or place. Given the many and conflicting tasks that are simultaneously attached to the exchange rate and the high uncertainty experienced in these economies, however, it seems prudent not to rely on a floating regime to do the job. This contention is justified both on theoretical grounds by the ‘overshooting’ literature,15 and on the empirical evidence of extreme volatility of exchange rates under floating regimes registered in some Latin-American countries. All this suggests that some direct control of the Central Bank over the evolution of the exchange rate, regardless of its specific nature, seems unavoidable. The fiscal policy is also subject to similar tensions between the short and long runs. With the latter in mind, the WC approach stands for suppressing all taxes on exports. However, export taxes are a vital source of revenues for the public sector in many Latin-American countries, and eliminating them without previously making sure that they will be replaced by other taxes, capable of generating an equivalent amount of fiscal revenues, can be extremely dangerous for both the fiscal balance and stabilization. Moreover, the lack of predictability of tax policy in such a context can have negative repercussions on export performance, which are at least as serious as those originating on export taxes. The recurrent frustrated attempts to definitively lift taxes on grain exports in Argentina clearly illustrate this point. There is a second question associated with this. As we have seen above, the rents stemming from natural resources in several Latin-American countries constitute a relevant source of public revenues. In some cases the state has direct command over the rents, because it has property rights on the resources. Mexico, Chile and Colombia are good examples of this situation. Sometimes, however, although it is certainly not the best alternative, the only available way to socialize these rents is to impose taxes on primary exports. Argentina might be in this situation. This raises a more general question, namely that of export promotion. The successful experience of East Asian NICs, as was mentioned above, was based on a ‘pick-the-winners’ strategy. Among other things— maintaining an undervalued currency was, needless to say, one of them—the strategy implied an active role of the state in selecting some leading economic sectors and promoting their sales abroad in several ways. In

TRADE REFORM IN LATIN AMERICA

109

Latin America, Brazil’s impressive export performance since the 1970s and the success of Colombia in expanding its ‘minor’ exports also point in the same direction. Even in Chile, export promotion was not entirely absent.16 Given the constraints that the current situation imposes on the use of tax exemptions and credit support, the prudent use of multiple exchange rates seems justified as a second-best solution. Recent failures of multiple exchange rate systems in some Latin-American countries should in fact be attributed to destabilization and government mismanagement, rather than to insurmountable deficiencies of the system itself. Multiple rates naturally call for some exchange controls. The failure of financial markets to spontaneously prevent the denationalization of savings without disruptive domestic consequences strengthens the need for exchange controls in order to limit capital movements. The inconvenience of liberalizing the capital account in the present context is even shared by some of the participants of the Washington Consensus.17 There remains, finally, the issue of import liberalization. According to what has been said above, one of the main goals of an outward-oriented growth strategy should be to increase imports, at least as fast as exports. Trade reform is also required to remove anti-export biases and provide more adequate signals to investment decisions. Most of the WC proposals on this subject are certainly valid, but they need some qualifications. Consider first the recommendations to eliminate all non-tariff barriers, reduce the average level of tariff protection and diminish sectoral differences in the tariff structure. Generally speaking, these are obviously sound recommendations. Even more so when the access to intermediate and capital goods imports at competitive international prices is at stake. Nevertheless, they deserve a few additional remarks. It should be noted, on the one hand, that most Latin-American countries have already gone a long way in the process of lifting non-tariff barriers and rationalizing their tariff structures. It is worth remembering, on the other hand, that the maintenance of some level of import protection can be justified at least on two grounds. First, tariff receipts represent a major source of fiscal revenues that cannot be dispensed with easily. Second, it is accepted both theoretically and empirically that infant industries require some kind of import protection, though on a strictly temporary and decreasing basis.18 Because of the comparative static nature of their approach, however, the most critical omission in the WC proposals on trade liberalization concerns its dynamic aspects. This is not a minor problem. The dramatic failure of the late 1970s Southern Cone liberalization attempts can be attributed precisely to dynamic factors.19 It is worth recalling the main lessons of those experiences. One of them refers to the appropriate sequence of tariff reductions. If the basic goal of trade reform is to provide adequate incentives for improving resource allocation and investment decisions, liberalization should begin with the intermediate and capital goods sectors. Otherwise, as experience shows, it can lead to massive imports of final consumer goods with devastating consequences on the balance of payments and the activity level. Needless to say, such an outcome can only have negative repercussions on ‘animal spirits’ and investment. There is also the critical issue of timing and opportunity of the trade reform. The above-mentioned conflict between the short and long run reappears with this issue. Trade reform is for the long run and policymakers should resist the temptation to use it for stabilization purposes. In fact, as also shown by experience, the combination of an abrupt liberalization and an overvalued domestic currency can severely impair expectations and become an explosive mix for both stabilization and growth.

110

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

This is not to deny the beneficial influence liberalization, as well as other long-term reforms, can have in consolidating stability. Rather, the point we want to emphasize is that a relatively stable environment, and particularly a high and stable real exchange rate should be considered basic prerequisites for a successful liberalization. Also, because of its long-term character, trade reform should proceed gradually. Gradualism is required to allow an orderly relocation of the productive factors, therefore reducing uncertainty and enhancing investment prospects. This was the approach taken by Colombia, the Latin-American country with the best macroeconomic performance throughout the 1980s. Finally, given the importance of imperfect competition, intra-industry specialization and economies of scale as determinants of current trends in international trade—forcefully pointed out by the new theories on this subject20—Latin-American countries are well advised not to rely on market mechanisms and price signals as the only instruments of an outward-oriented growth strategy. In fact, according to several authors, price signals did not play such a central role for investment decisions in the successful development processes that followed this strategy in East Asia.21 In this respect, the opportunities that regional integration can open for growth recovery in Latin America should not be dismissed. The process of integration initiated by Argentina, Brazil and Uruguay, therefore, represents a step forward in the right direction, though major results cannot be expected from this initiative before a stabler environment is reached in the area. APPENDIX 5.1 ADDITIONAL STATISTICAL INFORMATION Note The following tables (Tables 5.13 to 5.23) do not have text references in Chapter 5, but are included in order to provide additional statistical information on the Latin-American countries. Table 5.13 Gross Domestic Product and per capita GDP Argentina GDP Rate GDP/Pop Rate Brazil GDP Rate GDP/Pop Rate Colombia GDP Rate GDP/Pop Rate Chile

1980

1981

1982

1983

1984

1985

1986

1987

1988

100

93.0 −7.0 91.6 −8.4

87.6 −5.8 85.0 −7.2

89.9 2.6 86.0 1.2

91.9 2.2 86.8 0.9

87.7 −4.6 81.6 −6.0

92.8 5.8 85.3 4.5

94.5 1.8 85.7 0.5

91.7 −3.0 82.1 −4.2

95.6 −4.4 93.5 −6.5

96.2 0.6 91.9 −1.7

92.9 −3.4 86.8 −5.6

97.6 5.1 89.2 2.8

105.8 8.4 94.7 6.2

113.7 7.4 99.6 5.2

117.9 3.7 101.1 1.5

117.5 −0.3 98.7 −2.4

102.3 2.3 100.1 0.1

103.3 1.0 99.0 −1.1

105.3 1.9 98.7 −0.3

109.3 3.8 100.3 1.6

113.4 3.8 100.3 1.6

121.4 7.1 106.8 4.9

128.4 5.8 110.6 3.6

132.9 3.5 112.1 1.4

100

100 100

100 100

TRADE REFORM IN LATIN AMERICA

1980 GDP 100 Rate GDP/Pop 100 Rate Mexico GDP 100 Rate GDP/Pop 100 Rate Source: CEPAL(1989).

1981

1982

1983

1984

1985

1986

1987

1988

105.2 5.2 103.5 3.5

91.5 −13.0 88.4 −14.6

91.0 −0 .6 86.5 −2.2

96.5 6.0 90.2 4.3

98.8 2.4 90.8 0.7

104.1 5.4 94.1 3.6

109.7 5.4 97.5 3.6

117.5 7.1 102.7 5.3

108.8 8.8 106.3 6.3

108.1 −0.6 104.0 −2.2

103 .6 −4 .2 96.3 7 .4

107.3 3.6 97.5 1.3

110.1 2.6 97.7 0.2

105.9 −3.8 91.9 −5.9

107.4 1.4 91.1 −0.9

108.6 1.1 90.1 −1.1

Table 5.14 Rate of growth of exports (%) Argentina Value Volume Unit. Value Brazil Value Volume Unit. Value Colombia Value Volume Unit. Value

1981

1982

1983

1984

1985

1986

1987

1988

1980–8

14.0 17.7 −3.2

−16.6 −2.1 −14.9

2.8 13.6 −9.5

3.4 −8.5 13.0

3.7 19.9 −13.5

−18.4 −12.7 −6.6

−7.2 −11.3 4.7

43.1 24.9 14.6

13.6 38.9 −18.3

15.6 23.0 −6.0

−13.3 −6.9 −6.9

8.6 17.0 −7.2

23.3 19.4 3.2

−5.1 1.9 −6.8

−12.7 −13.7 1.2

16.8 19.4 −2.2

29.5 16.9 10.7

68.2 96.4 −14.5

−20.8 −10.8 −11.2

−1.4 −4.9 3.7

−4.6 −2.9 −1.7

15.1 11.2 3.5

4.4 10.8 −5.8

46.0 23.6 18.1

−1.5 22.2 −19.4

1.4 −2.6 4.0

30.5 49.3 −12.6

1981

1982

1983

1984

1985

1986

1987

1988

1980–8

−3.4 16.7 −17.2

3.4 5.0 −1.5

−4.7 0.8 −5.5

4.2 18.1 −11.8

10.4 6.6 3.5

24.4 5.8 17.6

35.0 5.0 28.5

49.9 61.5 −8.3

6.5 27.1 −16.2

5.1 19.4 −12.0

8.4 8.8 −0.4

−10.5 −9.2 −1.4

−26.0 1.9 −27.4

28.9 16.2 10.9

0.0 6.7 −6.3

26.5 112.0 −40.4

Chile Value −18.5 Volume −6.5 Unit. Value −12.8 Mexico Value 22.1 Volume 11.9 Unit. Value 9.1 Source: CEPAL (1989).

111

112

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

Table 5.15 Gross fixed investment as proportion of GDP at constant prices (%) Argentina Brazil Colombia Chile Mexico Source: CEPAL (1990). Note: *Preliminary data.

1970–9

1980–1

1982–4

1985–8

1989*

21.7 24.5 17.3 17.5 23.0

20.9 22.0 17.1 17.0 25.7

13.8 17.5 17.5 12.8 18.6

12.0 17.6 15.4 14.9 16.8

9.9 17.7 15.2 17.2 17.9

Table 5.16 Savings, investment and resource transfers (percentage of GDP at constant prices) Argentina National savings Foreign savings Domestic savings Real transfers Gross domestic investment Brazil National savings Foreign savings Domestic savings Real transfers Gross domestic investment Colombia National savings Foreign savings Domestic savings Real transfers Gross domestic investment Chile National savings Foreign savings Domestic savings Real transfers Gross domestic investment Mexico National savings Foreign savings Domestic savings

1980–1

1982–4

1985–8

16.1 6.1 20.0 2.2 22.2

7.8 6.7 19.6 −5.1 14.5

7.5 3.9 20.7 −9.3 11.4

17.9 5.0 23.5 −0.6 22.9

17.5 1.5 24.9 −5.9 19.0

17.4 0.5 23.9 −6.0 17.9

15.1 4.6 17.0 2.7 19.7

10.1 7.3 17.4 3.0 20.4

13.3 4.0 19.2 −1.9 17.3

9.9 12.8 15.6 7.1 22.7

7.6 5.2 16.5 −3.7 12.8

8.8 6.1 24.8 −9.9 14.9

21.0 4.7 24.9

12.2 6.5 21.1

16.5 0.3 26.7

TRADE REFORM IN LATIN AMERICA

Real transfers Gross domestic investment Source: CEPAL(1989).

0.8 25.7

−2.4 18.7

113

−9.9 16.8

Table 5.17 Chile: public sector deficit* (percentage of GDP) Consolidated public sector**

Non-financial public sector

1978 1.44 1.44 1979 4.61 4.61 1980 5.36 5.36 1981 0.39 0.39 1982 −3.52 −9.18 1983 −3.45 −7.71 1984 −4.51 −9.26 1985 −2.86 −10.11 1986 −1.45 −4.34 1987 0.25 −1.01 Source: Larranaga (1989). Note: * It does not include privatization receipts. ** Including debt services of financial public sector and the quasi-fiscal deficit. Table 5.18 Chile: revenues and expenditures of central government (percentage of GDP) Total expenditures (excl. debt services)

Debt services Direct taxes Indirect taxes Non-taxes revenue

1980 20.4 2.7 6.2 1981 23.3 1.6 6.3 1982 26.6 1.9 6.5 1983 25.8 2.6 4.6 1984 25.7 3.1 4.6 1985 26.0 6.5 4.2 1986 25.7 4.3 4.1 1987 24.0 4.3 3.7 1988 23.5 7.2 3.0 Source: Banco Central de Chile (1989).

14.3 15.9 15.6 16.8 17.9 19.4 19.8 20.2 18.3

1.9 2.9 2.5 1.2 1.0 1.4 2.0 2.7 3.7

Cooper revenues

Central government deficit

3.8 1.5 1.6 2.0 1.3 1.2 1.3 1.6 4.0

−3.1 −1.7 2.3 3.8 4.0 6.3 2.8 0.1 1.7

Table 5.19 Brazil: public sector expenditures in wages and investment (percentage of GDP) Wages 1976–80 1980

Investment

National accounts

Public enterprises

Total

National accounts

Public enterprises

Total

6.8 6.3

– 2.0

– 8.3

3.1 2.4

– 4.6

– 7.0

114

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

Wages

Investment

National accounts 1981 6.4 1982 7.0 1983 6.5 1984 5.5 1985 6.8 1986 7.2 1987 7.5 Source: Giambiaggi (1989).

Public enterprises

Total

National accounts

Public enterprises

Total

2.2 2.3 2.0 1.7 1.9 2.0 2.5

8.6 9.3 8.5 7.2 8.7 9.2 10.0

2.6 2.3 1.8 1.9 2.3 3.0 2.7

5.3 4.8 3.7 3.3 3.1 2.8 3.3

7.9 7.1 5.5 5.2 5.4 5.8 6.0

Table 5.20 Rates of inflation. Consumer prices (%) 1980 Argentina Dec/Dec 87.6 Annual average 100.8 Brazil Dec/Dec 86.3 Annual average 82.8 Colombia Dec/Dec 26.5 Annual average 27.2 Chile Dec/Dec 31.2 Annual average 35.1 Mexico Dec/Dec 29.8 Annual average 26.3 Source: CEPAL(1989).

1981

1982

1983

1984

1985

1986

1987

1988

1989

131.3 104.5

204.7 164.8

433.7 343.8

688.0 626.7

385.4 672.2

81.9 90.1

174.8 131.3

387.7 343.0

3731.0

100.6 105.5

101.8 98.0

179.9 142.0

208.7 196.8

248.5 326.9

63.5 143.7

432.3 231.7

1006.4 682.3

1476.1

26.7 28.1

23.9 24.6

16.7 19.8

18.3 15.9

22.7 24.6

20.7 18.6

24.6 23.5

28.3 28.5

27.1

9.5 19.7

20.7 9.9

23.1 27.3

23.0 19.9

26.4 30.7

17.4 19.5

21.5 19.9

12.7 14.7

21.1

28.7 27.9

98.8 58.9

80.8 101.9

59.2 65.4

63.7 57.7

105.7 86.2

159.2 131.8

51.7 114.2

18.2

Table 5.21 Capital flight (millions of current dollars) 1979 1980 1981 1982 1983 1984 1985 1986

Argentina

Brazil

Chile

Mexico

2995 6400 6494 2563 1415 −626 −1334 1099

481 2707 −404 484 106 1012 2966 6024

−213 −580 −149 734 38 957 nd nd

2590 5019 6619 6435 9724 3280 5327 2427

TRADE REFORM IN LATIN AMERICA

Argentina

Brazil

Chile

115

Mexico

1987 −393 14886 nd 7118 Sources: Argentina: Fanelli, Frenkel and Rozenwurcel (1990). Brazil and Mexico: Cardoso (1989). Chile: Arellano and Ramos (1987). Table 5.22 Argentina: monetary assets (percentage of GDP) Private M1

M2

(currency+demand deposits)

(M1+short-term deposits)

1980 7.5 1981 6.2 1982 4.8 1983 3.8 1984 3.6 1985 3.5 1986 5.6 1987 5.1 1988 3.2 1989 2.7 Source: Central Bank of Argentina.

28.4 28.0 19.6 11.2 10.3 10.9 17.1 18.5 14.7 10.6

Table 5.23 Brazil: monetary and financial assets (percentage of GDP) High-powered money 1980 3.4 1981 2.7 1982 2.9 1983 2.3 1984 1.7 1985 1.6 1986 3.1 1987 2.0 1988 1.3 Source: Giambiaggi (1989).

M1 (1) Public debt (2)

Savings deposits (3)

Term deposits (4)

M2 (1)+(2)

9.3 7.7 6.5 5.1 3.6 3.5 7.7 4.2 2.4

5.9 9.3 7.8 8.7 8.5 8.7 7.9 8.8 9.1

1.5 3.2 5.1 5.0 5.6 5.9 6.2 4.6 3.4

13.5 12.9 13.0 11.0 9.7 13.3 16.9 13.6 13.1

4.2 5.2 6.5 5.9 6.1 9.8 9.2 9.4 10.7

NOTES 1 Even the economic terminology referring to the situation and the needed reforms is suffering from the effects of the crisis. According to Feinberg (1990) ‘Washington did appropriate the language of structuralism but turned it on its head. Whereas in Latin America “structural flaws” meant market failures and “structural change” meant government action, in contemporary Washington it is government interventions that are structural distortions, and liberalization and deregulation that are the corresponding necessary structural reforms’ (p. 22).

116

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

2 Williamson (1990) defines ‘Washington’ as ‘the political Washington of Congress and senior members of the administration and the technocratic Washington of the international financial institutions, the economic agencies of the US government, the Federal Reserve Board, and the think tanks’. According to Williamson, however, ‘Washington does not, of course, always practice what it preaches to foreigners’ (p. 7). 3 See Fischer (1987), which analyses the main issues raised by the liberalization literature and the bibliography cited there. 4 See Kahn et al. (1986). 5 We shall follow here the interpretation of Balassa et al. (1986) because it is the more complete and the better informed and also because it represents quintessentially the Washington Consensus interpretation of the LatinAmerican experience with the ISI model. 6 See McKinnon (1973). 7 For an alternative and intellectually challenging interpretation of Korea’s development process see, for example, Amsden (1989). 8 See Guitian (1987). 9 See Corden (1990) for the case of Turkey and other successful stabilization experiences that received strong support from abroad. 10 More statistical information is given in Appendix 5.1 (p. 154). 11 See Meller (1990). 12 See Fanelli, Frenkel and Winograd (1987); Frenkel and Rozenwurcel (1990); Taylor (1990). 13 See Frenkel and Abadie (1990). 14 The same reasons that justify the use of the operational deficit as a target for fiscal performance apply here. 15 Dornbusch (1976) is the classic reference on this subject. 16 See, for example, Meller (1988) and Ffrench-Davis et al. (1990). 17 See Williamson (1990). 18 See, for instance, Krugman (1988) and Williamson (1990). 19 See Corbo et al. (1986) and the extensive literature there cited. 20 See, for instance, Krugman (1988) and the extensive literature on the subject mentioned there. 21 See, for instance, Sachs (1987), Taylor (1988), and Amsden (1989).

REFERENCES Amsden, A. (1989) Asia’s Next Giant: South Korea and Late Industrialization. Oxford and New York: Oxford University Press. Arellano, J.P. and Ramos, J. (1987) ‘Fuga de capitales en Chile: magnitud y causas’, Estudios CIEPLAN, No. 22, Santiago, Chile: CIEPLAN. Balassa, B., Buenco, G.M., Kuczynsky, P.P. and Simonsen, M. (1986) Toward Renewed Economic Growth in Latin America, Washington DC: Institute for International Economics. Banco Central de Chile (1989) Indicadores Económico Sociales, Santiago, Chile: Banco Central de Chile. Banco Central de la Republica Argentina, Boletín Estadístico, various issues, Buenos Aires: Banco Central de la Republica Argentina. Bande, J. and Ffrench-Davis, R. (1989) ‘Copper policies and the Chilean economy 1973–88’, CESCO Documento de Trabaio No. 4, Santiago, Chile: CESCO. Bhagwati, J. (1987) ‘Outward orientation: trade issues’, in V.Corbo, et al. (eds) Growth-Oriented Adjustment Programs, Washington DC: IMF and World Bank, pp. 257–90. Bodin de Moraes, P. (1990) ‘La tasa de ahorro nacional del Brasil en los años ochenta’, in C.Massad and N.Eyzaguirre (eds) Ahorro y Formación de Capital, Buenos Aires: Editorial GEL. Carciofi, R. (1990) ‘La desarticulación del pacto fiscal. Una interpretación sobre la evolución del sector público argentino en las dos últimas décadas’, Documento de la CEPAL No. 36, Buenos Aires: CEPAL.

TRADE REFORM IN LATIN AMERICA

117

Cardoso, E. (1989) ‘External factors in inflation stabilization: Brazil and Mexico compared’, (mimeo) Campinas, Brazil: University of Campinas. CEPAL, (1989) Estudios Económicos de America Latina 1988, Santiago, Chile: CEPAL. Corbo, V., De Melo, J. and Tybout, J. (1986) ‘What went wrong with the recent reforms in the Southern Cone’, Economic Development and Cultural Change, 34, 3. Chicago: University of Chicago. Corbo, V, Goldstein, M. and Khan, M. (eds) (1987) Growth-oriented Adjustment Programs, Washington DC: IMF and World Bank. Corden, M. (1990) ‘Macroeconomic policy and growth: some lessons of experience’. Annual Conference on Development Economics 1990, World Bank, Washington DC. Dornbusch, R. (1976) ‘Expectations and exchange rate dynamics’, Journal of Political Economy, December. Escude, G. and Guerberoff, S. (1990) ‘Ajuste macroeconómico, deuda externa y ahorro en la Argentina’, in C.Massad and N.Eyzaguirre (eds) Ahorro y Formación de Capital, Buenos Aires: Editorial GEL. Eyzaguirre, N. (1990) ‘Ahorro c inversion bajo restricción externa y fiscal: El caso de Chile’, in C.Massad and N.Eyzaguirre (eds) Ahorro y Formación de Capital, Buenos Aires: Editorial GEL. Fanelli, J.M., Frenkel, R. and Winograd, C. (1987) Stabilization and Adjustment Policies and Programmes. Country Study 12. Argentina, Helsinki: WIDER. Fanelli, J.M., Frenkel, R. and Rozenwurcel, G. (1990) ‘Ahorro, inversion y financiamiento en la Argentina y Filipinas: Un analisis comparado’, Documento CEDES No. 38, Buenos Aires: CEDES. Feinberg, R.E. (1990) ‘Comment’ in J.Williamson LatinAmerican Adjustment: How Much has Happened? Washington DC: Institute for International Economics, pp. 21–5. Ffrench-Davis, R., Leiva, P. and Madrid, R. (1990) ‘Evaluación de la política comercial en Chile’, (mimeo), Geneva: UNCTAD. Fischer, S. (1987) ‘Economic growth and economic policy’ in V. Corbo, et al. (eds) Growth-Oriented Adjustment Programs, Washington DC. IMF and World Bank, pp. 151–78. Fischer, S. (1990) ‘Comment’, in J. Williamson, Latin American Adjustment. How Much has Happened? Washington DC: Institute for International Economics, pp. 25–9. Fischer, S. and Husain, I. (1990) ‘Managing the debt crisis in the 1990s’, Finance and Development, June 1990. Frenkel, R. and Abadie, F. (1990) ‘The role of copper in the Chilean economic performance’, (mimeo), Buenos Aires: CEDES. Frenkel, R. and Rozenwurcel, G. (1990) ‘Restricción externa y generación de recursos para el crecimiento en la America Latina’, El Trimestre Económico 225, Mexico. Giambiaggi, F. (1989) ‘El desequilibrio interno’, in Perspectivas de la Economía Brasileira 1989, Rio de Janeiro: INPES. Guitian, M. (1987) ‘Adjustment and economic growth: their fundamental complementarity’, in V.Corbo, et al. (eds) Growth-Oriented Adjustment Programs, Washington DC: IMF and World Bank, pp. 63–94. International Monetary Fund Financial Policy Workshop (1981) ‘The case of Kenya’, Washington DC: IMF Institute. Kahn, M., Montiel, P. and Nadeem, H. (1986) ‘Adjustment with growth: relating the analytical approaches of the World Bank and the IMF’, World Bank Discussion Paper, on Development Policy Issues, Washington DC: World Bank. Krugman, P. (1988) ‘La nueva teoría del comercio internacional y los países menos desarrollados’, El Trimestre Económico 217. Larrañaga, O. (1989) ‘El deficit del sector público y la política fiscal en Chile 1978–87’, (mimeo) Santiago, Chile: CEPAL. McKinnon, R. (1973) Money and Capital in Economic Development. Washington DC: The Brookings Institution. Meller, P. (1988) ‘El cobre y la generación de recursos externos durante el régimen militar’, Estudios CIEPLAN No. 24, Santiago, Chile: CIEPLAN. Meller, P. (1990) ‘Chile’, in J.Williamson (ed.) Latin American Adjustment: How Much has Happened?Washington DC: Institute for International Economics. Michalopoulos, C. (1987) ‘World Bank programs for adjustment and growth’, in V.Corbo, et al., Growth-Oriented Adjustment Programs, Washington DC: IMF and World Bank, pp. 15–62.

118

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

Ortiz, G. and Noriega, C. (1990) ‘Rationalizing the public sector: the Mexican experience in 1982–90’, (mimeo), Mexico D.F.: Secretaria de Hacienda. Rezende, F., Afonso, J.R., Villela, R. and Varsano, R. (1989) ‘A questao fiscal’, in Perspectivas de la Economía Brasileira 1989, Rio de Janeiro: INPES. Sachs, J. (1987) ‘Trade and exchange rate policies in growth-oriented adjustment programs’, in V.Corbo, et al. GrowthOriented Adjustment Programs, Washington DC: IMF and World Bank, pp. 291–325. Selowsky, M. (1990) ‘Stages in the recovery of Latin America’s growth’, Finance and Development, June 1990. Tanzi, V. (1989) ‘Fiscal policy and economic reconstruction in Latin America’, IMF Working Paper, Washington DC: IMF. Taylor, L. (1988) ‘La apertura económica. Problemas hasta fines del siglo’, El Trimestre Económico 217. Taylor, L. (1990) ‘Gap disequilibria: inflation, investment, saving and foreign exchange’, in Income Distribution, Inflation and Growth: Lectures in Structuralist Macroeconomic Theory, (mimeo), Cambridge, Mass.: MIT. Villar, L. (1989) ‘Comercio exterior y políticas de ajuste en Colombia en la década de los’ 80s’ (mimeo), Bogota: FEDESARROLLO. Williamson, J. (1990) ‘What Washington means by policy reform’, in J. Williamson, Latin American Adjustment: How Much has Happened? Washington DC: Institute for International Economics, pp. 5–20.

6 Protectionist pressures in the 1990s and the coherence of North-South trade policies Juan A.de Castro

INTRODUCTION The growing difficulties faced today in explaining present patterns of North-South trade in manufactures are often the result of an incorrect perception of the real determinants of such trade, as well as of the way protectionist tensions are generated and percolate through the international trading system. Economic and trade policy decisions taken, either by developed or developing countries, are becoming, with the increase of interdependence, more interconnected than ever, a fact magnifying the potential cost of ill-conceived trade policies, in terms of their pervasive effects among trading partners. In addition, the evolution of forms in which North-South trade has been taking place in recent years, as a reflection of the evolving nature of comparative advantage, together with other interrelated factors, is increasingly leading to the appearance of various types of dichotomies in the trading system. It is precisely those dichotomies that render the present conceptualization and analysis of such new realities of trade more complex than ever. The relative success of some developing economies in achieving greater competitiveness and attaining increased market access in developed market-economy countries (DMECs) markets has not been an homogeneous process. This has led in recent years to the appearance of a dichotomy in the developing world. Some countries have succeeded better than others in external markets, attaining comparatively higher levels of economic growth and reinforcing their import capacity, most of the time, through outwardlooking types of development strategies. In contrast, other developing countries have encountered severe difficulties in implementing their strategies, either because of the strong pressures for protection prevailing in DMECs markets, which affect with great virulence the product composition of their exports, or because of the incipient industrial infrastructure still prevailing in many of those economies. As a reflection of this dichotomy, while some developing countries are succeeding in expanding their exports in the ‘dynamic’ area, and manage to face fewer obstacles to trade, others, unable to shift comparative advantage smoothly from sensitive to non-traditional sectors, are being penalized to a greater degree by the protectionist forces overhanging ‘grey’ areas of trade. In fact, it is structural convergence among trading partners that seems to be the prerequisite for any future development of greater free trade exchanges in the world economy.1 The interlinkages among trade policy actions and their effects are all indications of the overwhelming need to avoid trade policy incoherences if the international economy is to benefit from the future growth of world trade. The main incoherence observed today in the framing of trade policies in North-South relations is undoubtedly that of developed market-economy countries perpetuating trade restrictive protectionist policies against developing countries’ exports, and that precisely at a time when these countries are trying to implement the outward-looking trade and development strategies very often advocated by the former. Trade

120

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

policy incoherence threatens to sink developing countries’ export prospects in a trade-restrictive vicious circle likely to block the adoption of development strategies. Moreover, incoherence can also reverberate negatively on developed countries themselves through slower structural adjustment processes, with the effect of a prolonged contraction of developing countries’ import demand being induced in turn in their economies. In the coming years, the expected strong pressures for adjustment in the patterns of production and trade will undoubtedly exacerbate the consequences of such incoherences, to the detriment of the world trading system as a whole, but with particularly negative consequences for developing countries’ future economic growth. THE EXTENT OF INCOHERENCE OF NORTH-SOUTH TRADE POLICIES AND THE PROSPECTS FOR EXPORTS OF MANUFACTURES AND FOR PROTECTION IN THE 1990s As a recent IMF report has stated, while tariffs in developed countries have been reduced to relatively low levels, non-tariff measures have proliferated and possibly offset the effects of postwar reduction in tariffs (IMF 1988:36). Sharp increases in voluntary export restraints and similar arrangements have been observed recently, in particular in sectors already subject to quantitative restrictions. In addition, today’s international trade continues to be distorted by domestic economic measures that protect domestic producers at the expense of their foreign competitors. Considerations of factors explaining the adoption of defensive trade policies have frequently suffered, in the past, from too simplistic generalizations conferring the whole burden of protection on isolated factors. It must be recalled, however, that the decision-making process for protection is contingent on a very complex set of determinants whose underlying characteristics have proved to be essentially political in nature, and which remain very often alien to economic criteria. As an illustration, trade has sometimes been singled out for the adoption of defensive trade policies by developed market-economy countries, on grounds of the ‘disruptive’ effect that surges in imports may cause to factor markets. Trade per se cannot be regarded as the source of any disruptive type of pressure for adjustment, since it is rather the underlying determinants of changes in comparative advantage (different factor endowments, different productivity changes, product cycles or preference diversity) that generate such pressures (Greenaway 1983:186). It is political considerations that, in many cases, overemphasize the real impact that import upsurges may cause to sectors experiencing rapid structural changes and where the cost of adjustment is high. Protectionism is, most of the time, a response to the loss of international competitiveness brought about by differential rates of growth of factor productivity. In the wealthy periods of economic growth of the 1950s and 1960s, such shifts in competitiveness created relatively few adjustment difficulties, but recession, starting in the early 1970s, generated structural unemployment and led increasingly to more pressures for protection in DMECs. Anyhow, whatever the reasons for protection may be, levels of import penetration are today, in DMECs, one of the most advocated factors for ‘justifying’ the decision to protect. These import upsurges are frequently determined, in turn, by exchange rate deviations from trend, which appear to have had a powerful effect on the demand for protection in both the United States and the EEC (Grilli 1988). The dynamic performance of manufactured exports has been one of the main components of growth for developing countries and their exporters of manufactures in particular in past decades. As is generally recognized, the highly advocated outward-looking strategy, based on a relatively open economy and an overall emphasis on the promotion of manufactured exports, is a development model that generally leads to economic efficiency. However, after the 1980s opened with the worst global recession since the 1930s, a recession causing serious doubts about future economic growth and the potential buoyancy of DMEC

