Trade, Growth and Development: The Role of Politics and Institutions

TRADE, GROWTH AND DEVELOPMENT

TRADE, GROWTH AND DEVELOPMENT The role of politics and institutions Proceedings of the 12th Arne Ryde Symposium, 13–14 June 1991, in honour of Bo Södersten

Edited by Göte Hansson

London and New York

First published 1993 by Routledge 11 New Fetter Lane, London EC4P 4EE This edition published in the Taylor & Francis e-Library, 2003. Simultaneously published in the USA and Canada by Routledge 29 West 35th Street, New York, NY 10001 © 1993 Selection and editorial material, Göte Hansson. Individual chapters to their authors. 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 A catalogue record for this book is available from the British Library ISBN 0-203-16837-2 Master e-book ISBN

ISBN 0-203-26358-8 (Adobe eReader Format) ISBN 0-415-08760-0 (Print Edition) Library of Congress Cataloging in Publication Data Arne Ryde Symposium (12th:1991: Helsingør, Denmark) Trade, growth, and development: the role of politics and institutions: proceedings of the 12th Arne Ryde Symposium, 13–14 June 1991, in honour of Bo Södersten/edited by Göte Hansson. p. cm. Includes bibliographical references and index. ISBN 0-415-08760-0 1. Developing countries—Commerce—Congresses. 2. Developing countries—Commercial policy—Congresses. 3. Developing countries— Economic conditions—Congresses. 4. International trade— Congresses. 5. Economic development—Congresses. I. Hansson, Göte, 1950– . II. Title. HF4055.A76 1991 380.1´09172´4—dc20

93–19274 CIP

CONTENTS

List of figures List of tables List of contributors Preface

vii ix xi xiii

INTRODUCTION

1

Part I Recent developments in the theory of trade, growth and economic development 1 INSTITUTIONS, KNOWLEDGE, TRADE AND GROWTH: A PERSONAL ADDRESS Bo Södersten

13

2 THE NEW GROWTH THEORY, TRADE AND DEVELOPMENT: SOME BRIEF REFLECTIONS Pranab Bardhan

22

3 WAGE DISPERSION, INTERNATIONAL TRADE AND THE SERVICES SECTOR Ronald Findlay

28

Part II Latin America 4 SPAIN AND THE CONQUEST OF AMERICA: PROFITS, RELIGION AND FORCED LABOUR IN THE FIFTEENTH AND SIXTEENTH CENTURIES Mats Lundahl 5 DO TERMS OF TRADE MATTER? TERMS OF TRADE AND ECONOMIC PERFORMANCE IN CENTRAL AMERICA Tarmo Haavisto and Claudio Evald Nalin

v

43

71

6 ELEMENTS DETERMINING THE COMPETITIVENESS OF GUATEMALAN PRIMARY EXPORTS Lars Pira 7 FOREIGN FIRMS AND STRUCTURAL ADJUSTMENT IN LATIN AMERICA: LESSONS FROM THE DEBT CRISIS Magnus Blomström and Robert E.Lipsey

86

109

Part III Asia 8 EAST ASIAN DEVELOPMENT AND JAPANESE DIRECT INVESTMENT Thomas Andersson and Staffan Burenstam Linder 9 DIRIGISME OR FREE-TRADE REGIME? A HISTORICAL AND INSTITUTIONAL INTERPRETATION OF THE TAIWANESE SUCCESS STORY Christer Gunnarsson

135

154

10 TRADE AND DEVELOPMENT: THE EXPERIENCE OF KOREA AND TAIWAN Wontack Hong

184

11 ECONOMIC REFORMS UNDER AN IMPORT-SUBSTITUTION REGIME: THE EXPERIENCE OF BANGLADESH Alia Ahmad

203

Part IV Africa 12 EXCHANGE RATE POLICY AND MANAGEMENT IN THE CONTEXT OF ECONOMIC REFORMS IN SUB-SAHARAN AFRICA: TANZANIA AS AN ILLUSTRATIVE CASE Benno Ndulu

229

13 MACROECONOMIC ADJUSTMENT AND TRADE LIBERALIZATION IN KENYA, UGANDA AND ZAMBIA Arne Bigsten and Steve Kayizzi-Mugerwa

248

14 INTERNATIONAL TRADE AND ECONOMIC GROWTH: THE EXPERIENCE OF ETHIOPIA Göte Hansson

280

15 REGIONAL ECONOMIC INTEGRATION AND INTRAAFRICAN TRADE: ISSUES FOR DEVELOPMENT Daniel B.Ndlela

303

Index

329

vi

FIGURES

Figure Figure Figure Figure Figure Figure

1.1 3.1 3.2 3.3 3.4 5.1

Figure 5.2 Figure 11.1 Figure 11.2 Figure 11.3 Figure 11.4 Figure 11.5 Figure 13.1 Figure 13.2 Figure 13.3 Figure 13.4 Figure 13.5 Figure 13.6 Figure 13.7 Figure 13.8 Figure 14.1 Figure 14.2 Figure 14.3 Figure 14.4 Figure 14.5 Figure 14.6 Figure 14.7

Disparities in incomes and human development

Terms of trade for five Central American and five Nordic countries Frequency plot of terms of trade changes Current account and budget deficits Growth of exports Investments GDP growth by sector Effective rates of protection and effective rates of assistance

Development of GDP and agricultural production Trade and budget deficits Import composition Net barter terms of trade

Figure 14.8 Export composition Figure 14.9 (a) Per capita crop export and (b) per capita exports of hides and skins and live animals

vii

4 30 32 32 38 77 80 209 210 211 211 212 252 253 253 254 254 256 263 269 282 287 289 291 292 292 293

293 294

FIGURES

Figure 14.10 Figure 14.11 Figure 14.12

Debt service ratio and international reserves (a) Ethiopian exports: country composition (b) Ethiopian imports: country composition Development of real effective exchange rate and trade deficit

viii

295 298 299

TABLES

Table 5.1 Table 5.2 Table 5.3 Table 7.1 Table 7.2 Table 7.3 Table 7.4 Table 7.5 Table 7.6

Table 7.7

Table 7.8 Table 7.9 Table 7.10 Table A7.1 Table A7.2

GDP growth per capita, Central American and Nordic countries Mean values, standard deviations and extreme values The higher central moments Average annual growth rates of manufactured exports from developing countries Average annual growth rates of exports by US majorityowned manufacturing affiliates in developing countries Export propensities of US majority-owned manufacturing affiliates Change in total sales, local sales and exports in US majorityowned manufacturing affiliates in developing countries Export propensities in manufacturing in eleven developing countries Sales outside the host country by US majority-owned manufacturing affiliates as percentage of value of manufactured exports Industry distribution of sales outside the host country by US manufacturing affiliates in Latin America and developing Asia Average annual growth rates of exports by US majorityowned manufacturing affiliates in developing countries Export propensities of US majority-owned manufacturing affiliates in developing countries Changes in total sales, local sales and exports, US majorityowned affiliates in developing countries Value of manufactured exports by selected developing countries and country groups Sales outside the host country by US majority-owned manufacturing affiliates

ix

75 78 81 111 112 113 114 116

117

118 120 121 122 125 126

TABLES

Table A7.3 Table A7.4 Table A7.5 Table A7.6 Table A7.7 Table 8.1 Table 8.2 Table 8.3 Table 8.4 Table 8.5 Table 11.1 Table 11.2 Table 12.1 Table 12.2 Table A12.1 Table 13.1 Table 13.2 Table 13.3 Table 13.4 Table 13.5 Table 13.6 Table 13.7 Table 13.8 Table 13.9 Table 13.10 Table 13.11 Table 13.12 Table 15.1 Table 15.2 Table 15.3 Table 15.4 Table 15.5 Table 15.6 Table 15.7

Total sales by US majority-owned manufacturing affiliates Sales to the USA by US majority-owned manufacturing affiliates in developing countries Value of manufacturing output in producers’ prices Sales outside the host country by US majority-owned manufacturing affiliates in developing countries Total sales by US majority-owned manufacturing affiliates in developing countries Growth in income, growth in exports and average savings rates Debt service ratio, total direct investment and Japanese direct investment Variable definitions and descriptive statistics Results Summary of the expected and realized effects, sign and significance Effective rates of assistance by groups of industries Indices of industrial production by sector Effectiveness of devaluation in selected African countries Determinants of the real exchange rate in Tanzania Tanzania: real exchange rate and the fundamentals Price indices for Kenya Macroeconomic indicators for Kenya Kenya: balance of payments statistics Real sectorial and GDP growth rates in Kenya Price indices for Uganda Uganda: macroeconomic indicators Uganda: balance of payments statistics Sectorial and GDP growth rates for Uganda Price indices for Zambia Macroeconomic indicators for Zambia Zambia: balance of payments statistics Zambia: real sectorial and GDP growth rates World trade and share of developing Africa in world trade Developing Africa’s direction of trade Share of intra-African trade in total trade of developing African countries Intra-African trade (exports) Intra-African trade (imports) Intra-subregional trade matrix Value of intra-African trade by commodity classes

x

127 128 129 130 131 138 141 145 147 150 214 217 230 239 244 257 258 259 260 264 264 265 266 270 271 272 273 304 305 306 314 316 318 319

CONTRIBUTORS

Alia Ahmad is a research fellow of economics, Department of Economics, School of Economics and Management, University of Lund, Sweden. Thomas Andersson is a senior research fellow at the Industrial Institute of Economic and Social Research, Stockholm, Sweden. Pranab Bardhan is professor of economics, University of California at Berkeley, California. Arne Bigsten is professor of development economics, Department of Economics, University of Gothenburg, Sweden. Magnus Blomström is professor of economics, Stockholm School of Economics, Sweden and a research associate of the National Bureau of Economic Research, New York. Staffan Burenstam Linder is professor of international economics and President of Stockholm School of Economics, Sweden. Ronald Findlay is professor of economics, Columbia University, New York. Christer Gunnarsson is associate professor of economic history, School of Economics and Management, Lund University, Sweden. Tarmo Haavisto is a research fellow of economics, School of Economics and Management, University of Lund, Sweden. Göte Hansson is associate professor of economics, School of Economics and Management, University of Lund, Sweden. Wontack Hong is professor of international economics, College of Social Sciences, Seoul University, Korea. Steve Kayizzi-Mugerwa is a research fellow of economics, Department of Economics, University of Gothenburg, Sweden. Robert E.Lipsey is professor of economics, Queens College and the Graduate Center, City University of New York, and a research associate of the National Bureau of Economic Research, New York.

xi

CONTRIBUTORS

Mats Lundahl is professor of development economics, Stockholm School of Economics, Sweden. Claudio Evald Nalin is a research fellow of economics, School of Economics and Management, University of Lund, and an economist for OECD, Paris. Daniel B.Ndlela is an economic consultant in Bulawayo, Zimbabwe. Benno Ndulu is professor of economics, African Economic Research Consortium, Rockefeller House, Nairobi, Kenya. Lars Pira is Ambassador at the Embassy of Guatemala, Stockholm, Sweden. Bo Södersten is professor of international economics, School of Economics and Management, University of Lund, Sweden.

xii

PREFACE

This volume contains the papers presented at the 12th Arne Ryde Symposium, 13–14 June 1991, held at Elsinore, Denmark. The symposium was generously financed by the Arne Ryde Foundation which was established by Valborg and Sven Ryde, the parents of Arne Ryde who was an exceptionally promising young doctoral student at the Department of Economics, Lund University. Arne Ryde died in 1968, only 23 years old, after an automobile accident. Ever since 1973 the Arne Ryde Foundation has generously financed international symposia and lectures that are of great importance for economic research, in particular for doctoral students, research fellows, lecturers and professors at the Department of Economics, Lund University. The students and the staff of the department as well as other participants in the Arne Ryde Symposium are all very grateful to the Arne Ryde Foundation. In May 1990 I was asked by Professor Björn Thalberg at Lund University, a member of the board of the Arne Ryde Foundation, to organize the 1991 Arne Ryde Symposium. I was also asked to choose a topic in honour of Professor Bo Södersten, who celebrated his 60th birthday in June 1991. As a former student of Bo Södersten, I was deeply honoured by this request. The choice of topic for the symposium was straightforward. Professor Södersten’s major contributions to economic science have been in the field of international economics. In this field he is well known internationally through his textbook International Economics. Since its publication in the USA (1970) and in the UK (1971), this book has been translated into several other languages and has been used in a great number of universities all around the world. Partly the textbook builds on Södersten’s doctoral thesis from 1964, A Study of Economic Growth and International Trade, which provides an interesting general equilibrium analysis of the relationship between economic growth and international trade. Thus, considering that the choice of trade regimes is one of the most debated issues in relation to strategies of economic growth and development, the topic of the symposium was a more or less obvious choice. I am very grateful to all contributors who accepted the invitation and produced interesting papers on the various themes that were indicated in their respective invitations. Furthermore, I would like to express my gratitude to

xiii

PREFACE

Hans Falck, Ulla Olofsson, Keith Persson, Jeanie Petersson and Ann-Charlotte Sahlin who assisted me in practical matters before and during the conference. Finally, I would like to express my gratitude to Yves Bourdet, Mats Lundahl and Björn Thalberg for their support during various stages of the process leading up to this volume. Göte Hansson Lund

xiv

INTRODUCTION Göte Hansson

Issues of economic growth and development have been subject to a large number of theoretical and empirical analyses and have been debated intensively among politicians and economists also. The question whether international openness is an important part of a successful development strategy has been at the very centre of both research and the political debate after the Second World War and in particular during the past three decades. The nineteenth century, in contrast to earlier centuries, showed very high rates of economic growth in the western world. One important explanation for this rapid growth is that world trade grew rapidly during this century. Also, during the period after the Second World War, world trade has grown very rapidly and so has the production and incomes of industrial countries. However, only some of the less developed countries, concentrated in East and South-East Asia have succeeded in generating substantial growth in their per capita income, whereas most countries in sub-Saharan Africa have experienced falling per capita income. This is particularly true for the period after the economic shocks of the 1970s, e.g. the oil crises, the concomitant international recession and the increased interest rates. As a result, the gap in incomes between the industrialized world, the OECD countries and the newly industrialized countries (the NICs) on the one hand and the least developed countries on the other has grown sharply and continues to do so. A look at economic development during the last twenty-five years shows that GNP per capita in the least developed countries as a share of GNP per capita in the industrialized countries has fallen from 3 per cent in 1965 to just 1 per cent in 1989 (UNDP 1992: Table 7). Furthermore, the richest 20 per cent of the world’s population receive 82.7 per cent of the world income, the second richest 20 per cent receive 11.7 per cent, and the remaining 60 per cent receive 5.6 per cent, with just 1.4 per cent of the world income accruing to the poorest 20 per cent (UNDP 1992). What are the reasons for the poverty of the majority of the world’s population and why is the gap growing? In the literature, explanations refer to development strategies, the problems of internationalization for primary exporting countries, and the long-run development of terms of trade for less developed countries.

1

INTRODUCTION

During the last decade the debt crisis faced by many non-oil-producing less developed countries has also been advanced as an obstacle to development. The development debate during the last couple of decades has raised questions as to whether or not international trade and openness can work as an engine of growth. In a study by David Greenaway and Chong Hyun Nam from 1988 it is concluded that ‘We may not be entirely clear of the precise determinants of growth; the potential for trade strategy in discouraging growth is, however, rather more apparent’ (Greenaway and Chong Hyun Nam 1988:433). In debt crisis countries, the World Bank and the IMF have demanded various types of structural adjustment or economic recovery programmes as a prerequisite for favourable loans. These programmes have focused on various aspects for different countries. However, they have some common characteristics: liberalization of markets, trade liberalization, privatization, expenditure reductions, and correction of the exchange rate. Can these programmes be expected to take the debt crisis countries on to a track where trade can work as an engine of economic growth and development? As early as 1776 in An Inquiry into the Nature and Causes of the Wealth of Nations, Adam Smith observed and emphasized the role of foreign trade in solving the economic problems in a country with a rapidly growing population and limited agricultural resources. However, trade is not just seen as something positive in relation to growth and development. Economists such as Myrdal, Prebisch, Singer and Emanuell have all pointed to factors that recommend a more moderate or even reserved attitude towards international trade. According to the Singer-Prebisch argument against trade as an engine of growth, productivity growth (through technical progress) in less developed countries gives rise to a negative development of terms of trade. To make a critical mathematical analysis of this type of negative result and to answer a question originally raised by Mill (1962) concerning the results of technical improvements in the export branch were two of the objectives of Södersten’s doctoral thesis A Study of Economic Growth and International Trade. Södersten’s thesis, presented in 1964, also analyses the effects on trade of factor growth and what he calls unspecified growth. It does not analyse the causes of economic growth but rather focuses on how growth affects ‘a country’s trade and prosperity’ (Södersten 1964:14). Södersten’s thesis can be seen as an important critical assessment of the arguments postulating trade-generated deterioration of terms of trade and the quite striking phenomenon of impoverishing growth. Even if Södersten does not explicitly deal with the casual link between international trade and economic growth, he does so indirectly. Södersten was well aware of the limitation of his study in this respect and he recommended a much broader scientific approach to analyse the growth process: The growth process is obviously a complex result of economic, sociological, psychological, religious, etc. forces. It is hardly possible that economics alone could, even in principle, explain economic

2

INTRODUCTION

growth…. Yet, economists must study those parts of the whole that can be used as stepping-stones towards a more complete understanding of the determinants of growth (Södersten 1964:14) When one looks at the recent economic development in various parts of the world and also at recent economic theory on growth and development, the issue of the causes of economic growth is still puzzling and to some extent even more so today than in the early 1960s. Growth performance and standards of living differ largely between continents but also between countries within the same continent. This fact is a challenge of the utmost importance for social scientists, or, as stated in the World Development Report 1991, ‘Development is the most important challenge facing the human race’ (World Bank 1991:1). Grossman and Helpman, also, underline the importance of research in the field by stating: ‘It is the job of the social scientist to uncover the mechanisms that link structural and policy conditions to realized growth performance’ (Grossman and Helpman 1991:xi). This volume is one response to the challenges defined by Bo Södersten in his thesis back in 1964, which have now been highlighted by the present harsh realities and by a widening gap in the global economy. It contains both chapters on the recent developments in the theories of trade, growth and development, where factors such as knowledge and various types of institutions play an important role, and analyses of the experience of a great number of countries with different political and economic systems. Figure I.1 presents statistics on GNP per capita in twenty-six of the countries dealt with in this volume.1 The GNP per capita is expressed as an index where the average GNP per capita in the North, that is in the industrial economies, has index 100. In the figure it can be seen that very few developing countries have come close to the North in terms of per capita income while the majority of the countries dealt with have moved in the opposite direction. This is particularly the case for the African countries. The economic growth of a country is not synonymous with the development of the country. As noted by Chenery and Srinivasan, ‘It is generally agreed that “development” encompasses the reduction of poverty, improvements in the health and education of the population, and an increase in productive capacity as well as rising per capita income’ (Chenery and Srinivasan 1988:xi). To give an indication of the difference in relation to such non-pure economic variables the figure also presents the UNDP’s Human Development Index. 2 In terms of human development too the African countries tend to lag behind the rest of the world (UNDP 1992: Table 1). The empirical chapters analyse policy and performance in less developed countries around the world. Special interest is devoted to the role of politics and concomitant economic policies, in particular the choice of development strategy, and the role of institutions as important explanations for the economic success or failure of the respective countries.

3

INTRODUCTION

Figure 1.1 Disparities in incomes and human development Source: UNDP (1992: Tables 1 and 7)

PLAN OF THE BOOK The book is divided into four parts, one on recent theoretical developments and three on the experience of some countries in Latin America, Asia and Africa. Part I presents chapters on recent developments in the theory of trade, growth and economic development. In Chapter 1, ‘Institutions, knowledge,

4

INTRODUCTION

trade and growth’, Bo Södersten, in a personal address to the symposium, starts out from the fact that traditional growth models have difficulties in explaining the contemporary growing differences in growth rates and income levels between countries. Södersten also discusses how the new growth theories, which emphasize the role of knowledge and institutions, together with international trade, can provide plausible explanations for the observed differences in growth rates and income levels. Pranab Bardhan’s contribution in Chapter 2, ‘The new growth theory, trade and development: some brief reflections’, contains a short survey and discussion of the theoretical literature on trade, growth and development. Among the interesting issues dealt with is the result from the new growth theories that when accumulation of knowledge takes place mainly in rich countries, trade may reduce the profitability in research and development activities in poor countries and may drive these poor countries to specialize in traditional labour-intensive and possibly also stagnant industries. Bardhan’s survey underlines the importance of further theoretical research on the relationship between trade and growth, for instance on how various aspects of imperfect competition and multinational corporations affect this relationship. Bardhan concludes that: The theoretical models, for all their recent enrichment, have a long way to go before they can catch up with the complexity of the empirical reality.’ In Chapter 3, ‘Wage dispersion, international trade and the services sector’, Ronald Findlay incorporates skilled labour as an important factor of production into his analysis. This is done by adding to the traditional two-by-two general equilibrium model a third sector, services, that besides unskilled labour also uses skilled labour. Findlay analyses the effects on production structure and factor rewards of exogenous changes in international prices, factor endowments, technology, and trade policy. The analysis clearly reveals the importance of making a division of labour into skilled and unskilled labour. The analysis also shows that it is worth while to go beyond the pure theory of international trade and to use theories and models from the field of labour economics, such as the theory of human capital. The three chapters in Part I emphasize the role of institutions and knowledge in recent theories on trade and growth. Parts II to IV then analyse the experience of some less developed countries. Part II concentrates on the experience of some Latin American countries. First, in Chapter 4, ‘Spain and the conquest of America’, a historical perspective is presented. In this chapter Mats Lundahl analyses the Spanish conquest of America. He finds that the underdeveloped countries in the American continent did not adapt easily in the way predicted by the standard comparative cost doctrine of international trade. The driving forces when opening these countries to trade were non-neoclassical and the results had little to do with the predictions of standard theory. Among the facts that must be incorporated into any reasonably realistic explanation of how trade was initiated, as well as into the analysis of the impact that trade had in the American continent, Lundahl mentions that:

5

INTRODUCTION

(1) The opening of trade cannot be separated from the act of conquest…. (2) The forms that trade took during and after the conquest…were a direct outcome of events in Europe….(3) A steady flow of exports from the Americas could not be obtained without a thorough institutional reorganization of the indigenous society. (4) The new institutional forms were the result of a (partly zero-sum) game between three European actors: the Crown, the Spanish settlers and the Church. (5) The motives behind the conquest were not only economic but also religious. Given these facts, Lundahl develops an explanation that is largely an economic explanation of how trade was initiated between Spain and its American colonies and of some of the most important effects that this had on the Americas. One theory of trade and growth that has one of its originators in Latin America, Raúl Prebisch, and that has occupied many economists since the writings of Prebisch and Singer (UNDEA 1949) and Prebisch (1950) is the role of terms of trade in the process of economic growth and development. According to these authors, strategies based on the expansion of primary exports will lead to deteriorating terms of trade and cause growing hardship in the Latin American countries instead of growth through trade. In the political debate also, the international trade system resulting in an unfavourable development of the terms of trade for less developed countries is often claimed to be responsible for the poor performance of these countries. Tarmo Haavisto and Claudio Evald Nalin, in Chapter 5 ‘Do terms of trade matter? Terms of trade and economic performance in Central America’, analyse whether the long-run development of terms of trade can be given this critical role as an explanation for the poor development of five countries in Central America, i.e. Costa Rica, El Salvador, Guatemala, Honduras and Nicaragua. By comparing the economic growth of these countries with the growth of five Nordic countries that have experienced approximately the same development in terms of trade, Haavisto and Nalin show that it is difficult to attribute the observed poor economic growth in the Central American countries to the development of terms of trade. Instead, they conclude, one should look at differences in the way the state has performed, the security and stability of property rights and other institutional arrangements. In Chapter 6, ‘Elements determining the competitiveness of Guatemalan primary exports’, Lars Pira analyses the role of primary exports in the process of growth and economic development in one of the Central American countries, namely Guatemala, and what type of factors determine export competitiveness. It is shown that the opening up of exports can be explained by integrating the ‘vent for surplus’ and the ‘staples’ theories but that these theories have to be complemented by institutional analysis to explain how and why employment of labour, capital and land for exports has changed through history. It is also concluded that, compared with the successful North American exports which were highly dependent on British institutions, the legacy of Spanish institutions in

6

INTRODUCTION

Latin America led to relative backwardness and that Guatemala, for example, ended up depending more on nearby markets than on demand generated from Spain. One of the contemporary problems and obstacles to growth and development in less developed countries is the debt problem. In Chapter 7, ‘Foreign firms and structural adjustment in Latin America: lessons from the debt crisis’, Magnus Blomström and Robert E.Lipsey analyse the impact of foreign firms on economic performance in the Latin American indebted countries, Brazil, Chile, Colombia, Mexico and Venezuela. Among their conclusions is that the affiliates of foreign firms showed greater ability than Latin American firms to increase exports from debt-ridden countries and to change the market orientation from domestic sales towards exports. Their analysis also suggests that the foreign firms are better equipped than local firms to change the orientation of industries from importsubstitution to export-oriented industries. Part III presents and analyses the success stories of some countries in East and South-East Asia. In Chapter 8, ‘East Asian development and Japanese direct investment’, Thomas Andersson and Staffan Burenstam Linder analyse the economic success of East Asia. In this region large inflows of direct investment have paralleled the openness to trade. The focus of the study is on Japanese direct investment, which has increased tremendously during the 1980s. It is concluded that countries with a relatively low income, an export-oriented strategy and modest taxation appear to be favourable attractions for Japanese investment. Another interesting conclusion is that during the period 1979–89 the largest flows in absolute terms have been allocated to sectors with low technology and low skill intensity while it is the more advanced sectors that have received the greatest relative increases in the flows of Japanese investment. In Chapter 9, ‘Dirigisme or undistorted free-trade regime? A historical and institutional interpretation of the Taiwanese success story’, Christer Gunnarsson shows that in the case of Taiwan the success story is largely due to the strength of Taiwan’s politico-economic system. This strength, it is argued, has been manifested through its institutional efficiency displayed in the dichotomy between state and civil society and between politics and economics. The power of the autonomous state has been a guarantee of competition. Gunnarsson also discusses whether it will be possible to directly apply the Taiwanese model to other countries. He concludes that such a direct application will face considerable difficulties, one set of reasons being the political resistance that the introduction of a ‘Taiwanese’ model can be expected to meet. Chapter 10, ‘Trade and development: the experience of Korea and Taiwan’, by Wontack Hong, also analyses the success stories of two South-East Asian countries that have gone through drastic transformations from typical backward economies to export-oriented successful countries. Wontack Hong compares the experience of the two countries and focuses on such factors as how they institutionalized their respective export-oriented strategy and the characteristics of their respective financial systems. The chapter also contains a comparative

7

INTRODUCTION

evaluation of the growth performance of Korea and Taiwan, the effects of growth on income distribution, the role of interest groups, and a brief discussion of how the democratization process in the labour market can be expected to affect competitiveness and growth. It is concluded that with the process of democratization in Korea labour unions will play a crucial role in the future course of Korea’s growth, whereas in Taiwan labour unions do not present any problems since there is still a long way to go before constitutional democracy can be expected to be introduced there. Even if many Asian countries can be classified as success stories in terms of growth and development performance, as was seen in Figure I.1, this is not the case for all Asian countries. In the final chapter in Part III, Chapter 11, Alia Ahmad analyses the attempts at economic liberalization in Bangladesh. It is concluded that attempts to replace the import-substitution policy and to improve the poor economic performance have resulted in failures. The explanations presented are typical DUP activities, i.e. activities like lobbying, bribery, broad corruption, and government failure with regard to the provision of infrastructure, institutional and technical support. In spite of the differences in culture, the experience of Bangladesh has great similarities with the experience of African countries reported in Part IV. Having concluded in the previous parts that debt problems and terms of trade development per se cannot be claimed to provide satisfactory explanations for poor economic performance, and having found that economic policy, institutions and knowledge play a critical role for the potential of trade to work as an engine of growth, the chapters in Part IV analyse what has gone wrong in the African continent, with special emphasis on the role of politics and institutional factors. It is well documented that exchange rate misalignments are very frequent in less developed countries and the exchange rate regime is also one of the fields of economic policy on which the IMF and the World Bank focus when they design various types of recovery and adjustment programmes. Benno Ndulu, in Chapter 12, ‘Exchange rate policy and management in the context of economic reforms in sub-Saharan Africa: Tanzania as an illustrative case’, deals with exchange rate policy and management in sub-Saharan Africa, where during the past decade a large number of countries have conducted an active exchange rate policy. Ndulu focuses on the case of Tanzania and concludes that a flexible exchange rate regime that is able to respond to exogenous as well as domestic-policy-induced shocks is necessary to avoid deterioration in international competitiveness. Ndulu’s study suggests that an active exchange rate policy can play an important role in correcting the real exchange rate but that it should be complemented by macroeconomic policy (controls on credit expansion, reduction of exchange controls and trade liberalization) and institutional restructuring. Finally, Ndulu concludes that, since export performance in African countries is largely supply constrained, an appropriate incentive structure is a condition for obtaining supply responses, but that an enabling institutional structure and infrastructure that can

8

INTRODUCTION

support production and marketing are critical also in improving export performance and growth. Arne Bigsten and Steve Kayizzi-Mugerwa in Chapter 13, Macroeconomic Adjustment and Trade Liberalization in Kenya, Uganda and Zambia’, also analyse the response to the shocks of the 1970s. Their study covers three African countries, Kenya, Uganda and Zambia, and their economic reforms and performance during the 1980s. Even if the three countries all pursued largely similar adjustment policies based on the IMF/World Bank model, there have been wide variations in the results. In the analysis, Bigsten and Kayizzi-Mugerwa focus on the role of the economic structure, the nature and extent of the domestic control regimes, and the role of various pressure groups in the respective country. One of their interesting conclusions is that stability in government policy, which is necessary for the successful implementation of economic reforms, requires a sustainable underlying growth process, and that such a sustainable growth process cannot be wholly dependent on a single commodity. There is a severe risk that the modern sector can be wiped out by a bust, as has been the case in Zambia, but not in Kenya where the economy is more diversified. In Chapter 14, ‘International trade and economic growth—the experience of Ethiopia’, Göte Hansson analyses the economic situation in post-revolutionary socialist Ethiopia. With the help of a dual economy approach it is shown that the socialist policy and economic system discriminated heavily against individual peasants and thereby is responsible for the poor agricultural performance. The negative effects on agricultural performance of the socialist domestic policy were strengthened by the Ethiopian trade and exchange rate policy. International trade has not been allowed to work as an engine of growth. The poor agricultural performance, together with rapid population growth, has made increasing food imports necessary. As a consequence, Ethiopia experienced a growing balance of payment deficit and a rapidly growing debt. The chapter also analyses the potential and the weaknesses of the recent economic reform programmes that have been presented but only partially implemented. In the final chapter in Part IV, Chapter 15 ‘Regional economic integration and intra-African trade: issues for development’, Daniel Ndlela analyses the attempts at and potential of regional economic integration in Africa. Concern with Africa’s economic dependence on industrialized countries when it comes to trade and the slow or even declining trend in intra-African trade links has led African policy makers to reassert their determination to expand regional economic integration and intra-African trade. The African countries have established regional institutions to support these efforts. However, despite these institutions, Ndlela’s analysis shows that the level of intra-African trade continues to be very low. One of the conclusions drawn in the chapter is that: The success of regional economic integration in Africa will essentially depend on the political will of the member states in enabling the regional institutions to implement their programmes and to create an environment that enables maximum participation of non-governmental actors.’

9

INTRODUCTION

NOTES 1 2

The countries chosen are those for which comparable statistics are available. They are also countries that are dealt with more than marginally in the present volume, even though there are great differences in the depth of the analyses. The UNDP’s Human Development Index is an index which incorporates life expectancy, educational data and income indicators. See UNDP (1992:3) and UNDP (1992: Chapter 1) for more details and a discussion of the index.

REFERENCES Chenery, H. and Srinivasan, T.N. (1988) ‘Preface to the Handbook’, in H.Chenery and T.N.Srinivasan (eds) Handbook of Development Economics, vol. 1, Amsterdam: North-Holland. Greenaway, D. and Chong Hyun Nam (1988) ‘Industrialisation and macroeconomic performance in developing countries under alternative trade strategies’, Kyklos 41: 419–35. Grossman, G.M. and Helpman, E. (1991) Innovation and Growth in the Global Economy, Cambridge, MA: MIT Press. Mill, J.S. (1962) Principles of Political Economy Book III, New York. Prebisch, R. (1950) The Economic Development of Latin America and Its Principal Problems, New York: United Nations Economic Commission for Latin America. Södersten, B. (1964) A Study of Economic Growth and International Trade, Stockholm: Almqvist & Wiksell. UNDEA (1949) Relative Prices of Exports and Imports of Under-Developed Countries, New York: United Nations, Department of Economic Affairs. UNDP (1992) Human Development Report 1992, New York: Oxford University Press. World Bank (1991) World Development Report 1991, New York: Oxford University Press.

10

Part I RECENT DEVELOPMENTS IN THE THEORY OF TRADE, GROWTH AND ECONOMIC DEVELOPMENT

1 INSTITUTIONS, KNOWLEDGE, TRADE AND GROWTH A personal address Bo Södersten

INTRODUCTION The most basic fact about today’s world is that some countries are rich while the majority of mankind is poor. After the withering away of Soviet communism in its traditional form, the world today seems to comprise only two large areas: the developed or industrialized world with per capita incomes of 10,000 dollars or more and the less developed world with per capita incomes that are normally within the range 300–500 dollars. This is a question that has naturally concerned and interested economists for a long time. Some of us, I suppose, think that we broadly know the answer or answers to the question. Others would argue that the question is not really relevant; it is too broadly formulated for any answer to be given. Some economists would even go so far as to suggest that development theory should be banned as a field of study (and what there is to study should be left to anthropologists). It can be safely said that probably a majority of economists hold the view that trade usually plays a substantial role in determining a country’s opportunities for development. As far back as Adam Smith it was argued that the division of labour is only limited by the extent of the markets. International trade is undoubtedly the creator of markets and thereby of specialization. The fact that growth may have important effects on trade and lead to very different outcomes for relative prices and incomes is certainly not news to anyone. Some of us pursued that line of enquiry with some dedication in the late 1950s and 1960s. The questions we asked were perhaps somewhat narrow minded. But often the answers were very precise. It is also fairly self-evident that trade may also have important repercussions for development possibilities and as a prerequisite for economic growth. The law of comparative advantage, after all, teaches us that trade at least presents an opportunity to reap gains and improve a country’s position. Traditional growth and trade theory as we knew it was firmly embedded in the traditions of trade theory. Among its strengths was that it usually adopted a

13

RECENT DEVELOPMENTS IN THE THEORY

general equilibrium approach. Nevertheless its frequent use of a 2×2×2 form and a comparative static framework made it seem somewhat circumscribed and pedestrian. I will start by taking up two more recent, albeit quite different, approaches to the theme of trade and development. My starting point will be historically oriented. A new important paradigm has been launched by those historians, who have introduced an institutional approach to economic history. This approach is full of important insights and possible corollaries. It also allows trade to play a central part in the development of new institutions. Thereafter I will move on to examine a very different terrain. Traditional growth models have had difficulties in explaining both the differences in growth rates and the differences in income levels between countries which seem to me to be the most striking feature of our contemporary world. Traditional steady state models are not very interesting since they lead to the conclusion that incomes and growth rates will converge. That is manifestly not the situation in today’s world. If that were the case, the world would probably have become a much happier place to live in. The new type of theorizing that has emerged in recent years stresses the role of knowledge. One of the unsatisfactory aspects of earlier growth models was that accumulation of capital tended to come to a stop as the return on capital approached zero. In the new models knowledge plays both the role of a stock factor which creates a level effect that allows a high income and the role of a growth factor which can provide increasing returns to scale and rising marginal productivity. With the use of knowledge, hypotheses that explain differences both in the level and rates of growth of income may be generated. I will end my survey by commenting on the interrelationship between trade and knowledge and how knowledge can grow through trade. TRADE AND INSTITUTIONS The institutional approach to modern economic history takes its starting point around the year 1500. At this time the west did not have any advantage in economic or military matters over, let us say, the Islamic world or China. Islamic culture stood often on a higher scientific level and the Chinese were frequently more advanced in their technology. But now something started to happen in the west. Trade began to develop in a more systematic way. There were several reasons for the growing importance of trade. There had always been a need to bring supplies of various kinds from areas of surplus to regions of need. In the fifteenth and sixteenth centuries, an interlinked growth of trade and maritime technology took place, each feeding on the other in a significantly cumulative manner. It is important to realize that these activities were urban, not rural—hence the growing importance of a few, well-located cities. The development of trade was slow and painful. One important aspect related to the growth of markets which increased the scope for technical progress. Technical

14

INSTITUTIONS, KNOWLEDGE, TRADE AND GROWTH

progress was rather a response to economic needs and not in itself an independent force. We may say this early development of trade and technology had two important aspects: it was cumulative in nature and it made knowledge and technical progress an endogenous variable in development. The growth of trade stimulated the acquisition and need of knowledge which in turn improved trading methods and the scope for trade, which in turn increased the size of markets which in turn increased trade. After some two hundred years the west had changed, as had the comparative situation of nations. We should remind ourselves of the fact that all this took place before the era of the factory system which arrived much later towards the end of the eighteenth century. Hence trade was vital to economic development. However, increases in trade built on the acquisition and application of new knowledge. This systematic application of new knowledge fostered new institutions that were necessary for systematic increases in trade. These insights form the very core of the new paradigm. What was new and completely path breaking was that the west alone was able to build these new institutions. A fundamental prerequisite was that the west could construct a legal system. Trade involves transactions over time and merchants often became involved in disputes. Therefore they have to enter into agreements which are valid and whose consequences are as foreseeable as possible. No effective legal system existed when trade started to expand in the fifteenth century. It took a long time to devise and construct such a system. The English were pioneers in this field. After about two hundred years of trial and error, London had established courts which were reasonably effective and fair. When it became known that even a foreign merchant could expect a fair hearing in the defence of his rights at an English court, trade gained immensely. Other necessary prerequisites were the use of means of payment and the removal of medieval prohibitions and taboos about charging interest. Sea transport at that time involved great risks. To increase the opportunities for trade, it was therefore necessary to build up an insurance system. It is important to realize that as western capitalism emerged, it was centred around cities. The close link between trade and urbanization appears again and again. It seems to be the case that a critical mass of knowledge seems to be necessary for it to grow and spread in efficient ways. This is a point that will again emerge when we shortly return to the new, more esoteric and technical models of the interaction between economic growth and the role of knowledge. Another focal point is the diversion of the economic sphere from the political one. The predatory state dominated at the beginning of the modern era. The feudal lords and the princes and kings of that time always needed money for warfare, court pleasures and good living. They normally obtained the money and other resources through extortion and confiscation. The growing merchant class could to some extent counteract this by hiding their capital, smuggling and moving their

15

RECENT DEVELOPMENTS IN THE THEORY

enterprises to other cities and countries. However, it was vital to reach an agreement with the political powers in order that confiscation could be replaced by taxation. Over time this replacement of arbitrary exaction by a system of taxation would prove to be Pareto optimal. Accordingly the benefits of trade and accumulation could accrue to those who did the trading and accumulating. Property rights began to become established and private benefits and costs tended increasingly to reflect social benefits and costs. After some two hundred years the west had managed to establish an economic order based on legality, reasonable stability and property rights. The competing Islamic and Chinese cultures never managed to develop such institutions. The Arabs were skilful traders although they never managed to build a system that respected property rights. The ‘bureaucratic feudalism’ of the Asian kind never proved to be compatible with the emergence of capitalism and the long-run development of new institutions and new technology. The political sphere continued to prevail over the economic one. John Locke may be viewed as the political inspiration of the emerging new political world order and as the champion of the rising merchant class. Analogously Adam Smith rationalized in his writings 200 years of economic progress. His writings on economic development have usually been regarded as facile and naive. His basic message was not to worry about economic growth; it would take care of itself. The market would reward knowledge, adaptability, quick-wittedness and rational reasoning. Viewed in the light of the new paradigm with its stress on the importance of appropriate and efficient institutions, his writings may yet become vindicated and prove more full of insights than decades of Marxist and leftistinspired writings have made them out to be. KNOWLEDGE AND ECONOMIC GROWTH I will now move on to more theoretical types of approaches to growth and trade. Traditional growth theory of the Solow type works with production functions with constant returns to scale and falling marginal productivities. In this type of context, poor countries grow faster than rich ones. Growth and incomes converge and when growth has continued long enough, investment will stop as it no longer gives a positive return. This type of theory is not very satisfying as it obviously does not at all square with the basic facts of the huge differences between the rich and the poor countries. In recent years, a new type of growth theory has been launched. It seeks to explain both why some countries have higher levels of income than others and why some countries grow faster than others. Two factors which may not be new but are at least used in a novel and interesting fashion, particularly in relation to their interplay, are knowledge combined with the opening up of production functions which allows us to deal with increasing returns to scale. Pioneering work in this domain has been done by Paul Romer and Robert Lucas.

16

INSTITUTIONS, KNOWLEDGE, TRADE AND GROWTH

Knowledge has two aspects in particular, one that we might call the stock aspect and the other we can conveniently call the flow aspect. An individual may either work or have leisure. The latter activity is used for investment in human capital. An individual may therefore use his or her time for either work or education. The latter activity will increase the amount of human capital available to the economy. We might say that the accumulation of capital has two aspects, one of which we may call the external effect. The first simply means that the individual who receives training, let us say takes a PhD in economics, thereby increases his or her skill level. This will permit him or her to reap a higher reward by earning a higher salary than he or she would otherwise have done without this academic training. Furthermore, the average skill level in society increases. Even though no individual will have an effect on the overall level of a nation’s skill level, the total level will have an effect on overall productivity. The higher this level, the more productive will be all factors of production. The new types of theory often have a distinctly Schumpeterian flavour. An important concept is that of appropriability. Romer, for instance, assumes that each successful research project produces a patentable blueprint for its innovator. This in turn has the character of a ‘public good’ particular to the firm. But apart from this it also generates a contribution to the general stock of knowledge which is not appropriable. Therefore research has both an egotistical motive and an altruistic side effect. (Romer 1990a, b). In this version knowledge may be viewed as an external factor, an environmental bonus if you like. It should also be stressed that accumulation of human capital is a social phenomenon. I think we are all aware of the fact that it is very important to us who our colleagues are and we spend a lot of time and effort fighting about choosing who the right ones should be. We all want to be in milieux and environments that are ‘creative’. This certainly holds for academics but it holds equally for people in advertising or concerned with providing financial services as well as for musicians and actors. Talking about institutions, I stressed the role of cities in the development of these institutions. Now, again, we can see that cities are also the natural breeding grounds for the accumulation of human capital and hence for giving an increased value to the existing factors of production. The other aspect of accumulation of knowledge stresses our old friend ‘learning by doing’. In the former model, human capital could be accumulated only at the expense of giving up working time. When it comes to accumulation of human capital through learning by doing, human capital is instead product specific. We have a continuum of goods where new goods are continually being introduced. Learning by doing occurs rapidly at first, and then more slowly. There are diminishing returns to learning in each good but the ability to learn can be used in order to produce new goods. The more able and quick-witted an individual or a producer is, the better equipped he or she will be at playing the game of learning by doing.

17

RECENT DEVELOPMENTS IN THE THEORY

Demand conditions will also play a role that resembles the one we know from the old growth and trade models. The risk a producer always faces is that of diminishing terms of trade. If terms of trade do deteriorate, the alert producer will have to switch from that line of production into new ones. If the producer cannot do this, he or she faces the risk of falling into some form of immiserizing growth. These results are in complete accordance with those from the earlier growth and trade models we worked with in the 1960s. They again demonstrate the importance of adaptability for success. Naturally the more able you are to use and take advantage of learning by doing, the better off you will be. Another implication is that you should not be confined to excessively narrow bands of activity. Trade will, as I will shortly return to, greatly enhance your possibilities, as you will be able to specialize according to comparative advantage. Still, there is always a danger in specializing in mature products and lines of production. Learning by doing will pay off best if it is applied within a fairly broad spectrum of products. If a producer is highly specialized in, let us say, the production of a specific type of steel, the scope for learning by doing may be limited. If, on the other hand, the producer has a broader knowledge of manufacturing, let us say electrical equipment, and is engaged in various techniques for the transmission of electricity, the scope for improvement through gradual improvement and learning by doing may be large indeed. The development of Swedish industry with a limited number of multinational companies that have survived for a hundred years or more offers an interesting illustration of the new types of theorizing. KNOWLEDGE AND TRADE As we have seen the stock of knowledge is an important factor for determining a country’s level of welfare. Without too much exaggeration, we may say that a nation’s per capita income is determined by the amount of knowledge that the nation possesses. Hence, if a country is backward its complete productive capacity will be used to produce simple consumption goods. The level of education and research will also be low. The country will be unable to contribute on its own to the common stock of knowledge that exists. This is also a fundamental explanation of the fact that low-income countries have low growth rates and cannot break out of the poverty trap in which they are caught. The explanation in terms of learning by doing also fits nicely into this scheme. The low-income countries primarily produce simple consumption goods where the possibilities of further improvement via learning by doing have been exhausted. Referring to an earlier type of development theory, we may say that they are caught in a vicious circle from which they are unable to break out. It is quite easy to see how trade may have beneficial effects on growth according to the new knowledge models. This positive effect has at least two parts. Research is basically a good that is neither ‘rival’ nor ‘excludable’ according to the new jargon. If a person uses some new research result or a new

18

INSTITUTIONS, KNOWLEDGE, TRADE AND GROWTH

piece of knowledge, it is not thereby appropriated by that person or destroyed in the process. Nor is this type of knowledge ‘excludable’. It is hardly possible to exclude someone from the results of basic research. As the world economy becomes increasingly integrated one effect is that ‘double-work’ in research diminishes. If two reasonably large economies are integrated with each other the resources allocated to research in one country may produce the same amount of new knowledge that was produced in autarky by the resources of both countries. If the same amount of research is retained in both countries, the growth rate of knowledge and of new ‘designs’ will increase. This will have a positive effect on the growth rate. The effects on learning by doing will also be positive. Here the threat of diminishing returns is always present basically due to the fact that learning and instruction pays off best for the introduction of a new product. Since the opening up of trade or an expansion of trade means larger markets, the scope for scale effects and for the introduction of new products increases. International trade implies an exchange of knowledge. It also means that a country will gain access to the stock of knowledge capital through trade. Now it can be argued that human knowledge is not country specific. But on the other hand, it can also be argued that knowledge has no specific value until it is used. Trade theory is not taught in the same fashion all over the world. In fact the late Harry Johnson, when he once visited Lund, argued that trade theory was only taught in four or five places in the world. On that occasion Harry was in a good mood and was willing to include Lund on that list. Likewise South Koreans are probably much better at industrial espionage than North Koreans and the same probably holds for Swedes as compared with Norwegians. This does not primarily concern the fact that some people are more moral than others but simply illustrates the fact that if you are in the race and already producing and competing in the international market, the opportunities to improve through learning by doing are there and can usually be used quite easily to further your standing. CONCLUDING REMARKS Let me end then with two remarks of a more general nature. The first one is of a more technical nature. An important aspect of the new theories is that they make growth endogenous. This is essentially done by ‘the opening up’ of the production functions and by the introduction of non-convexities. The similarities between the more traditional growth and trade theory that some of us worked with in the 1960s and the new theories are not significant. The older theories basically explained the effects of growth on trade while the new ones are much more suited to explain the beneficial effects of trade on growth. In my 1964 book on growth and trade, I worked with what I called ‘unspecified’ production functions in the chapter on technical progress. It was

19

RECENT DEVELOPMENTS IN THE THEORY

demonstrated there that technical progress could well lead to increased marginal productivity for both factors of production. I thought that was a significant result. Unfortunately no one else seems to have thought so. Now with the new theories, a significant fact is that we can obtain increasing marginal productivities from the external effects of the accumulation of knowledge. The other remark of a broader nature concerns the fact that today there are three poles that completely dominate the world economy and world trade. They are the North American pole with the USA and Canada, the Western European pole and the East Asian pole with Japan and the four dragons. Together these three poles comprise some 800 million people who live in relative affluence. Outside is the other world of 4 billion people, the majority of whom live in abject poverty. It seems to me that the new theories I have mentioned provide the best explanation of these contemporary differences in living standards. One issue of long-standing debate in trade theory is that between an inwardlooking and an outward-looking trade policy—whether to opt for export push or import substitution. Most of us think that the arguments in favour of export push and openness are strong indeed. The traditional arguments have mostly been couched in static terms. A free trade policy forces a country to accept the prevalence of the international price structure. If a country wants to specialize according to its comparative advantage, it cannot tamper with the exchange rate but will have to find the ‘correct’ level for the exchange rate. The new type of theories provide new, strong arguments in favour of openness. The use of existing knowledge and the acquisition of new knowledge become factors of primary importance for a country’s well-being. For most countries, domestically produced technical progress usually forms a small, not to say negligible, part of global knowledge. But contacts with the rest of the world through trade are an excellent means to get access to new knowledge. Most countries in fact are forced to become free riders in terms of creating and acquiring knowledge. Restriction of trade will lead to backwardness, lack of knowledge and techniques, and in the end to poverty. The new theories also, in a way, vindicate and stress the importance of our old friend, the factor price equalization theorem. If a country wants to be among the world’s economic leaders, it will also have to be at the forefront of technological development and at least in some instances be the owner of exclusive designs and newly created knowledge. It is in fact only by owning some exclusive knowledge that a country can avoid having its factor prices equalized. Wages higher than the international average must be based on the ownership of superior technology. If we look at Swedish economic history, it can be used to illustrate many of the points of the new approaches. International trade certainly played a very important role in the years from 1870 to 1914 which were extremely important for Swedish development and provided the industrial structure that has dominated Swedish economic life up to the present time. Exchange with the leading

20

INSTITUTIONS, KNOWLEDGE, TRADE AND GROWTH

industrial nations was intense during this period and examples of imitation, importation of knowledge and learning by doing are numerous. During the last twenty years, the Swedish economy has not functioned so well. This is not primarily due to a slackening on the international trade front; it has more to do with an overexpansion of the welfare state and destruction of economic incentives. Here the institutional approach to economic history that I started to sketch takes on a considerable relevance. I am convinced that both the new approaches that I have briefly commented on will receive increased attention in the coming years. The implications of the new types of theorizing are important and well worth consideration. Research along these new avenues has only just started. We can in all probability expect new light to be thrown on both development issues and trade problems by using the new approaches in the years to come. BIBLIOGRAPHY Davis, Lance E. and North, Douglass C. (1971) Institutional Change and American Economic Growth, Cambridge University Press. Grossman, Gene M. and Helpman, Elhanan (1989) ‘Growth and welfare in a small open economy’, NBER Working Paper No. 2970, Cambridge, MA. ——(1990a) ‘Trade, innovation, and growth’, American Economic Review 80(2):86–91. ——(1990b) ‘Comparative advantage and long-run growth’, American Economic Review 80(4):796–815. ——(1991) ‘Trade, knowledge spillovers, and growth’, European Economic Review 35:517–26. Jones, E.L. (1987) The European Miracle, 2nd edn, Cambridge University Press. Lucas, Robert E. (1988) ‘On the mechanics of economic development’, Journal of Monetary Economics 22:3–42. ——(1990) ‘Why doesn’t capital flow from rich to poor countries?’, American Economic Review 80(2):92–6. Romer, Paul M. (1986) ‘Increasing returns and long-run growth’, Journal of Political Economy 94(5):1002–37. ——(1990a) ‘Endogenous technological change’, Journal of Political Economy 98(5): 70– 102. ——(1990b) ‘Are nonconvexities important for understanding growth?’, American Economic Review 80(2):97–103. Rosenberg, Nathan and Birdzell, L.E. (1986) How the West Grew Rich, London: Tauris. Södersten, Bo (1964) A Study of Economic Growth and International Trade, Stockholm: Almqvist & Wiksell. ——(1991) ‘One hundred years of Swedish economic development’, in M.Blomström and P.Meller (eds) Diverging Paths—A Century of Scandinavian and Latin American Development, Baltimore, MD: Johns Hopkins University Press.

21

2 THE NEW GROWTH THEORY, TRADE AND DEVELOPMENT Some brief reflections Pranab Bardhan

The theoretical literature on growth and trade flourished in the 1950s and 1960s, and Bo Södersten contributed handsomely to this. The literature first developed a framework for analysing the comparative statics of the effects of capital accumulation and exogenous technical change on a country’s terms of trade and balance of payments. Then in the second half of the 1960s the dynamic analysis of neoclassical growth models was extended to the case of the open economy. In the 1970s and almost up to the end of the 1980s the theoretical literature on growth and trade was somewhat inactive, except for the development of several NorthSouth models focusing on the impact on trade relationships of the various kinds of asymmetries between rich and poor countries (primarily in the structure of demand for importables and exportables, or in the labour markets). In the last few years there has again been a spurt in the literature flowing from the application of the so-called new growth theory following upon the leading contributions of Romer (1986, 1990) and Lucas (1988). The book by Grossman and Helpman (1991) is a major example of this application to an open economy. It is often said that the new growth theory has endogenized technical progress in contrast to the old growth theory, the central case of the latter being Solow’s (1956) growth model. This overlooks the tradition of endogenous growth in many of the growth models of the 1960s: apart from Arrow’s (1962) learning-by-doing model where learning emanated from the dynamic externalities of cumulated gross investment, there are the Kaldor-Mirrlees (1962) model where investment is the vehicle of technical progress and Shell’s (1967) model of inventive activity. Nor is the blurring of the distinction between capital accumulation and technical progress a new feature: it was the salient point of the Johansen-Solow type of vintage capital models as well as the many growth models of Kaldor. The idea of aggregate dynamic economies of scale in the form of the development of new inputs, expanding the productivity in the final goods sector using those inputs, which Romer borrows from Ethier (1982)—which actually goes back to Young (1928)—has in some sense already been formalized in vintage capital models where each new vintage of better

22

THE NEW GROWTH THEORY, TRADE AND DEVELOPMENT

machines expanded the range of higher-productivity inputs used in final goods production. I believe the major contributions of the new growth theory lie in combining all these with a tractable imperfect-competition framework which provides some (Schumpeterian) private motivation for investment in research and development. The major contribution of the open-economy models in this literature is to give us new insights on the effects of trade on growth. The East Asian success stories have given credence1 to the belief of many neoclassical economists in a positive relationship between ‘outward orientation’ and economic growth (although a rigorous empirical demonstration of the causal relationship between some satisfactory measure of outward orientation and the rate of growth is rather scarce). Standard neoclassical growth theory did not provide any such general theorem. There were three strands in the earlier theoretical literature which related trade policy to the rate of growth in a developing country. One is a simple extension of Solow’s model with an essential imported intermediate input, the growth in supply of which depends on the rate of growth of exports—see, for example, Khang (1968) and Bardhan (1970: ch. 4). In these models the steady-state rate of growth of the economy is different from the rate of growth of population and labour-augmenting technical progress, depending on the rate of growth of the country’s exports in the world market. The second strand,2 following an older tradition, formalized dynamic economies of scale associated with learning by doing (captured by the interfirm spillover effects of cumulated gross output) rationalizing an old argument for support of ‘infant’ industry producing import substitutes or new exports—see the models of Bardhan (1970: ch. 7) and Clemhout and Wan (1970). In the recent literature there have been two very important types of extension of this model. One is the model of Young (1991), which, on the basis of learning by doing that spills over across industries, although bounded in each industry, endogenizes the movement of goods out of the learning sector into a mature sector in which learning no longer occurs and thus gives a plausible account of an evolving trade structure. The other extension is by Krugman (1987) and Boldrin and Scheinkman (1988), where the learning effects enhance the existing patterns of comparative advantage; this may call for a deliberate trade policy that can orchestrate a breakout from such a historical ‘lock-in’. A similar model of hysteresis, based on selfreinforcing advantages not of learning but of headstarts in R&D, is developed in Grossman and Helpman (1991: ch. 8). The third strand in the earlier literature involved the effects of trade policy on the modernization of the capital stock, based on vintage capital growth models. The first models to link comparative advantage to endogenous differences in the economic life of capital between countries (lower wages in poorer countries allowing for the use of older, less productive, machines) were those of Bardhan (1966) and Bardhan (1970: ch. 5); Smith (1976) developed a more general model to include trade in second-hand equipment between rich and poor countries.

23

RECENT DEVELOPMENTS IN THE THEORY

Bardhan and Kletzer (1984) showed in terms of a simple vintage capital trade model with embodied technical progress how the question of a policy of protection accelerating or delaying modernization of capital stock (thus helping or dampening growth of labour productivity) depends, among other things, on the technological characteristics of the protected sector and does not have an unambiguous answer. The new growth theory literature has not yet significantly followed up on this strand of the earlier literature. In fact, borrowing as it does the Dixit-Stiglitz-Ethier ‘love of variety’ models in terms of a production functional where endogenous growth takes the form of simply extending the range of new inputs, it overlooks the endogenous process of the economic obsolescence of some inputs: it is not necessarily the width of the range of inputs that enhances productivity. A major result in the new literature is to show how economic integration in the world market, compared with isolation, helps long-run growth by avoiding unnecessary duplication of research. World market competition gives incentives to entrepreneurs in each country to invent products that are unique in the world economy—see the models of Rivera-Batiz and Romer (1991) and Grossman and Helpman (1991: ch. 9). One has, of course, to keep in mind the fact that sometimes these unique products are unique in the sense of product differentiation but not in the sense of any technological advance (it is well known, for example, that in the pharmaceutical industry a majority of the so-called new products are really recombinations of existing ingredients with an eye to prolonging patent protection, and that they are new, not therapeutically, but from the marketability point of view). Besides, the presumption in these models of a common pool of knowledge capital created by international spillovers of technical information is not often relevant for a poor country. When knowledge accumulation is localized largely in the rich country and the poor country is also smaller in (economic) size, the poor country may innovate less rapidly in long-run equilibrium with international trade than it does under autarky, as shown by Feenstra (1990) and Grossman and Helpman (1991: ch. 9). Trade may reduce the profitability of R&D in the poor country as it places local entrepreneurs in competition with a rapidly expanding set of imported, differentiated products and may drive the country to specialize in traditional, possibly stagnant, industries which use its relatively plentiful supply of unskilled workers. One should note that the relevant R&D for a poor country is, of course, more in the technological adaptation of products and processes invented abroad and in imitation. But even this kind of an R&D sector is usually so small that major changes in aggregate productivity and growth on the basis of the trade-induced general equilibrium-type reallocation of resources into or away from the R&D sector, as emphasized by Grossman and Helpman, will seem a little overdrawn if applied in the context of poor countries. In any case the ambiguity in the relationship between trade expansion and productivity growth in these general equilibrium models only confirms similar conclusions in careful partial equilibrium models, particularly when entry and exit from industries are not frictionless—see, for example, Rodrik (1992). The lack of diffusion of technology from rich to poor countries is often inter-

24

THE NEW GROWTH THEORY, TRADE AND DEVELOPMENT

preted in the literature as reflecting the frequent laxity in the enforcement of patents in poor countries and innovators in rich countries thus compelled to protect their ideas through secrecy. This, of course, ignores the cases of restrictive business practices of many multinational companies (like preemptive patenting and ‘sleeping’ patents where the old technology keeps on being used while the new patents are taken out simply to ward off competitors). Furthermore, the flow of technology through direct investments by multinational enterprises to a poor country is often constrained not so much by restrictive government policy in the host country as by its lack of physical infrastructure (the development of which in turn is constrained by the difficulty of raising large loans in a severely imperfect international credit market). In fact, while the new models of trade and growth bring into sharp focus the features of monopolistic competition particularly in the sector producing intermediate products and, in some models, the Schumpeterian process of costly R&D races with the prospect of temporary monopoly power for the winner— aspects which were missing in most of the earlier growth models—there are other important aspects of imperfect competition (like the case of ‘sleeping’ patents above or how international credit market imperfections shape the pattern of comparative advantage3) which need formalizing in the literature on trade and development. Finally, while the new literature has sharpened analytical tools and made our ways of thinking about the relationship between trade and growth more rigorous, it is high time that more attention is paid to the extremely difficult task of empirical verification of some of the propositions in the literature (I am leaving aside the largely vacuous cross-country regressions on the basis of very shaky, but easily available, international data that some of the new growth theorists have been playing with). While some beginnings have been made—see, for example, the study by Feenstra et al. (1992) on the basis of a sample of Korean industries— to confirm the hypothesis of the new growth models that the creation of new inputs generates continuous growth in total factor productivity, the evidence on the link between trade and productivity growth is still scanty and rather mixed. On the basis of a sample of semi-industrial countries in the World Bank project on ‘Industrial competition, productivity and their relation to trade regimes’, Tybout (1992) observes: ‘the lack of stable correlations (between trade and productivity) in sectoral and industry-level data is matched by a surprising diversity in the processes of entry, exit and scale adjustment’. The theoretical models, for all their recent enrichment, have a long way to go before they can catch up with the complexity of the empirical reality. NOTES 1

It should, however, be noted that the export boom in manufactures for Korea and Taiwan in the 1960s came before any significant trade liberalization. As Rodrik (1992) suggests, a realistic exchange rate policy and a generous programme of export

25

RECENT DEVELOPMENTS IN THE THEORY

2 3

subsidies, rather than trade liberalization per se, may be the key ingredients for successful export performance. The first and second strands have, in a sense, been combined and extended in the recent work of Quah and Rauch (1990). Kletzer and Bardhan (1987) show how more costly credit under imperfect information may drive a poor country away from specializing in sophisticated manufactured products which require more selling and distribution costs than traditional primary products.