PROTECTIONIST PRESSURES

121

markets for imports from developing countries, protectionist forces emerged, and revealed the high degree of incoherence still prevailing in trade policies that govern North-South exchanges of manufactured goods. The degree of trade policy incoherence is, in fact, the result of interactions among political considerations, macroeconomic conditions, the trade restrictive or trade liberalizing nature of the structural change and adjustment processes taking place in a given sector, and the trade-related pressures for protection or liberalization induced, in turn, by the magnitude and characteristics of trade expansion between trading partners. For the success of their development strategies, developing countries would need to expand greatly their exports in the coming years, mainly towards DMEC markets. The extent to which continued increases in the levels of import penetration (the share of imports in apparent consumption) are likely to be possible without inducing defensive trade policies, stifling, in turn, the growth of such exports, will determine the success or failure of such development models. As shown in Table 6.1, there is a general upward trend in the long-term evolution of the share of imports of manufactures from developing countries in the total consumption of the United States, the EEC or Japan. This trend became more pronounced in the recent 1985–8 period, particularly in the EEC and Japan, and essentially vis-à-vis imports from the fast-growing exporters of manufactures and the South and South-East Asian developing countries’ partners. Given the very complex nature of the determinants of protection in DMECs, and the rapid development of new forms of North-South trade in manufactures (North-South intra-industry specialization in production and trade), any projection on the expected pressures for protection must be approached with care. A look at the product composition of the value increments of developing countries’ manufacturing exports to DMEC markets in successive time-periods, shown in Table 6.2, confirms that such growth is, in the case of the United States, increasingly taking place in non-traditional products, while it is relatively more concentrated in traditional ones when the observed market for exports is either the EEC or Japan.2 The question that may be asked is if, for the DMECs taken as a whole, the pressures for protection against imports from developing countries in the EEC and Japan, during the 1985–8 period, might have been neutralized by trade liberalizing forces, which emerged from the type of product-mix exported by developing countries to the United States market? The similarity of the non-tariff measure (NTM) trade coverage ratio applied by DMECs, against developing countries’ manufacturing exports in 1988 to the one applied in 1985, may allow an affirmative answer.3 This trend may also indicate the need to develop further today’s new forms of North-South trade capable of granting, in the future, a certain standstill in non-tariff forms of protection by DMECs against imports from developing countries. Trade policy decisions are the result of a confrontation between pressures from industry and labour, on the one hand, and the disposition of the legislative and executive branch to confer protection on the other. Firms and labour provide the demand side of the market, while governing authorities provide its supply side (see Cline 1984:36). It is paradoxical to note that non-tariff protectionist responses to identical surges in import penetration vary widely, depending on the manufacturing sector concerned. The rapid technological progress, leading to forms of trade mainly driven by product innovation and differentiation, the internationalization of production processes and the globalization of markets are all factors that have engendered what may be called a ‘market dichotomy’, a division between a ‘dynamic’ area of international exchanges in manufactured goods and a ‘grey’ or ‘stagnant’ one. In the first area, North-South trade has already started to expand dynamically in a way similar to that in which intra-DMECs trade expanded in the 1960s and 1970s. This trade is increasingly characterized by vertical intra-industry specialization in the production and exchange of goods. On the other hand, in what may be called a ‘grey’ or ‘inter-industry’ or even ‘traditional comparative advantage’ area, a large portion of North-South trade, in the promotion of which comparative costs have a greater role to play, has continued

122

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

to perform in a sluggish manner. In this latter area, shifts in comparative advantage, affecting technologically mature industries, are taking place all the time, and adjustments to such shifts are difficult but inevitable. Wide cost differentials have threatened to displace high cost producers, leading increasingly to the widely accepted view that the expansion of one country inevitably results in contraction elsewhere. Thus, production in manufactured sectors in the ‘grey’ area tends to exceed trade and induces a surplus capacity, a factor which, together with the inadequacy and slowness of structural adjustment Table 6.1 Developing countries’ import penetration ratios1 of total manufactures in DMEC markets (percentages) United States 1975 World3 6.90 Developed market 4.92 economy countries Developing countries, 1.98 of which: Fast-growing 0.85 exporters of manufactures:4 Africa 0.09 Latin America 1.07 West Asia 0.04 South and South-East 0.77 Asia Oceania 0.01 European Economic Community5 World3 7.52 Developed market 5.50 economy countries Developing countries, 2.02 of which: Fast-growing 0.63 exporters of manufactures:4 Africa 0.50 Latin America 0.64 West Asia 0.11 South and South-East 0.72 Asia Oceania 0.05 Japan World 4.68 Developed market 3.04 economy countries

1980

1985

19882

9.12 6.21

11.43 8.07

13.12 8.40

2.91

3.36

4.72

1.50

2.31

3.08

0.15 1.31 0.02 1.42

0.13 1.18 0.04 2.01

0.28 1.68 0.04 2.72

0.01

0.00

0.00

10.09 7.10

12.37 8.73

13.23 9.03

2.99

3.64

4.20

1.01

1.27

1.78

0.66 0.89 0.22 1.18

0.81 0.95 0.43 1.41

0.87 0.92 0.44 1.95

0.04

0.04

0.01

5.43 3.42

5.04 3.27

5.84 3.18

PROTECTIONIST PRESSURES

123

United States 1975

1980

19882

1985

Developing countries, 1.64 2.01 1.77 2.66 of which: Fast-growing 0.68 0.85 0.95 1.41 exporters of manufactures: Africa 0.08 0.09 0.05 0.02 Latin America 0.29 0.31 0.25 0.21 West Asia 0.29 0.43 0.22 0.42 South and South-East 0.98 1.17 1.24 2.01 Asia Oceania 0.00 0.01 0.01 0.00 Source: Author’s calculations on the basis of OECD compatible trade and production data base (COMTAP) tapes. Note: 1 Import penetration ratios are defined as the ratio of imports from a particular source on apparent consumption (production+imports-exports), expressed in percentages. 2 Apparent consumption figures for the years 1986 to 1988 were estimated by using consumer expenditure indicators deflated by consumer price indexes from the IMF. World Economic Outlook, October 1988. Import figures for such years were derived from UNSO trade tapes. 3 The world is defined here strictly as the addition of developed market-economy countries and developing countries data. 4

Developing countries and territories that are fast-growing exporters of manufactures have been defined here restrictively as including: Republic of Korea, Singapore, Hong Kong, Mexico, Brazil and Argentina. 5 The European Economic Community indicators have, for reasons of data availability, been calculated on the basis of data for seven EEC countries: France, Federal Republic of Germany, United Kingdom, Belgium, Luxembourg, Italy and the Netherlands. 6 Excluding intra-EEC trade. Table 6.2 Changes in the structure of the value increment of DMECs’ imports of manufactures from developing countries 1973–80 Manufacturing sectors Total imports of manufactures: Increment in value (USSB.) Chemicals Textiles Clothing Iron and steel Other semi-manuf. Engineering prod. of which: Machinery for specific industries Office and telecom. equipment Road motor vehicles

1981–84

1985–87

All DMECs All DMECs All DMECs United States EEC %: %: %: %: %: %: %: %: %: %:

+49.1 100 7.3 6.7 21.2 3.5 9.4 34.2 2.2 10.0 1.2

+30.0 100 5.7 2.3 14.7 5.0 4.7 53.7 2.3 26.0 2.7

+64.3 100 3.3 5.3 19.4 1.9 9.5 42.5 2.8 12.0 6.5

+29.0 100 -0.1 2.8 14.9 -0.1 5.6 55.7 3.4 5.0 13.2

Japan

+22.7 +8.6 100 100 4.5 9.3 7.4 4.2 23.8 22.0 0.8 6.6 9.4 17.8 37.7 20.8 1.9 2.7 12.4 4.9 1.4 0.5

124

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

1973–80 Manufacturing sectors

1981–84

1985–87

All DMECs All DMECs All DMECs United States EEC

Japan

Other machinery and transport %: 11.8 16.3 10.7 13.3 1.8 5.5 equipment Household appliances %: 4.5 6.0 10.5 2.5 14.1 7.5 Other consumer goods %: 18.3 4.8 18.6 21.2 16.4 19.1 Source: GATT 1989 International Trade 1987–8, Volume II, Tables AB1, AB2, AB5, AB8 and AB9, Geneva: GATT.

processes in DMECs, has, in turn, favoured the blocking of trade and exacerbated trade conflicts in NorthSouth relations.4 What is increasingly evident is that intra-industry trade is becoming a large contributor to incipient North-South trade liberalization in certain industrial goods. While it is not our aim to draw a definite picture about what would be the state of non-tariff forms of protectionist pressures in the 1990s, an effort is made in the following pages to continue the research on today’s determinants of non-tariff forms of protection.5 The intention is to allow a first appraisal of the factors that are at the source of any acceleration or arrest of the 1990s protectionist threat. As illustrated by the ‘protectionist response surfaces’ shown in Figure 6.1 for the United States and the EEC respectively, the expansion of North-South mutual trade generates trade-related pressures for protection of a non-tariff nature; however, it can also be the source of pressures for trade liberalization, in particular if trade expansion is the result of an intra-industry specialization of production, leading to new forms of trade in manufactures among the developed and the developing countries concerned. It is increasingly evident that such new forms of trade are becoming a major contributor to incipient North-South trade liberalization forces emerging in certain industrial lines of production. These forms of specialization do not lead to pressures for protection, but rather they imply processes of product differentiation and product innovation, which are helping to alter the nature of modern trade without affecting the geographical location of industrial plants and thus avoid the exacerbation of the costs of structural adjustment in DMECs (see Falvey and Kierzkowski 1987). Moreover, in sectors where the expansion of North-South intra-industry trade is becoming important, transnational corporations (TNCs) are playing an active role and are reinforcing the emergence of new forms of industrial collaboration through, among other things, the development of international sub-contracting, joint ventures or technology agreements. In particular, the dynamic expansion of the off-shore assembly of goods illustrates how new forms of trade (intra-firm or intra-industry trade) promote such trade liberalization pressures. These forms of trade have been growing, in recent years, at rates considerably in excess of the growth of total manufacturing imports; this is particularly manifest in the case of United States trade in certain manufacturing products, with developing countries and territories such as Mexico, Taiwan (Province of China), Singapore, Hong Kong and Malaysia. Paradoxically, uncommon and illogical forms of protectionist response surfaces were found in the case of Japan, which are not presented here, but which reveal the very low degree to which NTM trade coverage ratios are an accurate description of the extent of protectionism against developing countries manufacturing exports in that market. They also illustrate the absence of a significant relation between the variables observed and the protectionist indicators used.6 A model of the determinants of protection (DOP model) has been developed in UNCTAD7 and was used here to examine probable incidence of protection in the 1990s and its magnitude vis-à-vis the manufacturing exports of developing countries. The equations of protection developed in the model showed, as expected, a right sign and significant results for the United States and the EEC. The more traditional character, already observed, of goods in exchanges between the EEC and the developing countries is also reflected in these

PROTECTIONIST PRESSURES

125

Figure 6.1 Non-tariff protectionist response surface1 against imports of manufactures from developing countries, as a function of trade and non-trade related factors in 1984 Source: Author’s Determinants of Protection model Note: North-South trade in manufactures gives rise to higher import penetration ratios (IPR) in the North but also, in some sectors, to greater dependence, by industries in the North, on southern markets. The import coverage on non-tariff measures (TCR) is relatively large in sectors with high import penetration ratios and low export dependence (XOP). In these sectors, inter-industry exchanges based on traditional comparative advantages are the predominant mode of international trade. Where high export dependence accompanies high import penetration, trade is mostly of an intraindustry nature and NTMs are relatively less significant. An important additional influence on TCRs is the relative dynamism of the sector. In mature, low growth sectors (CHP < O), structural adjustment processes are costly and, other things being equal, NTMs cover a relatively large proportion of imports. Conversely, in high growth sectors (CHP > O), structural adjustment is considerably easier and, consequently, TCRs tend to be lower. Of course, dynamic sectors also tend to be those in which intra-industry trade is important, while competition from imports is often strongest in slow growth sectors. Those interrelations are reflected in the protectionist response surfaces. 1Protectionist

response surfaces: the surfaces drawn are a representation in space of the influence of two variables on the ratio of trade covered by NTMs (in %) applied by the respective developed market economies in each of 81 manufacturing sectors. The form of each surface reflects the sensitiveness, across manufacturing sectors, of the trade coverage ratios (TCR) to import penetration (IPR) and export dependence (XOP) vis-à-vis developing countries, or to changes in the sectoral share of production over time (CHP).

equations of protection. The non-tariff protectionist response to upsurges in import penetration is, in the EEC, higher than in the United States, whose greater involvement in new forms of trade with developing countries makes import surges reverberate less on pressures for protection in that market. The model first gauged trade-related protectionist or liberalization pressures by extrapolating, under various assumptions,

126

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

the future growth of manufacturing imports and exports, from and to developing countries in each sector, as well as the future size of the respective DMEC markets. These projections are used to forecast, on the basis of equations of protection, the likely future trend of protection in each sector. In a second stage, and following the diagram shown in Figure 6.2, a rough classification of sectors in which exports to the United States and the EEC were significant was undertaken. For that purpose, trade-related pressures for protection or liberalization were compared with the inherent characteristics of the sectors, thus allowing a cluster exercise by allocating manufacturing sectors to four groups according to the real extent of non-tariff protection that developing countries may suffer in the two markets in the 1990s. A first look at clusters appearing in Tables 6A.1 and 6A.2 (see pp. 186 and 191) constructed on the basis of the diagram in Figure 6.2, suggests that not only the intensification, but the mere continuation and concentration of the present protectionist barriers in specific sectors are a serious matter of concern. Even if the 1990s bring DMECs closer to a tendency towards global standstill, high levels of protection are expected to remain in sectors already suffering from high barriers to trade. This will result in a harassment of traditional exports, which many developing countries, in particular those starting to develop their industrial bases, are concentrating

Source: Author’s Determinants of Protection model

Figure 6.2 Projeted protectionist pressures in manufacturing sectors of developed market-economy countries as a result of the expansion of trade with developing countries

PROTECTIONIST PRESSURES 127

128

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

upon. It is precisely the divergent degree of specialization in production and exports prevailing among developing countries at different stages of industrialization that is expected to amplify, in the 1990s, the present dichotomy, a dichotomy that exists between those countries exporting to sectors expected to suffer from and to surrender to trade or politically related pressures for protection in DMECs, and those that are not. If developing countries’ export expansion takes place, in the coming years, in the manufacturing sectors appearing in group I of Tables 6A.1 and 6A.2 such expansion will most probably induce, through the expected increase of import penetration ratios, trade-related pressures for protection, which will be exacerbated, in turn, by prevailing sectoral characteristics of a trade defensive nature.8 Higher protectionist tensions, or at least standstill, would be the prospect for those sectors in the coming years, with arguments for protection expected to remain as strong as they are today. Among such sectors, it is worth mentioning those in which current developing countries’ exports are the highest in the group. It is the case, for example, of already protected sectors, such as manufacturing of wearing apparel (ISIC 3220), manufacture of furniture (ISIC 3320), fabrication of metal products (ISIC 3819), manufactured production of leather (ISIC 3233) and canning and preserving fruits and vegetables (ISIC 3113) in the United States. In the EEC market, and in addition to the already-mentioned wearing apparel, canning and preserving fruits and vegetables, grain mill products (ISIC 3116), manufacture of vegetable and animal oils and fats (ISIC 3115) and spinning, weaving and finishing of textiles (ISIC 3211) are noteworthy. Moreover, in currently freetrade sectors, which are of importance to developing countries, pressures for protection may effectively induce the appearance of new barriers to trade. This could be, in particular, the case in sectors such as manufactures of footwear (ISIC 3240) in the United States, or manufacturing industries n.e.c. (ISIC 3909) in the EEC. It should be noted, finally, that eventual actions to revert protectionism in this group of sectors may be difficult to implement in the future, given the relatively high cost of the structural adjustment process in those sectors, compared with that prevailing in sectors of the other groups. However, in sectors appearing in group III, and despite upsurges of import penetration ratios and thus of trade-related pressures for protection projected for the 1990s, current levels of trade restrictiveness may even be lower in some cases, or at least remain as they are, given the ‘pro-trade’ characteristics prevailing in those sectors that are likely to reduce political forces in favour of protection. Among the main sectors that may benefit from such lower protection or standstill it is worth mentioning radio, TV and telecommunications equipment (ISIC 3832), and sugar factories and refineries (ISIC 3118) in the EEC; machinery and equipment n.e.c. (ISIC 3829) in both the United States and EEC markets, and the basic industrial chemicals (ISIC 3511), shipbuilding and repairing (ISIC 3841) and synthetic resins and plastic materials (ISIC 3513) in the United States market. In the case of some important free-trade sectors, which are expected to suffer trade-related pressures for protection, these pressures may be resisted and protection is not expected to occur, given the ‘pro-trade’ characteristics of these sectors (new forms of trade being developed, lower costs of adjustment, etc.). This is the case for sectors such as plastic products n.e.c. (ISIC 3560), electrical apparatus and supplies (ISIC 3839), electrical industrial machinery and apparatus (ISIC 3831) and jewellery and related articles (ISIC 3901) in the United States, and basic industrial chemicals (ISIC 3511), office computing and accounting machinery (ISIC 3825), and aircraft (ISIC 3845) in the EEC. All sectors in group III can be considered as promising sectors for the export expansion of developing countries, since efforts towards export diversification will match emerging new forms of North-South intra-industry specialization in production and trade in those sectors, all factors which encourage freer trade, despite expected import penetration upsurges in the 1990s. DMECs’ actions to impede the resurgence of protectionism or at least grant standstill in all these sectors will require more than just maintaining the present pace of structural adjustment. Research into possibilities of continuing to expand new forms of specialization in production

PROTECTIONIST PRESSURES

129

and trade with developing countries is also needed. In doing so, DMECs will favour forms of trade which, instead of provoking the complete relocation of industries in more competitive producing countries, would, on the contrary, induce complementary types of specialization in the same industry segments, a factor that would also reduce structural adjustment costs. North-South future trade prospects in manufactures are also contingent on the existence of trade liberalization forces, that is, on a trade expansion that is not expected to lead to greater pressures for protection in the 1990s, either because it consists mainly of sectors concentrated in new forms of exchanges with developing countries, or because the already prevailing protectionist levels would hinder any future upsurges of DMECs’ imports from them. This would be the case of sectors appearing in groups II and IV. However, the ‘anti-trade’ characteristics still prevailing in some of these sectors (group II) (slow and costly structural adjustment processes, persistence of traditional forms of exchange in goods, low export dependence, or the high political importance of the sector (higher than average employment)) may reveal the existence of a ‘political muscle’ for protection. Liberalization pressures being resisted, standstill or higher protection would be the estimated outcome for the 1990s in sectors of group II. It would be the case, in particular, for important sectors such as iron and steel basic industries (ISIC 3710), and slaughtering, preparing, and preserving meat (ISIC 3111), in the two markets analysed, and for spinning, weaving and finishing of textiles (ISIC 3211), and sugar factories and refineries (ISIC 3118) in the United States, and for sawmills, planing and other woodmills (ISIC 3311) in the EEC. In currently free-trade sectors appearing in group II, and despite expected trade-related liberalization forces, the nature and characteristics of sectors may also represent a future threat in favour of protectionism. This may be the case in important sectors such as grain mill products (ISIC 3116) and manufacture of vegetable and animal oils and fats (ISIC 3115) in the United States, as well as manufacture of furniture and fixtures (ISIC 3320) in the EEC. It is important to recall that any action for accelerating the pace of structural adjustment, if directed primarily at sectors in group II, will benefit developing countries’ trade expansion more rapidly, trade-related pressures for liberalization making trade-related arguments for protection groundless. Finally, if future developing countries’ export expansion in manufactures is concentrated in sectors of group IV, it would have a good chance of being granted standstill or even lower levels of protection in the 1990s, given the anti-protectionist influences of trade liberalization pressures mixed with the pro-trade characteristics of the sectors concerned. This will be the case in currently low protected sectors, such as radio, TV and telecommunications equipment (ISIC 3832) in the United States, and non-ferrous metals (ISIC 3720), motor vehicles (ISIC 3834) and plastic products n.e.c. (ISIC 3560) in the EEC. Moreover, trade is expected to remain free of protection in important sectors of export interest to developing countries, such as office, computing and accounting machinery (ISIC 3825), motor vehicles (ISIC 3843) and nonferrous metal basic industries (ISIC 3720) in the United States, and tanneries and leather finishing (ISIC 3231) and manufactures of pulp, paper and paperboard (ISIC 3411) in the EEC. Forms of specialization and exchange in sectors of group IV appear to allow them to avoid trade restrictive barriers of a non-tariff nature in the coming years, and, simultaneously, make them privileged candidates on which developing countries can concentrate their export diversification efforts. The extent to which developing countries’ exports are presently concentrated in each of the abovementioned groups can give a rough indication of the protectionist harassment they may encounter when increasing their exports to these markets in the 1990s. Thus, the distressing question is which developing countries will really benefit from the eventual export expansion possibilities of the 1990s. Information presented in Table 6.3 illustrates, four different developing country groups, the share of their current exports belonging to sectors classified in groups I+II (higher-protection or standstill) and III+IV (lower protection or standstill) of Tables 6A.1 and 6A.2 respectively. Given the more diversified nature of their exports, fast-

130

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

growing exporters of manufactures are, vis-à-vis other developing countries, relatively more concentrated in sectors with a lower probability of protection for the 1990s in the United States market. In contrast, in the EEC the distinction is almost non-existent, a fact that emphasizes the great potential for the development of new forms of North-South trade between the EEC and developing countries, particularly the more advanced among them, in the coming years.9 Proof that the more developing countries acquire a production structure convergent with that of the DMECs, the more they become involved in new forms of intra-industry specialization and trade, receiving the correct demand signals from DMECs, and the more they are able to avoid future DMECs trade policy incoherences, is also illustrated by the results presented in Table 6.3. As observed, not only the level of developing countries’ per capita income, but also that of the importance of the manufacturing industry in total GDP, is related to the extent of non-tariff protectionist response that these countries will face in expanding their manufacturing exports to DMECs in the 1990s. Even if high-income developing countries are today highly affected by protectionism, the increasing industrial convergence of their production structure with that of the DMEC countries, as well as the dynamic character of their exports, may grant them a relatively larger, freer entry into those markets in the 1990s, than that available to low-income countries with less convergent production structures, whose future penetration in such markets will be based on traditional comparative advantages and will be more concentrated in ‘sensitive’ (high adjustment costs, grey area of crisis) sectors. The pace of structural adjustment by DMECs for such low income countries may be the main determinant of their 1990s access

Source: Author’s calculations on the basis of the UNCTAD Determinants of Protection Model (DOP). Trade figures are from OECD compatible trade and production data base tapes for 1985. Non-tariff measures’ trade coverage ratios used come from the UNCTAD data base on non-tariff measures. Note: 1 Manufactures are here defined as the 81 categories at 4 digit of (ISIC 3), excluding petroleum refineries (ISIC 3530). 2 Refer to Figure 6.2 and Tables 6A. 1 and 6A.2 for a complete explanation of sector groups I, II, III and IV classified attending to the likely outcome of ntm’s protection in the 1990s. 3 Share of imports from developing countries included in sectors that in 1985 faced at ntm’s trade coverage ratio higher than 5 per cent. For ratios lower than 5 per cent the sector is considered as free. 4 Developing countries were classified in country groups according to GNP per capita indicators for 1986. The developing countries included in each group: lowincome (GNP/cap. 1,810$) are those included in World Bank, ‘World Development Report, 1988’, Table 1, p. 222, Washington DC: World Bank. 5 Developing countries were classified in two country groups according to their respective share of manufacturing production in total GDP in 1986. See World Bank, ‘World Development Report, 1988’, Table 3, p. 226, Washington DC: World Bank.

1

significant ntm’s forms of trade restriction or under free trade, and prospects for protection in the 1990s

Table 6.3 Concentration of DMECs’ imports of manufactures from developing countries classified by income levels, in sectors suffering

PROTECTIONIST PRESSURES 131

132

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

to DMEC markets. These facts are also illustrated in Table 6.4. In the United States market, the share of imports from Africa under protection is even expected to increase in the 1990s, while it is expected to experience nearly no change in the case of imports from Latin America. Conversely, imports from South and South-East Asia would seem to be able to avoid to a greater extent non-tariff forms of protection in the United States market in the 1990s. In the case of the EEC market, and given the more traditional character of its trade with the various developing regions, the phenomenon described is less marked and the restrictiveness of the market is not expected to change significantly vis-à-vis the various developing regions examined. CONCLUSIONS It is increasingly clear today, that while the evolution of the nature and complexity of new forms of NorthSouth trade pervades the international trading system, the way that trade policymakers tackle the issue of protectionism and trade-related aspects of structural adjustment seems often to remain alien to such new realities. At the dawning of the 1990s, the virtues of embarking on outward-looking development strategies, in order to maximize the benefits derived from greater comparative advantages in certain lines of products, or even in whole industrial sectors, are being preached to many developing countries. However, even if the philosophy surrounding such recommendations translated into concrete, successful cases of export expansion in manufactures during the 1960s and 1970s, the increasingly apparent incoherences in the shaping of trade policies by DMECs, at the beginning of the 1980s, are starting to raise serious doubts about the degree of homogeneity that may prevail among developing countries in their potential export success in DMECs’ markets in the next decade. The progressively incoherent character of DMECs’ protectionist mechanisms, mainly of a non-tariff nature, based in sectors in which many developing countries have concentrated their export supply capabilities, has started to create a marked dichotomy between these countries. On the one hand, some developing countries were able to develop, in the 1960s and 1970s, their incipient industrial bases, making use of the relatively lower degree of incoherences encountered in implementing their trade-based strategies. Those countries, mainly known today as the ‘fast-growing exporters of manufactures’, are already enjoying a relatively highly diversified industrial

Source: Author’s calculations on the basis of the UNCTAD Determinants of Protection Model (DOP). Trade figures are from OECD compatible trade and production data base tapes for 1985. Non-tariff measures’ trade coverage ratios used come from the UNCTAD data base on non-tariff measures. Note: 1 Manufactures are here defined as the 81 categories at 4 digit of (ISIC 3), excluding petroleum refineries (ISIC 3530). 2 Refer to Figure 6.2 and Tables 6A. 1 and 6A.2 for a complete explanation of sector groups I, II, III and IV classified attending to the likely outcome of ntm’s protection in the 1990s. 3 Share of imports from developing countries included in sectors that in 1985 faced an ntm’s trade coverage ratio higher than 5 per cent. For ratios lower than 5 per cent the sector is considered as free.

Table 6.4 Concentration of DMECs’ imports of manufactures1 from developing countries classified by regions, in sectors suffering significant ntm’s forms of trade restriction or under free trade, and prospects for protection in the 1990s

PROTECTIONIST PRESSURES 133

134

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

infrastructure, and are increasingly able to avoid harmful forms of trade restriction in DMEC markets, and to develop new forms of exchange with them, precisely through a less protectionist type of trade expansion. On the other hand, there is a large majority of developing countries with fragile industrial bases; in the case of these countries, the recent exacerbation of trade policy incoherences, reflected mainly in the emergence of the new protectionism in the 1980s, is threatening to frustrate the implementation of their present or future outward-looking development strategies. If the present situation continues, it is the 1990s development prospects of the latter countries that will be endangered, precisely at a time when they will need to reinforce their export outlets to counter the overwhelming debt-related disequilibrium in balance of payments that many of them may continue to suffer. Projections have shown that, for developing countries taken as a whole, no widespread surge in non-tariff forms of protection in DMECs is expected to arise vis-à-vis their manufacturing exports in the next decade. It may even be stated that these countries are closer today to the often requested standstill in protection that is generally recognized. However, even if standstill is the global projected path for protection in industrial exchanges taken as a whole, it may just be the illustration of a perpetuation of present incoherences. In other words, given the evolving forms in which North-South trade takes place, the chances of DMECs’ trade liberalization being open to some developing countries may be minimal, while the exports and development prospects of most of them will largely depend on the way DMECs render their trade policies less patently incoherent with developing countries’ ones. The need to avoid such incoherence becomes particularly acute when the dynamism that those countries have the potential to induce in world trade growth in the 1990s is looked at. It is evident that in certain ‘sensitive’ sectors, the future growth of exports from developing countries will run the risk of generating a strong protective response in DMEC markets. It must also be recalled that the implementation of outward-looking development strategies in developing countries will contribute to the mutually beneficial expansion of world trade. In the presence of such realities, DMECs should, instead of perpetuating the incoherent character of their present trade policies, undertake the necessary effective structural adjustment processes, by moving factors to more productive uses, try to achieve a greater international co-ordination of macro-economic policies, including trade policies, and engage in industrial specialization processes through product enhancement and innovation (see Hechstrate and Zepperwick 1988). There is no doubt that if such policies are implemented as DMECs become more coherent actors in the international trading system, DMECs will be more able to simultaneously ‘accept’ imports of traditional (labour-intensive, traditional comparative advantage, inter-industry) types of goods from developing countries at their initial stages of industrialization, to seek advantages by offering highly technological type of goods in the ‘dynamic area’ of international trade, and to develop new forms of exchange with the most dynamic exporters of manufactures among developing countries. The recently observed 1985–8 expansion of world trade, and the particularly dynamic role some developing countries are playing, illustrates the considerable scope open to these countries, in the 1990s, for a continued expansion of their exports at relatively rapid rates. The dynamic expansion of the evolving new forms of North-South trade is bringing to light the large potential DMECs have for absorbing increasingly diversified types of manufactured exports, in particular from the most ‘dynamic’ developing countries’ exporters. It is clear that, even if greater efforts for diversification would probably mean for this group of dynamic exporters some sacrifices and adjustments in traditional exporting sectors still enjoying comparative advantages in international markets, the expected standstill protectionist period, projected in DMEC markets for the 1990s, implies that more diversified export profiles will also provide a guarantee and greater security of continued access to those markets. A coherent and correct trade policy choice

PROTECTIONIST PRESSURES

135

(liberalization of trade regimes and the development of export incentives, among others) would, thus, be also of utmost importance to ensure this kind of market access. Finally, the impressive 16 per cent growth, in volume, of developing countries’ imports of manufactures in 1988 (see Gatt 1989) also emphasizes the important role developing countries’ import capacity can play in continuing to make future world trade flows more dynamic. That growth is to some degree a reflection of the extent to which today’s new forms of North-South trade exchange, relatively alien to comparative advantages in their traditional forms (intra-industry trade), can contribute to a certain avoidance of traderelated pressures for protection, despite future expansion of these countries’ exports, and to a certain relaxation of trade tensions in the international trading system. APPENDIX 6.1 THE DETERMINANTS OF PROTECTION MODEL (DOP) The determinants of protection model (DOP) is being developed in the UNCTAD secretariat. Since quantitative analysis of the decision to protect must, in order to be realistic, address the presence of nontariff barriers, the model uses protection data from the UNCTAD data base on non-tariff measures, either through the trade coverage ratio or the frequency ratio approach, to measure the trade restrictiveness in a given sector. The model considers only manufacturing sectors that are defined at the 4-digit International Standard Industrial Classification (ISIC) level of disaggregation, in order to be able to use NTM data simultaneously with the compatible production and trade data base of the OECD (COMTAP). In addition, UNIDO and UNSO data sources are used to complete the statistical data of the model. The DOP model uses equations of a cross-section ordinary least squares estimation nature. In the derived equations of protection for each DMEC market against imports from developing countries, the dependent variable of the extent of non-tariff protection is expressed as a function of demand and supply of protection variables. On the demand side: import penetration, export dependence or intra-industry trade, comparative advantage (labour productivity and changes in labour productivity). On the supply side: the political importance (share of manufacturing labour force), adjustment costs (changes in the share of employment, changes in the share of production and changes in import penetration) and others (share of production) are included. The selection of the most explicative and significant variables in each equation of protection is made through maximum Rsquared improvement of dependent variables technique. The United States and the EEC alone can be considered as sufficiently representative of the extent of DMECs protectionism since they represented nearly 80 per cent of the 1985–7 total DMECs’ increase in imports of manufactures from developing countries. Cross-section regression equations run for the model for 1984 showed the right sign and significance for the coefficients on import penetration (IPR:(+)), export dependence (XOP:(−)), structural adjustment costs (changes in the sectoral share in total production (CHP: (−)), and changes in import penetration over time (CIP:(−)).

Table 6A.1 United States: projected protectionist pressures in manufacturing sectors as a result of the expansion of trade with developing countries

136 FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

PROTECTIONIST PRESSURES

137

138

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

PROTECTIONIST PRESSURES

139

140

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

PROTECTIONIST PRESSURES

141

Source: Author’s Determinants of Protection Model (DOP) on the basis of the UNCTAD data base on non-tariff measures and of OECD compatible production and trade data (COMTAP) tapes and UNIDO production and employment data. Note: a See Figure 6.2 for a description of the rules according to which sectors were allocated to the four different clusters (I, II, III and IV). b The trade-related pressures for protection index (TRPPI) is an indicator of the extent to which the NTM trade coverage ratio in a given sector may be pushed upwards or downwards in the 1990–5 period by the mixed influences of projected growth of import penetration ratios of developing countries in the DMEC market, and of export dependence of DMECs on developing countries markets. The index is the ratio between 1995 NTM trade coverage ratio (TCR 95) projected by equations of protection of the UNCTAD DOP Model and the corresponding 1990 projection of that ratio: (TCR95/TCR90). c A manufacturing sector is defined as experiencing easier (+)/difficult and costly (−) processes of structural adjustment, if in the 1975–85 period the sectoral share it represented in total manufacturing production increased/decreased during that period, thus expressing the dynamic or new/ declining or mature character of the sector in question. d A: A manufacturing sector is said to be of an (intra-industry trade character (+)/inter-industry trade character (–)) if the intra-industry trade ratio prevailing in trade of the DMEC with developing countries was (higher/lower) than 50 per cent, or if the 1975–85 (increase/decrease) in that ratio would be (sufficient/ insufficient) to make the IIT ratio projected for the next ten years attain at least 50 per cent. B: A manufacturing sector is said to be potentially (benefited/not benefited) by the eventual resumption of developing countries’ import demand if the share of DMEC exports to developing countries in production was, in 1985, (higher/lower) than the average share recorded for all manufacturing sectors.

Table 6A.2 European Economic Community: projected protectionist pressures in manufacturing sectors as a result of t he expansion of trade with developing countries

142 FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

PROTECTIONIST PRESSURES

143

144

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

Source: Author’s Determinants of Protection Model (DOP) on the basis of the UNCTAD data base on non-tariff measures and of OECD compatible production and trade data (COMTAP) tapes and UNIDO production and employment data. Note: a See Figure 6.2 for a description of the rules according to which sectors were allocated to the four different clusters (I, II, III and IV). b The trade-related pressures for protection index (TRPPI) is an indicator of the extent to which the NTM trade coverage ratio in a given sector may be pushed upwards or downwards in the 1990–5 period by the mixed influences of projected growth of import penetration ratios of developing countries in the DMEC market, and of export dependence of DMECs on developing countries markets. The index is the ratio between 1995 NTM trade coverage ratio (TCR 95) projected by equations of protection of the UNCTAD DOP Model and the corresponding 1990 projection of that ratio: (TCR95/TCR90). c A manufacturing sector is defined as experiencing easier (+)/difficult and costly (–) processes of structural adjustment, if in the 1975–85 period the sectoral share it represented in total manufacturing production increased/decreased during that period, thus expressing the dynamic or new/ declining or mature character of the sector in question. d A: A manufacturing sector is said to be of an (intra-industry trade character (+)/inter-industry trade character (–)) if the intra-industry trade ratio prevailing in trade of the DMEC with developing countries was (higher/lower) than 50 per cent, or if the 1975–85 (increase/decrease) in that ratio would be (sufficient/ insufficient) to make the IIT ratio projected for the next ten years attain at least 50 per cent. B: A manufacturing sector is said to be potentially (benefited/not benefited) by the eventual resumption of developing countries’ import demand if the share of DMEC exports to developing countries in production was, in 1985, (higher/lower) than the average share recorded for all manufacturing sectors.

PROTECTIONIST PRESSURES 145

146

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

The regression equation for the United States was:

with t values for all variables significant at the 95%(**) and 99%(***) level. F value for the whole regression significant at the 99% level and R2=0.20. For the EEC, the corresponding equation was:

with t values for all variables significant from 90%(*) to 95%(**) and 99%(***). F value for the whole regression significant at the 99% level, and R2=0.21. Since the 1975–80 and 1980–5 periods reflected respectively, higher and lower than average growth of trade of the United States and the EEC with developing countries, the average 1975–85 growth period for production, imports and exports was taken in order to extrapolate 1990s imports, exports and size of the market variables to be introduced in the model. NOTES The author (who is Economist at the International Trade Programmes of the United Nations Conference on Trade and Development (UNCTAD), Geneva) would like to state that the views expressed in this chapter are personal and do not necessarily reflect those of the UNCTAD secretariat. He is grateful to Cato Adrian for his assistance in dealing with the data on non-tariff measures and their compatibility with other sources based on different nomenclatures, to Pilar Palazon for her statistical assistance and to Manuel Agosin for his valuable comments and suggestions. 1 The recent evolutions observed in international trade in manufactures increasingly show that intra-industry has been the more dynamic type of trade in North-South exchanges. Structural convergence among industrial and technological bases among trading partners is an explanation of such dynamic trends. See Tussie (1987). 2 As much as 56 per cent of the additional $US 30 billion increment of developing countries’ exports of manufactures to the United States in the 1985–7 period consisted of engineering goods (with as much as 13 per cent concentrated in motor vehicles components and parts) and only 17.6 per cent in textiles, clothing and iron and steel goods together. In contrast, the opposite situation is observed for the EEC and Japan, with around 40 per cent of the additional $US 30 billion (approx.) of developing countries’ exports of manufactures to the EEC and Japanese markets, together consisting of textiles, clothing and iron and steel, while only 33 per cent consists of engineering goods. See Table 6.2. 3 In fact, the average NTM trade coverage ratio affecting imports of total manufactures from developing countries in developed market economy countries’ markets remained at 15.9 per cent from 1985 to 1988. See UNCTAD (1988). 4 In the European Community, for example, the steel crisis of 1977 led to an explosion of anti-dumping cases initiated against steel imports, which continued in 1978 after the ‘state of crisis’ was declared by the Commission and the Davignon Plan launched. Similarly, in the case of the United States, the protectionist fever of the years 1977–9 originated largely in the crisis in the steel industry. (See Grilli 1988:322). 5 For a first approach to this subject, see de Castro, (1989). 6 Particular features of the Japanese economy, which may constitute forms of intangible protection, such as limitations in the distribution system, consumer preferences, overall higher level of comparative advantage in the production of goods, etc., have been advanced to explain past low levels of imports in Japan. For complementary

PROTECTIONIST PRESSURES

147

comments on the difficult appraisal of the real extent of protection in Japan, see Lawrence (1987), as well as Takeuchi (1988). 7 See a short description of the determinants of protection model (DOP) in Appendix 6.1 (p. 185) and preliminary results of the model presented in de Castro (1989). 8 Those characteristics are, mainly, high costs of structural adjustment, prevailing overall traditional forms of exchange with developing countries, a relatively low beneficial impact to be expected from the eventual resumption of developing countries’ import demand from the respective DMECs in that sector, as well as the importance of employment and its changes over time. They have been taken into account to decide the pro-trade or anti-trade nature prevailing on the demand for protection side, in those sectors, since they were all significant variables found in developing the equations of protection of the model. 9 In the United States, the share of fast-growing exporters of manufactures (FGE) in the total imports from developing countries, in sectors corresponding to projected standstill or lower protection groups (III and IV) is much higher than that recorded by other developing countries. FEGs countries recorded shares of 92 per cent in machinery and equipment (ISIC 3829), 87 per cent in plastic products n.e.c. (ISIC 3560), 76 per cent in radio, TV and telecommunications equipment (ISIC 3823) or 94 per cent in office, computing and accounting machinery and appliances (ISIC 3825), all main sectors in the two groups. In contrast, for the EEC, FGE countries recorded only 62 per cent of DMEC imports from developing countries of radio, TV, telecommunications equipment (ISIC 3832), 23 per cent of basic industrial chemicals (ISIC 3511), 5 per cent of non-ferrous metal industries (ISIC 3720) and 17 per cent of tanneries and leather finishing (ISIC 3231), all sectors of utmost importance for developing countries’ exports in groups III and IV towards that market.

REFERENCES Cline, W.R. (1984) ‘Exports of manufactures from developing countries: performance and prospects for market access’, Washington DC: The Brookings Institution. De Castro, J.A. (1989) ‘Determinants of protection and evolving forms of North-South trade’, UNCTAD discussion paper No. 26, June 1989, Geneva: UNCTAD. Falvey, R.E. and Kierzkowski, H. (1987) ‘Product quality intra-industry trade and imperfect competition’, in Protection and Competition in International Trade, essays in honor of W.H.Corden, New York: Blackwell. GATT (1989) Newsletter No. 60, March–April 1989. Greenaway, D. (1983) International Trade Policy: from Tariffs to the New Protectionism, London: Macmillam Grilli, E. (1988) ‘Macro-economic determinants of trade protection’, The World Economy 11(3) (September 1988). Hechstrate, H.J. and Zepperwick, R. (1988) ‘Distortions in world trade: recent development’, Intereconomics, November–December 1988. International Monetary Fund (1988) World Economic Outlook, October 1988, Washington DC: IMF. Lawrence, R.Z. (1987) ‘Import markets in Japan: closed markets or minds?’ Washington DC: The Brookings Institution. Takeuchi, K. (1988) ‘Japan’s market potential for manufactured imports from developing countries: a survey of the literature’ (draft), Washington DC: IECII, The World Bank (November 1988). Tussie, D. (1987) The Less Developed Countries and the World Trading System, London: Francis Printer. UN Conference on Trade and Development (1988) ‘Protectionism and structural adjustment’, TD/B/1196. Report by the UNCTAD secretariat, December 1988, Geneva: UNCTAD.

7 Inter-African integration and protection policies: unavoidable failure or missed opportunities? Jean Coussy

Attempts at integration in Sub-Saharan Africa (SSA) are generally considered as failures, although the explanation of these failures varies among authors. For some, failure was unavoidable because integration focused on protection, while others attribute it to specific circumstances, namely inefficient management of tariff protection. This chapter attempts to identify reasons why foreign trade policies failed to promote effective African integration. It resorts to historical perspectives, which stress the discrepancy between the coherence of theories and the empirical policies actually pursued. Thus, we might avoid the pitfall of embarking again on the debates of the 1960s, while omitting the lessons of the last thirty years. We shall follow the history of the three models of integration either successively or simultaneously applied in SSA: integration through tariff protection of infant industry, integration based on the protection of existing activities, and integration through trade liberalization. INTEGRATION THROUGH COMMON TARIFF PROTECTION OF INFANT INDUSTRIES The projects of African integration through common tariff protection of infant industries Initial conceptions of African integration During the 1960s, African integration was considered as a coalition against other regions or countries, and aimed at organizing complementarities on a national level. This conception resulted from a rejection of inter-African competition, which, at the international level, was perceived as generating losses (in terms of deterioration of terms of trade and creation of surplus capacities) and weakening bargaining power (against foreign purchasers of exported commodities and fund suppliers). At the domestic level, it was thought to increase production costs through duplication of investments. Consequently, the prevailing view was that integration should not result from customs unions replicating developed countries but through ex ante coordination of investment (Bourguinat 1968). This initial conception is closely linked to infant industries theories. Inter-African integration would simultaneously act as an educative protectionism and as a means to reduce costs by avoiding the creation of firms if prospective markets appeared too small. On a regional scale, the project was to achieve economies of scale, to induce learning by doing and external economies, to stimulate induced activities and reciprocal linkages effects between protected industries and induced activities. The objective would be to decrease

INTER-AFRICAN INTEGRATION

149

reliance on imports from the Rest of the World (RoW) and, in the long term, dependence on traditional exports, thanks to diversification of production. The ultimate target of integration, according to this theory, was then directly opposed to a liberal integration into the world economy. In this conception, African states were first to create complementarities through inter-state agreements fostering an inter-African division of labour (avoiding duplications) and to establish inter-African commercial preferential agreements guaranteeing outlet for production. This would have implied the elimination of obstacles to inter-African trade and the erection of obstacles against non-African competitors. However, as trade discrimination generates trade-diversion costs, at least in the short term, costs of protection ought to have been calculated beforehand, and arrangements devised for their distribution among participants. This would have implied the creation of an inter-African indemnity system avoiding injustices and/or breaches of agreements. Main objectives of integration through educative protectionism This simplified infant industry approach resulted, especially between 1960 and 1980 (and at a faster rhythm between 1970 and 1980), in protection of three kinds of activities: (a) Exploitation and first transformation of mineral and oil resources (iron and steel industries, oil refineries, petrochemical industries, fertilizer factories, paper-pulp plants, irrigation and hydro-electric dams, etc.) which are activities with large optimal dimension. African integration was thus expected to reduce the costs of protection of infant industries. But, since regional markets did not prove large enough and since agreement on the distribution of costs was seldom reached, these costs had generally to be met by the country where the activity was located. With high costs of investment and long recovery periods, protection stricto sensu had to be supplemented with subsidies, capital endowments, loans and preferential rates and other forms of incitement for foreign capital (fiscal exemptions, free provision of infrastructure, guarantees concerning profit remittances abroad, customs duties’ exemptions on inputs, etc.) which are nowadays heavily criticized because of their distorting effects. (b) Import-substituting policies focused on downstream consumer-goods industries close to markets (textile factories, car assembly lines, cement production, breweries and other agro-industrial activities, etc.). With a relatively small optimal dimension, they could sometimes survive on national or sub-regional markets, such as those constituted by institutional African unions. However, the distribution of the burden of protection was complicated by the cascading of protective tariffs, which increased costs to consumers, hence calling for compensation of consumer countries. (c) Protection was sometimes jointly planned for some short-process industries, which extended from first transformation to production of consumer goods (essentially agro-industrial firms: rice, sugar, some wheat, cassava, and textiles). But few complete processes were organized on a regional scale, because this would have entailed a complex network of subsidies (upstream) and tariff protections (downstream). The causes of failure of integration based on infant industry arguments Many of the firms launched in SSA during the 1960–80 period under the shelter of educative protectionism are now facing difficulties: under-utilization of productive capacities, deficits, lack of self-financing capacities, growing need for subsidies and protections, and threats of bankruptcy (which sometimes materialized). Competitiveness—which is the ultimate justification for educative protectionism— was generally not achieved.

150

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

Still, we should not yield to the prevalent antiprotectionist or afropessimist mood, and should bear in mind that some of these firms’ difficulties resulted from recessive macroeconomic conditions, with domestic markets shrinking, capacities to import inputs and capital goods falling, foreign funds drying-out and budget deficits precluding continuation of subsidies. The breakdown of the dynamics of growth often blurs the rationale for protection. We should recall, however, that the experience of large countries in other continents (Brazil and India, in particular) shows that positive effects of educational protectionism require more than a couple of decades to materialize. However, educative protectionism in SSA met with problems even before the economic crisis: there was very little investment coordination, inter-African commercial preference schemes were not respected, multinational firms underwent crises, distribution of costs of protection between participants caused disputes and effective payments to joint-ventures were often delayed. These failures were amplified by underestimation of the difficulties that integrated protection policies would have to face. Consequently, policy orientation shifted abruptly, and periods of optimistic project-planning were followed by massive close-downs in the face of high, unexpected and unplanned costs. One additional cause for failure was the belief that the positive effects, which List had only presented as a possibility, would automatically come about. All this fuelled an unwarranted optimistic view on educative protectionism, and led to incoherent policy timing and an increase in the number of objectives assigned to protectionist policies. The fragility due to previousness of borders was equally overlooked. The optimism of expectations concerning the efficiency of educative protectionism Initial optimism was probably the first cause for disappointment with educative protectionism. Expectations regarding scale economies were particular high, especially when we consider the total size of African markets (which add up to no more than Belgium’s GNP) and transport costs (which preclude any such summing-up). Optimal dimensions for heavy industries (iron and steel industry, petrochemistry) exceeded absorption capacities of the sole African market. Lighter industries (textile, agro-industry) could have reached their optimal size in African integration zones, but their development was often attempted on a national scale, and generally required high protection over long periods—although interesting possibilities had been uncovered by Pearson and Ingram (1980). Confusion between dynamic (which can be expected when markets grow) and static economies of scale (which correspond to present size of markets) was frequent, and transition costs arising when markets progressively expand were neglected. The scale argument was certainly responsible for a good deal of over-sizing and under-utilization of capacity, which resulted in production costs higher than those of foreign competitors. Expectations regarding training effects were equally over-optimistic, time-spans underestimated and automaticity too easily assumed. Learning by doing in SSA has not resulted in a decrease of costs: the employment of expatriate executives has remained high, pushing local management costs upward. Expectations regarding external economies were also disappointed. Exploitation and first transformation of domestic resources did not induce strong forward linkages, and even discouraged them by imposing negative effective protection on downstream activities. Downstrean activities, on the other hand, often required cascading tariffs, which discouraged backward linkages and domestic production of inputs, even in cases of local product processing (e.g. milk). Finally, mutual linkage effects between protected sectors and activities induced by these protected sectors were also victims of the automaticity delusion. In particular, too much reliance was put upon the creation of flows (between town and country or neighbouring countries), while the influence of prices (which created

INTER-AFRICAN INTEGRATION

151

conflicts) was neglected. When price-effects eventually depressed the terms of trade of agriculture and of suppliers of unprotected products, production incentives deteriorated and the conflicts on the distribution of costs of integration resumed. Timing problems of educative protectionism policy The coordinated protection of infant industries within an integrated multinational entity calls for a precise time-schedule, which was rarely respected in the integration and protection attempts in SSA during the 1960–80 period: (a) Premature creation of new activities often resulted from errors in forecasts on the growth rhythm of regional markets and the delays necessary for ensuring the procurement of machinery and inputs. (b) Delays in the execution of projects further extended the periods of deficit, protection and subsidy. These technical or administrative delays resulted in an increase of protection costs, a slow-down of the integration process and, sometimes, in attitudes rejecting integration. (c) Delays in the adaptation of tariff protection (most notably cascading tariffs) to the structural changes induced by economic growth hindered backward production linkages, which developed at a much slower pace in SSA than on other continents. (d) Instability of national protections is particularly high in African countries, owing to the high instability of domestic economies and of fiscal requirements in terms of tariff revenues, frequent political changes, and the general instability of inter-African relations. Such a lack of stability impedes successful implementation of educative protectionism or integration. (e) Protection and integration programmes were frequently interrupted following economic crisis and structural adjustments. This prevented them from lasting long enough to demonstrate positive results. The costs of protection were even considered as unnecessary losses, an attitude that played an important part in present-day negative assessments of this protection. The multiplicity of objectives assigned to protection measures Failures of protection measures to bring about African integration equally arose from the multiplicity of objectives that they were assigned. Most frequently, national objectives superseded regional ones: the common external tariff (against imports from Rest of the World) was only exceptionally put into practice and—where and when it existed—was often distorted by national fiscal policies (as in the case of the Union Douanière des Etats d’Afrique Central-UDEAC). Tariff and non-tariff obstacles (explicit or implicit, official or informal) violated the principle of free movement of goods within African regions, which not only slowed down the integration process but considerably increased the costs (and chances of failure) of protection, by reducing the economies of scale and multiplying the cases of under-utilization of productive capacities. This was aggravated during the 1970s, when the increase of mineral and oil revenues and the development of international indebtedness allowed each individual country to follow an autonomous investment policy (sometimes encouraged by private investors or foreign aid). As a consequence, cases of investment duplications increased in oil refineries, fertilizer factories, paper-pulp plants, and agro-industrial activities. Short-term economic shifts exerted a particularly strong influence on national policies: countries registering sudden increases in export earnings suddenly loosened their restrictions on imports from the RoW; domestic inflationary pressures resulted in abrupt bans on food exports to neighbouring African countries; countries facing a public finance deficit adapted their trade policies in order to maximize tariff

152

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

revenues, etc. These swings destabilized the protective system and slowed down the harmonization of national policies on foreign trade. The creation of rents was a frequent objective of trade policies, especially in the case of quantitative restrictions (import quotas granted without public auction). This sometimes led to decisions that directly opposed the needs of regional educative protectionism and introduced a permanent uncertainty which influenced negatively inter-African trade and investment. The regional boundaries’ perviousness to informal exchanges Lastly, the efficiency of educative and integration protectionism was hampered in many ways by ‘informal integration’. Legally protected (but in practice unprotected) firms had to face a contraction of markets. Recently, textile industries were affected by the import of second-hand clothes, rice producers were faced with the competition of illegal imports, etc. The permeability of frontiers is often mentioned as a definitive and inescapable obstacle to the implementation of any protective system in SSA, and is presented as an argument in support of devaluation. However, the discussions on this point are far from being closed (Vallée 1988). The disorganization, through informal integration, of planned integration results from attempts to create ‘rents’ by resorting to temporary measures, which are known to increase profits resulting from informal trade. Consequently, trade policies were voluntarily destabilized and made unpredictable, once again to the detriment of educative protectionism’s objectives. The penetration of African markets by products from the RoW, which is one of the major effects of informal trade, unbalances the integration projects by favouring consumer against producer countries, relatively high income countries against countries with an important production potential, and coastal economies against landlocked ones. In the end, ‘informal integration’ turns inter-African relations into an extension of relations with the RoW, which is the antithesis of the initial project of planned integration in a coalition perspective. INTEGRATION THROUGH PROTECTION OF EXISTING ACTIVITIES The project of African integration through educative protectionism that dominated the 1960s and 1970s exerted a strong influence on the development pattern in SSA. It resulted in taxation of export-oriented agriculture, substitution of imports for domestic food crops (due to exchange rate overvaluation, resulting from protectionist measures), reallocation of resources to less productive activities, decline of investments’ efficiency, external indebtedness, etc. The impact of educative protectionism—or rather of this form of protectionism wrapped up in educative arguments—generated a number of criticisms in the 1980s, when SSA faced financial crisis, food shortages and urban overdevelopment. One option then proposed was to resort not to protection of industry and emerging activities, but to a commonly planned protection of agriculture and other existing activities. Projects of integration The new protection projects Adapting protection to complementarities between African countries (especially between coastal and landlocked countries) was not a new idea. But it took the startling revelation of import-substitution failure to

INTER-AFRICAN INTEGRATION

153

promote patterns of protection which centred upon agriculture and resource-based industries (agro-industry and cottage industry). Ambitions to create new activities yielded to hopes of developing ‘autochthonous origin’ industries, of launching triangular operations (subsidizing and protecting inter-African trade with foreign aid support) and creating protected agricultural areas (such as the project of a Western African privileged cereal crop zone in Mindelo). These projects go beyond ‘conservative protectionism’, because they equally attempt to develop, and not just conserve, existing activities, inducing, when feasible, increases in productivity. The difference with the previous years’ educative protectionism is that they intend to build upon existing productive and social structures, and to avoid sudden shifts away from traditional specializations. The rationale for protection of existing activities The protection of existing activities, whose legitimacy, contrary to that of educative protectionism, was always questioned, has been recently accused of being no more than a manoeuvre aimed at saving senile industries. Without reopening the theoretical debate (see Weiller 1946), we shall simply note that, in SSA, this form of protection seeks to avoid the voluntarist adoption of industries designed in foreign economic systems, and thus to minimize consequent disturbances, such as destruction of rural and cottage industries, geographical destabilizations of activities, inter-African migrations and discontinuities in the development model. In other words, this form of protection could be justified by adequate economic calculation. Contrary to immediate liberalization or to protection of new activities, it seeks (not always successfully) to avoid the costs resulting from the regression of existing activities, accelerated obsolescence of specific machinery (especially agricultural equipment), degradation of the rural environment and dequalification of labour. These costs may be very high, as was recently shown by the experience of oil-producing countries which faced ‘Dutch disease’ and later had to bear the costs of reviving agricultural activities when oil revenues plummeted. Since regression costs are not accounted for in private calculations, it is legitimate—even within a strictly neo-classical perspective—for the state to intervene and compel decision-makers to take such regression costs into account, whether these are temporary (induced by the instability of comparative advantages) or permanent (created by bets on new comparative advantages). Furthermore, historical experience confirms that the balance of costs and benefits arising from such protection schemes did not necessarily impinge on growth prospects of countries that adopted them for lengthy periods. In some cases, the preservation of existing activities exerted favourable effects on infant industries (creation of secure and stable domestic outlets) and avoided the costs of disarticulation of productive structures and social destructuration. Thus ‘complex development’ models came into existence, with a sometimes low but regular growth rate. Even countries who paradoxically protected simultaneously segments of their industry and of their agriculture experienced such a development: the French economy in the long-run (Weiller 1946), newly industrialized Asian countries— who started under such ‘complex’ models—or some African countries, termed ‘conservative’, such as Cameroon in the 1960s (Coussy 1988). Difficulties met by projects of integration through protection of existing activities Although the projects of integration through protection of existing activities aroused interest among some donors (especially from the Sahel Club, the EEC, and bilateral foreign aid), objections, opposition and obstacles were strong enough to slow down implementation. Even the project of a Western African privileged cereal crop zone seems threatened before any policy discussion has begun (Giri 1989).

154

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

Technical obstacles and theoretical objections Transport costs, which hampered the integration of newly established sectors producing high scale economies, are a major obstacle to exchanges in agro-food industry, according to Delgado’s calculations (1988). Transport costs of cereals between producer and consumer countries (especially road rather than rail transport) widen considerably the producer/consumer price differential. Coastal countries are then tempted to take advantage of cheap sea transport and low world prices of cereals, meat and dairy products (which are subsidized by the EEC and the USA). The relative evolution of productivities, not only between developed countries and SSA but also between Asia and Africa, equally works against integration. One of the major risks of protecting existing activities when the rates of growth of productivity are so different is that integration will have to rely on increasingly inefficient activities, thus increasing the costs of African cooperation for consumer countries. Fixed rates of protection can encourage imports from the RoW when world prices drop temporarily, that is, in circumstances where national producers are most threatened by international competition. Flexible tariff schemes have been contemplated, but they meet resistance from finance ministries because they entail instability of customs duties’ revenues. In countries that import consumer goods, the irreversibility of consumption patterns can seriously hinder the effectiveness of tariff protection on food products. Such irreversibility has been clearly demonstrated in the case of Burkina-Faso (Delgado, Reardon and Thombiabo 1988), although reversibility of dietary habits has been observed in other countries, under different economic conditions (in the case of the Ivory Coast, see Requier-Desjardins 1989; Courade, Goi and Harre 1988). The risks of non-coordination are many: overly sectorial protection (as in projects where protection would be restricted to cereals) is one risk, while the opposite danger results from a possible adding-up of projects protecting existing activities with projects based on educative protectionism. This would end in complete non-selectivity of protection. Sociological and political resistance Technical obstacles to integration through protection of traditional structures were compounded by sociological and political oppositions, mainly: (a) The negative attitudes of consumers (especially in urban areas), who will bear the brunt of the costs of protection. Urban consumers have on many occasions successfully opposed high price policies entailed by integration. The recent Senegalese decision not to increase the price of rice, thus prohibiting all possible imports from Mali, is an example. (b) In contrast to the consumers’ negative stand, the resistance of taxpayers and Treasuries is usually neither direct nor conscious. But refusal to increase domestic taxation perpetuates the fiscal reliance on external trade taxation. Fiscal yields then depend on the volume of trade with non-African countries. Senegalese and Ivorian Treasuries, for instance, have a marked interest in maintaining a high volume of purchases from outside SSA. (c) In spite of official declarations, the lack of interest on the part of national decision-makers for integration projects favouring traditional activities is quite patent, especially when compared with the interest previously displayed for educative protectionism projects. We should note that this behaviour is in opposition to that of international donors, who tend to distrust import-substitution attempts but respond positively to the notion of agricultural protection. (d) Groups interested in trade with the Rest of the World equally oppose all projects that would reduce such trade: consumers, merchants, transporters, speculators, public administrators and civil servants often

INTER-AFRICAN INTEGRATION

155

have an interest, even in producer countries (and, a fortiori, in consumer countries), in preserving formal or informal imports from and exports to non-African countries. (e) Specialization strategies in informal transit operations appeared in certain coastal countries (Gambia, Benin), who have consciously adopted strategies facilitating imports from the RoW and ‘informal’ reexports to neighbouring African countries. These strategies dismantled Africa’s external protections and have in particular reduced the autonomy of neighbouring countries’ international trade policy (see the example of Senegal, in Egg, Igue and Coste 1988). (f) Finally, as in all continents, resistances to the harmonization of agro-industrial policies also stem from the political nature of agro-industrial strategies that concern land-holding systems, income distribution pattern, cost of living, inter-regional equilibrium, etc. INTEGRATION THROUGH THE LIBERALIZATION OF TRADE AND OF THE MOVEMENT OF FACTORS OF PRODUCTION Structural Adjustment Programmes (SAPs) dealt a decisive blow to integration projects based on common external protection. New projects emerged, which, by conviction or lip-service, were labelled ‘African integration’, but really turned out to be projects of liberalization of international movements of goods and factors of production. Arguments against integration through common external protection Theoretical criticisms The integration through external protection faces a number of criticisms. In the liberal line of thought, which opposes all commercial preferences, attempts at African integration were criticized, following J.Viner, as suboptimal arrangements that induce trade-diversion and generate negative static effects and anti-export biases. Although orthodox theory accepts the infant industry argument, its application to SSA was criticized because of the insufficient size of African domestic markets. Exports on the world market would be required to allow scale economies. Dynamic arguments in favour of educative protectionism were not even discussed. While previously these had been grossly exaggerated, prospects for influencing comparative advantages are nowadays completely forgotten. The liberal approach has constantly rejected the notion of protection of existing activities, seen as a conservative measure protecting declining industries. This argument was applied to the protection of noncompetitive agricultural activities. However, the debate between liberals and protectionists could never be settled. Among unresolved empirical questions was that of the existence of alternative employ ment opportunities for workers in sectors threatened by liberalization, or estimation of costs of regression. Disagreement also touched on conceptual matters, such as norms in relation to which costs and benefits of social destructuration should be appreciated. While protectionists expressed fears regarding the human costs of migrations, labour dequalification and so on, liberal optimism stressed the advantages of increased integration into the market. An important critical line stressed the incoherence of protection, and especially the dispersion of effective rates of protection (Greenaway and Milner 1986). The assessment of protective systems revealed the complexity and disorder of these rates of protection (with negative effective protection in priority sectors, lack of coordination between segments of an industry, disincentives to backward or forward linkages, etc.).