BIBLIOGRAPHY Arrow, K.J. (1962) ‘The economic implications of learning by doing’, Review of Economic Studies June. Bardhan, P. (1966) ‘International trade theory in a vintage capital model’, Econometrica October. ——(1970) Economic Growth, Development and Foreign Trade: A Study in Pure Theory, New York: Wiley-Interscience. Bardhan, P. and Kletzer, K. (1984) ‘Dynamic effects of protection on productivity’, Journal of International Economics February. Boldrin, M. and Scheinkman, J.A. (1988) ‘Learning by doing, international trade and growth: A note’, in P.W.Anderson, K.J.Arrow and D.Pines (eds) The Economy as an Evolving Complex System, Reading, MA: Addison-Wesley. Clemhout, S. and Wan, H. (1970) ‘Learning by doing and infant industry protection’, Review of Economic Studies January. Ethier, W.J. (1982) ‘National and international increasing returns to scale in the modern theory of international trade’, American Economic Review June. Feenstra, R.C. (1990) ‘Trade and uneven growth’, NBER Working Paper no. 3276, Cambridge, MA. Feenstra, R.C. Markusen, J.R. and Zeile, W. (1992) ‘Accounting for growth with new inputs: Theory and evidence’, American Economic Review May. Grossman, G.M. and Helpman, E. (1991) Innovation and Growth in the Global Economy, Cambridge, MA: MIT Press. Kaldor, N. and Mirrlees, J.A. (1962) ‘A new model of economic growth’, Review of Economic Studies June. Khang, C. (1968) ‘A neoclassical growth model of a resource-poor open economy’, International Economic Review October. Kletzer, K. and Bardhan, P. (1987) ‘Credit markets and patterns of international trade’, Journal of Development Economics October. Krugman, P. (1987) ‘The narrow moving band, the Dutch disease, and the competitive consequences of Mrs Thatcher: Notes on trade in the presence of dynamic scale economies’ , Journal of Development Economics October. Lucas, R.E. (1988) ‘On the mechanics of economic development’, Journal of Monetary Economics July. Quah, D. and Rauch, J.E. (1990) ‘Openness and the rate of economic growth’, mimeo., University of California, San Diego.

Rivera-Batiz, L.A. and Romer, P.M. (1991) ‘Economic integration and endogenous growth’, Quarterly Journal of Economics May. Rodrik, D. (1992) ‘Closing the technology gap: Does trade liberalization really help?’, in G.Helleiner (ed.) Trade Policy, Industrialization and Development: New Perspectives, Oxford: Clarendon Press. Romer, P.M. (1986) ‘Increasing returns and long-run growth’, Journal of Political Economy October.

26

THE NEW GROWTH THEORY, TRADE AND DEVELOPMENT

——(1990) ‘Endogenous technological change’, Journal of Political Economy October. Shell, K. (1967) ‘A model of inventive activity and capital accumulation’, in K.Shell (ed.) Essays in the Theory of Optimal Economic Growth Cambridge, MA: MIT Press. Smith, M.A.M. (1976) ‘International trade theory in vintage models’, Review of Economic Studies February. Solow, R.M. (1956) ‘A contribution to the theory of economic growth’, Quarterly Journal of Economics February. Tybout, J.R. (1992) ‘Linking trade and productivity: New research directions’, World Bank Economic Review May. Young, A. (1991) ‘Learning by doing and the dynamic effects of international trade’, Quarterly Journal of Economics 106: May. Young, A.A. (1928) ‘Increasing returns and economic progress’, Economic Journal 38: 527–42.

27

3 WAGE DISPERSION, INTERNATIONAL TRADE AND THE SERVICES SECTOR Ronald Findlay

INTRODUCTION Although I first met Bo Södersten in Sweden in the mid-1970s, our intellectual paths first crossed at MIT in the late 1950s and early 1960s, where both of us worked on the hot topic in the field at that time, which was trade and growth. Hicks’ famous inaugural lecture on the long-run dollar problem had inspired Harry Johnson, and his work in turn inspired several of the newcomers entering the field at that time, including Bo and myself. It was an exciting time. Bo’s subsequent dissertation, published as Södersten (1964), is the most thorough and detailed treatment of that stage of the literature available in print. Not only did he pull together the results based on the two-sector neoclassical model but also he related this work to the ‘development’ theories of Arthur Lewis, Raul Prebisch and others. He also linked the modern discussion to classical doctrine. This work can still be read profitably today by any student of international trade. Three decades have passed since then and many new methods and ideas have come and gone in the field of international trade. It seems to me, however, that the approach used in that early literature, the comparative statics of simple, stylized general equilibrium models, is still a powerful tool for dealing with the problems of today in the areas of trade and growth. To demonstrate this point, I will devote this essay to an attempt to show how a simple extension of the ‘workhorse’ model of the 1950s and 1960s can be used to throw light on several notable phenomena in recent economic experience. A striking pattern that has emerged in the US economy over the last decade or so is a sharp rise in the rate of return to education, reflecting a rise in the reward of skilled labour and a relative or even absolute decline in the real wage of unskilled labour. A considerable amount of research by labour economists, perceptively summarized and extended by Mincer (1991), has attempted to explain this by reference to the decline in the relative price of manufactured goods, particularly labour-intensive ones, as a result of the surge in imports from the newly industrializing countries and to other internal factors such as

28

WAGE DISPERSION AND INTERNATIONAL TRADE

technological progress and the relative growth of the services sector, which manifests itself in production, employment and exports. The role of international trade, in particular, has been stressed in an important paper by Murphy and Welch (1989). Owing to the unfortunate fragmentation of economics as a whole that specialization is bringing about, the fields of labour economics and international trade do not appear to have much formal contact despite the fact that international trade and income distribution are so closely linked. The following is a modest attempt to bridge the gap. THE MODEL general equilibrium model, with some The model that I propose to use is a special features that will render it particularly tractable. It is a hybrid with both Heckscher-Ohlin and Viner-Ricardo components, a variant of a model originally proposed by Fred Gruen and Max Corden (1970) and subsequently used by Anne Krueger (1977) in her well-known Graham Lecture. The application of the model to issues of human capital formation and international trade follows my earlier work in Findlay and Kierzkowski (1983) but the introduction of a ‘services’ sector is a new feature.1 The manufacturing sector produces two goods, X and Y, with two factors capital and labour, in the familiar Heckscher-Ohlin manner. Both production functions exhibit constant returns to scale, and X is more capital intensive than Y at any wage-rental ratio. The ‘services’ sector produces output as a constant returns function of two inputs also, but these are ‘skill’ and labour. It would be easy to introduce a specific form of physical capital into the service sector also but that will not be done explicitly.2 The production side of the model, that is to say the technology and the factor requirements and supplies, is therefore (3.1) (3.2) (3.3) (3.4) (3.5) (3.6)

Initially I assume that all three factors are in exogenously fixed supply. A later section will indicate how L can be ‘transformed’ endogenously into S through a process of education or human capital formation. Choosing X as the numeraire, let p denote the price of the labour-intensive manufactured good Y and π the price of Z, the composite output of the services sector. Let p be taken as determined by world market conditions, unless otherwise indicated. Since the manufacturing sector has a Heckscher-Ohlin

29

RECENT DEVELOPMENTS IN THE THEORY

structure with two inputs and two outputs the Stolper-Samuelson theorem applies, giving us (3.7) (3.8)

where w and r denote the real wage and the rental on capital respectively. Figure 3.1 is the Lerner diagram that depicts the relation between relative product prices, sectoral capital-labour ratios and the rewards of the factors. In competitive equilibrium if both X and Y are produced their factor costs must be equal, a condition represented by the common tangent to the isoquants for 1 unit of X and 1/p units of Y. The capital-labour ratios kX and kY are the slopes of the rays from the origin through the tangency points on the respective isoquants. These capital-labour ratios uniquely determine the marginal products of each factor in each good, and so give us w and r uniquely determined by p and the two production functions.

Figure 3.1

Let us now look at the services sector. Since the supply of the skilled labour input S is fixed we know that the marginal physical productivity of unskilled labour LZ in the Z sector is diminishing. Given any price π of Z relative to the numeraire X, profit maximization in the services sector will require (3.9)

Since the marginal productivity of LZ is diminishing, raising π must result in raising LZ, and hence Z, for any value of w. We thus obtain (3.10)

30

WAGE DISPERSION AND INTERNATIONAL TRADE

as the supply curve of the output of services as a function of its price relative to manufactures, with the expected positive slope. The output of our services sector is a conglomerate of not only non-tradable domestic services of the usual type but also professional and financial services that are tradable on world markets as well. As is well known, export of services is growing faster than export of goods in the USA and so this is an increasingly important component of US foreign earnings. We therefore specify two components, a domestic and a foreign, to the total demand for services, each one negatively related to price but positively related to the respective level of national income. We therefore have (3.11)

where Dz is total demand consisting of the home and foreign components and Ih and If are the corresponding income levels. If is exogenous but Ih is given by (3.12)

where v (π) is the reward per unit of the skilled factor. It depends upon p because (3.13)

is the marginal value product of the skilled input. It varies with π positively both because π is the coefficient of the marginal physical product of skilled labour and because this variable is an increasing function of LZ that itself varies positively with π. The equilibrium output of the services sector, and thus the equilibrium of the system as a whole, is determined by the requirement that (3.14)

which is the market-clearing condition for the service sector. Given If and of course p and the endowments K, L and S, both sides of (3.14) are functions only of π. The equilibrium price that clears the market for Z is denoted π *. The equilibrium of demand and supply in the services sector is depicted in Figure 3.2. In Figure 3.3 the downward sloping curve is the marginal value product of LZ evaluated at π*. The horizontal line corresponding to the wage w(p) determined in the manufacturing sector is the perfectly elastic supply curve of labour to the services sector. The intersection determines the equilibrium employment in the services sector corresponding to the equilibrium output of services Z*. This implies that equilibrium employment in the manufacturing sector is (3.15)

Define (3.16)

Since the capital-labour ratios kX and kY in X and Y are already known we can determine the fraction λ* of manufacturing employment in X from the equation

31

RECENT DEVELOPMENTS IN THE THEORY

Figure 3.2

Figure 3.3

(3.17)

The value λ* of λ that satisfies (3.17) gives us the equilibrium allocation of capital and labour between X and Y and hence the outputs X* and Y* of the two manufactured goods as well. Since the economy is an exporter of services balanced trade implies that it is a net importer of manufactured goods. We will assume that the labour-intensive good Y is definitely imported, while the capitalintensive good X could be exported or imported, depending upon the particular

32

WAGE DISPERSION AND INTERNATIONAL TRADE

application. Surpluses or deficits in the trade balance, on both ‘visible’ and ‘invisible’ trade, could be accommodated easily if they are specified exogenously. The Viner-Ricardo component of the model makes it extremely flexible so long as the simplified production structure is adhered to. Thus we may introduce a primary sector into the model with a specific input, ‘land’, that also draws on the common pool of unskilled labour. Domestic and foreign demands for the primary output could be specified and the equilibrium price, output and employment determined, along with the rental of land, in a manner entirely analogous to what we have done with services and skilled labour. A fall in the relative price of labour-intensive manufactures The first problem that we will investigate is the impact of a fall in p, the relative price of the labour-intensive manufactured good. From Figure 3.1 we see that more units of Y will now be equal in value to 1 unit of X than before. This will cause the common tangent to become flatter, leading to kX and kY both falling and thus to a fall in w and a rise in r. Turning now to Figure 3.3 the fall in w means that the horizontal supply curve of labour to the services sector shifts downwards. If π* is held constant this will imply an increase in employment LZ of unskilled labour to an extent equal to the elasticity of the marginal productivity curve. Thus, at each given value of p the fall in w leads to a rise in LZ and therefore of Z. The supply curve Z(π) in Figure 3.2 is therefore shifted to the right. Neglecting income effects on demand, which will be small relative to the impact on supply, the effect will be to lower π* and to increase and Z*. We now investigate what happens to v*, the reward of the skilled factor in the service industry. The effect can be seen to be ambiguous, depending upon the relative magnitudes of the decline in π* and the rise in induced by the fall in w, as indicated in equation (3.13). The more elastic is the world demand DZ for Z, the smaller will be the decline in π and the more favourable the impact on v*. In the limiting case of perfectly elastic world demand, so that π* is constant, v* must rise because of the increase in . The ratio of the rewards to skilled and unskilled labour is (3.18)

which must rise since π cancels out and the numerator increases while the denominator decreases because of the increase in LZ. The rate of return or education, by any of the standard formulae in the ‘human capital’ literature, is an increasing function of the relative reward to skilled labour and so will rise as a result of the fall in p, even if v* does not rise in absolute terms. The rise in implies a fall in L*m and hence a rise in κ∗m and λ*. Thus X* increases and Y* contracts because of the Rybczynski effect, in addition to the

33

RECENT DEVELOPMENTS IN THE THEORY

relative price effect of the fall in p with held constant. The Rybczynski effect associated with the withdrawal of labour from manufacturing to the services sector thus reinforces the impact of the relative price shift between the two sectors at the original level of employment. An increase in the stock of capital An increase in the stock of capital will leave w and r unchanged, since these depend only upon p so long as the manufacturing sector produces both goods X and Y in positive amounts. It does, however, lead to an increase in home national income Ih by the amount r times the increase in the stock of capital. The demand curve for services is shifted to the right while the supply curve Z(π) remains unchanged. As a result π*, , Z* and v* all rise in response to the increase in K. The rise in leads to a rise in and therefore to an increase in X* and a fall in Y* by the Rybczynski effect. The rise in v* with w constant implies a rise in the rate of return on education. Export of services will fall because of the rise in price induced by the increase in home national income. The increase in the output of the capital-intensive good X*, however, means that exports will rise or imports will fall, depending upon the direction of trade in this good. An increase in the supply of labour An increase in the supply of labour has results very similar to an increase in the supply of capital as far as the output and price of services, and the reward of skilled labour, is concerned. National income rises by the amount w times the increase in labour. This shifts the demand curve for services to the right, leaving the supply curve unchanged. There is a rise in π*, , Z* and v* as a result. The increase in , however, will be less than the increase in L, so will rise and will fall. This will reduce the output of the capital-intensive good X* and raise that of the labour-intensive good Y*. Exports of X will fall or imports of X will rise, depending on the direction of trade. Imports of Y will fall because the rise in supply due to the Rybczynski effect will be larger than the increase in domestic demand. An increase in the supply of skill An increase in the supply of skill affects both the demand and supply sides of the market for services. At constant p the increase in S will require a proportionate increase in LZ to keep the marginal product of labour in the Z sector equal to the real wage w. Thus the supply curve Z(π) is shifted to the right because of the proportionate expansion in both inputs. National income at constant π will increase by v* times the increase in S. This will shift the demand curve DZ to the right, but by less than the shift in supply, leading to a fall in π*. The fall in LZ

34

WAGE DISPERSION AND INTERNATIONAL TRADE

induced by the decline in π * means that some of the initial rise in L Z, proportionate to the increase in S, will be offset. Since both π* and the marginal physical product of skilled labour fall (because LZ has not risen commensurately with S), it follows that v* must fall, as well as the rate of return on education, because w is constant. The rise in means that falls and rises. Thus the Rybczynski effect on X* is positive and on Y* is negative. Imports of the labour-intensive good Y must rise and exports of X rise or imports of X fall. The export of services must rise. Technological progress in the capital-intensive sector Consider now the effect of technological progress in the capital-intensive manufacturing sector X. For simplicity let this be restricted to the ‘Hicks-neutral’ type, i.e. an upward shift in the output level associated with a given isoquant. In terms of Figure 3.1 this means that the isoquant for 1 unit of X will be shifted towards the origin, with the tangent at each point where it is intersected by a ray from the origin remaining parallel. In other words, both marginal productivities increase proportionality at any given capital-labour ratio so that the marginal rate of substitution remains the same.3 Since the relative price p between X and Y is still fixed by world market conditions, the common tangent to the 1X and (1/p)Y isoquants must now be flatter, indicating a fall in the wage-rental ratio and a decline in the capital-labour ratio kX and kY in each sector of manufacturing. The return to capital r therefore rises because of both the technical progress and the decline in kX, but the effect on the real wage w appears to be ambiguous since the technical progress raises it and the decline in kX lowers it. It can be seen, however, that the real wage w must in fact fall because of the condition that

(3.19)

The decline in kY means that ∂Y/∂LY has fallen and since p is constant this must imply that ∂X/∂LX and therefore w has fallen, despite the Hicks-neutral technical progress in the X sector. The fall in w induces an expansion in LZ at constant π and so shifts the supply curve of services Z(π) to the right. The technical progress constitutes an increase in real income at home, and so shifts the demand curve DZ(π) to the right as well, but not as much as the supply curve. The equilibrium outcome is therefore that π* falls while and Z* rise. The rise in raises v* while the fall in π* lowers it. The effect on the absolute level of v* is therefore ambiguous. By the some reasoning involved in the effect of a decline in p, however, we see from equation (3.18) that the ratio of v* to w in The services sector must rise and so the effect is to raise the rate of return to education. The output of X* rises because of the impact of the technical progress itself as

35

RECENT DEVELOPMENTS IN THE THEORY

well as the Rybczynski effect associated with the rise in . The output of Y* falls.

and consequent rise in

Technological progress in the services sector The effect of technological progress in the services sector will generally be to shift the marginal productivity of labour curve in Figure 3.3 to the right, holding π* constant at its initial level. The technical progress itself, as well as the associated increase in LZ, shifts the supply curve of services Z(π) to the right. The demand curve also shifts to the right because of the increase in domestic real income, but by less, so that the effect is for π* to fall but for , and Z* to expand. It is easy to show that the ratio of v* to w must increase and since w is constant as determined by the unchanged conditions in the manufacturing sector v* must go up. The rate of return to education therefore unambiguously rises. The withdrawal of more labour from the manufacturing sector raises X* and lowers Y*. Exports of X will increase or imports decrease, while imports of Y will rise. Exports of Z will increase. An increase in foreign national income Economic expansion abroad will shift the demand curve DZ(π) for services to the right leading to a rise in π*. This will lead to an increase in and Z* along the unchanged supply curve Z(π) of the services sector. The repercussion of the rise in on the manufacturing sector will be for X* to rise and Y* to fall, with the now familiar consequences for exports and imports of X and Y. Exports of Z will of course increase. There will be a definite increase in v* since both π* and rise. Since w remains constant the rate of return to education must rise. Protection of the labour-intensive manufacturing sector A tariff on the import of labour-intensive manufactures will raise the domestic price of X relative to Y above the world price p. In terms of Figure 3.1 this will mean that fewer units of Y will now be equal in value to 1 unit of X. The common tangent to the equal-value isoquants will therefore be steeper, indicating a rise in the real wage w and a fall in the rental on capital r. In Figure 3.3 the rise in w will mean an upward shift in the horizontal supply curve of labour to the services sector, a nd a consequent decline in LZ if π* is held constant. The supply curve of services Z(π) is therefore shifted to the left as a result of the protection for labour-intensive manufactures. The demand curve will also be shifted to the left since domestic real income falls because of protection, but this effect will be small relative to the shift in supply. The effect will therefore be for π* to rise and for and Z* to fall. Since π* rises and falls the effect on v* is ambiguous. The same argument as used earlier in connection with equation (3.18) indicates, however, that the fall in

36

WAGE DISPERSION AND INTERNATIONAL TRADE

means that the ratio of v* to w must fall. The rate of return to education must therefore fall as a result of giving protection to the labour-intensive manufacturing sector. The reduction in means that employment in manufacturing will rise. Output of the labour-intensive good Y will rise both because of the protection itself and the Rybczynski effect induced by the increase in . Output of X will correspondingly fall for both these reasons. Thus exports of Z and imports of Y will fall while exports of X will fall or imports rise depending upon the direction of trade in this commodity. ENDOGENOUS SUPPLY OF SKILL The model has so far been based on the assumption that the supplies of skill and raw labour, S and L, are each fixed. In reality, however, there is always an opportunity for raw labour to acquire skill through a costly process of human capital formation. A general equilibrium model of this process in the context of production and trade in a two-country world economy is presented in Findlay and Kierzkowski (1983). The reader is referred to this paper for the somewhat intricate details. In this section I shall briefly sketch how the extension to endogenous skill formation can be carried through and note the implications that follow. It is convenient to make a number of simplifying assumptions. All agents have uniform ability; N of them are ‘born’ at each instant and live for exactly T years. They can either work as unskilled labour earning the wage w as specified in the model, or spend a period of time θ acquiring an ‘education’, after which they can spend the remaining (T–θ) years of their life earning the skilled wage v*, also as specified in the model of the present essay. The sole cost of education is therefore the ‘foregone earnings’ that are sacrificed by anyone who chooses to become a skilled worker. Since all individuals have equal ability and there is perfect competition the lifetime earnings of both categories of workers must be equal when the system is in equilibrium. The excess of v* over w during the (T–θ) years of a skilled worker’s life must be just enough to compensate him or her for the loss of w over the time needed to acquire the skill. The equating of lifetime earnings is readily seen to imply that (3.20)

For given values of v* and w equation (3.20) can be taken to define δ as the implicit rate of return to education. In our model w is determined as a function of the parameter p, while v* is endogenously determined along with π* for given values of S and L. It has been demonstrated that (3.21)

37

RECENT DEVELOPMENTS IN THE THEORY

Figure 3.4

In Figure 3.4 the downward sloping curve RR´ shows the relation between the rate of return to education δ and ‘skill’ S. It slopes downward because raising S lowers v*, which from (3.20) must imply a lower δ since w(p) is fixed. The rise in S implies a lower L which also leads to a fall in v* and thus in δ. Suppose now that there is a constant rate of interest, equal to the rate of time preference ρ, in this economy. The level of skill S* can then be determined endogenously by the condition (3.22)

where S* is the unique value of S that satisfies (3.22). The corresponding value L* of unskilled labour can readily be determined from the given total population. In Figure 3.4 S* is determined by the intersection of the negatively sloped curve RR’, depicting δ(S), with the horizontal line indicating the fixed value of ρ. It has already been shown that the effect of technological progress in the capital-intensive sector, an increase in the capital stock and foreign national income in each case is to raise the rate of return to education for a given level of skill. Hence the RR´ curve is shifted upwards at each value of S and so the effect for a fixed value of p is to increase the endogenous skill level S*. A fall in p, the relative price of the labour-intensive manufactured good, also shifts RR´ upward and hence eventually raises S*. From equation (3.20) we see that if the rate of return to education δ is equal to the fixed rate of interest ρ then a fall in the real wage of unskilled labour, due, say, to a fall in the exogenous value of p reflecting increased foreign competitiveness,

38

WAGE DISPERSION AND INTERNATIONAL TRADE

would eventually lead to such a strong supply response of the skill level S* that v* must eventually fall in the same proportion as w(p). Thus the phenomenon observed in recent years of a decline in w(p) and a rise in v* is a disequilibrium phenomenon in the labour market that will induce an adjustment that leads to a proportionate decline in v* as well. The ‘good news’ of course is that the adjustment is made through a higher level of skill S and a lower level of raw labour L so that the average wage can go up, i.e. a higher proportion of the labour force gets the higher wage v* even though both v* and w have fallen. The rewards of physical capital also go up so the economy as a whole is better off as a result of the technical progress and of cheaper imports of labour-intensive goods. APPLICATION OF THE MODEL: US EXPERIENCE IN THE 1980s Thus far we have simply constructed a general equilibrium model and examined its comparative static properties. The motivation of the exercise, however, as stated in the introduction, is to see whether the model can be of any help in assessing the remarkable behaviour of wage differentials in the US economy during the 1980s. In addition to the basic papers by Mincer (1991) and Murphy and Welch (1989) cited earlier that provided the inspiration for the present effort, there is a very thorough and detailed evaluation of the relevant empirical findings by Bound and Johnson (1991).4 While there are many problems with identifying even highly ‘stylized’ facts with comparative statics exercises in simple general equilibrium models, it nevertheless seems clear that our model can account very satisfactorily for the behaviour noted and the explanations put forward by Mincer and Bound and Johnson. They find technical change to be biased towards skill, and to be the main factor associated with the rise in the relative wage of skilled workers, along with shifts in the composition of output towards skill-intensive sectors like financial and professional services and away from traditional blue-collar occupations like manufacturing. In terms of our model this is entirely consistent with the impact of (a) technological progress in the services sector itself and (b) technological progress in the capital-intensive manufacturing sector. The impact of a fall in the relative price of labour-intensive manufactured imports, emphasized by Murphy and Welch, works in the same direction. The increase in the supply of skilled workers noted in both papers as a mitigating factor on the extent of the change that would otherwise have taken place is also of course consistent with the results of our model on an increase in the relative supply of skilled workers. Growth in foreign demand for US service exports, while not considered by Mincer and by Bound and Johnson, is another factor that could have been working in the same direction, as implied by the results of our model. The surge in immigration, including illegals, which was probably biased in favour of unskilled workers, would also be consistent with raising the relative wages of skilled workers. In conclusion the case for a ‘dialogue’ between simple general equilibrium models of the trade theory type and the rich empirical tradition of the literature on

39

RECENT DEVELOPMENTS IN THE THEORY

human capital and labour economics seems to be well supported by our consideration of the wage dispersion phenomenon of the US economy during the 1980s. ACKNOWLEDGEMENTS This essay was largely written while I was a visiting professor at the Graduate Institute of International Studies, Geneva. I am grateful to Professor Hans Genberg and other colleagues at the Institute for their hospitality and assistance. I also thank my colleague Jacob Mincer for encouragement and advice while absolving him of responsibility for remaining defects. NOTES 1 2 3 4

See Bhagwati (1984) and Djajic and Kierzkowski (1989) for discussions of international trade in services as distinct from goods. Mincer (1991:17) states that ‘service industries are the major employer of educated workers’. See Findlay and Grubert (1959) and chapter 4 of Södersten (1964) for a more detailed analysis of the impact of technological progress on factor and goods prices in general equilibrium. See also the paper by Alan Krueger (1991) on the impact of computers on the wage structure.

BIBLIOGRAPHY Bhagwati, J.N. (1984) ‘Splintering and disembodiment of services and developing nations’, World Economy, 7:133–44. Bound, J. and Johnson, G. (1991) ‘Changes in the structure of wages in the 1980’s: An evaluation of alternative explanations’, mimeo., July. Djajic, S. and Kierzkowski, H. (1989) ‘Goods, services and trade’, Economica, February. Findlay, R. and Grubert, H. (1959) ‘Factor intensities, technological progress and the terms of trade’, Oxford Economic Papers February. Findlay, R. and Kierzkowski, H. (1983) ‘International trade and human capital: A simple general equilibrium model’, Journal of Political Economy December. Gruen, F. and Corden, W.M. (1970) ‘A tariff that worsens the terms of trade’, in I.A. MacDougall and R.H.Snape (eds) Studies in International Economics, Amsterdam: North-Holland. Krueger, A.B. (1991) ‘How computers have changed the wage structure: Evidence from microdate 1984–89’, mimeo., June. Krueger, A.O. (1977) ‘Growth, factor market distortions and patterns of trade among many countries’, Princeton Studies in International Finance no. 40. Mincer, J. (1991) ‘Human capital, technology and the wage structure: What do time series show?’, Working Paper no. 3581, NBER, January. Murphy, K. and Welch, F. (1989) ‘Wage differentials in the 1980’s: The role of international trade’, mimeo. Södersten, B. (1964) A Study of Economic Growth and International Trade, Stockholm: Almqvist & Wiksell.

40

Part II LATIN AMERICA

4 SPAIN AND THE CONQUEST OF AMERICA Profits, religion and forced labour in the fifteenth and sixteenth centuries Mats Lundahl

INTRODUCTION Exploitation and dependency are out of fashion among contemporary development economists—for obvious reasons. Structuralist, neo-Marxist and dependencia theorists never managed to construct a coherent framework of analysis that could be used to explain why some less developed countries did better than others (cf. e.g. Hunt 1989: chs 5–7). Group interests, on the other hand, are ‘in’. A great effort has been made to explain the existence of nonoptimal (in traditional welfare terms) policies and institutions, such as tariffs, by theories that are based on the explicit assumptions that conflicts exist between different economic actors, all pursuing their self-interests, and that governments act to balance these group interests in such a way as to maximize the likelihood of remaining in power (cf. e.g. Baldwin 1982, Findlay and Wellisz 1982, 1983, Becker 1983, Frey 1984, Magee et al. 1989). The present essay represents an attempt to link some of the elements that are currently ‘out’ with some of those that are ‘in’, within a trade context. The topic is an historical one: the Spanish conquest of America. How can some of the main features of the conquest be explained in economic terms and what were the main effects in America during the first century?1 These are questions that cannot be answered with the aid of the standard trade theory framework. The traditional theory of international trade basically deals only with Pareto-sanctioned behaviour. By concentrating on such products where the country has a low comparative cost of production, trading them for products which are more costly to produce at home, more goods can be made available to consumers at a lower real cost to the economy. In this sense, trade is superior to autarky. As is well known, this conclusion builds on the crucial assumption that the groups which stand to gain from the opening of international trade do also in fact compensate the losers. Most of the time, however, no actual compensation takes place. Thus, the opening of trade produces losers as well as gainers. But this

43

LATIN AMERICA

is not where the story of the opening of trade ends. The historical experience of how trade relations with other continents were initiated by Europe is quite eloquent on this point. What the Europeans found were undeveloped economies that did not adapt easily, in the way predicted by the standard comparative cost doctrine. These economies had to be opened to trade in a ‘non-textbook’ fashion. The driving forces were non-neoclassical and the results had little to do with the predictions of standard theory. Nowhere was this as clear as in the case of Spain and America. Any reasonably realistic account of how trade was opened and of the impact that this had in the American continent must make allowance for the following facts, (a) The opening of trade cannot be separated from the act of conquest. It was not a peaceful event, (b) The forms that trade took during and after the conquest-most importantly, exports of bullion from the New World—were a direct outcome of events in Europe: the Reconquista and the fiscal crisis in Spain, (c) A steady flow of exports from the Americas could not be obtained without a thorough institutional reorganization of the indigenous society, (d) The new institutional forms were the result of a (partly zero-sum) game between three European actors: the Crown, the Spanish settlers and the Church, (e) The motives behind the conquest were not only economic but also religious. In the following, an attempt will be made to examine all these elements in order to derive what is largely an economic explanation of the origins of trade between Spain and its American colonies and the important consequences of this trade for the Americas. Let us begin with the ‘European’ background. THE FISCAL CRISIS IN SPAIN

The Catholic Kings The Middle Ages in Spain were characterized by continuous border fights and major campaigns against the Moors. The country’s history in the fifteenth century is also heavily coloured by military events. First, there was internal strife which did not end until Spain was unified under the Catholic Kings. Ferdinand and Isabella were married in 1469. Isabella became Queen of Castile in 1474 and Ferdinand King of Aragon five years later. The Portuguese threat to Castile was finally eliminated a few years thereafter. Granada fell in 1492 and Spanish Navarre was incorporated in 1512. ‘It would be surprising if this rapid unification of Spain had not created the necessity for a mystique of empire’, writes Fernand Braudel (1975:670).2 Hence Spanish armies became engaged outside the Iberian peninsula. The Canary Islands had been successively drawn into the Spanish sphere of influence during the entire century and were incorporated with Castile in the 1490s. From these islands, the Spaniards periodically raided the African coast for slaves and gradually deepened the involvement of the country in that continent. Spain was

44

SPAIN AND THE CONQUEST OF AMERICA

also militarily engaged in Naples, particularly after the fall of Granada and the expulsion of the Spanish Jews in the same year. After a war with the French, the city came under Spanish sovereignty in 1504. The aggressive foreign policy of the Catholic Kings, which was largely based on the decisions of Ferdinand but on the resources of Castile, was a joint foreign policy that drew both on the Reconquista traditions of Castile and on the Mediterranean aspirations of Aragon (Kamen 1983:8). However, the Spanish venture had other driving forces as well. The Reconquista, and in particular the final struggle for Granada, had deep roots in the tradition of the crusades. Spain was defending the ramparts of the Christian faith, and this tradition was continued by the very process of expansion (Braudel 1975:671–2). Financial support for military activities on this scale was difficult to raise. Spain was in fact a poor country in the late fifteenth century.3 In Castile, ordinary state income rose from 151 million maravedís4 in 1481 to 269 million in 1496 and 320 million in 1510 while at the same time extraordinary revenue grew from 53 million in 1483 to 112 million in 1504 (Ladero Quesada 1970:783–4). In spite of the rise, however, it proved difficult to defray royal expenses. Henry Kamen (1983:49) summarizes the situation: the cost of maintaining royal residences and of ceremonial for ambassadors rose from 8 million maravedis in 1480 to 35 million in 1504; and military expenses within the peninsula amounted to over 500 million between 1495 and 1504. Foreign wars and royal marriages were exorbitantly expensive: the first expedition to Naples cost 88 million, the second 366 million; and arrangements for the journey to England and marriage of Catherine of Aragon cost 60 million. The growing diplomatic importance of Spain was also a burden: in the last ten years of her reign Isabella spent 75 million on ambassadors. The Catholic Kings had to carry out their undertakings going into debt. Since they lacked a recognizable economic policy, they were unable to create a consistent fiscal programme that would have yielded stable revenues (Elliott 1970:111, Kamen 1983:48, 51). The chief source of revenue, accounting for 90 per cent of revenue during the fifteenth century, was the 10 per cent sales tax known as the alcabala. Important groups in Spanish society were, however, exempt from this tax. Consequently the revenue it generated proved insufficient and had to be supplemented by taxes on the Mesta, the sheepowners’ guild, in exchange for privileges granted to this organization (Kamen 1983:48–51).5 The discovery of America by Columbus in 1492 did not bring any immediate financial relief. The island of Hispaniola, the only one penetrated to any significant extent during the reign of the Catholic Kings, did not yield much in the way of wealth. Placer mining of gold quickly came to an end there and the looting of golden treasure accumulated by the native peoples of the western part of the Tierra Firme never took place on a scale which was large enough to meet the

45

LATIN AMERICA

hopes of crown and conquistadores alike.6 Columbus and his followers had failed to find the quick route to Asia, and the wealth obtained in the Indies had not compensated for this failure. In 1510 Ferdinand had pointed to the New World as a source of wealth for more important ventures, e.g. the war in Africa (Hillgarth 1978:583). However, achievements fell far short of aspirations. What the Spaniards left behind certainly did not inspire much hope for future improvements. The discoveries of the great Aztec and Inca empires were just around the corner, but at the time it looked as if the New World could easily become more of a liablity than an asset in the future. Charles V This was the situation when Charles V took over in 1516. Charles faced the same type of problems as the Catholic Kings, only on a much larger scale.7 The country was involved in a conflict with France over Navarre, Burgundy and Italy. The struggle against the Muslims, viewed by Charles as a divine mission, continued in North Africa and in the Balkans, and when Protestantism began to spread throughout northern Europe, Charles, as German emperor, became involved there as well. The territory of the Netherlands was extended on two occasions. At the beginning of the reign, from 1517 to 1522, local revolts continued to take place in Spain itself, culminating in the Comuneros uprising of the Castilian cities. At the same time the Spanish expansion in America had started. In 1519, Cortés left Cuba for Mexico and in 1531 Pizarro left Panama for the conquest of the Inca empire in Peru. All this cost money. In theory each of Charles V’s territories should have been self-supporting (Kamen 1983:85). In practice, however, the large and growing scale of international commitments forced him to draw on the resources of all parts of his realm simultaneously. During the early part of his regency, first the Netherlands and then Italy were taxed. However, this provoked a tax revolt in Ghent in 1539 and a vigorous protest from the Spanish viceroy at Naples during the following year. The budget surpluses of Sicily, Naples and Milan—wealthy states—were all but wiped out (Braudel 1975:673). This put the main burden on Spain itself: Castile, Aragon, the Church and the New World (Lynch 1964:53–8, Elliott 1970:199–207, Kamen 1983:85–90).8 Aragon provided comparatively little, its laws forbidding higher taxation. The Church was in principle exempt from taxation, but still had to provide its share both of ordinary and extraordinary state revenue and ended up by being a major contributor. Castile was easier to tax than Aragon, since most taxes did not have to be passed by the Cortes. Even so, it proved difficult to increase them. The yield of the alcabala increased with inflation, but Ferdinand and Isabella had already allowed some towns to compound an encabezamiento, equivalent to the alcabala yield, and pay that instead. In 1534, Charles made this a general practice. That measure, however, made it easier for the Cortes to oppose tax increases, with the result that ordinary government revenues could not keep pace with inflation. Direct taxation had to complement the indirect one.

46

SPAIN AND THE CONQUEST OF AMERICA

This only left America. Whereas the Antilles had yielded only small amounts of precious metals, the continental mainland proved to be able to turn out a never ending stream of bullion. ‘During Charles’ entire reign over 15 million ducats’10 worth of bullion reached the imperial treasury (Kamen 1983:88), especially after the discovery of the silver of Potosí in Upper Peru (present-day Bolivia) and Zacatecas, Guanajuato, Taxco, Pachuca, Sombrerete and Durango in New Spain (Mexico) in the 1540s and 1550s (Wolf 1982:135). All these combined revenue sources did, however, not suffice. 11 The commitments of Charles V grew rapidly. At the same time, the price revolution had started. In Castile, prices doubled during the first half of the sixteenth century (Kamen 1983:99). The result of these two developments was the creation of a persistent budget deficit. The only way out of this predicament was borrowing. Annuities known as juros, already employed by the Catholic Kings, were used to this end, with the result that a large part of government income was in fact mortgaged, increasing from 36 per cent of normal revenue in 1522 to 68 per cent in 1556 (Castillo Pintado 1963:51). 12 In addition, and quantitatively more important, a foreign debt had been accumulated. The imperial treasury turned to foreign bankers—Germans and Italians—at ever higher interest rates, increasing from 17.6 per cent in the 1520s to a ‘horrifying’ 48.8 per cent three decades later (Kamen 1983:90). The financial crisis that Ferdinand and Isabella had managed to avert came to full maturity under Charles V. The burden fell heavily on the Castilian crown which was the guarantor of the loans. Important sectors of the Spanish economy became increasingly dominated by foreigners as a result of these loan transactions. By the time Charles resigned his crown, in 1556, Spain was on the verge of a suspension of payments. Philip II When Philip II took over the realm from his father, he attempted to clean the financial slates.13 In 1557, all payments from the Castilian treasury were suspended and all outstanding debts were converted to juros, yielding an interest of 5 per cent (Kamen 1983:90; cf. Lynch 1964:57). Spanish government finances were, however, not sound when Charles V handed over the reins of Spain to his son. This was to become apparent on several occasions during the next forty years. Most of the strain continued to come from Spain’s involvement abroad (Elliott 1970:231). ‘Philip II belongs to that group of rulers for whom foreign policy comes first. It could hardly have been otherwise, in view of the fact that he was master of the greatest empire yet seen’, writes Antonio Domínguez Ortiz (1971:64). Philip inherited the foreign policy of his father, but not all the resources. When Charles divided his empire, major adjustment problems arose in Spain. While the emperor could draw on German support both in terms of credit and soldiers, Philip was forced to build Spain into a world power from virtually nothing (Kamen 1983:129). Military activities continued in the Mediterranean

47

LATIN AMERICA

against the Turks and in the Netherlands, and in 1580, Portugal was annexed. At the same time, the Spanish empire was growing in the New World and in the Philippines. In 1588, the Great Armada was sent off to Britain, with disastrous results,14 followed by similarly unsuccessful attempts in 1596 (Ireland) and 1597 (England). The latter war efforts were highly important since they were basically directed towards the control of the Atlantic Ocean, the nexus with the New World (Braudel 1975:678). Spain also became involved in a war in France after the ascension of the Protestant Henry IV to the throne in 1589. On top of all this, revolts broke out in Granada in 1568 and in Aragon in 1591 (Elliott 1970:235–41, 277–84). Maintaining and extending the Spanish empire was a costly affair. Charles V had by and large managed without any major deployment of troops. He had not had to maintain a substantial Spanish army in Europe, relying instead on a force that included Spaniards only as one component among many others. Even the New World had been conquered by a handful of conquistadores rather than by a massive intervention of armed forces. Moreover, naval transport to the Americas was in the hands of private enterprise during the early sixteenth century (Elliott 1970:59, Kamen 1983:161).15 During the reign of Philip II, the situation was different. Spain was forced to increase its military force considerably until, in 1587, it was estimated that throughout his dominions, Philip II had over 100,000 men under arms (Kamen 1983:162). The navy had to be developed in a similar manner, beginning in the 1530s. In 1560, a Spanish fleet on its way to capture Tripoli was destroyed by the Turks at the island of Djerba. This made the navy an urgent priority over the next fifteen years—increasing the Spanish navy to about four times its size during the empire of Charles V. After the acquisition of Portugal, an Armada of the Ocean Sea (the Atlantic) started to come into existence as well (Kamen 1983:163). The size of the army and the navy could only be expanded at a rising cost. Such was the extent of the political commitments during the reign of Philip II that the state found itself in perennial financial trouble. In 1550, the ordinary revenue of Castile amounted to 1.25 million ducats, although this fell far short of what was needed. By 1598 the total annual military expenditure of Castile in Spain, the Mediterranean and Flanders was approximately 10 million ducats (Kamen 1983:164). Financing war expenditure on this scale called for extraordinary measures, especially since Philip had to begin his reign by a suspension of payments. The general policy was to attempt to make each war theatre bear its own costs. This was, however, only partially successful. In particular, the war in the Netherlands required far more than could be financed locally. The inevitable consequence was that the financial burden fell on Spain itself (see Lynch 1964:128–34). Here, Aragon continued to be fairly unimportant, whereas the church increased its contribution more than fourfold. This, however, did not free Castile from the main burden. Between 1559 and 1577 regular revenue increased almost threefold (whereas the price level only rose by 25 per cent). However, during the last two decades of the sixteenth century, tax income stagnated at its 1580 level while, in

48

SPAIN AND THE CONQUEST OF AMERICA

the meantime, inflation rose. After the defeat of the Great Armada, in 1588, a financial crisis was triggered which necessitated the introduction of additional taxes (Kamen 1983:165–6). At this point, the New World had started to yield handsome returns to the Crown, particularly the silver mines of Potosí, Zacatecas and Guanajuato. During the reign of Philip II the royal quinto (from 1504 the share of the Crown was fixed to one-fifth of the species production) together with American taxes produced more than 64.5 million ducats, while a further 8 million were seized from private traders. Nevertheless, even in the best years, no more than one-fifth of all government income came from bullion (Kamen 1983:166; cf. Elliott 1984b: 322). However, the real importance of silver lay elsewhere. The imperial adventures of Philip II (as well as those of Charles V) were financed by large-scale borrowing, which would hardly have been possible without the steady inflow of liquid American silver. The latter could then be used to attract the bankers of a Europe which had been desperately starved of precious metals for several centuries (Elliott 1989:23, Chaunu 1977:312, Elliott 1970:183). Nevertheless, government revenue had to be supplemented in a number of ways. The jurisdiction over common land and towns was sold to the nobility. Nobility privileges and offices of different types were also put on the market. Combining all the revenue sources, the total annual income of Philip II more than tripled from 1559 to 1598, from some 3 million to more than 10.5 million ducats. However, this still fell far short of expenditures. It has already been mentioned that payments were suspended in 1557. Three years later, there was another financial crisis and another suspension. The 1560s were a difficult decade with interest payments on the juros on a high level, absorbing in 1565 no less than 84 per cent of income. In 1574, it was calculated that the current debts of the government amounted to 74 million ducats while income in that year was a mere 5.6 million. This, in turn, led to a third suspension of payments in 1575 and to an increased reliance on international credit after experiencing new problems on the Netherlands front the following year. The last suspension during the reign of Philip II took place in 1596, as the king attempted to free Castile from Genoese financiers (Elliott 1970:285–7, Kamen 1983:166–8).16 DRIVING FORCES IN EXPLORATION AND CONQUEST: THE THEORY AND PRACTICE OF EMPIRE

Pecuniary motives ‘One should not believe that the Castilians went to America only out of a pure desire for missionary perfection’, writes Jaime Vicens Vives (1975:288). From the foregoing, it should be quite evident that the Crown had a serious financial stake in whatever discoveries that were made in the New World, one which furthermore tended to increase over time as government finances deteriorated.

49

LATIN AMERICA

At the beginning of the sixteenth century, Ferdinand, whose main interest in the Indies lay in the revenue which could be obtained there (Parry 1966a: 60), repeatedly urged the Hispaniola governors to put more Indians to work in the gold mines, to bring in more Indians from islands that lacked gold and to keep onethird of all Indians working in gold production at all times (cf. Simpson 1929: ch. 3 for details). Given the financial difficulties of the Crown, obtaining gold became the most important official objective, one which warranted sacrificing the natives in the process (Sauer 1966:196–7). An estimated quarter to a third of the members of each work gang died during the six to eight months they had to spend in the gold fields (Sauer 1966:148). The size of royal revenue was intimately connected with the quantities of gold and other precious metals that could be produced. This, in turn, was wholly dependent on forced labour. The pressure on the Crown to obtain revenue was not to decrease during the course of the sixteenth century, as we have seen in the preceding sections. As Charles V exhausted his European revenues and increased the national debt, this pressure to an increasing extent fell on American treasure (Lynch 1964:53), and, ‘driven by those necessities that know no laws’ (Pierson 1975:69), Philip II resorted to whatever methods he could think of to extract government revenue from the new territories. The same pecuniary motives were present, and to no smaller extent, among the individuals involved in the conquest. From the very beginning of his voyages, Columbus—himself obsessed with fortune—had spread the rumour of fabulous wealth (Sauer 1966:290–1).17 In Spain, there was a ready supply of people who wanted to get rich quickly and who were thereby prepared to cross the Atlantic. Ralph Davis (1973:41) summarizes: Fighting men and early settlers sought gain for themselves. […] Their main body was made up of seamen, soldiers from the wars in Granada and Italy, rustics who had drifted into Seville in search of a living and were now scooped up by the New World adventure, criminals and debtors escaping from their troubles. The leadership was drawn from a minor nobility that regarded rank and social distinction as attributes dependent on the possession of land and wealth, which were themselves, ideally, the fruits of valour in successful battle; and their restless and fortune-seeking attitudes were absorbed by their followers. Even artisans and agriculturists who had been carefully selected to go and exercise their skills in America succumbed to the prevailing ethos. The fantasy of gold for the taking lured them all, and they followed this will-o’ -the wisp from island to island and to far corners of continental America. The conquistadores as a rule had left without any help from the Spanish government—at their own expense—in search of riches (Parry 1966b: 57).18 The most important military conquests were private enterprises. The fact that their landfalls were privately financed influenced the behaviour of the conquistadores very strongly, focusing their attention almost exclusively on finding gold (and subse-quently—and more importantly—also silver). Being reprimanded by a

50

SPAIN AND THE CONQUEST OF AMERICA

priest for his brutal treatment of the Indians instead of teaching them the gospel, Francisco Pizarro answered: ‘I have not come for any such reasons. I have come here to take away from them their gold.’19 The fate of Hispaniola and the other Antilles in this respect is instructive. In 1512, placer gold was found in Cuba. The following year, the audiencia at Santo Domingo was already complaining about the exodus from Hispaniola, asking that an end be put to the emigration of artisans and miners (Sauer 1966:186). The outflow of Spaniards accelerated as the mainland was being explored, particularly after 1517, not only from Hispaniola but from the other islands as well, as the gold supply was exhausted (Davis 1973:42). Finally, when the word spread of the gold and other riches of Mexico after 1521, the islands were almost depopulated. Not even a law extending capital punishment for leaving could diminish the attractiveness of the precious metals (Logan 1968:30). In Mexico, the same story repeated itself with some variations. Cortés’ expedition was a privately financed one (Elliott 1984a: 179). Economic returns consequently played a large role for Cortés himself as well as for his men. However, since there was not enough precious metals in Tenochtitlán to satisfy all the Spaniards who were attracted there, large numbers of them continued under new leaders: into the jungles between Mexico and the Panama isthmus, across the deserts to California, southwards to Colombia (where gold mines were discovered in 1537), and, not least, after 1532, to the riches of Peru. All the Spanish settlements in Mexico, with the exception of Mexico City and perhaps Vera Cruz as well, had been drained of most of their European population a few years after their foundation (Davis 1973:42). Spreading the faith. The mixture of motives Of course, one should not think of the conquest of America simply in materialistic terms. As Ralph Davis has pointed out, Isabella of Castile ‘was an unusually pious woman in a pious age, and the conversion of an Indian population was gratifying to her’ (Davis 1973:39–40).20 Charles V was clearly orthodox in his religious views (Lynch 1964:74) and about Philip II, John Lynch (1964:172–3) writes that The sincerity of his religious life could hardly be questioned […] Religion pervaded all aspects of Philip II’s life and work. In making political decisions which affected his own conscience he invariably consulted his religious advisers.’ There is little doubt that Spanish society during the sixteenth century was one where religion and the Church were of utmost importance (Lynch 1964:109). All the expeditions of exploration and conquest included a priest (Kamen 1983:93). In addition, a series of papal bulls between 1493 and 1523 obliged the Castilian crown to promote the conversion of the Indians in the new provinces (Barnadas 1984:512).21 Personal desire for adventure, glory and greatness on the one hand and a need to leave difficult circumstances behind on the other were clearly important factors (Davis 1973:40, Vicens Vives 1972:287–8, Kamen 1983:92–3). However, for

51

LATIN AMERICA

many the prospect of gain was the only objective (Davis 1973:40). As Carlo Cipolla (1965:136) has expressed it, ‘Religion supplied the pretext and gold the motive.’ Contemporary observers shared this opinion. Bernal Díaz del Castillo, who participated in the conquest of Mexico with Cortés and later became the chronicler of this feat, summarized the mixture of motives in a classic statement (1943:394): ‘We came here to serve God, and also to get rich.’ In this, he echoed his leader who just before laying siege to Tenochtitlán in 1520 stated that his primary motive was the spiritual conquest of the Indians (Hanke 1949:133), but who also confessed that ‘I and my companions suffer from a disease of the heart which can be cured only with gold’ (López de Gómara 1966:58). The disease quickly became epidemic and it was in the Americas to stay. Without the slightest doubt, precious metals—gold and silver—constituted one of the most potent driving forces in the process of exploration, conquest and reorganization of economic life. In 1601, the Viceroy of New Spain worried about the colonization of New Mexico: ‘Great difficulty can be foreseen…in its satisfactory settlement and growth as long as there are no mines to stimulate the greed that will carry forward and facilitate the whole business.’22 The precious metals were to trouble the Spaniards throughout the colonial period. The mixture of motives in the conquest—government revenue, private gain, missionary zeal—created a problem. Somehow, the three had to be reconciled with each other and a policy which took all three into account had to be worked out. To a large extent, the three forces were contradictory. Naturally the Crown, being in financial distress, wanted as much as possible of what the New World produced. This could, however, only be obtained either at the expense of the colonists, assuming the pie to be shared to be fixed, or at the expense of the Indians, who would have to shoulder a larger work burden if both the Crown and the colonists were to be kept content. This problem was in turn further complicated by the mission of spreading the Catholic faith. In their wars against the Muslims in the Mediterranean, the Spanish did not have to pay much attention to the legal status of Turks and Arabs who were enemies of Christianity and could therefore be enslaved by anybody with the approval and sanction of both the Church and the Crown. In America, matters were different. The Indians were in principle innocents who should be converted to Christianity, not Muslim enemies. They were furthermore under the crown of Castile. Thus, their legal status, their rights and obligations, somehow had to be defined and this had to be balanced against the necessity to reward the Spanish pioneer settlers and to attract new immigrants in a manner which brought suitable revenue to the Crown. To complete this picture, the need for labour both to extract minerals and to provide food for the settlers was compounded by the Spanish disdain for manual work. The exercise of manual work or trade was not considered a worthy pursuit. Instead, ambitions were directed towards obtaining noble status, and being able to live on rent income. Even those who were for example traders were struggling to escape el deshonor de trabajo and make it into the landed gentry instead (Lynch

52

SPAIN AND THE CONQUEST OF AMERICA

1964:107–8, Elliott 1970:32, 64). The idle hidalgo was the ideal, and this ideal was carried across the Atlantic as well. Although the hidalgos constituted a minority among the Spanish settlers, their spirit dominated the conquest enterprise. The Spanish community that grew up in the New World was characterized by a conspicuous absence of manual labourers and peasants of European extraction (Parry 1966a:173). Those who left for America did so to improve their lot. ‘All the Spaniards…even the most miserable and unfortunate, want to be señores and live for themselves, not as servants of anyone, but with servants of their own’, wrote a contemporary observer.23 The ‘theory of empire’. The battle over the encomienda The task of reconciling the motives attracted a lot of intellectual attention. During the course of the sixteenth century, a more or less elaborate set of doctrines, a ‘theory of empire’ regarding the New World, emerged in Spain.24 This theory sprang from the variety of practical, administrative and juridical problems encountered by the Crown. During the reign of the Catholic Kings, America received much less attention than Africa (Hillgarth 1978:578), but as time passed, the importance of the new discoveries increased, and there was an increasingly intense discussion of the principles that should govern Spanish action. Here, the Church played a central role. Within the Aristotelian and Thomist traditions, scholastic methods were brought to bear on questions regarding the Spanish empire in America (Elliott 1984b:305). One of the central questions in the debate was what rights of conquest the Spanish had in the New World. Ultimately these rights rested on the fact that the Spaniards had discovered America for Europe. The claims were further substantiated by four successive papal bulls in 1493. The Treaty of Tordesillas between Spain and Portugal in the following year regulated the respective spheres of the two countries. The treaty superseded the bulls but still left the instruction that the Spanish in the course of their American activities were to convert the natives to Christianity. This was a duty that the Spanish crown could not deny to carry out, but it also gave rise to a host of questions (Parry 1966a:139): To what extent…were they authorised to use secular means to this spiritual end? Could the duty of conversion be held to justify armed conquest, the deposition of native rulers—if indeed the Indians had legitimate rulers—the assertion of Spanish sovereignty over the Indians in general? This was the central question. If it could be answered affirmatively it gave rise in turn to subsidiary questions. If the Indians should be reduced, by a just conquest, to the position of subjects of the Spanish Crown, what legal and political rights remained to them? Should they be converted by compulsion? Were they to be subject to Spanish courts and Spanish law, civil or ecclesiastical? Might they be commended to individual Spanish feudatories, deprived of their land, put to forced labour, enslaved?

53

LATIN AMERICA

In the debate over the right of conquest, one strand of thought maintained that the Indians were free individuals who could not be converted without their own agreement. Another held that if they refused to tolerate the peaceful preaching of the gospel, there was an automatic justification for conquest. Others invoked the doctrine of universal papal dominion whereby the infidels could keep their lands and possessions only if authorized to do so by the Church. In the case of noncompliance, the Pope could appoint Christian rulers. Others held that punitive expeditions could be sent out against the Indians if the Indians attacked missionaries or to prevent such attacks, or in the case of human sacrifice. Natural law was invoked. The Christians were to ‘exercise a paternal and benevolent guardianship’ should the Indians be ignorant of this law or openly defy it, indulging in tyranny, human sacrifice or bestiality (Parry 1940:23). The Aristotelian view that some men were by nature free and others servile was also brought to bear on the relations between Spaniards and Indians. Thus, a theoretical case for Spanish conquest was made, a case that rested on several different arguments. This case was sternly defended by the government. Once the conquest had been justified, the question of the treatment of the Indians remained. When Columbus returned to Hispaniola on his third voyage in 1498, he was faced with an uprising which had broken out among the colonists the year before. This uprising Columbus failed to put down, and in order to settle the matter, he had to enter into negotiations with the rebels and agree to allocating the Indian communities to the Spaniards. This act constituted the beginning of the repartimiento (encomienda) system which was to constitute a most important part of Spanish policy in the Americas for a long time to come.25 In 1503, the Governor of Hispaniola, Nicolás de Ovando, obtained a royal decree which sanctioned the practice begun by Columbus (Zavala 1935:4). The repartimiento system, later (and most commonly) known as encomienda (Sauer 1966:101), had its roots in medieval Spain where it had been created as an administrative device to be used during the Reconquista of land from the Moors (Chamberlain 1939). It was then a temporary grant of manorial rights and of jurisdiction over the conquered territories and it entailed the right of feudal services of the peasants living on the land (Korth 1968:2; cf. Elliott 1984a:165). In the New World, the encomienda had a different character. It no longer referred to land rights but to Indian villages ‘commended’ to a Spaniard who was entitled to a tribute from ‘his’ Indians. During the early period of Spanish expansion, this frequently took the form of labour obligations. In return, the encomendero had to provide military defence, maintain the village clergy and protect and instruct the Indians (Parry 1940:8).26 This system, giving the Spaniards authority over the Indians, lent itself to a great deal of abuse. In the Antilles, those Indians who had survived the conquest were worked to death in the gold fields or were so debilitated that they easily fell prey to inevitable cultural shock resulting from the ‘encounter’ with the Spaniards and, above all, to hitherto unknown European epidemic diseases. Columbus arrived in Hispaniola in 1492. The size of the indigenous population at that time is

54

SPAIN AND THE CONQUEST OF AMERICA

uncertain. Estimates range from around 100,000 to about 8 million.27 The exact figure, however, is of minor importance. What matters is the decrease of population over time. The repartimiento of 1514 enumerated 22,726, excluding children and elderly. This figure contained an unknown number of people born outside the island (Sauer 1966:200–1). In a period of twenty years, the Indian population had been all but wiped out. Unfortunately, Hispaniola was not unique in this respect; the Indian populations of both Cuba and Puerto Rico shared the same fate, and the Bahamas and some of the Leeward Islands were more or less systematically raided for labour to be transported to the larger islands, with the result that, for example, the Bahamas were virtually emptied of people (Sauer 1966:157–60, 186–94). Naturally, this demographic catastrophe did not go unnoticed. Above all, it was the encomienda that became the subject of the heated debate which ensued. In 1511, the Dominican friar Antonio de Montesinos delivered a sermon in which the citizens of Santo Domingo were exposed as living in sin because of their treatment of the Indians.28 The following year, the Laws of Burgos reserved the right to ‘commend’ Indians to the Crown, emphasizing the free status of the Indians but at the same time deciding that most of them were in practice not capable of using their liberty (Simpson 1929:ch. 4, Hanke 1949:23–5, Parry 1966a:177–8). Bartolomé de las Casas. The New Laws Montesinos’ sermon had been agreed by the whole Dominican community in Santo Domingo, and when Bartolomé de las Casas in 1515 left Cuba to go to Spain as a defender of Indian rights, subsequently (in 1522) joining the Dominican ranks, the order had obtained a terrific champion of the native cause. Las Casas was to dedicate the rest of his long life (he died in 1566) to this cause.29 More than any other contemporary writer, he insisted upon the liberty of the Indians. Conversion should never be the result of force or coercion but of the free, full acceptance of the Christian faith (Parry 1940:48). The monarch, in the opinion of Las Casas, received his authority from God. The royal office was invested with high and difficult duties and these duties had to be performed faithfully. A legitimate king could only rule in the interest of common welfare. This included the Indians who were the subjects of the Spanish crown and thus enjoyed all the guarantees of liberty and justice of the Castilian laws (Parry 1940:54–5). In practice, few of these principles had been observed. According to Las Casas, the encomienda granted to private individuals a jurisdiction and authority over the Indians, something that was strictly royal privilege. The institution was alien to reason, natural law and Castilian laws alike.30 Las Casas was not fighting an easy battle. Nothing came out of the Laws of Burgos. The Spanish settlers persisted in their habits of exploiting the Indian labour force in various ways (Parry 1966a:176), and in addition the Consejo de Indias by experience had found out that the encomienda was a highly efficient

55

LATIN AMERICA

instrument of settlement (Parry 1966a:179). This view was naturally reinforced by the conquistadores, as when Cortés pointed out that the conquests in Mexico would be impossible to hold without the aid of the encomienda institution (Simpson 1929:ch. 6, Zavala 1935:40–50; cf. also Parry 1966a:180, Elliott 1984b:149). Everywhere in the New World, the Crown’s officials often took a laissez-faire attitude. Balancing this, however, was the realization among the councillors, eagerly upheld by the powerful Dominican lobby at the court of Charles V, that the encomienda was a serious cause of Indian sufferings. In practice, the opposing influences created a situation where the Crown on the one hand would attack and take away the privileges of the encomenderos and on the other restore and even strengthen them, in a stop-and-go process that, however, in the long run tended to define the rights of the encomenderos in ever more restrictive terms.31 During these vacillations on the part of the Crown, the encomienda system spread to New Spain and to Peru, where in both cases it no doubt attracted new settlers. Gradually the system came to be viewed as a permanent one by those who benefited from it. In 1542, however, it suffered what at the time appeared to be a severe setback. In that year, the New Laws of the Indies, drafted with the aid of Las Casas and other Dominicans, were promulgated. This code contained an absolute prohibition of slavery. The employment of Indians as carriers or in personal service was prohibited as well. Finally the very institution of the encomienda was abolished (Parry 1940:30, Hanke 1949:91–2). In itself this meant very little. Laws had fallen into desuetude in the New World on many previous occasions. However, at the same time, the Crown let it be known that the New Laws were not to be taken as a mere piece of paper that could be quietly forgotten. The new code was meant to be seriously implemented. The Crown, increasingly absolutist, wanted to reserve the direct control of the Indians for itself and insisted on their personal liberty, fearing that independent encomenderos would with time constitute a hereditary aristocracy. The Dominicans had gained the ear of Charles V.32 The New Laws triggered a stream of protests from the colonists who feared that their incomes would dwindle and that the Indians would rebel (Hanke 1949:95–105). At court, the settlers’ lobby put an extraordinary effort into influencing the emperor’s councillors, by persuasion or by bribes, and Cortés and his fellow believers organized a strong resistance to Las Casas and the Dominicans (Simpson 1929:ch. 12, Elliott 1984b:307–8). The events of the early 1540s showed how far creole society had already diverged from the society of metropolitan Spain. The leading colonists were as blind to the motive forces of Spanish politics as Spanish statesmen were to the realities of colonial life’, writes J.H. Parry (1966a:182). The Valladolid debate: theoretical victory, practical defeat The same year as the New Laws were promulgated, a theoretical defender of the

56

SPAIN AND THE CONQUEST OF AMERICA

case of the colonists opposed Las Casas and the Dominicans: Juan Ginés de Sepúlveda, who that year had written his Democrates Alter.33 Ginés de Sepúlveda presented the case of natural aristocracy, whereby the higher races should govern the lower, and its corollary of natural servitude, pointed out that the Indians lived in defiance of natural law and that their very failure to resist the Spanish onslaught was a proof of their inferior state. For conversion to be possible, the natives had to be put under civilized government and supervision even if they themselves did not consent to it. Conversion did, however, in no way make for equality between Indians and Spaniards, nor could it be used as an argument in favour of political independence. The Dominicans managed to make the Crown withhold the permission to publish Democrates Alter. The work was criticized heavily. Ginés de Sepúlveda was publicly exposed as an advocate of naked slavery, a thought which was abhorrent to most Spanish theologians and jurists of the time. Finally, in 1550, by imperial order, a formal debate, continued the following year, took place beween Ginés de Sepúlveda and Las Casas in Valladolid, over the former’s work (see Hanke 1949:ch. 8, for details). Ginés de Sepúlveda lost his case and failed to obtain permission to publish his book. His theory was judged to be inhumane and the encomienda was taken to represent an illegitimate private authority over the Indians, one which created an aristocracy with vested interests that stood in the way of the rights of the Crown in Indian affairs. Hereby, the power of the Crown over the settlers in the New World was greatly strengthened. When it came to the Indians, on the other hand, Las Casas’ victory had virtually no practical effects. Already in 1545, the provisions in the New Laws that abolished the encomienda were revoked. However, the encomienda that hereby came into effect differed from the previous system in that it did not entail any right to Indian labour. Instead, the villages paid tributes to the encomenderos (Parry 1940:44–5). This, however, did not mean that the Indians escaped labour service. The discovery of silver at Potosí and Zacatecas in combination with a devastating epidemic in New Spain in 1545 and 1546 made for an acute labour problem for the Spaniards which somehow had to be solved, since mining created a great demand for labour. The demand for labour increased even further in response to the application of the mercury amalgamation process, which was taken into use in New Spain in the late 1550s and in the Andes at the beginning of the 1570s, and which enabled the refining of low-grade ores (Bakewell 1984a:115). The solution consisted in compulsory recruitment, known as repartimiento in New Spain and mita in Peru. In this system, Indian villages and towns had to supply a fixed proportion of its males for work for a given number of weeks, in rotation, on such tasks for which the use of repartimiento labour had been authorized, e.g. mining. For this, the Indians were paid a low wage (Parry 1966a:186–7).34 To a certain extent, the repartimiento was complemented by a policy of ‘congregating’ those Indians that did not already live in villages or towns into such units.35 Large-scale congregation was undertaken both in Peru in the 1570s and in all settled parts of New Spain and New Galicia (northern and

57

LATIN AMERICA

north-western Mexico) at the end of the sixteenth century and the beginning of the seventeenth (Parry 1966a:188–9). Gradually, a rapprochement was established between the representatives of the Church and the encomenderos. One has to keep in mind that Las Casas did not typify what most representatives of the clergy felt. Already in the 1540s, the Church and the encomenderos had begun to join forces. The Mendicant orders received financial assistance from the encomenderos who, in turn, understood that missions had a stabilizing effect on indigenous society and that tribute would be larger if the Indian communities could be made to prosper. However, most important of all was the fact that the encomenderos needed church support in their struggle with the Crown and its officials (Parry 1966a:166). During the second half of the sixteenth century, the Spanish debate on the theory of empire gradually faded out, in spite of the efforts of Las Casas and other Dominicans, to be replaced by a more pragmatic view. In the decade following the promulgation of the New Laws it became increasingly clear that the crown would in the future seek advice on colonial policy from its colonial administrators and not from the peninsular theorists (Parry 1940:57–8). The new repartimiento system became general and imperial practice ceased to pay any attention to theory (Parry 1940:74–5). Morover, the principle of obedecer pero no cumplir (obeying but not complying) was all too often adhered to (Elliott 1984b:303). In Upper Peru, Potosí grew extremely rapidly, to the point where in 1570 it was the largest town in the New World, with a population of 120,000, outnumbering furthermore any town in Spain itself. The mine—from the 1570s to the 1620s quite probably the richest one in the world (MacLeod 1984:364) -quickly earned a reputation as a devoradora de mitayos.36 Beginning in 1573, in the face of an increasing shortage of labour as the population declined, the viceroy had forced labour drawn from Indian villages across an enormous area extending 800 miles from north to south and a maximum of 250 miles from east to west (Cole 1985:9). ‘The cerro of Potosí resembled, mutatis mutandis, an immense concentration camp’, summarizes Pierre Chaunu (1977:310). The mitayos at Potosí were to work for a maximum of eighteen weeks every seven years, but reality was completely different. As the Indian population dwindled and the inhabitants of the districts subject to mita service deserted their houses and moved to zones not included in the mita, the Indians became increasingly exposed to overwork (Cole 1985:25–32). Around 1600, in some cases not one year out of seven but one out of two had to be spent at Potosí (Bakewell 1984a:125). THE EXPLANATIONS It should be clear from the preceding, historical, sections of the present essay that the opening of commercial relations between Spain and its colonies in the New World did not take place in the smooth, frictionless way envisaged by latter-day textbooks in international trade theory. Trade was opened through an

58

SPAIN AND THE CONQUEST OF AMERICA

act of military conquest and new institutions were designed to ensure that forced native labour was forthcoming. The purpose of the last section is to provide some, tentative, economic explanations of some of the main features of this process. Let us begin by focusing on the main decision makers and their goals. Lewis Hanke concludes his investigation of the Spanish struggle for justice in the conquest of the Americas by pointing to the conflict between the revenue maximizing goal and the evangelizing one (Hanke 1949:173). Throughout the present essay, we have emphasized this conflict, pointing to the divergent interests of the three main strategic decision makers in the conquest of America: the Crown, the Spanish settlers and the Church. We have, however, also dealt with the partial consensus of church and crown interests in religious matters and the struggle against the encomienda system,37 as well as the partly common mundane interests of church and settlers. In this concluding section, we will attempt to put these strands together and in addition see to what extent the interests of the Crown and the settlers coincided—a theme which has not been touched in the foregoing. Beginning with the Crown, it is realistic to assume that the overriding goal was the maximization of revenue (net present value of tax revenue). Throughout the period dealt with in the present context, the Castilian crown was in financial difficulties as a result mainly of costly warfare inside and outside Europe. Second, the Spanish settlers in America were actors who attempted to maximize their personal wealth, as described above. Finally, as we have already noted, the Church may to some extent also be seen as an economic actor—with the same type of goals as the Crown or the settlers (cf. also Batchelder and Sanchez 1988). We cannot, however, deal only with the purely economic rationality of the three strategic decision makers, but as we have seen above, in order to make the picture complete, the religious element must be added. The Castilian crown was under an obligation to spread the Christian faith to the Indians and the Church very definitely took evangelization seriously. If we are to understand the events sketched above, it cannot simply be assumed that economic rationality was left to work entirely without any ‘moral’ restrictions. The decisions taken by the three groups of actors were interdependent in various ways. In the first place, as has been pointed out in the foregoing, the various conquests were privately financed ventures, sanctioned (most of the time) by the Crown, although the Crown did not contribute any resources. In the case of success, the settlers were given the rights to exploit the area they had conquered, provided that the Crown received its legal share of the output. In this way, the interests of the two coincided. With a given share accruing to the Crown, the common goal was to maximize the net present value of production. However, there is also a second way of looking at the situation. If output instead is assumed to be given, the interests of the settlers and the Crown were in conflict, since the share of each then depended on the share of the other in a zerosum fashion. Thus, both for the Crown and for the conquistadores there

59

LATIN AMERICA

existed a tradeoff between obtaining a larger share of a given pie and enlarging the pie with given shares. Strictly speaking, only the conquistadores could enlarge the pie, but they were not allowed to do this without the consent, at least ex post, of the Crown. It must then be noted that the Crown had a much smaller stake in increasing the size of the pie than the settlers. With a share of in principle one-fifth of the net output from mining, i.e. a minority share, ceteris paribus, it must have been more lucrative for the Spanish sovereigns to attempt to increase the royal share of what was already produced in the New World by means of different types of taxation than to concentrate on growth-promoting measures. It is easy to see why. If it is assumed that the Crown attempted to increase its revenue at the expense of the share of the conquistadores by, say, 100 million ducats, this would not be without costs in terms of output forgone, since the resources employed could have been used, for example, for increasing mineral production. Thus, output would be lost in the redistribution process. On the other hand, this would affect the Crown as well, in the form of a loss of revenue. Then, the crucial question is how much revenue the Crown would be prepared to forgo in order to obtain the 100 million. The answer to this is of course a maximum of 100 million (marginal revenue equalling marginal cost), but with a one-fifth share of net output, the Crown would then be prepared to sacrifice 500 million ducats’ worth of output for society as a whole. Conversely, to obtain 100 million from growth, total output would have to be made to grow by five times as much, which was presumably much more difficult. The conquistadores on the other hand, whose share would amount to four-fifths, would be prepared to impose a much lower cost on society in the event of a redistribution in their favour: 125 million. Thus, one would expect that, ceteris paribus, at any given moment, the Crown would have stronger motives than the conquistadores to prefer the redistribution of a given pie to its enlargement and the conquistadores had a much stronger incentive to employ indigenous labour to increase output (cf. Olson 1982:41–7). Given the above complex of goals, constraints and conflicts, it is possible to offer certain hypotheses regarding some of the economic mechanisms that were instrumental in producing some of the most important features of the process explored by the present essay. By and large, six main questions emerge: 1 Why was the conquest of America left largely in private hands? 2 Why did the Crown choose to operate on a share contract basis with the conquistadores? 3 Why was the economic system that existed during the first century following the discovery and conquest to a large extent built on unfree, as opposed to free, labour? 4 Why was the encomienda system chosen instead of outright slavery? 5 Why were encomienda rights only temporary and not perpetual? 6 Why was the repartimiento (mita) system chosen for silver production and not the encomienda?

60

SPAIN AND THE CONQUEST OF AMERICA

In order to answer the first question, it is sufficient to note that during the first century after the discovery of America, the Spanish rulers did not dispose of a sufficiently large military strength to be able to employ regular soldiers in the conquest. The costly wars in Europe absorbed too much of Spain’s total resources. Thus, an alternative simply had to be sought which involved a joint venture with private interests. The contract between the Crown and the conquistadores was a share contract, with the former (from 1504) taking one-fifth and the latter, four-fifths. To understand why this particular contract form was chosen, and not one where the Crown or, alternatively, the conquistadores, was paid a fixed sum, the question of risk must be dealt with. As is well known from the literature on sharecropping (cf. e.g. Griffin 1974:22–6), a pure ‘wage’ contract would have put the entire risk burden on the Crown, being the residual claimant, and a pure ‘rent’ contract would have made the conquistadores bear the entire risk. The type of risk that we are dealing with here was substantial. As we have already pointed out, it took thirty years (until the conquest of Mexico) before the New World began to yield satisfactory returns. The probability was very high that any given expedition would result in a failure. Hence, it was natural that the conquistadores—the only ones who could produce revenue—should want to shift some of the risk to the Crown, by paying a share of the net output rather than a fixed sum, in return for the right to exploit the New World. (The Crown, as we already know, was not in a financial position where it could hire the services of the conquistadores against wage payments.) Against this reasoning, it may be argued that, if no contract costs exist, it is always possible to arrive at a mixture of pure wage and pure rent contracts which in terms of risk sharing is equivalent to what can be achieved through share contracts (Stiglitz 1974). However, in the case that we are analysing, contract costs did indeed exist and these could be substantially reduced by resorting to a standardized contract: the quinto real (cf. North 1981:26). Third, conquest ventures could be based either on free or on compulsory labour. In the foregoing, we have seen how the mentality of the Spaniards in general, and of those who went to the New World in particular, was one of shunning manual labour and aspiring to become hidalgos. In addition, there was plenty of land to be had in the New World, provided that the territories in question could be conquered. Thus, given the sanction from the Crown, land was practically free, especially when the indigenous population began to dwindle. Putting these two facts together immediately takes us to Evsey Domar’s hypothesis regarding the causes of slavery and serfdom: out of three elements—a non-working landowner class, free land and free labour—only two, but never all three, can exist simultaneously (Domar 1970). This provides the solution to our third problem: free labour was out of the question. Our fourth question is why the encomienda was preferred to a system based on outright slavery. (The enslavement of Indians, other than those captured in ‘just war’, was prohibited (Parry 1940:9).) An answer to this may be given in terms of

61

LATIN AMERICA

contract costs. The newly conquered territories in America had to be defended, both against external intruders (rival European powers) and against Indian uprisings. The Crown lacked the resources to do this. Hence if the Spanish settlements were to be defended, the organization of defence would have to take place via the voluntary efforts and the co-operation of the conquistadores. As Ronald Batchelder and Nicolás Sanchez have pointed out, co-operation was possible to a certain extent because a given capability to suppress revolts of a group of given size could be used as a threat against a number of such groups, provided that the latter groups were not in the position where they could join forces (where this was prohibitively costly) (Batchelder and Sanchez 1988). Thus, the conquistadores could agree to share resources in order to minimize defence costs. In the absence of government regulation, this would have involved mutual agreements between groups of settlers and the number of such agreements that would have to be concluded would have increased rapidly with the number of signatories involved. Increasing the number of conquistadores in an area from n individuals to n+1 in principle increased the number of contracts to be concluded by n, provided that all individuals could be of mutual assistance. The Crown had an incentive to act as a free rider here and leave defence completely in the hands of the conquistadores. Defence is a collective good for those defended and, limiting the reasoning to the defence of actual and potential output, the share of the Crown amounted to one-fifth, while that of the conquistadores was four-fifths. The incentive to organize defence was much stronger for the latter than for the Crown (Olson 1965:ch. 1). Thus, a system based on regular slavery would have been left to voluntary agreements between the slave-owners when it came to organizing defence. However, the Crown could minimize both its own defence costs and the contract costs of the conquistadores. We have an analogy to the Coase explanation of the existence of the firm here (Coase 1937). The role of the Crown was that of the entrepreneur in the Coase theory. By imposing a forced contract on a limited number of encomenderos (analogous to the ‘factors’ in the Coase explanation) to provide defence the number of contracts to be concluded was greatly reduced.38 The mere reduction of transaction costs does, however, not suffice to explain why encomiendas were given solely as temporary rights, and not in perpetuity. It can even be argued, with Coase, that a perpetual grant should have been preferred because this would have led to long-term agreements between the Crown and the encomenderos and hence to even fewer contracts than with temporary grants. To understand why the temporary arrangement was preferred, other reasons have to be invoked. One such reason is to be found in the financial situation of the Crown. The rate of time preference for output was higher for the Crown, being in financial distress (possibly to the point of myopia), than for the encomenderos (cf. Fisher 1930). While the latter were constantly lobbying for perpetual rights, by granting temporary rights only, the rate of time preference of the encomenderos could be increased to the point where it coincided with that of the Crown in order to yield the desired output and crown revenue per annum.

62

SPAIN AND THE CONQUEST OF AMERICA

On the other hand, we have already noted the incentive for the Crown to attempt to enlarge its share of output as compared with increasing the growth rate. However, this would not have been without costs. In a situation where the administration was weak, evasion of taxes was not too difficult, and the incentives for the encomenderos to hide some of the output would increase with the tax rate. Thus, a balance had to be struck between the gross addition to revenue that an increased rate of taxation would lead to and the increased cost of supervision and control. Judging from the historical evidence, increasing revenue by increasing output while maintaining a constant tax rate was the better alternative for the Crown. During the course of the sixteenth century, the encomienda system was gradually abolished and encomendero rights reverted to the Crown. In the Antilles, the supply of both precious metals and Indians was depleted to the point where in 1542, in connection with the New Laws, the Antillean Indians were exempted from all kinds of taxes and personal services. Charles V ordered ‘that they be left in peace so that they can more easily multiply’.39 In the 1570s, the Crown had won the battle against the encomienda in New Spain (Gibson 1977:68) and in Peru the institution had declined at the beginning of the seventeenth century (Villamarin and Villamarin 1975:75). Before that, however, a technological revolution in silver mining had taken place: the mercury amalgamation process. As we have already pointed out, this increased the demand for labour drastically. Both in New Spain and in Upper Peru (Potosí), mining had rested on a mixture of slaves, encomienda and free workers (Bakewell 1984a:127). However, the acute shortage of hands in both regions which was brought about by the new techniques in combination with the dwindling labour force called for new methods: forced draft labour, repartimiento (mita).40 However, the encomienda was also built on forced labour. Why, then, was the repartimiento substituted for the encomienda or, rather, why was the encomienda not employed to increase the production of silver? An encomienda solution could have been of two kinds: temporary or perpetual. As witnessed in the Antilles, the former variety would have entailed a rapid depletion of the labour force. This, however, was out of the question. The example of Potosí is very clear on this point. The amalgamation process introduced at the beginning of the 1570s required a significant long-term investment in crushing mills which, in turn, called for a steady, not declining, supply of labour (Cole 1985:7–8). A perpetual encomienda, on the other hand, would quite probably have led to a rate of extraction which would have been too slow for a crown that was in deep financial trouble. Thus, the Potosí mita represented an ‘intermediate’ solution, one which in addition kept labour costs at a level far below that which would have been required to attract free wage labourers. The legislation in favour of the Indians that was enacted with the assistance of the Church was largely unobserved in practice, and the substitution of the repartimiento (mita) system for the encomienda did not do much to change the fate of the Indians. Its main effect was to increase the supply of indigenous

63

LATIN AMERICA

labour. In addition, the substitution helped the Crown obtain a more direct control over labour use, strengthening its bargaining position vis-à-vis the colonists. Finally, the move may be interpreted as a means of placating the Church, since the abuse of the Indians in many ecclesiastical circles was closely identified with the encomienda institution per se rather than with actual practice. CONCLUSIONS The above finishes our discussion of the conquest of America by Spain and the economic system developed in the colonies. It should by now be quite obvious that economic relations were established between Spain and its colonies in a fashion that had very little to do with the way envisaged by the standard theory of international trade. Above all, the institutional setting was completely changed in the New World, as a result of an interactive game between three actors whose interests partly coincided and partly did not. At the end of the first century after the discovery of America, however, all the peninsular actors had made (pecuniary and spiritual) gains, the distribution of which was the result of the complex operation of the new system. For the conquered, on the other hand, the conquest was a tragedy. While it is certainly not possible to maintain that the labour obligations imposed by the Spaniards on the Indians led to wholesale destruction of the indigenous population (Davis 1973:54–5, Sánchez-Albornoz 1984:9–10, Gibson 1984:402, Cole 1985:26) (except perhaps in the Antilles and in the murderous Huancavelica mine), they did contribute to a larger complex of problems. Labour obligations, in combination with malnutrition caused by disruption of indigenous food supply systems due to the European intrusion, as well as outright cruelty, made the Indians extremely susceptible to hitherto unknown diseases, introduced by the Europeans (no less than fourteen major epidemics swept Mesoamerica and up to seventeen the Andean region between 1520 and 1600) (Wolf 1982:133; cf. Wachtel 1984:212–13). Together, these reduced the Indian population drastically everywhere in the New World where there was contact between the two races.41 We have already mentioned the fate of the Antilles. What happened in central Mexico is subject to substantial controversy, but only as far as magnitudes are concerned, not when it comes to the general trend. Thus, the Indian population was reduced from somewhere between 11 and 25 million in 1519 to 750,000 in 1630 (Sánchez-Albornoz 1984:4). In present-day Peru, a population of some 3.3 million at the beginning of the conquest was reduced to about 670,000 in 1620 (Cook 1981:94). Everywhere the pattern was the same. Contact with the Europeans and their diseases brought disaster for the Indians. ‘In its extent this phenomenon is without parallel in the modern history of the world’s population’, writes Nicholás Sánchez-Albornoz (1984:7),42 and it is difficult to escape Ralph Davis’ (1973:55) conclusion that, at least during the first century following the conquest, ‘Spain’s principal gift to America was the destruction of its people’.

64

SPAIN AND THE CONQUEST OF AMERICA

ACKNOWLEDGEMENTS This study was finished during my stay as a research fellow in Uppsala, at the Swedish Collegium for Advanced Study in the Social Sciences (SCASSS). Financial support was obtained from the Swedish Council for Research in the Humanities and Social Sciences and from the Swedish Agency for Research Cooperation with Developing Countries (SAREC). This is gratefully acknowledged. Thanks are also due to Ronald Findlay, Magnus Mörner, Christina Rapp and Mauricio Rojas for comments on a draft version. NOTES 1

2

3 4 5 6 7 8 9 10 11 12

The essay draws on several sources of inspiration, like Kindleberger’s (1951) insistence on the importance of group behaviour in the context of international trade, Cipolla’s (1965) emphasis of (especially) the role of the guns (but also sails) in the European expansion, Frank’s (1967) insistence on the expropriation of ‘ecomomic surplus’ by ‘metropolitan centers’ from ‘peripheral satellites’, Griffin’s (1969) and Hymer and Resnick’s (1971) analyses of how, in the historical perspective, mercantilist trade produced large ‘deadweight losses’ outside Europe, Wallerstein’s (1974) and Chaunu’s (1977,1983) accounts of the intimate connection of economic events inside and outside Europe in the sixteenth century, Barratt Brown’s (1974) historical examination of the international division of labour against a Keynesian and Marxist background, North’s (1981) grandiose attempt to bring a transaction cost approach to bear on the historical evolution of economic institutions, Bhagwati’s (1982) discussion of directly unproductive profit-seeking activities, and, finally, the attempt by Batchelder and Sanchez (1988) to employ a transaction cost framework to analyse the encomienda institution. (This does, however, not necessarily mean that I subscribe to all of the views held by these authors.) The Portuguese had invaded Castile in 1475, King Alfonso V claiming the Castilian crown. (See e.g. Hillgarth 1978:354–63, for details.) Navarre was formally joined to the crown of Castile at the Cortes of Burgos in 1515. The connection between the Reconquista and the Spanish expansion outside Europe is dealt with, for example, in the articles contained in Johnson (1970). Ladero Quesada (1973) analyses the finances of the Castilian crown in the fifteenth century. From 1497, the main gold coin in Castile was the ducat, but accounting around this time was done in units of maravedís, with 1 ducat equal to 375 maravedís (Kamen 1983:xi). The classic study of the Mesta is the one by Klein (1920). The story of the early explorations, conquests and frustrations is told in Sauer (1966). Koenigsberger (1971) gives a thorough account of the wars of Charles V and Philip II. A detailed study of the finances of the Spanish crown under Charles V is made in Carande(1990a, b, c). Hamilton (1934) constitutes the ‘classic’ study of the shipments of bullion from the New World. For a brief evaluation of Hamilton’s work, see Lynch (1964:123, note). Cf. note 3, above. From the beginning of the sixteenth century, the ducat took over as the unit of account (Kamen 1983:xi). Elliott (1970:199) refers to the ‘hand-to-mouth existence’ of the Crown under Charles V. This work traces the history of the juros.

65

LATIN AMERICA

13 14 15 16 17 18 19 20 21 22 23

24 25 26

27

28 29 30 31 32 33 34 35 36 37

The royal finances during the reign of Philip II are examined in exhausting detail in Ulloa (1977). Cf. also Ruiz Martín (1968). See Fernández-Armesto (1988), for an account of the reasons for the failure. Spain’s trade routes with America are scrutinized in every possible detail in the monumental work by Chaunu and Chaunu (1955–9), including Arbellot et al. (1957). ‘The crown’s American revenue rose from about 11 per cent of total income in 1554 to about 20 per cent in 1598’ (Lynch 1964:128–9). Columbus’ obsession with gold runs as a red thread through all his voyages. Cf. Morison (1942), passim and Sauer (1966), passim. For examples, see Sauer (1966:113–14, 169, 190, 236) and Lynch (1964:150). Quoted by Hanke (1949:7). ‘Isabella’s faith was fervent, mystical and intense’ (Elliott 1970:103). This effort was clearly successful, at least in terms of the number of converts: ‘in America, perhaps seven to eight million newly baptized at the end of the 16th century’ (Chaunu 1977:7). Quoted by Bakewell (1984a:150). The Franciscan Fray Gerónimo de Mendieta. Quoted by Elliott (1984a:190). Possibly, this may have to do with the extensive employment of slave labour in southern Spain (and Portugal) during the centuries immediately preceding the conquest (cf. e.g. Chaunu 1983:142 and Pike 1967). On the other hand, the extent of Spanish contempt for manual work has frequently been overstated (cf. Mörner 1991). The debate surrounding the emergence of this theory is discussed in Parry (1940). Cf. also Parry (1966a:ch. 7) and MacLachlan (1988). Sauer (1966:93–101) gives the details. Simpson (1929) provides an account of the system in New Spain (Mexico). Zavala (1935) deals with the spread of the system to other parts of the Spanish empire in the New World. The term encomienda has been the subject of much confusion in the Latin American context. In particular, it has been maintained that land rights were contained in the encomiendas, but this is not the case. Cf. Frank (1979:10–11) and the references cited there. A discussion of contemporary estimates of the population of Hispaniola can be found in Cook and Borah (1971:ch. 6), who after scrutinizing the information arrive at a probable total of 8 million. This figure is, however, almost certainly a gross overestimate. Cf. the literature referred to in Bethell (1984). See Las Casas (1965:vol. 2:438–44) for a contemporary account, and Hanke (1949:ch. 1). Cf. also Sauer (1966:197). For a biography of Las Casas, see e.g. Hanke (1965). For more references, see Bethell (1984:594–7). Quoted in Parry (1940:53). See Parry (1966a:ch. 9) for details. Elliott (1970:74) stresses that ‘it is highly improbable that so much would have been achieved if the Spanish Crown had not already been predisposed in favour of Las Casas’ ideals for less altruistic motives of its own’. For details regarding Ginés de Sepúlveda’s theories, see Parry (1940:ch. 3). Cf. also Ginés de Sepúlveda (1941). Frank (1979:ch. 3) discusses the system in New Spain. The practice was also known as reducción. Cf. Mörner (1970:43–51) for details. Details can be found, for example, in Hanke (1956). Bakewell (1984b) and Cole (1985) deal specifically with the mita. In this context it should be noted that the interests of the Crown superseded those of the Church when the two conflicted. As noted by Colin MacLachlan, ‘By the sixteenth

66

SPAIN AND THE CONQUEST OF AMERICA

38

39 40

41 42

century, Crown control over the Spanish Church exceeded that of any other European kingdom, whether Protestant or Catholic’ (MacLachlan 1988:7). Cf. also MacLachlan (1988:29–32) for details. Only 36 encomiendas have been documented in the Valley of Mexico (Gibson 1977:68) and in the entire former Inca empire the maximum amounted to about 500, around 1540 (Mörner 1985:38). Thus, the vast majority of settlers never became encomenderos. Quoted by Zavala (1935:38). The main use of the draft labour system was in Potosí and Huancavalica. New Spain offered more of a problem since the Indian population in many of the major mining districts was too dispersed or too rebellious to be easily drafted. Thus, wage labour, for cost reasons, quickly became the prevailing mode of operation in northern New Spain (Bakewell 1984a:126–8). See e.g Sánchez-Albornoz (1984:7–14) for a discussion of these factors. Sánchez-Albornoz (1984:4–7) gives the details for other areas in Latin America than those mentioned in the text. Cf. also the articles in Denevan (1976) as well as Mörner (1989).

BIBLIOGRAPHY Arbellot, Guy, Bertin, Jacques, Chaunu, Huguette and Chaunu, Pierre (1957) Séville et l’Atlantique (1504–1650). Première partie: statistiques, Tome VII, Construction graphique, Paris. Bakewell, Peter (1984a) ‘Mining in colonial Spanish America’, in Leslie Bethell (ed.) The Cambridge History of Latin America, Volume 2, Colonial Latin America. Cambridge. ——(1984b) Miners of the Red Mountain: Indian Labor in Potosí, 1545–1560, Albuquerque, NM. Baldwin, Robert E. (1982) ‘The political economy of protectionism’ in Jagdish N.Bhagwati (ed.) Import Competition and Response, Chicago. Barnadas, Josep M. (1984) ‘The Catholic Church in colonial Spanish America’, in: Leslie Bethell (ed.) The Cambridge History of Latin America, Volume 1, Colonial Latin America, Cambridge. Barratt Brown, Michael (1974) The Economics of Imperialism, Harmondsworth. Batchelder, Ronald W. and Sanchez, Nicolas (1988) ‘The encomienda and the optimizing imperialist: An interpretation of Spanish imperialism in the Americas’, Department of Economics, University of California, Los Angeles, Working Paper # 501, September. Becker, Gary S. (1983) ‘A theory of competition among pressure groups for political influence’, Quarterly Journal of Economics vol. 98. Bethell, Leslie (1984), ‘A note on the American population on the eve of the European invasions’, in The Cambridge History of Latin America, Volume 1, Colonial Latin America, Cambridge. Bhagwati, Jagdish N. (1982) ‘Directly unproductive, profit-seeking (DUP) activities’, Journal of Political Economy vol. 90. Braudel, Fernand (1975) The Mediterranean and the Mediterranean World in the Age of Philip II, 2 vols, 2nd revised edn, London. Carande, Ramón (1990a) Carlos V y sus banqueros. 1. La vida económica en Castilla (1516–1556), 4th edn, Barcelona. ——(1990b) Carlos V y sus banqueros. 2. La hacienda real de Castilla, 3rd edn, Barcelona. ——(1990c) Carlos V y sus banqueros. 3. Los caminos del oro y de la plata, 3rd edn, Barcelona.

67

LATIN AMERICA

Castillo Pintado, Alvaro (1963) ‘Los juros de Castilla. Apogeo y fin de un instrumento de crédito’, Hispania vol. 23. Chamberlain, Robert S. (1939) ‘Castilian backgrounds of the Repartimiento-Encomienda’, Contributions to American Anthropology and History vol. 5. Chaunu, Huguette and Chaunu, Pierre (1955–9) Séville et l’Atlantique (1504–1650), 8 vols, Paris. Chaunu, Pierre (1977) Conquête et exploitation des noveaux mondes (XVIe siècle), 2nd edn, Paris. ——(1983) L’expansion européenne du XIIIc au XVe siècle, 2nd, corrected, edn, Paris. Cipolla, Carlo M. (1965) Guns and Sails in the Early Phase of European Expansion 1400– 1700, London. Coase, Ronald H. (1937) ‘The nature of the firm’, Economica, NS vol. 4. Cole, Jeffrey A. (1985) The Potosí Mita, 1573–1700: Compulsory Indian Labor in the Andes, Stanford, CA. Cook, Noble David (1981) Demographic Collapse. Indian Peru, 1520–1620, Cambridge. Cook, Sherburne F. and Borah, Woodrow (1971) Essays in Population History: Mexico and the Caribbean, Volume One, Berkeley, CA. Davis, Ralph (1973) The Rise of the Atlantic Economies, Ithaca, NY. Denevan, William M. (ed.) (1976) The Native Population of the Americas in 1492, Madison, WI. Díaz del Castillo, Bernal (1943) Historia verdadera de la conquista de la Nueva España, vol. 2, ed. Ramón Iglesia, Mexico. Domar, Evsey D. (1970) ‘The causes of slavery or serfdom: A hypothesis’, Journal of Economic History vol. 30. Domínguez Ortiz, Antonio (1971) The Golden Age of Spain 1516–1659, New York. Elliott, John H. (1970) Imperial Spain 1469–1716, Harmondsworth. ——(1984a) ‘The Spanish conquest and settlement of America’, in Leslie Bethell (ed.) The Cambridge History of Latin America, Volume 1, Colonial Latin America, Cambridge. ——(1984b), ‘Spain and America in the sixteenth and seventeeth centuries’, in Leslie Bethell (ed.) The Cambridge History of Latin America, Volume 1, Colonial Latin America, Cambridge. ——(1989) Spain and Its World 1500–1700. Selected Essays, New Haven, CT. FernándezArmesto, Felipe (1988) The Spanish Armada. The Experience of War in 1588, Oxford. Findlay, Ronald and Wellisz, Stanislaw (1982) ‘Endogenous tariffs, the political economy of trade restrictions, and welfare’ , in Jagdish N. Bhagwati (ed.) Import Competition and Response, Chicago. ——(1983) ‘Some aspects of the political economy of trade restrictions’, Kyklos vol. 36. Fisher, Irving (1930) The Theory of Interest, New York. Frank, Andre Gunder (1967) Capitalism and Underdevelopment in Latin America. Historical Studies of Chile and Brazil, New York. ——(1979) Mexican Agriculture 1521–1630. Transformation of the Mode of Production, Cambridge. Frey, Bruno (1984) International Political Economics, Oxford. Gibson, Charles (1977) Los aztecas bajo el dominio español, 3rd edn, Mexico. ——(1984) ‘Indian societies under Spanish rule’, in Leslie Bethell (ed.) The Cambridge History of Latin America, Volume 2, The Colonial Period, Cambridge. Ginés de Sepúlveda, Juan (1941) Tratado sobre las justas causas de la guerra contra los indios, Mexico. Griffin, Keith (1969) Underdevelopment in Spanish America. An Interpretation, London.

68

SPAIN AND THE CONQUEST OF AMERICA

——(1974) The Political Economy of Agrarian Change: An Essay on the Green Revolution, Cambridge, MA. Hamilton, Earl J. (1934) American Treasure and the Price Revolution in Spain, 1501– 1650, Cambridge, MA. Hanke, Lewis (1949) The Spanish Struggle for Justice in the Conquest of America, Philadephia, PA. ——(1956) The Imperial City ofPotosí. An Unwritten Chapter in the History of Spanish America, The Hague. ——(1965) ‘Bartolomé de las Casas, historiador. Estudio preliminar’, in Fray Barto-lomé de las Casas, Historia de las Indias, Tomo I, Edición de Agustin Millares Carlo, 2nd edn, Mexico. Hillgarth, J.N. (1978) The Spanish Kingdoms 1250–1516. Volume 11, 1410–1516, Castilian Hegemony, Oxford. Hunt, Diana (1989) Economic Theories of Development: An Analysis of Competing Paradigms, New York. Hymer, Stephen H. and Resnick, Stephen A. (1971) ‘International trade and uneven development’, in Jagdish N.Baghwati, Ronald W.Jones, Robert A.Mundell and Jaroslav Vanek (eds) Trade, Balance of Payments and Growth. Papers in International Economics in Honor of Charles P.Kindleberger, Amsterdam. Johnson, H.B., Jr (ed.) (1970) From Reconquest to Empire: The Iberian Background to Latin American History, New York. Kamen, Henry (1983) Spain 1469–1714. A Society of Conflict, London. Kindleberger, Charles P. (1951) ‘Group behavior and international trade’, Journal of Political Economy vol. 59. Klein, Julius (1920) The Mesta: A Study in Spanish Economic History, 1273–1836. Cambridge, MA. Koenigsberger, H.G. (1971) ‘The European Civil War’, in The Habsburgs and Europe, 1516–1660, Ithaca, NY. Korth, Eugene H. (1968) Spanish Policy in Colonial Chile. The Struggle for Social Justice, 1535–1700, Stanford, CA. Ladero Quesada, Miguel Angel (1970) ‘Les finances royales de Castille à la veille des temps modernes’, Annales vol. 25. ——(1973) La hacienda real de Castilla en el siglo XV, Seville. Las Casas, Bartolomé de (1965) Historia de las Indias, 3 vols, ed. Agustín Millares Carlo, 2nd edn, México. Logan, Rayford W. (1968) Haiti and the Dominican Republic, London. López de Gómara, Francisco (1966) Cortés. The Life of the Conqueror by His Secretary, transl. and ed. Lesley Byrd Simpson from Istoria de la conquista de México, printed in Zaragoza, 1552, Berkeley, CA. Lynch, John (1964) Spain under the Habsburgs, Volume One, Empire and Absolutism 1516–1598, Oxford. MacLachlan, Colin M. (1988) Spain’s Empire in the New World. The Role of Ideas in Institutional and Social Change, Berkeley, CA. MacLeod, Murdo J. (1984) ‘Spain and America. The Atlantic trade 1492–1720’, in Leslie Bethell (ed.) The Cambridge History of Latin America, Volume 2, Colonial Latin America, Cambridge. Magee, Stephen P., Brock, William A. and Young, Leslie (1989) Black Hole Tariffs and Endogenous Policy Theory. Political Economy in General Equilibrium, Cambridge. Morison, Samuel Eliot (1942) Admiral of the Ocean Sea. A Life of Christopher Columbus (single volume edition), Boston, MA. Mörner, Magnus (1970) La Corona española y los foráneos en los pueblos de indios de América, Stockholm.

69

LATIN AMERICA

——(1985) The Andean Past. Land, Society and Conflicts, New York. ——(1989) ‘Evolución demográfica de Hispanoamérica durante el período colonial’, in Guillermo Marón (ed.) Historia general de América, vol. 14, Carácas. ——(1991) ‘Acerca de los motives de los emigrantes a Indias’, Paper presented to the Symposium América, Continente de la Libertad, Universidad del Norte, Barranquilla, Colombia, 22–24 July 1991 . North, Douglass C. (1981) Structure and Change in Economic History, New York. Olson, Mancur (1965) The Logic of Collective Action. Public Goods and the Theory of Groups, Cambridge, MA. ——(1982) The Rise and Decline of Nations. Economic Growth, Stagflation, and Social Rigidities, New Haven, CT. Parry, J.H. (1940) The Spanish Theory of Empire in the Sixteenth Century, Cambridge. ——(1966a) The Spanish Seaborne Empire, London. ——(1966b) The Establishment of the European Hegemony: 1415–1715. Trade and Exploration in the Age of the Renaissance, 3rd edn, New York. Pierson, Peter (1975) Philip II of Spain, London. Pike, Ruth (1967) ‘Sevillian society in the sixteenth century. Slaves and freedmen’, Hispanic American Historical Review vol. 47. Ruiz Martín, Felipe (1968) ‘Las finanzas españolas durante el reinado de Felipe II (Alternativas de participación que se ofrecieron para Francia)’, in Cuadernos de Historia, Vol. 2, Anexos de la revista Hispania, Madrid. Sánchez-Albornoz, Nicolás (1984) ‘The population of colonial Spanish America’, in Leslie Bethell (ed.) The Cambridge History of Latin America, Volume 2, The Colonial Period, Cambridge. Sauer, Carl Ortwin (1966) The Early Spanish Main, Berkeley, CA. Simpson, Lesley Byrd (1929) The Encomienda in New Spain. Forced Labor in the Spanish Colonies, 1492–1550, Berkeley, CA. Stiglitz, Joseph E. (1974) ‘Incentives and risk sharing in sharecropping’, Review of Economic Studies 41. Thompson, Earl A. (1974) ‘Taxation and national defense’, Journal of Political Economy vol. 82. Ulloa, Modesto (1977) La hacienda real de Castilla en el reinado de Felipe II, Madrid. Vicens Vives, Jaime, con la colaboración de Nadal Oller, Jorge (1975) Manual de historia económica de España, 9th edn, Barcelona. Villamarin, Juan A. and Villamarin, Judith E. (1975) Indian Labor in Mainland Colonial Spanish America, Newark, DE. Wachtel, Nathan (1984), ‘The Indian and the Spanish conquest’, in Leslie Bethell (ed.), The Cambridge History of Latin America, Volume 1, Colonial Latin America, Cambridge. Wallerstein, Immanuel (1974) The Modern World System. Capitalist Agriculture and the Origins of the European World-Economy in the Sixteenth Century, New York. Wolf, Eric R. (1982) Europe and the People without History, Berkeley, CA. Zavala, Silvio A. (1935) La encomienda Indiana, Madrid.

70

5 DO TERMS OF TRADE MATTER? Terms of trade and economic performance in Central America Tarmo Haavisto and Claudio Evald Nalin

INTRODUCTION Development economists have basically, and often in different ways, tried to answer the question: ‘Why are poor countries poor?’ Without disparaging the achievements of the discipline, it would perhaps have been more fruitful if a different question had been posed: ‘Why are some countries rich?’1 After all, economic stagnation and decline rather than growth have constituted the fate of most societies throughout history. Douglas North has provided us with a framework for explaining why stagnation has been more or less a ‘normal’ feature of human society (North 1981) and a key to understanding why certain countries gradually have become more and more efficient (North 1990). He has shown why institutions and their capacity to reduce transaction costs are crucial for the way economies perform. According to North, the difference between rich and poor countries is mainly to be found in differences in institutions. Development economics has, until recently, given scant consideration to the role of institutions in the performance of economies. This is true of the bulk of the research on the economic problems of Latin America, much of which has focused on factors external to these economies. The discussion on the development of these countries’ terms of trade is a case in point. It has been a recurrent theme that Latin American economies, and developing countries in general, face both a secular decline and an extreme short-run variability in their terms of trade. This, in turn, is said to hamper their growth and has been used as an argument in favour of protectionism. While the intensity of the debate may have abated recently, the subject is still given considerable political and scientific attention. This essay deals with movements in the terms of trade of a number of Central American and Nordic countries. Using annual data for the period 1920–84, the following questions are examined. First, is there a discernible difference in the long-run terms of trade between these developing Central American countries and the rich Nordic countries? Second, is there a difference in the variability in

71

LATIN AMERICA

the terms of trade between the two groups of countries? Third, are the terms of trade of either group of countries more subject to intense shocks, negative or positive, than the other? The ultimate purpose of the analysis is to shed light on whether changes in the terms of trade may help to explain the relatively poor growth performance of the developing countries of Central America. The motive for selecting Central America for this enquiry is that data has recently become available for these countries for the long period from 1920 to 1984 (Bulmer-Thomas, 1987). The comparison with the Nordic countries may be justified on the grounds that the Nordic countries, like the Central American ones, are fairly small economies. The structure of the Nordic countries in terms of the role of agriculture in the economy, exports based on natural resources, etc., during the first decades of this century was very similar to that of the Central American countries today. The Nordic and Central American countries differ, however, in terms of their overall economic performance. The growth of GDP per capita has been considerably higher in the Nordic countries than in the Central American countries over the period from 1920 to 1984. The conclusion drawn from this analysis is that the difference in terms of trade movements hardly helps to explain the difference in overall economic performance between the two groups of countries. TERMS OF TRADE AND THE DEVELOPING COUNTRIES The notion that the developing countries face a secular deterioration in their terms of trade is associated with Raul Prebisch (1949) and Hans Singer (1950), who claimed that the terms of trade would move against primary goods. They used a two-sector model whose basic traits were that (a) price and income elasticities are low for primary goods, and (b) the structures of product and factor markets differ in some important respects between rich and poor countries. The Prebisch-Singer thesis has received considerable attention but has also been subject to severe criticism. Bo Södersten (1964), for instance, has argued that their thinking is difficult to fit into a consistent analytical framework. North is also critical of this approach and extends his criticism to the dependency theory: both the writings of the Economic Commission for Latin America (ECLA) and the dependency theory explain the poor performance of Latin American economies on the basis of the international terms of trade with industrialized countries and other conditions external to these economies. Such an explanation not only rationalizes the structure of Latin American economies, but also contains policy implications that would reinforce the existing institutional framework. (North 1990:99–100) Several empirical studies (see e.g. Spraos 1980, Diakosavvas and Scandizzo

72

ECONOMIC PERFORMANCE IN CENTRAL AMERICA

1991; also Krueger 1984, World Bank 1986, Evans 1987) have examined the long-run development of the terms of trade. The empirical picture that emerges appears rather ambiguous. Results do not, however, support the idea of secular deterioration as a universal pattern for developing countries. The determination of secular movements in the terms of trade depends on the time period examined in the various studies, on the individual commodities and countries included, as well as on the different measures applied to the terms of trade. We are left without conclusive evidence on the issue of secular changes: Despite the appeal of the Singer-Prebish thesis, the evidence over the long term has not been supportive of the basic idea of a consistent secular deterioration of either primary product prices as a whole or low-income countries as a group. (Lewis 1989:1548) Furthermore, there is no clear link between changes in the terms of trade and overall economic performance. For example, Mark Gersovitz and Christina Paxton (1990) show that there is no simple and clear-cut relationship between the prices of the commodities exported by African countries and the overall economic performance of these countries. Despite the lack of clear evidence of a downward trend in the terms of trade of developing countries, this remains a very popular theme in the eyes of many Third World governments. For instance, within the framework of the United Nations Conference on Tariffs and Trade (UNCTAD), delegates from developing countries continue to deplore the alleged deterioration in developing countries’ terms of trade, as the following statement exemplifies: Given the continuing downward trend in commodity prices and also the prevailing unpredictability in the commodity market, it is evident that much needs to be done by the [UNCTAD] member states. It calls for renewed multilateral co-operation based on strong political will and broad understanding of the interests of the whole community in reversing the negative trends in commodity markets.2 A similar view is put forth by Gert Rosentahl (1989), Executive Secretary of the Economic Commission for Latin America and the Caribbean, who attributes the poor economic performance of the region to an adverse international economic environment in general, and to deteriorating terms of trade in particular. INSTABILITY OF PRIMARY PRODUCT PRICES In addition to the alleged secular decline in the terms of trade for primary products, the instability of primary product prices has been an important issue both for development economists and for international organizations, as is also

73

LATIN AMERICA

reflected in the quotation above. Instability in primary product markets is considered to be another external factor confronting poor countries which may negatively affect their potential for growth and development. Numerous studies of the variability of commodity prices have been undertaken (e.g. MacBean 1966, Behrman 1978, Krueger 1984, Helleiner 1986). While differences in results have emerged the main conclusions indicate that: • developing countries have generally been subject to larger fluctuations than developed countries; • fluctuations have been substantial in both manufactured and primary product earnings; • greater commodity concentration of export earnings leads to greater earnings instability; • quantity fluctuations have generally been the dominant source of instability in export earnings (Lewis 1989:1551). The next question then is whether fluctuations are inherently bad. Instability is said to damage growth by complicating the task of development planning, by reducing the size of investment, and by even reducing the efficiency with which investment resources are used (Massell 1970:618). The evidence from developing countries as a group is not clear. Most studies suggest that fluctuations in export earnings are not associated with the slowing down of economic growth. Anne Krueger (1984) has suggested that given the multitude of other factors that affect the rate of growth, the impact of instability of export earnings may be hard to distinguish. HIGHER ECONOMIC GROWTH IN THE NORDIC COUNTRIES THAN IN CENTRAL AMERICA Our essay is concerned with five Central American countries (Costa Rica, El Salvador, Guatemala, Honduras and Nicaragua) and five Nordic countries (Denmark, Finland, Iceland, Norway and Sweden). In Table 5.1, the rate of real GDP growth per capita over the period from 1920 to 1979 is shown for these countries. The long-run growth rate of per capita income for the Nordic countries has been higher than the corresponding rate for all of the five Central American countries. None of the Nordic countries has an average growth rate below 2.2 per cent, while none of the Central American countries exceeds 2 per cent. The observed low average growth rate for the Central American countries for the period as a whole does not imply that these countries have not experienced periods of high growth. During the 1920s and 1930s, the economic growth rate in Guatemala was higher than the observed growth rate for any of the Nordic countries. Furthermore, the growth record of Costa Rica since the 1950s is approximately at the same level as that of the Nordic countries.

74

ECONOMIC PERFORMANCE IN CENTRAL AMERICA Table 5.1 GDP growth per capita. Central American and Nordic Countries (average annual changes, per cent, 1920–79)

Sources: For Central America Bulmer-Thomas (1987). For Nordic countries the growth rates of the early years of our study can be found in Hjerppe (1985) for Finland, Krantz and Nilsson (1975) for Sweden, Bjerke (1966) for Norway, Ölgaard (1966) for Denmark. The more recent years are provided by the statistical yearbooks of each country. For Iceland the measures were kindly made available for us by the statistical department of the Central Bank of Iceland

THE SECULAR DEVELOPMENT OF TERMS OF TRADE The relatively low long-run GDP growth for the Central American countries is often attributed to unfavourable movements in their terms of trade. As mentioned above both the secular development of terms of trade and their short-run variability allegedly contribute to this poor performance. By use of the method of descriptive statistics we examine whether systematic differences in the terms of trade pattern can be observed between the Central American countries, on the one hand, and the Nordic countries, on the other. We look at net-barter terms of trade and the sources of the data are the same as those used for Table 5.1. Following ‘textbook’ methodology in statistics, the descriptive characteristics of data are location, dispersion and shape.3 The corresponding measures for location and dispersion are the mean and standard deviation,4 respectively, while the shape (of a frequency) is measured by higher central moments, skewness and kurtosis. If the differences in the long-run economic performance between the Nordic countries and the Central American countries could, at least in part, be attributed to changes in the terms of trade, we would expect: 1 2 3

that the Central America countries had a less favourable long-run terms of trade trend than the Nordic countries, and/or that the terms of trade of Central American countries shows a higher variability than the Nordic countries terms of trade, and/or that the Central American countries have a higher frequency of extreme terms of trade shocks and, also, more frequent negative shocks than positive ones.5

Thus, we are mainly concerned with possible differences between the two groups of countries rather than the size of the observed values per se.

75

LATIN AMERICA

Let us start with the long-run terms of trade. The changes that have occurred in the terms of trade of all ten countries between 1920 and 1984 are shown in Figure 5.1. The pattern that emerges from Figure 5.1 is that no sizeable difference in the long-run development of terms of trade is discernible between the two groups of countries. It would be difficult to single out the group of the Central American countries from our sample of ten countries by visual observation. We can hardly identify any country in the sample that has clearly fared worse than another. Additional information from the time series is obtained by calculating the mean values for the logarithm of the observed terms of trade values. These are presented in Table 5.2 which also provides standard deviation for terms of trade observations and extreme values. Over the whole period, one of the Nordic countries, Denmark, and three Central American countries experienced a negative change in their secular terms of trade. However, differences between countries are fairly small and average decline exceeds 1 per cent per year only in the case of Nicaragua (–1.5 per cent). For all other countries, the mean value is between ±1 per cent. The highest positive value is observed for Iceland, with an average annual growth of +1 per cent. Mean values for eight of the ten countries are very close to zero. In our view, the fairly small differences in the secular movement of terms of trade can hardly explain the sizeable difference in the long-run growth rate between the countries in our sample. The terms of trade performance of Nicaragua is somewhat less favourable than that of the other countries. However, it should be noted that the drastic deterioration during the first year of our study, from 1920 to 1921, accounts for most of the difference. During the subsequent period, the Nicaraguan terms of trade have not developed in a strikingly different way from those of the other countries. If one argued that this deterioration explains the poor performance of the Nicaraguan economy, one should also explain why an adequate structural change in response to the price fall in 1921 has not taken place during the subsequent six decades. TERMS OF TRADE VARIABILITY Turning to terms of trade fluctuations, we note from Table 5.2 that three of the countries have standard deviations below 10 per cent. For those three Nordic countries, i.e. Finland, Norway and Sweden, changes in the terms of trade have rarely reached a value above 10 per cent; 68 per cent of all observations are below 8.2, 6.5 and 6.1 per cent, respectively. For the majority of the countries, including the two remaining Nordic countries as well as four of the Central American countries, the standard deviation varies between 10 and 20 per cent. It is only in the case of El Salvador that the volatility exceeds 20 per cent (21.8 per cent). Costa Rica has experienced terms of trade fluctuations of the same magnitude as Nordic countries like Denmark and Iceland. Although the volatility is higher for the group of five Central American

76

Figure 5.1 Terms of trade for five Central American and five Nordic countries 1920-84, log scale Sources: Table 5.1 In the Swedish case the terms of trade data is provided by Fridlizius (1964)

LATIN AMERICA Table 5.2 Mean values, standard deviations and extreme values

countries than for the Nordic countries, the difference between the two groups appears rather small. Iceland, with a standard deviation at the same level as the Central American countries, 13.6 per cent, has a GDP growth performance which is significantly higher than for any of the Central American countries. TERMS OF TRADE SHOCKS So far we have only been concerned with mean values and standard deviations of changes in the terms of trade. We have, thus, implicitly assumed that terms of trade changes are normally distributed. However, time series may contain more information which is relevant for the possible impact of terms of trade. Table 5.2 indicates that the Central American countries as a group have experienced both lower minimum values and higher maximum values than the Nordic countries.6 If the broad span between minimum and maximum values also indicates that the Central American countries have suffered more frequently from extreme shocks than the Nordic countries, then that pattern should be reflected in different shapes in the distributions of terms of trade changes between the two groups of countries. If we have more observations (than would follow from the normal distribution) several standard deviations away from the sample mean, this would influence the peakedness of the distribution. Furthermore, if the distribution is asymmetric, e.g. the Central American countries have been exposed to more frequent negative shocks, the distribution becomes skewed, and the coefficient of skewness would differ between the two groups of countries. Frequent and large negative shocks could be particularly harmful to economic growth. Such negative terms of trade shocks could for example, by leading to dramatically reduced profits, be a strong disincentive to investment.7 The skewness of terms of trade changes measures the departure from symmetry. If changes in terms of trade had been symmetric (symmetric changes

78

ECONOMIC PERFORMANCE IN CENTRAL AMERICA

on both sides of the mean value), then the measure of skewness would be equal to zero. If the left tail of the distribution is elongated, i.e. we have skewness towards the low values, the measure is negative. If terms of trade changes are skewed towards high values, the series exhibit positive skewness, i.e. relatively high frequency of high positive changes in the terms of trade. If a country displays a high value of kurtosis, indicating extreme shocks, and if, at the same time, the value of skewness is negative, we have a shape that reflects negative extreme shocks. For the sake of illustration, Figure 5.2 shows the frequency plots of terms of trade changes for two countries, Costa Rica and Nicaragua. The frequency plot for Nicaragua is more peaked than for Costa Rica (Nicaragua has more extreme outliers than Costa Rica). In Table 5.3 this is seen from the higher value of kurtosis for Nicaragua than for Costa Rica. Figure 5.2 also shows that the extreme outliers in the Nicaraguan case are negative shocks, reflected in Table 5.3 by negative skewness coefficients. As seen in Table 5.3, the terms of trade series are skewed to the right for all countries, with the exception of El Salvador and Nicaragua. Note also that the observed negative skewness of El Salvador is very low. Therefore, only Nicaragua shows a pattern of negative terms of trade shocks, measured by a negative skewness. Table 5.3 also suggests that the coefficients of skewness are higher for the Nordic countries than for the Central American countries. However, we also note that two of the Central American countries, Honduras and Costa Rica, have coefficients on the same level as three Nordic countries, Denmark, Finland and Iceland. Only the Norwegian and Swedish values are considerably higher. With high kurtosis, however, the measure of skewness is highly sensitive for one or two ‘outliers’, and the negative skewness for Nicaragua reflects the great fall in terms of trade from 1920 to 1921. We cannot observe any systematic difference in peakedness between the two groups of countries. Though the distribution is rather peaked for Honduras and Nicaragua, it is no worse than for Denmark. The distributions within each group vary more than the distributions between the two groups of countries. The kurtosis coefficients lie between 4.330 and -0.328 for the Nordic countries, while the corresponding figures for Central America are 4.367 and -0.364. A normality test of distribution confirms this conclusion. At a 5 per cent level, ‘normality’ is accepted for two of the Central American countries, Costa Rica and Guatemala, and for one of the Nordic countries, Iceland. WHY DIFFERENT ECONOMIC PERFORMANCES? It appears as if changes in the terms of trade cannot provide an explanation of the difference in growth performance between the Nordic and the Central American countries during the period examined in this essay. We have to look elsewhere to find out why growth has differed, and, in our view, we ought to turn to institutions and the incentives that they create.