156

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

One outcome of these assessments was to show that the objectives of protection, especially concerning African integration, were very seldom respected. Critiques, however, sometimes provided approximate and inappropriate arguments. Static criticisms were voiced against measures intended to induce dynamic objectives—for instance, measures intended to modify comparative advantages were judged in terms of static Domestic Resource Cost calculations—while provisional situations were criticized on dynamic efficiency arguments (for example, cascading tariffs criticized for not promoting domestic production of inputs). Similarly, assessments of protectionist policies were based upon the value that indicators (Domestic Resource Costs, effective rates of protection, equilibrium rates of exchange) had at the time of assessment, that is, when countries were in such difficulties that they had to resort to SAP. This stressed the present inappropriateness of policy, but did not warrant the retrospective critique of past policies which, unfortunately, has often occurred. Thus, the assessment of past policy-courses taken in a different economic environment was distorted by deliberate or involuntary anachronisms. Such anachronisms are especially dangerous because they attribute to errors of judgement what is really the result of an interruption of growth, largely explained by exogenous factors. Strategic objectives of protection also came under heavy criticism. Stated objectives were reinterpreted and demystified: the accumulation objective shown not to conceal a motive of implicit-taxation of unprotected sectors, the industrialization objective to favour the creation of bribe-generating state-owned enterprises, and advocation of protection could cover up inefficient management and redistributive motives. Finally, the numerous obstacles to African integration were shown to draw policies away from officially stated integration objectives. This was actually the starting-point for new integration projects: by removing such obstacles, liberalization would constitute the basis for an alternative integration scheme through increased commodity trade and mobility of factors of production. Dismantling existing protections One major prerequisite for integration through market mechanisms is the suppression of obstacles to trade that had gradually accumulated over the past decades. The market would then be relied upon to create interAfrican complementarities and stimulate inter-African competition. Liberalization of factors of production (labour and capital) is simultaneously advocated. In this view, quantitative restrictions should be abolished and protection brought about by rate-ofexchange policy alone. When this proves unfeasible (as in the CFA zone) or if, for fiscal reasons, tariffs cannot be reduced, taxation should tend toward uniformity in order to minimize distortions. Devaluation can also be approximated through the adoption of uniform rates of export subsidies and import taxes. Within such a non-selective framework, one would have to disregard long-term objectives, which justified protection of infant industries, and regression and human costs, taken into account in protection of existing activity schemes. In this scheme non-discrimination complements non-selectivity: liberalization of relations with the RoW and of inter-African relations will proceed apace. African integration is not a coalition against the RoW any more, but one way of ensuring Africa’s integration into the world market. This philosophy of integration is then, at least in its extreme liberal version, in complete opposition to the former projects. Although this shift in objectives is covered up in diplomatic language, the central notion is that of competition and not coalition any longer. African countries become competitors in their home-markets and abroad, both as export-goods suppliers and as demanders of capital. African economies are not to be insulated through common protection schemes, but compelled to compete on the world market.

INTER-AFRICAN INTEGRATION

157

Uncertainties regarding the role of protection in SAPs Initial formulations of SAPs were openly hostile to integration through common protection. After ten years of structural adjustment programmes, we still cannot assess their impact on integration. First of all, SAPs, which were elaborated and discussed at a national level, contained no specific set of measures on integration (Boidin 1988). And, second, the challenge to protection policies proceeded with delays, twists and turns, diversions by governments and progressive adaptation to resistance of men and events. Reduction and rationalization of protection under Structural Adjustment Following critical assessments of effective protections, all countries undergoing SAP reduced and rationalized protection according to liberal recommendations. On the import side, quantitative restrictions were dismantled or reduced, tax and subsidy schemes were simplified, dispersion of effective protections were narrowed, highest rates of protection were edged-off, and sometimes cascading tariffs were reduced. On the export side, export taxation was reduced, negative protections were withdrawn, public monopsonies were abolished and real rates of exchange were lowered (or, in CFA zones, export subsidies were established). Domestic markets were equally reformed, and subsidies (to consumer goods, agricultural inputs and state-owned enterprises) were curtailed. Although the prime motive for these reforms was not to further African integration, the reforms did influence integration by removing some obstacles to inter-African trade. Prospects for ‘informal integration’—in so far as they resulted from disparities between policies—were weakened, ‘informal flows’ (when resulting from comparative advantages) were turned into formal ones, de facto harmonization of policies allowed removal of distortions, and common external protection and coalition-enhancing measures were suppressed. Persistence of previous protection and integration measures Differences in timing of reforms, however, did leave some previous protection measures untouched and prevented actual harmonization of national policies. Quantitative restrictions, which were to be removed first, were sometimes maintained or reintroduced either during balance of payments emergencies or in order to reinstall protection (e.g. prohibition of cereals imports in Nigeria, or rice imports in Madagascar). Reduction of nominal rates of protection conflicted with fiscal revenue motives, which called for maintaining high tax-rates (duties on rice imports in Senegal and Côte d’Ivoire both yield tax revenue and allow, through equalization operations, to support the domestic price of rice). And even though rates of effective protection have been rationalized and simplified on paper, a detailed study of customs system reform in Zaire (Siggel 1986) showed results to be much more ambiguous than expected. Although the World Bank constantly insisted upon the reduction of export taxes, national authorities often took advantage of devaluations to maintain or even increase them, as Madagascar did on coffee. Substitution of rate of exchange for tax and subsidy protection could not apply in the CFA zone, where assortments of taxes and subsidies, which are bound to prove selective and discriminating, were implemented. Administrative and informal obstacles to African integration, rooted in past practices, still survive, and transport costs have increased rather than decreased, owing to fuel price-increases, demanded in SAPs, and road deterioration dating far back, before SAPs.

158

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

Significantly, informal integration has survived, thriving on maintained obstacles to trade, policy disparities, and pessimistic anticipations on devaluating countries (expectations of vicious circles of devaluation and inflation) (Geronimi 1989). Gradual recognition of the needs for integration and protection by SAPs On integration and protection, as on most other points, Structural Adjustment programmes at first aimed at introducing a strict rationality, breaking with previous policy experiences. Gradually, however, risks and dangers of policy discontinuity were recognized, and some former objectives and methods were reintroduced within SAP frameworks. Contributions of domestic policies to African integration—which are now taken into account when assessing adjustment experiences— were an example. Possible negative impacts of domestic policy reforms upon the workings of African multinational enterprises were first considered. Then, conscious regional harmonization of national SAPs were sought from inception. Finally, regional SAPs were contemplated (as in the case of UDEAC). Furthermore, and equally important, exceptions and amendments to principles of liberalization were admitted when they contradicted ongoing integration moves. Such was the case in the CFA zone, where some specific state interventions were maintained. Such was equally the case for a few joint enterprises, such as Air Afrique or the Sahelian Dams. An even stronger departure from initial principles was accepted when the principle of a common external tariff was admitted in discussions on UDEAC (even though such an admission might eventually turn out to be a mere lip-service). Such exceptions to initial adjustment principles might well be only temporarily accepted by international organizations who still aim at bringing African countries to face international competition. The presence of ulterior motives on issues such as the future of the CFA zone or the phasing-out of common external tariffs is probably more than a simple possibility. The rhetorical use of the language of integration in liberal documents could well be a formal concession to be followed by further legitimations of the opening-up of African economies to world trade. The lack of convergence of national policies One of the paradoxes in the evolution of national trade policies is that, in spite of the stated ambition that all countries should adopt a similar model, differences between policies not only survive but, on occasion, widen. Some differences are inherited from past policies. Countries who previously invested in exploitation and first transformation of natural resources (e.g. the Nigerian steel industry) will attempt to keep these running. This will probably spark conflict with international organizations, the outcome of which will remain uncertain for a while and will depend on respective negotiating power. The outcome will prove even more uncertain when it touches on import-substitution industries: some countries will chose to close down formerly protected activities, some will maintain protection on downstream activities alone (e.g. assembly-lines), while others will attempt to protect upstream industries (Nigeria, for instance, choosing to produce locally inputs for its agro-industry). Differences between national policies can also arise from different abilities to resort to comparative advantages. Coastal countries are, more than ever before, tempted to reject common protection of

INTER-AFRICAN INTEGRATION

159

agriculture, which is increasingly desired by landlocked countries. Oppositions on rice policy, for instance, are very strong between Senegal and Côte d’Ivoire, on one side, and Mali, on the other. Four elements at least account for the perpetuation of such differences between policies. First, countries embarked on adjustment at different periods of time, and early adjusters might presently be reaping comparative advantages. Second, various countries are not hit by the same elements of a manifold crisis. For instance, some countries will ban cereal imports to alleviate foreign currency payments, others will impose import taxes for fiscal reasons, and others will liberalize imports to curb inflation. Third, liberalization has many degrees, and countries might attempt to outbid one another through labour legislation and investment code reforms. Such is the case with the multiplication of free trade areas (Mauritius, Togo) which exceed SAPs’ recommendations on liberalization. Finally, as noted earlier, some countries have specialized in informal transit (Gambia, Benin) and derive their profits by outbidding their neighbours (Senegal and Nigeria) in terms of opening-up. Risks of conflicts arising from inter-African competition Inter-African competition, stimulated by liberal-inspired structural adjustments, could go further than the breaking-up of interstate coalitions and end in open confrontations between policies. These will result at best in ‘benign neglect’ regarding consequences to neighbours’ situations and at worst in a zero-sum game, where every single amelioration is bought at the neighbours’ expense. The crisis of commodity agreements has already sharpened. Foreign debt negotiations have been conducted without any concertation. Competition for foreign capital induces ever more favourable conditions for foreign investors, as in Ghana (Leenhardt and L’Hériteau 1988). The search for nontraditional exports could equally result in renewed competition on RoW markets, as is already the case between Nigerian and neighbouring countries’ textile industries. The acuteness of competition on the African market itself, depressed as it is by prolonged crisis and adjustment policies, might be exacerbated by the fact that it does not come to grips with private enterprises freed from public intervention, but involves new forms of state interventions (through open or disguised subsidies, concurrent devaluations, political pressures on wage costs, etc.). Such conflicts might assume extreme forms unless some compensation schemes for losers are built into regional SAPs. CONCLUSION Three coherent doctrines arose in the debate on African integation, but since none of them was implemented coherently, it is almost impossible to decide whether failures of integration are unavoidable, or result from missed opportunities or adverse circumstances. It is, a fortiori, even more difficult to make predictions regarding the future utilization of protection strategies. Past incoherences Examination of past trade policy reveals a story of incoherence and mismatch between objectives and decisions. The structure of rates of protection displays a number of inconsistencies and aberrations, and demonstrates a general failure to comply with policy objectives. Whether we believe, along liberal lines, in the congenital irrationality of state interventions, or whether we retain hope of possible improvements of the

160

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

state’s capacities, previous experience demonstrates how difficult it has been for African states to control their protection systems. These incoherences do not result from instrumental difficulties alone, but also from the multiplicity of objectives implicitly assigned to trade policies. The feeling of irrationality and disorder in the rates of protection is somewhat tempered when we consider that the integration and protection objectives were not the only, nor perhaps the main objectives, of trade policy. These policies have equally, and perhaps most noticeably, followed motives that were short-term, financial, circumstantial, social and so on. Which means that, contrary to hopes—expressed in SAPs—of achieving minimal protection through uniform rates, multiplicity of objectives might well introduce a good deal of complexity into trade policies. Scrutiny of trade-policy objectives equally reveals the existence of unstated, or even hidden, motivations, such as nationalistic drives, active state participation in ‘informal integration’ or creation of rents ‘a la’ Krueger in favour of administrative clientèles. We cannot say whether such phenomena, and the resulting obstacles to African integration, are permanent or transitory. On that point, the only thing the economist can do is to avoid falling back on the kind of naïvete that provides normative economics recommendations—which untiringly rehearse the principles of optimal integration-oriented trade policies, and overlook thirty years of ambiguous behaviour by African states. And also to avoid the converse form of naïvete, which believes that denunciation of policies that hamper integration will, in itself, correct them. The debate on the future of protectionist interventions in African integration This debate revolves around two extreme positions. On the liberal side, SAPs do not see, or hope for, any future for such policies that would necessarily produce the same effects as in the past. On the opposite side, those in favour of state-monitored integration think that protectionist instruments have not so far been given their right place, that a more systematic protectionism could achieve what has failed in the past and actually represents the only hope to achieve integration. Recent experiences do not provide answers to this debate. However, three remarks can be made: If the aim of protectionism really was integration, then everybody will agree that it was misapplied. This does not mean that protectionism is unfit for promoting integration, but shows that expectations have been overoptimistic, and over-optimism in itself, as seen earlier, was a cause for failure. Furthermore, many beneficial effects expected from protection are almost out of reach in the case of SSA, due to the exiguity of markets, high transport costs, high costs of qualified labour and perviousness of borders. Now, although these factors underscore the need for prudence when assessing the efficiency of protectionist instruments in the construction of African integration, it does not follow that their failure should be considered inevitable. First of all, many criticisms (disregard of comparative advantages, lack of upstream or downstream effects, overdimensioning of investments, etc.) tie together two different things: badly planned policies, and premature interruption of policies, abandoned before they could yield positive results. We should also refrain from systematically blaming state intervention for bad results. Many of these results do in fact stem from insufficient insulation, which deprives states of effective intervention instruments against adverse external developments, such as foreign actors’ roles in indebtedness, fluctuations of prices of primary commodities and instability of comparative advantages. Finally, whatever the real mistakes and failures of protectionist interventions, it seems fairly improbable that these should disappear. Some forms of intervention will reappear sooner or later, for reasons accepted by all. It seems completely unlikely that African states would be the only ones not to try and modify their comparative advantages through external protection and integration. And we cannot imagine that social

INTER-AFRICAN INTEGRATION

161

consequences of economic destructuration and delocalization arising from integration through the market will not prompt interventions from African states or from donors. Other interventions will equally occur, based on more disputable grounds, such as properly political coalitions and discrimination against the RoW, financial objectives, emergency interventions following crises, containment of fraud and smuggling, resort to state power in inter-African competition, etc. This reemergence of interventionist tendencies has in fact been observed during the first years of implementation of the most liberal SAPs. African countries, as all other countries, resort to empirical mixes between liberal and protectionist measures, heavily influenced by constraints specific to each country (Hibou 1990), and between various forms of protection. It is quite uncertain, however, whether these compromises will enhance African integration or not. REFERENCES Boidin, J. (1988) ‘La cooperation régionale a l’épreuve de l’ajustement structurel’, Courrier ACP-CEE112. Bourguinat, H. (1968) Les marches communs des pays en développement, Geneva: Droz. Courade, G., Goi, I. and Harre, D. (1988) Evaluation des habitudes de consommation des produits alimentaires en Côte d’Ivoire, Paris: Document ORSTOM. Coussy, J. (1988) ‘Le conservatisme de la politique économique du Cameroun depuis l’indépendance’, in P.Geschiere et P.Koenings (eds) Colloque sur l’économie politique du Cameroun, Leiden Research Reports No. 35, Leiden University. Delgado, C.L. (1988) Questions a propos d’un espace regional protégé pour les céréales, Washington DC: IFPRI. Delgado, C.L., Reardon, T. and Thombiabo, T. (1988) ‘La substitution des céréales locales importées—La consommation alimentaire des ménages a Ouagadougou’, Ouagadougou: Documents Cedres. Egg, J., Igue, J. and Coste, J. (1988) ‘Echanges régionaux, commerce frontalier et sécurité alimentaire en Afrique de l’Ouest: Méthodologie et premiers résultats’, Document INRA-UNB-IRAM. Geronimi, V. (1989) ‘Les échanges parallèles’, Working paper, Paris: CERED/LAREA. Giri, J. (1989) ‘L’espace regional céréalier: myths ou réalité de demain’, in Club du SahelEspaces céréaliers régionaux, Paris: OECD. Greenaway, D. and Milner, S.R. (1986) ‘Estimating the shifting of protection across sectors: an application to Mauritius’, Industry and Development 16. Hibou, B. (1990) ‘Analyse comparée des politiques de protection et de regulation des marches céréaliers ouestafricains’, Paris: Document INRAUNB-IRAM. Leenhardt, B. and L’Hériteau, M.F. (1988) ‘Situation macro-économique et monétaire au Ghana’, Paris: CCCE. Pearson, S.C. and Ingram, W.D. (1980) ‘Economies of scale, domestic divergences and potential gains from economic integration in Ghana and the Ivory Coast’, Journal of Political Economy 88(51). Requier-Desjardins, D. (1989) L’alimentation en Afrique, Paris: KarthalaPusaf. Siggel, E. (1986) ‘Protection, distortions and investment incentives in Zaire: a quantitative analysis’, Journal of Development Economics 22. Vallée, O. (1988) ‘La dimension monétaire des échanges entre les pays du Golge du Bénon (Nigeria, Benin, Togo)’, Paris: Document INRA-UNBIRAM. Weiller, J. (1946)Problèmes d’Economie Internationale, Vol. 1, Paris: PUF.

8 Import liberalization and the new industrial policy in Senegal Anne-Marie Geourjon

Trade policy occupies an important place in the Structural Adjustment reform programmes adopted by LDCs during the 1980s. It has an important impact on fiscal revenues, especially in African countries,1 and opens up possibilities of influencing the supply side through improvements in the industrial incentives system. Studies undertaken prior to the adoption of reform programmes assessed the protection policies adopted in the 1960s and 1970s. They all reached the same conclusion, namely that over-protection of the domestic market led to major distortions which were incompatible with the necessary reallocation of productive resources in favour of tradeable goods. Consequently, the main objective of industrial policy reforms has been to reduce the protection of the domestic market and to reinforce export promotion. Among the measures contemplated, those aiming at import liberalization are the most difficult to put in practice, because they endanger established and well-rooted interests. This chapter, based upon the experience of Senegal, illustrates the problems met by import liberalization within the framework of a global reform of industrial policy. Before proceeding, it will be necessary to recall the situation of the industrial sector in Senegal on the eve of these reforms, and to specify the salient features of the new industrial policy (NIP) adopted in 1986. THE WEAKNESSES OF THE SENEGALESE INDUSTRY IN THE EARLY 1980s, AND THE ADOPTION OF THE NIP Senegalese industry developed before that of other African countries, because it was meant to supply industrial goods to the former AOF (Afrique Occidentale Française) whose capital was Dakar. Following Independence, industrial firms suffered from a shrinking of the market owing to new national boundaries and competition from newly established industries in neighbouring countries, especially in Côte d’Ivoire. Trade and industrial policies adopted since 1960 aimed at reinforcing import-substitution (essentially on consumer goods) by offering local firms a high degree of protection. The situation in the early 1980s Total entry charges2 on imports as specified in customs tariffs have always been fairly high. Moreover, towards the end of the 1970s, this legal tariff protection showed a tendency to increase. However, if we consider the evolution of the rate of collection of import duties and taxes (levied on CIF value), we find it has dropped sharply, falling from 28 per cent in 1970–2 (arithmetical average of the three years) to 22 per cent in 1979–81 and 16 per cent in 1983–5 (Geourjon 1988). This discrepancy between legal tariffs and the taxes actually collected results from the proliferation of exemptions granted during this period by the government for various reasons (Investment Code, suspensive customs regulations, ‘exceptional’

IMPORT LIBERALIZATION IN SENEGAL

163

exemptions, etc.). In 1984, while the average legal rate stood at around 86 per cent, imports were in effect taxed at less than 20 per cent. In that year, the loss of revenues to the Treasury totalled 148 per cent of total revenues from customs. Given such implementation conditions, tariff policy contributed to increase distortions among industrial activities. In order to compensate for these distortions, like many other LDCs (Krueger 1978) the Government of Senegal, under pressure from local producers, increasingly resorted to quantitative restrictions.3 Thus Senegal’s trade policy evolved into a complex and highly incoherent system during the early 1980s. Moreover, the generalized practice of exemptions and this high level of total entry charges was an encouragement to fraud. In order to assess the impact of the protection system as it stood in 1986, the rates of effective protection for each industrial sector were calculated. The results obtained (given in Table 8.1) highlight the importance of distortions generated by the system, and especially the over-protection granted in certain cases: the highest rate amounts to 8.25,4 which means considerable economic and financial advantages, while the lowest stands at −34.44. Activities whose effective protection is negative are either those for which international value-added is negative—which amounts to a loss of foreign currency for the national economy —or those who are penalized by the cost of local inputs (energy) or by advantages granted to other sectors. For instance, biscuit and confectionery firms purchase sugar from the Compagnie Sucrière Sénégalaise (a firm benefiting from a special convention—most advantageous clause of the Investment Code) at 313 CFA francs per kilo, which is three times the world price. The evolution of Senegalese industry has been far from satisfactory. While the Government expected this sector to contribute significantly to economic development (growth, employment, etc.), and although it was relatively larger and more diversified than in other African countries in the 1960s, the annual growth rate of the industrial sector declined in the 1970s. Between 1970 and 1982, the manufacturing industries’ average annual growth rate was 0.8 per cent. Compared with that of other African countries (Côte d’Ivoire: 5.4 per cent, Cameroon: 8.4 per cent, Kenya: 9 per cent), this was a very poor performance, even taking into account the fact that the industrial sector in Senegal had been established for a relatively long time. It is worth completing this rapid assessment of the overall situation of the industrial sector with a few remarks concerning tradeable goods. Protection was highest on import-substituting activities, especially those producing consumer-goods, while nothing in the incentive structure favoured the use (or production) of local intermediate goods. Sheltered from competition for far too long, these over-equipped importsubstituting firms often produced poor-quality goods at high prices. These were spurned by the local consumer, who preferred imported or smuggled goods. On the whole, production was well below productive capacity. Excessive protection contributed to curbing this sector’s development, which, in any case, would have been handicapped in the long run by the small size of the domestic market. The exporting sector met with difficulties from the late 1970s onwards. Exports other than groundnuts fell by 20 per cent between 1977 and 1984 (only fish exports progressed).5 This poor export performance is mainly due to the overvaluation of local currency in the late 1970s, but equally results from excessive protection of the domestic market and high production costs. The protection policy, so advantageous for import-substitution industries in the 1960s, hindered the export sector, both by making production from the home market more attractive and by allowing wages to increase. Production costs are particularly high in Senegal (Barbier 1989). Labour is more expensive than in most other countries of the CFA

164

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

Table 8.1 Rate of effective protection by sector Sector

Domestic market CEAO market Other exports Total

TINNED TUNA-FISH (subsidized) TINNED TUNA-FISH (non-sub) FROZEN FISH (sub) FROZEN FISH (non-sub) BISCUIT-TRADE (sugar wp) BISCUIT-TRADE (sugar cp) CONFECTIONERY (sugar wp) CONFECTIONERY (sugar cp) DAIRY PRODUCE Natural milk and yoghurt Unsweetened condensed milk Sweetened condensed milk FLOUR-MILLS TINNED VEGETABLES (tomato) OIL-WORKS TEXTILES (NPC output=tariff) TEXTILES (NPC output fraud)* Spinning-mills (tariff) Spinning-mills (fraud)* Raw fabrics Printed fabrics (tariff) Printed fabrics (fraud)* Hosiery (tariff) Hosiery (fraud)* SHOE INDUSTRY SOAP MANUFACTURE PAINT INDUSTRY COSMETICS MATCHES PAPER CARDBOARD PACKAGING CEMENT-FIBRE ARTICLES (tariff) CEMENT-FIBRE ARTICLES (NPC 1.0) VEHICLES (commercial) METALLIC PACKAGING BICYCLES AND MOPEDS ELECTRIC BATTERIES SUGAR BATTERIES AND ACCUMULATORS

1.00 1.00 −0.15 −0.01 0.07 −0.30 −0.36 −0.58 0.00 2.44 0.94 −0.01 4.26 1.05 −0.05 0.10 0.69 0.20 0.18 0.14 0.05 1.16 0.64 2.41 1.02 −0.71 3.88 −0.36 2.29 0.70 0.21 0.68 −0.32 2.77 1.05 6.76 0.40 −2.47 −2.08

0.00 0.00 −0.15 −0.29 0.25 −0.25 −0.31 −0.29 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.27 0.27 0.37 0.37 −0.15 0.27 0.26 7.10 7.40 0.19 0.00 0.00 0.62 −0.25 0.28 0.00 0.00 0.00 0.00 0.00 0.00 0.90 0.00 0.00

0.35 −0.27 0.40 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 −0.67 0.00 0.00 −0.16 0.03 0.03 0.01 0.00 0.41 0.36 0.35 0.34 0.50 0.08 −0.17 0.23 −0.03 −0.25 −0.17 0.43 −0.09 −0.11 0.00 0.07 −0.87 −0.04 0.00 0.00

0.36 −0.25 0.28 −0.05 0.12 −0.26 −0.36 −0.55 0.00 2.44 0.94 −0.03 4.26 1.05 −0.13 0.08 0.40 0.07 0.11 1.10 0.09 0.94 0.51 1.48 0.64 −0.69 3.67 −0.28 1.65 0.48 0.43 0.65 −0.32 8.25 0.80 6.68 −0.38 −2.47 −2.08

IMPORT LIBERALIZATION IN SENEGAL

165

Sector

Domestic market CEAO market Other exports Total

PESTICIDES AND PLANT-PROTECTION PRODUCTS DETERGENTS AND OTHER SOAPS SALT GALVANIZED OR ENAMEL HOUSEHOLD GOODS CLINKER AND VARIOUS PHOSPHATES LIMESTONE PHOSPHATES TOBACCO

0.07 0.18 −0.96 0.85 −0.12 −0.21 1.22

PAPER WRAPPINGBAGS FERTILIZERS FISH FLOUR PETROLEUM PRODUCTS OTHER TEXTILE PRODUCTS CONSTRUCTION MATERIALS PREFABRICATED BUILDINGS OTHER CONSTRUCTION MATERIALS PLASTIC (PVC) ARTICLES FLEXIBLE PLASTIC FOAM METAL SCAFFOLDING SOFT DRINKS AND BEER MINERAL WATER AGRICULTURAL EQUIPMENT IRON RODS, FOR CONCRETE, ETC. PHARMACEUTICA L PRODUCTS SAWN TIMBER LABELS AND BOXES CATTLE FOOD GAS AND OXYGEN

−0.10 0.00 0.47 0.57 −0.10 0.00 0.44

−0.10 −0.08 −0.02 −0.03 −0.04 −0.11 0.27

−0.27

−0.04

0.00

−0.26

−0.18 0.00 −48.16

−0.41 0.00 0.00

−0.14 −0.33 2.91

−0.21 −0.33 −34.44

0.00

0.00

0.30

0.00

−5.12

0.00

0.00

−5.12

0.29

0.00

0.00

0.29

−0.01

0.00

0.00

−0.01

−0.22

−0.07

−0.12

−0.22

−0.40

0.00

0.00

−0.40

−0.09

0.00

−0.18

−0.09

0.75

0.00

0.00

0.75

0.53 0.07

0.00 0.08

0.00 0.08

0.53 0.07

0.92

0.51

0.01

0.80

−0.31

0.42

−0.04

−0.15

3.04 0.52

0.00 0.00

0.00 −0.06

3.04 0.16

0.84 −0.23

−0.04 −0.10

−0.09 0.00

0.72 −0.23

−0.37 0.17 0.26 0.61 −0.04 −0.12 0.96

166

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

OTHER PRINTED −0.26 0.00 0.00 −0.26 MATERIALS BATTERIES FOR −2.08 0.00 0.00 −2.10 ACCUMULATORS Source: Rapport sur l’étude de la structure des incitations industrielles, Dakar: Ministère de l’Economie et des finances, Direction de la Prévision et de la Confoncture, May 1986. Note: * Price comparison with prices applied in fraudulent transactions. NB: Domestic added-value at negative world prices for fertilizers, petroleum products and sugar.

zone: in 1986, the legal minimum hourly wage was 184 CFA francs compared to 114 CFA francs in BurkinaFaso, 109 CFA francs in Mali and 72 CFA francs in Togo. In Côte d’Ivoire it was higher (207 CFA francs), but so was labour productivity. Senegalese labour legislation is a major restraint for employers who, among other restrictions, may not dismiss personnel without authorization. Entrepreneurs are also penalized by the cost of technical inputs, especially energy and water (the price of a kWh charged to industrialists was approximately 55 CFA francs, compared to 30 CFA francs in Côte d’Ivoire). Finally, industries compelled to use inputs produced locally by officially designated firms pay these well above world market prices. Awareness of these difficulties led the government to adopt an export-subsidy scheme in August 1980, which, however, proved insufficient to overcome these handicaps. The undeniable weaknesses of Senegalese industry at the beginning of the 1980s, recognized both by the government and foreign donors, meant that structural adjustment required a change in trade and industrial policy. The New Industrial Policy (NIP) and import-liberalization measures In Structural Adjustment programmes (SAP II and SAP III), the Senegalese Government, together with the World Bank, defined a programme of reforms aimed at simultaneously rationalizing the system of protection and industrial incentives. The NIP proposal was adopted on 10 February 1986. Initially, all the measures were to be applied within a period of three years (1986–8), and industrial recovery was expected to begin by 1989. The programme’s various measures can be grouped under five headings: 1 Rationalization of the protection system, i.e. liberalization of imports. The objective is not to abolish protection totally but to implement a less distorting trade and industrial policy (tending towards greater neutrality) whose effects could be controlled. This meant simplifying the system and striving for a single tariff that would be both lower and homogeneous. This equally meant abolishing a number of practices that created confusion and complication (such as mercurial values).6 Exclusive reliance upon tariff policy was seen as a necessary first step. This would later allow the industrial sector to be directed through a modification of tariffs, gradually shifting away from neutrality. Import liberalization is a prerequisite for such a policy and is also consistent with the fiscal objectives: lower and simpler tariffs should reduce smuggling and fraud. Import tariffs are also to be lowered and ‘codes de précision’7 abolished. In order to clarify the system, mercurial values should no longer be used for protection purposes. Finally, all quantitative restrictions would be progressively dropped. 2 Improvement of export promotion measures. Export subsidies will be rationalized, and so will suspensive customs regulations. In order to encourage exports, bureaucratic procedures will be simplified;

IMPORT LIBERALIZATION IN SENEGAL

167

studies will also be undertaken to examine the national institutions for export promotion (trade centres, insurance companies, etc.). Finally, the quality of goods produced will be more strictly controlled, with a view to developing export markets. 3 Amendment of the Investment Code and the repeal of preferential market agreements. The reform of the Investment Code should reduce the advantages formerly granted to government-approved firms. The issue of ‘enterprises conventionées’8 (i.e the most protected ones) is a delicate one: even if the government does sign new contracts, running agreements cannot be cancelled without a legal settlement. The principle of their re-negotiation has been admitted and audits will take place for each specific case. 4 Improvement of firms’ economic environment, especially through liberalization of prices and distribution. One of the measures contemplated concerns the liberalization of imports, namely simplification of the procedures necessary to obtain an import-export licence. 5 Modification of labour legislation, particularly restrictive for employers, in order to make the conditions of employment, wages and dismissal more flexible. Overall, this reform package is internally coherent and consistent with the initial objective, which is to improve industrial incentives within the framework of the adjustment programme. It appears intellectually sound, given the diagnosis of the industrial sector made at the beginning of the 1980s. However, closer examination reveals a number of weaknesses: Many of the measures do not go any further than announcing a reform or a study of a specific issue. Given the time span necessary for implementation, one could not realistically expect that measures aimed at supporting industrial activity and reducing costs (re-negotiation of protection conventions, reform of labour legislation and reduction of social security charges) would be put into effect within the originally planned three years. But, on the other hand, measures relating to deprotection (tariff reform, elimination of quantitative restrictions), a number of which were conditions laid down by the World Bank, were to be implemented much faster. Industry would then be submitted to international competition before it could benefit from supporting measures. Such measures, as well as reforms of the banking system aimed at increasing investment, should have been taken prior to deprotection. Furthermore, according to the reform programme, imports were to be liberalized and quantitative restrictions abolished within two years. Timing of deprotection reforms is a difficult question (Steel 1988). In 1985–6, the World Bank’s trade policy doctrine was largely influenced by the works of A.O.Krueger, who favoured a rapid liberalization of the economy in order to avoid the political disturbances that would arise if reforms were long drawn out. This argument is valid, but we must remember that brutal changes entail great economic and social costs and that the speed of reform depends on the absence of rigidities, especially in the labour market (Michaely 1986). This is not the case in Senegal, where legislation is strict. As noted by J.P. Barbier (1989) ‘cultural factors cannot be overlooked. African culture, which prizes highly group solidarity, tradition, consensus, oral tradition and ways of life based on non-monetary exchange, cannot integrate concepts of competitiveness (and hence of competition) as quickly as other, e.g. Asian, cultures. The integration of the time element in policy design is crucial.’ In practice, the ‘time’ factor in economic adjustment is difficult to master. National authorities, with the assistance of international creditors, must simultaneously take into account short-term (financial) requirements and medium-term perspectives (sectorial policies). The World Bank negotiates the terms of adjustment with the government and, according to the decisions reached, grants a loan enabling the country to meet short-term commitments. Under these conditions, one understands that some elements of the industrial policy of Senegal were hastily negotiated to speed up World Bank funding. Predictably, problems overlooked in the negotiation phase would erupt during implementation.