79

Figure 5.2 Frequency plot of terms of trade changes

ECONOMIC PERFORMANCE IN CENTRAL AMERICA Table 5.3 The higher central moments

As mentioned in the introduction, the economic structure of the Nordic countries at the turn of the nineteenth century was very similar to that of the least developed countries of today (Blomström and Meller 1991). Agriculture was still the dominant activity in the 1870s and played an important role during the first decades of the present century. As much as 50 per cent of the Swedish labour force was still employed in agriculture at the beginning of this century. Primary exports have played an important role for the Nordic countries as well as for the Central American countries. A main difference between the Central American and Nordic countries lies in the way that the state and institutions have performed. In the introduction, we referred to Douglas North, who has pointed out the crucial importance of stable and secure property rights for economic growth. While stable and secure property rights have gradually developed in the industrialized countries, in most developing countries insecurity seems to prevail. According to North ‘the inability of societies to develop effective, low-cost enforcement of contracts is the most important source of both historical stagnation and contemporary underdevelopment in the Third World’ (North 1990:54). This is not the place to make a thorough examination of institutional differences between the Central American and Nordic countries. But let us very briefly highlight certain differences in the behaviour of the state in the two regions, and in particular focus on the difference in property rights in agriculture. There seems to be a striking difference in the way that the state has behaved in Central America and in the Nordic countries, respectively, during the present century. In Central America, rulers appear to have been mainly concerned with their personal income and welfare and a predatory pattern of behaviour has not been uncommon in the region. In countries like Haiti and the Dominican Republic, dictators like François Duvalier, Jean-Claude, and Rafael Trujillo accumulated huge fortunes and most of their policies seem to have served this

81

LATIN AMERICA

purpose. The incentives created in these countries were not such as to stimulate economic growth (Lundahl and Vedovato 1989). In the countries included in this essay, the Somozas in Nicaragua perhaps stand out as the most conspicuous predatory rulers. Anastasio Somoza accumulated a fortune of about 400 to 500 million dollars. Richard Millet (1977:198) observes ‘Graft was common in Nicaragua, but Somoza refined and systematized it’. But also in countries like Guatemala, El Salvador and Honduras, various caudillos have been engaged in fights over the spoils of government. Bulmer-Thomas (1987:17) contends that control of the presidency in Central America has been the surest and quickest way to enrichment. He also notes that The struggle for power and the spoils of office assume particular ferocity in countries where the opportunities for enrichment through economic activities are seriously limited’. The recent experience of the Nordic countries is quite different. Here, during the last centuries, the state has gradually carried out changes that have stimulated economic growth. Stable and secure property rights have developed and the state has created institutions that can enforce contracts.8 Property rights in agriculture represent a striking difference between the Nordic countries and Central America. Swedish society at the start of the nineteenth century was still a genuine peasant society, in the sense that over 90 per cent of the population lived, worked and died in the countryside. Property rights in agriculture had been changing since the early eighteenth century and went on changing during the nineteenth century. At the beginning of the eighteenth century, over 30 per cent of the land was farmed by peasants who were masters of their own production. By 1850 this share had increased to over 60 per cent (Österberg 1990:271). Several centuries ago, the Swedish peasants had established channels for contact with authority and had institutions of their own. Conditions were similar in other Nordic countries. Österberg (1990:272) claims that in the Nordic countries (Denmark, Norway, Finland and Sweden) ‘peasants undeniably achieved some success in their action vis-à-vis the central government’ and could for example present complaints to the several hundred existing courts. Land reforms took place not only in Sweden but also in Denmark (from the end of the eighteenth century) and in Norway. Finland was a latecomer, with major reforms in the 1920s. The reforms created small and medium-sized privately owned farms with secure property rights. The Nordic countries had a legal system that could enforce contracts and guarantee property rights. In Central America, individuals have instead been subject to arbitrary behaviour from their rulers and even today insecurity appears to be a prominent feature in the countryside. Surveys in Honduras and Nicaragua find that most of the rental contracts are unwritten and usually cover less than one year (Brocket 1988:69). Contracts are frequently not renewed. One technique that has been applied throughout the region has been to rent out uncleared land at a nominal price for a short duration. At the end of the contract, the land, now cleared by the renter, has been converted to cattle grazing and the former renter must move to

82

ECONOMIC PERFORMANCE IN CENTRAL AMERICA

new land. Earlier this century, peasants were pushed away from their land on a large scale. Somoza, for example, became Nicaragua’s largest landowner by simply dispossessing previous owners from their land. Trujillo did the same thing in the Dominican Republic. It is obvious that such conditions have not provided the stable and secure conditions needed for long-run investment and economic growth. In Costa Rica peasants did not experience the same mass scale of land dispossessions suffered by peasants in the rest of Central America. In Costa Rica there is also a relatively large proportion of independent small farmers, a factor which has been said to have contributed to Costa Rica’s political stability. Such conditions could explain the good growth performance of Costa Rica compared with that of the other countries in the region. CONCLUDING REMARKS This essay has examined secular movements in the terms of trade for the period from 1920 to 1984 for five Nordic and five Central American countries. We have looked at the secular growth of the terms of trade, its variability and the frequency of extreme shocks. On the basis of this analysis, we find it difficult to attribute the observed differences in overall growth between the two groups of countries to differential movements in the terms of trade. It is our belief that one should look instead at differences in the way that the state has performed, the security and stability of property rights, etc., in order to explain divergences in growth performance. ACKNOWLEDGEMENTS Nalin thanks SAREC for financial assistance. NOTES 1 2 3 4

5 6

North claims that: ‘The disparity in the performance of economies and the persistence of disparate economies through time have not been satisfactorily explained by development economists, despite forty years of immense effort’ North (1990:11). Statement by the Indian ambassador as spokesman for the Group of 77 in preparation of UNCTAD VII. See e.g. Spanos (1986:23–7). We use the standard deviation of changes in terms of trade, instead of the volatility of the residual term from a regression of terms of trade against trend (as is done in several previous studies). Our argument is that the assumption of a particular time pattern for a trend (e.g. linearity) poses certain problems. We therefore use the familiar method used in empirical studies in the field of financial economics and study volatility in terms of the standard deviation of changes. This latter argument has not, as far as we have observed, been dealt with in any depth in the literature. There is no evident link between extreme values of terms of trade changes and GDP

83

LATIN AMERICA

7 8

growth rates. Costa Rica with the lowest span between minimum and maximum values (shorter than the Danish span and in line with the Finnish) among the Central American countries has undeniably experienced the region’s highest growth rate (although of course lower than that of Denmark and Finland). But El Salvador with the largest span of all countries has the next highest growth rate of the Central American countries, only slightly below Costa Rica’s. The problem can be accentuated by deficient capital markets which could make it difficult to finance sharp temporary falls in profits. The question of why institutions have become more efficient in the Nordic countries than in Central America is far beyond the scope of this essay. See North (1990) for a fruitful framework for discussing institutional change.

BIBLIOGRAPHY Behrman, J.R. (1978) Development, the international economic order, and commodity agreements, Reading, MA: Addison-Wesley. Bjerke, J. (1966) Trends in Norwegian Economy 1865–1960, Oslo: Statistisk Sentralbyra. Blomström, M. and Meller, P. (1991) ‘Issues for development: Lessons from ScandinavianLatin American comparisons’, in M.Blomstrom, and P.Meller, (eds) Diverging Paths: Comparing a Century of Scandinavian and Latin American Economic Development, Baltimore, MD: Johns Hopkins University Press. Brocket, C.D. (1988) Land, Power, and Poverty. Agrarian Transformation and Political Conflict in Central America, Boston, MA. Allen & Unwin. Bulmer-Thomas, V. (1987) The Political Economy of Central America since 1920, Cambridge, MA: Cambridge University Press. Diakosavvas, D. and Scandizzo, P. (1991) ‘Trends in the terms of trade of primary commodities, 1990–1982: The controversy and its origins’, Economic Development and Cultural Change vol. 39. Evans, D. (1987) ‘The long-run determinants of North-South terms of trade and some recent empirical evidence’, World Development vol. 15. Fridlizius, G. (1964) ‘Sweden’s export 1850–1960- a study in perspective’, Economy and History vol. VI. Gersovitz, M. and Paxton, C. (1990) ‘The economies of Africa and the price of their exports’, Princeton Studies in International Finance no. 68, October. Helleiner, G.K. (1986) ‘Outward orientation, import stability and African economic growth: An empirical investigation’, in S.Lall and F.Stewart (eds) Theory and Reality in Economic Development, London: Macmillan. Hjerppe, R. (1985) The Finnish Economy 1860–1985. Growth and Structural Change, Bank of Finland Publications. Kenen, P.B. and Voivodas, C.S. (1972) ‘Export instability and economic growth’, Kyklos vol. 25. Krantz, O. and Nilsson, C.-A. (1975) Swedish National Product 1861–1970, New Aspects on Methods and Measurement, Lund: CWK Gleerup. Krueger, A.O. (1984) ‘Trade policies in developing countries’, in R.W. Jones and P.B. Kenen (eds) Handbook of International Economics, vol. 1, Amsterdam: NorthHolland. Lewis, S.R., Jr (1989) ‘Primary exporting countries’, in Handbook of Development Economics, vol. II, Amsterdam: North-Holland. Lundahl, M. and Vedovato, C. (1989) ‘The state and economic development in Haiti and the Dominican Republic’, Scandinavian Economic History Review vol. 37. MacBean, A. (1966) Export Instability and Economic Development, Cambridge, MA: Harvard University Press.

84

ECONOMIC PERFORMANCE IN CENTRAL AMERICA

Massell, B.F. (1970) ‘Export instability and economic structure’, American Economic Review vol. 60. Millett, R. (1977) Guardians of the Dynasty, Maryknoll, NY. North, D.C. (1981) Structure and Change in Economic History, New York: Norton. ——(1990) Institutions, Institutional Change and Economic Performance, Cambridge, MA: Cambridge University Press. Ölgaard, A. (1966) Growth Productivity and Relative Prices, Studier fra Köpenhavns Universitets Ökonomiske Institut, vol. 10. Österberg, E. (1990) ‘Compromise instead of conflict? Patterns of contact between local peasant communities and the early modern state in Sweden in the sixteenth to eighteenth centuries’, in M.Lundahl and T.Svensson (eds) Agrarian Society in History. Essays in Honour of Magnus Mörner, London: Routledge. Prebisch, R. (1949) Economic Progress in Underdeveloped Countries, United Nations Economic Commission for Latin America. Rosentahl, G. (1989) ‘Some thoughts on poverty and recession in Latin America’, Interamerican Studies and World Affairs vol. 31. Singer, H. (1950) ‘The distribution of gains between investing and borrowing countries’, American Economic Review, Papers and Proceedings, vol. 40. Södersten, B. (1964) A Study of Economic Growth and International Trade, Stockholm: Almqvist & Wicksell. Spanos, A. (1986) Statistical foundations of Econometric Modelling. Cambridge, MA: Cambridge University Press. Spraos, J. (1980) ‘The statistical debate on the net barter terms of trade between primary commodities and manufactures’, Economic Journal vol. 90.

85

6 ELEMENTS DETERMINING THE COMPETITIVENESS OF GUATEMALAN PRIMARY EXPORTS Lars Pira

INTRODUCTION Ever since the colonial period, developing countries have been exchanging primary export products for manufactured imports from the more developed parts of the world. In many of the colonies, this trade pattern was established by the mother country’s demand for agricultural and mining products which were not available at home. But contrary to other colonies, where trade ties with the mother country remained after independence (Frey 1984:2–3), Latin American independence from Spain meant that exports could be directed to other more profitable markets. The period after independence coincided with rapid industrial growth in Europe, increasing demand for raw materials and agricultural products. Latin America was able to seize this opportunity, and the high rates of growth experienced during this time were due to a rapidly expanding export sector, which had managed to become increasingly integrated into the world market system (Glade 1986:7). Although the demand from Europe helped the Latin American countries increase exports, the decision of these countries to produce has also depended on transportation costs and adaptive changes in factor markets, enabling them to become and remain cost competitive in certain products. These changes were not the same for all countries: some were endowed more than others with natural resources; export crops were produced by allocating and managing land resources differently; labour was imported in some parts, while the labour supply was abundant in others; the flow of capital from the industrializing economies, which was crucial for the growth of export production, played a very dynamic role and its implications differed according to area and product (Glade 1986:22–46). By analysing the historical evolution of the Guatemalan export economy, this essay demonstrates the importance of theoretical analysis in explaining how a

86

GUATEMALAN PRIMARY EXPORTS

small Latin American country initiated trade and has been able to maintain competitiveness. Basically, the essay will attempt to answer the following questions. What was the relative importance played by access to markets and adaptive changes of factors in the determination of competitiveness? What has governed the adaptive change in the employment of factors and why have the paths followed in different situations differed? STAPLES THEORY AND INSTITUTIONAL ANALYSIS The primary importance of demand for the initiation of trade is specified in the ‘staples’ theory, which tells us that exports of staples cannot take place in the absence of an international market. A staple is defined as an export product largely composed of a natural resource and whose production does not necessarily require elaborate processing, nor involve large amounts of labour and rare skills.1 Transportation costs determine whether the primary commodity can have access to the market at a competitive price and technological change will render resources cost competitive in world trade.2 Once all these elements have been defined, the production for exports of a particular staple ‘will be allocated among countries according to their relative production costs, viewing costs as dependent in a very broad fashion on the physical traits of various regions’ (Caves and Nolton 1959:35). Contrary to the Heckscher-Ohlin model, which assumes given factor endowments, an important condition for the ‘staples’ theory is the existence of factor mobility across boundaries. Factor mobility is also specified in the ‘vent for surplus’ theory (Myint 1958), but factors in this case are employed in order to take advantage of idle resources prior to initiating trade. This has been the case in many Latin American countries, where foreign investment and entrepreneurship and sometimes imported labour were important to set in motion vast areas of formerly unused land for the production of staples (Glade 1986). The ‘vent for surplus’ and the ‘staples’ theories can be integrated by including into the model the concept of an endogenous land frontier (Findlay and Lundahl, forthcoming). The ‘stylized facts’ of this integrated model tend to conform in the opening up process of trade, no matter whether it analyses: (a) ‘regions of recent settlements’, where production of almost ‘empty’ lands was induced through substantial inflows of capital and labour from Europe, such as was the case with the temperate zones of the Americas and Australia; (b) ‘plantation economies’, where production of staples was initiated by both domestic and foreign capitalists and entrepreneurs, who employed both local and immigrant labour, in what previously had been mostly subsistence economies; or (c) ‘peasant export economies’, where native peasants turned from subsistence to export production (Findlay and Lundahl, forthcoming). The premises underlying the ‘staples’ and the ‘vent for surplus’ theories cannot by themselves explain, however, why the historical evolution of trade has differed so drastically between any two countries, despite similar demand

87

LATIN AMERICA

patterns, transportation costs and resource endowments when the opening of trade took place. The different paths followed by each country in the development of an export sector can in great part be attributed to the social and political arrangements adopted by institutions, which have governed the adaptive change in the employment of factors. Institutions are created in order to reduce transaction costs involved in defining, protecting and enforcing property rights to goods, since markets are imperfect and information is incomplete (North 1990). However, owing to the uncertainty of organizing transactions involving trials, experiments and innovations, institutional arrangements could sometimes become more costly than the use of the market mechanism (Coase 1960). Moreover, costs incurred in the production of a staple, which also depend on costs of monitoring and policing the employment of factors, do not necessarily become socially optimal, if the institutional framework does not permit all of the interest groups involved in production to lobby for their interests (North 1990:48). Thus, production decisions resulting from the same demand conditions and resource endowments and employing the same physical inputs of land, labour and capital could take different paths in any two countries due to the particular role played by institutional arrangements (North 1990:113–17). The inclusion of institutional analysis does not preclude the use of any of the theoretical tools already developed: it only extends the neoclassical scarcity/ competititon postulates and the central role played by incentives to incorporate ‘incomplete information and subjective models of reality and the increasing returns characteristic of institutions’ (North 1990:112). Since the study of institutions evolves in a historical path-dependent pattern, it helps us understand the role played by the factors or set of factors which the ‘staples’ and ‘vent for surplus’ theories employ to explain export competitiveness. This includes not only the relative importance played by these factors in a particular time frame, but also why and how this importance changed over time. EARLY ATTEMPTS TO ESTABLISH EXPORTS 1520–1870 The establishment of an export economy in colonial Guatemala can be divided into two phases. The first could be described as one of plundering for gold and slaves, where the Spaniards had no intention of settling in the conquered areas. The second, which was devoted to the search for alternative exports by the settlers following the arrival of the first conquerors, was characterized by a long-lasting process of trial and error until a key export product was found (MacLeod 1980). Demand and transportation costs The initiation of Guatemalan exports depended primarily on demand opportunities. The main staple during the colonial period was indigo, especially

88

GUATEMALAN PRIMARY EXPORTS

during the second half of the eighteenth century, when Spanish merchants increased demand for this dyestuff to supply the growing European textile industry (Jones 1939:148,197–200). But except for small quantities of balsam and sarsaparilla, there were hardly any other products exported to Spain during this time. Most Guatemalan exports were directed to satisfy demand originating in Latin America, especially in Mexico. Indigo was already grown and exported in Guatemala for more than a hundred years before being shipped to Europe and the cacao export boom experienced in the Guatemalan lowland at the end of the sixteenth century was due to growing Mexican demand. However, attempts to become integrated into world trade during the colonial period and the first fifty years of independence were sporadic and short lived and, for the most part, subsistence production dominated the local economy. Demand in Latin America was limited and increased supply from other producers reduced competitiveness. Cacao was outcompeted by increased access to the Mexican market of cheaper and lower-quality cacao from Guayaquil and Venezuela. Indigo exports to Europe lost their competitive position by the end of the eighteenth century because of lower-price exports from India. Cochineal, which was introduced as a substitute to indigo during the post-independence period from 1821 to 1871, also lost its competitiveness when synthetic dyestuffs were introduced (Woodward 1981:92). One important reason why Guatemala was unable to establish a more stable export economy outside Latin America was that transportation costs were too high in relation to the price the product could fetch once the market was reached. The further away the colony the more likely it was to produce an export which was small in bulk and light in weight so that an adequate profit margin could be ensured despite a long voyage.3 Guatemala did not have any significant deposits of silver and gold, and it had been more profitable to export slaves to Panama and Peru and cacao to the growing Mexican market than to ship balsam and sarsaparilla, for example, to the mother country. Not only had there never been enough demand in Spain for these products to become major exports, but unstable prices could never guarantee that costs would be covered (MacLeod 1980:55–7). The small shipments of indigo to Spain were possible because bulk and weight were relatively limited in relation to the price in the final market and sufficiently large profit margins could be obtained (MacLeod 1980, Smith 1959). Adaptive change of land, labour and capital Like demand conditions and transportation costs, the way factors were adaptively used in determining competitiveness during the colonial period changed according to product. Exports of cacao and indigo took place as specified in the ‘vent for surplus’ theory. Competitiveness in cacao exports was successfully gained by controlling the population which already produced a surplus of this staple prior to the arrival of the Spanish conquerors. An initial thrust for exports of indigo took place by setting into production large amounts of unused land resources.

89

LATIN AMERICA

Although work in the cacao plantations was not as physically demanding as in the Caribbean sugar mills, many skilled workers were required to tend to the very delicate high-quality Guatemalan cacao fruit. The encomienda system4 provided an effective way of securing cheap labour without forcing the Indian population out of its climatic and cultural environment. The Indians were able to return to their villages to tend to their corn and other food production while supplying the encomendero with labour and tribute in kind to make cacao exports a thriving business (MacLeod 1980:95). Epidemics reduced the population under the encomienda system, however, and the acute labour shortage which ensued reduced the competitive edge gained by Guatemalan cacao. Although proximity to the Mexican market had been an important factor for the successful exports of cacao, lower quality cacao from Guayaquil and Venezuela became more cost competitive even though produced further away (MacLeod 1980:63–4,131). The encomienda system permitted the Spanish cacao traders to enjoy a monopsony in the demand for labour without involving land rights (Lundahl 1993). Acquisition of large tracts of land by Spanish descendants took place particularly during the eighteenth century when indigo became an important staple. Land was largely available during this period of time because it laid fallow and abandoned after epidemics had taken their toll (Carmack 1979:39). While the employment of large tracts of unused land made possible the initial production of indigo, the use of labour became increasingly determined by relative scarcities. Competitiveness depended on how effective producers were in coercing a scarce labour supply to harvest indigo on land whose use presented little opportunity cost to the economy. Unlike the cacao tree, which needed careful tending by a constant workforce, the more weather- and disease-resistant indigo plant did not require much attention during its growth period and labour shortage problems only affected the one or two months of the year when harvesting took place (Smith 1959). As a matter of fact, one of the reasons why this dyestuff initially became an important staple in colonial Guatemala was because it was able to resist depressed periods with low populations better than other crops. Implementation of the repartimiento (conscription) system permitted the landlords to force the indigenous population down from the mountains only for employment during the harvesting season.5 Other arrangements, such as debt peonage, were employed to guarantee the small amount of permanent workers needed on the farm.6 Some technical improvements were introduced with indigo production. Since better-quality dyes always fetched a higher price, growers were concerned about improving agricultural practices, such as employing better seed, as well as treating land as an exhaustible resource. Certain skills and dexterity were required by labour employed in the dyeworks in order to assure quality and prevent losses during the transformation process. However, the modernization process undertaken by indigo growers, and even those involved in cochineal production and trade during the first half of the nineteenth century, ‘had extremely restricted

90

GUATEMALAN PRIMARY EXPORTS

impacts on both land use and labor patterns’ (McCreery 1976). Exports of these products played only a marginal role as an engine of growth for the Guatemalan economy. Export enterprises were never developed to the same degree as the Caribbean plantations or Andean mining activities. The role of public and private institutions Institutions played an important role in determining export markets and the adaptive change of factors during the colonial period. For example, regulations imposed on trade and production by Spanish institutions had negative consequences for export growth in Guatemala. Since the main interest of the mother country was to maximize its stock of precious metals, in order to improve the seriously deteriorated financial position of the Crown,7 countries without important mineral resources, such as Guatemala, became relegated to a marginal position with respect to trade. Guatemala contributed to the Spanish crown by means of a heavy tax burden on indigo production and exports, which reduced competitiveness of this product even further (Smith 1959, Facio 1965). In order to control the trade flow with the colonies, Spain had established in 1503 the ‘Casa de la Contratación’, a trading house which by exerting its monopoly position did not permit the colonies to open trade links with other countries (Hamilton 1948:40–3). As a consequence, Guatemala was not able to take advantage of the more profitable European markets for its export of indigo. Since indigo growers were not able to trade with other than Spanish merchants, these always managed to turn conditions to their advantage, such as when they established lower prices by manipulating quality grading (Floyd 1961). Trade and navigation reforms introduced by Spain after 1765 only worsened trade conditions for Guatemala. Spanish colonies had to continue to trade exclusively with the mother country and trade between the colonies was prohibited even further, which meant that ships were no longer allowed to change course or unload in a port other than the one specified when leaving Spain (Acuña 1980:26–7). This situation increased the share of costs and reduced profit margins, since transportation often had to be arranged through Mexican ports.8 In view of the above situation, trade was often illicitly organized through Belize to bypass Spanish traders and gain access to the expanding British manufacturing industry, which provided a profitable market for Guatemalan natural dyes (Cohen 1980:32). The Spanish monopoly control on trade, which hampered the expansion of Guatemalan indigo production, together with the increasing ties being established with British traders, is one of the most important factors contributing to the success of the independence movement (Facio 1965). It is also not surprising to find that after independence from Spain ‘Great Britain and Central America gravitated toward one another as each sought in the other a partial solution to pressing economic problems’ (Naylor 1969:364).

91

LATIN AMERICA

Certain domestic reforms were launched by Spain some fifty years prior to independence in order to reduce political tensions in colonial Guatemala. Domestic growers and traders were allowed to form their own guilds, such as the Indigo Grower’s Society in 1772 and the Consulado of Guatemala in 1794. The first was formed to increase the power of indigo growers when negotiating terms and conditions for financing and determining prices. The second was created by domestic merchants to break the monopoly power of Spanish traders and safeguard themselves from the indigo growers’ newly acquired control over the determination of prices (Smith 1946, Smith 1959, Woodward 1981:25–6). The Indigo Grower’s Society was replaced by the Economic Society after independence. But until 1870, these institutions maintained the same colonial structure and power groups. Guilds and associations were formed because one group wanted to strengthen its position vis-à-vis another within the colonial elite. The Consulado of Guatemala was particularly slow in adopting new techniques and in entering other lines of production and its monopoly on trade proved to be a serious impediment for the further expansion of the export economy (Woodward 1981:92–3). As already mentioned, availability of labour was secured through institutional arrangements, such as encomienda and repartimiento. The reason why these institutions were able to exist was because the marginal cost of maintaining labour was lower than the marginal benefit of its contribution to production.9 The relative bargaining power of labour (property rights) has thus depended on how much the marginal cost of metering and monitoring has increased in relation to the marginal benefit obtained in production. According to Evsey Domar, in relatively archaic agricultural structures with land, labour and landowners, where capital and management have played an insignificant role, the elements of free land, free labour and non-working landowners cannot all exist simultaneously (Domar 1970:21). As long as land was free, the non-working Spanish settlers used the encomienda system to extract a surplus from labour. When labour became scarce, land ownership became important, but in order to increase their leisure time and extraction of surplus, landlords recurred to other institutional methods for procuring cheap labour, such as repartimiento and debt peonage. The marginal cost of enforcing institutional arrangements to control the labour force became higher, at times, than the marginal benefit obtained. The loss of competitiveness of cacao was largely due to the increase in the cost of enforcement when the size of the labour force, needed to tend to the very delicate and highquality cacao plant, was reduced by disease. The relative cost of employing the repartimiento system for indigo harvesting increased as well because of a reduction in the size of the population (MacLeod 1980:174–5). Attempts to substitute part of this labour with imported black slaves were unsuccessful (Smith 1959:189). The export competitiveness of Guatemalan indigo was thus partly lost due to an increase in production costs relative to those in India.

92

GUATEMALAN PRIMARY EXPORTS

Institutional arrangements for controlling labour became less important in cochineal production. In this instance, peasants became free producers on an increasingly scarce amount of land and resources. The non-working landowner diminished in importance and profits depended on the relative bargaining power of the peasant vis-à-vis the powerful trader. EXPORT-LED GROWTH 1870–1955 It was not until the promotion of coffee production by the Liberal government in the 1870s and the establishment of banana export production by the United Fruit Company (UFCO) during the first decade of the twentieth century that Guatemala was able fully to integrate its economy into world trade and establish the necessary conditions for export-led growth. High-quality coffee was able to grow because of the required elevation and bananas were produced efficiently and at low cost on the Atlantic coast, where rain and humidity are constant all year round. However, the successful initiation of exports cannot be explained by a simple availability theory based on climatic and geographical considerations. As in the staples theory, international demand was the catalyst, and market access at competitive prices was possible due to the modernization of transportation. The degree of cost competitiveness depended also on the adaptive change of labour, land and capital. Demand and transportation costs Guatemalan coffee and banana exports were spurred by rising demand brought about by the emerging middle and working class in industrial Europe and North America. By the end of the first decade of the twentieth century, Guatemala was one of the most important producers of Milds, the highest-quality coffee in the world.10 Consumer acceptance of bananas was accompanied by efficient retail marketing, facilitating the introduction of this perishable fruit in the most remote areas of North America and Europe.11 Advances in overseas shipping allowed for prompt delivery and lower transportation costs than prior to mid 1800 (Mosk 1958:168). For example, competitiveness of Guatemalan coffee was discouraged for a time by monopoly prices charged by the Panamanian Railroad and the Pacific Mail Steamship Company. Modernization of the shipment service brought about alternative arrangements which decreased the cost of transporting coffee to Europe to about half the price.12 Also, technological improvements in transportation made it possible to export the perishable banana fruit.13 One of the main bottlenecks for exports had been the inadequate domestic transportation facilities. The roads available when the coffee economy was first introduced in Guatemala were essentially those which had remained from the colonial period.14 Domestic transportation improved continuously after 1870 and, especially, during the first two decades of the twentieth century, when the

93

LATIN AMERICA

UFCO undertook investments in the railroad, port and telecommunication systems. Adaptive change of land, labour and capital While technological change in transportation, which allowed access to the new emerging markets, set the conditions for promoting export-led growth, production of coffee and bananas could rapidly take place by expanding a largely unutilized agricultural frontier in a ‘vent for surplus’ fashion. As a result, the initial stage of coffee and banana export production did not involve a significant opportunity cost for the economy, since the cultivable area had little alternative use. Coffee production expanded primarily on expropriated church property which formerly was unused by the clergy and, contrary to some nineteenth century Liberal regimes in Latin America, the Guatemalan one did not have to resort to sweeping land confiscations of productive rural communities for this purpose (McCreery 1983:739, Facio 1965:35–6). The initial growth of banana export production involved the expansion of the agricultural frontier in new lands which were not subject to the traditional system of land tenure, since the area was largely uninhabited because of the high incidence of tropical diseases (Castillo 1966:15–16). However, further expansion of coffee production during the twentieth century increasingly involved land hitherto destined for other crops. As a result of this, production of basic grains and other agricultural products for internal consumption, such as sugar, cotton and cacao, dwindled (Grieb 1979:154–5, Jones 1939:189). The extent to which banana production in later years developed at the expense of alternative production is less clear. One thing is certain. The UFCO acquired larger amounts of land, including on the Pacific coastal plain, much of which was left uncultivated for extended periods to ward off the Panama disease.15 The expansion of the agricultural frontier dedicated for coffee production was made possible by reintroducing colonial recruitment methods for the use of labour: repartimiento for seasonal employment and debt servitude for permanent employment (McCreery 1983:736). 16 However, recruitment of labour was different in banana production, since, owing to the unhealthy Atlantic coast, no fixed supply of this factor existed in the area. If securing an adequate supply of cheap labour for the coffee economy was difficult, ‘the political elite clearly felt that not even coercion would be sufficient to induce labour to move from the highlands to the coastal region’ (Bulmer-Thomas 1987:12). Thus, initially, the banana enclave economy depended on imported black labour from the Caribbean. Whereas the bulk of labour used in coffee production was seasonal, since the coffee plant requires limited upkeep and no additional investments for long periods of time, in the case of bananas, where labour could be utilized fully during the course of the year, the maintenance of a reserve of underemployed workers

94

GUATEMALAN PRIMARY EXPORTS

became less important. The almost full utilization of labour all year round, together with the benefits accrued from large-scale operations, made the banana industry an efficient enterprise enjoying high rates of productivity. The attainment of economies of scale and the full utilization of factors were possible because banana production could be organized in line rather than in parallel. The social organization of agriculture was definitely different for coffee and bananas than for cacao and indigo. Even though large amounts of land, together with the ability to secure labour, were a prerequisite for initiating exports for all of these staples, coffee and bananas differed from earlier ones with respect to the adaptive role played by capital in order to maintain competitiveness in world trade. Although indigo production modernized the Guatemalan export economy, it only served a limited market, the infrastructure was mostly deficient and only domestic technology was applied. Moreover, production was carried out with small amounts of capital, generated usually by the domestic merchants. The coffee estates, on the other hand, served a wider world market and adopted new techniques in production and transportation, introduced mainly by foreign entrepreneurs and financed by foreign lending institutions (McCreery 1981:56). Thus, indigo production remained of the hacienda type, whereas coffee production resembled that of the plantation.17 But the true plantation economy was established by the UFCO.18 The role of public and private institutions The changes which took place to promote export-led growth would not have been possible if it were not for the initial role played by Liberal public institutions. The social and economic reforms undertaken during the 1870s were crucial for expanding coffee exports. Laws were changed such that unused land owned by the Church could expediently be distributed as individual properties with legal titles. Ecclesiastical taxes and church exemptions were banned in order to direct financial resources to agricultural investments rather than to unproductive activities (Herrick 1974:126–33). A currency reform was also carried out which replaced the outdated system inherited from Spain (Herrick 1974:138–49; Young 1958:114–24). Considering that coffee has a gestation period of three to five years, requiring a considerable initial investment, the availability of credit was important for its production (McCreery 1981:57). The Liberals instituted the first Guatemalan bank, and a system of mortgage loans and commercial transactions was established. The government also improved domestic marketing infrastructure and facilitated imports of know-how (Herrick 1974:149–55). These activities were conducted through the Ministry of Development, an institution which replaced the archaic ‘Consulado de Guatemala’, formerly backed by the Conservatives to defend the economic interests and monopoly power of the merchants (McCreery 1981:18–19). While Liberal public institutions played a crucial role in the formation and

95

LATIN AMERICA

initial modernization of coffee production, the continuing growth of coffee production and exports was due to the efforts of private foreign initiative. The contribution of government institutions for promoting banana export production was practically nil, since foreign entrepreneurship was the driving force behind the adaptive change in technology. In reality, foreign enterprises were the ones responsible for the construction of infrastructure and, even, the operation of the transportation system, because the Guatemalan government defaulted and could not easily obtain foreign credit (Herrick 1974:258).19 Private institutions in the importing country were also crucial for the continuing success of exports, both for the production and marketing of coffee.20 A clear example is the evolution of the banking system. The National Bank, which was formed initially with the primary objective of assisting the farmers who benefited from land redistribution, defaulted because of misuse of funds.21 Even though private banks were founded following bankruptcy of the National Bank, most of the credit for coffee production was provided, at least until the First World War, by European banking houses. These houses also established the link between the producer and the consumer, making coffee marketing a successful enterprise.22 Thus, contrary to the monopoly of Spanish institutions during the colonial period, which hampered the expansion of Guatemalan exports, the role played by public and foreign private institutions during export-led growth was crucial for the insertion of coffee and bananas in world trade. Private domestic institutional arrangements, however, continued to be much the same as during colonialism. Contrary to many foreign private entrepreneurs, who undertook technological innovations, domestic landowners continued to apply outdated techniques.23 However, any productivity increases due to modernization did not produce the desired trickle-down effect with respect to labour, particularly in coffee production. Since variable costs in the production of coffee include almost only labour costs,24 competitiveness continued to depend to a great extent on how effective the system was in keeping wages low. As was the case during the colonial period, producers were able to maintain low labour costs because institutional arrangements permitted them to obtain a higher marginal benefit than the marginal cost incurred when metering and monitoring labour.25 As long as labour continued to have a low opportunity cost and the employer could act as a monopsonist, the bargaining power of seasonal workers continued to be depressed. Such an implicit arrangement could be maintained because of the institutionalized distribution of land, the minifundio-latifundio system. The arrangement worked for seasonal employment because the harvesting of staples did not coincide with those months used by the workers to till their own land. As emphasized above, the Indian population was able to keep much of its land despite the expansion of coffee. Moreover, the cultivation of corn, the most important basic food grown by the Indian population, did not coincide with the coffee harvesting season (McCreery 1983:738–9). Thus, since the minifundio was

96

GUATEMALAN PRIMARY EXPORTS

fragmented and unproductive, this part of the population became increasingly dependent on seasonal employment on the latifundio in order to satisfy its basic needs. The institutional arrangement to keep low wages for permanent workers was dictated directly by the government. The bargaining power of these kinds of workers was reduced even further than during the colonial period, when the Spanish authorities actually enacted laws to prevent the use of forced labour, even though these laws were bypassed many times by indigo growers. After 1870, the government officially institutionalized a kind of debt peonage, 26 but this arrangement was supplanted during the second quarter of the 20th century by stricter forms of labour coercion, such as the ‘Vagrancy Law’. If prior to this law a labourer remained idle by virtue of staying out of debt, now he was compelled to avoid being charged with vagrancy.27 EXPORT DIVERSIFICATION 1955–85 Notwithstanding the successful insertion of coffee and bananas in the world market, the depression years and the Second World War showed what unstable reliance on one export product meant for the acquisition of foreign exchange.28 Moreover, despite favourable post-war conditions in the coffee market, there was a general feeling that decreasing income elasticity of demand and increasing competition from other producers would saturate the market. In the case of bananas, demand was growing relatively less than for other fruits (Federal Trade Commission 1954, ECLAC 1958). For this reason, a process of primary export diversification was initiated in the 1950s.29 The government was able to encourage cotton production with relative ease. Not only had post-depression revival taken advantage of expertise and equipment available since the beginning of the century, but potential domestic growers were also induced to enter the market by a decrease in US post-war production. Attempts to increase domestic sugar production showed signs of immediate success and, by the end of the 1950s, a surplus began to be exported (Banco de Guatemala 1960:34). Likewise, meat exports were initially promoted by substituting imports to satisfy domestic demand. Growing demand and improvements in transportation costs As was the case with export-led growth, primary commodity diversification during the three decades following 1950 occurred largely in response to growing demand in consumer nations, many of which were new non-traditional markets. Even exports of former traditional products, such as coffee, were able to expand due to increased access to less saturated markets—notwithstanding a rapidly falling income elasticity of demand—especially where Milds were in high demand, such as in Northern Europe (Singh et al. 1978:36–7). Rapidly rising post-war real incomes generated an increase in beef

97

LATIN AMERICA

consumption in the United States (Williams 1986:84) Guatemalan beef had many advantages for the growing fast food industry, since grain-fattened beef from the United States was too expensive to be employed in this kind of market. It was also more convenient to import Central American beef than that from South America because it was free of foot and mouth disease (Williams 1986:84–7). Favourable world demand conditions were also vital for inducing export production of Guatemalan cotton. The United States, which was the largest cotton exporter in the world, became uncompetitive when the government reduced subsidies. This caused a considerable reduction in the production of cotton, which maintained favourable prices throughout the end of the 1940s and the first half of the 1950s (Castillo 1966:49). Improvements in the transportation system were crucial for gaining access to the growing markets, especially to improve the competitiveness of new products such as cotton and beef.30 Cotton production requires well-maintained roads to attend expediently to constant pest outbreaks and to deliver the more delicate cotton ball to the gin. The availability of efficient modern roads was also crucial for transporting cattle to the slaughterhouse, and from there to the packing plant and the port of export. The use of refrigerated transportation from the packing plant to the port was also important for successful exports of beef.31 Adaptive change of land, labour and capital Unlike earlier periods, when large amounts of unused land together with the availability of low-cost labour provided an initial thrust for export production, the expansion of cotton, beef and sugar production was carried out on an increasingly restricted agricultural frontier. It has been demonstrated that cotton and beef production on the Pacific coastal plain was undertaken in many instances by disposing of a great part of the forestry resources.32 However, in certain areas of the Pacific coastal plain, cotton and sugar experienced a rapid initial growth by taking advantage of unused land resources, very much like that specified in the ‘vent for surplus’ theory. Cotton grown in areas formerly used by the banana company, where land was productive and, thus, more appropriate for agricultural production, may come as a good example.33 The basic difference between export-led growth and post-war diversification of certain staples, particularly with respect to cotton and beef, lay in how effectively producers were able to implement technological improvements on an increasingly restricted agricultural frontier. Although technological change, primarily in transportation, had been important for the initial success of coffee and banana export production, the period after the Second World War differed in the sense that new techniques were also applied to land and production to improve productivity and product acceptance. The massive construction of roads and infrastructure facilitated considerably the application of new technological discoveries.34 High levels of cotton production were attained not only because of the kind of

98

GUATEMALAN PRIMARY EXPORTS

soil employed—especially in those areas which were inherited from the banana plantations where black earth, the most productive variety, prevailed—but also because of the use of soil management techniques, appropriate seed varieties35 and effective pesticides. Beef export production was successfully established by introducing (a) high-yielding pastures and soil management techniques; (b) imported cattle stocks, such as pure-bred Brahman and Santa Gertrudis, both of which were meatier and highly resistant to pests and tropical heat; (c) artificial insemination, which reduced breeding costs; (d) animal health care to prevent diseases; and (e) modern meat packing plants, which could meet the volume demands and health requirements of the US market.36 While new technological improvements were introduced, competitiveness continued to depend on low labour costs. With a low opportunity cost of labour, manual picking was also employed in those instances where harvesting can be undertaken by machines. This is the case for cotton, where the use of labour for harvesting allowed for better selection of cotton balls and ensured higher yields, since buds ripen at different stages. The employment of cheap unskilled labour was also responsible for price competitiveness in sugar and cardamome, not to mention the newly established beef export industry (Williams 1986:85). The role of public and private institutions The scope of action of government institutions in the production of primary export diversification was enhanced even further after the Second World War. Above and beyond improving ports and roads, the government also began to offer direct assistance in the construction of industrial and packing plants, vital to rendering cotton, meat and sugar exports competitive (Williams 1986:20–4, 92). The creation of the National Institute of Production (INFOP) played a crucial role in promoting the cultivation of cotton (Williams 1986:93–4). Government price policies were important, at least at the outset, to foster production of new agricultural products for exports. Although price controls were employed earlier to prevent widespread loss of income when prices dropped, they were either implemented by a large importer or producer, as was the case with the price floor established by the United States with respect to coffee imports during the Second World War, or implemented briefly by the government to avoid a collapse in production. But they were rarely implemented as part of a conscious strategy to induce production of an export commodity. The government also played an important role in providing credit to newly established agricultural enterprises. During the 1960s and 1970s, credit incentives to promote export diversification were channelled through INFOP and the Agricultural Development Bank (BANDESA). The latter was important when small and medium-sized producers were involved, as was the case with the promotion of cardamome, providing about 90 per cent of total credit during the latter part of the 1970s. It also financed the modernization of existing crops, such as coffee, produced by a large number of small farmers (Banco de Guatemala

99

LATIN AMERICA

1979, 1980). The availability of large quantities of credit was also important for raising cattle.37 However, as witnessed in the case of coffee during the last quarter of the nineteenth century, the government played a major role only at the initial stage of the promotion process. Although INFOP was responsible for most of the indirect and direct financing for cotton during the first years, cotton growers began to rely increasingly on credit supplied by private banks which, by the end of the 1950s, replaced government financial institutions to this end. The price support programme for cotton established by INFOP in the early 1950s was relatively short lived when stocks began to build up. Foreign transportation and marketing techniques were improved by both domestic and foreign private individuals and big business conglomerates. But the difference between the pre- and post-war periods was the increasing ability of domestic producers to take advantage of technological innovations and to control certain stages of the marketing chain. Cotton production and ginning in Guatemala have been undertaken by private domestic growers, who have also controlled the marketing of by-products for domestic consumption, such as cotton seed and cotton cake. Most beef producers have been nationals, who were also responsible for applying the technology required for raising cattle. In the case of Guatemalan coffee and sugar, local producers have gained greater control over exports and have been able to exert some influence in the search for new markets. Transportation and marketing from the port of export have been controlled, until today, by major traders, in the case of cotton, and big corporations, in the case of bananas and beef. Beef exports have resembled those of bananas, the only difference being that most of the production has been undertaken by domestic cattle-raisers. But further processing in modern meat packing plants and domestic and overseas transportation in specially designed containers are activities which have been carried out by the vertically integrated food processing firm (Williams 1986:110). Contrary to the changing role played by private institutions with respect to capital, the competitiveness of exports continued to depend on low labour costs. Although no laws and regulations were applied after the Second World War in order to tie labour explicitly to the farm, since permanent workers were now considered to be free, the producer could keep wages down by providing a plot of land for the worker’s own subsistence.38 By securing two basic conditions over time, namely ‘the possibilities of exploiting family labour in subsistence plots that have no costs for the employer and paying the worker for his effective labour only when it is needed’ (de Janvry and Garramón 1977:209), the producer continued to secure a cheap supply of labour. The minifundio-latifundio system continued to provide cheap labour for seasonal employment.39 Some evidence points to the fact that demographic growth gave rise to further subdivision of subsistence farms in the highlands, where most seasonal workers cultivate for personal consumption, which guaran-

100

GUATEMALAN PRIMARY EXPORTS

teed a larger supply of labour for the new cotton and sugar-cane activities (Fletcher et al. 1970:61–2). The skewed distribution of land also strengthened the bargaining power of large landowners who acted outside government regulations to limit attempts by workers to voice their plight (Adams 1970:364). However, maintaining low labour costs proved to be counterproductive in the long term, since in certain instances it made producers postpone the application of new production techniques. If the competitiveness of certain staples, such as cotton and beef, was gained through the introduction of new technological discoveries, there were other products, such as coffee, where technological improvements were marginally applied.40 It has been suggested that development programmes aimed at improving labour productivity and incomes in the subsistence sector in Guatemala would ‘create pressure for improved efficiency in export production’ (Fletcher et al. 1970:86–7). It is also true that the application of new techniques did not imply efficient use of all land in all cases. First, in the absence of any land tax, big landholders tended to employ only part of their land for production (Fletcher et al. 1970:123). Studies have even demonstrated that, before the 1980s, coffee yields varied inversely with the size of holdings, even though the smaller producers had less access to credit and technology (Griffin 1976:157–61). Second, not all applications of new techniques meant efficient production in the long term. This is particularly the case with cotton, where competitiveness increasingly depended on pest management costs in relation to cotton prices fetched in the market (Pira 1981:19–20). Production was not necessarily privately efficient, even though yields were higher than in any other part of the world.41 CONCLUSIONS The ‘staples’ theory helps us to understand why Guatemalan export production would not have been generated without an initial demand and why transportation costs were crucial in determining whether the staple would have access to the market at a competitive price. This competitiveness depended also on the relative costs of land, labour and capital. Furthermore, the opening up of exports during the colonial and export-led periods, and to a certain extent during the diversification process of the 1950s, can be explained by integrating both the ‘vent for surplus’ and the ‘staples’ theories through the concept of an endogenous land frontier, where large amounts of previously unused land were employed for the production of a staple, without causing a significant opportunity cost for the economy. However, there is no generalized pattern of the role played by labour and capital when expanding the land frontier. A large number of permanent workers were needed to tend to cacao, something which was not necessary in indigo production. Whereas coffee harvesting depended on the availability of large supplies of seasonal workers, banana export production was carried out with imported labour employed on a permanent basis. Cotton and sugar production

101

LATIN AMERICA

required large numbers of workers during the harvesting season, but the extensive cattle farms needed only a negligible amount of mostly permanent workers. The role played by capital also changed. Whereas during the colonial period, its participation was negligible, the opening up of coffee exports could not have been possible without the modernization of transportation and infrastructure. Capital played an even more important role in the initiation of the vertically integrated banana export activity, as well as in the diversification process after the 1950s. The difference in the employment of capital also enables us to define the social organization of agriculture. The closest example to the plantation economy starts with the expansion of coffee, but it is even more closely defined when the highly technically advanced banana company initiates production. This study also shows that the ‘staples’ and ‘vent for surplus’ theories must be complemented by institutional analysis to explain how and why the employment of land, labour and capital for Guatemalan exports has changed adaptively or followed a certain pattern through history. The different stages characterizing the evolution of the Guatemalan export economy, until this century, are not entirely historically independent. This has been the case particularly with labour’s contribution to render coffee exports competitive, where employment methods often followed the same pattern as under colonial rule. Unlike other interpretations dealing with the legacy of Spanish colonial rule regarding the use of labour (Torres-Rivas 1981, Martínez-Peláez 1981), institutional analysis studies the dynamic factors underlying the adaptive change of labour to explain why: (a) contracting labour was different for each export product, because of different requirements and unavailability in specific periods; and (b) institutional arrangements changed because of an increase in the relative bargaining power of labour in specific periods. Government and private institutions were also important in initiating exports and maintaining competitiveness of staples. But while government support, which included the availability of credit and technical know-how, coupled with the provision of the essential infrastructure for marketing, was important at the outset, competitiveness was secured by private entrepreneurship and skills rather than state ‘dirigiste’ measures. The role of private institutions also changed through history. The competitiveness of coffee and banana exports relied on imported, rather than domestic, capital and skills. This situation gradually changed after the Second World War, when export competitiveness increasingly relied on the ability of domestic entrepreneurs to adopt new techniques. Although this study does not pretend to make a comparative analysis of the historical evolution of exports between countries, consideration of the role of institutions in the initiation of exports and the determination of competitiveness in Guatemala enables us to explain why path dependence produced divergent outcomes between, for example, Latin America and North America, even though conditions were the same for both areas when exports opened up. This divergence happened ‘right from the beginning, reflecting the imposition of the institutional patterns from the mother country upon the colonies and the radically divergent

102

GUATEMALAN PRIMARY EXPORTS

ideological constructs that shape the perceptions of the actors’ (North 1990:102). Not only were institutional arrangements different with respect to the use of land, labour and capital, but the initial market opportunities reflected from the outset the institutions which governed the economic and political life in the mother country. Whereas British institutions were crucial for the successful development of North American exports, in the case of Latin America the legacy of Spanish institutions was a major cause for relative backwardness. These institutions lagged behind in the modernization process since they merely reflected an economy which was increasingly falling behind the industrialization race in Europe. While Britain demanded raw materials for its growing industrial sector, Spain’s main concern was to amass taxes and precious metals from Latin America to meet financial commitments. Small Latin American countries lacking greater amounts of precious metals, such as Guatemala, ended up depending more on nearby markets than on demand generated from the mother country.42 Spanish institutional arrangements also contributed historically to the fact that producers, even until today, have been able to secure an adequate supply of cheap labour. However, the ‘staples’ theory has taught us that these kinds of arrangements could condition the continuing growth of staple exports in the long term. Drawing on the Canadian experience, this theory proves that the ‘growth inducing factor was also dependent in a large measure on their respective production functions and the resulting distribution of income’ (Bertram 1963:163). Unlike in Canada, where ensuing income distribution has been responsible for the modernization and continuing growth of the export sector, in Guatemala the perpetuation of cheap labour to maintain competitiveness has become, in certain instances, a deterrent for the continued modernization of the agro-exporting sector. As long as wages remained low and the producer was able to obtain an adequate profit margin, there was little incentive in the long term to secure the competitiveness of certain staples by continuing to apply new techniques. ACKNOWLEDGEMENTS A first draft of this essay was presented at the 5th Annual Conference of ASERCCA, Aix en Provence, France, October 1989. NOTES 1 2

3

For a review of the staples theory, see Lundahl (1991). Although certain countries can have an advantage by owning certain key resources, a simple availability theory is rather an exception to the rule, since it does not explain the importance of access to markets and adaptive changes in the factors involved in production. For a review of the availability theory, see Findlay (1970:76–8). This is why, for example, ‘Sugar cane, a bulky product requiring elaborate processing and handling by the standards of the period, might, if grown in Cuba for instance, still show a profit in Seville after a voyage of 70 days, whereas if grown in the Mexican

103

LATIN AMERICA

4

5 6

7 8

9 10 11 12 13 14

15 16 17 18

19

lowlands around Veracruz, after a voyage of 130 days from Seville it was far less likely to be profitable’ (MacLeod 1984a:367). As elsewhere in Latin America, the encomienda system ‘was developed under the cloak of doing good to the Indians. Those entrusted with allotments to natives (encomendero) assumed the discharge of certain duties toward them the chief of which was to bring them into the Catholic faith. In return the encomenderos received the right to demand labor or certain payments in kind from those delivered into their care’ (Jones 1939:115). ‘If the encomienda was, in part, a way for an emerging elite to control and share a major resource, labour, then repartimiento drafts were a way to ration an increasingly scarce supply’ (MacLeod 1984b:225). Which was ‘used to recruit and discipline a permanent labour force, often in mines or on haciendas, but advances against future work were also used to recruit village labour for seasonal work, or to sustain an incipient “free” rural proletariat during the times of the year when it was not needed on the plantation’, (MacLeod 1984b:230). For a detailed account about the financial position of the crown, see Lundahl (1993). As a matter of fact, Mexico, due to its importance as a major trading colony, was able to gain a more competitive position than Central America with respect to exports. For example, cochineal did not emerge as a major colonial export in Guatemala because Mexican comparative advantage in the production of this dyestuff inhibited its expansion in Central America (MacLeod 1980:170). Even in the extreme case of the master-slave relationship, ‘the owner must devote resources to monitoring and metering a slave’s output and critically applying rewards and punishments based on performance’ (North 1990:32). The share of Guatemalan exports of Milds relative to total world exports increased continuously from 7.0 per cent in 1885–9 to 14.0 per cent in 1905–9 (Mosk 1958:170). Especially the US market, which during the first quarter of the twentieth century, absorbed three-quarters of total world banana imports and had a per capita consumption which surpassed that of citric fruits (Wolf 1959:13). Increased shipping competition made coffee transportation charges to Europe fall relatively more than those to North America, even though the ship had to reach Europe via the Straits of Magellan (McCreery 1976:445–8). For the characteristics of the banana fruit which influence and condition its transport requirements, see Clairemonte (1975:141). Although the opening of the Panama railroad in 1856 provided a link whereby Guatemala was able to export through its Pacific ports, ‘the inadequately maintained roads delayed shipments, immobilized capital, and raised export expenses, hampering Guatemala’s ability to compete internationally’ (McCreery 1976:443). During the second quarter of the current century only one-third to one-quarter of the land owned by this company was cultivated (May and Plaza 1959:90–4). The practice of repartimiento had virtually disappeared during the first half of the nineteenth century, since production of cochineal was carried out by self-sustained agricultural units requiring small quantities of capital labour. This classification follows Wolf and Mintz’s attempt to classify the social organization of agriculture between hacienda and plantation (Wolf and Mintz 1957:380–412). Where the production unit had to be organized ‘so as to be capable of absorbing large amounts of capital, as well of repaying capital used with large profits…use of land, labour and technology on the plantation thus must be geared to the production in bulk with maximum efficiency and utilization’ (Wolf and Mintz 1957:396). For example, the transportation network required to market coffee produced in the

104

GUATEMALAN PRIMARY EXPORTS

20 21 22

23 24

25 26

27 28 29 30

31

32

northern part of Guatemala, where the highest yields were obtained, was almost entirely developed by German initiative, when it became clear that the government was not able to promote its construction (McCreery 1981:64–6). Banana exports grew in importance when the UFCO-owned railway company continued the construction of the railroad line to the Atlantic coast (May and Plaza 1959:4–14). Although the state-owned National Coffee Institute promoted quality coffee for exports, it did so only during the 1930s and its role remained very limited (Grieb 1979:60). Such as financing war against El Salvador (McCreery 1981:58). The European banking houses, most of which were German, made an arrangement with the grower, whereby he would sign a contract for the future delivery of his produce at an established price. As late as the mid 1930s, they were responsible for almost 65 per cent of total coffee export value, while the corresponding share for Guatemalan exporters was only 5 per cent (Jones 1939:208, Munro 1967:267–8). It has been demonstrated that foreign coffee producers enjoyed much higher levels of productivity than their national counterparts (Jones 1939). Until today, the harvesting of coffee has been done manually for two reasons: (a) it is a perennial crop and cannot be harvested with threshing machines; and (b) the quality of the coffee bean depends largely on when it ripens, and this can vary on a single plant. A study has shown that there was a tendency for real wages to decline during the period 1875–1935 with only small recoveries experienced around the 1920s (McCreery 1983:748–50). Such as the ‘Reglamento de Jornaleros’ of 1877 and the ‘Ley de Trabajadores’ of 1894. Not only were money advances made in order to coerce labour to provide services, but employers manipulated ‘the accounts so that workers might be kept almost perpetually in debt, and thus required by law to work continuously on the plantation’ (Whetten 1961:120). According to the ‘Vagrancy Law’ an ‘idle’ worker had to provide 100 to 150 days of free and obligatory labour, reducing the unit cost of production in the coffee plantations (Jones 1939:162–4; McCreery 1983:757–8). In reality it was coffee which accounted for most of the export-led growth in the economy (Bulmer-Thomas 1987). This analysis does not include industrial production, even though Guatemala was also able to expand its gamut of industrial exports to Central America. Cotton production in Guatemala had lagged behind El Salvador and Nicaragua because the road connecting the Pacific coastal plain was not modernized until the 1960s. Roads developed at a later date, such as the one in northern Guatemala in the late 1970s, promoted cattle raising (Williams 1986:23, 93). One of the most important innovations in beef transportation during the 1960s was the ‘fishy-back’ land-and-sea trucking service of Coordinated Caribbean Transport, Inc., which as early as 1964 was transporting a weekly average of four 20-ton trailer-loads of fresh beef from Guatemalan packing plants for export to the United States (Parson 1965:157–8). One study shows that in Escuintla, where most beef production has taken place, the share devoted to pasture in 1964 was as much as 42 per cent of total available land, whereas only 16 per cent of this area remained in forest. The comparative shares in the mid 1950s were 12 and 48 per cent respectively. Another study shows that between 1950 and 1963 the eight most important cotton-producing municipalities in Guatemala lost 380,000 acres of forest, of which 220,000 acres were transformed into pasture land and the rest, the better land, was used for cotton (Williams 1986:53,113).

105

LATIN AMERICA

33

34 35 36

37

38 39

40

41

42

It has to be emphasized that, contrary to the operations carried out on the Atlantic coast, banana production on the Pacific coastal plain, which had been initiated in the 1930s as a response to the threatening Panama disease, used land extensively. Cotton made use of this land more efficiently. These improvements have made the Pacific coastal plain, where most of the cotton, sugar and cattle grows, the most productive agricultural area in Guatemala (Parson 1965:158). Which have played an important role in the determination of fibre strength, productivity and resistance to weather and pests (Parson 1965:153). The techniques were not introduced until the late 1950s, despite being known twenty years earlier, because of obsolete transportation and laws which prevented cattle from being sold in the municipal slaughterhouse with a large enough profit to induce reinvestments in herd or pasture improvements. It was not until credit was available and road improvements, modern packing plant and refrigerated transportation facilities were introduced that cattle owners began to employ modern farm management techniques (Williams 1986:95). Credit for cattle was provided by World Bank, AID and IDB loans, which were channelled through projects executed by government institutions. These loans were granted during the 1960s and 1970s in addition to loans provided by these same institutions to undertake road building for the efficient transportation of beef to the port of export (Williams 1986:93–5). This arrangement is possible as long as ‘the opportunity cost of the total land given in usufruct to the worker is less than the value of production the worker can generate on it through use of family labor’ (de Janvry and Garramón 1977). A census carried out in 1950 showed that while the minifundio, too small an area to satisfy the basic needs of a family, represented over 88 per cent of the amount of farms, but incorporated only around 14 per cent of the total area of farms, the lati-fundia, farms of over 900 hectares, made up for only 0.1 per cent of the amount of farms, but employed almost 41 per cent of total farm area. This has been demonstrated with productivity rates in Guatemala which, during the 1980s, were less than half of what they were in a country like Costa Rica, where new techniques in production have been employed owing to relatively higher labour wages (SIECA 1987). Guatemalan cotton production is a perfect example of how technology improved productivity, but where marginal gain from the use of this technology even became negative since no profit-maximizing economic threshold was defined by the growers with respect to the use of pesticides. Moreover, the indiscriminate use of pesticides represented a considerable social cost to the society as a whole (Pira 1981:128–33). This position contrasts with the relatively static historical interpretation of many dependency writers, since complete appropriation of a surplus only took place during the early phases of colonization, when the Spaniards’ sole aim was to plunder without settling in the newly discovered regions.