168

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

IMPLEMENTATION OF THE IMPORT LIBERALIZATION PROGRAMME According to the initial schedule, imports were to be liberalized on 1 July 1988. This section looks into the implementation process. It examines first the time-schedule effectively followed, and moves on to analyse diversions from initial objectives realized during implementation. The time-schedule of import liberalization The reduction of import taxes progressed according to plans (law 86– 36 of 4 August 1986). Following the wish expressed by the customs administration, this was done in two stages: July 1986, and July 1988. However, this was a concession on their part, as they would have preferred to stick to the 1986 reform. The second reduction was imposed by the World Bank. Imports are subjected to two separate forms of taxation in Senegal: customs’ duty, which stands at a uniform rate, is supplemented by fiscal duties, which can take four different rates (reduced, standard, increased and special) depending on the nature of the goods imported. The first reform, in 1986, left taxation of inputs untouched (raw materials and semi-finished products) and these as previously continued to carry the reduced fiscal rate. Harmonization of tariffs was brought about by the reduction of duties levied on finished products. These went on carrying the increased fiscal duty rate, which was lowered by 15 points in 1986. The ‘protection differential’ (i.e. the difference in total import duties between finished products and intermediate goods) was reduced from 40 points in 1985 (before the NIP) to 25 in 1986 in order to reduce effective protection. In addition to the rate cuts (reductions of 5 points on customs’ duty and reduced fiscal rates, and of 15 points on the increased fiscal rate), the last stage of the tariff reform included a reclassification of products in order to establish a progressive taxation according to the degree of processing: raw materials would continue to be taxed on a reduced fiscal basis, but semi-finished products would be subject to the standard fiscal rate. Under these conditions, the protection differential between finished products and raw materials would stand at 20 points but, as from 1 July 1988, the differential for firms whose inputs are semi-finished products would amount to only 10 points. Mercurial values were to be eliminated in order to clarify the tariff system. In 1986, it was decided that mercurial values would not be used for protection purposes any more. However, they were to be retained until July 1987, but use would be restricted to control under-invoicing and to check dumping. The workings of the ‘precision codes’ system were also to be investigated pending abolition. This was not done, however, and ‘precision tariffs’ are still used today to tax some goods at an above-legal rate. This is the only measure of the tariff protection reform not to have been implemented. The time-schedule for gradual abolition of quantitative restrictions was adhered to equally strictly. In fact, some decisions were even taken before the planned date. The last stage (announced for January 1988) was delayed, but by a month only (February of 1988). According to the initial programme, imports of noncompeting goods were to be liberalized first. Since those goods were not produced domestically, impact on local industry would be minimal. The following stages, however, were expected to penalize Senegalese industry. Liberalization was phased accordingly, starting from least sensitive sectors and gradually moving into most vulnerable ones: textiles, the shoe industry and school supplies. Quantitative restrictions on industrial goods are now abolished, save for the notable exceptions of goods produced by enterprises benefitting from preferential conventions (‘entreprises conventionnées’), such as sugar and cement, since these conventions have not been renegotiated. Except, then, ‘codes de precision’ and ‘entreprises conventionnées’, all the import-liberalization measures were adopted, including simplification of administrative procedures necessary to obtain import-

IMPORT LIBERALIZATION IN SENEGAL

169

export licences. But supporting measures intended to help industrialists adapt to the new situation (such as the reform of labour legislation, or actions to reduce the cost of factors of production) were not adopted, except for a few reforms of the business environment. We can then wonder whether, under the pressure of industrialists, the Government of Senegal has not deliberately attempted to limit the impact of deprotection on industry. From deprotection to tariff reprotection As from June 1987, and especially since July 1988, i.e. after the second tariff reduction and the total elimination of quantitative restrictions, protection was reinstated on some industrial sectors. Senegalese authorities had come to realise the harmful effects of deprotection upon industry. Among these, the abrupt abolition of quantitative restrictions allowed cheap imports in, against which domestic industry (such as textiles), weakened by previous excessive protection, could not compete. The design and implementation of supporting measures were equally proving more difficult than expected. The principle of abolition of quantitative restrictions was maintained, however. Instead, tariff policy was called upon to reinstate protection. A number of instruments, respecting at first the 1986 legal framework, were used. ‘Codes de precision’ allowed ‘exceptional’ reductions of fiscal duties on inputs, thus increasing the protection differential. This became a frequent practice, and customs authorities managed to maintain a differential of no less than 20 points on all sectors. This is, of course, not what had been initially planned by the law 86–36. ‘Safeguard’ mercurial values were established to mitigate the effects of the lift of quantitative restrictions and reduced import duties on finished products. These protection-oriented mercurial values applied to most locally produced goods. This system, which had been abandoned in 1986, was reinstated in contradiction to the initial objective of NIP.9 ‘Minimum de perception’ is another device that was resorted to. It establishes a threshold on duties to be collected on a specified good. It is normally used to consolidate revenues on goods with an uncertain taxbasis (in cases such as dumping or imports of second-hand or downgraded goods). It can also be used to protect an industry and, if high enough, can prove very effective. In July 1988, it was substituted for quantitative restrictions on four products to protect a local firm (CAFAL). Finally, in August 1989, the government modified the 1986 legislation and increased the uniform rate of customs’ duty by 5 points, bringing it back to its pre-1985 level (i.e. prior to the NIP). After these protective measures were adopted, the government, realizing that there was no short-term way to amend the labour code or reduce the costs of factors of production, reimposed a degree of tariff protection to limit the unforeseen economic and social costs of deprotection. Imports then were not totally liberalized, as had been planned within NIP. But since protection is now brought about by different instruments, the industrial incentive system has changed. GLOBAL ASSESSMENT OF THE EXPERIENCE The initial idea of the reform was to clarify and simplify the protection system in order to monitor more closely its consequences. Although quantitative restrictions in effect have been abolished, one can wonder whether the new protection measures (extensive use of ‘protective’ mercurial values and of exceptions to the tariff system) have not actually increased the complexity of the system.

170

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

Precise assessment of this new system’s impact on the industrial sector would require an evaluation of the effective rates of protection of various economic sectors. But since the relevant information is not available, we can only reflect upon possible consequences of this reform as it is presently applied. The effects on industry10 The industrial sector’s situation has constantly deteriorated since the first measures of deprotection were implemented. According to the Statistical Department, the industrial production index (groundnut oil works excluded) dropped by 4 per cent between the first quarter of 1987 and the first quarter of 1988 and by 20.5 per cent between the second quarters of 1987 and 1988. Many firms went into bankruptcy: 23 firms closed down between February 1986 and November 1988. Although often considered the sole cause of the deterioration in the results of the industrial sector since 1986, changes in trade policy do not alone explain the recent evolution of industry, which may be no more than the continuation of the trend of the 1970s. Indeed, many firms were already in difficulties before the shift in policy: Bata, for instance, applied before 1986 for authorization to reduce its work force. Many industries with obsolete equipment seized the opportunity offered by the NIP to invest in other activities. The deprotection officially announced by the Government and partially implemented within a few months partly explains the rapid decline of foreign private direct investment. The absence of positive measures in favour of industry and the adoption of an unfavourable Investment Code also contributed to this decline. Indeed, 1985, 1986 and 1987 were years where investment flows were negative, with disinvestment accelerating in 1987.11 Since it lowered expected gains for black marketeers, the 1986 reduction in tariff rates probably reduced fraud. But the informal sector in Senegal paradoxically seems to be flourishing, which suggests that local firms are subjected to strong competition from fraudulent imports. Now, if the reduction of tariffs did indeed discourage black marketeers, the simplification of import-export licence formalities in May 1987 allowed a greater number of operators to engage in international trade (since these reforms were implemented, the number of import-export licences issued has multiplied by four). This in turn has multiplied the number of potential defrauders. The pre-1987 bureaucratic formalities restricted delivery of import-export licences to a small group of people, all of whom belonged to the privileged classes, thus maintaining fraud within some predictable proportion. All in all, the import liberalization measures adopted in 1986 and 1987 probably lowered the proportion of fraudulent transactions undertaken by each operator, but increased the number of such operators. And, with an increased number of potential defrauders, the 1989 rise in tariffs might well spark-off a new round of illegal import practices. The complete elimination of quantitative restrictions has contributed to massive imports of undervalued goods (under-invoicing, export subsidies in the country of origin, dumping, downgraded and second-hand goods). Indeed, under a quota system, importers are few and well-known. This allows some form of control on the imports of restricted goods, even within a generally fraudulent environment. Massive imports of consumer goods at abnormally low prices amount to unfair competition against which local firms, particularly vulnerable, cannot stand. Mercurial values and minimum fixed charges were used to compensate for undervaluation of imports, but proved insufficient. The reforms adopted penalized industrialists rather than traders. Nonetheless, the reintroduction of quantitative restrictions is undesirable: they represent a loss of fiscal revenue and contribute to corruption. Moreover, such a measure would call into question the principle of a purely tariff protection, which the NIP has firmly established.

IMPORT LIBERALIZATION IN SENEGAL

171

Impact of NIP on the evolution of behaviours One positive result of import-liberalization within NIP was to bring about a change in the attitudes of involved agents, that is, the government, the local industrialists and the World Bank. After introducing measures intended to deprotect industry, the government reinstated a certain degree of protection through tariff measures. Various reasons explain this change of attitude. The Government was sorely pressed by short-term commitments when it adopted NIP in 1986. It was unable, therefore, to adapt the standard World Bank programme to the specific needs of Senegal. As implementation proceeded, the programme’s shortcomings, which the government could not ignore, became increasingly apparent. Economic and social costs resulting from the brutal deprotection of local industry threatened to spark-off social disturbances. Therefore, it was politically expedient to revert policies, because the positive support measures could not be implemented. Since the official announcement of NIP, the private sector has reacted by pointing out to policy decisionmakers and politicians the negative consequences that deprotection would entail unless compensating measures were taken. The many critical comments made by the CNPS (‘Centre National de Patronat Sénégalais’, the Senagalese Business Confederation) definitely influenced the Government’s decision to reintroduce a certain degree of protection. Over-protection granted to industries during the 1970s had almost completely eliminated competition. Industrialists usually enjoyed monopolistic or oligopolistic situations that did not encourage them to improve their competitiveness. The adoption of the NIP, and especially import liberalization measures, shook them out of lethargy. Irritated at not having been consulted at the programme’s planning stage, the private sector promptly counterattacked and organized itself. The NIP mobilized Senegalese employers, both national and foreign, who joined forces within the CNPS in order to define means of action and adopt a common attitude. The reform was strongly criticized by industrialists, but the episode made them face their responsibilities and realize the urgency of adaptation to international competition. The World Bank’s attitude has also evolved since 1986. The reform presented in the SAPs II and III contained measures concerning the deprotection of industry, elimination of mercurial values, abolition of ‘codes de precision’ and implementation of accompanying measures. In practice, the programme differed greatly from the initial project. The World Bank seems to have gained some awareness of the reform’s shortcomings and of the existence of constraints specific to Senegal. It now tolerates the use of protective mercurial values and other tariff techniques as a way to increase tariff protection without revising rates upward. It has also accepted the August 1989 increase of customs duties. However, for all its divergences from the initial project, the industrial policy reform maintains its initially stated objectives. CONCLUSION The initial guidelines for import liberalization proved difficult to maintain when put into practice. However, the attempts to enforce them (partially successful since quantitative restrictions were abolished) did yield a positive result: attitudes of different actors are converging. Concertation now seems more feasible. This could lead to jointly undertaken and more efficient efforts to devise ways of attaining the programme’s initial objectives. Indeed, in spite of implementation difficulties, liberalization is a prerequisite for improved resource allocation and, as such, a necessary step in structural adjustment. Its success greatly depends on additional measures taken to increase firms’ competitiveness and facilitate their funding. The timing of the programme is also an important factor.

172

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

Senegal’s experience also shows that quantitative restrictions can be a very effective import-control instrument in countries where quality and price-controls by customs’ administration is inadequate. However, the notion of a purely tariff protection should not be repudiated: it is easy to administrate and supervise, and creates fiscal revenues. Other systems than physical control of imports, in particular mercurial values, can be used to check import-undervaluation, and to raise revenues. The elimination of quantitative restrictions should proceed gradually, starting with the least vulnerable sectors. The ensuing transition period, which will vary among countries, would give time to organize an effective price and quality control system on imports, and to develop international cooperation to eliminate under-invoicement. NOTES This chapter is based on a preliminary version of a report prepared for Elliot Berg Associates as part of a study assessing Senegal’s structural adjustment experience. The report was undertaken for USAID (United States Agency for International Development). 1 Import duties and taxes represented, on average, 25 per cent of the government’s global revenues in sub-Saharan African countries (simple arithmetic average for three years: 1983, 1984 and 1985), compared with approximately 16 per cent for other LDCs as a whole (Geourjon 1988). 2 These rights include customs duties, strictly speaking, and fiscal taxes; the latter vary greatly according to the type of goods concerned. VAT is not an entry charge, as it is levied on imported and locally produced goods alike. 3 In 1984, 160 products were controlled. 4 An effective protection rate equal to 8.25 means that domestic added-value can exceed international added-value by 825 per cent. 5 For various reasons (especially the dilapidated state of buildings, the competition of factory ships, and the overexploitation of fishing grounds) this branch of activity has been in difficulty since 1989. 6 Mercurial values are officially fixed price-lists used as reference when calculating duties and taxes collected by customs. The system is used in order to avoid under-invoicing, as well as to enable the authorities to take action concerning domestic prices. It can also be used to modify the tariff protection applied to certain goods without modifying rates. 7 A technique allowing for any given product’s tariff, as specified in legal texts, to be modified according to convenience, thus multiplying the number of exceptions. 8 According to the Investment Code, such enterprises are granted special privileges regarding tax exemptions, price-setting, protection on output and so on. These privileges are specified in a written document, the ‘Convention’, which is legally binding. 9 However, the reduction of fiscal duty rates was not reconsidered but maintained. 10 These effects arc analysed in depth in Chambas and Geourjon (forthcoming). 11 World Tables, World Bank.

REFERENCES Barbier, J.P. (1989) ‘Reflexions sur la compétitivité: Comparaisons AfriqueAsie’, Notes et Etudes CCCE26, (May 1989). Chambas, G. and Geourjon, A.M. (forthcoming) ‘The effects of a highly controversial reform on the industrial sector: the New Industrial Policy in Senegal’. To be published in the conference proceedings of the Development Studies Association and the Development and Project Planning Centre, Bradford: University of Bradford. Geourjon, A.M. (1988) ‘La protection commerciale’, in P. and S.Guillaumont (eds) Strategies de développement comparées zone franc et hors zone franc, Paris: Economica.

IMPORT LIBERALIZATION IN SENEGAL

173

Krueger, A.O. (1978) Foreign Trade Regimes and Economic Development: Liberalization, Attempts and Consequences, Cambridge, Mass.: Ballinger. Michaely, M. (1986) ‘The timing and sequencing of a trade liberalization policy’, in A.Choksi and D.Papageorgiou (eds), Economic Liberalization in Developing Countries. Oxford: Blackwell. Steel, W.F. (1988) L’ajustement de la politique industrielle en Afrique sub-saharienne, Finances et développement, March 1988. World Bank (1987) World Development Report, Washington DC: World Bank.

9 Impact of price-based and quantity-based import control measures in Nigeria T.Ademola Oyejide

INTRODUCTION Interest in the empirical analysis of changes in the structure of imports of less developed countries derives from at least two considerations. First, an understanding of the causes and consequences of such changes could provide an insight into the role of imports in the process of economic development. Second, the extent to which changes in the structure of imports are policy-induced could provide useful information to the planner and policy-maker. Such a knowledge may assist in determining whether and how import control measures of different kinds may be used as instruments for bringing about a given pattern of development. It is generally believed that LDCs tend to deliberately shift their import structure away from consumer goods towards intermediate and capital goods in their process of economic development. Two important reasons apparently lie behind this behaviour. One is that, because their economic development strategy is based on industrialization via import-substitution, it is easier to replace imported consumer goods through domestic production than it is to tackle intermediate and capital goods, for which domestic markets are usually much smaller and production technology may be more sophisticated. A second reason often cited is the need to spend scarce foreign exchange only on ‘essentials’, and consumer goods are not usually thought of as falling within this category. In either case, deliberate import control measures lead to increased domestic production of consumer goods and a decline in the relative share of this category of imports. Bhagwati and Wibulswardi (1972) provide empirical evidence, on the basis of a cross-country study, which supports the proposition that LDC import structure tends to shift away from consumer goods. This evidence is derived from a comparison of relative shares among various groups of commodity imports. This chapter provides a more detailed analysis of the Nigerian experience during the 1960–88 period, and compares the impact of policy-induced changes in the structure of imports over sub-periods during which price-based import control measures prevail with those over which quantity-based import controls are effective. In the process, inferences are drawn regarding the relative effectiveness of both types of import control regimes. This is followed by a description of the structure of imports, and the evolution of Nigeria’s import control regimes from 1960 to 1988. Then, analytical models show the effects of import control regimes, and some of these are used as the basis of an empirical analysis of the effects of price-based versus quantity-based import control measures.

IMPORT CONTROL MEASURES IN NIGERIA

175

THE GROWTH AND STRUCTURE OF IMPORTS Nigeria’s imports grew quite rapidly during the 1960–88 period. As Table 9.1 shows, the value of imports rose from 0.43 billion naira in 1960 to 0.55 billion naira five years later; it reached 0.76 billion naira in 1970. Nigeria experienced an oil boom, starting in 1973, and the expanded aggregate demand associated with the boom spilled over into increased demand for imports. This explains the sharp increase in the value of imports to more than 3.7 billion naira in 1975. The boom-induced increase in imports peaked in 1980, when total value of imports was just over 9 billion naira. Subsequently, as the world oil market slumped from mid-1981, Nigeria’s imports also fell to less than 8 billion naira in 1986, but imports rose rapidly again and reached a value of almost 25 billion naira in 1988. The growth of imports is much better appreciated, perhaps, by looking at the index in Table 9.1. Based on 1970, the value of imports almost doubled in the 1960–70 decade; but it rose more than Table 9.1 Growth of Nigerian imports 1960–88 1960

1965

1970

1975

Value 431.8 550.6 756.4 3721.5 (million naira) Index 57.1 72.8 100.0 492.0 (1970=100) Source: Central Bank of Nigeria, Annual Reports, various years.

1980

1985

1988

9095.6

7932.9

24900.4

1202.5

1048.8

3292.0

four-fold during the 1970–5 period. In fact, during the 1970–80 period, imports increased twelve-fold, while between 1980 and 1988 imports increased almost three-fold. As the value of imports rose, so did the structure of imports (see Table 9.2). This shows that during the first half of the 1960s, consumer goods accounted for about 48 per cent of total imports, while capital goods took 27 per cent, leaving industrial raw materials with 25 per cent. In the next 15 years, the structure of imports had changed substantially. The share of consumer goods had fallen by almost 35 per cent from 48 per cent to 31 per cent, while the shares of both capital goods and industrial raw materials had increased. In particular, the share of capital goods gained 13 percentage points, while industrial raw materials’ share was up by 3.5 percentage points. The structure of imports was turned around again during 1980–3 as the share of consumer goods rose to almost 42 per cent at the expense of capital goods and industrial raw materials, whose shares fell to 33 per cent and 26 per cent respectively. Finally, during 1987– 8, consumer goods imports lost ground considerably as their share of total imports fell to less than 30 per cent. The major gainer this time was the industrial raw materials imports, whose share rose above 36 per cent; the share of capital goods was also 37 per cent. This analysis establishes the point that the structure of Nigeria’s imports has been subject to rather frequent and substantial changes during the 1960–88 period, and that these changes have not necessarily favoured one category of imports to the exclusion of the others. This analysis raises questions regarding the objectives of import control policy during the period, and the consistency with Table 9.2 Structure of Nigerian imports, 1960–88 (in percentages) Consumer goods

1960–65

1973–76

1980–83

1987–88

47.6

31.2

41.8

26.5

176

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

1960–65

1973–76

1980–83

Capital goods 27.4 40.3 32.5 Industrial raw materials 25.0 28.5 25.7 Source: Computed from original data in Central Bank of Nigeria’s Annual Reports.

1987–88 37.2 36.3

which that policy has been articulated and implemented, as well as its efficiency. EVOLUTION OF THE IMPORT CONTROL REGIME Trade policy may be expected to reflect both the broad strategy of economic development chosen by a country over a given time period and the particular constraints imposed by others, and perhaps overriding, macroeconomic considerations. As it concerns imports, Nigerian trade policy has taken two basic forms in the period under consideration. Changes in the customs tariff schedule, in the form of duties and taxes, were predominant in the earlier part of the 1960– 1988 period. Quantity-based and quantitative import restrictions, exemplified by import licensing and import prohibitions or bans, were more liberally used subsequently. When Nigeria achieved political independence in 1960, import duties on most products were quite low; probably these duties served purely revenue-generating purposes only. The predominant tariff rate was 20 per cent ad valorem, although some processed products attracted 33.3 per cent duty rates. The increase from the standard 20 per cent was considered necessary, according to the budget speech of 1960, ‘in order that local manufacturers should not be at a disadvantage because of the increased duties already imposed on some of the principal ingredients’ of these processed products. In spite of this, however, the budget statement of the Minister of Finance was able to claim, with justification, that the Nigerian market was ‘freer [in 1960] than at any time during the past twenty years’. As it eventually turned out, this statement could just as aptly apply to the next fifteen years. Within the next five years, customs tariff rates rose rapidly so that the range increased to between 40 per cent and 66.6 per cent. The budget speech of 1961 offered a justification for what was to become, in essence, the future direction of import policy by claiming that ‘by and large, increases [in import duties] were imposed upon goods consumed by the better-off sections of the community’. The second half of the 1960s was a difficult period for Nigeria; growing political tensions eventually culminated in a civil war, which started in 1967 and raged until January 1970. The balance of payments problems associated with the war effort clearly dominated all other considerations. Even though tariff rates on many imported consumer goods were raised to 50 per cent, others stayed in the 40 per cent to 66.6 per cent range. But the major radical departure from the past was the introduction and liberal use of quantitative restrictions on consumer goods imports. Thus, between 1965 and 1970, the government physically restricted the importation of food items such as wheat and wheat flour, fruits, sugar, meat and poultry, largely through the imports licensing system. After the civil war, acute shortage of consumer goods in the face of large pent-up demand made it difficult to continue restricting their importation. But the balance of payments situation did not leave much room for manoeuvre. Hence, between 1970 and 1973, many of the quantitative import restrictions that had emerged during the civil war remained in place. Between 1973 and 1977, the balance of payments situation improved dramatically as the era of the oil boom dawned. This permitted a general relaxation of import controls. Hence, quantitative import restrictions were removed and tariff rates sharply reduced, so that, once again, many imported consumer products attracted customs duties in the range of 0 per cent to 20 per cent.

IMPORT CONTROL MEASURES IN NIGERIA

177

More specifically, duties on wheat, milk, cream and salt were reduced to 0 per cent, while duties on maize, rice, fish, meat and tomatoes fell from between 40 per cent and 60 per cent to 10 per cent. The budget speech of 1974 offered a rationalization of food import liberalization in the following words: ‘as the drought situation in parts of the country has created a situation where food in larger quantities than ever before may have to be imported, if only temporarily, import duties have been reduced on essential food items such as rice and maize without endangering the healthy development of local agriculture’. It is thus clear that between 1973 and 1977, as import capacity improved, the focus of trade policy shifted towards concerns of food security in the face of rising domestic inflation and the more liberal reliance on food imports as a way of addressing these concerns. This liberal import regime lasted only for as long as the balance of payments situation remained comfortable. The down-turn in the world oil market between 1977 and the early part of 1979 ended the experiment. Once again, tariff rates were revised upwards, and quantitative import restrictions again assumed a position of central importance in the arsenal of trade policy measures. New import prohibition orders covered a wide range of goods. Import licensing covered additional commodities including sugar, beef, wheat, rice, cereal flours, and other grains. The 1978 budget speech explained these measures essentially as an attempt ‘to give a boost to domestic production’. The world oil market improved again between 1979 and 1980, but this brief rally was permanently reversed by mid-1981. Subsequently, Nigeria’s macroeconomic and financial problems worsened considerably. Import controls, therefore, became further entrenched, so that between 1981 and 1986, most import items were either subject to import licensing or outright prohibition. Back in 1976, the Central Bank’s Annual Report had indicated that ‘the specific objectives of fiscal policy were directed against the high price levels, and included ensuring that essential commodities were imported in reasonable quantities to make up for short-falls in domestic production in the face of vastly increased and rising domestic spending’. But, in 1982, in spite of the fact that inflation continued to be a major problem, the Central Bank’s Annual Report of that year indicated that ‘policy measures applied in 1982 were designed to reduce the level of importation so as to minimise the drain on the nation’s external reserves, generate increased revenue and protect local industries’. In spite of the attempts to rationalize the import control regime after the launching of the structural adjustment programme in 1986, several important consumer items remained banned, including rice and rice products, fruits, poultry, vegetables, vegetable oils and animal oils. A major change in macroeconomic policy that should be noted, however, is that, from late 1986, Nigeria moved from a fixed exchange rate regime to a market-determined floating exchange rate system. This arrangement caused a massive devaluation of the naira and placed a high premium on free market forces for the determination of domestic prices. In spite of the remaining quantitative import restrictions, therefore, the quantity of various categories of imports that actually came in became largely price-determined. MODELLING THE EFFECTS OF IMPORT CONTROLS An attempt to estimate empirically the effects of import controls can be based on several different models. Of particular relevance in this respect are Keynesian analyses and models, which show that import controls can generate several interesting effects. Ocampo (1987), using basic Keynesian models, has analytically derived three such effects. These include a positive employment effect that accompanies the induced importsubstitution, involuntary disabsorption and crowding out of exports as domestic expenditures expand in the face of supply-constrained production of exportables, which compete with importables. It is also shown in Ocampo’s work that the relation between domestic economic activity and the strength of import controls

178

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

depends rather crucially on the induced-import-substitution parameter generated by the typical Keynesian model. While Keynesian models indicate the analytical significance of the induced-import-substitution effect of import controls, several important issues still need to be explored at the empirical level. These include questions regarding how much of the induced-import-replacement occurs and whether this is transformed into permanent import-substitution. Other questions concern how import controls change the structure of imports through changes in (a) the market shares of different import commodity groups, and (b) changes in overall market size. It turns out that these issues can be pursued, at the empirical level, by using an easily quantified market share model (Edozien and Oyejide 1973, 1977). The market share analysis is based on a comparison of import values and shares over two periods. It assumes that a country’s demand for imports is primarily dependent on the country’s income level and prices of the imported commodities in relation to domestic prices. It follows, from this basic assumption, that changes in imports can be explained in terms of changes in income or relative prices, or both. If neither of these changes, imports should remain unchanged. In the same way, deliberate policies aimed at importsubstitution with respect to a particular category of imports would tend to induce a decline in the relative share of this import group out of total imports, essentially by changing relative prices, in a discriminatory manner, against this particular commodity group. In other words, induced import-substitution occurs largely because the imposition of higher tariff rates and/or quantitative restrictions on a specific commodity group raises its real import price in relation to both domestic substitutes and other import categories that are not affected by the import restrictions. Table 9.3 displays the model’s building blocks. This table assumes that if there is no change in relative prices between any two periods, total imports (M) and national income (Y) are tied together in a stable relationship such that, in period I, M1=m1Y1, and, in period II M2=m2 Y2. Since there is no change in relative prices, m1=m2 and, hence, M2=m1Y2, implying that the difference between M1 and M2 is due entirely to the change in income between the two periods. In the same way, the structure of imports is assumed to change only in response to relative prices, assuming that preferences, tastes, technology, etc., are taken as given. Thus, imports of commodity A in period II is given by a1M2 if income is ignored and a1m1Y2 if neither of the ratios(a and m) changes. The actual change in the value of commodity A imports between Table 9.3 An import market share model Period ACTUAL (a) Aggregate Total imports GDP Commodity A imports (b) Proportions Total imports/GDP Commodity A imports/ Total imports HYPOTHETICAL Total imports, assuming no change in imports/GDP ratio Import of A, assuming no change in import structure

I

II

M1 Y1 A1

M2 Y2 A2

m1=M1 Y1 a1=A1/M1

m2=M2/ Y2 a2=A2/M2

M1=m1 Y1 A1=a1M1

m2Y2 a1M2

IMPORT CONTROL MEASURES IN NIGERIA

179

Period Import of A, assuming no change in import structure and import/GDP ratio

I

II

A1=a1m1Y1

a1m1Y2

two periods (A2–A1) can be decomposed into several component parts as follows: (eq. 1) The first of the three components on the RHS of equation 1, that is, (a1M2–a1m1Y2) can be taken as a measure of the import displacement effect (IDE). The second component (A2–a1M2) measures the import market size effect (IMSE), while the third component (a1m1Y2–A1) indicates the overall market size effect (OMSE). Import displacement occurs when both parameters a and m fall so that the import of commodity A falls below the norm. When only parameter a declines, it reflects the effects of a discriminatory policy against the import of commodity A, which is manifested in the form of a change in the structure of imports. For commodity A,this translates into a negative IMSE. Within the framework of this model, the general proposition or hypothesis to be tested is that, in the face of economic development generated by an import-substitution-industrialization strategy maintained by an import control regime, there tends to be a decline in the share of consumer goods out of total imports. Furthermore, the import displacement effect (IDE) and the import market share effect (IMSE) are expected to be negative in the case of consumer goods import but positive in relation to both capital and intermediate goods. If the economy grows steadily and no across-the-board import compression measures are imposed, OMSE can be expected to be signed positively for all categories of imports. Put in tabular form, the sign expectations are listed in Table 9.4. Table 9.4 Sign expectation Import category

Measure of effect

Expected sign

Consumer goods IMSE OMSE Capital goods IMSE OMSE Industrial raw materials IMSE OMSE

IDE − + IDE + + IDE + +

–

+

+

EMPIRICAL ANALYSIS The market share model described above is used to analyse changes in the structure of Nigeria’s imports from 1960 to 1988. But this period is broken into a number of shorter time-intervals as a means of reflecting definite changes in the import control regime. Thus the sub-periods covered by 1960–5, 1973–6 and 1980–3 are, in general, referred to as periods of import liberalization; while during 1966–72, 1977–9 and, particularly, 1984–6 stringent import control measures held sway. The 1987–8 period is something of a hybrid. Import liberalization started, in broad terms, with the adoption of a structural adjustment programme

180

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

in 1986. The massive currency devaluation that accompanied this process essentially replaced quantitybased with price-based import controls, but some significant import bans continued to exist. Period averages are used in the computations because the effects of specific import control measures take some time to work themselves out. Hence, Table 9.5 is based essentially on differences between successive period averages. The periods indicated in the Table 9.5 Estimated effects of import control (in million naira) Period

Effect

Consumer goods

Capital goods

Industrial raw materials

I IMSE OMSE Total Change II IMSE OMSE Total Change III IMSE OMSE Total Change IV IMSE OMSE Total Change V IMSE OMSE Total Change VI IMSE OMSE Total Change

IDE −91.4 131.3 −5.4 IDE −117.9 427.3 628.5 IDE −197.4 1200.7 1263.0 IDE 1375.8 1601.3 2185.1 IDE −856.0 1559.1 −2279.7 IDE −662.8 −592.9 1309.4

−45.1 48.7 74.1 96.7 319.1 165.7 421.9 960.7 259.7 592.2 1532.3 2459.9 −792.0 −1625.9 2693.1 −264.7 −2982.8 245.6 1212.2 −861.3 2565.1 256.5 2051.5 5427.7

26.1 37.3 67.6 81.1 373.1 −62.1 372.2 639.0 335.4 −394.8 1083.7 926.1 −1331.9 250.2 1104.5 709.5 −2319.1 610.4 958.5 −264.9 3119.7 406.2 1960.3 5347.6

−23.8

328.9

237.2

−645.2

−1833.8

2981.1

table are I (1960/5–1966/72); II (1966/72–1973/6); III (1973/6– 1977/9); IV (1977/9–1980/3); V (1980/3– 1984/6); VI (1984/6– 1987/8). Table 9.5 reveals a number of interesting results. One of these is that, as expected, IMSE was negative for consumer goods during all but one (i.e. period IV, 1977/9–1980/3) of the periods covered by this analysis. But it was also negative in relation to capital goods in period V and industrial raw materials in periods II and III. This result indicates that, during periods of restrictive import control, consumer goods suffered relative declines in their import shares and that, as the trade regime became more liberalized, consumer goods recaptured some of the lost ground at the expense of industrial raw materials, in some periods, and of capital goods, in others. Over the 1960–88 period, the aggregate IMSE against consumer goods amounted to 550 million naira or less than 10 per cent of average 1987–8 imports.

IMPORT CONTROL MEASURES IN NIGERIA

181

Import displacement occurred with respect to the three categories of imports in periods I, IV and V. In this aggregate, total import displacement over the 1960–88 period was worth ten billion naira or 40 per cent of average import 1987–8. This shows that the import displacement effect of import controls in Nigeria was quite large. It is worth noting, however, that in terms of sectoral distribution, only 38 per cent of the import displacement was accounted for by consumer goods; capital goods carried 37 per cent of the burden of import displacement, while industrial raw materials contributed another 25 per cent. What may be responsible for this distribution is that a large proportion (71 per cent) of the aggregate import displacement occurred as the import control regime became more restrictive between 1980–3 and 1984–6. The blanket and across-theboard nature of the import restrictions imposed during this period meant that trade policy was motivated primarily by crippling balance-of-payments problems rather than an import-substitution-industrialization objective. A pair-wise comparison of the results for periods V and VI indicates the difference in effects between price-based and quantity-based import control measures. In period V, stringent quantitative import restrictions generated large import displacement effects and large overall reductions in imports affecting all import categories. In period VI, however, a more price-based (lowered tariffs plus large devaluation) import control regime generated a markedly different kind of result. Import displacement of the previous period was more than compensated for and, although consumer goods suffered losses with respect to import and overall market shares, value of imports increased across the board. CONCLUDING REMARKS In the Nigerian experience during the 1960–88 period, restrictive import control measures, particularly of the quantity-based type, have generated significant import displacement effects. In the process, the consumer goods sector has tended to suffer losses in terms of import market share. However, this general tendency has been blunted somewhat and the import displacement effect reversed largely because trade policy has neither been stable nor consistent over time. The fundamental objectives of enhancing economic development through industrial protection and import-substitution (see Oyejide 1975) has often given way to other pressing macroeconomic considerations. The rapid reversals of policies and frequent changes in policy instruments have caused inconsistencies that the results of this study clearly demonstrate. REFERENCES Bhagwatti, J. and Wilbulswardi, C. (1972) ‘A statistical analysis of shifts in the import structure in LDCs’, Bulletin of the Oxford Institute of Economics and Statistics, 34 (2): 229–40. Edozien, E.C. and Oyejide, T.A. (1973) ‘Import restrictions in Nigeria and their impact on imports from Japan’, The Nigerian Journal of Economic and Social Studies, 15 (2): 157–70. and (1977) ‘A statistical analysis of shifts in Nigeria’s import structure’, The Journal of Business and Social Studies, new series, 1(1): 1– 12. Ocampo, J.A. (1987) ‘The macroeconomic effect of import controls: a Keynesian analysis’ Journal of Development Economics, 27:285–305. Oyejide, T.A. (1975) Tariff Policy and Industrialization in Nigeria, Ibadan: Ibadan University Press.

10 Import liberalization in Kenya Jean-Marc Fontaine

This chapter deals with the effects of import-liberalization in Kenya. The introduction recalls some features that make Kenya a somehow untypical SSA country, in that preconditions for import-liberalization are probably more fully met there than elsewhere. First, we put the objectives and problems of import-liberalization in Kenya into perspective, and present some expectations on the influence of import reform. Then, the impact of devaluation and importliberalization on the structure of imports is analysed, focusing on the behaviour of imports of inputs for the manufacturing sector. We next discuss determinants of domestic input-substitution, and show that devaluation induces input-substitution while liberalization slows it down. This is followed by an examination of the relationship between imported input availability, productivity and exports in the manufacturing sector, which reaches the conclusion that, owing to the importance of regional markets for manufactured goods exports, the expected liberal sequence does not apply. We then investigate the evolution of the impact of devaluation on imports over time. This instrument is found to be less and less effective, and contractionary outcomes of devaluation become more and more probable. SOME UNTYPICAL FEATURES OF KENYA Owing to its dominant position in the British East-African colonization scheme (Brett 1973), marketrelations extended into the economy earlier than in neighbouring countries. Although access to export-crops was prevented by the white settlers, African natives engaged in cash-crops as soon as the opportunity arose (Lamb 1974; Leo 1984) and today Kenya is more familiar with the workings of market relationships than most of its African counterparts (Adam and Fontaine 1987). Industrialization, which proceeded quite soon along classical import-substitution lines, brought about a rather unusual form of alliance between foreign and local capital (Swainson 1980). Although this had neither been planned or expected, the manufacturing sector, which operated within a protectionist framework established in agreement with the World Bank, started exporting to neighbouring markets in the mid-70s (Ngeno 1987). After Independence (‘Uhuru’) in 1963, the country adopted a particular brand of market-oriented ‘African Socialism’ quite favourable to private business, with only limited direct intervention from the state, and resorted to prudent and conservative policies (King 1979). From the mid-70s onward, although the economy was subjected to the same kind of shocks as other African countries (oil crises in 1973 and 1979, export booms in 1976–7, and a mini-boom in 1986, acute balance of payments crisis starting in 1979–80, stabilization episodes, structural adjustment programmes, drought and food shortages in the early 1980s), overall macroeconomic management avoided gross distortions of agricultural prices (Jabara 1985; Mosley

IMPORT LIBERALIZATION IN KENYA

183

1986), while budget expansion and currency overvaluation remained within manageable proportions (Killick 1984). All this explains why Kenya was considered as the African country where adjustment policies had the best chances: if they failed here, then they would fail everywhere else in Africa.1 Which makes Kenya something like a test-case for liberalization. Given their above-average chances for success, the workings of external-trade—and especially import— liberalization measures in Kenya will probably not be typical of other African countries. On the other hand, they will clearly illustrate both potential and limits of such reforms. OBJECTIVES AND PROBLEMS OF IMPORT LIBERALIZATION Starting from gradual imposition of obstacles to import from the early 1970s, liberalization was attempted in 1974, 1980, 1984 and 1985.2 The 1974 and 1980 attempts were promptly reversed when balance of payments abruptly deteriorated. The liberalization objective was reaffirmed in the 1986 Sessional Paper and Budget Speech, and globally maintained since then, in spite of occasional administrative interferences and limited reversals. Import controls had two aims. The first was to globally compress imports, in order to alleviate trade balance deficits and domestic demand for currency. This was achieved through global arrangements, such as advance-deposit schemes (whereby importers had to freeze the domestic currency equivalent of imports in a special Central Bank account for some time, usually six months) and constraining financial arrangements (aggravation of credit conditions to potential importers). The second aim was to maintain imports of essential goods at a predefined desired level. This was achieved by ranking them on what amounted to a priority list (import-schedules) and issuing importlicences accordingly. Licences to import essential goods, especially agricultural inputs and industrial intermediary goods, were easiest to obtain, while at the other end imports of luxury consumption goods and locally manufactured goods were discouraged or banned. This did not necessarily mean that ‘priority’ goods could be imported freely. Other restrictions existed: fertilizer imports, for instance, were limited to a few (seven) agreed importers, who thus had the power to influence domestic supply (e.g. on fertilizers, see Lele et al. 1989). Finally, administrative delays and complications were consciously used to regulate import demand following foreign exchange (forex) reserves and requirements. According to importers (private interviews), the duration of validity of import-licences was often shortened in periods of forex shortage, and issue of import licences was commonly suspended between October or November and January each year (and before IMF Stand-By Review meetings or Paris Club negotiations.) Motives for liberalizing imports were manifold. The Kenyan Government’s stated philosophy always has been a fairly liberal one (Killick 1984; Bierman and Fontaine 1987). The World Bank equally constantly pressed for import-liberalization, and is responsible for the disastrous 1980 experience—which was interrupted before the year had elapsed. The longer-run motives concern industrial efficiency and the desire to subject local manufacturers to increased competition from outside, with a view of strengthening export potential. Simplification of administrative red-tape—which implied relaxation of controls—was equally seen as speeding up procedures and allowing greater flexibility and adaptability. So far, the context of import-liberalization almost reads as the set of favourable assumptions postulated in textbooks. But Kenya is not a textbook country, and, in spite of its ‘exemplarity’, shares many common features with other LDCs that will complicate liberalization processes.

184

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

First of all, it is a country where import capacity is well below import-requirements. Between bad and good years, exports cover 40 to 60 per cent of imports. Thus, imports have to be compressed to keep the external gap within financing possibilities, inducing an overall excess domestic demand for imports. Now, given that all demands cannot be satisfied, decisions have to be taken concerning which goods should or should not be imported. Since various goods have unequal strategic importance—scarcity of imported semi-luxury consumer goods, for instance, typically induces less perturbations than scarcity of inputs, if only on employment— and since market power (i.e. ability of demanders to influence prices) does not reflect strategic importance, especially where income distribution is skewed, substitution of price mechanisms for quantity controls—which is the essence of liberalization—will shift the import structure in favour of consumer goods. In a country where industry heavily relies on imported inputs, the domestic supply of inputs might contract, thus inducing import-strangulation (African Development Bank 1990) and, possibly, recession (Diaz-Alejandro 1981). The stronger the previous restrictions have been, the stronger will be the liberalization shift in import structure (Ocampo 1987). Which means that, even if one agrees with the ultimate objective of importliberalization, timing and sequencing will be delicate: abolition of controls, in effect, amounts to a reversal of previous priorities which, if abrupt, can prove harmful to manufacturing activities (Fontaine 1992). Second, the dismantling of controls will leave rate-of-exchange policies as the main (or sole) regulator of imports, which can have two sets of consequences. The first might be to accelerate devaluation. One of the reasons advanced for Kenya’s comparatively good record on currency valuation is domestic compression of forex demand resulting from import controls (Sharpley and Lewis 1990). To remove control is to forgo a policy instrument and to increase the weight of adjustment on the rate of exchange policy. The sparking of a vicious circle centred upon devaluation is another possibility. If devaluation is frequently resorted to and becomes an ordinary policy instrument, actors will adopt their plans (Geronimi 1989) and ‘rational expectations’ on devaluations might develop. Which could bring the question of contractionary effects of devaluation (Krugman and Taylor 1978) into the discussion. IMPORT LIBERALIZATION AND THE STRUCTURE OF IMPORTS This section examines the impact of liberalization upon the structure of imports. It starts from a global import demand function, which is later compared with demand functions for three categories of imports, namely manufactured consumer goods, industrial inputs and capital goods. Impact of liberalization is measured by the inclusion of a dummy variable. The behaviour of total imports For the 1970–86 period, factors influencing total imports were:3 1 2 3 4

The real GNP. The nominal rate of exchange. Previous years’ export purchasing power in terms of imports. The purchasing power of reserves at beginning of year.

Results are summarized here in elasticity form (Eqs 10. la and b).

IMPORT LIBERALIZATION IN KENYA

185

Figure 10.1 Kenya: imports of processed consumer goods, industrial inputs and capital goods, 1970–88 (in millions of 1980 Kenyan shillings) Source: Kenya, Statistical Abstracts

where ε denotes total imports’ elasticity, in relation to: Y=real GNP, Tc=nominal rate of exchange expressed in shillings/SDR; PL=previous year’s export values deflated by current year’s overall import-price index, and R1=foreign exchange (forex) reserves at the beginning of the year deflated by previous year’s overall import-price index. The first three variables are traditional, but we should note the presence of R1, (forex reserves) which indicates that foreign reserve availability acts as a constraint on imports (Khan and Knight 1986). This ranks Kenya among import-compressed economies. (For a description of effects of import compression, see African Development Bank 1990 or Taylor 1988, Ch. 2.) Note also that the variable representing purchasing power of reserves is the value of reserves at the beginning of the year deflated by the previous year’s import prices. This shows that decisions regarding import licences are taken ‘ex ante’, that is, at the beginning of the year, on the basis of information on prices available at the time, which clearly indicates that these decisions operate as binding quotas. One should also note the lagged effect of export-earnings (last years’ exports deflated by present year import-prices), which arises from domestic delays in the distribution of export earnings. Inclusion of a dummy variable for years of liberalization signals an overall increase of total imports of 12 to 13 per cent in liberalization years. Analysis of the structure of imports Three categories of imports were analysed, manufactured consumer goods, industrial inputs and capital goods, whose chronological evolution is shown in Figure 10.1. Econometric analysis yielded the following results in terms of elasticities with respect to the rate of exchange (see Eq. 10.2):

186

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

Figure 10.2 Kenya: year-to-year rate of growth of manufacturing output, 1969–87 Source: Kenya, Statistical Abstracts

where ε stands for elasticity and subscripts i, c, k and m for imports of inputs, consumer goods, capital goods and total imports, and t refers to the nominal rate of exchange. We first note the differential impact of liberalization upon various classes of imports. Capital goods are not affected, while consumer goods and inputs increase by 20 per cent and 17 per cent. If, however, we note that liberalization increases the total volume of imports by almost 13 per cent, then the differential impact of liberalization upon imports of consumer goods and of industrial inputs is 7 per cent for consumer goods against 4 per cent for inputs. As was expected, consumer goods do benefit more from liberalization than inputs, but the shift is relatively moderate. Since imported input availability increases in periods of liberalization, the assumption of import-strangulation in manufacturing activities seems unlikely. We should note, however, the differences between rate of exchange elasticities. Total imports and imports of capital goods have roughly unit elasticities, while consumer goods are more sensitive to devaluations (εo, t= −1.14) and inputs even more (εi, t= −1.45). If import structure were to be regulated solely through the rate of exchange, then we would expect imports of industrial inputs to react more strongly to devaluations than do imports of consumer goods. In other words, control of imports through devaluation alone might result in discrimination against inputs to industry. Thus, the hypothesis of importstrangulation of industrial activities, which was rejected, could be replaced by a softer version, such as, say, gradual import-choking. This, of course, very much depends upon the behaviour of the manufacturing sector, and especially on possibilities of substitution of home-produced for imported inputs. INPUT SUBSTITUTION AND THE BEHAVIOUR OF THE MANUFACTURING SECTOR Simple graphical analysis does not reveal any marked link between liberalization and the level of manufacturing activity. Growth rates (see Figure 10.2) underwent brutal changes with two peaks (1971–2 and 1977), corresponding to Mwai Kibaki’s expansionist budget policy in 1971 and the mid-1970s coffee boom. A ‘mini-peak’ is just discernible in 1986, probably corresponding to the 1986 mini-coffee boom, in

IMPORT LIBERALIZATION IN KENYA

187

an otherwise moderately ascending trend with manufac- turing rates of growth ranging round 5 per cent since 1984. Expectations regarding import-liberalization and manufacturing activity Import liberalization can influence the workings of the domestic manufacturing sector in many ways. On the supply-side, productivity should increase/decrease following the improvement/deterioration in domestic availability of imported inputs (assumed to be of a better quality than their home-produced substitutes) and should increase following competition on output (which should induce rationalization effects). On the demand side, by diverting domestic demand to imported goods (finished products as well as inputs), it should contract outlets and push the level of activity downwards. Theoretical expectations are thus indeterminate. The balance between positive and negative factors will depend upon the relative importance of supply-side or demand-side constraints. One reasonable expectation, based upon the distinction drawn by Taylor (1988) between ‘exhilarationist’ and ‘stagnationist’ diagnoses, is that liberalization will exert globally positive effects in economies constrained from the supply-side, and will prove ineffective or counterproductive if demand-side constraints dominate. Now, as hinted above (Figure 10.2), diagnoses on Kenyan industry are generally of the ‘stagnationist’ kind. Analyses by Van der Hoeven and Vandemoortele (1987), Sharpley and Lewis (1990) and Coughlin (1988) unambiguously point to demand expansion as the prime mover of manufacturing activity. This is confirmed by statistical analysis (see Eq. 10.3) where the level of manufacturing production is seen to depend primarily on domestic demand and regional outlets (and displays a negative time-trend, capturing the progressive loss of dynamism). This leads to low expectations concerning the impact of importliberalization upon manufacturing production. We will now try and isolate the various elements of the liberalization sequence. Imported input availability and input-substitution We first concentrate on the relation between imported input availability and manufacturing activity. The first striking thing is the almost complete absence of relation between these two variables. Statistical analysis does not reveal any significant correlation between them (see Eqs 10.4a and b) and graphical comparison of indices of manufacturing production and of imports of input (see Figure 10.3) shows that, after the 1973–6 shock, manufacturing production develops steadily even though imports of inputs fall. This suggests that Kenyan industry substituted for imported inputs from the mid-1970s onward. There is indeed some evidence of input substitution (World Bank, 1987, Vol. I, pp. 52 ff.), which however did not extend very far backwards to capital equipment goods (Matthews 1985; Coughlin 1988). But the nature of this substitution is not very clear. Although the distinction is not watertight, we can try to separate deliberate from forced (or involuntary) substitutions (see Ocampo 1987). Deliberate substitution results from conscious plans, and typically responds to a shift in relative prices of imported/domestic manufactured goods. Involuntary substitution, on the other hand, is a reaction to forced circumstances, typically import-restrictions or compression. The touchstone would be reversibility: if the trend of input-substitution is reversed when import-restrictions are relaxed, we would suppose that some involuntary element has been at work. Both are probably at work here.

188

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

Figure 10.3 Kenya: index of manufacturing activity and volume index of imports of inputs, 1971–88. Moving average (2 years) Source: Kenya, Statistical Abstracts

Statistical analysis shows input-substitution to depend on relativeprice movements (i.e. to respond to a deterioration in the international terms of trade of the domestic manufacturing sector), as shown by the high correlation between the index of input-substitution and the relative price of imported inputs to domestic manufacturing output (see Eq. 10.3a). Input-substitution would then appear as a deliberate reaction to the evolution of international import-prices and devaluation. But there is equally an involuntary component in this substitution. It is partly reversible and slows down whenever the external constraint eases or import controls weaken, i.e. during the 1976–7 coffee-boom and import-liberalization periods (see Figure 10.4 and Eq. 10.4b). This reversibility does not result only from shifts in input demand from domestic to imported inputs allowed by import-liberalization. ‘Indirect’ substitution equally plays an important part. The relaxation of import-controls increases the variety of imported goods and machinery, which become as a whole more heterogeneous. Consequently, the specifications of comparable spare-parts (e.g. for tractors, pumps, cars, trucks, buses, etc.) become more diversified, preventing the demand for any of them to reach a viable size for home production (Coughlin 1985, 1988). Input-substitution then seems principally of the ‘deliberate’ kind, although the ‘Ocampo effect’ (involuntary substitution and reversibility in periods of liberalization, plus direct and indirect substitution) is present. Although it isn’t dominant, it is not negligible. The relative potency of both effects is roughly −1 (for liberalization) to +7 (for devaluation) (see Eq. 10.4c). This means that import-liberalization opposes input-substitution, while devaluation (through an increase in PR) will enhance it. Two instruments of the liberal package thus exert conflicting effects upon inputsubstitution. LABOUR-PRODUCTIVITY, IMPORTED INPUT AVAILABILITY AND MANUFACTURED EXPORTS The liberalization-productivity sequence is theoretically quite clear. Import-liberalization should increase productivity in the manufac-turing sector through rationalization pressures (induced by increased competition) and improvement in imported input availability. This, in turn, should result in increased

IMPORT LIBERALIZATION IN KENYA

189

Figure 10.4 Kenya: imported input-substitution (MISUB) and liberalization periods (vertical lines)

exports of manufactured goods, which, by expanding the scale of production, should further increase productivity. The actual picture, however, is quite different (see Eqs 10.5, a to d). As expected, the level of productivity4 is influenced positively by the level of domestic demand,5 which might indicate economies of scale and learning effects, and the liberalization dummy is positive, which could capture ‘rationalizing’ influences. But two features contradict expectations. Output per head is negatively correlated with the proportion of manufactures in total exports. And, when input-substitution is brought into the model, it carries a positive sign. All this points to the possibility of an untypical sequence linking expansion of manufactured exports, imported input availability and productivity. Pending further research, two tentative and rather intuitive explanations are presented here. The positive association between productivity increases and input-substitution could be explained by economies of scale and/or learning effects. Substituting for previously imported inputs is a new activity that requires a lot of innovative ingenuity. And, since designation of products is presumably quite strict (because they have to fit into previously designed processes), this must have resulted in adoption, adaptation or substitution of technology, thus increasing labour-productivity.6 If, furthermore, import-controls reduced the span and diversity of imported goods, rationalization and standardization of input-substitution might have accelerated economies of scale. The relationship between manufactured exports and productivity could be explained by the nature and direction of manufactured exports. These are essentially regional exports to neighbouring markets7 and, as such, were not influenced by liberalization but by fluctuations in regional relationships. Figure 10.5 shows the striking parallelism between the shares of manufactured goods and of regional exports in total exports. Statistical analysis singles out two explanatory variables for the export-orientation of the manufacturing sector, that is, the volume of regional trade8 and the import-content of manufacturing activity (Eqs 10.6a and b), while inclusion of the liberalization dummy proved constantly non-significant. Goods exported on the regional markets are simple goods (textiles, soap, pyrethrum products, crockery, processed food) and goods with a high import-content (pharmaceuticals, cement, fertilizers, petroleum products). If output per head in these sectors is lower than average, an increase in the share of exports in total manufactured goods production stimulates relatively less efficient segments of industry, thus slowing down the evolution of productivity and relatively import-intensive productions, thereby increasing the average import-content of the manufacturing sector.9

190

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

Figure 10.5 Kenya: shares of exports to regional market (XPTA/XTOT) and of manufactured exports (XMNF/XTOT) as percentages of total exports

We should note that this does lead to some ‘import-strangulation’ hypothesis in the Khan and Knight (1986) line, whereby exports are negatively influenced by import-controls. But it does not fit in with the liberal argument, since exports do not follow from productivity increases brought about by importliberalization but from a demand-pull factor, that is, the situation of regional markets. This point will not be elaborated further, but it does indicate the need to adapt the liberal argument when a country has both a dual structure of exports (South-North, with comparative advantage on the worldmarket, and South-South, with comparative advantage on a regional, globally protected, market; see Stewart 1984) and a dual manufacturing structure (domestic-oriented, with relatively modern inputsubstituting activities of the assembly-type, and oriented toward regional export, import-intensive and relatively less modernized). DEVALUATION FATIGUE? If quantitative regulations are lifted and the range of rates of import taxes narrowed, the rate of exchange will be the main instrument commanding import demands. One last question should then be asked regarding the effectiveness of devaluation as an import-regulating instrument. The impact on the structure of imports can be predicted from the relative values of rate-of-exchange elasticities of various categories of imports. Since it is higher for imported inputs than for other imports, inputs will react strongest to devaluation. If, as hypothesized earlier, this induces input-substitution and labour-productivity increases, it would globally strengthen the Kenyan manufacturing sector. But, since import-liberalization was seen to slow down and disrupt the substitution process, input-substitution will require protection. Which is simply one way of saying that the infant-industry argument applies strongly to input-substituting activities. The rate of exchange might also be losing its effectiveness as a global import-controlling instrument.10 Although the period is not long enough to apply a proper Chow-test, the elasticity of imports with respect to the nominal rate-of-exchange11 seems to fall over time. If we define various time-periods (1 for 1965–77, 2 for 1974–85 and 3 for 1975–87), regression gives the following results (see Eqs 10.7):

IMPORT LIBERALIZATION IN KENYA

191

The first conclusion is that devaluation is less and less effective in curbing imports. The instrument wears out, for three possible reasons. One is that imports have gradually been brought down close to the incompressible level. This would be consistent with the gradual increase of the coefficient of export purchasing power, which shows imports to be more and more dependent on forex income as relative prices (rates of exchange) lose importance. Another is that, as devaluations become more and more frequent, they are considered as a normal policy instrument, and as such may be subject to a ‘rational expectations’ phenomenon, which reduces their effectiveness. The third one is that donor organizations (EEC, World Bank, Bilateral Aid) have multiplied ‘Commodity Schemes’ since 1986, whereby imports of some specified goods are ensured through specific programmes with ear-marked funds.12 This, of course, disconnects imports from macroeconomic variables. The second conclusion is that, as elasticities of imports fall, the probability that devaluation will induce domestic contraction will increase. Although the Marshall-Lerner condition does not apply here,13 the essential intuition remains valid. If imports (and exports) are rigid, devaluation will cause a degradation in the trade balance. If, on top, the trade balance is initially in deficit, as it is in Kenya, domestic contraction will follow (Hirschman 1949; Krugman and Taylor 1978). CONCLUSION The behaviour of imports of Kenya appears to have been quite typical of import-substituting countries14 with imports of processed consumer goods falling while that of inputs increased, a feature which was reinforced by import-controls (import-licensing or constraining global arrangements). Import-compression is another feature that explains the behaviour of imports, which were found to be constrained by the amount of forex reserves. The alleviation of controls resulted in an overall increase in imports and a shift in their composition, with imports of processed consumer goods benefiting more than inputs, while capital goods were unaffected. However, since total availability of imported inputs increased owing to the effect of liberalization on aggregate import availability, the import-strangulation hypothesis is not fully vindicated. Furthermore, manufacturing activity is more influenced by demand than supply-side factors, and Kenyan manufacturing sectors managed to compensate for the decrease in imported input availability through some form of input-substitution. This input-substitution reacted strongly to the deterioration of the terms of trade of the manufacturing sector and to devaluation. However, it was slowed down by import-liberalization, which, inter alia, prevented rationalization of production in the input-substituting sector. Since input-substitution exerted a positive influence on labour-productivity, there is a good case here for protection and import-controls based on infant-industry arguments. The relationship between the evolution of labour-productivity, exports of manufactured goods and availability of imported inputs is rather unusual. While liberalization exerted a globally positive influence on average labour-productivity in the manufacturing sector, availability of imported inputs slowed it down but benefited manufactured exports. This could mean that increased availability of imported inputs benefited more the export-oriented than the domestic manufacturing sectors. The paradoxical element here is that this result both confirms and contradicts the liberal thesis. It confirms it, in that import-liberalization allows imports of export-oriented inputs to increase, thus pushing exports of manufactured goods up. It contradicts it, however, in that the sectors producing manufactured exports are probably not the most efficient ones, so that labour-productivity is negatively

192

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

correlated with the export-orientation of the manufacturing sector. All in all, import-liberalization, through increased imported input availability, seems to favour exports of manufactured goods on regional, globally protected markets, and not on the world market. The resulting international specialization does not then meet liberal expectations. Finally, the regulation of imports through the sole rate of exchange might come up against some form of ‘devaluation fatigue’. If such was the case, exclusive reliance on rate-of-exchange policy could spark-off contractionary spirals. Both as an economy that has exhausted first-stage import-substituting opportunities and as an importcompressed economy, Kenya would probably benefit more from an industry-oriented rationalization of import-controls than from total liberalization of foreign trade. APPENDIX 10.1 ECONOMETRIC RESULTS IMPORT BEHAVIOUR (a) Total imports 1970–86 (Eq. 10.1a) R2 cor=0.94; D.W=3.02 (Eq. 10.1b) R2 cor=0.97; D.W=2.47 where log in front of a variable denotes logarithm, M stands for real total imports, Y real GNP, T is the nominal rate of exchange expressed in Kenyan shillings/SDR (yearly average), PL is purchasing power of previous year’s export in terms of current year’s import prices, i.e. export receipts lagged one year deflated by current import-price index. R1is foreign reserves’ purchasing power lagged one year. DL is a dummy variable. It is equal to 1 for years of import liberalization (1974, 1980, 1984–7), and 0 otherwise. Unless stated otherwise, data were taken from Kenya’s Statistical Abstracts; rates of exchange from International Financial Statistics, IMF. Coefficients of log Y and log T are in line with expectations, and correspond to elasticities of 0.57 and −0. 98 for income and rate-of-exchange.A1 Inclusion of PL (selected after various trials on various specifications of the function) reveals the influence of the lag in the distribution of export proceeds. Inclusion of R1 indicates that reserves operate as a constraint, which confirms that Kenya is an import-compressed economy. The lag indicates that import

IMPORT LIBERALIZATION IN KENYA

193

quotas are established at the beginning of the year, on the basis of information about import-prices then available. (b) Various categories of imports (Eq. 10.2a)

(Eq. 10.2b)

(Eq. 10.2c) where Mi, Mc and Mk stand for volumes of imports of industrial inputs, of processed consumer goods and capital goods.A2 Imports of capital goods are insensitive to liberalization (no statistical significance). The specification of functions calls for a few comments. Only imports of inputs (Mi) are sensitive to forex reserves, taken here to indicate administrative controls (import licensing), while consumer and capital goods are not (presence or absence of R1and statistical significance). In the case of consumer goods, this might arise from the fact that they are a residual, with more or less constant allocation of forex. In the case of capital goods, this might be due to financing from own funds, which allows the bypassing of restrictions. Consumer goods imports are not influenced by Y, but PL (exports’ purchasing power) is very significant. This probably captures the reactivity of smallholders’ demands to export receipt variations (Haugerud 1981; Kabagambe and Williams 1982; M.Adam, in Adam and Fontaine 1987). The same phenomenon explains why Mi (imports of inputs) do not react to PL (export’s purchasing power). Their demand follows production of domestically manufactured goods, which are relatively insensitive to export receipts. MANUFACTURING ACTIVITY (Eq. 10.3) where log in front of a variable denotes logarithm, MNF stands for the volume of manufacturing production, YNET for GNP net of MNF, as a proxy for aggregate demand, XPTA for exports to neighbouring countries (reconstituted for periods prior to actual operation of PTA) (PTA:Preferential Trade Area), TIM is time. All variables are expressed in constant 1980 shillings. Inclusion of variables capturing liberalization measures (DL, availability of imported inputs, weight of imported inputs in manufacturing production) did not produce any results, with very poor t-Student.

194

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

IMPORT INPUT-SUBSTITUTION 1970–87 (Eq. 10.4a) MISUB=1–(MPI/MNF) is an index of substitution of imported inputs, conventionally fixed at 0.5 in 1970. MPI is the index of import of inputs and MNF the index of manufacturing activity (1980=1). MNF was taken from the World Bank’s World Tables. PR=DfMNF/DfMPI, where Df stands for Deflator (1980=1), measures the evolution of relative prices of domestic manufactured output to imported inputs. Inclusion of DL yields: (Eq. 10.4b) Note that liberalization opposes input-substitution, while devaluation (through its impact on PR) enhances it. Relative impacts are measured by: (Eq. 10.4c) where subscript N indicates a normalized variable.A3 LABOUR-PRODUCTIVITY IN THE MANUFACTURING SECTOR 1970-86 (Eq.10.5a) where MNFW stands for manufacturing output per head (Source: World Tables, World Bank), and XMFX is the proportion of manufactured goods in total exports (% of current values). Alternatively, replacing MNF by MISUB, yields: (Eq. 10.5b) Both models are roughly equivalent. Since MISUB and MNF are highly correlated, they cannot both be included in the same model. A semi-logarithmic model yields: (Eq. 10.5c) and (Eq. 10.5d) The rate of growth of productivity thus seems correlated to MISUB rather than to MNF. EXPORT ORIENTATION OF THE MANUFACTURING SECTOR 1970–1987.

IMPORT LIBERALIZATION IN KENYA

195

(Eq. 10.6a) where logXMFM is the logarithm of the share of manufactured exports in total manufacturing output, XPTA the volume of exports to PTA countries and MIM the average import-content of manufacturing output (index M1/index MNF). Normalized variables yield the following coefficients: (Eq. 10.6b) GRADUAL LOSS OF EFFECTIVENESS OF RATE OF EXCHANGE POLICY Regressions conducted on one same modelA4 for different periods yield: Period 0: 1965–85 (Eq. 10.7a) Period 1: 1965–77 (Eq. 10.7b) Period 2: 1974–85 (Eq. 10.7c) Period 3: 1975–87 (Eq. 10.7d) The elasticity of imports with respect to the rate of exchange and to GNP seems to fall over time, while the reactivity to export purchasing power (PL) increases. This could mean that imports are growing increasingly insensitive to devaluation as well as to variation of GNP, indicating rigidity. The increase in the coefficient of export purchasing power could indicate that the forex constraint is increasingly binding. NOTES FOR APPENDIX 10.1 A1 This is lower than estimates by R.L.Friesen (1975). Difference might arise from the difference of periods analysed (see equations 10.7). However, the ratios of price elasticity to income elasticities are comparable in both his and our study. A2 UN classification ‘Grandes categories économiques’, nominal Kenyan shillings deflated by Kenyan price indices SITC No.6 for M1, „ No.7 for Mk and No.8 for Mc, 1980=1; sources Kenya Economic Surveys and Economic Abstracts (various years). A3 (variable–mean)/standard-deviation. The constant disappears, and the regression retains the same statistical characteristics: R2, t-Student, D.W., F-stat, etc…

196

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

A4 Specification differs from previous tested models: inclusion of forex reserves was not possible for periods prior to 1970. The number of observations did not allow periods to be separated totally. This is also the reason why the Chow-test was not applied.