BIBLIOGRAPHY Acuña, V. (1980) ‘La reglamentación del comercio exterior en América Central durante el siglo XVIII’, Mesoamérica vol.1. Adams, R. (1970) Crucifixion by Power: Essays on Guatemalan National Structure, 1944– 1966, Austin: University of Texas Press. Banco de Guatemala (1960) Memoria Anual. ——(1979) Informe Económico, April-September. ——(1980) Informe Económico April-June.

106

GUATEMALAN PRIMARY EXPORTS

Bertram, G. (1963) ‘Economic growth in Canadian industry, 1870–1915: The staple model and the take-off hypothesis’, Canadian Journal of Economics and Political Science vol. 29. Bulmer-Thomas, V. (1987) The Political Economy of Central America since 1920, Cambridge: Cambridge University Press. Carmack, R. (1979) La Historia Social de los Quichés, Guatemala: Editorial José Pineda Ibarra. Castillo, C. (1966) Growth and Integration in Central America, New York: Praeger. Caves, R. and Nolton, R. (1959) The Canadian Economy: Prospect and Retrospect, Cambridge, MA: Harvard University Press. Clairemonte, F. (1975) ‘Bananas’, in Cheryl Payer (ed.) Commodity Trade of the Third World, London: Macmillan. Coase, R. (1960) ‘The problem of social cost’, The Journal of Law and Economics October. Cohen, I. (1980) Notas sobre el Transfondo Histórico del Desarrollo Centroamericano, mimeo. (CEPAL/MEX/ODE/35). Domar, E. (1970) ‘The causes of slavery or serfdom: A hypothesis’, Journal of Economic History vol. 30. ECLAC (1958) ‘Evolution y perspectivas del mercado internacional del banano’, Boletín Económico de América Latina vol. 3. Facio, R. (1965) La Federación de Centroamérica: Sus Antecedentes, su Vida y su Disolución, San José: ESAPAC. Federal Trade Commission (1954) Economic Report of the Investigation of Coffee Prices, Washington, DC: United States Printing Office. Findlay, R. (1970) Trade and Specialization, Harmondsworth: Penguin. Findlay, R. and Lundahl, M. (forthcoming) ‘Natural resources, “vent for surplus” and the staples theory’, in Gerald Meier (ed.) From Classical Economics to Development Economics: Essays in Honor of Hla Myint, Oxford: Oxford University Press. Fletcher, L. et al. (1970) Guatemala’s Economic Development: The Role of Agriculture, Ames, A: The Iowa State University Press. Floyd, T. (1961) ‘The Guatemalan merchants, the government, and the Provincianos, 1750– 1800’, Hispanic American Historical Review vol. 41. Frey, B. (1984) International Political Economics, New York: Basil Blackwell. Glade, W. (1986) ‘Latin America and the international economy’, The Cambridge History of Latin America, Volume 4, Cambridge. Grieb, K.J. (1979) Guatemalan Caudillo: The regime of Jorge Ubico, Athens, OH: Ohio University Press. Griffin, K. (1976) Land Concentration and Rural Poverty, London: Macmillan. Hamilton, E. (1948) ‘The role of monopoly in the overseas expansion and colonial trade of Europe before 1800’, American Economic Review vol. 38. Herrick, T. (1974) Desarrollo Económico y Político de Guatemala: 1871–1885, Guatemala: Editorial Universitaria de Guatemala, de Janvry, A. and Garramón, C. (1977) ‘The dynamics of rural poverty in Latin America’, The Journal of Peasant Studies vol. 4. Jones, C.L. (1939) Guatemala: Past and Present, Minneapolis, MN: The University of Minnesota Press. Lundahl, M. (1991) ‘Staples trade and economic development’, EFI Yearbook, Stockholm. ——(1993) ‘Spain and the conquest of America: Profits, religion and forced labour in the fifteenth and sixteenth centuries’, this volume. McCreery, D. (1976) ‘Coffee and class: The structure of development in liberal Guatemala’, Hispanic American Historical Review vol. 56. ——(1981) Desarrollo Económico y Política National: El Ministerio de Fomento de

107

LATIN AMERICA

Guatemala, 1871–1885, Guatemala: Centre de Investigaciones Regionales de Centroamérica. ——(1983) ‘Debt servitude in rural Guatemala’, Hispanic American Historical Review vol. 63. MacLeod, M. (1980) Historia Socio-Económica de la América Central Española 1520– 1720, Guatemala: Piedra Santa. ——(1984a) ‘Spain and America: the Atlantic trade 1492–1720’, The Cambridge History of Latin America, Volume 1, Cambridge. ——(1984b) ‘Aspects of the internal economy of colonial Spanish America: Labour; taxation; distribution and exchange’, The Cambridge History of Latin America, Volume 2, Cambridge. Martínez-Peláez, S. (1981) La Patria del Criollo, San José: EDUCA. May, S. and Plaza, G. (1959) La United Fruit en América Latina, Mexico: National Planning Association. Mosk, S. (1958) ‘Economía cafetalera de Guatemala durante el período 1850–1918: Su desarrollo y signos de instabilidad’, Seminario de Integración Social (ed.) Economía de Guatemala, Guatemala: Editorial del Ministerio de Educación. Munro, D. (1967) The Five Republics of Central America: Their Political and Economic Development and their Relations with the United States, New York: Russell and Russell. Myint, H. (1958) ‘The classical theory of international trade and the underdeveloped countries’, The Economic Journal vol. 68. Naylor, R. (1960) ‘The British role in Central America prior to the Clayton-Bulwer Treaty of 1850’, Hispanic American Historical Review vol. 40. North, D. (1990) Institutions, Institutional Change and Economic Performance, Cambridge: Cambridge University Press. Parson, J. (1965) ‘Cotton and cattle in the Pacific lowlands of Central America’, Journal of Inter-American Studies vol. 7. Pira, L. (1981) ‘Towards integrated control of cotton pests in Guatemala: An economic analysis’, Unpublished MA Thesis, McGill University, Montreal, Canada. SIECA (1987) Centroamérica: Evolución y Situación Actual de los Productos Básicos de Exportación, DAG/SIECA, Guatemala. Singh, S. et.al. (1978) Café, Te y Cacao, Madrid: Editorial Tecnos. Smith, R. (1946) ‘Origins of the Consulado of Guatemala’, Hispanic American Historical Review vol. 26. ——(1959) ‘Indigo production and trade in colonial Guatemala’, Hispanic American Historical Review vol. 39. Torres-Rivas, E. (1981) Interpretación del Desarrollo Social Centroamericano, San José: EDUCA. Whetten, N. (1961) Guatemala: The Land and the People, New Haven, CT: Yale University Press. Williams, R. (1986) Export Agriculture and the Crisis in Central America, Chapel Hill, NC: The University of North Carolina Press. Wolf, J. (1959) ‘Evolución y Estructura del Mercado Bananero Mundial’, Boletín Mensual de Economía y Estadística Agrícolas vol. 8. Wolf, E.R. and Mintz, S.W. (1957) ‘Haciendas and plantations in Middle America and the Antilles’, Social and Economic Studies vol. 6. Woodward, R.L. (1981) Privilegio de Clase y Desarrollo Económico, San José: EDUCA. Young, J.P. (1958) ‘Monedas y finanzas Centroamericanas’, Seminario de Integración Social (ed.) Economía de Guatemala, Guatemala: Editorial del Ministerio de Educación.

108

7 FOREIGN FIRMS AND STRUCTURAL ADJUSTMENT IN LATIN AMERICA Lessons from the debt crisis Magnus Blomström and Robert E.Lipsey

INTRODUCTION In the backwash of the debt financing of the 1970s and the worldwide recession of the early 1980s, many developing countries, particularly in Latin America, ran into serious debt-servicing difficulties. To cope with the debt crisis, radical policy changes were introduced. These ‘structural adjustment programmes’ generally included deregulation and privatization of the economies, and opening them to external markets and competition.1 The shift from inward to outward orientation has involved shifting production from domestic to export markets. In this study, we examine the extent to which a number of heavily indebted Latin American countries have redirected their sales of manufactured goods to world markets and the role of multinational corporations (MNCs) in this shift. We are particularly interested in investigating whether affiliates of multinational firms are better equipped to redirect their sales than local firms in developing countries.2 It is not difficult to think of reasons why MNCs should have an easier time switching markets than other firms. MNCs have internal markets already set up to supplement external markets. This could give them opportunities to switch their sales of finished products or components from local to overseas markets. MNCs also presumably have greater access than local firms to market information, distribution channels and international marketing skills, all of which facilitate access to export markets. There are a number of possible measures of ‘export performance’ by countries and groups of firms that one could use. One is the rate of growth of exports. Another might be the growth of exports to developed country markets, the more competitive part of world markets and those in which the exporters would be sure to earn convertible currency. Still another would take account of the industries in which exports were originally concentrated and measure the degree to which the countries’ or firms’ export growth exceeded or fell short of what it would have been if it had simply kept up with the world growth of exports in each industry.

109

LATIN AMERICA

A somewhat different way of judging export performance would be to examine the proportion of output exported, or the propensity to export, assuming that one of the tasks of adjustment is that of shifting sales from domestic to export markets. An ambiguity in changes in the propensity to export is that a rise could be achieved by increasing exports from increasing production, by moving sales from local to export markets without changing production, or simply by reducing domestic sales or even reducing both export and domestic sales, but the latter more rapidly. Clearly, the last would be the least desirable form of adjustment from the point of view of the host country. The study is organized as follows. The next section examines the export growth of five heavily indebted Latin American countries for which data on US multinationals’ trade is available, and compares them with a less heavily indebted group of Asian countries for which we also have MNC data. We take 1977–82 to represent the period before the debt crisis and periods after 1982 to represent the debt crisis era. The third section investigates the same measures of export performance for US affiliates in these countries during the same periods. Changes in export propensities of US affiliates and their host economies are discussed in the fourth and fifth sections, respectively. The following section looks at US affiliates’ export performance by industry group, and, finally, the last section concludes the study. THE EXPORT PERFORMANCE OF COUNTRIES The manufactured exports of all developing market economies as a group were increasing rapidly in nominal terms, expressed in US dollars, in the five years before the debt crisis (Table 7.1). The countries that were not later labelled as heavily indebted raised their manufactured exports (in $US) by over 16 per cent per annum between 1977 and 1982, while the heavily indebted countries increased their exports by about 15 per cent per annum. After 1982, the rate of increase was only 2½ per cent per year for the more indebted group through 1986, but then rose rapidly to almost 12 per cent for the period through 1988. For the less indebted countries, the rate of increase from 1982 through 1988 reached almost 18 per cent. A somewhat different picture of the two periods is given by the second set of three columns in the table, which shows the same export growth in terms of the purchasing power over manufactured products exported by developed countries to developing countries. In these terms, the less indebted countries are seen to have enjoyed not only substantial real export growth in manufactures between 1977 and 1982, but even faster growth in 1982–8 than before: a rise in the growth rate by over 15 per cent. The heavily indebted countries are shown to have had slower real manufactures export growth in the 1977–82 period than the less indebted countries, and declines in real manufactured exports from 1982 to 1986. In the three restructuring countries, Chile, Colombia and Mexico, real export growth from 1982 to 1988, especially after 1986, outran that of the less indebted countries.

110

STRUCTURAL ADJUSTMENT IN LATIN AMERICA Table 7. 1 Average annual growth rates of manufactured exports from developing countries (%), various periods, 1977 to 1988

a

Hong Kong, India, Malaysia, Singapore, South Korea, Taiwan and Thailand. Source: Table A7.1 Lipsey et al. (1991)

THE EXPORT PERFORMANCE OF MANUFACTURING AFFILIATES OF US FIRMS The export performance of US affiliates in the same countries and in two geographical areas during the same periods is described in Table 7.2. In the heavily indebted Latin American countries, these affiliates expanded their exports more rapidly than their host countries in most periods, but the data are ambiguous for 1982–8. In real terms, their export growth was somewhat higher in 1982–8 than before the debt crisis. This acceleration was accounted for largely by affiliates in Mexico but it probably extended to Chile as well; in Brazil there was a deceleration of export growth by US affiliates after 1982. And that slowdown of growth was also evident for affiliates in Latin America as a whole. In the less indebted group of Asian developing countries, the rate of export growth in real terms by US affiliates was lower after 1982 than before. Although the individual country data (not shown here) are somewhat skimpy, the comparisons with aggregate export growth in these countries suggest that the US multinationals’ affiliates kept up with their countries’ export growth before 1982, but lagged after that, and that was also the case for US affiliates in the developing

111

LATIN AMERICA Table 7.2 Average annual growth rates of exports by US majority-owned manufacturing affiliates in developing countries, various years, 1977–88

a

Hong Kong, India, Malaysia, Singapore, South Korea, Taiwan and Thailand. Source: Table A7.2

Asian countries as a group. In other words, although US affiliates in Asia did well in terms of exports during the 1980s, locally owned Asian firms and other multinationals in the region did better. An important difference between the two groups of developing countries is that while US firms are the major multinationals in Latin America, and therefore among the heavily indebted countries, that is not the case among the Asian countries listed. In these, Japanese firms play more of a role than US firms. Thus, the US affiliates are a good approximation of multinationals in general in Latin America, while in the Asian countries, a large part of the exports by multinationals were by Japanese firms and are not included in our data.3 THE EXPORT PROPENSITIES OF AFFILIATES OF US MULTINATIONALS There has been a long-term trend towards greater export orientation among the foreign affiliates of US firms. It has been present in affiliates in both developed and developing countries, but the export propensity of affiliates in developing countries was, in 1977, less than 60 per cent of that in developed countries, and in

112

STRUCTURAL ADJUSTMENT IN LATIN AMERICA Ta b l e 7 . 3 E x p o r t p r o p e n s i t i e s ( e x p o r t s a s p e r c e n t a g e o f t o t a l s a l e s ) o f U S majority-owned manufacturing affiliates, various years, 1977–88

a

1985. Hong Kong, India, Malaysia, Singapore, South Korea, Taiwan and Thailand. Sources: Tables A7.2 and A7.3

b

1982, still only about 60 per cent as large (Table 7.3). After 1982, however, the export propensities of LDC affiliates leaped, rising by almost 50 per cent to over 80 per cent of that of affiliates in developed countries in 1986 and almost 90 per cent in 1987 and 1988. That sudden, dramatic shift was centred in the heavily indebted countries, a fact that suggests the possibility that the debt crisis may have had some role in the transformation. US affiliates in the Asian countries, especially those not heavily indebted, have had high export propensities throughout the period since 1977. There is only a little difference in export behaviour between the first and later years, on average, although there are some substantial shifts in individual countries. To the extent that there was any change, it was a reduction in export propensities. However, all, except those in India, had export propensities above 50 per cent in every year. Among affiliates in the heavily indebted countries, the course of events was very different. They were far less export oriented than those in the less indebted group in 1977. After 1982, however, those in almost every heavily indebted country moved strongly towards exporting. By 1988, the average export propensity in the heavily indebted countries was more than twice that of 1977. An indication that indebtedness, rather than geographical location, was a determinant of this shift is the fact that affiliates in the one heavily indebted Asian country that we have data for (not shown here), The Philippines, shifted towards exporting

113

a Hong Kong, India, Malaysia, Singapore, South Korea, Taiwan and Thailand. Source: As Table 7.3

Ta b l e 7 . 4 C h a n g e i n t o t a l s a l e s , l o c a l s a l e s a n d ex p o r t s , va r i o u s p e r i o d s , 1 9 8 2 – 8 , i n U S m a j o r i t y - ow n e d m a n u fa c t u r i n g a ffi l i a t e s i n d eve l o p i n g c o u n t r i e s

STRUCTURAL ADJUSTMENT IN LATIN AMERICA

almost as much as the Latin American affiliates did (see Blomström and Lipsey 1990). How was this shift in the orientation of sales carried out? In the long run, such a shift could conceivably be accomplished by increases in production largely or entirely dedicated to export markets, without any reduction in local sales. In the short run, a large expansion in exports would presumably have to be at the expense of local buyers unless the producers had excess capacity. Between 1982 and 1986, the increasing export propensities of US manufacturing affiliates in developing countries as a group, and especially in Latin America and in heavily indebted countries, were more the result of decreasing sales, and particularly decreasing local sales, than of increasing exports (see Table 7.4). While the exports of US affiliates in Latin America increased by $1.8 billion (or 38 per cent) between 1982 and 1986, their local sales decreased by almost $9 billion, or 25 per cent over the same period. Among the heavily indebted countries, the decrease in affiliates’ local sales ($8.3–8.6 billion) was almost four times the increase in exports ($2.2–2.5 billion). In developing Asia, in the same years, both exports and domestic sales grew, but export growth was much larger; the export growth came from increased production rather than from declining domestic consumption. However, by 1988, the picture had changed considerably. Total sales of affiliates were higher than in 1982 in Latin America as a whole and in heavily indebted countries as a group, and in all but one of the individual countries. There were still decreases in local sales in both groups of countries and for affiliates in two of the five heavily indebted countries, but these were far outweighed by increases in exports. Thus, by 1988, most of the growth of exports by US affiliates was coming from increased production rather than from decreased local sales. The reversal after 1986 can be seen more sharply in the data for 1986–8, when the growth of affiliate local sales in Latin America, in the heavily indebted countries as a group, and in the three restructuring countries, Chile, Colombia and Mexico, outpaced export growth by large margins. In contrast, affiliates in the less indebted Asian countries continued to increase exports more than local sales. EXPORT PROPENSITIES OF COUNTRIES We would like to compare affiliate behaviour with that of domestic firms across regional and indebtedness groupings. However, we do not have regional production data and are limited to some comparisons between affiliates and their host countries for individual countries. The production data we do have (gross output at producers’ prices) are, for at least some of the countries, incomplete, and incomplete to a changing extent over time. The country export propensities derived using these output figures as denominators are almost certainly overstated, both because the output estimates

115

LATIN AMERICA Table 7.5 Export propensities in manufacturing (exports as percentage of value of output) in eleven developing countries, various years, 1977–88

a

Maquiladora exports to USA estimated from US data. Corresponding 1982 ratio is 7.7%. Chile, Colombia, Mexico and Venezuela. The 1982 ratio corresponding to 1977 is 10.1%. c Hong Kong, Malaysia, Singapore, South Korea and Taiwan. d Hong Kong, Malaysia, Singapore and Taiwan. e The propensities over 100% in the shortcut calculations reflect the fact that the published data are on a general trade basis, including re-exports. The data from the UN tapes are domestic on the special trade basis, including only domestic exports. The high ratios for Singapore are more of a puzzle, but probably reflect the large entrepôt trade also. Sources: Tables A7.1 and A7.5 b

are incomplete in coverage, often omitting small plants, and because most are in producers’ prices, or even factor cost, while exports add costs of transportation to the border and other costs incurred by intermediaries such as wholesalers. The national export propensities for manufactures estimated from these data (Table 7.5) show clearly the much higher export orientation of the less indebted Asian countries’ manufacturing industries, as compared with those of Latin America. However, the average export propensity of manufacturing in heavily indebted countries, starting from a much lower level than those of the less indebted group, more than doubled between 1982 and 1986. The greatest increase, by far, was for Mexico.4 It resulted not only from the almost 100 per cent rise in export values in US dollars but even more from the incredibly sharp fall in

116

STRUCTURAL ADJUSTMENT IN LATIN AMERICA Table 7.6 Sales outside the host country by US majority-owned manufacturing affiliates as percentage of value of manufactured exports, twelve developing countries, various years, 1977–88

a

Corresponding 1982 ratio is 16.8. Corresponding 1982 ratio is 12.1–12.5. Corresponding 1982 ratio is 12.2–12.8. Sources: Tables A7.1 and A7.2

b c

the dollar value of production, which must have been far greater than any decline in physical output. Even aside from Mexico, all the export propensities rose between 1982 and 1986, except that for Venezuela. Thus, while export propensities in heavily indebted countries were still relatively low in 1986, some force had apparently encouraged a shift from domestic sales to exporting. Although these country export propensities for Latin America appear almost as high as, or even higher than, those for US affiliates (Table 7.3) more often than not, that comparison is inconclusive because the denominators in the country ratios (Table 7.5) are seriously understated. However, it does appear from the

117

LATIN AMERICA

comparisons of changes that the US affiliates raised their export propensities more between 1982 and 1986 than the countries did, and especially more than non-affiliates did. In Asia, on the other hand, the country export propensities, except for the two entrepôt countries, Hong Kong and Singapore, are almost all lower than those of US affiliates, despite the overstatement of the country ratios. Since the country ratios in Asia were rising and the affiliate export propensities were falling, the two sets of ratios were coming closer over time. Another way of examining the role of the US affiliates is through their shares in exports of manufactures. Within Latin America, and within the heavily indebted countries as a group, they increased their share substantially from 1977 to 1986, but lost ground in the next two years (Table 7.6). In the Asian countries and the less indebted countries as a group, in contrast, the shares of US affiliates in exports were declining most of the time. The decline was slight outside of Hong Kong and Taiwan, and may have been offset by increasing Japanese shares, but there was a substantial decrease in those two host countries. Only in Malaysia, among this group of countries, were US firms’ shares of country exports increasing. It seems, then, that the strong export performance of US affiliates relative to local firms after 1982 was not a worldwide phenomenon, but was concentrated in the heavily indebted countries. And that export performance of US affiliates consisted more in the degree to which they shifted the direction of their sales towards exports than in the rate of growth of their exports. EXPORT PERFORMANCE BY INDUSTRY GROUP Since the US affiliates in Latin America are active in a very different set of industries from those in South-East Asia, their export performance could to some extent reflect the worldwide fortunes of their industries rather than the Table 7.7 Industry distribution of sales outside the host country by US manufacturing affiliates in Latin America and developing Asia, 1982 (%)

Sources: Blomström and Lipsey (1990), Tables A7.2 and A7.6

118

STRUCTURAL ADJUSTMENT IN LATIN AMERICA

debt circumstances or macroeconomic policies of their host countries. (It should be added, however, that the industry distribution of US firms’ investment and production may itself be a reflection of the openness or importsubstitution orientation of the host countries.) The 1982 distributions of US affiliates’ exports from the two regions are shown in Table 7.7. For the affiliates in Asian countries, mostly less indebted, three-quarters of exports were in electrical machinery, and the largest part of the rest was in non-electrical machinery, both relatively fast-growing sectors in world trade. Exports by affiliates in Latin American countries were much more evenly distributed across the seven industry groups. The growth rates of US affiliate exports in the two areas, by industry, shown in Table 7.8, point up the fact that foods and metals, which accounted for 23 per cent of exports by Latin American affiliates but for less than 4 per cent of Asian affiliates’ exports in 1982, were slow-growing industries for US affiliates in developing countries worldwide before 1982. Machinery and transport equipment, fast-growing industries, were 40 per cent of Latin American affiliates’ exports and 88–89 per cent of exports by Asian affiliates. The sharpest contrast between the two areas was in chemicals, in which export growth accelerated for Asian affiliates, while in Latin America the industry switched from being the fastest growing in exports in 1977–82 to a declining exporter after 1982. It would appear that the Latin American affiliates would have done considerably better in exporting, on average, if they had been in the same industries as the Asian affiliates. Within industries, their exports did not seem to grow particularly slowly in 1977–82 relative to Asian or all developing country affiliates; they did better than developing country affiliates in general in three out of seven industries. Even in 1982–6, they had faster export growth in two industries. While the growth rates of affiliate exports were not distinctively different between Latin America and Asia or developing countries in general, there was one respect in which there were clear regional differences. Export propensities of affiliates in Latin America were lower than those in Asia or developing countries in general in almost every industry in every period (Table 7.9). To some extent, that difference may have reflected the fact that Latin American markets were larger than Asian markets, but it may also reflect a difference in policies towards trade: a more inward orientation in Latin America. However, there was a distinct shift towards export orientation by Latin American affiliates in all industries from 1977 to 1988. Some of the shift began in 1977–82, but it accelerated after 1982 in all industries except chemicals. Affiliates in Asia were also becoming increasingly export oriented after 1982 in most of the industries for which we can make the calculation, but in some cases they had little room to increase the share of production they sold abroad because it was already so high. The more inward initial orientation of Latin American affiliates may have given them more scope than Asian affiliates had for switching sales from domestic to foreign markets after 1982. However, data for Asian affiliates are too heavily suppressed to permit many comparisons. The largest absolute

119

LATIN AMERICA Table 7.8 Average annual growth rates (%) of exports by US majority-owned manufacturing affiliates in developing countries, by region and industry, various years, 1977–88

Source: Table A7.6

increase in exports by Latin American affiliates, in the transport equipment industry, was accompanied by a similar, but even larger, decline in domestic sales between 1982 and 1986, and therefore by a drop in total sales. But by 1988, total sales were well above the 1982 level and only half the export increase was a switch out of local sales (Table 7.10). The next largest export

120

STRUCTURAL ADJUSTMENT IN LATIN AMERICA Table 7.9 Export propensities (exports as percentage of sales) of US majority-owned manufacturing affiliates in developing countries, by region and industry, 1977, 1982 and 1986–8

Sources: Tables A7.6 and A7.7

increase, in electrical machinery, came more out of increased production, and in non-electrical machinery, by 1988, local sales had risen even more than exports. In foods, chemicals, metals and other manufacturing, exports and domestic production both stagnated or, more often, declined from 1982 to 1986, so that we cannot say there was a shift of output from domestic to export markets. By 1988, exports had recovered dramatically, mostly via a shift from local to export markets, except in chemicals. The data are poor for Asian affiliates because there is so much suppression in the source, but there is little indication there of any major shifting of sales among markets. The growth of exports was accompanied either by a growth in domestic sales, or by declines in domestic sales that were too small to account for much export growth.

121

LATIN AMERICA Table 7.10 Changes in total sales, local sales and exports, US majority-owned affiliates in developing countries, by region and industry, 1982–6 and 1982–8

Sources: Tables A7.6 and A7.7

122

STRUCTURAL ADJUSTMENT IN LATIN AMERICA

CONCLUSIONS The distinctive feature of the export performance of US multinationals’ affiliates in heavily indebted Latin American countries after the debt crisis began was their sharp shift from selling in host country markets to exporting. One aspect of this shift was the faster growth of their exports than of the exports of their host countries. Another was the sharp rise in their export propensities between 1982 and 1986, larger than for the host countries, although the host countries too were shifting their sales to export markets. In both these measures, the behaviour of affiliates in the heavily indebted Latin American countries differed distinctly from that in the less indebted Asian developing countries, so that it cannot be explained as a general characteristic of the affiliates or of developing countries in general. Much of the shift in markets by US affiliates involved more reductions in host country domestic sales than increases in exports, and the affiliates’ shares in exports did not rise substantially during this period. The implication is that the affiliates were quicker than domestic firms to reduce their local sales, and they achieved much of their gains in export propensities that way. It would not be surprising that multinationals reacted this way if debt problems caused host countries to restrict the conversion of local currency income to US dollars, since there would be little incentive for US firms to accumulate depreciating local currencies. Thus, the flexibility of the multinationals’ affiliates, as compared with other firms in their host countries, in the face of the debt crisis was shown by their ability to increase exports from debt-ridden countries rapidly, their ability to change the orientation of these affiliates from domestic sales towards exports, but also by their ability to withdraw from the domestic markets when they became less attractive or more risky. Another lesson from this period is that events take time to unfold. The story of the responses to the debt crisis looked very different in 1988 from that of 1986. Over the longer period, but not the shorter one, the export growth of the three Latin American countries undergoing ‘structural adjustment’ outran even that of the Asian developing countries. And over the longer period, but not the shorter one, most of the gains in affiliate exports came from rising production rather than falling host country consumption. The results from this study also suggest that multinationals carry a potential to export from production facilities that were largely set up to serve local markets in host countries (i.e. those in Latin America), and that these firms are better equipped to convert import-substituting industries to exporting than local firms are. Future research should investigate whether the firms’ responses could be explained by changes in their host countries’ exchange rates and in trade, balance of payments, and investment policies.

123

LATIN AMERICA

ACKNOWLEDGEMENTS The research reported here is part of the NBER’s programme in International Studies. Blomström’s participation in this project was supported by the Swedish Council for Research in the Humanities and Social Science. We are grateful to Osman Sari and Zhang Qing for programming and statistical work, and to James Hayes for the preparation of the manuscript. Any opinions expressed are those of the authors and do not necessarily represent the views of the National Bureau or the sponsoring agency. NOTES 1 2 3 4

For a discussion of the Latin American development model that led to the debt crisis, see Blomström and Meller (1991). For a more general discussion of the role of trade in the structural adjustment required by the Latin American debt crises, see Edwards and Savastano (1988). Blomström et al. (1988) and Table A7.3. See López (1991) for a detailed study of the Mexican structural adjustment process.

124

STRUCTURAL ADJUSTMENT IN LATIN AMERICA

APPENDIX A

Table A7.1 Value of manufactured exports a by selected developing countries and country groups, 1977, 1982, 1986–8 ($ million)

a

Manufactures are defined to match the industry classifications in US Department of Commerce (1985). The conversion from SITC Revision I commodity codes given in UN tapes to industry groups and manufacturing as a whole, described in Blomström et al. (1988), has been revised here to remove certain fish and shell fish (SITC 031), rice (SITC 042), green coffee (SITC 07111), tea and mate (SITC 074), and spices (SITC 075) from manufactured products and from manufactured food products. b Estimated from 1987 assuming the same rate of growth as for total exports. c NBER estimates. Export data were not shown in the UN trade tape and exports were estimated from partner country imports. d Corresponding 1982 figure is 6,113. e Totals on UN trade tape basis are 42,038 for 1987 and 54,628 for 1988. f Totals on UN trade tape basis are 23,223 for 1987 and 30,380 for 1988. g Totals on UN trade tape basis are 169,022 for 1987 and 216,108 for 1988. Sources : UN trade tapes and Lipsey et al.(1991). Shortcut estimates are from various issues of United Nations Yearbook of International Trade Statistics and Commodity Trade Statistics.

125

LATIN AMERICA Table A7. 2 Sales outside the host country by US majority-owned manufacturing affiliates, various years, 1977–88

Sources: US Department of Commerce (1981), Tables IIIF5 and IIIH3; (1985), Tables IIID3 and IIIE3; (1989), (1990) and (1991), Tables 29 and 36

126

STRUCTURAL ADJUSTMENT IN LATIN AMERICA Table A7.3 Total sales by US majority-owned manufacturing affiliates, various years, 1977–88

Sources: US Department of Commerce (1981), Table III.F5; (1985), Table III.D3; and (1989), (1990) and (1991), Table 29

127

LATIN AMERICA Table A7.4 Sales to the USA by US majority-owned manufacturing affiliates in developing countries, various years, 1977–88

Sources: US Department of Commerce (1981), Tables III.F5, III.H3, III.H4 and III.H5; (1985), Tables III.D3, III.E3, III.E4 and III.E5; (1989), (1990) and (1991), Tables 29, 36, 37 and 38

128

STRUCTURAL ADJUSTMENT IN LATIN AMERICA Table A7.5 Value of manufacturing output in producers’ prices, twelve developing countries, 1977, 1982, 1986–8

a

Data for Thailand cover only reporting firms and are not consistent in coverage. Hong Kong, Malaysia, Singapore, South Korea and Taiwan. c Hong Kong, Malaysia, Singapore and Taiwan. Sources: For production data in own currency, United Nations (1983), (1988), (1989) and (1990): Total manufacturing less ISIC 353 (petroleum refineries) and 354 (petroleum, coal products), and, for Taiwan, from Republic of China (1988), Table 94, and (1 990), Table 149. Exchange rates for all countries except Hong Kong are from the IMF International Financial Statistics Yearbook. Exchange rates for Hong Kong and Taiwan are from the Federal Reserve Bulletin, United Nations (1991), and Republic of China (1987) and (1989) b

129

LATIN AMERICA Table A7.6 Sales outside the host country by US majority-owned manufacturing affiliates in developing countries, by industry group and region, various years, 1977–88 (units: $ million)

Sources: US Department of Commerce (1981), Tables III.F5, III.H3, III.H4 and III.H5; (1985), Tables III.D3, III.E3, III.E4 and III.E5; (1989), (1990) and (1991), Tables 29, 36, 37, 38 and 53. US imports from affiliates have been used in some cases as an alternative to affiliate sales to the USA

130

STRUCTURAL ADJUSTMENT IN LATIN AMERICA Table A7.7 Total sales by US majority-owned manufacturing affiliates in developing countries, by industry group and region, various years, 1977–88

Sources: US Department of Commerce (1981), Table III.F5; (1985), Table III.D3; (1989), (1990) and (1991), Tables 29, 30, 36, 38 and 53

131

LATIN AMERICA

APPENDIX B: References to Appendix A Republic of China (1987) Quarterly National Economic Trends, Taiwan Area, Republic of China, no. 37, May, Directorate-General of Budget, Accounting, and Statistics, Executive Yuan, Taiwan, ROC. ——(1988) Statistical Yearbook of Taiwan, 1988 (in Chinese), DirectorateGeneral of Budget, Accounting, and Statistics, Executive, Yuan, Taiwan, ROC. ——(1989) Quarterly National Economic Trends, Taiwan Area, Republic of China, no. 45, May, Directorate-General of Budget, Accounting, and Statistics, Executive Yuan, Taiwan, ROC. ——(1990) Statistical Yearbook of Taiwan, 1990 (in Chinese), DirectorateGeneral of Budget, Accounting, and Statistics, Executive Yuan, Taiwan, ROC. United Nations (1983) Yearbook of Industrial Statistics, 1981. ——(1988) Yearbook of Industrial Statistics, 1986. ——(1989) Yearbook of Industrial Statistics, 1987. ——(1990) Yearbook of Industrial Statistics, 1988, New York: United Nations. ——(1991) Yearbook of International Trade Statistics, 1989, New York: United Nations. US Department of Commerce (1981) US Direct Investment Abroad, 1977, Washington, DC: Bureau of Economic Analysis. ——(1985) US Direct Investment Abroad: 1982 Benchmark Survey Data, Washington, DC: Bureau of Economic Analysis. ——(1989) US Direct Investment Abroad: Operations of US Parents and their Foreign Affiliates, Revised 1986 Estimates, Washington, DC: Bureau of Economic Analysis. ——(1990) US Direct Investment Abroad: Operations of US Parents and their Foreign Affiliates, Revised 1987 Estimates, Washington, DC: Bureau of Economic Analysis. ——(1991) US Direct Investment Abroad: Operations of US Parents and their Foreign Affiliates, Revised 1988 Estimates, Washington, DC: Bureau of Economic Analysis. BIBLIOGRAPHY Blomström, Magnus, Kravis, Irving B. and Robert E. Lipsey (1988) ‘Multinational firms and manufactured exports from developing countries’, NBER Working Paper no. 2493, January. Blomström, Magnus and Lipsey, Robert E. (1990) ‘Foreign firms and export performance in developing countries: Lessons from the debt crisis’, NBER Working Paper no. 3412, August. Blomström, Magnus and Meller, Patricio (eds) (1991) Diverging Paths: Comparing a Century of Scandinavian and Latin American Economic Development, Baltimore, MD: Johns Hopkins University Press. Edwards, Sebastian and Savastano, Miguel (1988) ‘Latin America’s intraregional trade: Evolution and future prospects’, NBER Working Paper no. 2738, October. Lipsey, Robert E., Molinari, Linda and Kravis, Irving B. (1991) ‘Measures of prices and price competitiveness in international trade in manufactured goods’, in Peter Hooper and J.David Richardson (eds) International Economic Transactions, Studies in Income and Wealth, Chicago: University of Chicago Press for the NBER, 1991. López, Julio (1991) ‘Contractive adjustment in Mexico 1982–1989’, Banca Nazionale del Lavoro Quarterly Review no. 178, September.

132

Part III ASIA

8 EAST ASIAN DEVELOPMENT AND JAPANESE DIRECT INVESTMENT Thomas Andersson and Staffan Burenstam Linder

INTRODUCTION The phenomenal development in East Asia separates this region from the rest of the developing world. Many, sometimes contradictory, explanations have been put forward. Some have emphasized the efficiency of market forces (see e.g. Burenstam Linder 1986), others the active role of governments (Sachs 1986). What is the least controversial is that a relatively great openness to trade has been a key factor in establishing high investment efficiency, savings and growth (Balassa 1982, 1991). Still, certain countries, such as the Philippines, have not succeeded in successfully pursuing such a strategy. It remains to be fully understood what has enabled most of the East Asian countries to do so. One set of issues concerns the role of capital inflows and how they relate to openness (cf. Aizenman 1991). Following great problems with their external balances and mounting debts, many developing countries have virtually ceased to have access to international capital markets. A feature characterizing most countries in East Asia is the relatively large inflow of direct investment, i.e. equity investment undertaken by multinational corporations. While there have been extensive western and Chinese investments for a long time, in the 1980s the bulk was provided by Japanese firms. Following the appreciation of the yen and higher costs in Japan, a range of economic activities have been transferred from Japan to its Asian neighbours. In contrast to portfolio investment, direct investment constitutes not primarily a transfer of capital, but of firm-specific factors related to technology and skills in management, distribution and so forth (see Dunning 1977 or Caves 1982). The bundle of factors provided remains under the control of the foreign investor, who also carries the risk of failure. The recipient country generally gains because of the investors’ inability to capture all the rents associated with their activities, which spill over to domestic agents through wages, taxes, influences towards sharper competition, etc. At the same time, there may also be negative effects, especially when direct investment is motivated by barriers to trade, such as the establishment of monopolies, suppression of

135

ASIA

domestic entrepreneurs, and an anti-competitive bias of technology. Other negative effects may take the form of depletion of natural resources and negative external effects on the environment. Japanese direct investment in East Asia has been viewed in different ways. According to some, it serves as a locomotive for growth in the region as a whole. The popular ‘flying wild geese’ concept captures the idea that the transfers of goods and services within East Asia generate development by restructuring the division of labour (cf. Shinohara 1972). Kojima and Ozawa (1984) maintained that Japanese direct investment would be more favourable to social welfare in the host countries than US investment, owing to a greater emphasis on export orientation. Kojima (1990) further argues that Japanese direct investment is in line with comparative advantage, and consequently acts as a catalyst for growth and economic development in general. According to Naya (1990), direct investment has contributed to the expansion of exports in East Asia, but the policies of the host countries would have been crucial for this impact. Others argue that Japanese direct investment brings economic and political suppression, and depletes the resource basis of poor Asian countries (cf. Nester 1990). The dominance of Japan creates a certain fear of dependency, and some resentment of the cultural influence it may inflict. There are at least three reasons why the role of Japanese direct investment is still associated with question marks. First, there have been misconceptions and a general lack of understanding between Japan and the west, and the activities of the Japanese have been viewed with suspicion. This was true even before the country’s military expansion in the early part of the twentieth century, and it has remained true since. Second, the spurt in Japanese direct investment has come late, and there has not been time to study it sufficiently. Third, there have so far been limited and unsatisfactory data on Japanese direct investment, especially regarding its sectorial and industrial composition. To obtain a clearer picture of the distribution and motivation of Japanese direct investment across sectors and countries, this study examines Japanese data which has not previously been official. A fundamental question concerns whether Japanese direct investment is motivated by economic conditions rather than, for example, cultural or institutional attractions. If the latter were prominent, it would not be possible to relate the pattern of direct investment to economic conditions. Thus, we will examine the usefulness of various economic variables to explain the pattern of Japanese direct investment in East Asia. It will be noted that this investment may be motivated by both ‘push’ effects in Japan and ‘pull’ effects in East Asia. The former include higher wages, a stronger yen, or government planning on the part of MITI.1 Examples of the latter are better infrastructure and higher labour productivity in East Asia, or protected host country markets. An interrelated issue concerns what kind of direct investment East Asia obtains. Some observers view Japan as running the risk of being ‘deindustrialized’ because Japanese companies move activities throughout the industrial spectrum to countries with lower production costs. Others argue that

136

EAST ASIA AND JAPANESE DIRECT INVESTMENT

advanced activities are dependent on mutual linkages between industries in the home market, and consequently do not leave Japan (cf. Shinohara 1989). In examining the motivations of Japanese direct investment, we will take note of its sectorial composition interpreted in terms of skill and technology intensity. This is also potentially important for the impact on the social welfare of the host countries. Finally, to what extent can other developing countries emulate, or at least learn from, the development in East Asia? Given that Japanese direct investments are desirable, is it possible for other developing countries to attract them successfully as well? If economic variables are of crucial importance, is it necessary to have achieved a high income level, high growth or an open economy, before Japanese direct investment can be attracted? Or are such characteristics a consequence of direct investment? The essay is organized as follows. The next section reviews the development performance of the Asian newly industrialized economies (ANIEs) and the countries in the Association of South-East Asian Nations (ASEAN),2 and the policies they have pursued. The third section surveys the importance of inward investments, and presents ‘push’ and ‘pull’ factors which may have spurred Japanese direct investment in East Asia. Data and hypotheses for empirical testing are presented in the fourth section. The following section reports the results of some simple regressions and the final section summarizes and discusses implications for other developing countries. BEHIND THE EAST ASIAN GROWTH MIRACLE The ANIEs have undergone a remarkable development in the last decades, whether measured in per capita income or other welfare indices. It has also become evident that the countries in ASEAN, with the exception of the Philippines, are following them with a good deal of success. Together with the People’s Republic of China, the four ANIEs and the other four major members in ASEAN are here referred to as ‘East Asia’. Of course, the region also includes Indo-China and North Korea. Under the period studied, these countries have had centrally planned economies to a greater extent than China, and received or allowed little direct investment. Hence, they are excluded in the following. Table 8.1 compares the performance of East Asia with the major economies in South Asia, which have also been more successful in economic terms than most other developing countries in the 1980s. The average annual rate of growth, the growth of exports and average gross domestic savings are given for 1980–8. As can be seen, East Asia has generally had higher growth in terms of income, but the Philippines is a clear exception. The same applies to export growth, although both the Philippines and Indonesia (which suffered a fall-back in oil exports) were exceptions in this case. In South Asia, Pakistan stands out as similar to East Asia in terms of both income and export growth. The record of the other countries in South Asia is less impressive, but still respectable. The most substantial difference

137

ASIA Table 8.1 Growth in income, growth in exports and average savings rates

a

For 1981–8. Sources: World Bank (1990) and Asian Development Bank (1983, 1990)

between the regions appears in the savings rate, where the Philippines again forms a (partial) exception. The savings rate has traditionally been high in East Asia, and the gap with other developing countries has widened in the 1980s. A voluminous literature has sought to explain why the economic performance of the ANIEs and most of ASEAN differs markedly from other developing countries. Many studies have focused on factors that are unique to the region. These include Confucian values, the economic skills of the Chinese and Chinese minorities, the Japanese ethos, the industriousness and high level of education in the region, lack of raw materials, abundant US aid, political stability, and so forth. A basic problem with these explanations is their ad hoc nature. The region is heterogeneous in every way, with countries differing in history, constitution, race, size, resource endowments and climate. However, high growth is the typical pattern, with the exception of the Philippines. Indo-China and North Korea display a very different record. China itself made impressive progress from 1979 onwards, when it adopted an ambitious reform programme. Small-scale rural enterprises grew spontaneously from a modest size to achieve a total labour force of some 90 million. These semi-private enterprises accounted for about one third of China’s industrial output as of 1989. After the events in Tiananmen Square, the

138

EAST ASIA AND JAPANESE DIRECT INVESTMENT

country temporarily halted both its reforms and economic progress. This pattern points towards a major influence of economic policies, although the importance of other factors should not be denied.3 While the specific design of policies differs, those countries which have succeeded share certain basic characteristics. The most clearcut and general difference from other developing countries is that the ANIEs and ASEAN have pursued relatively more export-oriented policies (Balassa 1982, 1991). This has included promotion of sales on world markets, exposure to competition and utilization of scale advantages. Other developing countries have rather biased their incentive systems away from world markets and comparative advantage. Using indices for price distortions, it has been shown that the ANIEs and ASEAN have generally had less distorted prices than other developing countries (World Bank 1983). With the exception of Hong Kong, the ANIEs and ASEAN have not practised laissez-faire. Exports promotion has not taken place through a general liberalization of imports, but there has been a combination of exchange rate management and fiscal incentives for exports. In addition to the provision of infrastructure and other minimal functions, the governments can be said to have intervened in five broad areas. They have: 1 2 3 4 5

set up public enterprises to produce private goods; claimed responsibility for macroeconomic stability; intervened in the distribution of ownership through land reform in some cases; engaged in industrial development through industrial incentives; and encouraged inward direct investment, although generally not allowed it in all activities.

Of these five areas, it is clear that the first one has been a relatively unimportant factor in the growth of the region. The state-owned companies are not particularly ‘small’ (e.g. compared with Latin America), but their share of industry is small relative to that of the private sector, and their growth has been low throughout, with the exception of the oil sector in Indonesia. The importance of the second and third factors has been emphasized for Japan, Taiwan and Korea (Sachs 1986). Regarding macroeconomic policy, government budgets have been kept in balance, which has contributed to high savings and low inflation rates, among other things. Nominal exchange rates have been adjusted so as to keep real exchange rates at reasonable levels. The strong public finances have allowed governments not to promote industrial growth at the cost of taxing agriculture which, rather, has been protected compared with industry. At the same time, the land reforms in South Korea and Taiwan created, as in Japan, a conservative peasantry of independent proprietary farmers which has lent support to the national governments. In this respect, the Philippine government has not followed. The involvement of governments in the management of private industries is

139

ASIA

the hardest to evaluate. Korea, for example, has supplemented moderate protectionism with selective export promotion, achieving neutrality on the whole. Whether the economic performance would have been even better without this kind of interference we do not know. It is clear, however, that the measures used have generally aimed at stimulating the efficiency of markets. Incentives have promoted activities in line with comparative advantage: at one point labour intensity, at another capital intensity. Input and output prices have been carefully maintained at the level of the world market, so as to encourage competition. The objective has not been to make life easy for business, but subsidies have been connected to demanding standards for progress in terms of output, exports or investment in R&D (Amsden 1991). Meanwhile, governments have intervened to promote the bargaining position of domestic firms relative to foreign ones, and facilitated the exploitation of static and dynamic scale advantages (Hong 1990). While this clearly applies to the ANIEs, there have been more of the usual protectionist policies in ASEAN, giving rise to vested interests which seek to maintain their privilege of not having to be competitive on world markets. This applies particularly to the Philippines under Marcos, and to Indonesia. Concerning the last area of government influence, direct investment has been promoted through provision of various incentives, such as tax holidays and tariff reductions. Some of these have been implemented in special free trade zones. Such incentives are unlikely to exert a major influence on the location decisions of multinational firms, partly because ‘all’ countries can be expected to do their best to attract them. See Andersson (1991), where competition between potential host countries is analysed. After systematically reviewing incentives and investment behaviour during the last decades, Contractor (1990) similarly draws the conclusion that host country incentives have not influenced the pattern of direct investment in any significant way (see also Farrell 1985, UNCTC 1988). The risk of forced divestment or expropriation is an exception, as it may discourage direct investment from being undertaken. Since the late 1970s, expropriations have become virtually non-existent in developing countries, and they have certainly ceased to be seen as a threat in East Asia. But if government incentives do not account for the expansion of direct investment in East Asia, what does? Of course, policies still matter owing to their impact on trade and development in general. Before examining the investment behaviour of the Japanese, the next section discusses the background of foreign investment flows in East Asia. INWARD INVESTMENT Foreign investment is often said to have helped to set growth off in East Asia. Most notably, there was a great deal of foreign development assistance from the United States to Japan, South Korea, Taiwan and Indonesia. Equally large or even larger official flows to other countries have not been followed by any similar spurt in growth, however. Foreign development assistance alone does not explain the

140

EAST ASIA AND JAPANESE DIRECT INVESTMENT Table 8.2 Debt service ratio, total direct investment and Japanese direct investment

Sources: Calculations on the basis of World Bank (1990), Asian Development Bank (1990), UNCTC (1983) and the Ministry of Finance (1990)

difference in development performance between East Asia and other regions of developing countries. Leaving foreign development assistance aside, how large have the inflows of external capital been? Table 8.2 compares East and South Asia in terms of debt service ratio in 1981 and 1989, the total stock of inward direct investment in percentage of GDP 1979, and the stock of Japanese direct investment in percentage of GDP as of 1979 and 1989. For three of the ANIEs, the debt service ratios were zero throughout. The 21.7 per cent recorded by Korea in 1981 reflects that this country absorbed a good deal of foreign lending in the early stages of industrialization, but the other ANIEs did not. The ASEAN countries, on the other hand, resemble South Asia in terms of debt service ratio. Comparing export earnings, the borrowing of both the ANIEs and ASEAN in the international credit markets stands out as very modest in relation to that of other developing countries. Thus, not a great deal of foreign borrowing has enabled the ANIEs and ASEAN to outperform other developing countries. Countries which are relatively scarce in capital have a high rate of return on the margin, and, hence, should import capital from abroad. However, it cannot be

141

ASIA

taken for granted that capital is invested effectively. Countries are sovereign, and may consume rather than invest, or invest so as to enhance the prestige of the ruling elite rather than generate an economic return. Common examples are the build-up of industries, monuments or buildings which are conspicuous but unproductive, or military spending which strengthens the government’s grip on power. From the late 1970s onwards, current account deficits and the accumulation of large external debts have gone hand in hand with low rates of growth and a deteriorating standard of living for large parts of the population in Africa, Latin America and South Asia. Accepting direct investment, in contrast, East Asia has allowed capital exporters to maintain a relatively high degree of control over the funds provided. Direct investment has played a major role in East Asia for a long time. In the wake of colonization there were at first substantial European investments in the extraction and trade of natural resources, especially in Indonesia, Malaysia and Singapore. The USA dominated after the Second World War, focusing on the same countries, as well as on the Philippines and Taiwan. Comparing direct investment from all sources, Table 8.2 includes only the year 1979, as data for the different countries is not available for the late 1980s from a single source.4 As of this year, East Asia had attracted considerably more direct investment than South Asia. The closed Chinese economy was an exception, and Thailand and Korea had also attracted a modest amount. Singapore, Malaysia, Hong Kong and Indonesia, on the other hand, had received exceptionally large direct investments. For Indonesia, the oil sector, which accounted for perhaps as much as 50 per cent of the stock of all foreign direct investment in the late 1970s, is not included. The notable difference between the regions stands out in comparisons of East Asia with most other developing countries as well.5 The effects of capital flows on factor productivity and domestic savings are still a matter of controversy. Concerning the Asian countries, however, Gupta and Islam (1983), Lee et al. (1986) and others have found that the private flows have had a favourable impact on savings as well as factor productivity. It is, indeed, plausible that direct investment plays an important role in the East Asian development process. Following the debt problems of most developing countries, the creditor side has responded with demands for reform before providing additional funds. Meanwhile, the attitudes towards direct investment have become more favourable, as reflected in the decline of nationalizations, reduced taxes and the provision of investment incentives throughout the Third World (Andersson 1991). Still, the flows of direct investment to all developing countries have declined along with the drying up of portfolio investments. While the developing world held abut 28 per cent of the total stock of direct investment in the late 1970s, its share had shrunk to about 22 per cent in the mid 1980s (UNCTC 1988). Only East Asia has enjoyed increasing inflows. At the same time, the west’s domination in the region has been broken by the expansion of the Japanese. In the two columns on the right, Table 8.2 shows the remarkable increase in Japanese direct investment that has taken place in East Asia in 1979–89. This

142

EAST ASIA AND JAPANESE DIRECT INVESTMENT

data, which is reported by the Ministry of Finance in Tokyo, is subject to pitfalls which will be commented on in the subsequent section, but should not bias the comparison over time. Between 1979 and 1989, the stock of Japanese direct investment in relation to the GDP of the host countries at least doubled in three of the ANIEs and in all four countries in ASEAN. The most remarkable increases were recorded in Thailand (from 1.2 to 5.1 per cent), Hong Kong (4.2 to 17.3 per cent), Singapore (7.6 to 22.2 per cent) and Malaysia (2.4 to 6.7 per cent). Until the late 1970s, overseas direct investment made up only about 1 per cent of total fixed capital formation in Japan, which can be compared with 3 per cent in Germany or 4.5 per cent in the USA. In 1989 the share surpassed 6 per cent in Japan, which is more than for all other major industrialized economies. The sectorial focus, which used to be mining, natural-resource related investment and manufacturing, shifted towards finance, real estate, transportation, commerce and services. In terms of regions, there has been a major shift away from the developing to the developed countries, first the USA and more recently the European Community. Obviously, the 1980s have seen new motivations for Japanese companies to invest abroad. From the viewpoint of East Asia, changes within Japan itself, or in Japan’s relations with other countries, can be regarded as ‘push’ factors. The following partly interrelated factors belong to this category: 1

2

3 4

5

Financial deregulation and integration. The overhaul of Japan’s foreign exchange law in 1980, and financial liberalization in general, has enabled Japanese financial institutions to expand their activities on a worldwide scale, including the undertaking and servicing of direct investment. Meanwhile, the development of the Euro-currency and the Euro-bond markets together with financial deregulation in the USA and the UK have provided opportunities for Japanese financial institutions to absorb short-term borrowing and convert it to long-term lending and investment. Most went into US securities in the early 1980s, but direct investment has become an increasingly important destination in the late 1980s. Current account imbalances. Huge surpluses on the Japanese side, especially after 1983, have spurred capital outflows. These have also fuelled political pressure on Japan from the USA and the EC to reduce its exports, further inducing direct investment for production in importing countries or exports from third countries. Currency alignments. An appreciating yen vis-à-vis the dollar has favoured production by Japanese companies abroad from 1985 onwards. Rising costs of labour, land and other inputs in Japan. Robotization and upgrading of technology have limited the impact, but it has become relatively more favourable to move especially labour-intensive production to countries with lower costs. Adaptation in organization. The Japanese have become more capable of adapting their organizations to local conditions, applying this to the USA and Europe as well as to Asia.

143

ASIA

The above factors speak for a general increase in Japanese direct investment, which we have seen in the 1980s. In the developing world, however, the main destination has been East Asia. The bottom part of the two columns on the right in Table 8.2 shows that South Asia has attracted virtually no direct investment from Japan. East Asia probably enjoys an advantage due to its geographical location, but this is hardly the only reason why the Japanese focus so heavily on this region. What are the ‘pull’ factors attracting direct investment in East Asia? At least five have been put forward in the literature: 1 2

3

4 5

Favourable macroeconomic conditions, such as high growth and moderate inflation. Sound economic policies, particularly open trade regimes and low taxes. In some of the ASEAN countries, however, trade barriers have also attracted direct investment. The ANIEs and ASEAN countries have had their currencies more or less pegged to the dollar, which has made operations in these countries more attractive for Japanese firms. Political stability, including a high priority for economic development at the national level. Institutional and cultural conditions, creating a work ethic which could readily be taken advantage of by the Japanese.

Many of the above factors, ‘push’ as well as ‘pull’, are likely to have played a role. Still, we need a clearer picture of the distribution and motivations for the Japanese investments in East Asia. The next section presents data for empirical testing, and sets up the hypotheses to be examined in this study.

DATA AND HYPOTHESES Seeking to explain the motivations of Japanese direct investments in East Asia in greater detail, we investigate two dependent variables below. On the one hand, we study the absolute flows of Japanese direct investments for 1979–89 (y1) and on the other their relative increase for 1979–89 (y2). The former is referred to as model 1, the latter as model 2. For both dependent variables, the basis of the data is Ministry of Finance (Tokyo) statistics. It should be noted that this is based on approved, not actually implemented, investments. There is often a discrepancy between the two, with some of those approved not being undertaken at all, and others being delayed a few years. In addition, some types of loans and acquisitions of bonds are included, while repayments of these loans are not (neither are the withdrawals of equity investment). This might seriously affect panel data, and particularly investigations of mutual interactions between investment behaviour and country developments. It is less troublesome for the present study, as the determinants of the average increase of approvals over an extended period of time should reflect those of the average actual implementation of investment.

144

EAST ASIA AND JAPANESE DIRECT INVESTMENT Table 8.3 Variable definitions and descriptive statistics

Because there were no Japanese direct investments at all in the People’s Republic of China in the early 1980s, this country recorded an infinitely large relative increase in direct investment. Thus, from the start China was excluded from model 2. Inclusions of the country in model 1 worsened the results, which may indicate that this country was too different from the market economies in the region to be readily analysed within the same framework. The investment flows have been divided into five sectorial aggregates; three in manufacturing and two in non-manufacturing. These have been ranked in the following order: 1 2 3 4 5

Basic non-manufacturing (forestry, fishing and mining). Basic manufacturing (food and textiles). Resource-processing manufacturing (pulp and paper, chemicals and ferrous and non-ferrous metals). Advanced manufacturing (machinery, electronics and transport machinery). Commerce and services (commerce, banking and insurance, services and transport).

This division aims to reflect increasing skill and technology intensity, as well as decreasing dependency on labour costs. Thus, we create a discrete variable (z1), ‘sector’, which takes a value between 1 and 5 in accordance with the sectorial classification (see further below). Direct investment is generally believed to be influenced by ‘macroeconomic’ variables, such as growth, income level, domestic markets, openness in trade and investments, and taxes. Together with the sectorial variable, these are used as our explanatory variables. However, the formulation of hypotheses hinges on the special characteristics of the Japanese economy in the period studied, and how they relate to East Asia. The explanatory factors are presented in Table 8.3,

145

ASIA

together with some descriptive characteristics. The reasons for their inclusion, and their expected influences on the dependent variables, are as follows: •

•

•

•

• •

Average growth (x1) is expected to exert an unambiguously positive effect on direct investment, as it creates a greater domestic market as well as more favourable socioeconomic conditions. If such a positive influence is actually obtained, however, we cannot rule out a simultaneity problem, in the sense that the Japanese investments may have spurred high growth rather than the other way around. The income level (x2) is normally positively related to direct investment. However, the Japanese direct investments in East Asia should partly be motivated by lower labour costs, which suggests that the income level exerts a negative influence on both dependent variables. The economy’s size (x3) can be expected to exert a positive influence on the first dependent variable, since a greater economy should attract more direct investment in absolute terms. However, the relative increase should be negatively influenced, as a large economy should have attracted a good deal of investment before already, and therefore be smaller. Openness, or exports as percentage of GDP (x4), should exert a positive impact on investment flows in both models. In case a positive impact prevails we again have a simultaneity problem, which is returned to below. If trade barriers would rather have attracted investments, we would instead expect a negative impact. The level of taxation in relation to GDP (x5) indicates higher costs for business in general, and should exert a negative influence in both models. The sectorial variable (z1) should exert a negative influence if East Asia has attracted primarily direct investments which rely on low technology and low skill intensity. If the increasing attractiveness of East Asia for high skill- and technologyintensive sectors dominates, however, we instead expect a positive influence. This should be most probable in model 2, which measures relative increase.