NOTES 1 ‘It seems that if successful stabilization and good working relationships with the Fund are not feasible in Kenya, it is unclear where else in Africa they might be achieved.’ (Killick 1984:166). 2 The periodization of import-controls presented here results from the following sources: Ngeno 1987; Kenya Government Economic Surveys, Kenya Budget speeches; IMF Trade and Exchange Restrictions (various dates). The effectiveness of measures was checked in interviews with importers in 1987. 3 See detailed results of statistical analysis in Appendix 10.1 (p. 264). 4 Approximated here through the real output per employee index presented in the World Bank’s World Tables. 5 Inclusion of variables representing demand or level of production yield better statistical results than inclusion of a time variable, so that the level of production, and not a time-trend, appears as the relevant explanatory variable. 6 From that point of view, relaxation of import-controls in so far as it slows down input-substitution, could have exerted adverse influences on productivity. 7 UNIDO (1984) estimated the proportion of manufactured exports to regional markets to be somewhere between 56 per cent and 87 per cent, depending upon the definition of sectors adopted (see also Godfrey 1987). 8 Approximated through the volume of exports to PTA countries (data reconstructed for periods prior to 1984). 9 Although Lall, Khanna and Alikhani (1987) found a positive correlation between skill intensity and manufactured exports to LDCs (and Africa) for Kenya, a partial verification on World Bank (1987) figures shows that among the four sectors where export expansion unambiguously proved to influence growth (Beverages and Tobacco, Clothing, Plastic products, Non-metallic mineral products, Vol. I, p. 6) only one (Beverages and Tobacco) had a lower than average import-content (Table IV.2, Vol. I., p. 38), while all four displayed lower implicit productivity indexes than average (Appendix, Table 28, Vol. II, p. 341). Out of the ten sectors that were expected to increase exports according to the 5th Development Plan, three only display higher than average productivity indexes (Glassware, Processed petroleum products, and Textiles) but higher-than-average import-content, and three only have below-average import-content (Leather, Processed food and Paper products). 10 For a general argument in the case of Kenya, see Godfrey 1987. For general implications of devaluation policies in the case of Africa, see Assidon and Jacquemot 1988. 11 We chose to use the nominal rate of exchange rather than real (or real effective) rate of exchange because it is the only proper policy-instrument. Real rates depend on variables that are outside direct policy control, that is, inflation differentials (and direction of trade, if effective rates are considered). Although the net effect of devaluation should be captured by real rates, the policy impact, which we are interested in, is captured by nominal rates. If, for instance, devaluation systematically sparks off inflation and maintains the effective rates unchanged, this will mean that the instrument, i.e. nominal devaluation, is ineffective. 12 This points to an inconsistency between policy recommendation and actual practice: the very same agency (i.e. the World Bank) who insists on import-liberalization also implements ‘Commodity Schemes’ that are, in effect, a recognition of the impossibility of ensuring proper supply of imports through the sole workings of a market mechanism. If speech was brought in line with effective practice, analysis would gain in realism. 13 Due to the stringent assumptions it needs, especially regarding infinite export-supply elasticities. This condition does not apply to African countries. See Bond (1983, 1987) and Fontaine (1987). 14 See Chapter 9 by A. Oyejide in this book for further elaboration.

REFERENCES Adam, M. and Fontaine, J.M. (1987) ‘Kenya, analyse de la fillière café (mimeo), Paris: CNRS.

IMPORT LIBERALIZATION IN KENYA

197

African Development Bank (1990) African Development Report 1990, Abidjan: ADB. Assidon, E. and Jacquemot, P. (1988) Politiques de Change et Ajustement en Afrique, Paris: Ministère de la Cooperation. Biermann, W. and Fontaine, J.M. (1987) ‘Ajustement structurel et stabilisation: Kenya et Tanzanie dans les années ‘80’, Tiers-Monde 28:109. Bond, M. (1983) ‘Agricultural responses to prices in Sub-Saharan African countries’, IMF Staff Papers 30:4. Bond, M. (1987) ‘An econometric study of primary commodity exports from developing country regions to the world’, IMF Staff Papers 34:4. Brett, E.H. (1973) Colonialism and Underdevelopment in East Africa: The Politics of Economic Change, 1919–1939, London: Heinemann. Coughlin, P. (1985) ‘The Kenyan steel and steel-related industries: a programme for domestic reliance and exportpromotion’. Monograph, Industrial Research Project, Nairobi: Economics Department, University of Nairobi. Coughlin, P. (1988) ‘Toward a new industrialization strategy in Kenya?’ in P. Coughlin and G.Ikiara (eds) Industrialization in Kenya, London: Heinemann. Coughlin, P. (1990) ‘Kenya: moving to the next phase?’ in R. Riddell (ed.) Manufacturing Africa, London and Portsmouth: James Currey & Heinemann. Diaz-Alejandro, C. (1981) ‘Southern Cone stabilization plans’, in W.Cline and S.Weintraub (eds), Economic Stabilization in Developing Countries, Washington DC: The Brookings Institution. Fontaine, J.M. (1987) ‘Les projets de liberalisation des agricultures africaines’, Economie et Sociétés, 7:185–208. Fontaine, J.M. (1992) ‘Bias overkill? Removal of anti-export bias and manufacturing investment: Ghana 1983–9’, in C.Kirkpatrick et al., Industrial Trade and Policy Reforms in Developing Countries, Manchester: Manchester University Press. Friesen, R.L. (1975) ‘The determinants and implications of the demand for imports: an econometric study of Kenya’, Eastern Africa Economic Review 7:49–63. Geronimi, V. (1989) ‘Les échange parallèles’ (mimeo), Paris: CERED/ LAREA, Université de Nanterre. Godfrey, M. (1987) ‘Stabilization and structural adjustment of the Kenyan economy, 1975–85: an assessment of performance’, Development and Change, 18:595–624. Haugerud, A. (1981) ‘Economic differentiation among peasants’ households: a comparison of Embu coffee and cotton zones’, IDS Working Papers No. 383 (mimeo), Nairobi: IDS. Hirschman, A. (1949) ‘Devaluation and the trade balance: a note’, Review of Economic and Statistics 23. Jabara, C.L. (1985) Agricultural pricing policy in Kenya’, World Development, 13:5. Kabagambo, D. and Williams, K.G. (1982) ‘The impact of the coffee-boom in Meru district’, Kenya Research Project, Working Paper No. 2 (mimeo), Nottingham: Institute of Planning Studies, University of Nottingham. Khan, M.S. and Knight, M.D. (1986) ‘Import compression and export performance in developing countries’, Development Research Department Discussion Papers, Report No. DRD 197 (mimeo) Washington DC: World Bank. Killick, T. (1984) ‘Kenya 1975–81,’in T. Killick (ed.) The Quest for Economic Stabilisation: The IMF and the Third World, London: Heinemann and ODI. King, J.R. (1979) Stabilization Policy in an African Setting: Kenya 1963–73, London: Heinemann. Krugman, P. and Taylor L. (1978) ‘Contractionary effects of devaluation’, Journal of International Economics 8: 445–66. Lall, S., Khanna, A. and Alikhani, I. (1987) ‘Determinants of manufactured export performance in low-income Africa: Kenya and Tanzania’, World Development 15(9). Lamb, G. (1974) Peasant Politics. Conflict and Development in Murang’a, Lewes, Sussex: Julian Freedman. Lele, U., Christiansen, R. and Kadiresan, K. (1989) Fertilizer Policy in Africa, MADIA Discussion Papers No. 5, Washington DC: World Bank. Leo, C. (1984) Land and Class in Kenya, Toronto: University of Toronto Press. Matthews, R.G. (1985) ‘Machinery manufacture in the formal sector’, IDS Working Paper No. 425, Nairobi: IDS.

198

FOREIGN TRADE REFORMS AND DEVELOPMENT STRATEGY

Mosley, P. (1986) ‘Agricultural performance in Kenya since 1970: Has the World Bank got it right?’ Development and Change, 17:3. Ngeno, N. (1987) ‘Kenya’s Industrial Exports: market conditions and domestic policies’, in Kenya Economic Association, Kenya’s Industrial and Agricultural Strategy towards the Year 2000, Nairobi: Kenya Economic Association. Ocampo, J.A. (1987) ‘The macroeconomic effects of import controls: a Keynesian analysis’, Journal of Development Economics, 27:307–38. Sharpley, J. and Lewis, S. (1990) ‘Kenya: the manufacturing sector to the mid-1980s’, in R.Riddell (ed.) Manufacturing Africa: Performance and Prospects of Seven Countries in Sub-Saharan Africa, London: Janes Currey & Heinemann. Stewart, F. (1984) ‘Recent theories of international trade: some implications for the South’, in H.Kierkowski (ed.) Monopoly Competition and International Trade, Oxford: Oxford University Press. Swainson, N. (1980) The Development of Corporate Capitalism in Kenya, London: Heinemann. Taylor, L. (1988) Varieties of Stabilization Experiences, Oxford: Clarendon Press, p. 30 ff. UNIDO (1984) Kenya, Industrial Development Review Series (UNIDO/ IS.459), Vienna: UNIDO. Van der Hoeven, R. and Vandermoortele, J. (1987) ‘Kenya’, Country Study No. 4, Stabilization and Adjustment Policies and Programmes, Helsinki: WIDER. World Bank (1987) Kenya: Industrial Sector Policies for Investment and Export Growth, Report No. 6711-KE, May, Washington DC: World Bank.

11 Industrialization strategies, foreign trade regimes and structural change in Turkey 1980–8 Nurhan Yentürk-Coban

Between 1963 and 1977 (with the exception of a short period of liberalization), governments in Turkey adopted import-substitution industrialization strategies (ISI); due to scarcity of foreign currency, these were abandoned in 1978–9. The year 1980 marks the beginning of a new period, characterized by a consensus on the necessity to develop the economy’s export capacity and loosen the foreign exchange constraints, although not on the policy measures required. On 24 January 1980 a liberalization policy was officially adopted, but the policies actually implemented since then have tended to be more protectionist than liberal. Short-term measures allowed exports to increase, but did not bring about a structural change in industry, although this would have been necessary to cope with international competition. This chapter analyses the reasons why such a structural change has not occurred. The first section analyses briefly the concepts of import-substituting and export-led industrialization, and expounds policy requirements of both strategies. This is followed by an analysis of the measures adopted during the 1963–79 ISI period and presents the situation prevailing on the eve of the adoption of export-led industrialization policies. It focuses on two aspects: the efficiency of ISI, and the influence of ISI on exports and imports level and structure. Next, we discuss the economic policies adopted since 1980, during what was called the ‘period of economic liberalization’. Different measures are presented and their short-term and long-term impact on the structure of foreign trade is assessed, stressing the relationship between exports, manufacturing value-added, utilization of excess capacity and export subsidies. Finally, we consider the structural weaknesses of the Turkish economy and possible measures to increase international competitiveness. THE CONCEPTS OF IMPORT-SUBSTITUTING AND OF EXPORT-LED INDUSTRIALIZATIONS With the exception of a short period of liberalization, policy in Turkey between the Second World War and the late 1970s has been oriented towards ISI. From 1980 onwards, a strategy of so-called ‘economic liberalization’ was adopted. Before analysing these two periods of industrialization in Turkey, three comments must be made concerning ISI and exportled industrialization strategies. 1 ISI should not be confused with an interventionist policy, and neither should export-led industrialization be considered as a liberal one. Export-promotion can be brought about by specific incentives, resulting from state intervention. We should recognize the need for LDCs to resort to protectionist measures before confronting competition on the world market.

200

STRUCTURAL CHANGE IN TURKEY: 1980–8

ISI and export promotion, however, require different sets of protectionist measures. ISI protectionist policies (such as high tariffs compensating for high costs, global protection and overvalued rates of exchange) are inappropriate for export promotion, which should prepare industry for international competition. This can justify an ‘educative’ protectionist policy, limited, sectorial and temporary. But policies required to open up an economy and promote exports do not systematically amount to ‘laissezfaire.’ 2 This does not mean that export-promotion policies are justified by ISI’s ‘weaknesses’. Most of today’s manufactured exporting LDCs have in effect adopted ISI policies at some early stage of their industrialization. Without them, it would have been extremely difficult for these LDCs to grow out of the first stages of industrialization and accumulate technological know-how. These two strategies should not then be considered antinomic, but rather seen as two complementary stages, one leading to the other. Transition from ISI to export-led industrialization results as much from domestic constraints as from modifications in the world economy, which command the international division of labour. 3 Both external and internal dynamics determine the integration of underdeveloped countries into the world economy. We should then put the transition from ISI to export-orientation in a historical perspective and relate it to the developments of the world economy. Various international outlooks have allowed various industrialization strategies, and ignoring them would result in a partial and incomplete analysis. IMPORT-SUBSTITUTION INDUSTRIALIZATION IN TURKEY: THE PLANNING STAGE 1963–79 From 1963 to 1979 the Turkish economy was regulated by a planning process directed towards economic development. The strategy was to bring about a new balance between public and private sectors and to accelerate the latter’s development. Public enterprises reduced their involvement in the production of consumer goods, allowing private capital to develop in that sector, and concentrated in the production of intermediate goods. This in turn allowed them to supply the private sector with cheap intermediate goods, which was expected to encourage private investment. Like other LDCs, Turkey adopted ISI policies to develop domestic production of previously imported manufactured goods. In order to enhance the production of consumer goods by the private sector a number of classical protectionist measures were adopted: high import taxes, overvalued exchange rate, restrictions on imports of consumer goods. Production of consumer durables was especially responsive to high tariff protection. This highly protected sector developed without taking either quality or price criteria into consideration, a feature which allowed industrial manufacturers to accumulate capital rapidly. During the ISI period, the growth of the consumer goods industry led to an improvement in labour qualifications and in technological know-how. This was particularly true in the consumer durable goods industry, and in the upstream sectors, that is, in production of equipment necessary for making refrigerators, automobiles and washing-machines. During this period, industry relied heavily on imports of capital and intermediate goods, as shown by the high correlation between the growth of industrial production and the availability of these imports (see econometric analysis in Appendix 11.1). The rate of exchange was consequently overvalued to push domestic production costs down. Overvaluation was actually so high that imports turned insensitive to devaluation. Prices of imports were so low that imports did not fall when the currency was devalued (see regressions presented in Appendix 11.2).

276

201

Structure and composition of imports During the ISI period the share of imported intermediary (MI) and capital and equipment goods (ME) in total imports rose, while that of consumer goods (MC) decreased. On the foreign trade level, the objective of ISI policies was to reduce the amount of imports in relative terms, but Table 11.1 shows that the ratio of imports compared to GNP progressed from 9.4 per cent to 12.0 per cent. Table 11.1 also indicates the composition of imports, which were almost totally made up of capital and intermediate goods. Export structure and foreign trade deficit The overvaluation of domestic currency penalized exports during the ISI period and thus accentuated the trade deficit. During this period, Turkey’s exports were essentially agricultural, and persistent overvaluation of the exchange rate operated as an instrument transferring resources from the agricultural to the industrial sector. During the ISI period, the ratio of exports (E) to imports fell from 0.53 to 0.30. The contribution of exports to GNP also dropped (Table 11.2). Table 11.2 shows that, during the 1963–77 period, both the ratio of exports to imports and to GNP decreased. During the ISI period, Table 11.1 Composition of imports and proportion of imports (M) to GNP* Years

ME

MI

MC

%ME

%MI

%MC

M/GNP

1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977

315 245 241 341 324 367 338 417 497 751 967 1248 1961 2239 2255

336 266 306 341 328 361 376 431 547 697 986 2320 2574 2733 3363

37 26 25 35 34 36 33 37 44 60 84 152 203 156 178

0.46 0.46 0.42 0.47 0.47 0.48 0.42 0.44 0.46 0.48 0.46 0.33 0.41 0.44 0.40

0.49 0.49 0.53 0.47 0.48 0.47 0.47 0.45 0.51 0.45 0.47 0.61 0.53 0.53 0.58

0.05 0.05 0.04 0.06 0.05 0.05 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.03 0.03

0.094 0.068 0.068 0.072 0.062 0.060 0.058 0.070 0.085 0.093 0.098 0.124 0.130 0.123 0.120

Source: ISE, Annual Statistical Bulletin 1970–88. All import figures in million US$. Note: *Data concerning 1978 and 1979 have been omitted because they correspond to years of crisis. Table 11.2 Ratio of exports to imports and GNP, and trade deficit 1963

E/M

E/GNP

E-M

E-M/GNP

0.53

0.050

−320

0.044

202

STRUCTURAL CHANGE IN TURKEY: 1980–8

E/M

E/GNP

E-M

1964 0.77 0.052 −126 1965 0.81 0.055 −108 1966 0.68 0.049 −228 1967 0.76 0.047 −162 1968 0.65 0.039 −268 1969 0.67 0.039 −264 1970 0.62 0.043 −360 1971 0.63 0.053 −394 1972 0.57 0.053 −678 1973 0.63 0.062 −769 1974 0.41 0.050 −2246 1975 0.30 0.038 −3338 1976 0.38 0.047 −3169 1977 0.30 0.036 −4043 Source: ISE, Annual Statistical Bulletin, 1970–88. E-M in million US$.

E-M/GNP 0.016 0.013 0.023 0.015 0.021 0.019 0.026 0.031 0.040 0.036 0.074 0.091 0.076 0.084

the trade deficit expressed in terms of GNP widened from 0.04 to 0.09. Which means that, in 1977, Turkey’s external financial requirements amounted to 9 per cent of its GNP. The deficit was covered by funds provided by the United States and multilateral organizations, by external debt and by immigrant workers’ transfers from abroad. The second half of the 1960s and the early 1970s witnessed the end of Keynesianism on the international level, the breakdown of the international monetary system, and the end of the United States’ economic hegemony. These abrupt changes and the ensuing tensions in the world economy, followed by the two oil shocks, forced LDCs to increase their exports and to open up their economy. However, up to 1977, Turkey lived as if these shocks had not taken place. Currency remittances of migrant workers, by providing foreign exchange, allowed the postponement of decisions on export- policy. Table 11.3 shows that this source of hard currency represented 40 to 120 per cent of the foreign trade deficit between 1965 and 1975. Up to the foreign exchange crisis of 1978–9, the Turkish economy maintained ISI policies, compressing consumer goods’ imports to a strict minimum and allocating foreign currency revenues Table 11.3 Immigrant workers’ remittances (WRT): value (million US$) and percentage of trade deficit 1964 1965 1966 1967 1968 1969 1970 1971

WRT

WRT/E-M

9 70 115 93 107 141 273 471

0.07 0.65 0.50 0.57 0.40 0.53 0.76 1.20

276

1972 1973 1974 1975 1976 1977 Source: Senesen et al. 1986.

WRT

WRT/E-M

740 1183 1426 1312 983 982

1.09 1.54 0.64 0.39 0.31 0.24

203

to the purchase of capital and intermediary goods. During the whole of the ISI period, one of the state’s most important functions was to supply foreign currency for essential imports. During this period, manufacturers benefited in effect from rents arising from availability of hard currency at an overvalued exchange rate, while domestic demand was kept increasing through an expansionary wage policy. So we can easily understand that the ISI period was a time of social peace: industrialists could obtain their imported inputs without any major difficulty and were heavily protected, while domestic demand was vigorous enough to absorb relatively expensive goods. As long as the foreign currency influx allowed the import of inputs required by the growth of industry, ISI policies were considered both by bureaucrats and industrialists as a permanent strategy. The economic structure (characterized by its high dependency on technological imports, high production costs compared with industrialized countries, and disincentives to export arising from an overvalued exchange rate) was seen as indefinitely self-perpetuating. However, this situation proved to be unsustainable: from 1975, foreign currency reserves dropped drastically, leading the Turkish economy into a dead-end. ISI policies were reconsidered and gradually abandoned from 1978–9 onward. EXPORT-INDUCED INDUSTRIALIZATION IN TURKEY 1980–7 Towards the end of the 1970s, when foreign currency shortages prompted the curtailment of ISI policies, export stimulation was generally seen as the only possibility for Turkey to restore its trade balance. The main problem was to devise appropriate means to drive Turkish industry from total protectionism to international competitiveness. After two years of economic crisis (1978–9), the government announced on 24 January 1980 its ‘economic liberalization’ policy. In spite of its apparent resolution, the government was conscious that Turkish industry could not immediately confront international markets. The policies that were implemented at the time clearly reveal the hidden facet of liberalism: relatively high customs barriers, export subsidies, and severe limitation on workers’ strikes: the so-called ‘liberal’ economic stabilization programme contained a fair number of measures that are the opposite of liberal tenets. This section analyses the measures implemented from 1980 onwards to open the economy, and focuses on the inability of the Turkish economy to undergo structural change even though its exports have increased considerably.

204

STRUCTURAL CHANGE IN TURKEY: 1980–8

Measures adopted to open the economy Even though protective tariffs were reduced after 1980 (Table 11.4), the protection to industry still remained significant. Most of the protected industrial activities are in sectors that would otherwise either collapse or go bankrupt if they were to confront international competition directly.’ The compression of domestic demand was one of the main motives for turning to international markets. After 1980, employees and civil servants lost about 25 per cent of their purchasing power. On a 1976 =100 basis, the average salary stood at 60.5 in 1980, at 55.0 in 1982 and at 46.3 in 1985 (Kepenek 1986). After the decline in domestic demand, subsidies were the second most important incentive to exports. Direct subsidies allowed industrialists to export at prices below cost. However, industry did not undergo any noticeable structural change. Cost-reducing efforts— through rationalization of production, technological innovation or quality increases—were made unnecessary by export subsidies. Devaluation of the Turkish pound is one of the few measures in line with liberal philosophy actually implemented. As from 24 January 1980, after a number of drastic devaluations (Table 11.5), the official and black market exchange rates were finally brought into line. Between 1980 and 1988, the Turkish pound depreciated, from 77.54 to 1,200 Turkish pounds per dollar. The overvaluation policy penalizing exports was abandoned and, from 1980 onward, devaluation allowed a strong expansion of exports (see Econometric Analysis, Appendix 11.2). The sluggishness of structural change in Turkish industry since 1980 The measures adopted during this period constituted a turning-point for exports (Table 11.6). Both volume and structure of exports were Table 11.4 Protective tariffs 1980 1981 1982 1983 1984 1985 1986 Source: Togan et al. (1988).

0.763 0.763 0.763 0.763 0.489 0.489 0.489

Table 11.5 Evolution of the exchange rate Years

T.Pound/$

1977 1978 1979 1980 1981 1982 1983 1984

18.09 24.63 35.21 77.54 112.20 165.07 230.33 375.10

276

Years

T.Pound/$

1985 1986 1987 Source: ISE, Annual Statistical Bulletin, 1977–88.

528.48 660.88 1200.00

205

profoundly modified. We note, in column 3 of Table 11.6, that manufactured goods represent three-quarters of total exports. One important feature of the economic evolution appears in Table 11.7, which shows the ratio of total exports (E), manufactured exports (EM) and manufacturing value-added (YM) to GNP. It is immediately clear that the increase in manufactured goods’ exports was achieved without an equivalent increase of the manufacturing Table 11.6 Evolution of exports (E) and share of exported manufactured goods (EM) in total exports E 1977 1753 1978 2288 1979 2261 1980 2910 1981 4703 1982 5746 1983 5728 1984 7134 1985 7958 1986 7454 1987 10190 Source: ISE, Annual Statistical Bulletin, 1977–88.

EM

EM/E

586 621 785 1047 2290 3429 3658 5145 5995 5324 8065

0.33 0.27 0.35 0.36 0.49 0.60 0.64 0.72 0.75 0.72 0.79

Table 11.7 The ratio of total exports (E), manufactured exports (EM) and manufacturing value-added (YM) to GNP 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987

E/GNP

EM/GNP

YM/GNP

0.036 0.044 0.036 0.053 0.081 0.108 0.114 0.146 0.152 0.125 0.219

0.012 0.012 0.012 0.019 0.039 0.065 0.079 0.105 0.114 0.089 0.173

0.17 0.18 0.17 0.16 0.17 0.17 0.18 0.18 0.19 0.19 0.19

206

STRUCTURAL CHANGE IN TURKEY: 1980–8

E/GNP

EM/GNP

YM/GNP

Source: ISE, Annual Statistical Bulletin, 1977–88.

industry’s contribution to GNP. We note the following features of the 1980–8 period in Tables 11.6 and 11.7: (a) an increase of total exports in GNP, (b) an increase of exports of manufactured goods (EM) compared with total exports (E) and GNP, (c) a stabilization of manufacturing value-added’s (YM) contribution to GNP. These three features indicate a lack of deep evolution in the Turkish economy, in spite of increased exports. One point explaining this lack of dynamism is the nature of export incentives adopted, especially subsidies, which produced only short-term effects. Subsidies are not directed at manufacturing firms, where they could promote technological change, increases in productivity and a reduction in costs but are paid out to commercial firms. Furthermore, the Turkish government has sponsored concentration among firms dealing in foreign trade. About ten firms presently share Turkey’s foreign trade, and actually collect export subsidies. This has triggered off a heated debate between industrialists, who produce, and commercial firms, who actually export. Closer examination of the structure of exports and imports (Table 11.8) reveals another reason why economic structure was not altered by the development of manufactured goods exports. Between 1980 and 1988, the ratio of exports to imports progressed from 37 to 71 per cent. However, the structure of imports constantly favoured imports of intermediate goods at the expense of capital goods. Table 11.8 Value of exports (E), imported capital goods (ME) and intermediate goods (MI), and export/import ratio (E/ M) E

ME

1977 1753 2255 1978 2288 1590 1979 2261 1597 1980 2910 1581 1981 4703 2207 1982 5746 2324 1983 5728 2317 1984 7134 2659 1985 7958 2603 1986 7454 3474 1987 10190 3817 Source: ISE, Annual Statistical Bulletin, 1977–88.

MI

E/M

3363 2877 3377 6158 6547 6332 6675 7624 7836 6673 9180

0.30 0.50 0.45 0.37 0.53 0.65 0.62 0.66 0.67 0.67 0.71

There is a strong statistical correlation between the increase in exports and that of imported intermediate goods, which suggests that the increase of exports largely depended on the utilization of excess capacities (see econometric analysis in Appendix 11.3). This hypothesis is equally confirmed by the analysis of imports’ structure (Table 11.9).

276

207

During the ISI years, capital and intermediate goods were imported in roughly the same proportion (50: 50). After the crisis years, when the exchange rate started to depreciate, the share of imported capital goods fell to 20 per cent following a drop in new investment. This illustrates one pitfall of the export-promotion policies pursued: the good export-performance of the 1980s was achieved through higher utilization of productive capacity, but did not stimulate investment. Without new investment, technological improvement could not take place or structural change in the economy occur.2 It is also clear from Table 11.9 that the share of intermediate goods in total imports increased during the export-expansion period (1980– 8). When all excess capacity has been mobilized, the increase in Table 11.9 Capital goods (ME) and intermediate imports (MI) as a proportion of total imports (M) ME/M

MI/M

1970 0.44 1971 0.46 1972 0.48 1973 0.46 1974 0.33 1975 0.41 1976 0.44 1977 0.40 1978 0.35 1979 0.31 1980 0.20 1981 0.25 1982 0.26 1983 0.25 1984 0.25 1985 0.23 1986 0.31 1987 0.27 Source: ISE, Annual Statistical Bulletin, 1977–88.

0.45 0.51 0.45 0.47 0.61 0.53 0.53 0.58 0.62 0.67 0.78 0.73 0.71 0.72 0.71 0.69 0.60 0.65

exports will require an increase in imports of capital goods as well as intermediate inputs. According to Istanbul’s Chamber of Commerce, the present rate of utilization of productive capacities stands around 85 per cent, which can be considered as full employment in LDCs. If we assume a downward rigidity of domestic demand, which is at near incompressible level, and if no specific measures are taken to promote structural changes, total imports, both of capital goods and of industrial inputs, can be expected to increase pari passu with exports from now on. CONCLUSION Under ISI, a private manufacturing sector developed in the durable consumer goods sector. This produced a qualified labour force and technological know-how, especially in the assembly-line type of production.

208

STRUCTURAL CHANGE IN TURKEY: 1980–8

Brought about by import restrictions, absolute protection favoured the creation of an industry that operated without any consideration for quality or costs. The overvaluation of the exchange rate penalized exports, while domestic prices of imported intermediate and capital goods were kept down artificially. Consequently, import-substitution did not extend to capital and intermediate goods, which in turn meant that manufacturing production remained largely dependent on imported technology and inputs, and Turkey’s trade balance was constantly in deficit. As long as expatriate workers’ remittances ensured a regular inflow of foreign exchange, ISI was considered as a viable permanent strategy. The Turkish economy was thus able to perpetuate this structure up to 1978–9, when forex shortages became acute. Increased exports were then seen as the only solution. In order to strengthen the export potential and prepare Turkish industry to face international competition, various policy options could have been contemplated. A ‘liberal’ stabilization programme was actually implemented by the Turkish government between 1980 and 1988. Export promotion policies, brought about by domestic demand compression, devaluation and direct subsidies, did increase export volume but failed to alter the economic structure of the country. Three reasons can account for this failure. First, from 1980 onward, although both total exports and the share of manufactured goods in total exports increased over this period, the manufacturing industry’s growth rate remained unchanged. There was no correlation between the growth rates of the manufacturing sector and of manufactured exports. In fact, increased exports largely resulted from domestic demand compression, so that neither the structure nor the export-capacity of Turkish industry was modified. Second, the alignment of prices of exported goods on international prices was not brought about by a reduction in domestic production costs but by export-subsidies and devaluation alone. Domestic production costs thus remained above prices on international markets, because productivity remained low, production did not reach its optimal size, and new technologies were not introduced. Export subsidies had only short-term effects because they were directed to trading firms rather than to productive enterprises. They carried no incentive in favour of structural change, but artificially reduced export prices. Finally, the manufacturing sector remained very dependent on imported inputs. The post-1980 growth of exports resulted from an increased utilization of production capacities, which in turn increased imports of intermediate goods. The improvement of the export/import ratio did not only result from the increase in exports but equally, and more worryingly, from a slow-down in capital goods imports. This resulted from a lack of new investments, which could have contributed to a modification of the economy’s structure. With productive capacity now almost fully employed and domestic demand close to its incompressible level, we can expect the continuation of the same export-promotion policy to increase imports of capital as well as of intermediate goods. Any long-term solution that would make structural change possible would have to determine which industries play a central role as suppliers of inputs to exporting industries and then attempt at establishing them in Turkey. These would obviously be sectors producing capital and intermediate goods. Their creation in the presence of international competition entails a certain amount of official support. Some form of educative protectionism, both sectorial and reversible, could prepare infant industry to face international competition while taking cost and productivity criteria into account.

276

209

APPENDIX 11.1 Assessment of the variables influencing GNP The variables used are: YF=GNP expressed in Turkish pounds, in constant prices; YF1=GNP expressed in Turkish pounds, lagged one year; MIEF1=imports of capital and intermediate goods, expressed in constant dollar prices, deflated by the capital and intermediate goods’ imports deflator; WR1=foreign currency remittances by immigrant workers, lagged one year. 1971–9:9 observations (11.1) The first regression’s results are fairly good, with R2=0.99. The variable MIEF1 indicates that the imports of capital and intermediate goods of the previous year are statistically significant in explaining the GNP’s increase, as are also YF1 and WR1. 1980–7:8 observations (Eq. 11.2) Where MIF1=imports of intermediate goods lagged one year; EF=exports. For the period corresponding to export-led industrialization, equation 11.2, whose variables are exports and imports of intermediate goods, gives the best results with R2=0.971. APPENDIX 11.2 Assessment of the exchange rate’s influence on imports The variables used are: MF=total imports in constant dollar prices; ERI=imports’ exchange rate. 1970–87:10 observations (Eq. 11.3)

1970–7:8 observations (Eq. 11.4) 1980–7:8 observations

210

STRUCTURAL CHANGE IN TURKEY: 1980–8

(Eq. 11.5) The three regression equations for the 1970–87, 1970–7 and 1980–7 periods show that the exchange rate’s effect on total imports is effectively positive. This indicates that the Turkish pound’s devaluation does not reduce the amount of total imports. APPENDIX 11.3 Assessment of the variables influencing exports The variables used are: EF=exports in constant dollar prices, deflated by the exports’ price index; MIF=imports of intermediate goods in constant dollar prices, deflated by the intermediate goods’ deflator; ERE=exports’ exchange rate (Turkish pounds per dollar); DD=domestic demand in constant Turkish pounds; TRMF=export subsidies in constant Turkish pounds. 1970–87:18 observations (Eq. 11.6) Equation 11.6 indicates that, since the early 1970s, the increase in exports has been influenced by the exchange rate’s devaluation and by export subsidies, and by the domestic demand’s slowdown. 1980–7:8 observations (Eq. 11.7) The second regression covering the 1980–7 period produces statistically significant results with R2=0.995. Domestic demand’s effect has a negative value, which shows that domestic contraction increased exports. The exchange rate’s influence on exports is equally obvious. In equation 11.7 we have also added the imports of intermediate goods as a restricting variable indicating the utilization of productive capacities and exerting a positive effect on exports subsequent to 1980. NOTES 1 On the subject of tariff protection of the various sectors, see Togan et al. (1988). 2 For a more detailed analysis of this hypothesis in the clothes industry, which is the most important component of exports, see Duruiz and Yentürk-Coban (1988).

REFERENCES Duruiz, L. and Yentürk-Coban, N. (1988) Technological and Structural Change in the Turkish Clothing Industry, Ankara: Social Science Association.

276

211

Insel, A. (1987)‘Liberalisation autoritaire: double échec en Turquie, Annales du Levant, 2, Paris. ISE (1977–88) Annual Statistical Bulletin, Ankara: ISE. Kepenek, Y. (1986) Cumhuriyet, 24 January 1986. Senesen, U. (ed.) (1986) Türkiye Ekonomisi için bir Ekonometrik Model Denemesi (An Economic Model for the Turkish Economy), Istanbul: Chamber of Commerce. Togan, S. (ed.) (1988) Eternal Economic Relations of Turkey, Istanbul: Foreign Trade Association of Turkey.

12 Exchange rates and subsidies in an export-promotion policy: the case of South Korea Mario Lanzarotti

THE CONTRACTIONARY AND INFLATIONARY EFFECTS OF DEVALUATIONS Several theoretical and empirical studies oppose the use of systematic devaluation as a means to maintain exports’ competitiveness. Devaluations can result in contractionary economic effects. One standard view is that a devaluation will depress the trade balance if the Marshall-Lerner condition is not fulfilled. Hirschman (1949) showed that, if it occurs in the situation of an initial trade deficit, devaluation can have depressive effects even though the sum of imports’ and exports’ price-elasticities is superior to unity. DiazAlejandro (1963) suggested that similar consequences may appear if devaluation provokes a regressive distribution of income, that is, by benefiting those who have the highest marginal savings-propensity. Krugman and Taylor (1978) identified a third depressive effect, which appears when devaluation redistributes income in favour of the state, which—in the very short term—is supposed to have a very high propensity to save. The Diaz-Alejandro effect probably does not apply to South Korea. Available estimates for this country invalidate the idea of a higher savings-propensity among higher income social classes (Lee 1985; Kim and Park 1977). Similarly, the absence of export taxes and the frequent tariff exemptions mean the KrugmanTaylor effect is relatively improbable. In return, even though estimates of import and export price-elasticities show their sum to be relatively high, the Hirschman effect is likely to occur, because devaluations took place only after the trade balance had been in deficit for a long while. On the other hand, devaluations may have an inflationary effect when imports are non-competitive and mainly composed of intermediate and capital goods. This applies more readily to the Korean case where imports of consumer goods are small and imports of inputs are fairly insensitive to exchange-rate variations (Koo 1977). This most certainly explains the Korean authorities’ reluctance to resort to devaluation. Indeed, by increasing the cost of inputs, devaluations may have inflationary effects if prices are submitted to markup procedures, which seems to be the case in Korea, according to price equations proposed by Van Wijnbergen (1982). This is probably a common feature of LDCs: in a study concerning more than fifteen LDCs, Taylor (1988) found this to be true in almost all cases. The high level of external indebtedness also made devaluation a delicate matter in South Korea. The immediate effect of any devaluation is to increase the local currency value of outstanding debt, which could therefore endanger local firms’ standing. This was the case after the maxi-devaluation of 1971, which had to be followed by a general reduction of rates of interest and the renegotiation of firms’ domestic debt on favourable terms.

292

213

Figure 12.1 Nominal rate of exchange (Won/US$)

Figure 12.2 Real exports and real effective rate of exchange (RERE) (Millions of 1965 US$ and Won/US$)

NOMINAL AND REAL EFFECTIVE EXCHANGE RATES The evolution of the exchange rate illustrates the South Korean aversion to devaluation. A unified exchange rate was adopted in 1964. Up to the 1980s, its evolution in nominal terms was characterized by a few maxidevaluations, followed by periods of relative stability. This gives a curve in the form of steps (Figure 12.1). Depending on the period considered, the phases of stability are explained by the use of fixed parities or by the Bank of Korea’s interventions in a floating exchange-rate system. Three of the four maxi-devaluations (1964, 1971, 1974, 1980) were imposed by largely exogenous factors. The first preceded the definitive abrogation of the multiple exchange rate system that characterized the import-substitution period, while those of 1974 and 1980 were the consequences of the two oil shocks. As for the 1971 maxi-devaluation, it was decided after a phase of spontaneous devaluation in a flexible exchange rate context and was part of a stabilization programme signed with the IMF. From June 1972 onward, South Korea adopted a fixed-parity system. On the contrary, the evolution of the real effective exchange rate does not show stability but appreciation between maxi-devaluations. In Figure 12.2 we can see that, during the phases of over-valuation, real exports continued on their upward trend. This is a first indication of the role played by subsidies.

214

AN EXPORT PROMOTION POLICY: SOUTH KOREA

THE ROLE OF SUBSIDIES Two aspects will be examined. The first is the quantitative importance of subsidies; the second is their real impact on the evolution of exports. Subsidies are used here as a generic name to cover all measures adopted to promote exports. They include direct subsidies, tariff and fiscal exonerations and reductions, concessional interest rates, preferential exchange rates, and preferential relaxation of import controls for exporters. This last element means that measures adopted in an import-substitution perspective are, in fact, used to promote exports; it encompasses various measures offering exporters the possibility to import, and to sell locally, goods that are otherwise subject to import restrictions or totally banned. These are generally omitted in attempts to quantify subsidies, because of the difficulties raised by their valuation. However, certain fragmentary assessments suggest they were relatively important during the 1960s (Koo 1977). A second and more important factor is equally omitted from the calculations of subsidies, that which pertains to preferential rates of interest. If we estimate the interest rates’ subsidy on the basis of the official banking system’s normal interest rates, an important aspect is overlooked, namely the privilege that automatic access to formal-sector funding constitutes in a context of rationed credit facilities (Westphal and Kim 1977). Given its function as an alternative source of funding, it would therefore seem more logical to use the interest rates of the parallel financial market as a reference. This is what we have done for the period 1963–75.1 This is justified by the fact that, when they obtained fundings superior to their own needs (a fairly frequent situation), exporters invested the sums in excess over their needs in the informal financial markets (Cole and Park 1983). By doing so, they made a profit equal to the difference between the interest rate applied to export credits and those prevailing on the parallel market. We used this differential here in order to measure the subsidization through rates of interest. Table 12.1 presents a comparison between our own results2 and those obtained in other studies. Each statistical series gives the percentage corresponding to subsidies in local currency (per export dollar) compared to the nominal exchange rate (expressed in wons per dollar). If we use the informal market’s interest rates as a reference, total subsidies (over the 1963–75 period) represent on average a 35 per cent increase over the nominal exchange rate, which is still an underestimate. We must also note that the series of Westphal and Kim, and of Wang and Suh are fairly close to one another as from the mid-1960s, and inferior to Koo’s. The difference, quite important for some of the years considered, gives a partial idea of the subsidy content of the relaxation of import control for exporters. Finally, all series show that, as from 1972, subsidies clearly tend to dwindle. In order to measure the impact of subsidies on exports, we used the following model: Table 12.1 Evolution of subsidies per export dollar (as a percentage of the nominal exchange rate) Years

I

II

III

IV

V

1963 1964 1965 1966 1967 1968 1969 1970

58.7 49.1 23.4 27.8 35.0 37.0 36.0 37.9

44.8 29.2 12.0 16.9 22.7 29.1 27.7 29.9

45.7 21.3 14.7 19.0 23.1 28.1 27.8 28.4

11.2 10.6 15.1 18.6 23.0 28.8 28.7 29.3

– – 17.8 20.5 28.6 31.0 30.3 33.8

292

Years

I

II

III

IV

1971 37.2 31.9 29.6 30.7 1972 31.8 28.1 26.9 26.8 1973 27.9 24.1 23.7 23.4 1974 28.0 21.6 21.2 21.2 1975 31.4 15.5 16.7 15.7 1976 – 19.6 – – 1977 – 21.0 – – 1978 – 22.3 – – I Our own calculations II Wang (1983) III Westphal and Kim (1977) IV Suh (1977) V Koo (1977); these figures include an estimate of the advantages proceeding from one of the imports’ control relaxing systems (wastage allowance).

215

V 32.4 – – – – – – –

where EXP=volume of exports, TER=real effective exchange rate in wons per dollar, SUB=total subsidies expressed in wons. Two regressions were run. The first covered the 1963–75 period with the subsidy effect of concessional interest rates calculated according to the criterion proposed earlier in this paper. The second, covering years 1962–78, relied on the data supplied by Wang (1983), who estimated the discounted interest rates according to the normal rates applied by the official banking system. The results obtained in both cases are comparable (Table 12.2). The elasticities of exports with respect to the real effective exchange rate and to subsidies are positive and statistically significant. Secondly, the β coefficients, calculated on the basis of normalized variables ([variablemean of variable]/standard deviation) in order Table 12.2 Estimate of exports’ elasticities α1 (t) α2 (t) α3 (t) β1 β2 R2 R2cor DW

1963–75

1962–78

−10.3 (−5.1) 0.8 (1.9) 0.7 (20.2) 0.1 0.9 0.99 0.99 1.63

−6.8 (−3.7) 1.1 (3.2) 0.6 (33.3) 0.1 0.9 0.99 0.99 1.57

216

AN EXPORT PROMOTION POLICY: SOUTH KOREA

Figure 12.3 Real subsidies and real effective rate of exchange (RERE) (Won/US$, 1963–75) (Own calculations)

to allow comparison of elasticities, indicate that exports are more elastic to subsidies than to the real effective exchange rate. Consequently, in South Korea, subsidies seem to have constituted an effective instrument for promoting exports, at least as far as their volume is concerned. THE RATIONALITY OF THE SUBSIDIZED EXPORTS POLICY It is now necessary to take a closer look at the criteria determining the quantitative importance and the qualitative profile of subsidies. As far as the quantitative aspect is concerned, an essential criterion for attributing subsidies seems to have been that of compensating the real effective exchange rate’s fluctuations, in order to smooth its evolution. This can be shown by comparing the evolution of subsidies per export dollar in real terms with that of the real effective exchange rate. Here again, the comparison was made both on the basis of our own calculations for the 1963–75 period (Figure 12.3), and from the data supplied by Wang (1983) for 1962–78 (Figure 12.4). These two variables appear to have, on the whole, evolved in opposite directions up to 1973. Subsidies increased when the real effective exchange rate fell, and dropped when the opposite took place. There existed, therefore, a compensation between the two. Regarding qualitative aspects, two different phases must be examined separately. In the 1960s and early 1970s, subsidies were essentially aimed at increasing exports’ gross value, with little consideration either for domestic value-added or the creation of new productive capacities. In this first phase, the main policy instruments were import tax exemptions on imported inputs destined for the production of export goods and relaxation of import controls for exporters, two measures which tend to discourage backward production linkages. In this phase, export credits were essentially short-term ones, aimed at financing production rather than creating productive capacities, and no discrimination between products was brought in. A change in perspective took place during the 1970s: henceforth, economic policy sought to achieve a stronger integration of domestic production and focused on the development of specific sectors. Under these conditions, the qualitative profile of subsidies changed. Tariff exonerations on imported inputs and also on imported capital progressively disappeared, while advantages granted to the exporting sectors were extended to their domestic suppliers; export credits discriminated among sectors and the share of long-term funding increased.

292

217

Figure 12.4 Real subsidies and real effective rate of exchange (RERE) (Won/US$, 1962–78) (Wang’s calculations)

This evolution shows quite clearly that it is possible to draw on the array of subsidies and tailor measures to specific objectives. One objective worth mentioning, and which Korean authorities were quick to pursue, is the specialization among export activities in order to separate production from international trading. This distinction is justified by the fact that both activities constitute separate professions requiring specific skills. In the 1960s, no one could engage in import-export activity without being granted an international trader’s licence, which could not be obtained or retained unless some minimal export performance was achieved. In other words, in order to be able to import (a very profitable activity), one had first to export. Traders also enjoyed fiscal and tariff exonerations and subsidized funding along with other specific advantages. They could thus be awarded a export-monopoly either for a particular commodity or on a specific market; or they could be granted the right to use part of the currency earned from export to import certain goods normally banned. The KOTRA (Korean Trade Promotion Corporation) was also founded, its mission being to promote Korean products abroad. This institution has a vast international network and is financed by a special tax of 0.55 per cent levied on all imports; it is to be noted that, once again, this framework can either be used to discourage imports or to promote exports. Traders were also ranked according to their export performances, a high rank yielding additional advantages. During the 1970s, Korean authorities decided to stimulate the creation of ten or more important international trading firms. By observing the example of the imposing Japanese Sogo Shoshas (who channelled a significant share of Korean exports), they realized that economies of scale were in fact possible when penetrating international markets, and that traders should not over-specialize but rather pursue horizontal integration. This set of measures fostered the formation of a vast international sales force,3 which is no doubt an important element of Korea’s export success. CONCLUSION One can then conclude that in South Korea subsidies have, in part, been a substitute for devaluations. Devaluations were avoided for the best because of their inflationary (and on occasion, stagflationary) and financial impacts on an economy characterized by the rigidity of input imports, a permanent trade deficit and a high level of external indebtedness. Subsidies also constituted an efficient means of promoting exports, first, because of their quantitative impact on the volume of exports and, second, owing to the

218

AN EXPORT PROMOTION POLICY: SOUTH KOREA

selectivity they allowed. Subsidies did indeed allow the profile of exports to be modulated by introducing deliberate sectorial discrimination, something that devaluation of a unique exchange rate cannot obtain. Subsidies, however, present certain dangers. There is no guarantee a priori that the administrative machinery will take the best decisions; it is also necessary to avoid the transformation of subsidies into a permanent necessity. In the South Korean case, these two stumbling-blocks were avoided. Not only did exports experience a strong development, but subsidies proved to be a reversible practice. Indeed, the quantitative importance of subsidies began to dwindle during the 1970s. This tendency continued during the 1980s, when the trade balance displayed important surpluses. The revaluations experienced by the Korean currency toward the end of the 1980s seem to bring to a close the period of subsidy concessions as an alternative to devaluations. NOTES 1 Our calculations are also an under-estimate as we have only taken short-term funding into consideration. 2 In calculating the concessional interest rates applied to export credits we took into consideration the information supplied by Wang (1983) on: direct subsidies (1963–4), the premium to the dollar sold by exporters (1963–4), fiscal and tariff rebates and exonerations (1963–75), and the implicit subsidy in the extended lead time for the payment of import taxes (1975). 3 In 1977, there were 450 international trading firms in South Korea with at least one foreign branch, and twelve of them had more than ten such branches each.

REFERENCES Cole, D. and Park, Y.C. (1983) ‘Financial development in Korea 1945– 1978’, Cambridge, Mass.: Harvard University Press. Diaz-Alejandro, C. (1963) ‘A note on the impact of devaluation and the redistributive effect’, Journal of Political Economy71 Hirschman, A. (1949) ‘Devaluation and the trade balance: a note’, Review of Economics and Statistics23. Kim, C.K. (ed). (1977) Planning Model and Macroeconomic Policy Issues, Seoul: Korea Development Institute Press. Kim, M.J. and Park, Y.C. (1977) ‘A study on the savings behavior 1953– 1972’, in C.K.Kim (ed.) (1977) Planning Model and Macroeconomic Policy Issues, Seoul: Korea Development Institute Press. Koo, B.H. (1977) Foreign Exchange Policies: an Evaluation and Proposals, in C.-K. Kim (ed.) (1977) Planning Model and Macroeconomic Policy Issues, Seoul: Korea Development Institute Press. Krugman, P. and Taylor, L. (1978) ‘Contractionary effects of devaluation’, Journal of International Economics8. Lee, M.H. (1985) ‘Determining factors of household saving in Korea’, Quarterly Economic Review, The Bank of Korea, September. Suh, S.T. (1977) ‘Growth contribution of trade and the incentive system’; in C.-K.Kim (ed.) (1977) Planning Model and Macroeconomic Policy Issues, Seoul: Korea Development Institute Press. Taylor, L. (1988) Varieties of Stabilization Experience, Oxford: Clarendon Press. UNIDO (1987) The Republic of Korea, Industrial Development Review Series, Vienna: UNIDO. Van Wijnbergen, S. (1982) ‘Stagflationary effects of monetary stabilization policies: a quantitative analysis of South Korea’, Journal of Development Economics10. Wang, Y.K. (1983) ‘Export assistance regimes in the Pacific Asian developing countries’. Paper presented at the Pacific Co-operation Task Force Workshop on Trade of Manufactured Goods, Seoul, quoted in UNIDO (1987), The Republic of Korea, Industrial Development Review Series, Vienna: UNIDO. Westphal, L. and Kim, K.S. (1977) ‘Industrial policy and development in Korea’; World Bank Staff Working Paper 263, Washington DC: World Bank.

Index

Africa, regional integration in: as protectionism, 199–210, 217–19; as trade liberalization measure, 210–17; see also individual countries agriculture: and African regional integration, 201, 206; prices in 30, 72; and trade policy reform, 56–7, 69–70 allocative efficiency, 3–4, 100 anti-dumping measures, 74 Argentina: balance of payments problems, 125–6, 127–8, 129; debt problems, 124, 130, 131; fiscal policy, 136–40, 145, 152; prospects for growth, 146; trade policy, 54, 66, 69; ‘Washington Consensus’ policies in, 121 ASEAN (Association of South-east Asian Nations), 73

prospects for growth, 146; trade policies, 49, 126–7, 151, 152 budgets, see fiscal policy Burkina Faso, consumption patterns in, 208 Cameroon, development experience of, 207 capital, see investment capital flight, 139 Central American Common Market, 73 CFA franc zone, 53, 66, 214, 215 Chile: balance of payments problems, 125–6, 127, 129; debt problems, 124, 130, 131; exchange rate policy, 63, 66; fiscal policy, 135–6, 144–5; prospects for growth, 146; structural adjustment policies, 35; trade policy, 51, 53, 63, 64, 65, 69, 71, 74, 105, 127, 152; ‘Washington Consensus’ policies in, 123 China, trade policy in, 65, 66 collusion, 105 Colombia: balance of payments problems, 124, 129; debt problems, 124; exchange rate policy, 66; fiscal policy, 132–5, 144; prospects for growth, 146; trade policy, 63, 70, 125, 152 competition: and productivity, 101, 103–6 complex development model, 207 concentration in markets, 103–4 consumption: irreversibility of patterns of, 208; and structural adjustment, 36–7 contestable markets theory, 104 Côte d’Ivoire (Ivory Coast): Costa Rica, trade policy in, 53

balance of payments problems: 29, 124–31; and stabilization policies, 148, 150; and structural adjustment, 36–9; and trade policy reform, 60, 61 Bangladesh, trade policy in, 51, 54, 69 bargaining theory, 34 Benin, trade policy in, 209, 216 Bolivia: devaluation, 63; prospects for growth, 149–50; trade policy, 51, 63, 64, 69, 71; ‘Washington Consensus’ policies in, 122, 123 Brady initiative, 123 Brazil: balance of payments problems, 125–6, 129; debt problems, 124, 131, 141; exchange rate policy, 66; fiscal policy, 136–7, 140–2, 145; 219

220

INDEX

consumption patterns, 208; exchange rate policy, 66; fiscal policy, 209, 214; manufacturing industry, 57, 58; pricing problems, 47; trade policy, 54, 69, 105–6, 214,216 credibility problems, 73–4 current account, see balance of payments debt problems, 119, 124, 130, 131–2, 150, 291 demographic problems, 48 deregulation, 119 determinants of protection (DOP) model, 174, 185–96 developing countries, trade policies of: and rise of protectionism, 166–84; see also individual countries DUPe models, 9, 79–80, 81 Dutch disease, 207 duty waivers, 68 duty-free zones, 68 dynamic efficiency, 100–3 East African Common Market, 72 ECLA, 115 Economic Community of West African States, 72 Ecuador, structural adjustment policies in, 32–3, 41 educative protectionism, 200–5 efficiency, see productivity Egypt, trade policy in, 102 European Community: 73; import penetration of, 168, 170; protectionist pressures in,167, 169, 171–2, 174, 176– 81 exchange rates: and fiscal policy, 62; multiple, 152; and stabilization policies, 149, 151; and structural adjustment policies, 30, 31, 212; and trade policy, 51, 53, 56, 60, 62, 63, 65–6, 74, 252, 261–2, 280–1, 290, 291; ‘Washington Consensus’ on, 119 export processing zones, 68 export-orientation: arguments for, 9–10, 48–9, 93,168; industrialization by, 275, 280–6; policies for, 66–70, 151, 226 external balance, see balance of payments fiscal policy: 131–45; and African regional integration, 209, 214;

and stabilization policies, 148; and trade policy reform, 60,61,62, 68,152; ‘Washington Consensus’ on, 118,120 food assistance programmes, 74 foreign exchange, see exchange rates Gambia, trade policy in, 209, 216 General Agreement on Tariffs and Trade (GATT): 65, 74; Uruguay Round of, 71–2 Ghana: agriculture in, 57; devaluation in, 53, 62, 63; fiscal policy, 60,62; manufacturing industry, 57, 58; structural adjustment policies, 32–3, 40–2, 44; trade policy, 48, 53, 69, 70, 73 government and state: control of expenditure of, 30, 118; debts of, 131; developmental role of, 27, 43, 82–5, 111, 117; (‘Washington Consensus’ critique of, 116–18); rent-seeking models of, 79, 89–90 growth, economic: 145–6; and productivity, 101; and stabilization policies, 147–50; and structural adjustment policies, 36–9; and trade policy, 55–6, 167–8; and ‘Washington Consensus’, 112–23 Guinea: devaluation in, 53; trade policy in, 69 Guyana: structural adjustment policies, 32–3, 40–3, 44; trade policy, 51 Haiti, trade policy in, 51 Hong Kong: industrialization, 44; trade policy, 65, 68 import discipline hypothesis, 104–5 import substitution: 47–8, 93, 201, 262; industrialization by, 6, 275–9, 285; ‘Washington Consensus’ critique of, 113, 114–18 incomes policy, 149 India, trade policy in, 47, 68 Indonesia, trade policy in, 68 industry, see manufacturing infant industry protection: 47, 80–1, 100,119;

INDEX

African regional integration and, 199–205 inflation: stabilization policies, 148, 149; and trade policy reform, 60, 61 informal sector, 232 institutional reform: and structural adjustment, 33; and trade policy, 54, 57 interest groups, 209, 233 interest rates: and structural adjustment policies, 30; ‘Washington Consensus’ on, 118–19 International Monetary Fund (IMF): 2; and Brady initiative, 122–3; on non-tariff barriers, 166; Polak model, 30; and stabilization policies, 29, 31, 120, 147; and trade policy reforms, 51, 93; and ‘Washington Consensus’, 111, 113 investment and capital: 133, 136, 142; and structural adjustment, 36–9, 40–3, 226; and trade policy reform, 69, 232; in ‘Washington Consensus’, 119, 122–3 Italy, industrialization in, 84 Ivory Coast, see Côte d’Ivoire Jamaica: structural adjustment policies, 31, 32–3, 40–2; trade policy reform, 62 Japan: import penetration of, 168, 170; industrialization in, 83, 84; protectionist pressures in, 169, 173 Kenya: devaluation, 261–2; manufacturing industry, 250, 255–9; pricing problems, 47; structural adjustment policies, 32–3, 40–2, 44; trade policy, 54, 62, 249–63 Korea, South: exchange rates, 291, 295, 298; industrialization, 83, 84, 114; stabilization policies, 123; trade policy, 53, 63–8 passim, 84, 85, 102, 292–8 labour markets, 69, 74, 225, 227 Laos, devaluation in, 63

221

liberalization of trade: 5–10; African regional integration and, 210–17; credibility and sustainability of, 73–4; critique of, 10–15; expected gains from, 46–9; experience and implications of, 50–9, 97–103, 226–34, 249–63, 280–6; incoherence of, and rise of protectionism, 166–84; measures for, 65–71; sequencing and timing of, 9, 59–65, 79–90, 153–4, 227–8; and stabilization policies, 60–2, 150–4; and structural adjustment, 32, 93; trading partners and, 71–3; ‘Washington Consensus’ on, 119 location of industry, 69 Madagascar: manufacturing industry, 57; trade policy, 51, 63, 68, 70, 214 Malawi: structural adjustment policies, 32–3, 37, 40–2, 44; trade policy, 54, 69 Malaysia, trade policy in, 69, 70 Mali, trade policy in, 70, 216 manufacturing industry: encouragement of development of, 44, 83–7, 97, 221– 5, 250; by export-orientation, 275, 280–6; by import-substitution, 6, 275–9, 285; by infant industry protection, 47, 80–1,100,119, 199– 205; productivity in, 98–106; trade policy reforms and, 57, 58, 68–9, 98–103, 226–8, 231–4, 255–9 market(s): competition in, 103–4; imperfections in 3–4; overregulation of, 116–17; structural adjustment policies 30 marketing boards, 70 Mauritius: structural adjustment policies, 35; trade policy reform, 62,68,216 mercurial values, 226, 229, 230 Mexico: balance of payments problems, 125–6, 127–8, 129; debt problems, 124, 130; devaluation, 63;

222

INDEX

fiscal policy, 62, 136–7, 142–4, 145; prospects for growth, 146, 149–50; trade policy, 51, 53, 63, 64, 65, 68, 70, 71, 74; transport problems, 48; ‘Washington Consensus’ policies, 122 migration, 48, 57 mining, 69, 127, 136, 200 monopoly, 3–4, 104 Morocco: fiscal policy, 62; trade policy, 54, 68, 70 Mozambique, trade policy in, 64 Nigeria: agriculture in, 57; devaluation in, 63; fiscal policy, 60, 62, 65; imports of, 238–40, 245–7; manufacturing industry, 57, 58, 65, 215, 216; trade policy, 64–5, 70, 73, 214, 240–2, 247–8; pricing problems, 47 non-tariff barriers: 46, 57, 70, 72, 153, 166, 169, 171; see also quotas oil crisis, 29 oligopoly, 104 Olivera-Tanzi effect, 139, 140 ‘overshooting’ theories, 151 Peru, trade policy in, 54 Philippines: structural adjustment policies, 31, 32–3, 35, 40–2, 44; trade policy, 54, 69 Polak model, 30 political weakness: 139; and structural adjustment, 34–5; and trade policy, 54, 167, 209–10 price(s): agricultural, 30, 72; instability, 4–5; misalignment of domestic and world, 46, 47 privatization, 119 productivity: 98–9; and domestic competition, 103–6; and trade policy, 99–103, 208, 259–60, 263 property rights, 119 protectionism: African regional integration and, 199–210; arguments against, 6–9, 210–12;

arguments for, 13–15, 47–9; incoherent trade policies and rise of, 166–84; model of determinants of, 174, 185–96; welfare cost of, 100; see also non-tariff barriers; quotas; tariffs quotas: 70, 241–2, 247, 252; arguments against, 8–9, 11, 79, 81–2; arguments for, 11–12; removal of, 229–30, 232–3 rationalization of industries, 104, 105 regional policies, 69 regional trading groups: 72–3,115, 154; protectionism and, 199–210, 217–19; trade liberalization and, 210–17 regression costs, 207 rent-seeking models, 9, 79–80, 81, 89–90 reserves of foreign currency, 89 savings, 290 scale economies, 104, 202–3 Second Best, Theory of, 4,30 Senegal: fiscal policy, 209, 214; industrial policy, 221–8, 231–4; pricing problems, 47; trade policy, 70, 209, 214, 216, 222, 226–34 sequencing of reforms: in trade policy, 9, 59–65, 79–90, 153–4, 227–8; in ‘Washington Consensus’, 120–1,122 Sierra Leone, fiscal policy in, 62 Singapore: industrialization in, 44; productivity in, 102; trade policy, 68 Somalia, fiscal policy in, 62 Sri Lanka, devaluation in, 63 stabilization policies: 27, 30, 34, 40–2; and economic growth, 147–50; and trade reform, 60–2, 150–4; and ‘Washington Consensus’, 112–23 state, see government static efficiency, 99–100 strategic trade policy theory, 49 structural adjustment policies: 2, 27–8, 43–5, 93; effectiveness of, 35–9; experience of, 39–43, 210, 213–17;

INDEX

223

implementation of, 28–35; see also liberalization of trade, stabilization policies structuralism, 27, 43 subsidies: to exports, 66–7, 292–8; to infant industries, 81 supply-side policies, 27

unemployment, 167 United States of America: import penetration of, 168, 170; protectionist pressures in, 167, 171–2, 174, 176–81 Uruguay: trade policy, 69; ‘Washington Consensus’ policies in, 122, 123

Taiwan: industrialization, 83, 84; trade policy, 65, 66, 68, 84–5 Tanzania: 2; devaluation in, 53; trade policy in, 60, 63, 70 tariffs: 81–2, 228–9, 230, 240, 242–5; arguments for, 8–9, 11; reform and removal of, 53, 62, 63–4, 70–1, 87, 89, 153, 226 taxation: 133, 138–9, 142, 143, 145, 209; and trade policy reform, 62, 70, 152, 214; ‘Washington Consensus’ on, 118 technical change: and industrialization process, 10, 83–4 technical efficiency, 100 technocracies, 35 Thailand: structural adjustment policies, 32–3, 41,42; trade policy, 68, 70 Togo, trade policy in, 216 trade: volumes of, 5 trade policies, see export-orientation; import-substitution; liberalization; protectionism training schemes, 74, 203 transport facilities, 48, 63, 208 Turkey: exchange rate policy, 66; industrialization, 276–86; stabilization policies, 123; structural adjustment policies, 32–3, 35, 41, 42, 44; trade policy, 51, 53, 68, 69, 102, 276–86

wages: 225; and structural adjustment policies, 40, 41,42 ‘Washington Consensus’ policies: 111–23; critique of, 123–4, 150–3 West African Economic Community, 72 World Bank: and Brady initiative, 123; on state intervention, 27; and structural adjustment policies, 29, 30–45; and trade policy reform, 51, 69, 93–4, 97, 105, 227–8, 234; and ‘Washington Consensus’, 111, 113

UDEAC (Union Douanière des Etats d’Afrique Centrale), 72, 204, 215 Uganda: fiscal policy, 62; trade policy, 60

X-efficiency, 100, 103 Yugoslavia, trade policy in, 51, 54, 66,102 Zaire: devaluation in, 63; fiscal policy, 62; trade policy, 63, 214 Zambia: fiscal policy, 62; manufacturing industry, 58; trade policy, 51, 54, 73–4 Zimbabwe, trade policy in, 51, 54