Concerning the timing of the explanatory variables, we do not know whether investors consider past observations or form rational expectations by looking ahead. In order not to include data which did not influence investment decisions in the period of study, we disregard the first and the last years. Thus, average growth is calculated for 1980–8. The other explanatory variables indicate ‘level’ rather than change. For these we use the ‘last’ year, 1988, as firms would have liked to base their investment decisions on the state of affairs towards the end of the period of study. The bulk of the total investments was actually undertaken in the years 1987–9. EMPIRICAL TESTS Each one of the explanatory variables is believed to influence investors’ expectations regarding the profitability of investment opportunities. However, they are merely proxies for the particular circumstances that are crucial for

146

EAST ASIA AND JAPANESE DIRECT INVESTMENT Table 8.4 Results

individual firms. Thus, we have no basis for more complex functional forms. This means that we test the following (1)

where represents the two models. The disturbance term fulfils the usual assumptions required for ordinary least-squares estimation. The x are the

147

ASIA

explanatory variables and z the sectorial variable which is discrete, taking a value from 1 to 5 in accordance with the sectorial specification above. The results of the specification in (I) are given in Table 8.4. For the second model, which measures relative increase, a logarithmic transformation of the explanatory variables was undertaken to limit the influence of extreme observations. The same transformation was checked for model 1 as well, but did not improve the outcome, wherefore only the untransformed relation is reported in that case. As can be seen, model 1 resulted in an adjusted R2 close to zero, and there were no significant effects at the 5 per cent level for any variable. Model 2, on the other hand, produced an adjusted R2 of 0.61, and significant effects except for one variable. As expected, the rate of growth was found to increase the flow of investment, while the size of the economy and the tax rate reduced it. The income level did not have any significant impact. The sectorial factor exerted a positive influence, indicating that the relative increase in investment was stronger in advanced sectors. As mentioned, a positive influence from the sectorial variable was expected to be the most probable in model 2. The ‘openness’ variable did not have the expected positive effect, but it was negative and significant at the 5 per cent level. This indicates that ‘openness’ was less of a motivation for Japanese direct investment in the 1980s than it had been before. This might be due to the difference in policies between countries, with the Philippines and Indonesia attracting direct investment by protecting markets. Owing to the lack of a positive effect we do not have to deal with the problem of causality that is associated with this variable. For the future it remains important to clarify under what circumstances there actually is a positive impact of ‘openness’ on investment, and when there is a positive impact of investment on ‘openness’. While model 2 generated a satisfactory outcome, the same cannot be said of model 1. One possible reason is that the influence of some country characteristics on the absolute size of the investment flows is likely to be different in different sectors. A higher income level, for example, is unlikely to motivate direct investment in mining, but it should stimulate more investment in financial services. To account for this, we multiply each independent variable by the sectorial variable, creating a series of multiplicative variables. These are expected to account for a major impact on the absolute size of the investment flows, measured in model 1. In model 2, they may matter less, as direct investments before 1979 were also affected by such multiplicative effects. The estimates obtained with the multiplicative variables included in model 1 are presented in the third section of Table 8.4, still without logarithmic transformation (logs increased adjusted R2, but led to fewer significant effects). As expected, the multiplicative variables did not add much to model 2.6 Owing to a problem with heteroscedasticity, following the construction of the new variables, we used White’s (1980) estimator to compute heteroscedasticityconsistent standard errors. The t-values obtained with this method are given in the fourth column. As can be seen, the results of model 1 turn out considerably stronger with the multiplicative variables included, and controlling for the heter-

148

EAST ASIA AND JAPANESE DIRECT INVESTMENT

oscedasticity affects the outcome only marginally.7 Adjusted R2 is only 0.29, but all effects have the expected signs and are significant at the 5 per cent level, except for the rate of growth (and the multiplicative growth-sector variable), which has the expected positive sign but is insignificant. The sectorial factor exerts a negative influence, meaning that there are smaller flows in absolute terms in advanced sectors. The multiplicative variables throughout have the reverse signs of the respective macroeconomic variables taken alone. So far, we have not included any examination of the extent to which different countries, or kinds of countries, in themselves exert any impact on the investment flows. Such influences would be expected to the extent that institutional and cultural factors are of major importance. Unfortunately, this cannot be investigated in any thorough way without panel data, i.e. year-by-year data on dependent as well as independent variables, where each country’s influence can be captured through a fixed effect. Such data enables a detailed examination of shifts over time and interdependency in the flows to different countries. At the same time, it is extremely sensitive to the formulation of expectations, i.e. whether these are based primarily on past events or foresight concerning the future. Thus the present study has focused on the average trends in Japanese direct investments for 1979–89. Tentative tests of country-specific effects were undertaken by dividing the countries into groups, and including them as dummies in the above regressions. Three dummies have been used, representing the following country groups: 1 2 3

the city states, Hong Kong and Singapore; the other ANIEs, Taiwan and Korea; and the large ‘growth economies’ in South-East Asia, Malaysia, Thailand and Indonesia.

The Philippines was not classified, being atypical of East Asia in terms of political, social and economic stability. Various observations of, for example, attitudes towards foreign investors among the general public, as well as in the bureaucracies, lead us to expect that 1 and possibly 3 would be most associated with a positive effect on direct investment, while 2 may be less so. Rerunning the most successful tests, displayed in Table 8.4, with country-group dummies included, no results were obtained which supported these hypotheses, however. There is consequently no indication that non-economic, country-specific factors would exert a major influence on the investment decisions of the Japanese in East Asia. On the contrary, it appears that a good deal of the variation in these investment flows can be attributed to the influence of economic factors. To study countryspecific influences in depth and detail we need to explore panel data, however. SUMMARY AND CONCLUDING REMARKS The ANIEs and the countries in ASEAN, with the exception of the Philippines,

149

ASIA Ta b l e 8 . 5 S u m m a r y o f t h e e x p e c t e d a n d r e a l i z e d e ff e c t s , s i g n a n d s i g n i fi c a n c e ( n o n - s i g n i fi c a n c e i n p a r e n t h e s e s , 5 p e r c e n t l ev e l )

have recorded a favourable development compared with other developing countries during the last decades. Government policies have in most instances actively stimulated competitiveness in world markets and given rise to less distorted prices. Concerning capital inflows, neither,the foreign development assistance nor the borrowing in the international capital markets have been exceptionally large, but there has been considerably more direct investment than in other regions in the developing world. It is plausible that this has contributed high investment efficiency, growth in income and exports, and high savings. From the early 1980s, the bulk has been provided by Japanese firms, which between 1979 and 1989 at least doubled their stock in relation to GDP in two of the ANIEs, and in all of ASEAN. The reason is not government incentives directed specifically at foreign firms, but there are ‘push’ effects in Japan as well as other ‘pull’ effects in East Asia. The study has further examined the economic motivations behind Japanese direct investment in East Asia, using data not previously official. Both the absolute size of Japanese direct investments in 1979–89 (model 1), and their relative increase between these years (model 2), have been studied across the ANIEs and ASEAN. The most successful results obtained with ordinary least squares for each model are presented in Table 8.5. The absolute flows were larger in sectors based on low technology and skill intensity, but the relative increase was greater in advanced sectors. Thus, there is still more Japanese direct investment in less advanced activities, but those in more advanced ones seem to be catching up. The lower the level of income in the host country the larger the absolute flow of investment, although there was no significant impact on the relative increase. The level of taxation was negatively related to investment, as expected. Concerning growth and openness, neither of these variables exerted the expected significant effects in both models. In fact, openness was negatively related to the relative increase in model 2. For this reason, we abstain from dealing with the problem of determining whether direct investment is attracted by

150

EAST ASIA AND JAPANESE DIRECT INVESTMENT

growth and openness, or giving rise to it. Resolving such issues requires, in any case, studies of panel data. The inclusion of country-group dummies, finally, gave no indication that country-specific factors would have exerted a significant positive influence on Japanese direct investment within East Asia. Regarding implications for other developing countries, one must be careful in extending the results to other regions and types of countries than those represented in East Asia. To the extent that the findings are applicable to other parts of the developing world, countries with relatively low income but acceptable growth rates and modest levels of overall taxation should stand a chance of attracting Japanese direct investment. It may be argued, on the contrary, many countries have an inevitable drawback in terms of their geographical location. Although this is a valid argument for those which are land-locked or isolated islands, it hardly applies to the majority of countries. Moreover, it is likely to lose a great deal of whatever importance it may once have had. The ongoing development process in East Asia continuously shifts factor costs and comparative advantages, to which not only Japanese companies adapt. In recent years, there have been rapidly increasing flows of direct investment from Korea, Taiwan, Hong Kong and Singapore to the less developed countries in the region. As East Asia proceeds along the path of growth, higher labour costs and economic restructuring, there will be new opportunities for other regions. This is not to say that direct investment should serve as a substitute for foreign development assistance and other capital flows to less developed countries. In most cases, it is unrealistic to expect direct investment on such a grand scale. In addition, a satisfactory infrastructure as well as economic and social stability is necessary for countries to be attractive locations for multinational firms on a long-term basis. It can also be noted that a sound economy makes host country regimes more capable of negotiating mutually acceptable terms with foreign investors. The economic debacle in countries like the Philippines and Indonesia may have contributed to their strategies to attract foreign investors by protection, giving rise to concentrated domestic markets dominated by multinational firms. It is well known that such investments may be detrimental to the welfare of the host country (see e.g. Batra 1986), while investments attracted under free trade should exert a more favourable impact. It is beyond the scope of this study to examine the welfare implications of direct investment in detail, however. The fact that large markets, high income levels or country-specific attractions have not been crucial for obtaining Japanese direct investment within East Asia may in any case be viewed as promising from the perspective of other developing countries. NOTES 1

2

MITI, the Ministry of International Trade and Industry, has announced an ‘International Cooperation Program’, the goal of which is to make Japanese companies cut their exports from Japan, increase imports and raise direct investment overseas. The ANIEs are the People’s Republic of Korea, Taiwan, Hong Kong and Singapore.

151

ASIA

3 4 5 6 7

Except for Singapore, ASEAN consists of Thailand, Malaysia, the Philippines, Indonesia and Brunei. The last country is excluded here owing to its small size and the special characteristics of its economy. Little (1981) concluded that the success of the ANIEs is entirely due to ‘good policies and the people’. These are not really separate explanations, rather ‘good people’ have pursued good policies, and good policies have been facilitated by ‘good people’. Note that the debt service ratio and the stock of direct investment in relation to GDP is not really comparable. Moreover, the data on direct investment is highly uncertain, and varies between sources. Only Mexico, Brazil and Argentina have received direct investment on a scale that is comparable with that recorded by East Asia. Foreign borrowing accounted for much larger capital flows in these countries, however. This estimation is not reported here. The only tangible difference is that the sectorial variable ceased to be significant and that adjusted R2 increased to 0.67. Some of the variables ceased to be significant at the 5 per cent level but remained so at the 10 per cent level. We abstain from this qualification in the further reasoning.

REFERENCES Aizenman, J. (1991) ‘Inward versus outward growth orientation in the presence of country risk’, Economica 58 (229): 57–77. Amsden, A. (1991) ‘Diffusion of development: The late-industrializing model and greater East Asia’, American Economic Review, Papers and Proceedings, pp. 282–6. Andersson, T. (1991) Multinational Investment in Developing Countries, A study of nationalization and taxation, London: Routledge. Asian Development Bank (1983) Key Indicators of Developing Member Countries of ADB, Manila.——(1990) Asian Development Outlook, Manila. Balassa, B. (1982) ‘Development strategies and economic performance: A comparative analysis of eleven semi-industrial economies’, in B.Balassa, et al. (eds) Development Strategies in Semi-industrial Economies, Baltimore, MD: Johns Hopkins University Press. ——(1991) Economic Policies in the Pacific Area Developing Countries, London: Macmillan. Batra, R.N. (1986) A general equilibrium model of multinational corporations in developing countries’, Oxford Economic Papers 38:342–53. Burenstam Linder, S. (1986) The Pacific Century: Economic and Political Consequences of Asian Pacific Dynamism, Stanford, CA: Stanford University Press. Caves, R. (1982) Multinational Enterprise and Economic Analysis, Cambridge: Cambridge University Press. Contractor, F.J. (1990) ‘Do government policies towards foreign investment matter? An empirical investigation of the link between national policies and FDI flows’, mimeo., Graduate School of Management, Rutgers University. Dunning, J.H. (1977) ‘Trade, location of economic activity and the MNE: A search for an eclectic approach’; in B.Ohlin, P.-O.Hesselborn and P.M.Wijkman (eds) The International Allocation of Economic Activity: Proceedings of a Nobel Symposium, Stockholm, pp. 395–418, London: Macmillan. Farrell, T. (1985) ‘Incentives and foreign investment decisions: An opposing view’, The CTC Reporter 20: Autumn. Gupta, K.L. and Islam, M.A. (1983) Foreign Capital, Savings and Growth, Boston: Reidel. Hong, W. (1990) ‘Export-oriented growth of Korea: A possible path to advanced economy’, International Economic Journal 4:97–118.

152

EAST ASIA AND JAPANESE DIRECT INVESTMENT

Hughes, H. (ed.) (1988) Achieving Industrialization in East Asia, Cambridge: Cambridge University Press. International Monetary Fund (1988) ‘Issues and developments in international trade policy’, Occasional Paper 63, Washington, DC. Kojima, K. (1990) Japanese Foreign Direct Investment, Hitotsubashi University, Tokyo. Kojima, K. and Ozawa, T. (1984) ‘Micro- and macro-economic models of direct foreign investment: Toward a synthesis’, Hitotsubashi Journal of Economics 25:1–20. Lee, J., Rana, P.B. and Iwasaki, Y. (1986) ‘Effects of foreign capital inflows of developing countries of Asia’, Asian Development Bank Staff Paper, 30, Manila. Little, I. (1981) ‘The experience and causes of rapid labor intensive development in Korea, Taiwan Province, Hong Kong and Singapore; and the possibilities of emulation’, in E.Lee, (ed.) Export-led Industrialization and Development, Singapore: ILO, Asian Employment Programme, pp. 23–47. Ministry of Finance (1989 and 1990) Kokusai Kinyu Kyoku Nenpo, Tokyo. Naya, S. (1990) ‘Direct foreign investment and trade in East and Southeast Asia’, in R.W. Jones and A.O.Krueger (eds) The Political Economy of International Trade, Oxford: Basil Blackwell, pp. 288–312. Nester, W.D. (1990) Japan’s Growing Power over East Asia and the World Economy: Ends and Means , London: Macmillan. Sachs, J. (1986) ‘External debt and macroeconomic performance in Latin America and East Asia’, Brookings Papers 2:523–73. Shinohara, M. (1972) Growth and Cycles in the Japanese Economy, Institute of Economic Research, Hitotsubashi University, Tokyo. ——(1989) ‘High yen, overseas direct investment, and the industrial adjustments in the Asia-Pacific area’, in W. Klenner (ed.) Trends of Economic Development in East Asia, Berlin: Springer. United Nations, International Trade Statistics Yearbook, New York, various issues. United Nations Centre on Transnational Corporations (UNCTC) (1983) Transnational Corporations in World Development, Third Survey, New York. ——(1988) Transnational Corporations in World Development, Trends and Prospects, New York. United States Bureau of the Consensus, Statistical Abstract of the United States, New York, various issues. White, R. (1980) ‘A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity’, Econometrica 48:817–38. World Bank (1983 and 1990) World Development Report, New York.

153

9 DIRIGISME OR FREE-TRADE REGIME? A historical and institutional interpretation of the Taiwanese success story Christer Gunnarsson

INTRODUCTION Much has been written, for good or ill, about the economic ‘miracle’ of East and South-east Asia in general and about South Korea and Taiwan in particular. The aim of this study is to discuss the causes of the economic miracle in one of these countries, namely Taiwan, from an historical and institutional point of view. Almost all theoretical schools have presented their own strongly ideologically biased favourite interpretations of the Taiwanese success story. It has been idealized by free marketeers (liberals) and slandered by radical critics (dependistas). It is regarded sometimes as the ideal market economy with minimal state intervention, sometimes as an example of the need for aggressive dirigiste intervention. All such interpretations are gross oversimplifications. Neither the invisible hand nor the hard fist can serve as sole explanations of the Taiwanese miracle. The strength of Taiwan’s politico-economic system is rather its institutional efficiency displayed in the dichotomy between autonomous political and economic spheres, by virtue of which the autonomous state power has become a guarantee of economic competitiveness—the mainstay of the market economy. Purely from the standpoint of structural economics, there is no reason to believe that Taiwanese development cannot serve as a model for other underdeveloped countries. To apply the Taiwanese model to other economies in practice, however, appears considerably more difficult. The implementation of development strategy in Taiwan has been dependent on a series of specific historical circumstances, if not accidental events. These circumstances made it politically possible in Taiwan to carry through an economic policy that encountered stubborn political resistance in other countries no matter how theoretically interesting the strategy may have been. This means that any discussion of institutional efficiency ought to be placed in its proper historical context. THE MYTHS OF THE TAIWANESE MIRACLE Far the most common interpretation, and the one most familiar to the layman, has

154

THE TAIWANESE SUCCESS STORY

been that Taiwan is an ideal example of uncontrolled capitalist development based on the free play of market forces. Liberals and certain of the radicals are agreed on this interpretation while diverging on the evaluation of the result. Private enterprise, it is said, has been given free rein here, while the state’s involvement in the economy has been reduced to a minimum.1 Taiwan is regarded in the liberal interpretation as a classic example of the motto that ‘he governs best who governs least’.2 The role of the state ought to consist in creating favourable conditions for free enterprise and adjusting economic policy to the demands of the market, while the size of the public sector ought to be minimized (Balassa 1988, Hughes 1989, Landau 1983, 1986). The switch from import substitution to export orientation in the late 1950s is regarded as a particularly crucial factor in the Taiwanese success story. The opening of the economy to foreign competition forced Taiwanese industry to modernize its technology while at the same time foreign capital was being attracted to invest in the country. Taiwan’s incorporation into the world economy on market terms is thus held to be the chief cause of modernization, in glaring contrast to other underdeveloped countries which screened themselves off from world market competition and so preserved an obsolete politico-economic structure (Balassa 1978, Riedel 1987; Kuo et al. 1981a,b). Paradoxically enough, the notion that Taiwan’s economy has been completely ruled by the market and led by the export sector has been shared by the miracle’s severest critics among radical theorists. In accordance with the dependency school’s world view, these regard Taiwan’s economy as totally dependent on the world market. Taiwanese economic expansion is regarded as entirely export led and controlled by multinational companies. A combination of such factors as growing world trade, low production costs and an influx of foreign capital and American aid has operated to the benefit of Taiwanese capitalism (Fernandez 1980, Frank 1982, Cline 1982). The critics stress particularly that Taiwan is a special case and that its success is of an accidental and temporary nature. Taiwanese industry is trapped in its role as an exporter of light consumer goods while lacking the capability to catch up with the industrialized countries in terms of technological capacity. Government policy is regarded simply as an instrument of indigenous monopoly capitalists and multinational companies. Consequently the thesis that economic growth has also led to an equalization of incomes is rejected. The liberal idealization of the free market and the dependency school’s derogatory account of the same phenomenon were for a long time the two clearly dominant views of the Taiwanese economy. There is no doubt, however, that both views are superficial and largely erroneous. They both have the character of ideological manifestos or dogmas rather than serious analyses of the Taiwanese economy. Even though these schools hardly have a prominent place in serious research on the subject of Taiwan, it may be worth while to devote some time and space to a critical examination of their most important theses. The more empirically detailed and theoretically advanced analyses are

155

ASIA

few and have after all been conducted within a narrow circle of specialists. Outside this circle ideologically slanted notions persist, helping to cement a number of the prejudices standing in the way of an understanding of Taiwan’s economic ‘miracle’. THE FREE MARKETEERS In the liberal glorification of the free market obviously the active role played by the Taiwanese government is deliberately disregarded. The role of government, it is said, has been to provide proper institutional arrangements, not to intervene directly. There are a series of problems and unexplored assumptions underlying this interpretation. First, the assumption is made that the government has followed the ‘correct’ economic policy on all occasions (Ranis 1978, Kuo, et al. 1981a,b; Myers 1986, Scitovsky 1986). The picture that emerges from this explanation is of the invisible hand, guiding the economy in conformity with the most ideal assumptions of economic theory. At the same time we see the picture of a government, skilful and far-sighted if not clairvoyant, succeeding at every point in interpreting the invisible hand’s innermost purposes and fancies. Second, it ignores the fact that direct state intervention in the economy is still considerable. It is true that the share of industrial value added accounted for by state-owned enterprises fell from nearly 50 per cent in the mid 1950s to slightly more than 10 per cent in 1990, but the state still does control key sectors. Energy production (oil, gas, electricity) is completely monopolized by the state, as also is the production of wine and spirits. The state also controls large parts of the important sugar and steel industries (Crane 1987). State-owned enterprises are usually large and concentrated in capital-intensive heavy industry, as is shown for example by the fact that during the 1980s they accounted on average for more than 30 per cent of the industrial sector’s entire capital formation (Crane 1987). Of Taiwan’s ten largest industrial firms no fewer than nine are state owned. The sales of the biggest firm, Chinese Petroleum Corporation, are seven times larger than those of the biggest private firm and thirty times larger than the sales of the biggest subsidiary company of a multinational concern (Crane 1987). Third, government policy does not seem to contain any clear conception of any contradiction between the retention of a large state sector and the simultaneous encouragement of private enterprise. On the contrary, in spite of a recently growing interest in privatization state-owned enterprises continue to be allotted an important role in the government’s industrial policy. State-owned enterprises have been used as a central instrument of the government’s plans for national development and restructuring of industry. State-owned enterprises have frequently been assigned a pioneering position in key sectors. Massive capital sums have been invested in activities which, because of large capital requirements and small prospects of short-term profitability, have been of little interest to private enterprise. State-owned enterprises have been considered capable of

156

THE TAIWANESE SUCCESS STORY

shouldering responsibility for the initial investments necessary to prepare the way for private successors (Nunnenkamp 1981). During the industrial restructuring of the 1970s and 1980s, state-owned enterprises have had a central role with respect to technological upgrading and the introduction of new products (Wu Yuan-li 1985, Gold 1986). State-owned enterprises were considered capable, by virtue of their size and the state guarantees available, of accepting short-term losses while investments were being made with a view to rationalization and reconstruction (Kuo 1983). Whether this view of the spearheading industrial adjustment concurs with actual facts or merely reflects a view delineated by the government might, however, be a matter of controversy. It is far from certain that a large state industrial sector is optimal or that restructuring is best achieved through its being led by state-owned firms. Private business and industry argue just as frequently that state-owned firms have brushed aside or blocked private initiatives. Others contend that on the whole the targeted policy for industry has been a failure (Balassa 1988, Dollar and Sokoloff 1989, Hughes 1989). Fourth, Taiwan’s trade policy is not as liberal as liberal rhetoric might lead one to believe. It is true that the abandonment of total import substitution at the end of the 1950s was an important milestone in the modernization of Taiwanese industry. The Taiwanese miracle could scarcely have occurred without the great export successes, and the latter would not have come about had not trade policy been reformed in the direction of greater openness. In some respects the government has indeed relied on a relatively undistorted trade regime. The exchange rate has been kept stable and undistorted; there has been no discrimination against exports; wage levels have been determined by the market; and real interest rates have been positive. Effective exploitation of Taiwan’s comparative cost advantages in external markets has undoubtedly been an important element in expansion. However, the unlocking of trade policy has taken place gradually, and liberalization of imports over the whole spectrum has not been on the agenda until the late 1980s. There is innumerable evidence of the way in which Taiwan has discriminated against potentially competing imports with the aid of tariff barriers and import controls which have been eased only gradually and reluctantly (Jacoby 1966, Werner 1985, Wu 1989a,b).3 Right through to the 1980s, indigenous industries were being protected effectively by means of import duties. The average nominal customs tariff on imported goods was as high as 55 per cent up to 1974, after which date it was gradually reduced to just over 20 per cent in the late 1980s (Wu 1989a,b). Although it is true that Taiwan’s effective rate of protection was considerably lower than in other LDCs it is also true that until the 1980s trade policy bore more resemblance to an import-substitution regime than to a free trade regime. In fact, true trade liberalization is a recent phenomenon. Export promotion, which took the form of tax and customs exemptions for firms with export-orientated business, can therefore be regarded as an exception in a predominantly protectionist trade policy. The other side of Taiwan’s trade policy is therefore that it effectively

157

ASIA

managed to prevent its comparative disadvantages from being exposed. The enormous trade surplus accumulated over time would scarcely have come about without this partially protectionist trade policy. Fifth, the free marketeers idealize the accuracy and timing of government economic policy. Everything the government has done has been ‘right’: it has all been done at the right moment and in full conformity with an ideological conviction of the necessity of trade liberalization (Fei and Ranis 1979, Kuo et al. 1981a, b, Kuo 1983, Scitovsky 1986). But this is scarcely more than an ex post facto rationalization. Results are being confused with intentions. As we shall show, there has not really been any total consensus on the utility and necessity of liberalization. The gradual liberalization has been enforced in the course of a struggle between different fractions inside the political and administrative leadership and under strong pressure from outside (Haggart and Pang 1989). Moreover, the results of the policy have not always accorded with its intentions. It cannot therefore be asserted with certainty that the Taiwanese government’s economic policy has been optimal always and in every respect. It is even possible that economic progress might have been faster had liberalization been carried through at a swifter pace and over a wider spectrum. Sixth and finally, although Taiwan is rightly held to be a classic example of growth with equity it is not clear what the concept of ‘growth with equity’ stands for. Does it mean that growth itself created equality of distribution or does the latter result from direct measures of redistribution policy? In general redistributive policies have not been applied in Taiwan. The tax system, the most important instrument of income transfer, does not seem to have played any significant part in changes of income distribution in any direction (Ranis 1978). Therefore the more equal distribution appears instead actually to have been created via the ‘trickle-down effect’, which also seems paradoxical since economic growth in most underdeveloped countries appears to lead to an increasingly lopsided income distribution, at any rate during periods of rapid industrialization (Kuznets 1955). If the trickle-down theory is correct it accords with the view that a minimum of market regulation has led both to a higher rate of growth and to improved distribution. This might seem paradoxical in light of the fact that the state has not given private enterprise free rein. In reality the economy is hedged about with a multitude of regulations (covering for example finance and credit, the establishment of businesses, import permits and foreign establishments) which, if they had been fully complied with, would certainly have had an influence on growth opportunities as well as on distribution. From this would follow either that redistributive interventions have not been effectuated or that the trickle-down effect is non-existent. This type of discussion, however, leads us only into blind alleys. The Taiwanese success story is not an either-or case of market-led or stateled growth; it is a case of both intervention and free market mechanisms. With respect to distribution a more reasonable interpretation would be that the trickledown effect has operated in Taiwan because some initial redistributive measures

158

THE TAIWANESE SUCCESS STORY

were taken by the government. The important thing then is to deter-mine which of the interventions were effective in this sense. An important concept in this context is ‘equality of opportunity’, which means that the politico-economic structure of the nation is sufficiently homogeneous for a majority to have been afforded the opportunity of participating in economic life as producers and consumers. Normally this presupposes a minimum of distortive interventions in economic life. A factor that helped to create equality of opportunity in Taiwan was that the state’s regulations were at any rate less discriminatory vis-à-vis private business than is customarily the case in other developing countries. Perhaps the tax system does not have any direct redistribution effects, but what is equally important is that it does not discriminate against small and medium-sized firms as happens in other newly industrialized countries, including South Korea. Here too, therefore, the principle of equality of opportunity applies to a greater degree than in other underdeveloped countries. However, it is important to point out that even though the concepts of ‘equality of opportunity’ and ‘equality of income’ are different, they are causally connected. If there is to be a measure of equal distribution of opportunity, some degree of equality of income is necessary when the growth process starts. In order to create equal opportunities, therefore, some form of redistribution of economic resources is usually needed (Senghaas 1985). In Taiwan, income distribution was already relatively equal in the 1950s, mainly as a result of the implementation of a comprehensive land reform. As we shall see, the land reform contributed actively to industrialization, for example by diffusing economic incentives down to the lowest strata of society. The reform equalized incomes in the rural areas and created an equality of opportunity forming the foundation of the developmental process which can be characterized as growth with equity. THE DEPENDISTAS If the neo-liberal theory has idealized and misrepresented the picture of the Taiwanese miracle in one way, the calumnies of which adherents of the dependency school have been guilty are no less notable. The eagerness to deny every positive feature of this ‘bulwark against communism’ has been so great that gross distortions of the truth have sometimes been perpetrated. To begin with, Taiwan is far from being the biggest beneficiary of American aid in Asia. The fact is that the correlation between scale of aid and rate of growth is negative in East and South-East Asia. With the exception of South Korea, the biggest beneficiaries are those countries which have had the poorest economic development (Rana 1987). During the 1970s and 1980s Taiwan received the least development aid of any country in the whole of East and South-East Asia (Rana 1987). During the 1950s, on the other hand, American aid was considerable (Ho 1978). In certain years it amounted to sums corresponding to half the national product. American aid was indeed very important during the first half of the

159

ASIA

1950s. Most of the aid was earmarked for military purposes, which undoubtedly helped to consolidate the Kuomintang’s power position in Taiwan. For the rest, American aid was focused on educational purposes, expansion of the infrastructure, and agricultural reform. The aid programme also helped to consolidate the orientation towards public investment and state enterprise during the 1950s. All this was of great importance in shaping the Taiwanese politico-economic ‘model’, especially with respect to the role of the state in the economy, but it cannot explain the long-term expansion. For even more significant is the fact that direct American economic aid began winding down from the end of the 1950s and ceased altogether in 1965. The acceleration of development thereafter can hardly be explained in terms of aid. As we shall argue later on, it is more probable that the withdrawal or threatened withdrawal of American aid had a greater long-term effect on the economy than the actual existence of the aid. The termination of aid enforced an economic liberalization which subsequently brought a gradual diminution of the state’s direct involvement in the economy. Second, the equal distribution of incomes in Taiwan is a dilemma for adherents of the dependency school. Since by definition capitalism is unequal while socialism is equal, it surely cannot be true that income distribution is as equal as or more equal than in socialist countries. But this is a brute fact just the same. Comparisons made with Cuba show that with respect to distribution Taiwan can measure up well to this flagship of Third World socialism. To this it may be added that since Taiwan has a much higher per capita income it is obvious that the poorest classes in Taiwan have a higher material living standard than the worst-off Cubans. Third, the proponents of dependency theory exaggerate Taiwan’s foreign dependence in one important respect, namely the role of direct foreign investment. Taiwan has not been such an Eldorado for multinational companies as is often believed (Amsden 1985, Hamilton 1983, Ranis and Schive 1985). Over the entire period 1951–88, direct foreign investment did not account for more than about 5 per cent of capital formation, which is most readily described as extremely modest in the perspective of an underdeveloped country (Wu 1989b, Rana 1987). Up to 1965 the government maintained de facto an unsympathetic attitude to foreign investment. Since the start of Taiwan’s industrial progress must be dated prior to that year this means that direct foreign investment had no significance at all during the critical initial phase of industrialization (Ranis and Schive 1985). During the 1970s and, in particular, the late 1980s the influx of foreign capital increased, but it is still only in certain branches such as the electronics industry that this has had any critical significance for production and employment. Fourth, the dependency theory holds that since Taiwan’s economic progress has been founded on the exploitation of comparative advantages in labourintensive light consumer goods industries, the economy will remain caught in a labour-intensive trap (Frank 1982, Hammer 1984). The economy will be

160

THE TAIWANESE SUCCESS STORY

dependent on foreign technology, and its capacity to upgrade the industrial structure technologically by its own efforts will be insufficient. This view is based on the assumption that Taiwan’s economy will be unable to adapt when the comparative advantages are changed. Even though there is some justification for the allegation that Taiwan’s industry is dependent on imported technology and is hardly renowned as a technological innovator, its ability to upgrade its technology and restructure its industry still ought to be very considerable. This is partly because of the government’s deliberate efforts to transform industry and to encourage businesses with higher added values. As has been observed, stateowned enterprises play a not insignificant role in this. Another important factor, however, is that the Taiwanese economy, unlike the South Korean, is diversified. It has many strings to its bow, which in itself means there is a good basis for expansion. The high educational standard is a further factor of significance. The restructuring of industry is intrinsically a problem for Taiwan in the 1990s and it is not one which will solve itself, but there is nothing to indicate that Taiwan is any worse equipped than others in this respect (Fransman 1986). Some recent studies have indicated that Taiwan’s industrial restructuring and technological upgrading is well under way (Schive 1991). Fifth, the dependency theorists contend that Taiwan is a special case which cannot constitute a model for successors (Fernandez 1980, Frank 1982, Cline 1982). This is based on the unrealistic assumption of immutable comparative advantages. This is a view which is both falsified by the facts and remarkably inconsistent with the rest of the argument. Taiwanese industry’s need for restructuring would not have arisen if the comparative advantages had not changed to the detriment of Taiwan and the benefit of other Asian ‘newcomers’. Industrialization is proceeding apace today in a number of other Asian countries. Perhaps development is not following a Taiwanese ‘model’, but Thailand, Indonesia and China have made important inroads in the American market for textiles, for example, which traditionally has been Taiwan’s domain. Thus Taiwan is not a special case in terms of the ability to exploit comparative advantages. Finally, there emerges from the simplest variant of the dependency theory a misleading picture of the role of the state in the economy. Entirely in accordance with the rest of the thesis of extreme foreign dependence, the state is held to be a guardian of the interests of multinational companies and a lackey of the United States (e.g. Gates 1979). By laying too much stress on the dominance exercised over the economy by foreign companies while pointing at the same time to the provisional and repressive character of the KMT government, it is easy to conjure up a picture of a state apparatus in the service of foreign interests. Such an ‘analysis’ does not supply any real knowledge either of the character and role of the state in the economy or of Taiwan’s possible dependent relationship vis-à-vis the world outside. All in all these attempts at explanation illustrate the intellectual collapse of the dependency school. It has led to conceptual hair-splitting instead of enhancing our understanding of the issues under debate. If dependency is to be a meaningful

161

ASIA

analytical category, it must be specified in concrete terms. To say that Taiwan is less dependent than other underdeveloped countries is of course nonsense in ordinary terms. It is difficult to conceive of any greater foreign dependence than that under which Taiwan has laboured both politically and economically. Taiwan has not been able to shape its own model in total isolation, independently of foreign interests and pressures. On the contrary, the country’s extreme dependence on the world outside is an important explanatory factor in itself. On the other hand there is truth in the notion that dependence in the more specifically economic sense has been less than in the case of other underdeveloped countries. The existence of a strong state power has resulted in part from the fact that the safeguarding of private foreign special interests has not been the main objective of the policy vis-à-vis Taiwan. In this sense, then, Taiwan is a special case of relative independence despite its enormous dependence on the outer world in other respects. The conclusion must be that it is meaningless to speak of dependency as a concept which can be determined in absolute terms in some sort of quasi-quantitative sense. THE STATISTS As will have become apparent earlier in this account, the question of the state’s role in the economy is not unimportant. On the contrary, it needs to be considered in detail if we are to arrive at a reasoned understanding of Taiwan’s economic development. But in such an analysis it is not enough to pin one’s faith to ideological convictions of whatever tendency or to ready-made theoretical schemes in which the Taiwanese actuality loses real significance among the strivings to demonstrate the superiority of the theory preferred. Modern works on the Taiwanese miracle also lay stress on the active role of the state in a way quite different from what has been the case previously. The state is regarded neither as the tool of a bourgeois capitalist class nor as a lackey of imperialism but rather as an autonomous entity vis-à-vis civil society including economic interest groups and classes (Amsden 1979, 1985, Barret and Whyte 1982, Crane 1987, Cummings 1984, Evans and Pang 1987, Gold 1986, 1987, 1988, Gunnarsson 1987, Haggart 1986, Haggart and Cheng 1987, Haggart and Pang 1989, Hamilton 1983,1987, Johnson 1981, 1986, Wade 1984, 1988a,b, 1990, White and Wade 1988, Wynn 1982). These analyses have had different starting points and have focused on a variety of features of the state-society relationship. Some authors, such as Evans and Pang, Johnson, and Wade and White, have laid stress on special institutional features within the East Asian politico-administrative power apparatus. Others, such as Haggart et al. and Hamilton, have taken a special interest in the relationship between classes/groups within civil society (Lauridsen 1990). This research has indisputably resulted in a number of very interesting studies which have overthrown the more simplistic hypotheses of both free-tradeism and dependency theory. Even traditional liberal economists are increasingly critical

162

THE TAIWANESE SUCCESS STORY

today of the conception of Taiwan as an ideal model of laissez-faire capitalism (e.g. Chen 1989, Ranis 1989, Bradford 1986). But like most other economists they have assumed the existence of effective institutions rather than actually analysing their evolution and role in the economy. The increasingly prominent place given to political and institutional factors in the analysis of economic change is indeed to be welcomed. All the same there is a risk that too intense a focus on the state and on economic policy as objects of study may exaggerate the state’s significance while the dynamics of private business and industry slip into the background. There is especially a risk of forming a lopsided picture of the interplay between the dynamics of the market economy and the processes of political and institutional change. All economic change may come to be regarded as the result of more or less deliberate manifestations of political will and strategies. From such a perspective the state becomes not just an autonomous and leading actor but, de facto, the very determinative actor. Not infrequently, therefore, the judgements made of the state’s role in the economy become just as positive and laudatory in tone as in the ultra-liberal perspective. The difference is that while the liberals emphasize liberalization and deregulation as the engine of growth, the state theorists attach greatest importance to the active state’s regulatory policy or to co-operation between the state and the business and industrial world. There is a risk here of overstressing the state’s regulatory role and, especially, of noticing only the positive aspects of state intervention while overlooking possible negative effects, A serious analytical problem facing anybody who wishes to study the role of the government in economic performance is that the object of study itself, i.e. the state, is extremely apt to issue statements about its own role. Such statements are often confusing and not seldom contradictory. Taiwan’s political leaders are normally anxious to point out that they advocate market economy, competition and free enterprise. However, the actual content of these concepts appears to be of less importance. The concepts are rather used as slogans and as ideological window-dressing to westerners. The same leaders seldom hesitate to speak of Taiwan as a democracy in spite of the fact that society was under martial law for almost four decades. Today, trade liberalization has become a catchword although few political leaders would accept or even understand the principle of comparative costs (Chou 1991). The political elite is also eager to emphasize its own leading role in the modernization of Taiwan. In this context it is pointed out that the economy has been efficiently government directed and planned. Small wonder then that the interpretations of the Taiwanese miracle diverge even when they are derived from the same source of information. The special point about Taiwan is the very combination of a strong and autonomous state apparatus with a flexible and expansionist free enterprise sector. It is the dynamics of the interaction between these two levels and categories of actors that need to be analysed in detail. Otherwise it is easy to fall into a sterile state-versus-market polemic.

163

ASIA

What do we mean, then, when we say that the state is autonomous? A state apparatus can be the instrument of a certain class or group, but it can also be an independent actor protecting its own interests, which are different from the interests of the dominant class. Yet again the state can also be a ‘third party’ mediating between a variety of conflicting interests in the economy and/or a guardian of common national interests. Irrespective of which role the state plays, it can never be regarded as autonomous in the sense of being a monolith disengaged from social and economic, internal and external processes and conflicts. The state is a part of society whether it is an instrument of a particular class/group or whether its character and actions are to be regarded more as a reflection of social processes and conflicts. The state’s structure, its motives for action, its power instruments, its legitimacy and its effectiveness are all functions of a specific social context. When we speak of the state as an autonomous phenomenon, therefore, this means that it is not the direct instrument of a specific social class or of other special interests. Then is an autonomous state in this sense to be regarded as a special case or as a necessary condition of the functioning of an economic system? To say that all economic systems need a functioning state power and effective institutions is a truism since institutional efficiency per se is a measure of economic efficiency. More specifically, however, a market economy without functioning game rules is doomed to degenerate into a racketeer economy. Thus a strong state power is needed in order to set the rules of the game and to ensure their enforcement. To be able to perform these tasks effectively in the long term the state always needs to be autonomous in the sense that it must not degenerate into a tool of private special interests because these will endeavour to guard positions once attained, which impairs the flexibility of the economy in the long run. This does not mean that an autonomous state is always efficient. It can just as easily pursue a policy which reduces rather than raises socio-economic optimality. This may be the result of incompetence or of the ruling elite’s protecting its own special interests. There are extreme instances in the Third World where the state apparatus has been wholly privatized by a small autonomous power elite which uses its monopoly of power to fortify its own position politically and economically (Lundahl and Vedavato 1986). Furthermore, a strong or autonomous state does not necessarily have to intervene directly in the economy. State intervention is more the rule than the exception in the Third World irrespective of whether the economic system is called capitalism or socialism. In the vast majority of cases the state’s involvement has been directly detrimental to the economy. Regulation has distorted the allocation of resources and state and private monopolies have arisen, while economic incentives have been withheld from the great mass of potential economic actors (World Bank 1987). But these state apparatuses are seldom autonomous in our meaning of the word. On many occasions the state has been the tool of different kinds of ‘distributional coalitions’ which penetrate the state apparatus for the purpose of their own gain (Olson 1982). Thus regulation is

164

THE TAIWANESE SUCCESS STORY

conducive to rent-seeking behaviour and non-productive profit maximization (Krueger 1974, Bhagwati 1982, Srinivasan 1985). In point of fact rent-seeking or monopolistic behaviour is an inherent tendency of the private sector. However ardently and frequently entrepreneurs may declare their enthusiasm for free competition, there are few who will forgo a chance of appropriating monopoly profits or exploiting special privileges. In every economy the state power has an important role in preventing monopolization and rent seeking. At the same time it is the very existence of the state power that creates the opportunity of reaping advantage from privileges, monopoly rights and secure rents. Therefore the state can easily degenerate into an organ which protects private vested interests, which in turn guarantee the continued existence of the political elite and the bureaucracy. Therefore the liberalization needed in many underdeveloped countries is a virtual ‘de-privatization’ of the state. Accordingly it is not to be wondered at that the countries where free enterprise has been given the widest scope are also the ones which have enjoyed the most auspicious economic development. But free enterprise has not flourished in the absence of a strong regulatory state power. On the contrary, expansion of the market economy has been dependent on a strong state which both guaranteed compliance with the game rules of the economy and avoided being overly penetrated by special economic interests. Taiwan is a striking example of a state which is autonomous in this sense. The state in Taiwan is scarcely more interventionist than the state in Latin American countries, but it is indisputably more autonomous. National economic development has always been the main purpose of state action, but this purpose has been brought to fulfilment via the market and freedom of enterprise. The state has been the generalissimo of development, but private enterprise, meaning chiefly the small and medium-sized firms, has been its army. The discussion so far has touched on the autonomous state and its necessity in a dynamic market economy. What still remains, however, is to explain how the Taiwanese state was able to achieve an autonomous position and why this autonomy was used in a manner that was effective and conducive to the development of the economy. But we cannot present any universally valid theory capable of answering these questions. Neither is such a theory really necessary. Each specific historical case must be analysed in terms of its own particular circumstances. Either any general theory will be so general that in order to explain each individual case supplementary explanations will still have to be invoked, or else one will finish up by making purely tautological statements, e.g. that the state is autonomous because it has not been penetrated by special interests. Therefore when discussing the reasons for the rise of an autonomous state power in Taiwan and its relations with the market economy it is of importance to consider a number of questions concerning the state’s character, motives and particular actions in their specific historical context. Some questions stand out as central to an understanding of the special relationship between the state and the business and industrial world in Taiwan. To begin with, how is it that the

165

ASIA

government party, the KMT, which achieved notoriety as a predatory regime on the Chinese mainland, became transformed into an autonomous developmental state of gigantic dimensions? Second, why did a regime which by tradition was sceptical towards capitalism and free enterprise place its trust to an increasing degree on those very forces instead of relying on increased interventionism? Third, what factors have been at work in Taiwan to prevent the strong state power from having a stifling and stunting effect on the economy as is so frequently the case in other underdeveloped countries where state interventionism is effectively an instrument of kleptocracy and plunder? These questions lead on to a concluding discussion touching the question of whether Taiwan can be regarded as a model for economic development in other underdeveloped countries. We shall now go on to discuss these questions in the order specified. THE KUOMINTANG AND MODERNIZATION If the prospects of economic development in Taiwan looked very gloomy at the end of the Second World War, this was not merely the result of the effects of the war. What was equally disquieting was that in 1945 Taiwan became a protectorate under the Chinese nationalist regime, the Kuomintang (KMT). That a regime which had so manifestly failed economically on the mainland and which had developed into perhaps the most notorious plunder-regime of its time should possess both motivation and capacity to lead the rebuilding and modernization of the Taiwanese economy must have appeared almost totally illusory to contemporaries. Matters were hardly improved, moreover, by the fact that the party was headed by Generalissimo Chiang Kai-shek, who lacked economic training and was more at home in the conservative, anti-modernization wing of the KMT. But strange are the ways of history: the miracle happened. Shortly after the KMT had established its headquarters in Taipei following the defeat at the hands of the communists in 1949, a programme of land reform was introduced which has no counterpart in the Third World in terms of consistent implementation and effectiveness. The concentration of land in Taiwan was not as palpable as in China, but about half of the agricultural workforce consisted of landless labourers or tenant farmers on sharecropping contracts (Ho 1978). First of all a rentreduction programme was introduced, then the land which had been confiscated from the Japanese was sold off, and lastly, in 1953, a ‘Land to the Tiller’ programme was initiated which in effect signalled the final elimination of the class of great landowners. The land reform was perhaps far the most important institutional reform associated with the creation of the special Taiwanese ‘model’. It not only brought about a redistribution of existing incomes but above all it created better and relatively equal sources of livelihood for the large and growing agrarian population. The land reform was a necessary prerequisite of the modernization of agriculture, whose increased production laid the foundations of industrialization. Modernization not only gave bigger harvests enabling the rapidly growing popu-

166

THE TAIWANESE SUCCESS STORY

lation to be fed. It also generated incomes which could be transferred to industrial purposes and helped to provide surplus labour for industry. No one denies the contribution of agriculture to the early industrialization. But it is possible to hold differing opinions regarding the dynamic relation between the transformation of agriculture and the growth of the industrial sector. One commonly held view lays emphasis on the extractive aspect, i.e. agriculture’s role as a source of capital and reserve of labour (Amsden 1985, Apthorne 1979, Ka and Selden 1986). It is beyond all doubt that industrialization after 1965 was attended by a great exodus of labour from actual agricultural production. It is also quite clear that tax and price pressure on the agricultural sector was heavy during the 1950s. Taxes on agriculture were considerable and agricultural prices fell relatively to industrial goods. These factors in combination made it possible for large capital resources to be transferred from agriculture to the state and the industrial sector. This picture of agriculture as a source of capital accumulation and labour supply in the industrial sector is correct on the whole. It is also fully in line with the view that Taiwan’s industrialization was export led as a necessary consequence of a limited home market. But if this were the whole truth about the success story it would really be justified to speak of a Taiwanese ‘miracle’. The question is whether economic history shows any examples of an industrialization process, with equal or improved income distribution taking place simultaneously with the exploitation and ‘punishment’ of the agricultural sector. In such instances inward-looking strategies have usually foundered because of the limited size of the home market, while export-orientated strategies have reinforced regional, structural and income imbalances in the economy. Instead, a central factor in Taiwan’s industrialization was that modernization of agriculture put increased incomes and higher purchasing power into the hands of the rural population, thereby enabling it to form a market for industrial goods (Oshima 1986). The question is how this could happen when agriculture was being so heavily exploited at the same time. The answer lies in the increased productivity resulting from the restructuring of agriculture. The land reform did not involve the replacement of the semi-feudal system of cultivation with a capitalistic form of organization, i.e. highly mechanized farming with large cultivating units. On the contrary, before 1965 cultivation units were extremely small, around 1 hectare or, more often, less. This could be thought to have meant the preservation of unmechanized, low-productivity farming, but this was of less importance in the early phase of industrialization, chiefly because the land reform was accompanied by a series of institutional changes at the local, regional and national levels which brought positive effects in terms of diffusion of technology, credits and infrastructure. American aid played an important role in this. Perhaps the most important institutional change was the organization of the socalled Farmers’ Associations at the local level (Oshima 1986). A Farmers’ Association was a sort of independent financial association with local management, which organized credit and marketing and helped with the introduction of new

167

ASIA

technology. Through the land reform and its accompanying institutional reforms the marginal cost of land was reduced, which facilitated investment in new technology. Moreover, the incentives reached down to the producers at the bottom level instead of being confined to the big landowners as before. The organizational, technological and institutional changes in agriculture gave rise to a surplus of labour which it was possible to transfer to the industrial sector. In the initial phase this was an important stimulus to the growth of rural-based small-scale industries, which had ample access to cheap labour. However, these rural industries were closely associated with agriculture. Despite the fact that some of the rural population left primary agricultural production, this does not mean that they left the land, or their families, in the physical sense. By virtue of the fact that sons and daughters of peasants continued to form part of the rural household the incomes brought in through work outside agriculture accrued to the agricultural household (Fei and Ranis 1979). This enabled the households’ purchasing power to be strengthened despite the heavy taxation and the reduced relative prices of agriculture. In other words, the strengthened purchasing power of agricultural households was able to form an important market for the emergent rural-based small-scale industries. The rise of these small rural industries, so typical of Taiwan, can hardly be explained in any other way. Thus we see that redistribution of the land was an important factor in creating equal opportunities. In this respect Taiwan is indisputably a special case. In most other underdeveloped countries which initiated rapid industrialization, inequality of opportunity, and of access to financial resources, was extremely high right from the start. In these cases industrialization served to reinforce inequality and social cleavages more than anything else. Thus the ‘trickle-down effect’ did not function, since the distribution of incomes became more and more lopsided as per capita income increased (Kuznets 1955, Adelman and Morris 1973). In Taiwan, where financial resources were more equally distributed because of the land reform, opportunities for the individual were greater from the very beginning, and this caused Taiwanese industrialization to take on a propitious shape. We can really speak here of ‘balanced development’, i.e. an industrialization that took place interactively with the modernization of agriculture.4 This was especially important during the initial phase of industrialization. In later days when the industrial sector had become self-sufficient, more highly localized in urban centres and more dependent on exports, the relative significance of agriculture as a reserve of labour, a source of capital and a market gradually diminished. From the early 1960s onwards the agricultural sector was increasingly faced with a labour shortage, which necessitated technological improvements and an increase in the size of cultivation units. Then why did the KMT carry out a land reform in Taiwan? One explanation may be the insight gained from the experiences on the mainland. The land reform promised but never implemented by the party had formed the best breeding ground for the peasant-based communist revolution. This experience must have carried much weight when the party was planning an economic strategy the over-riding aim

168

THE TAIWANESE SUCCESS STORY

of which was to fight the spread of communism. Perhaps the phenomenon of political leaders learning from the mistakes of themselves and others is not very common, but the situation in Taiwan did develop in a very special way. The first provisional KMT government which took over after the Japanese defeat in 1945 certainly developed very quickly into a predatory regime of the worst kind. Japanese assets were confiscated and transferred into state ownership. However, these state-owned enterprises were run as private firms in the negative sense that their assets were stripped out and transferred into private hands. Nepotism and discrimination against native Taiwanese businesses through monopolistic control over transport, credit and the supply of input goods soon led to fierce antagonisms between the KMT and the local Taiwanese population. In February 1947 a revolt broke out which threatened to undermine the KMT’s economic and political control over Taiwan. In the short term this problem was solved through the KMT’s total military supremacy. Taiwanese political leaders were murdered, driven into exile or silenced in various ways, and thousands of other opponents were killed (Haggart and Pang 1989). Between 1945 and 1949, then, the KMT acted in the manner that might be expected bearing in mind the party’s track record on the mainland. However, when Chiang Kai-shek and his newly appointed governor Chen Cheng arrived in Taiwan the situation changed drastically. Chen Cheng was already known from China days as an ardent reformer. He had tried to bring about certain land reforms in Hupeh province in China, and he was the one leader who saw most clearly, perhaps, the link between economic development, political power and legitimacy (Haggart and Pang 1989). It was already clear by 1949 that in the long run the KMT’s power position in Taiwan could not be sustained by military means alone if at the same time the regime lacked legitimacy among the Taiwanese. The best means of acquiring legitimacy was the practical implementation of the KMT’s programme, in which land reforms occupied a key place. In other words it was time to go over from political rhetoric to practical action. Another explanation of the reform policy may be that for the first time in the parly’s history the KMT had now achieved a position of power from which its party programme could actually be put into effect. The KMT’s power position in China had been undisputed only in Nanking province. The rest of China had been controlled during the 1920s and 1930s by local warlords with whom the KMT cooperated but over whose policy it lacked real influence (Gold 1987). The KMT therefore fell de facto into alliance with the landowner class, whose elimination had been promised in the original party programme. In Taiwan, the instruments of power were infinitely bigger. The KMT had full military control over the whole island and had no historical ties at all with the local landowner class (The Taiwan Development Experience 1985). The elimination of the latter, therefore, was not a particularly controversial question on either ideological or practical grounds. To this we may add that the class of big landowners was not specially large in number, nor did it have full control over the peasant population. This was because the Japanese, during the colonial era, had already implemented reforms which

169

ASIA

had strengthened the position of peasant proprietors and tenant farmers (Ka 1987).5 These causes, i.e. the political and military power monopoly on the one hand and the necessity of reform on the other, combined to convert the KMT from a parasite to a development-orientated party. THE KUOMINTANG AND CAPITALISM Taiwan’s economic success, as everybody knows, has been achieved within the framework of a capitalist economic system. Even though this is self-evident, there is an equally self-evident paradox in the fact that the KMT’s party programme contains strongly anti-capitalistic elements. It is not partiality towards state intervention as such that is difficult to understand in Taiwan but rather how it has come about that a party whose interest in free enterprise was traditionally lukewarm could be so successful in fostering the growth of capitalism. The KMT’s ideological roots are to be found in the ideas of Dr Sun Yat-sen, most notably in his book on The Three Principles of the People (national development, democracy, prosperity of the people). The third principle is the controversial one in this context. On the one hand it can be interpreted as ‘prosperity of the people’ but it could also mean ‘socialism’. Fundamental elements of Sun’s thinking were his stress on equality, his aversion to concentration of economic power, and his conviction that capitalism must be regulated or limited. This ideological heritage is hardly the most appropriate one to claim the allegiance of a regime said to pin its faith on free market forces (Sun 1924). However, there are contradictions in Sun’s ideas which give scope for pragmatic interpretations. For he also says that the power of the state should be limited if possible, and that functions which can be performed better by private enterprise should be placed in private hands and protected by law (Haggart and Pang 1989). This dualism in the view taken of the state-market relation is very apparent in the KMT’s practical policy. The development strategy of the 1950s undoubtedly placed its emphasis on the restrictive features, on equality and planning, which found expression in the land reform and the total concentration on import substitution. Even during later decades when import substitution was gradually being abandoned and free enterprise allowed a looser rein, we find, as noted earlier, strongly interventionist elements in the economy. The most interesting phenomenon, however, is the change of economic policy signalled by the switch to export promotion. Even though deregulation of imports has been slow there is no doubt that the process initiated in 1958 opened the way for an increasing emphasis on free enterprise and private initiative. What we see at this point in time is a shift in the KMT’s strategy. In the official written history this shift is portrayed as a manifestation of the government’s ‘realizing’ that the possibilities of import substitution were ‘exhausted’ and therefore ‘deciding’ to go over to an export-led strategy. It is possible and even probable that the realization of the limitations and

170

THE TAIWANESE SUCCESS STORY

drawbacks of import substitution did influence the reform decision. The pace of growth had been slackening since the middle of the 1950s. During 1956 and 1957 consumption per capita scarcely increased at all. Industry was facing big problems as well. At the same time as the degree of self-sufficiency increased to over 80 per cent in such branches as textiles, timber products and the cement and petroleum industries, dependence on imports had risen with respect to machinery and transport equipment (Lin 1973). This had led to a substantial deterioration of the balance on current account and to problems of foreign exchange reserves. It was also obvious that import substitution had favoured indigenous monopolists in a way that was harmful to the national economy. Inefficiency, corruption and rentseeking behaviour generally were phenomena of common occurrence (Yin 1954a,b). So far the KMT government showed few signs of being more efficient or more autonomous than most other contemporary LDC governments. It would be naive to believe that the change of politico-economic course took place in a condition of total consensus or free choice. The Taiwanese leadership class was split into two wings at this time, one of which adhered doggedly to regulatory ideas and took a negative view of private enterprise while the other was more development orientated and more pragmatic in its view of the character, aims and possibilities of state intervention. The conservative wing did not merely express an ideologically rooted mistrust of the business capabilities of the private entrepreneurial class. It was also allied with established industrial interests which were afforded special privileges by protectionism. There was also an element of ethno-political antagonism here. Most of the entrepreneurs who received special privileges under import substitution were immigrants from mainland China. The local Taiwanese entrepreneurial class was weak and a prey to discrimination. The fear of liberalization therefore reflects an aversion to permitting any strengthening of the latter group’s economic position, since this could have political consequences (Haggart and Pang 1989). The eventual victory of the more change-orientated and liberal wing did not result primarily from its having managed to persuade its opponents of the expediency of the reform policy. What was much more important was that the liberal wing succeeded in obtaining the support of the top leadership of the KMT, notably that of Chiang Kai-shek himself. The reformist wing had its foremost actors among the younger American-educated ‘consultants’ such as John Fei, Anthony Koo and Gregory Chow (Evans and Pang 1987). These had been trained at American universities, where they acquired a more liberal, proentrepreneur outlook than was current among the KMT. Alongside these there were American economic advisers with a broadly similar view of how Taiwan should be modernized. Thus the reformist wing was very strong, but its exertions would still scarcely have succeeded had it not managed to get the ear of Chiang Kai-shek. In order to do this direct American pressure was needed. The American government’s attitude to the KMT was also undergoing a change at this time. The primary purpose of American aid, of course, was to prevent a communist assumption of

171

ASIA

power. Initially this meant that the aid was directed to military purposes. During the late 1950s, however, the American approach was gradually modified. The US government explicitly declared its unwillingness to help in any military reconquest of China. It was the development of Taiwan that was the goal instead. In this connection the strengthening of the economy became the primary instrument of development. The United States intended to develop Taiwan as a showcase of capitalism, as a bulwark against communist expansion in East Asia (Jacoby 1966, Haggart and Pang 1989). The American interest was therefore of a broad strategic nature. But there were no ties with private American economic interests that had to be protected. The most powerful means of bringing pressure to bear, of course, was the declaration of intent to change the alignment of aid along with the imposition of special conditions for continued economic assistance. Thus the paradox is that it was not American aid as such that underlay the Taiwanese miracle but rather the direct threat of its withdrawal and its final termination. Also in later states of trade liberalization external pressure appears to have been a decisive factor. This is true for the liberalization in 1972–3 which was necessiated by the diplomatic isolation facing Taiwan after Nixon’s visit to Peking. It is even more apparent for the liberalization beginning in 1986–7 when the US governemnt threatened Taiwan with open trade war (Chou 1991) THE KUOMINTANG AS A DEVELOPMENTAL STATE In broad terms the view of capitalism and free enterprise taken by the KMT may be said to have become gradually more positive. But this does not prevent the regulatory element from still being strong, perhaps even too strong. As has been remarked, one cannot simply accept uncritically the thesis that the Taiwanese economic miracle would have been impossible without the strong stateinterventionist element. Indeed it is quite possible that state regulation has to some extent also impeded free enterprise. A not uncommon proverbial saying in Taiwan is that ‘the economy has succeeded despite the best efforts of the government’ (The Taiwan Development Experience 1985). But it does seem to be a fact that the element of regulation has not had directly harmful effects. Here we find one of the most interesting aspects of the Taiwanese political economy. What has to be explained, first, is how the state has been able to maintain a strongly autonomous position vis-à-vis private economic and political interest groups. Second, one may wonder how the market economy has been able to be so expansive and dynamic in the face of such large-scale state intervention. The Taiwanese state has largely managed to avoid being an arena for rentseeking distribution coalitions (Gold 1987). This does not mean that the danger of degeneration has not been present. Without doubt the potential for emergence of a privilege state was enormous during the protectionist era of the 1950s. There was just as great a risk as in other underdeveloped countries that the state would become an organ of private vested interests (Yin 1954a,b). The reform policy of

172

THE TAIWANESE SUCCESS STORY

the late 1950s therefore undoubtedly stands out as a turning point in Taiwan’s economic history. By withdrawing support from certain monopoly interests and laying them open to competition the state succeeded in averting the emergence of a home industry with feeble technological dynamics and poor international competitiveness. The political base of these entrepreneurial interests was also snatched away. Since then the gradual liberalization effected chiefly through the enforced reforms of 1972 and 1987 has made rent-seeking behaviour less and less lucrative. Equal opportunities have come to mean more than special rights. Taiwanese practice has thus come de facto to live up more and more to the precepts of liberal economic doctrine. The paradox, however, is that this has not happened in the absence of a regulatory state power but on the contrary could scarcely have happened without a strong autonomous state power that refused to be the errand-boy of particular entrepreneurial interests. How has the Taiwanese state been able to become so autonomous? This cannot really be understood unless due weight is given to Taiwanese society’s functional division of labour and power between political and economic elites, between mainland Chinese and indigenous Taiwanese, between technocrats and the private economy, and between military and civil forces (Evans and Pang 1987). An important factor underlying the division of power is that the indigenous economic elite has been excluded from political influence to a large extent because it is Taiwanese and lacks roots in the KMT. The political elite, on the other hand, consists of persons and groups linked to the KMT, which in its turn traditionally has little part in economic activity, except of course for its dominance over the state enterprises. This means that the great mass of entrepreneurs have been denied political rights in exchange for a considerable degree of freedom in the economic sphere. On the opposite side political, military and administrative power has been a sufficient incentive in itself to prevent politicians, bureaucrats and the military from interfering in the economy for their own gain. Broadly speaking there is little mobility between the different spheres of activity. Just as entrepreneurs do not involve themselves in politics the military remain military and the bureaucrats bureaucrats. It is paradoxical on the face of it that the political elite encouraged the growth of an indigenous and potentially oppositional entrepreneurial class while at the same time refraining from rent seeking themselves. But this is a paradox only if one takes a functionalist approach and views the result as a manifestation of a deliberate and desirable policy. The strengthening of the Taiwanese middle class has hardly been an aim of the KMT, but it has been the result. The solution to the paradox is that this is a price which the KMT has had to pay in order to develop the economy, which was the means of maintaining its political hegemony. We have already dealt with the question of the KMT’s motives in choosing the path of development rather than of predation. As regards ability to implement this strategy as well, the distribution of power is an absolutely crucial factor. The political dictatorship is sometimes cited as decisive in preventing the negative impact of distribution coalitions on the economy. But the existence of a political

173

ASIA

dictatorship does not suffice as an explanation. For most predatory regimes are really dictatorships which have demolished and devastated the national economy for their own gain. The fact that this can happen is a result of the mingling of political, military, administrative and economic power in a self-reinforcing degenerative process on every plane. More factors than the dictatorship have been at work to prevent the degeneration of the state and collapse of the economy in Taiwan. First of all, the KMT is an elitist top-governed party organized on Leninist principles. With the consolidation of the KMT regime in Taiwan, a centralization of party power under Chiang Kai-shek was brought about for the first time, and so also was uniformity with regard to the party’s political targets and its strategy of modernization (Haggart and Pang 1989). Whereas during the China period the party was torn apart by rival wings, during the Taiwan period it has been led firmly and efficiently from the top. Rival wings seeking coalitions with civil forces in order to strengthen their own positions have scarcely existed. One might even wonder whether the centralization of political power in a homogeneous and top-governed party in Taiwan has not been a factor just as important as the existence of an autocratic state power. Second, the entrepreneurial class has been excluded from political influence because it consists of indigenous Taiwanese. Thus, an entrepreneurial class whose interests had to be safeguarded has hardly existed. In the KMT’s judgement, such a class could have constituted a serious political danger if allowed to exert influence over political life. In Taiwan there is no organized, formalized forum of collaboration between the state and the elite of the business and industrial world (as in Japan and South Korea). For those who want to exercise pressure the only available channels to the political/administrative elite are informal and indirect. Therefore Taiwan cannot be regarded as a corporative politico-economic system. Third, private enterprise has been unable to organize distributive coalitions not only because of the ban on organization but also because of the structure of the economy itself and the conditions of enterprise. Private business and industry are compelled to live in a hard competitive world where competition comes not only from abroad but also from home. Capital concentration is relatively slight too. Instead, what is typical of Taiwan is the number of small and medium-sized firms. However eager their desire to form cartels may be, they have not succeeded in building pressure coalitions for rent-seeking purposes. Fourth, there is a very special American interest in Taiwan. In contrast to many other underdeveloped countries, Taiwan has had no really important private foreign economic interests to be safeguarded. Therefore the American advisers have been able to act in accordance with textbook economic theory and their own ideological convictions rather than making it their first priority to ensure that the KMT’s policy protected private American interests. Thus the American advisers had an autonomous status vis-à-vis both local social forces amd private interests back home. Their autonomy is not unlike the autonomy of the KMT’s state vis-àvis the Taiwanese civil community.

174

THE TAIWANESE SUCCESS STORY

By dealing with the question of the autonomy of the state we have also partially answered the question of the dynamics of the market economy. The fact that the market’s dynamics have not been stifled by vested interests is largely attributable to the autonomy of the state. But this is not the whole truth. We have already remarked that despite the gradual liberalization, Taiwan’s trade policy has had more of the character of an import-substitution regime than of a free trade regime. This obviously applied primarily to the 1950s, but in high degree also to the succeeding two decades of export promotion. The ability of the Taiwanese export industries to exploit their comparative advantages is the key to an understanding of the expansion after 1960. The question is, however, how these comparative advantages could arise under a protectionist import substitution regime. According to the infant industry argument it may be justified at certain periods for a country to protect the build-up of its own industrial base capacity provided that the home market is not saturated. The capacity built up behind the tariff wall can be used in later periods for offensive drives in export markets (e.g. Myrdal 1957, Lewis 1955). In most instances this strategy does not work, because protectionist measures tend to work to the advantage of certain businesses and individual interests selectively. Costreducing technological innovations are therefore not stimulated. When the protected industry is exposed to foreign competition it collapses. Accordingly it becomes an end in itself, both for private vested interests and for governments, to prolong and enlarge the protective shield. This is a common experience of import substitution in LDCs (Little et al. 1970). Naturally there have been such tendencies in Taiwan too, especially during the 1950s, but even so Taiwan must be regarded as an exception. For the special factor in Taiwan is that industries which had been protected from foreign competition proved to be highly competitive when exposed to foreign competition. Even during the import substitution of the 1950s, industries grew up which later revealed themselves to be competitive on the first contact with foreign markets. This, then, does not chime in with the thesis that import substitution always delays and impedes the introduction of cost-reducing innovations. However, the Taiwanese model of import substitution bears a different appearance from that of most other models. In the majority of cases import substitution is implemented via the selective protection of favoured businesses and firms on a home market with a feeble growth dynamic. Selective protectionism has occurred in Taiwan as well, but import substitution has been applied on a home market that was growing vigorously first as a result of land reform and then of export successes. The home market’s growth dynamic stimulated internal competition despite the fact that foreign competitors were shut out. This took place not in accordance with the principles of free trade theory but in accordance with mercantilist principles of external protection simultaneously with internal encouragement of the free market economy. The outcome would scarcely have been so successful without modernization of agriculture during the 1950s and export promotion from the late 1950s

175

ASIA

onwards. Both these measures were conditional on the existence of a strong autonomous state power. But the relationship between the state and the business and industrial world is more complex than this. One important aspect of industrial development is the financing of industry. It is typical of Taiwan that the money and capital markets are undeveloped. Trade in stocks and bonds has been limited, even up to the end of the 1980s. Business financing has instead been effected via loans within the framework of the formal financial system (banks, investment companies, insurance companies, etc.) and through the medium of loans raised on the informal and semi-illegal financial market. Here is another of the paradoxes of the Taiwanese miracle. The state exercises strong control over the banking system, by means of which the granting of credit to private business and industry can be controlled. The great source of capital formation is the enormous scale of household savings, constituting something like 80 per cent of total savings in a country where the savings ratio amounts to about 25 per cent. The financial system’s role is to channel these savings into productive investments. Through its control over the banks the state can in this way also exercise a measure of control over investments. The efficiency of this system can be called into question, however. State-owned enterprises and large private firms have been favoured while small and medium-sized firms have been treated unfairly. Moreover, export firms have been given special treatment at the expense of non-exporting firms. This must have worked to the benefit of the exportpromotion policy, but at the same time many critics contend that the state’s control over the financial system impaired the flexibility of the economy instead. Lending, it is argued, has had a conservative and defensive character. ‘Riskminimizing’ has been rewarded more than risk taking, most notably because of the high demands for security on loans. All this does, it is true, support the thesis of the autonomous interventionist state in Taiwan, but this in itself does not explain the flexibility of the market economy. Above all it does not explain how small and medium-sized firms, which form the heart of Taiwan’s economy, have been able to cope with their financing. In order to understand this we must look at a phenomenon which stands outside the scope of state control, namely the informal credit market. The informal financial sector consists of finance companies, more or less legal, which help to circulate capital generated in the official financial sector. Estimates have shown that as much as 35 per cent of all private sector finance emanates from the informal financial sector (Shea et al. 1989). If to this we add the fact that larger private firms usually avail themselves of the official system, it seems clear that a very large part of the financing of small and medium-sized firms is accomplished without any sort of control on the part of the state.6 In other words we find here a sector standing outside the state’s patronage and control but capable even so of being flexible and expansionist. If the private sector’s inability to penetrate the state apparatus constitutes the one component of the Taiwanese state-market dichotomy, the state’s inability to penetrate and regulate large parts of the private

176

THE TAIWANESE SUCCESS STORY

sector accordingly constitutes the other and perhaps at least equally important component. Thus, with this obvious dual character of Taiwan’s economy in mind some new light could be cast on the seemingly endless state-contra-market controversy. On the one hand Taiwan’s economy is characterized by a sector consisting of fairly large and state-sponsored companies which have entered into what might be called a ‘mercantilistic partnership’ with the government.7 On the other hand we find a sector of small and medium-sized companies which operate with little or no government support. Instead of following government dictates or advice they have rather come to develop into ‘avoiders’ of the government. These companies operate in ‘traditional’ Chinese manners via informal institutional arrangements. They are able to function efficiently as long as production costs and transaction costs can be maintained low enough to allow for production to take place on a small-scale basis. TAIWAN IN TRANSITION At the moment, in fact, the Taiwanese political economy finds itself in a period of upheaval. The economy is under heavy strain despite the continued high growth rate. The traditional export industries are losing market shares because of eroded comparative advantages. This has led much of industry to transfer its operations to low-cost countries such as Thailand, Malaysia, Vietnam and China. Direct Taiwanese investments in South-East Asia have an aggregate volume exceeded only by Japan’s. 8 The ability of Taiwan to upgrade its technology and make its industries competitive in the domain of high technology is still an open question. Another and equally important factor for change is the civil society’s gradual penetration of the state apparatus. This movement manifests itself inter alia in the so-called democracy movement and threatens to undermine the KMT’s position of absolute power. An undercurrent of this movement is Taiwanese nationalism, one extension of which is the desire for a Taiwan politically independent of China. Such a movement obviously rejects the legitimacy of the KMT. The question is how the Taiwanese model will be able to survive when the functional division of power ceases and politics and economics become interwoven. There is a considerable risk in giving different interest groups, in the name of democracy, the opportunity of penetrating the state apparatus. This possibility is also opened by the fact that business and industry have become more and more concentrated and in recent times increasingly penetrated by foreign capital. It is probable, then, that the Taiwanese model is in the course of dissolution and that the autonomy of the state is coming to an end. So while other LDCs struggle with their attempts to copy a Taiwanese development model in which autocratic state power is a vital ingredient the Taiwanese themselves are faced with the problem of industrial restructuring and upgrading under conditions of weakening state authority.

177

ASIA

CONCLUSION The Taiwanese economic miracle scarcely permits of being summarized and explained with a few simple strokes of the pen. Taiwan’s political economy is far more complex and multifaceted than appears from the picture which we have given of various theoretical schools all the way from liberals to dependency theorists. One conclusion which can be drawn is that Taiwan is neither an Eldorado for laissez-faire capitalism nor an example of dirigiste capitalism. The strength of Taiwan’s economy lies in the existence on the one hand of a strong state power which fosters competition in the private sector but does not allow itself to be governed by private interests and, on the other, of a strong private sector which largely keeps itself free of state control. Is there a Taiwanese model then? For a development strategy to be capable of serving as a model for others it is important first of all that it should consist of a set of components which can be determined ex ante but which can also explain ex post why development took a particular form. If one cannot discern the presence of a coherent strategy, decisively guiding and influencing economic development in predetermined directions, it is hardly meaningful to speak of a model. The second criterion of a model is that it must be capable of being transplanted elsewhere. However, this capability is determined by both structural and historical conditions. Certain structural prerequisites must exist for it to be meaningful to think at all about transplantation of a model to another environment. Even though two countries may have relatively similar structural conditions, however, there could still be other obstacles to the transplanting of a development model. Historical circumstances will always determine whether it is politically possible to implement a particular development strategy even when the structural economic conditions are in place. It is sometimes said that LDCs have little to learn from the Taiwanese experience and that any attempt to compare Taiwan with ‘average’ LDCs is superficial. This may be true to some extent, but the worst we can do is to be superficial in our interpretations of the Taiwanese success story. The Taiwanese model is the child of special historical circumstances. Viewed in its entirety it resembles no other model of economic development that has been tried out in the developing world. Even so it is not without interest as an example for others. For it points to the possibility and perhaps even the necessity of both effectively functioning social institutions and guidance mechanisms on the one hand and a functioning individual-orientated market economy based on the principle of ‘equal opportunity’ on the other. In stressing the latter point we cannot avoid the question of redistribution of income. In fact the Taiwanese case lends more support to the case for ‘redistribution with growth’ than to the argument for ‘growth first—distribution later’. LDCs need a functioning market economy but also a state power which is both strong and autonomous in relation to private interest groups. In underdeveloped

178

THE TAIWANESE SUCCESS STORY

countries, in fact, the market economy does not function if this is not the case. To find in practice a way of implanting the Taiwanese model in underdeveloped countries appears very difficult. All the same it is perhaps a socioeconomic dynamic of Taiwanese ‘type’ that must be created for the market to be able to function in other developing countries. NOTES 1 2

3

4

5

6 7

8

For a critical assessment of this argument see Wade (1990). A view which has been adduced in recent years is that South Korea’s somewhat lower growth rate results from excessively slow and inadequate liberalization, dirigisme and a high degree of public ownership (Hughes 1989). This, to put it mildly, is a bold conclusion drawn on the basis of an average difference of 0.3 per cent in growth rate between Taiwan and South Korea during the period 1965–85. For example, Wu shows that as late as 1970 no fewer than 41 per cent of all industrial goods were classified as ‘controlled’, which means that they were protected from foreign competition. This control system was abolished almost completely in 1972, since which time customs duties are the only form of protection remaining. We are not referring here to ‘balanced development’ in the sense used by structuralists (Nurkse, Lewis and others). Their development strategy involved the encouragement of industrial development, the role of agriculture being to supply industry with cheap labour. We have not discussed in detail the significance of Japanese colonialism for Taiwan’s economic development. This subject would merit an article to itself. The Japanese programme for Taiwan’s development was very comprehensive. Investments were made in infrastructure and human capital, and great efforts were made to modernize agriculture. Oshima (1986) shows for example that during the 1920s and 1930s agricultural productivity actually increased faster in Taiwan than in Japan. Ruralbased small-scale industries owned by Taiwanese were already burgeoning around the expansive agricultural sector as early as the 1930s. This was an important factor in the creation of the post-war Taiwanese model. Another factor was the Japanese effort to promote industrial development, chiefly the processing of raw materials. In this sector great state monopolies arose which the KMT regime later on was able to take over. This is a factor which may explain the considerable scale of state-run enterprise during subsequent periods. As regards the shaping of the state apparatus and bureaucracy, on the other hand, colonialism probably played a smaller part. The state apparatus was entirely Japanese and the bureaucracy likewise. The model subsequently developed on Taiwan therefore has more to do with the KMT’s ideology and practice than with the Japanese heritage. The only real control to speak of is of course the fact that the official financial sector, which is state controlled, is the sector where most of the loan capital is originally created. In his excellent study Governing the Market, Robert Wade presents a convincing case for the close co-operation between the government and these larger companies (see Wade 1990). However, Wade’s study offers little insight into the problems and behaviour of the smaller companies in the sector which is found outside the government sphere. These new phenomena are the subject of analysis in the research project: ‘Exports of capital from East Asian NICs’ undertaken by the present author. For preliminary findings see Gunnarsson (1992).

179

ASIA

BIBLIOGRAPHY Adelman, I. and Morris, C. (1973) Economic Growth and Social Equality in Developing Countries, Stanford, CA: Stanford University Press. Amsden, A. (1979) ‘Taiwan’s economic history: A case of etatism and a challenge to dependency theory’, Modern China 5:3. ——(1985) ‘The state and Taiwan’s economic development’, in P.Evans, D.Reuschemeyer and T.Skocpol (eds) Bringing the State Back In, Cambridge: Cambridge University Press. Apthorne, R. (1979) ‘The burden of land reform in Taiwan’, World Development vol. 7. Balassa, B. (1978) ‘Exports and economic growth: Further evidence’, Journal of Development Economics no. 5. ——(1988) ‘The lessons of East Asian development: An overview’, Economic Development and Cultural Change, vol. 36. Barret, R. and Whyte, M.K. (1982) ‘Dependency theory and Taiwan: Analysis of a deviant case’, American Journal of Sociology vol. 87. Bhagwati, J.N. (1982) ‘Directly unproductive profit seeking (DUP) activities’, Journal of Political Economy vol. 90. Bradford, C.I. (1986) ‘East Asian “Models”; Myths and lessons’, in J.P.Lewis and V. Kallab (eds) Development Strategies Reconsidered, New York: Transaction Books. Browett, J. (1985) ‘The newly industrializing countries and radical theories of development’, World Development vol. 13. Chen, E. (1989) ‘East and Southeast Asia in the world economy: Issues, problems and prospects’, Copenhagen Papers in East and Southeast Asia. Cheng Wei-Yuan (1987) Economic dependency and development of Taiwan: Cross sectional and longitudinal analyses, Conference paper 3–5 September. Chou, T.C. (1985) ‘The pattern and strategy of industrialization in Taiwan: Specialization and offsetting policy’, Development Economics 23:(2). ——(1991) ‘From neo-mercantilism to liberalization–The role of ideology in revising Taiwan’s trade policy’, Public Administration no. 23. Cline, W. (1982) ‘Can the East Asian model of development be generalized?’, World Development vol. 10. Crane, G.T. (1987) The adjustment capacity of the state in Taiwan: State-owned enterprises and the oil shock of the 1970s, Conference paper 3–5 September. Cummings, B. (1984) ‘The origins and development of the Northeast Asian political economy: Industrial sectors, product cycles and political consequences’, International Organization vol. 38. Dollar, C. and Sokoloff, K. (1989) Industrial policy, productivity growth and structural change in the manufacturing industries. A comparison of Taiwan and South Korea, Conference paper, Taipei 4–5 December. Evans, P. (1987) ‘Class, state and dependence in East Asia’, in F.Deyo (ed.) The Political Economy of the New Asian Industrialism, Ithaca, NY: Cornell University Press. Evans, P. and Pang Chien Ku (1987) State structure and state policy: Implications of the Taiwanese case for newly industrializing countries, Conference paper 3–5 September. Evans, P., Reuschemeyer, D. and Skocpol, T. (eds) Bringing the State Back In, Cambridge: Cambridge University Press. Fei, J.C. and Ranis, G. (1979) Growth with Equity; The Taiwan Case, Oxford: Oxford University Press. Fernandez, P. (1980) ‘Third World industrialization. A new panacea?’, Monthly Review vol. 32. Frank, A.G. (1982) ‘Asia’s exclusive models’, Far Eastern Economic Review no. 25. Fransman, M. (1986) ‘International competitiveness, technical change and the state: The machine tool industry in Taiwan and Japan’, World Development vol. 14.

180

THE TAIWANESE SUCCESS STORY

Galenson, W. (ed.) (1978) Economic Growth and Structural Change in Taiwan, Ithaca, NY: Cornell University Press. Gates, H. (1979) ‘Dependency and the part-time proletariat in Taiwan’, Modem China vol. 5. Gold, T. (1986) State and Society in the Taiwan Miracle, Armonk, NY: Sharpe. ——(1987) Changing relations between the state and the private sector in Taiwan and mainland China, Conference paper, Taipei, 3–5 September. ——(1988) ‘Entrepreneurs, multinationals and the state’, in A.Winckler and S. Greenhalgh (eds) Contending Approaches to the Political Economy of Taiwan, Armonk, NY: Sharpe. Gunnarsson, C. (1987) ‘Utvecklingsstaten i Öst- och Sydöstasien’ (The developmental state in East and Southeast Asia), Den Ny Verden no. 15. ——(1989) The new institutional economics as a framework for analysing the state in developing countries, Conference paper, Taipei, 8–9 December. ——(1992) The political economy of restructuring, Conference paper, Taipei, 2–3 March. Haggart, S. (1986) ‘The newly industrializing countries in the international system’, World Politics vol. 38. Haggart, S. and Pang Chien-Kuo (1989) The transition to export-led growth in Taiwan, Conference paper, Taipei, 4–5 December. Haggart, S. and Tun-jen Cheng (1987) ‘State and foreign capital in East Asian NICs’, in F. Deyo, (ed) The Political Economy of the New Asian Industrialism, Ithaca, NY: Cornell University Press. Hamilton,C. (1983) ‘Capitalist industrialization in the four little tigers of East Asia’, in P. Limqueco, and B. McFarlane (eds) Neo-marxist Theories of Development, London: Croom Helm. ——(1987) ‘Can the rest of Asia emulate the NICs?’, Third World Quarterly 9: (4). Hammer, H.-J. (1984) ‘Comment on dependency theory and Taiwan: Analysis of a deviant case’, American Journal of Sociology vol. 89. Ho, S. (1978) Economic Development in Taiwan 1895–1970, New Haven, CT: Yale University Press. Hughes, H. (1989) ‘Catching up: The Asian newly industrializing economies in the 1990s’, Asian Development Review vol. 7. Jacoby, N. (1966) US Aid to Taiwan: A Study of Foreign Aid, Self-Help and Development, New York: Praeger. Johnson, C. (1981) ‘The Taiwanese model’, in J.Hsiung, (ed.) Contemporary Republic of China. The Taiwanese Experience 1950–1980, Cornell, NY: American Association for Chinese Studies. ——(1986) ‘The non-socialist NICs: Asia’, International Organization 40: (2). ——(1987) ‘Political institutions and economic performance: The government-business relationship in Japan, South Korea and Taiwan’, in F.Deyo (ed.) The Political Economy of New Asian Industrialism, Ithaca, NY: Cornell University Press. Ka, Chih-ming (1987) ‘Land tenure, development and dependency in colonial Taiwan 1895–1945’, Dissertation, State University of New York at Binghamton. Ka, Chi-ming and M.Selden (1986) World Development vol. 14. Krueger, A. (1974) ‘The political economy of a rent-seeking society’, American Economic Review vol. 64. Kuo, S.W.Y. (1983) The Taiwan Economy in Transition, Boulder, CO: Westview Press. Kuo, S.W.Y., Ranis, G. and Fei, C.H. (1981a) Rapid Growth with Improved Distribution in the Republic of China 1952–1979, Boulder, CO: Westview Press. ——(1981b) The Taiwanese Success Story, Boulder, CO: Westview Press. Kuznets, S. (1955) ‘Economic growth and income inequality’, American Economic Review vol. 45.

181

ASIA

Landau, D. (1983) ‘Government expenditure and economic growth: A cross-country’, Southern Economic Journal vol. 49. ——(1986) ‘Government and economic growth in the less developed countries: An empirical study for 1960–1980’, Economic Development and Cultural Change vol. 35. Lauridsen, L. (1990) From newly exporting countries to newly industrializing countries— The developmental role of the state. Comparing Taiwan and Thailand, Conference paper, Mön, Denmark, 30 September-3 October. Li, K.T. (1988) The Evolution of Policy behind Taiwan’s Developmental Success, New Haven, CT: Yale University Press. Li, K.T. and Yu, T.S. (1982) Experiences and Lessons of Economic Development in Taiwan, Taipei: Academica Sinica. Lin, C-Y. (1973) Industrialization in Taiwan 1946–1972, New York: Praeger. Lewis, W.A. (1955) Theory of Economic Growth, London: Allen & Unwin. Little, I.M.D. (1981) ‘The experience and causes of rapid labour intensive development in Korea, Taiwan province, Hong Kong and Singapore and the possibilities of emulation’, in E. Lee, (ed.) Export-led Industrialization and Development, London: ILO. Little, I.M.D., Scitovsky, T. and Scott, M. (1970) Industry and Trade in Some Developing Countries, Oxford: Oxford University Press. Lundahl, M. and Vedovato, C. (1986) ‘Statens betydelse för utvecklingen i fattiga länder’ (The role of the state in the economic development of poor countries), Ekonomisk Debatt no. 1. Myers, R. (1986) ‘The economic development of the Republic of China on Taiwan 1965– 1981’, in L. Lau,(ed.) Models of Development: A Comparative Study of Economic Growth in South Korea and Taiwan, San Francisco: ICS. Myrdal,G. (1957) Economic Theory of Underdeveloped Regions, London: Methuen. Nunnenkamp, P. (1981) ‘The efficiency of state-owned enterprises in the manufacturing industry of Taiwan’, Academia Economic Papers. Olson, M. (1982) The Rise and Decline of Nations, New Haven, CT. Oshima, H.T. (1986) Economic Growth in Monsoon Asia, Tokyo: Tokyo University Press. Rana, P.B. (1987) ‘Foreign direct investment and economic growth in the Asian and Pacific region’, Asian Development Review vol. 5. Ranis, G. (1978) ‘Equity with growth: How special is the special case?’, World Development vol. 6. ——(1989) ‘The role of institutions in transition growth: The East Asian newly industrializing countries’, World Development vol. 17. Ranis, G. and Schive, Chi (1985) ‘Direct foreign investment in Taiwan’s development’, in W.Galenson, (ed.) Foreign Trade and Investment: Economic Development in Newly Industrializing Asian Countries, Madison, WI: University of Wisconsin Press. Reuschemeyer, D. and Evans, P. (1985) The state and economic transformation: Towards an analysis of the conditions underlying effective intervention’, in P.Evans, D.Reuschemeyer and T.Skocpol (eds) Bringing the State Back In, Cambridge: Cambridge University Press. Riedel, J. (1987) ‘Economic development in East Asia: Doing what comes naturally’, in H. Hughes (ed.) Explaining the Industrialization Success of East Asia, Cambridge: Cambridge University Press. Schive, Chi (1990) The Foreign Factor. The Multinational Corporation’s Contribution to the Economic Modernization of the Republic of China, New York: Hoover Press Publication. Schive, C. (1991) ‘How Taiwan overcame its Dutch disease problem’, Unpublished paper, Taipei. Scitovsky, T. (1986) ‘Economic development in Taiwan and South Korea 1965–1981’, in

182

THE TAIWANESE SUCCESS STORY

L.Lau (ed.) Models of Development: A Comparative Study of Economic Growth in South Korea and Taiwan, San Francisco: ICS. Senghaas, D. (1985) The European Experience: A Historical Critique of Development Theory, London: Berg Publishers. Shea, Jia-Dong and Ya-Hwei Yang (1989) Financial system and the allocation of investment funds, Conference paper, Taipei, 4–5 December. Srinivasan, T.N. (1985) ‘Neoclassical political economy, the state and economic development’, Asian Development Review vol. 3. Sun Yat-sen (1924) The Three Principles of the People (transl.), China Publishing Co. The Taiwan Development Experience and Its Relevance to Other Countries (1985) Taipei: SRI International. Wade, R. (1984) ‘Dirigism Taiwanese style’, IDS Bulletin vol. 15. ——(1988a) ‘State intervention in “outward-looking” development: Neoclassical theory and Taiwanese practice’, in G.White (ed.) Developmental States in Asia, London: Macmillan. ——(1988b) ‘The role of government in overcoming market failure: Taiwan, Republic of Korea and Japan’, in H.Hughes, (ed.) Achieving Industrialization in East Asia, Cambridge: Cambridge University Press. ——(1990) Governing the Market: Economic Theory and the Role of Government in East Asian Industrialization, Princeton, NJ: Princeton University Press. Werner, R. (1985) ‘Taiwan’s trade flows: The underpinning of political legitimacy?’, Asian Survey vol. 25. White, G. and Wade, R. (1988) ‘Developmental states and markets in East Asia: An introduction’, in G.White (ed.) Developmental States in Asia London: Macmillan. World Bank (1987) World Development Report 1987, Oxford: Oxford University Press. Wu, Rong-I. (1971) The Strategy of Economic Development. A Case Study of Taiwan, Louvain. ——(1989a) Foreign direct investment and technology transfer, Conference paper, 4–5 December. ——(1989b) ‘Trade liberalization and economic developemnt in Taiwan, ROC’, in Social Sciences Development, Social Sciences Publication Series no. 1, National Chung Hsing University, Taipei. Wu Yuan-li (1985) Becoming an Industrialized Nation, New York: Praeger. Wynn, S. (1982) ‘The Taiwanese economic miracle’, Monthly Review vol. 33. Yin, K.Y. (1954a) ‘Adverse trend in Taiwan’s industrial development’, Industry of Free China no. 2. ——(1954b) ‘A discussion of industrial policy in Taiwan’, Industry of Free China no. 1.

183

10 TRADE AND DEVELOPMENT The experience of Korea and Taiwan Wontack Hong

INTRODUCTION Taiwan, in the late 1950s, and Korea, in the early 1960s, both commenced exportoriented growth by promoting the export of labour-intensive manufactured products. As a result, both Korea and Taiwan went through drastic economic transformations from typical backward economies to so-called export-oriented NICs. The objective of this study is to analyse the similarities and differences between the experiences of these two countries in their export-oriented growth. The next section briefly highlights the initial conditions in Korea and Taiwan prior to their commencement of export-oriented growth. The third section delineates the necessity of export-promotion efforts and the actual beginning of export-oriented growth strategy. The fourth section describes how these countries institutionalized the export-oriented regime, the fifth section examines the financial systems that have been maintained, and the final section attempts a comparative evaluation of the growth performances of these two export-oriented economies. THE INITIAL CONDITIONS One may amplify the differences in the initial conditions of Taiwan from those of Korea by emphasizing the facts that Taiwan enjoys a multicrop subtropical climate, that there was a sudden influx of 1½ million highly educated middleclass refugees of distinct characteristic traits into Taiwan from the mainland in the late 1940s, and that Taiwan could benefit from the extensive overseas Chinese network and a well-experienced government establishment from the beginning (i.e. 1949). This section, however, focuses more on the similarities in the initial conditions of Korea and Taiwan. By the end of Japanese colonial rule, only 2 per cent of the Korean population over 14 years of age had completed secondary school education (see Kim 1990:146). The literacy rate was 13.4 per cent in 1945 (see Scitovsky 1985). After the Second World War, however, there was an outburst of educational zeal among

184

KOREA AND TAIWAN

the Korean people, even during the chaotic period of 1945–53 and especially in the 1950s and 1960s. The number of enrolments in elementary schools increased by 8.2 per cent per annum during 1945–52 (increasing from 1.37 million people in 1945 to 2.37 million people in 1952), and yet, due to the heritage of a colonial educational obscurantist policy, elementary school enrolment as a percentage of corresponding age groups (i.e. 6–11 years old) amounted only to 60 per cent and the illiteracy rate still stood at 78 per cent in 1953. The number of enrolments in colleges and universities increased by 23.4 per cent per annum in 1945–52, and yet the enrolment as a percentage of corresponding age groups amounted to a mere 3.1 per cent in 1953 (see Kim 1990:146, McGinn et al. 1980:5–6). In 1953, the share of manufacturing value added in Korea’s GNP amounted only to 9.0 per cent. As late as 1963, the share of manufacturing in total employment and GNP amounted to 7.9 per cent and 14.7 per cent, respectively. Hence Mason et al. (1980:250) could not but state that ‘[d]espite the relatively rapid growth of manufactured output after 1953, Korea, on international comparison, had a manufacturing sector smaller than the average for countries of similar per capita income and a volume of exports very much smaller’. The so-called ‘Japanese colonial heritage’, i.e. the introduction to the Korean people of modern technology and management as well as modern social-overhead capital facilities, seems to have been no better than those in, say, India, Indonesia or Vietnam that were inherited from the colonial rule of the British, Dutch and French, respectively. The so-called ‘growth potential’ (in the form of human capital, in particular) left behind by the Japanese at the end of the Second World War was no better than those left behind in more than a hundred underdeveloped countries by the western colonial powers (cf. Mason et al. 1980:2–6). The handful of heavy industries and electrical power plants that emerged during the colonial period and the important mineral deposits were mostly located in the northern part of Korea. The southern part was dominated by agriculture and the manufacture of a few light consumption goods. The Korean economy, which was initially designed as a colonial economy dependent on Japan and then further crippled by the separation of the North from the South in 1945, found itself fully engaged in reconstruction work in the wake of the Korean War (1950–3). In the 1950s, Korea, with less than $750 of per capita GNP (in 1989 dollar prices) and more than two-thirds of the entire population engaged in the primary sector, possessed all the familiar characteristics of an extremely underdeveloped economy. Japan ‘designed’ Taiwan as a major supplier of foodstuffs. Until 1936, industrialization in Taiwan was non-existent. The Japanese industrialization of Taiwan after 1937 gave rise to aluminium plants, metal-refining plants, cement plants, small-scale fertilizer plants and alcohol production as a by-product of sugar refining. Basic consumer goods, including textiles, were imported from Japan. Ginsburg (1952) states: At the close of the War the economic structure of the island was near

185

ASIA

collapse. The 20,000 Japanese agricultural technicians had gone, fertilizers were scarce…. One of the key problems in Taiwan today is the replacement of Japanese technicians with Chinese [from the mainland]. Unfortunately, relatively few of the mainlanders possess the necessary technical skills, and those who do too often have been trained for topechelon rather than middle-echelon jobs. In Taiwan, the illiteracy rate among the population over 6 years of age amounted to 55 per cent in 1946 (see Hou 1988:8). 1 According to the Household Registration data, the average number of years of schooling for all persons aged 15 years and over was only 2.8 in 1946–50 (see Hou and Chang 1981:486). As of 1952, three years after the relocation of the Kuomintang government (accompanied by 1.5 million refugees) to the island of Taiwan, only 1.4 per cent and 8.8 per cent, respectively, of the population over 6 years of age were attending or had completed college-university education and secondary education, and the illiteracy rate still amounted to 42 per cent (see CIECD 1973:7).2 In 1954, the enrolment rate by school-age population group amounted to 91 per cent for the primary level (6–11 years of age) but for the secondary level (12–17 years) it amounted only to 22.4 per cent for males and 9.7 per cent for females and, for the tertiary level (18–24 years), it amounted to a mere 1.9 per cent for males and 0.3 per cent for females. In Taiwan, the pre-war production level is believed to have been restored during 1948–51 (see Liang and Liang 1981:314). And yet, the share of manufacturing in Taiwan’s total employment and NNP (Net National Product) in 1952 amounted to 12.4 per cent and 10.8 per cent, respectively, which rose to 14.1 per cent and 15.5 per cent in 1958. Hence Galenson (1981:70) asserts that ‘Taiwan began its development march from an initial condition that was not superior to those of many less developed countries’. Both in Korea and Taiwan, however, there were rapid improvements in the overall level of education in the 1950s, and consequently both countries came to have a relatively well-educated labour force by the time they commenced exportoriented growth. In Korea, the enrolment as a percentage of corresponding age group expanded from 60 per cent to 86 per cent for elementary schools, from 21 per cent to 33 per cent for junior high schools, from 12 per cent to 20 per cent for senior high schools and from 3.1 per cent to 6.4 per cent for colleges and universities during 1953–60.3 The illiteracy rate fell from 78 per cent to 28 per cent during the same period. In Taiwan, the enrolment rate increased from 91 per cent to 95 per cent for elementary schools, from 22.4 per cent to 33.9 per cent and from 9.7 per cent to 16.6 per cent for males and females, respectively, for secondary schools, and from 1.9 per cent to 4.0 per cent and from 0.3 per cent to 0.9 per cent for males and females, respectively, for tertiary schools during 1954–8.4 Furthermore, in Korea, as a result of the Korean War, almost the entire ablebodied male population had to go through basic military training and serve several years in the modern military establishment. Many economists believe that

186

KOREA AND TAIWAN

the regimental experiences of discipline must have been conducive in facilitating the transformation of traditional farm labour into modern factory workers. The extraordinary size of the military establishment maintained in Taiwan must also have produced similar effects. THE INEVITABILITY OF EXPORT PROMOTION After 1950, both Korea and Taiwan depended heavily on US grant-in-aid to finance their imports: about 42 to 47 per cent of Taiwan’s commodity imports during 1953–6 and about 82 to 85 per cent of Korea’s commodity imports during 1956–8 were financed by US aid. Foreign capital inflow to the two countries occurred only after 1961. Although Taiwan’s First Four-Year Economic Reconstruction Program for the attainment of economic independence (FFYP 1953–6) envisaged US aid to 1956 in decreasing annual amounts, the Program was in fact well supported by US aid. The ROC Taiwan government put into effect the Agricultural Four-Year Plan in late 1953 in conjunction with the FFYP in order to promote agricultural import substitution and to increase exports, ‘thus building up foreign exchange with which to buy the most needed products’ (US DC, 1959:15). With the beginning of the more ambitious Second Four-Year Economic Reconstruction Program (SFYP:1957–60), however, the absolute magnitude of US grant-in-aid to Taiwan indeed started to decline (and eventually came to an end in 1968), accentuating the problem of chronic foreign exchange shortages. Although the Program itself envisaged a drastic cut in US aid (i.e. US aid in 1960 was expected to be about 45 per cent of the 1956 level), and although Taiwan ended up receiving a large amount of US aid which was more than it anticipated, it is still true that at the very moment when the need for foreign exchange to finance long-term economic development began to expand, the supply of US aid to Taiwan began to decrease (see US DC, 1959:15–16). As a result, while the FFYP included only ‘improvement in balance of payments’ in its list of major targets, the SFFP added ‘export promotion’ to the list (see Li and Yeh 1981:174). The export target presented in the SFYP was, however, rather modest: exports were to increase by 41 per cent above the 1956 level (i.e. from $130 million in 1956 to $182.5 million in 1960). Even this modest target was not achieved: exports increased only by 34 per cent during 1956–60. The main objective of the SFYP was merely to achieve a self-supporting economic unit through industrialization, with agricultural development serving as a base. According to Li and Yeh (1981:170–1). ‘[b]y the time the Third Plan [1961–5], which incorporated the Nineteen-Point Program of Economic and Financial Reform as well as the Statute for the Encouragement of Investment of 1960, was under preparation, the eventual phasing-out of the U.S. economic aid program had become a matter of time

187

ASIA

On the other hand, in Korea, US aid reached its peak in 1957 and then its absolute magnitude started to decline rapidly thereafter. In the midst of such a declining trend, the newly established military government in Korea launched the ambitious First Five-Year Plan (FFYP:1962–6) in 1962. The ROC Taiwan government envisaged relatively declining traditional exports such as sugar and rice (to markets such as Japan established through prewar trade relationships) and a somewhat broadened export base involving pineapples, tea, bananas, aluminium, petroleum, cement and cotton textiles. The FFYP of the Korean government also envisaged an expanded share of manufactures in total commodity exports from 15 per cent in 1960 to 22 per cent in 1966. Apparently both the Korean government and the ROC Taiwan government keenly recognized the need to enhance foreign exchange earnings but, at the beginning, neither of them seems to have conceived the expansion in foreign exchange earnings through the single-minded promotion of labourintensive manufactures exports. The fact that Taiwan was the second most densely populated country in the world and Korea the third most did not particularly attract the attention of policy makers in these countries. Both governments simply promoted exports in general, but due to the obvious limitation in the export potential of primary goods the benefits of various export promotion measures ultimately fell mostly upon labour-intensive manufacturing.5 In the 1950s, no one had heard about the so-called ‘outward-looking’ ‘exportoriented’ growth strategies; familiar slogans then were ‘balanced growth’ or ‘unbalanced growth’. As Little (1979:474) summarizes: ‘Almost all development economists were then emphasizing industrialization through protection, the necessity of controlling imports, the price mechanism in general, and the dangers of relying on trade with the imperialist powers’. In order to relieve the problem of serious foreign exchange shortages, the governments of Korea and Taiwan began to promote foreign-exchange-earning activities through export promotion and then, somewhat unexpectedly, they witnessed the expansion of exports in labourintensive manufactures. The miraculous achievement of these governments was the fact that they were capable of recognizing immediately that growth through the export promotion of labour-intensive manufactures was the best possible growth strategy for them and, most importantly, they put an all-out effort into pursuing that strategy.6 INSTITUTIONALIZING THE EXPORT-ORIENTED REGIME Compared with most other developing countries, both Korea and Taiwan were blessed by having been subject to less insurmountable socio-political obstacles to institutionalizing incentive schemes to promote export activities. By the early 1950s, both countries had already eliminated the tradition-bound landlord class through successful land reforms. Both had experienced an import-substitutionoriented regime, but their experiences were not long enough to have generated

188

KOREA AND TAIWAN

extremely powerful vested interest groups entrenched in the import-substituting activities geared to the captive domestic markets. Although the policy to promote import substitution has never been discontinued in Taiwan, and the average nominal tariff rate for all importable items amounted to 55.7 per cent as late as in 1974 (see Hou 1988:10–11), the ROC Taiwan government was able to institutionalize the bulk of the export promotion system by the early 1960s. In July 1955, the Regulations for Rebate of Taxes on Export Products provided the rebates of import duty, defence surtax and the commodity tax levied on export products and on raw materials used in export production activities. In 1956, manufacturers could retain up to 80 per cent of the foreign exchange they earned from exports and use them on the imports of raw materials needed to produce export products.7 In 1957 a liberal and flexible export loan programme was introduced (at an interest rate of 0.99 per cent per month for a maximum term of 1 year) in order to promote exports of agricultural, industrial and handicraft products. The multiple exchange rates were gradually unified at a more realistic level, ultimately settling at the rate of $NT40T per $US1 in June 1961. Both in Korea and Taiwan, the extremely overvalued domestic currency in the 1950s had discouraged exports and necessitated severe quantitative import controls that encouraged import-substituting activities (more than offsetting the impact of the overvalued domestic currency on imports). The unified realistic exchange rate in the 1960s, however, not only encouraged the expansion of non-traditional exports but also reduced the need for extreme quantitative import controls. See also Tsiang (1984:306–7). The Third Four-Year Plan (1961–5) of Taiwan introduced the Nineteen-Point Economic Financial Reform that aimed to provide preferential treatment to investment activities, to unify exchange rates, to reduce import controls and to encourage export expansion (see Kuo et al. 1981:74). The Statute for Encouragement of Investment was enacted on 10 September 1960 pursuant to the nineteen-point reform measures. The Statute provided new firms or old firms undertaking expansion (of 30 per cent or more in production capacity), which conformed to the Statute’s criteria, a five-year exemption from corporate income taxes or an accelerated depreciation allowance at the choice of the firm.8 The new industrial establishments eligible under the Statute were also exempted from import duties on machinery and equipment imported for their own use, and corporation profit reinvested for productive purposes was deductible from taxable income (see Yu and Chen 1981:284, Balassa 1981:417). According to Kuo et al. (1981:76), the average rate of tax reduction and refunds on income tax, stamp tax, customs duties and commodity tax increased from 1.5 per cent in 1955 to about 12 per cent in 1960–1, about 23 per cent in 1962–5 and to about 30 percent in l965–70.9 Export activities were exempted from business and related stamp taxes. Manufacturing and mining firms that exported more than 50 per cent of their outputs could deduct 2 per cent of total export earnings from taxable income and the tax rate itself was reduced by 10 per cent. (This income tax reduction scheme was abolished in 1970.) Exporters could retain their foreign exchange earnings for the

189

ASIA

importation of raw materials and machinery and could sell such import rights to other firms. (This system was also terminated in 1970.) Special low-interest loans were provided to finance the import of raw materials and other export-related production activities.10 Manufacturers’ associations financed direct subsidies to export in such industries as cotton spinning, iron and steel, rubber products, woollen fabrics, paper products and monosodium glutamate (see Liang and Liang 1981:328). Economic and commercial counselling offices were established in several foreign countries, and the China Productivity Center was established. Government also provided such export-promotion facilities as export inspection, technical and trade consultation services, and market research. That is, the government started to encourage the exploration of new export products, including processed agricultural products. According to Kuo (1988:41): ‘The response of industry toward export expansion in the early 1960s was still slow. In 1965, the Statute [for the Encouragement of Investment] was revised and its scope was expanded. The Kaoshung Export Processing Zone was set up…. Development strategy at this time became entirely export-oriented.’ See also Ho (1978:187–98). According to Kuo and Fei (1985:56–7), Taiwan had maintained fairly high real interest rates during 1952–70, but the high rates of profit induced high rates of domestic investment, i.e. the average real interest rate decreased from about 21 per cent in 1951–3 to about 10–12 per cent in 1954–65 and then to 8–9 per cent in 1966–70, while the rate of profit in manufacturing rose steadily from about 9 per cent in 1951–3 to about 16 per cent in 1954–6, to about 28 per cent in 1957–62 and then to about 35 per cent in 1963–70.11 The real interest rates were high but not so excessively high as to discourage investment. After 1958, tariffs on imports were reduced and the controls on foreign exchange allocation became less strict. According to Liang and Liang (1981:325), the ratio of net customs revenues to imports was reduced from about 42 per cent in 1955 to about 28 per cent in 1960 (the gross rate, which included those rebated to exporters, was reduced from about 43 per cent to about 32 per cent). The quota on permissible imports was abolished. According to Yu and Chen (1981:287), the ‘restrictions on the import of raw materials and, to only a slightly lesser extent, on the import of machinery equipment were very substantially eased as of 1958 for manufacturers who wanted to export’. According to Ho (1978:198), the post-1958 reform, by changing the critical price relationship, made it more profitable for domestic producers to export, i.e. ‘producing for export has become…more profitable than producing for the domestic market’.12 After the early 1960s, the Korean government not only maintained a realistic unified exchange rate, but also mobilized fully the preferential tax system, preferential subsidy system and administrative support system in order to reward those who contributed to export expansion (see Hong 1979). In Korea, the subsidies mostly took the form of rationing the low (often negative) real interest rate domestic and foreign bank loans to selected activities and firms. Furthermore, the relative rates of return on import-substituting activities in Korea were significantly lowered by the reduced overall quantitative import restrictions

190

KOREA AND TAIWAN

(see Hong 1989). The imports of materials directly required for export production activities were set free from virtually any kind of government restrictions (until the early 1970s): as far as export activities were concerned, Korea had maintained a free import regime. Mason et al. (1980:262) characterizes the approach of the Korean government in pursuing export-oriented growth strategy as follows: Government policy is also highly pragmatic in the sense that it shows no hesitation in devising means most appropriate to the end in view without significant ideological bias. The result is a balance between market forces and direct government intervention and between government ownership and reliance on private entrepreneurship. Where the market works, fine; where it does not, the government is quick to intervene. Policy is not only pragmatic but particularistic in the sense that the activities of a single firm may form the object of government intervention. Throughout the 1960s and 1970s, the Korean government had rewarded individual entrepreneurs with preferential tax treatment and subsidized credit rationing mostly in proportion to their export performance. As a measure of each entrepreneur’s contribution to national economic growth, the gross export value may have been an economically much less sensible index than the net export value (value added) and may have unduly encouraged the dependence of Korea’s export production activities on imported inputs.13 And yet, the index had the advantage of simplicity. Maximizing the receipt of government subsidies implied maximizing the gross export performance. With such a clear-cut simple performance test, it was relatively easy for every entrepreneur fully to mobilize energies and ingenuities to achieve the goal (see Hong 1990b). As a result, ever since 1962, the actual export performance has always exceeded the planned export target in Korea. Within the Taiwan government, there were conflicts between the conservative group that wanted to pursue a state-oriented economy and the liberal group that wanted to pursue a market-oriented economy. Little (1979:475) states that ‘the whole Chinese tradition was against laissez-faire and free trade…[m]oreover most high officials and ministers had been in government during World War II and were accustomed to controls’. Since the mainlanders had to appease the native Taiwanese in economic affairs, however, the government chose a marketoriented economy with, as a compromise, government ownership of some important industries. According to Little (1979:475), the intention of the Taiwan government ‘was not to create laissez-faire conditions for the whole industry, let alone the whole economy. They created a kind of dual economy in which exports, but only exports, could be manufactured under virtually free trade conditions—a policy soon copied by Korea.’ Hou (1988:19) on the one hand states that ‘since the [Taiwan’s] exportpromotion policies were by and large directed to all industries without special

191

ASIA

favour to any particularly selected or targeted industries, entrepreneurs were essentially left alone to develop those industries whose products they could sell profitably in foreign markets’, and then also states that Taiwan has always had an import-substitution policy. The basic philosophy underlying this policy or strategy is that an economy will undergo certain stages of development, and at each stage there are certain key industries [such as integrated steel mills, large shipyards, and petrochemical plants] which through various linkages will bring about development of the entire economy. This strategy also assumes that government officials know what those key industries are and what policy measures should be adopted to develop those industries. The import substitution policy adopted in Taiwan embodied all the above elements. (Hou 1988:26) The Statute for Encouragement of Investment granted incentives only to firms producing iron and steel, aluminium, copper, petrochemicals, non-electrical machinery, transport equipment, processed food, pulp and paper, rubber processing, various non-metallic minerals, textiles, prefabricated housing units and several miscellaneous items. The granting of incentives to firms appearing on this positive list has been further subject to limitations on minimum capacity, amount invested, manufacturing process or decisions by the Ministry of Economic Affairs that have enhanced the scope of government discretionary action (see Balassa 1981:417).14 The Korean government has always maintained sector-specific importsubstitution policies, and yet the export-promotion policies were, as in Taiwan, not so conspicuously sector specific until the mid 1970s when it began targeting actively various selected export industries. After the mid 1970s, the Korean government granted various tax exemptions and channelled low-interest domestic or foreign bank loans to hand-picked entrepreneurs, mostly big business groups, who undertook those investment projects selected for them by the government. When the projects failed, the government had to assume the responsibility of repaying the bank loans. The Korean government had extensively socialized the investment risk for the selected entrepreneurs, and such an arrangement invigorated the animal spirit of the big business groups, inducing them to indulge in aggressive expansion through the fail-safe government-sponsored investment activities. With an alarming number of failures, however, this targeting-cum-risksocialization practice of the Korean government had been very much reduced by the 1980s. Though much less conspicuous, however, the Taiwan government has also practised targeting by promoting selected strategic industries through fiveyear income tax holidays, tax-exempt retained earnings, a maximum ceiling of 20 per cent business income tax on key capital and technology-intensive enterprises, tariff exemption on machinery imports, low interest loans and accelerated depreciation allowances.

192

KOREA AND TAIWAN

According to Gold (1986:87), ‘Taiwan did not develop a laissez-faire, freemarket economy; the state retained multiple controls and only granted what seemed like free-market activities within strict bounds’, and ‘[t]here was a lot more going on than getting the prices right’. Both in Korea and Taiwan, however, the government has refrained from excessive direct controls of economic activities and allowed market forces to function, admittedly within bounds, so that the energies of the workers and entrepreneurs could be fully mobilized for export-oriented activities. The merits of the competitive market mechanism have been, if not fully, well exploited as to sustain the dynamism of both economies throughout the 1960s, 1970s and 1980s. There were numerous government failures, but as a whole and in a relative sense, the government has mostly tried to make up the (supposed) market failures. Both in Korea and Taiwan, the government regulations on economic activities and interfirm competition have been combined into a discretion-cumcompetition market system. The necessity of external competition in international markets helped prevent the system from degenerating into a rent-seeking one and maintained the dynamism of each economy (see Hong 1990b). Tsiang (1984:318) summarizes the mechanism of ‘economic takeoff in Taiwan as follows: The relative stability of prices and the fairly attractive interest rates for savings deposits restored and stimulated the traditional thrifty habit of the Chinese people. …additional tax measures such as exempting from personal income tax the interest income from saving and time deposits with maturity terms of two years or more, and exempting from corporation income tax profits that were plowed back for investment. …rapidly enlarged the inflow of voluntary savings, which provided noninflationary financing for the domestic investment opportunities created by the new export-encouragement policies. Most importantly, the ROC Taiwan government, just like the Korean government, had severely suppressed labour union activities, and hence were able to maintain perfectly competitive labour markets and a market-clearing wage rate. Until very recently, neither in Korea nor in Taiwan were trade union movements allowed to negate the competitive advantage of abundant and cheap labour.15 FINANCIAL SYSTEMS AND ALLOCATION OF INVESTMENT FUNDS Ever since the early 1950s, the financial system in Taiwan (including the postal savings system) apparently has successfully performed the task of mobilizing savings, but achieved questionable performance in allocative efficiency, i.e. in the allocation of loanable funds.16 Since the early 1980s, the financial system in Korea has also begun to perform the task of mobilizing savings.17 Just like in Taiwan, however, Korea’s financial system by itself cannot claim to have maintained a high allocative efficiency.

193

ASIA

Most of the full-service domestic banks in Taiwan have been managed by the government and hence have been bound by all kinds of laws and regulations and vulnerable to the interventions of government authorities and interest groups. The big three state-owned banks control nearly 50 per cent of deposits, enjoying a three-percentage-point spread between the interest rates applied on deposits and loans. Both in Korea and Taiwan, financial institutions have been tightly controlled, or, one may rather say, directly managed by the government. There are two important differences however: (a) Korea maintained real interest rates below 5 per cent per annum on average until 1982, while Taiwan maintained real interest rates higher than (plus) 5 per cent per annum on average on bank deposits and loans since the early 1950s; and (b) the Korean government has very much concentrated bank loans to big private business groups, while the Taiwan government, though favourably financing government-owned corporations, has not ‘intentionally’ concentrated bank loans to big business groups. The high interest policy in Taiwan has apparently encouraged the use of more labour-intensive methods in production activities than in Korea. According to Scitovsky (1985), the policy has also contributed to Taiwan’s egalitarian income distribution: the typical lender is a small saver, the typical borrower is the corporation, often the large corporation, so that high interest rates favor the low income saver and limit the profits of business enterprise…[a]nother advantage is that it [high interest rate] limits profits which restrains the rate at which the size of the individual enterprise grows. In neither country, however, the real interest rates have been market-clearing ones which could equate the demand for and supply of loanable funds, although the rates of implicit subsidies associated with credit rationing have been much smaller in Taiwan than in Korea. The relative share of deposits and loans of the (government-managed) commercial banks has been declining not only in Korea, but also in Taiwan, while the share of other, private, non-monetary financial institutions such as investment and trust companies or life insurance companies has been steadily increasing. Since the late 1970s, curb market lending activities in Korea have rapidly merged with the activities of financial institutions such as investment and trust companies. This has significantly reduced the risk associated with lending in informal sectors and has resulted in a rapid expansion in the activities of private, non-bank, financial institutions. Taiwan has shown a rather slow rate of expansion in the share of assets owned by ‘Other Financial Institutions’ if we exclude the postal savings systems. The curb market in Taiwan never seems to wither: during 1964–86, it provided about 37 per cent of funds for private enterprise and its share expanded in the 1980s.18 The fact that business firms and individuals are still willing to pay higher costs to get curb loans in both Korea and Taiwan implies that there still exists subsidized

194

KOREA AND TAIWAN

credit rationing. This also implies that, for the politically less influential businessmen, the real (effective) cost of borrowing from the organized financial institutions, taking into account the compensatory deposit or collateral, must be much higher than the official nominal cost. The government-operated commercial banks still maintain monopolistic powers, and the activities of secondary, non-bank, financial institutions are still very much restricted by the government both in Korea and Taiwan. In Taiwan, in spite of interest rate liberalization (on deposits) enforced after January 1986, one could only observe widening differences between lending rates and deposit rates. Indeed, a liberalization on interest rates (on bank deposits) without an accompanying liberalization in the financial sector in general does not seem to activate the ‘market principle’. Taiwan’s financial system needs modernization and rationalization. And yet, in Taiwan, the theoretically desirable financial policies are not implemented because the processes of decision making and implementation are hampered by vested interest and corruption.19 Similar arguments should also be applied to Korea’s backward financial system (see Hong 1986, 1987). In terms of secondary-market trading, the bond markets barely exist in both Korea and Taiwan. In Korea, direct financing has increased tremendously during 1987–9 and the market value of stock outstanding became equivalent to about 60 per cent of GNP in 1989. Although Korean entrepreneurs have recently been actively raising investment funds from capital markets, most of them still hesitate to go public, being afraid of losing their traditional control of firms. In Taiwan, the equity market was set up to provide a legal place to trade shares of state-owned companies. The government still imposes stiff regulatory controls on firms issuing shares and entrepreneurs still hesitate to raise investment funds from capital markets. The stock market is still full of retail investors pursuing speculative short-term capital gains in a casino-like atmosphere. EVALUATING THE RELATIVE PERFORMANCE OF KOREA AND TAIWAN Both Korea and Taiwan have pursued export-oriented growth strategies with all kinds of policy measures which, in hindsight, were not always the best possible ones. Many economists, however, think that the set of policies adopted by the Korean government must be deemed to have been less efficient or less desirable than that adopted by the ROC Taiwan government because Korea had to suffer chronic inflation that amounted to nearly 20 per cent per annum on average until 1982, underutilized domestic savings potential, cumulating external debts, worsening (or non-improving) distributive equity, and concentration of economic power in the hands of a small number of big business groups. Both in Korea and Taiwan, export-oriented growth generated massive new employment and rapid increases in real wage rates (in Taiwan, since the early 1960s and in Korea since the late 1960s) that contributed to the improvement of the overall distributional equity. In Korea, however, the deliberate concentration of economic power in the

195

ASIA

hands of a small number of big business groups as well as the extremely preferential tax treatment that had been provided for non-labour nonentrepreneurial incomes (such as those associated with the transactions of real estate or financial assets) had a seriously adverse impact on distributional equity. In Korea, the income shares of the upper 20 percentile class and the lower 40 percentile class were 44 per cent and 18 per cent, respectively, in 1985, while those in Taiwan amounted to 38 per cent and 22 per cent, respectively, in 1987 (see CEPD 1989, EPB 1988).20 By establishing a large number of industrial parks that provide infrastructure facilities and land and buildings on a rental basis, the ROC Taiwan government has encouraged many people with entrepreneurial talent to establish themselves as independent businessmen. In Korea, the existence of government-supported large firms made it more difficult and expensive for small newcomers to enter the market. The Taiwan government has also actively encouraged the decentralization of industrial activities, which has enabled Taiwan to maintain a more labour-intensive growth pattern than Korea, minimizing the cost of excessive urbanization and achieving a more equitable distribution of income (see Ranis 1979:222–3, Hong 1987).21 Recently, however, the policies adopted by the Taiwan government began to be criticized because of their apparent failure to promote large enterprises based on legitimate scale economies and their failure to utilize properly the vast available investment funds (i.e. domestic savings) for investment in domestic production activities. The absence of large-scale enterprises in Taiwan becomes very conspicuous when one makes the comparison with Korea or Japan. If Korea has sacrificed efficiency by excessively promoting large-scale business groups. Taiwan can be regarded to have sacrificed efficiency by failing to generate large enterprises on the basis of legitimate scale economies. The financial system of Taiwan could support a few large-scale government-owned enterprises, but it seems to have failed to extend adequate support for export-oriented large private enterprises.22 Banks in Taiwan make loans on the basis of excessively conservative principles mostly requiring real estate as collateral and hence often being accused of mimicking the pawnshop. That is, Taiwan’s financial system has been capable of mobilizing large domestic savings but has not been capable of adequately channelling the savings into investment activities characterized with substantial scale economics. On the other hand, Taiwan has had trade account surpluses every year since 1971 except 1974 and 1975. Before 1981, the trade-surplus/GNP ratio was around 4 to 5 per cent (except in 1972 and 1978 when it reached around the 8 per cent level). Since 1981, the magnitude of the trade surplus expanded rapidly, reaching 22.5 per cent in 1986 and 21.5 per cent in 1987.23 As of 1988, it still amounted to 11.6 per cent of GNP. The fact that a large proportion of domestic savings have not been utilized for domestic investment purposes, while most small firms still have to depend on curb loans, implies that Taiwan’s financial system has failed to provide low-cost investment funds to all capable entrepreneurs even in the midst of an excessive abundance of loanable funds.

196

KOREA AND TAIWAN

Taiwan has maintained substantial export surpluses since the late 1970s, and yet the foreign exchange rate policy has been determined by vested interest groups which have firmly established themselves in the course of export-oriented growth. The artificially determined exchange rates have been generating artificial profits to vested export producers. Even with such enormous current account surpluses, the Taiwan government had resisted appreciation of the NT dollar until 1986, and some are still arguing for a devaluation of the NT dollar. Further, the relative size of public investment in Taiwan has declined, the real interest rates were not allowed to decline significantly to stimulate investments (for fear of inflation associated with a trade surplus), and tariff and non-tariff barriers have stayed essentially unchanged until very recently. Although both Korea and Taiwan have a backward financial intermediation system with low allocative efficiency, some people speculate that the Korean economy will maintain more dynamism because of the presence of large-scale firms which are in a better position to undertake expensive R&D activities, whereas Taiwan’s economy is dominated by small enterprises which have limited absorptive capability for the available investment funds. The painful experience of ultra-hyperinflation in pre-war mainland China seems to have had a lasting mental impact on the top decision makers of the ROC Taiwan government as to maintain an ultra-conservative attitude on monetary and fiscal policies. By the 1980s, the lack of proper public investment in SOC (Social Overhead Capital) facilities has resulted in extreme traffic congestion on highways, in city centres and at port facilities, deteriorating education and housing facilities, inadequate power supply and overall deterioration in the conditions for everyday living and general production activities. Belatedly, the ROC Taiwan government has decided in 1991 to invest $303 billion for SOC facilities (consisting of 779 projects) over the coming six-year period. These investment activities are expected to help Taiwan significantly to replace the role of Hong Kong after 1997. On the other hand, the Korean government has also minimized government expenditures on SOC facilities in the 1980s in order to pull down the annual inflation rate to the level of Taiwan’s. This anti-inflation policy was successful but only at the cost of deteriorating SOC facilities that constitute the most visible bottleneck for sustained growth of the Korean economy by the early 1990s. Until the late 1980s, Korean entrepreneurs were free from labour disputes because the authoritarian Korean government had always been ready to intervene on their behalf. With the progress of democratization in Korea, however, the government could not but maintain a more neutral stance while the labour union movement became militant and more systematic. At the present time, nobody can be sure whether Korea’s labour unions will become more like US-European ones or more like Japanese ones. The recent aggressive labour disputes have seriously damaged the international competitiveness of Korean industries. Depending on how the labour movement develops, the future course of Korea’s growth will be markedly different. Taiwan has a long way to go to achieve constitutional

197

ASIA

democracy. As a result, Taiwan is still free from any kind of serious labour dispute, but no one is sure how long the peaceful façade can last. The Taiwan economy, however, is dominated by family-style small and medium-sized firms and hence workers harbour little sense of grievance that is likely to arise from regimental oppression within privately owned gigantic company groups. One might, therefore, foresee little possibility of Taiwan’s workers imitating the militant behaviour of Korean workers. ACKNOWLEDGEMENTS This essay is a revised version of the paper (Hong 1991b) presented at the Center of Asian Studies Conference on ‘The Future of Industrial and Trade Development in Hong Kong,’ held at the University of Hong Kong on 4–6, January 1989. NOTES 1 The Literacy rate was 21.3 per cent in 1940. 2 The ratios increased to 1.8 per cent and 11.2 per cent, respectively, in 1958 and to 9.4 per cent and 42.6 per cent, respectively, in 1986. Data from CEPD (1987:7). 3 The enrolment rate further increased to 103 per cent by 1970 for elementary schools. In 1986, the enrolment rate for junior high schools reached 99 per cent, that for senior high schools reached 83 per cent and that for colleges and universities reached 25.5 per cent. The illiteracy rate became so insignificant after the mid 1970s that the government ceased to collect data on it (see Kim 1990:147, Table 1). 4 The enrolment rates by school-age population group increased, for the primary level, to 100 per cent in 1978; for the secondary level, to 78.7 per cent for males and 73.8 per cent for females in 1978; and for the tertiary level, to 13.5 per cent for males and 9.3 per cent for females in 1978. Data from Statistical Yearbook of the ROC (1979), Directorate General of Budget, Accounting & Statistics, Executive Yuan, ROC. 5 According to Tsiang (1984:306): ‘Taiwan’s sugar exports were practically fixed for her by the international sugar agreement that allotted world market share annually for each participating sugar-producing country. Her rice exports went exclusively to Japan, and the quantity and price were fixed each year by direct negotiation between the two governments. Thus these two major exports [which constituted 78 per cent of Taiwan’s commodity exports in 1957] were confronted with literally zero demand elasticity with respect to the exchange rates.’ Hence it used to be believed that (Tsiang 1985:35) ‘devaluation would greatly worsen the terms of trade and exacerbate domestic inflation by imparting a strong cost-push to the price level.’ 6 Compared with Korea, Taiwan was relatively better endowed with exportable agricultural resources such as sugar, rice, tea, bananas and citronella oil, and was also in a better position to earn substantial foreign exchange through the export of processed agricultural products such as canned pineapple. Hence the export drive of the ROC Taiwan government seems to have been relatively less vigorous than that of the Korean government. The share of manufactures exports in Taiwan’s total commodity exports increased from 8 per cent to 43 per cent over the four-year period of 1958–61, but in Korea it jumped from 15 per cent to 46 per cent in the two-year period of 1962–3. 7 This measure was abolished in 1958 because export producers could freely import the necessary raw materials and, furthermore, the new exchange rate applied to exports became very favourable. In 1958, the rate for export exchange settlements became the

198

KOREA AND TAIWAN

basic official rate of $NT 24.58 plus the official posted rate for the exchange settlement certificate (ESC) of $NT 1.05 per $US 1 which could be disposed of in the open market at higher rates. After 1959, the ESC issued by the monetary authority was freely negotiable in the market and could be used for various permissible imports. 8 The maximum rate of business income tax for such firms did not exceed 18 per cent (compared with 32.5 per cent for ordinary firms). See Kuo et al. (1981:75). Since 1977, the start of the period of tax exemption could be postponed by up to three years (see Balassa 1981:417). 9 The rate of reduction on income tax alone increased from 2.4 per cent in 1961 to 21.5 per cent in 1962 but then the rate fell to a level of about 15 per cent after 1969. 10 Balassa (1981:403) notes that in Taiwan ‘[w]hile short-term preferential [export] credits have long been provided, they are not granted automatically as in the case of Korea…under current regulations export credits could amount to one-fifth of the exports of the previous 12 months, in November 1977 the actual ratio was only 2.9 percent…compares with a ratio of 12.3 percent in Korea [cf. Kuo et al. 1981:79– 81]…. The amount available for medium-term export credit is also limited in Taiwan, and export credits with a maturity of over 5 years will be provided only after the establishment of the proposed Export-Import Bank.’ In Korea, the short-term export credits began to be provided only with collateral in the late 1970s. 11 Until 1964, Taiwan’s unemployment rate exceeded 6 per cent and real wages did not rise significantly. Furthermore, the decentralized industrialization enabled the utilization of part-time farmers and young female workers at low cost. According to Kuo and Fei (1985:58), these factors contributed greatly to the increased profitability in 1952–70. 12 Ho (1978:197) notes that ‘[a]t first, administrative complexities and continued government control over foreign exchange and imports reduced the effectiveness of these export incentives. However, since 1962 additional reforms and further relaxation of controls have made the export incentives progressively more effective.’ 13 The import content of exports amounted to about 40 per cent in Korea and about 58 per cent in Taiwan. See Scitovsky (1985) and Hong (1989). 14 The Statute was revised more than a dozen times by 1980, updating the categories and criteria of strategic productive enterprises for special encouragement. In 1980, the statute stated its objective as: (a) to encourage those investments fostering the improvement of commodity quality, the diversification of exports, and the enhancement of value-added exports; (b) to upgrade industrial structure through the development of technology- and capital-intensive industries; and (c) to encourage both the opening up and the efficient utilization of major natural resources (see Yu and Chen 1981:284). After 1977, selected firms could retain profits up to 100 per cent (see Balassa 1981:419). 15 In Korea, entrepreneurs have been able to maintain a system of long working hours by instituting low base wage rates and high extra rates for overtime work. 16 The interest income earned from deposits longer than two years was tax exempt until 1981. After 1981, interest income up to $NT 0.36 million per year was exempt from income tax. As for interest income from the curb market no tax is paid at all. See Hou (1988:17). More than 90 per cent of postal savings in Taiwan had to be re-deposited with the Central bank and whose absolute magnitude, by the end of the 1970s, exceeded the total outstanding amount of rediscount and refinancing facilities extended to financial intermediaries by the Central Bank. 17 In Japan, the savings/GNP ratio came close to or exceeded the 30 per cent level after the beginning of the 1960s; in Taiwan, after the early 1970s; and in Korea after the early 1980s. The usual econometric analyses fail to explain satisfactorily such a high and rising savings rate in terms of conventional variables such as real interest rate,

199

ASIA

18 19

20

21

22

23

level of income, growth rate of income, distribution of income or inflation rate, and hence let economists seek answers in culture and tradition, the absence of a fully fledged social security system, bonus systems, undeveloped capital markets, expensive housing costs, social instability, and so on. (See Sun and Liang 1981:403– 25.) Though one often finds a negative relationship between interest rates and savings in conventional econometric approaches, we may still be able to contend that, within a steady-state framework, if Korea had maintained a positive real interest rate regime since the early 1950s, it could also have experienced such a high saving regime characterized by rapid progress in financial deepening and monetization as in Taiwan by the early 1970s. See also Kuo et al. (1981:80). Tsiang (1981:265) states that: ‘If it is felt that the banks should play a bigger role in financing the industrial investments, the obvious thing to do is to set them free to compete for the big pool of total available savings and to permit freer entry into the financial intermediation, rather than to supply the banks with more reserve money in order to let them expand the money supply.’ He continues: ‘It seems to me, therefore, a big political mistake that the Central Bank was recently (1980) put directly under the jurisdiction of the cabinet and subject to the constant surveillance by the legislature. There is clearly a danger that from now on monetary and foreign exchange policies will be determined largely by popular clamour rather than by expert opinion. Powerful vested interest groups that control the mass media and have a strong lobby in the legislature would inevitably distort the monetary policy in their own favor.’ (Tsiang 1981:26 8.) As Hou (1988:20) notes: ‘A particular feature of export expansion in Taiwan is that small and medium enterprises (roughly with less than 100 employees) have played an important role in developing foreign markets. For the period from 1978 to 1985, export earnings of small and medium enterprises constituted about 65% of total export earning of Taiwan. The ratio was probably much higher in the early years.’ In Korea, the share of small and medium enterprises in total export earnings has constituted less than 40 per cent. By the mid 1970s, about two-thirds of the average farm household income in Taiwan came from non-farm activities. That is, farming in Taiwan became a small-scale parttime activity with little hope of attaining international competitiveness in productivity. Scitovsky (1985) notes that ‘[t]he proportion of workers employed in manufacturing who lived in rural areas as part of farm households and commuted daily on a seasonal or full-time basis grew steadily and constituted over half of the work force by mid1960s’. Balassa (1981:399) notes that ‘[p]ublic enterprise generally have first claim on investment credit…. Also, large private firms manufacturing [highly physical capitalintensive] intermediate products [which sell mostly in the protected domestic market] have greater access to credit than the generally smaller firms producing machinery and equipment.’ Balassa (1981:413) argues that ‘generally, it is not desirable to run a current account surplus that involves the outflow of capital from a capital-poor country to capitalrich countries. With the social marginal productivity of capital estimated at 15 percent… Taiwan can borrow at a rate only slightly above LIBOR that hardly exceeds zero in real terms’.

BIBLIOGRAPHY Academia Sinica, The Institute of Economics (1981) Conference on Experiences and Lessons of Economic Development in Taiwan, Taipei: Academia Sinica.

200

KOREA AND TAIWAN

Balassa, Bela (1971) ‘Industrial policies in Taiwan and Korea’, Weltwirtschaftliches Archiv, Bank 106:55–77. ——(1981) The Newly Industrializing Countries in the World Economy, New York: Pergamon. Council for Economic Planning & Development (CEPD) (1987) Taiwan Republic of China: Economic Development. ——(1989) Taiwan Statistical Data Book. Council for International Economic Cooperation and Development (CIECD) (1973) Taiwan Statistical Data Book: 1973, Taipei: Executive Yuan, ROC, June. Economic Planning Board (EPB) (1988) Social Indicators in Korea. Economic Research Center, Council for U.S. Aid (ERC, CUSA), (1963) Taiwan Statistical Data Book: 1963, Taipei: Executive Yuan, ROC, June. Galenson, Norton, S. ‘Taiwan: A resource analysis of an oriental society’, Economic Development and Cultural Change, June 110–31. Galenson, Walter, (ed) (1979) Economic Growth and Structural Change in Taiwan, Ithaca, NY: Cornell University Press. ——(1981) ‘How to develop successfully: The Taiwan model’, in Conference on Experiences and Lessons of Economic Development in Taiwan, Taipei: Academia Sinica. Ginsberg, N.S. (1952) ‘Taiwan: A resource analysis of an oriental society’, Economic Development and Cultural Change, June: 110–31. Gold, Thomas B. (1986) State and Society in the Taiwan Miracle, Armonk, NY: Sharpe. Ho, Samuel P.S. (1978) Economic Development of Taiwan 1860–1970, New Haven, CT: Yale University Press. Hong, Wontack (1979) Trade, Distortions and Employment Growth in Korea, Seoul: KDI Press. ——(1986) ‘Institutionalized monopsonistic capital market in developing economy’, Journal of Development Economics 21:353–9. ——(1987) ‘Theory of interest and the steady-state rate of return on capital’, International Economic Journal Autumn:87–90. ——(1989) ‘Factor intensities of Korea’s domestic demand, production and trade: 1960– 85’, International Economic Journal Summer: 97–113. ——(1990a) ‘Market distortions and polarization of trade patterns: Korean experience’, in Jene K. Kwon, (ed.) Korean Economic Development, New York: Greenwood Press, 115–28. ——(1990b) ‘Export-oriented growth of Korea: A possible path to advanced economy’, International Economic Journal Summer: 97–118. ——(1991a) ‘Import restriction and import liberalization in export-oriented developing economy: in light of Korean experiences’, in Lawrence B.Krause and Kihwan Kim (eds) The Liberalization Process in Economic Development, Berkeley, CA: University of California Press. ——(1991b) ‘Comparative study of the industrialization experience of Korea and Taiwan’, Proceedings of the International Conference (4–6 January 1989) on the Future Industrial and Trade Development in Hong Kong, Center of Asian Studies, University of Hong Kong, July. Hong, Wontack and Park, Yung Chul (1986) ‘Financing export-oriented growth in Korea’, in Augustine Tan and Basant Kapur (eds) Pacific Growth and Financial Interdependence, Sydney: Allen & Unwin. Hou, Chi-ming (1988) ‘Strategy for economic development in Taiwan and implications for developing economies’, paper presented at the Conference on Economic Development Experiences of Taiwan, Taipei, June 8–10. Hou, Chi-ming and Chang, Ching-hsi (1981) ‘Education and economic growth in Taiwan: The mechanism of adjustment’, Academia Sinica 469–531.

201

ASIA

Keesing, Donald B. (1967) ‘Outward-looking policies and economic development’, Economic Journal June. Kim, Linsu (1990) ‘Korea: The acquisition of technology’, in Hadi Soesastro and Mari Pangestu (eds) Technological Challenges in the Asia-Pacific Economy, Sydney: Allen & Unwin, pp. 145–57. Kuo, Shirley, W.Y. (1988) ‘The achievement of growth with equity’, paper presented at the Conference of Economic Development Experiences of Taiwan, Taipei, 8–10June. Kuo, Shirley, W.Y. and Fei, C.H. (1985) ‘Causes and roles of export expansion in the Republic of China’, in W.Galenson (ed.) Foreign Trade and Investment, Madison, WI: University of Wisconsin Press. Kuo, Shirley W.Y., Ranis, Gustav and Fei, John C.H. (1981) The Taiwan Success Story, Boulder, CO: Westview Press. Li, Kwoh-ting and Yeh, Wan-an (1981) ‘Economic planning in the Republic of China’, in Conference on Experiences and Lessons of Economic Development in Taiwan, Taipei: Academia Sinica, pp. 169–95. Liang, Kuo-shu and Liang, Ching-ing Hou ‘Trade and incentive policies in Taiwan’, Conference on Experiences and Lessons of Economic Development in Taiwan, Taipei: Academia Sinica, pp. 313–49. ——(1985) ‘Trade strategy and industrial policy in Taiwan’, Academia Economic Papers September: 329–66. Little, Ian M.D. ‘An economic reconnaissance’, in W.Galenson (ed.) Foreign Trade and Investment, Madison, WI: University of Wisconsin Press, pp. 448–507. McGinn, N.F., Snodgrass, D.R., Kim, Y.B., Kim, S.B. and Kim, Q.Y. (1980) Education and Development in Korea, Cambridge, MA: University Press Harvard. Mason, Edward S., Kim, M.J., Peerkins, D.H., Kim, K.S. and Cole, D.C. (1980) The Economic and Social Modernization of the Republic of Korea, Cambridge, MA: Harvard University Press. Ministry of Finance (MOF) (1989) ’88 White Paper on the Reform of Tariff Rates (in Korean), Bureau of Tariff, February. Myint, H. ‘Comparative analysis of Taiwan’s economic development with other countries’, in Conference on Experiences and Lessons of Economic Development in Taiwan, Taipei: Academia Sinica, pp. 107–30. Ranis, Gustav ‘Industrial development’, in W. Galenson (ed.) Foreign Trade and Investment, Madison, WI: University of Wisconsin Press, pp. 206–62. Scitovsky, Tibor ‘Economic development in Taiwan and South Korea: 1965–85’, Food Research Institute Studies XIX:(3). Sun, Chen and Liang, Ming Yih (1981) ‘Saving in Taiwan, 1953–1980’, in Conference on Experiences and Lessons of Economic Development in Taiwan, Taipei: Academia Sinica, pp. 403–25. Tsiang, S.C. (1981) ‘Monetary policy of Taiwan’, in Conference on Experiences and Lessons of Economic Development in Taiwan, Taipei: Academia Sinica, pp. 249–69. ——(1984) ‘Taiwan’s economic miracle: Lessons in economic development’, in A.C. Harberger (ed.) World Economic Growth, San Francisco: ICS Press. ——(1985) ‘Foreign trade and investment as boosters for take-off; The experience of Taiwan’, in V.Corbo, A.O.Krueger and F.Ossa (eds) Export-Oriented Development Strategies, Boulder, CO: Westview Press, pp. 27–56. US Department of Commerce (US DC) Investment in Taiwan: Basic Information for United States Businessmen, Washington, DC. Yu, Tzong-shian and Chen, Ting-an (1981) ‘Fiscal reforms and economic development’, in Conference on Experiences and Lessons of Economic Development in Taiwan, Taipei: Academia Sinica, pp. 275–305.

202

11 ECONOMIC REFORMS UNDER AN IMPORT-SUBSTITUTION REGIME The experience of Bangladesh Alia Ahmad

INTRODUCTION Bangladesh is one of the poorest countries in the world with a per capita income of $US 210 (World Bank 1992). The per capita income has remained more or less stagnant for many years as a result of the slow rate of growth of GDP and a rapid rate of population growth. After independence from Pakistan in 1971, it took several years for Bangladesh to recover from the damages of the Liberation War during 1970–1. The First Five Year Plan 1973–8 initiated the huge task of reparation but it was not until 1975 that GDP was restored to its 1969 level. Hence, the 4 per cent average annual growth rate during this period led to a marginal improvement in the country’s standard of living. In the early 1980s, the country was faced with severe economic problems like mounting external debt, huge fiscal and current account deficits and low rates of growth of value added in key sectors of the economy. These problems were precipitated by international recession, declining terms of trade and a stagnant flow of external resources. To cope with these problems, the World Bank and the IMF recommended Bangladesh to undertake major policy reforms. Concessional loans by the World Bank and the IMF’s Enhanced Structural Adjustment Facilities (ESAF) were made conditional on policy reforms being undertaken. The purpose of this study is to demonstrate that the reforms in the 1980s have had very little effect on the growth of the economy although the short-run goals of stabilization have been partially met. The main reason, as will be argued below, is that a fully-fledged reform has not been possible because of the emphasis placed on industrialization through import substitution. This policy has intensified distortions in the economy which, in turn, have resulted in an inefficient allocation of resources and a level of output lower than the country’s potential. The outline of the study is as follows. The next section describes the evolution of the import-substitution regime and the main characteristics of the reforms in the 1980s; the third section considers the macroeconomic effects of the reforms and the fourth section analyses the effects of import substitution in the key

203

ASIA

sectors. The fifth section deals with the motives behind import substitution in Bangladesh and the reasons for the frustrated reforms. The last section contains a summary and conclusions. IMPORT SUBSTITUTION AND INDUSTRIALIZATION IN BANGLADESH UNTIL 1980 Import substitution in Bangladesh was a legacy which the country inherited. Hence, it is necessary to take a historical view in order to understand the interplay of various factors in the evolution of the import substitution policy. During the British colonial period, Bangladesh lost its comparative advantage in the manufacture of cotton textiles and the economy became more specialized in the production of jute in which she had a comparative advantage.1 Institutions and infrastructure were created to facilitate the export of jute, but with regard to education, infrastructure and the institutions which were essential for industrial development, the country remained underdeveloped. The period 1947–71 After independence from British rule, economic development in Bangladesh was mainly influenced by the trade and industrialization policies adopted for the whole country, i.e. Pakistan. At the time of independence (1947), there were very few industries in East or West Pakistan. Consequently, most of the demand for consumer goods was met by imports. In the early 1950s, especially after the Korean War (1952), Pakistan faced a severe shortage of foreign exchange. In view of Pakistan’s unwillingness to devalue the currency, strict control of imports and foreign exchange were adopted in order to cope with balance of payments problems.2 Industrialization first became the dominant strategy and the Industrial Policy Statement of 1949 emphasized the role of the private sector which was weak at that time. However, a group of merchants and traders who migrated from India after Partition took advantage of these policies and the changed circumstances. Hence, numerous import-substituting industries were established mainly around Karachi which was the capital of Pakistan during that time. In East Pakistan, a number of agro-based industries such as jute, cotton textile and food manufactures also began to develop slowly. Economists who have studied Pakistan’s industrialization are of the opinion that import substitution in the 1950s did not create too much distortion or an inefficient industrial structure because the rules were of uniform nature. A similar industrial structure (simple consumer and intermediate goods based on local raw materials) would have grown up under a less controlled regime (Lewis 1970). For several reasons, there was a shift towards a more complicated pattern of import substitution in the late 1950s. First of all, the emergence of the military regime under the leadership of Ayub Khan provided not only a degree of stability but also a clear-cut development strategy based on a mixture of planning and the

204

THE EXPERIENCE OF BANGLADESH

private sector. By this time, Pakistan had also developed its civil service, bureaucracy and the planning machinery. All these factors enabled the government to attract more foreign aid. President Ayub Khan aspired to develop the country through industrialization by redistributing resources from agriculture to industry and to a handful of large entrepreneurs. A complex system of tariffs, quotas, import licences and foreign exchange control was designed, which affected different industries in a discriminatory fashion. From the late 1950s to the mid 1960s, Pakistan adopted certain export-promotion measures such as the Export Bonus Scheme which enabled the exporters of manufactures to sell their foreign exchange at a higher than official exchange rate and to import inputs freely. However, the system in general worked through direct controls in spite of the recommendations by the Planning Commission for indirect controls (Islam 1981). The modest attempts at liberalization and export promotion in the early 1960s were also tempered by the foreign exchange crisis during the war with India in 1965 when more direct controls were imposed. Although the rate of growth of manufacturing industries was very high as a result of the industrialization drive in the 1960s, Pakistan developed an inefficient industrial structure which could not compete in the international market without heavy subsidies from the government (Islam 1981). The utilization of the huge capacity which was created by the import licensing system remained low. The worst part was that this strategy of development created extreme social and political pressures (inter-household and regional inequality increased sharply) which ultimately led to the break-up of the nation and the creation of Bangladesh in 1971. Post-liberation period The major change in the 1970s was increased government intervention in the economy spurred by the developments of the 1960s. The economic domination of East Pakistan by West Pakistani industrialists (most of the large industries were owned by West Pakistanis) created a mistrust among East Pakistanis against capitalist development. Thus, after the Liberation War, the Awami League Government led by Sheikh Mujibar Rahman followed a socialist path which reflected the wishes of different groups such as workers, students, political leaders and the intelligentsia. The major industries, banks and insurance companies were nationalized and strict state control prevailed over international trade and payments. Import substitution became reinforced due to increased state control. Towards the second half of the decade, especially after the assassination of Sheikh Mujib in 1975 and the military coup, conservative forces gained ground in Bangladesh, and certain moves towards liberalization took place. For example, the ceiling on private investment in the industrial sector was raised from Taka 2.5 million to Taka 30 million in 1974 and later on to Taka 100 million in 1975.

205

ASIA

Between 1975 and 1980, two major devaluations were undertaken. However, extensive government control in the trade sector continued. THE REFORMS IN THE 1980s In the early 1980s, Bangladesh agreed to adopt various stabilization and liberalization measures in order to qualify for concessional loans from the World Bank and the IMF. The reform package included fiscal reform, financial liberalization, maintenance of a realistic and flexible exchange rate, together with trade liberalization, reduced government intervention and improved management of public enterprises. The main objective of the reform was to stabilize the economy, remove distortions and achieve the steady economic growth which could alleviate poverty. During the ten years since the reforms were initiated, it is evident that only some of them have been carried out. Since import substitution has remained the dominant strategy and has constrained the growth of the economy, we shall concentrate on the actual reforms with respect to industries, trade and foreign exchange. Industrial policies The most pervasive reform in the industrial sector was the privatization of public sector enterprises in the large-scale manufacturing sector. After the adoption of the New Industrial Policy (NIP) in 1982, a massive sale of public enterprises to private owners took place. Over 650 public enterprises were transferred to the private sector during the period 1981 to 1986 and the share of the public sector in fixed industrial assets declined from 85 per cent to 40 per cent (World Bank 1990).3 Second, in order to promote private activities procedures for sanctioning investment and the import of necessary inputs were also simplified. Private entrepreneurs were given various fiscal and financial incentives such as tax holidays, cheap loans and different types of allowances under the Revised Industrial Policy (RIP) 1986. Specialized credit institutions known as Development Finance Institutions were responsible for extending credit to private industrialists at concessional rates.4 Third, industrialization policies have continued to stress import substitution as before. Although specific measures have been taken to boost exports of selected products (which will be discussed later in this section), most of the industries in both the public and private sectors are still geared to the domestic market. The dominance of the import-substitution strategy together with certain counteracting measures to boost exports is reflected in the nature of the reforms with respect to trade and foreign exchange.

206

THE EXPERIENCE OF BANGLADESH

Trade policies Until the early 1980s, Bangladesh had a highly restrictive import regime with extensive quantitative restrictions—either quota or complete bans on certain items. This was administered through a narrowly defined positive list of eligible imports for which licences were issued by the Ministry of Commerce and a tightly controlled official foreign exchange market. In the 1980s, a certain relaxation of controls took place in the following areas: 1

2 3 4

5 6

In 1986, the introduction of a progressive reduction of the number of items covered by the list of banned items from 39 per cent of import categories at the four-digit SITC level at the end of 1986 to 28 per cent at the end of 1988. A gradual replacement of quota restrictions by tariffs starting from 1982. Simplification of the tariff structure by a reduction of the number of tariff rates from twenty-four to eleven in 1988. In 1988, reduction of nominal tariffs on final goods from 200 per cent to 125 per cent. However, tariffs were also reduced on raw materials and intermediate imports even before 1982. From 1986, wage earners abroad were allowed to import freely a wide variety of consumer and intermediate goods. From 1987–8, some of the imports could be financed through foreign exchange acquired in the Secondary Foreign Exchange market or through exports of manufactures.

It should be mentioned that many of these measures did not become effective until 1989. Throughout the 1980s, there were consequently high tariffs on final goods but low tariffs on inputs and quantitative restrictions. In the early 1990s the list of banned and restricted imports was substantial and nominal tariffs were high on final goods. In spite of the simplification of import rules, the system remains complex and bureaucratic. Exchange rate policy Bangladesh has pursued a flexible exchange rate policy since mid 1979, which meant that the taka was periodically devalued. After 1983, the US dollar replaced pound sterling as the intervention currency and the value of the taka was pegged to a basket of currencies of her major trading partners.5 The aim of this change was to facilitate a gradual adjustment of the nominal exchange rate to the fluctuations in the currencies of Bangladesh’s major trading partners. The taka has, accordingly, been devalued several times during the 1980s in response to domestic inflation. Since January 1992, the value of the taka has been 38.08 to the US dollar. Flexibility is also reflected in the greater involvement of the secondary foreign exchange market in financing imports. In addition to the official nominal exchange rate there is a special higher rate for the benefit of wage earners abroad (Wage Earners’

207

ASIA

Scheme—WES). Although with the adjustment of the nominal exchange rate the wage earners’ rate has been adjusted upwards, the foreign exchange market has become more unified because the differences between the official exchange rate, the wage earners’ rate and also the free market rate have narrowed during the 1980s. In spite of these improvements, the taka remains overvalued. The rate of depreciation of the currency has been barely enough to counteract the high inflation in Bangladesh in comparison with her major trading partners.6 The real effective exchange rate has fallen. This has been the case particularly during 1988–9. The gap between the official nominal rate and the free market rate has increased in recent years. The continuation of import control through licences, quotas and high tariffs also indicates the overvaluation of the taka. Policies towards exports In order to counteract the disincentives to export that have been created by import-substitution policies, certain measures have been undertaken to encourage export-oriented industries. These export-oriented industries may emerge out of private initiatives and/or promotional efforts taken by the government. For example, garments industries have sprung up from private initiatives facilitated by both improved domestic policies and the comparative advantage of Bangladesh’s cheap labour.7 Some large industries, such as the paper and fertilizer industries, which were geared from the beginning to the domestic market have also ventured into the export market in recent years. On the other hand, the government often identifies certain sectors as having a comparative advantage. These sectors (toys, electronics, frozen food, leather goods) are then called ‘thrust sectors’ for which various incentives are provided (Bangladesh Economic Survey 1989/90:152–3). In addition to the reforms in the foreign exchange market, incentives are provided to foreign investors to invest in the Export Processing Zone.8 Domestic exportoriented industries are also given facilities such as simplified import rules, duty drawbacks on imported inputs, suppliers’ credit and Export Benefit Entitlement which allows exporters to use a certain percentage of their export earnings to import inputs and final goods without conversion into domestic currency. Summary The major policy changes during the 1980s consist of the privatization of public sector enterprises, modest attempts to maintain a realistic exchange rate and introducing changes towards a unification of the foreign exchange market. Progress with respect to trade liberalization had remained limited until the end of 1988. Although several export-promotion measures were undertaken, import substitution under heavy protection continues to be the dominant strategy. Reforms in other areas did not materialize except in agriculture where subsidies on fertilizers were reduced and private trading was allowed.

208

THE EXPERIENCE OF BANGLADESH

MACROECONOMIC EFFECTS OF REFORMS One of the major goals of the economic reforms was to achieve macroeconomic balance. This was considered to pave the way to economic growth. The macroeconomic indicators show that Bangladesh has achieved some success in reducing deficits in the current and fiscal accounts (Figure 11.1). As a percentage of GDP the current account deficit has fallen from 10 per cent in 1981 to 6.7 per cent in 1989. The reduced gap in the government’s budget is less impressive, but still a success. It fell from 9.0 per cent to 7.2 per cent between 1981 and 1989. An analysis of the factors behind this improvement can be made using information regarding exports, imports and the budgetary position of the central government. There has been a substantial improvement in the current account balance. The deficit in absolute terms fell from 1,428 million US dollars in 1981 to 1,376 million US dollars in 1989. The figures were particularly low in 1984 and 1987. One of the main factors behind this reduction was increased earnings in the export sector and the remittance of migrant workers, both of which have been boosted by the flexible exchange rate policy and relaxed import rules.9 The annual average rate of growth of exports (Figure 11.2) during 1981–8 was 11 per cent. The major contributor was the non-traditional export sector such as garments and frozen food with a 29 per cent average annual rate of growth at 1981 prices. The flexible exchange rate policy and a certain relaxation of import controls in the second part of the 1980s have had a positive effect on imports of consumer goods. This was exacerbated by an appreciation of the real exchange rate and expansionary monetary and fiscal policies in the latter part of the 1980s.10 Food imports increased at the rate of 17 per cent annually. However, figures for total imports in Bangladesh show that in comparison with exports, imports have grown

Figure 11.1 Current account and budget deficits (per cent of GDP) Source: Ministry of Finance and Bangladesh Bank

209

ASIA

Figure 11.2 Growth of exports (per cent) Source: Bangladesh Bureau of Statistics

at a low rate of 3.9 per cent annually. This was mainly due to a negative rate of growth (–2.0 per cent) of imports of capital goods and a marginal increase (0.9 per cent) in the imports of intermediate goods. It shows that the improvement in the current account balance took place as a result of export growth and reduced imports of investment goods. The demand for imports was also contained to a certain extent through credit restrictions in the private sector which affected private investment. The curtailment of imports of capital and intermediate goods has caused a decline in the rate of investment in the economy (Figure 11.3). Gross domestic investment fell from 15.9 per cent of GDP in 1981 to 11.8 per cent in 1988. It is evident from the central government financial data that the government sector bears the major burden of the recession in the economy. The improvement in the budgetary position of the government was made possible by a curtailment in development expenditures which fell from 10.2 per cent of GDP in 1981 to 5.9 per cent in 1989, whereas current expenditure has grown from 5.6 per cent to 8.7 per cent during the same period. The fall in annual development expenditure occurred due to the failure to generate surplus revenue. Some of the foreign-aided projects did not materialize because of the inadequacy of local funds. The decline in annual development expenditure adversely affected investment in the private sector which fell from 9.5 per cent of GDP in 1981 to 5.9 per cent in 1988. The slow growth of investment is reflected in the rate of growth of the economy which declined drastically after 1987 (Figure 11.4). While the poor performance of agriculture in three consecutive years due to bad weather is the main reason behind slow GDP growth, the manufacturing sector also performed poorly. The contribution of the non-tradable sectors such as construction, utilities and services, which include trade and social services, is strikingly high. The structural changes in the economy between 1981 and 1990 consist of a declining

210

THE EXPERIENCE OF BANGLADESH

Figure 11.3 Investments (per cent of GDP) Sources: Bangladesh Bureau of Statistics and World Bank estimates

share of agriculture and manufacturing with a corresponding rise in the share of the non-tradable sector. The slow rate and unhealthy pattern of growth of the economy is not only due to a declining rate of investment but also the way resources have been allocated and used. We shall argue below that an inefficient allocation and use of resources have largely been the result of import-substitution policies which allow distortions to permeate the product and factor markets. While our discussion is mainly limited to the manufacturing sector where import-substitution has been pursued, the adverse effects of import substitution on other sectors of the economy are also considered.

Figure 11.4 GDP growth by sector (average annual growth rates) Source: Bangladesh Bureau of Statistics

211

ASIA

COSTS OF IMPORT SUBSTITUTION It is well known that import substitution under high protective tariffs and import quotas entails heavy costs to society which are classified as production costs, consumers’ loss and losses from DUP activities (Krueger 1978, Bhagwati 1982). The extent of the costs depends on the height and the structure of protection in the trade sector and distortions in the factor markets which may be either created and/ or sustained through deliberate policies. We shall now consider the costs inflicted by import-substitution policies on the economy in Bangladesh. The nature of protection A conventional method of measuring the extent of protection is that of effective rates of protection (ERP) which takes into account import quota and tariffs on both inputs and outputs. The Harvard Institute for International Development and the Planning Commission of Bangladesh have studied 21 major industries in Bangladesh and found that the average effective rates of protection received by these industries have steadily increased in the 1980s compared with the mid 1970s (see Figure 11.5). The actual protection received by large industries depends on both tariffs and quotas on imports and domestic policies which affect factor prices. In addition to protection in the trade sector, industrialists in Bangladesh are given various fiscal and financial incentives which distort factor markets. Capital market distortions are reflected in the effective costs of capital faced by different producers. The financial market in Bangladesh has been found to be dualistic (Chowdhury and Rahman 1989). In the formal sector, institutional loans are provided by commercial banks (nationalized) and specialized

Figure 11.5 Effective rates of protection and effective rates of assistance (per cent) Source: Sahota and Huq (1991)

212

THE EXPERIENCE OF BANGLADESH

institutions at subsidized rates of interest which vary between 10 and 14 per cent. The informal financial sector, on the other hand, operates under market forces although signs of imperfections are visible in the form of segmentation of the market and monopoly profits (Chowdhury and Rahman 1989). In both the public and private sectors, large enterprises meet their demand for credit through the formal market. Although it has been the policy of the government to supply cheap credit to small enterprises in industry and agriculture, only a small part of their credit needs are met by institutional loans. Small enterprises depend mainly on their own savings and loans from the informal market where the interest rates vary between 50 and 100 per cent. The main point is that distortions arise because the costs of and access to institutional loans faced by different borrowers are not determined by economic factors such as risk but by the power and influence of the borrowers. Empirical studies indicate that small farmers and manufacturers are discouraged from borrowing from the institutional sources because the officials demand bribes which also cause substantial delays (Bhattacharya 1991, Islam 1989). On the other hand, large industries were able to borrow heavily from the Development Finance Institutions (DFIs) throughout the 1980s and the recovery rates for these loans are extremely low. In spite of the repeated warnings of the donors, the government has not been able to deal with the defaulters of the DFIs which are mainly supported by foreign aid (Sobhan and Sen 1989). Imperfections in the capital market are exacerbated by differential access to foreign exchange and imported inputs faced by different producers. Import licences and foreign exchange are easily procured by large enterprises in the public and private sectors compared with small enterprises. The glaring example is the handloom sector which is in constant difficulties regarding the supply of imported raw cotton (N.Chowdhury 1989). The wage/rental ratio faced by different producers is affected by both capital and labour costs. Labour cost is affected by the minimum wage laws existing in the country and the degree of unionization among the workers. According to government regulations in Bangladesh, different floors on wages are set according to the particular industry and skills of workers. However, more than 60 per cent of the industries are not covered by such regulations (Anderson et al. 1991). Unionization is also limited to a few industries, mostly large, nationalized ones and with a low proportion of female workers.11 It is difficult to ascertain whether the labour market is distorted because of the lack of sufficient information regarding the wage-setting pattern in different enterprises. Available evidence suggests that while wages are generally higher in the covered sector, minimum wage laws are not followed uniformly.12 Some studies indicate that public sector enterprises incur a higher cost of labour due to trade union pressure (Hossain and Jamal 1989). Private sector enterprises in the same industry also have to pay the same wage, for example jute mills in the public and private sectors cannot differ too much in terms of employment and wage rates of workers. On the other hand, the garment industries pay a much lower wage

213

ASIA Table 11.1 Effective rates of assistance by groups of industries, 1989

Source: Sahota (1991)

214

THE EXPERIENCE OF BANGLADESH

rate. This has been possible partly because most of the workers are women who have weak bargaining power and also because of the strong inducement among employers to reduce labour costs to withstand international competition. The study of Anderson et al., on the other hand, indicates that unionization had no effect on wage increases but has caused work stoppages. The effects of work stoppages on output did not affect employment, which may be due to the attempts of the firms to substitute more contract labour for wage labour. On the whole, distortions in the labour market seem to be less than in the capital market in view of the excess supply of unskilled workers, the low degree of unionization, the low minimum wage and its limited coverage in Bangladesh. To take account of the effects of both domestic policies and trade policies, the Harvard Institute for International Development and Employment and Small Enterprise Policy Planning (HIID/ESEPP) study has used another measure, called the effective rates of assistance (ERA). This measure reflects the differences between the value added when tariffs and various subsidies (on capital, foreign exchange, tax exemptions) are used and the value-added in their absence. Recognizing the fact that it is difficult to measure accurately the extent of protection received by industries through such methods, the results of this study generally confirm that the protection received by industries was higher in the 1980s than in the 1970s. Second, these rates have been not only high but also highly discriminatory among different industries and firms. Generally, largescale industries producing for the domestic market enjoyed higher ERA than export-oriented activities and agriculture. Within a given industry, different branches faced different degrees of protection. Thus, as in Table 11.1 the protection enjoyed by the cotton yarn industry ranged between 113 and 513 whereas the handloom industry experienced negative actual protection (due to smuggling) against a statutory rate of 153 (Sahota 1991).13 The HIID study has not found any difference in the ERA enjoyed by large versus small enterprises. The study explained that large firms’ advantages in fiscal and financial matters over small firms are outweighed by the negative effects of labour regulations which the small firms can easily avoid (Anderson et al. 1991). One should be sceptical of this result in view of the fact that in Bangladesh distortions are more severe in the capital market than the labour market. We are going to consider below the costs to society arising from the discriminatory policies of the government. Production losses Under competitive conditions, the introduction of tariffs or quantitative restriction can give rise to production losses because resources are pulled into the protected sector from their more efficient uses. In a monopoly situation, the effects can be different with the extent of losses depending on the incentive structure and the behaviour of the monopolists. Bangladesh has a monopolistic or oligopolistic situation in the formal

215

ASIA

manufacturing sector which has been created through conscious policies. During the time of Pakistan, industries (jute and cotton textiles, fertilizer, sugar, paper, steel and food manufacturing) were established through government initiative under heavy trade protection and domestic incentives. The monopolistic tendencies and import-substitution efforts were intensified after liberation in 1971 when even the few enterprises which were privately owned were nationalized. The move towards privatization started from the mid 1970s when the ceiling on private investment was raised, and in the beginning of the 1980s almost half of the jute and textile mills were transferred to private hands. But this has not resulted in a reduction of monopolistic elements and increased competition. Foreign competition is eliminated through trade restrictions while domestic competition is hindered through a system of industrial licensing and various fiscal and financial privileges directed at specific groups of entrepreneurs. It should be mentioned that the public sector still plays a major role as it owns nearly half of the fixed industrial assets. An inefficient industrial structure has also emerged from trade and industrialization policies. The capital/labour ratios in most industries such as basic metal, petroleum, cement, paper and fertilizer are very high. Some of these industries such as basic metal and paper are capital intensive by nature and inappropriate for Bangladesh. But even the agro-based or natural resource-based industries which were considered to have good prospects of being efficient in the long run, such as jute, cotton, sugar, cement, fertilizer and food manufacturing, exhibit high capital intensity. Distorted factor prices and import licensing rules encouraged capital-intensive techniques and the overexpansion of capacity. The overexpansion of capacity is an acute problem with jute textiles. The total world demand for jute products was overestimated by not taking into account the competition from synthetic substitutes. There have also been specific problems with respect to the supply side. The quality of raw jute in Bangladesh deteriorated because of the neglect of jute research and discrimination against jute growers, especially in the 1960s and 1970s. Now Bangladesh cannot compete with India in quality and price. In the face of stiff competition and sluggish demand, the appropriate policy would be to close down some of the most inefficient mills. This has been difficult because of the fear of displacing the public sector employees and consequent social unrest. Inefficient private enterprises can also continue because of the lenient attitude of the government for similar reasons. Jute, however, is an export industry and cannot be considered on the same basis as other import substituting industries like cotton, sugar or food manufacturing. Because of the absence of price support for its goods in the international market, jute experiences negative protection compared with domestic-oriented industries. Low-capacity utilization is a major problem in most of the large import substituting industries which face a limited domestic market. The problem of limited demand can be attributed, in turn, to their inability to supply quality products at competitive prices. The protection through high tariffs seems to have

216

THE EXPERIENCE OF BANGLADESH

backfired because it has induced smuggling of various consumer goods which are often cheaper and of better quality. In spite of this the inefficient firms working at low capacity in both the public and private sectors continue to survive because of the support of the government. The inefficient use of resources within the industrial sector is further reflected in the fact that small and cottage industries use highly labour-intensive, primitive methods because of the lack of capital, while large firms have unused capacity. Studies of small firms in metal fabrication and handlooms confirm that technological upgrading is essential to raise the productivity of these activities which have the potential to increase output along labour-intensive lines (Bhattacharya 1989, N.Chowdhury 1989). The restructuring of industries along more labour-intensive lines and an evolution of a more efficient industrial structure require restoration of competition and correct factor prices. It is evident from the rapid growth of garment industries in recent years that sound policies Table 11.2 Indices of industrial production by sector (1973–4=100)

Source: Bangladesh Bureau of Statistics, Monthly Statistical Bulletin of Bangladesh

217

ASIA

such as the relaxation of controls in the foreign exchange market do reduce distortions and promote labour-intensive activities. These problems are clearly reflected in the pattern of growth of the manufacturing sector. Its share of GDP has remained stagnant at below the 10 per cent level after three decades of industrialization through import substitution. The annual rate of growth of value added in manufacturing was only 1.3 per cent during 1981–6 when the major reform (privatization) took place. The indices of production in major industries (Table 11.2) such as textiles, paper, petroleum, basic metal and non-electrical machinery fell. Some improvement has been possible in certain products such as cement, fertilizer, electrical machinery and beverages excluding tea. Some of the policies in the 1980s such as the flexible exchange rate, the unification of the foreign exchange market and various export-promotion measures have exerted positive effects on the growth of exports. The traditional exports like raw jute, tea and leather have benefited from a flexible exchange rate and lower export duty. Non-traditional exports—garments and frozen foods— have experienced phenomenal growth as a result of the additional measures described earlier. While this is certainly a desirable outcome, one has to bear in mind that the export sector is only a very small part of the whole manufacturing sector. The share of the two major non-traditional export industries in the total value added by the manufacturing sector is only 2.5 per cent. The linkage effects of the main non-traditional export (garments) on the economy are not too great because of its high import intensity. The net export earnings from garments is much less than is shown by the export figures. This is mainly because of the failure of the domestic textile industry to supply high-quality material at competitive prices. Effects on agriculture It is well known that import-substituting industrialization results in discrimination against agriculture. In the early phase of industrialization (between 1960 and 1975), agriculture in Bangladesh suffered a great deal from a highly overvalued exchange rate, low output prices and taxes on agricultural exports. Some of these problems have been obviated with the adoption of the flexible exchange rate and the reduction of export taxes. However, because of the pursuit of import substitution policies agriculture still suffers from discrimination. The output prices for several agricultural products are still below the international prices. While agriculture receives subsidies on credit and inputs, the volume of credit and the magnitude of subsidies are less than they are for industry. The negative effects of import-substitution policies also arise from the incentives among farmers to divert private agricultural surplus to trading and speculative activities (which are due to trade protection)14 and the inability of the industrial sector to supply simple consumer goods which may serve as incentive goods.

218

THE EXPERIENCE OF BANGLADESH

Summary In Bangladesh, resources have been channelled to import-substituting large-scale industries at the cost of small-scale, export-oriented industries and agriculture. Output would have been higher in less distorted market conditions. It is evident from our discussion that this policy has not resulted in any dynamic gains such as economies of scale. In addition to losses from misallocation and the underutilization of resources, losses also arise from inefficient use of resources due to organizational failure and a lack of incentives to reduce costs. Proper utilization of capacity and the efficient use of resources would have been possible if the producers faced a larger market and were subject to competition. A gradual movement towards a more appropriate industrial structure is also being hindered by the industrial policies of the government which emphasizes new investments in the public sector (Fourth Five Year Plan 1990–95). Consumers’ losses Consumer loss arises from the substitution away from the protected sector where the prices of consumer goods are raised by the amount of the tariffs. Free trade and competition would have resulted in lower prices and higher consumption. Information, although incomplete, suggests that smuggling on a large scale has a certain depressive effect on prices. However, competition in smuggling may not be high enough to drive prices down to the level without tariffs. This means that consumers do suffer some losses which are, however, lower than is implied by the tariffs and quotas (HIID/ESEPP Project 1990). Losses from DUP activities Some of the losses due to trade restrictions may appear as transfers, for example higher tariffs result in government revenue while consumers’ surplus is reduced because of the higher prices. However, this transfer may turn into welfare losses if import licensing gives rise to rent-seeking and revenue-seeking activitites (Bhagwati 1982). Such losses may result from both legal and illegal activities. Legal activities such as rent-seeking (created through licences) import trading can divert resources away from productive activities. In Bangladesh, the pattern of private investment and the growth of the service sector, especially trading in imports, confirm the existence of legal DUP activities because of import substitution behind prohibitive tariffs. Several studies on agriculture indicate that farmers are more interested in investing their own surplus in urban activities than in farming because of the higher rate of return in the case of the former (IFRI and BIDS 1985, Hossain 1988). Besides legal DUP activities, various illegal activities are common such as over- and underinvoicing of imports and exports, bribing customs officials and

219

ASIA

smuggling. The welfare effect of smuggling is ambiguous because it may increase welfare when it eliminates tariffs and legal trade through lower domestic prices. On the other hand, welfare may be reduced if smuggling uses up additional resources which could otherwise be used productively.15 When smuggling and legal trade coexist, an unequivocal loss of welfare can result because consumers do not gain from lower prices, the government loses revenue and additional resources are used up in the process of smuggling (Pitt 1981). In Bangladesh, smuggling and legal trade seem to coexist because importers, in order to camouflage their illegal activities, bring some goods through legal channels while prices in the local market are almost equivalent to world prices plus the tariff. Corruption also results from certain export-promotion policies. A system which basically encourages import substitution and yet tries to correct the bias against the export sector contains various anomalies (Krueger 1978). In Bangladesh, a discriminatory tariff structure designed to protect the producers serving the domestic market exists side by side with the facilities for duty-free imports of raw material for export-oriented industries. This has resulted in both high administrative costs and corruption. It is not uncommon that inputs imported for the use of export industries find their way into the domestic market because duties on these products keep the price high.16 This system ensures that exporters can make a higher profit from domestic trading than from export activities which also involve risk. Promotion of exports through administrative means rather than prices (correct exchange rate) leads to various inefficiencies and corruption. In spite of some simplifications in recent years, exporters complain about procedural complexities, and corruption with respect to the allocation of quotas for garment exports is also a common phenomenon (Ahmad 1993).17 Another area where policies have created further scope for corruption is cheap credit given to industrialists. The study of Sobhan and Sen indicates that default rates among the industrialists is so high that the existence of the financial institutions is in jeopardy. Cheap loans which are not repaid often result in capital flight rather than in industrial investment (Sobhan and Sen 1989). The recent import liberalization has also facilitated capital flight through the overinvoicing of imports of machinery and parts (The Economist Intelligence Unit 1991). It is difficult to substantiate the above statements about corruption because statistics are neither available nor can be generated.18 However, it is quite likely that a complex system based on direct control can breed corruption and inefficiency in an economy which lacks administrative skills and well-established institutions. It is now time to enquire why import-substitution policies are pursued in spite of these losses. THE MOTIVES BEHIND CONTINUED IMPORT SUBSTITUTION A strong commitment on the part of policy makers is one of the important conditions for a successful programme of economic reform. If the pressure for

220

THE EXPERIENCE OF BANGLADESH

stabilization and liberalization of the economy comes from an external source such as the donors, and the government does not perceive the need for strong and radical reforms, then the result will be an unsustainable and weak reform (Michaely et al. 1991). In the case of Bangladesh, the pressure for reform came from the donors who perceived the crises much more seriously than the policy makers. Bangladesh agreed to their proposal only to receive concessional loans and as a result the less painful measures were undertaken. Stabilization was achieved through tight fiscal policy which mainly reduced capital expenditure and certain subsidies, the restriction of credit to the private sector and the maintenance of a flexible exchange rate which was facilitated by the remittance of foreign exchange by migrant workers. The reluctance was much stronger with respect to trade, financial and fiscal matters which affect economic agents at the micro level. There can be both economic and non-economic reasons for choosing a gradual reform. The economic reasons may be stated in the following manner. First, since the government is heavily dependent on import duties, trade liberalization without a tax reform would only increase the budget deficit and thwart the goal of stabilization. Second, trade liberalization is not expected to lead to a rapid reallocation of resources from the import competing sector to other activities because of the immobility of factors, especially capital, in the short run. The sluggish response may also arise from the predominance of public sector enterprises in the large industry sector which operate under a soft budget constraint. There is also a possibility that resources would move to non-tradable activities owing to the lack of information about the export market in a backward economy. Third, in the absence of a movement of resources, the possibility that the import-competing sector may remain in business by improving its performance may be considered. However, it may take a long time. Fourth, there is a fear that unemployment will increase because of the possible contraction of the importcompeting sector not compensated by the expansion of other activities. The fear of social costs is often aggravated by frequent natural disasters in Bangladesh. Lastly, there is also an inherent belief among different groups in Bangladesh that import substitution behind protective barriers is a necessary condition for industrialization. This attitude seems to be influenced by the experience of neighbouring India which has achieved some success in import-substitution in consumer goods irrespective of the costs involved. While the above arguments are valid to a certain extent, socio-political factors provide a more plausible explanation. Over the years, the import-substitution regime has created numerous vested interests such as commercial importers, industrialists, top executives of various corporations or public enterprises, government officials in different ministries and divisions, banks or financial institutions. The military, which ruled Bangladesh for a long time, has also developed strong connections with these vested interest groups. All of them gained from the system through various rent-seeking and revenue-seeking

221

ASIA

activities. It is obvious that they would resist any trade liberalization which may curtail their economic and political power. The most important change in the 1980s, the privatization of public enterprises, did not lead to any improvement in efficiency (Sobhan and Mahmood 1991). On the other hand, it has created a more unequal distribution of wealth and income.19 This has been mainly due to distortions in the financial market. Inefficiencies and loss of potential output are bound to continue as long as large enterprises are not subject to strict financial discipline,. After coming to power, the new government showed its determination to deal with defaulters by publishing their names and by taking legal action against them. However, this enthusiasm seems to have waned, and most of the loans have still not been recovered. Reform of the large public or private enterprises is also difficult owing to the rigidities in the labour market which arise from strong union pressure. Given the monopolistic conditions prevailing in the domestic market, one of the ways of forcing the producers to become more efficient is to expose them to foreign competition through trade reform. Although certain simplifications of import rules have taken place in recent years, major reforms with respect to the reduction of tariffs have not materialized because trade reform is closely connected with the political economy of reforms in other areas. The government budget is heavily dependent on taxes on imports, as sales tax on imports and import duty account for more than half of the total tax revenue (Chowdhury and Hossain 1988:16). Trade reform requires that import duties have to be replaced by new taxes (mainly direct), higher rates of taxes, a broadening of the tax base and improvements in the efficiency of tax collection. Raising the rates of direct taxes and broadening the tax base in the short run may be difficult owing to the very low level of income of the large majority of households and to the informal nature of the economy. Tax revenue may, however, be increased substantially if corruption and inefficiency in tax collection are tackled. This is a political problem as it mainly concerns the rich who hold powerful positions in society. Among the recent changes only value-added tax which was introduced in July 1991 is noteworthy. No significant change has taken place with respect to other measures such as fighting corruption or reducing tax evasion. This is evident from the extremely low tax/GDP ratio, 8 per cent, in Bangladesh in 1991 compared with India, 17 per cent, Pakistan, 11 per cent, and Sri Lanka, 15 per cent (Finance Minister’s Budget Speech, June 1991). Another closely associated problem is the emergence of inefficient institutions, i.e. the pattern of behaviour or informal rules, in response to the system. For example, a person familiar with Bangladesh society would notice how the practice of offering and taking bribes and of evading rules has become widespread and socially accepted. The inertia of such habits and institutions has serious implications for economic growth and welfare because they tend to outlive the system. Even if the government attempts to liberalize the economy there will always be certain rules. An economy cannot function efficiently if people refuse to comply with the basic rules (North 1991).

222

THE EXPERIENCE OF BANGLADESH

The motivation for continued import substitution in Bangladesh and the problem of liberalization are, therefore, more political and institutional in character than economic. SUMMARY AND CONCLUSIONS The two main objectives of the policy reforms in Bangladesh are stabilization and long-term growth of the economy. The above analysis of the performance during the 1980s indicates that the short-run goal of stabilization has been partially achieved at the cost of growth. First of all, the reduced budget and current account deficits occurred as a result of declining investment. Second, the available investment funds have not been allocated and used efficiently, mainly because of the persistent reliance on import substituting industrialization. Industrialization requires the fulfilment of three conditions: proper incentives, development of technological capability and an efficient institutional structure (Lal 1991). These cannot be achieved through sole dependence on either the market or the state. The removal of distortions in the product and factor markets has to be complemented with government intervention in the build-up of technological capability, legal and financial institutions and basic infrastructure. The reform attempts in Bangladesh have failed to fulfil these conditions for industrial development. The government has followed an extreme form of import substitution based on government intervention which has distorted the markets and allowed inefficiency. While prices have never been right, government failure with regard to the provision of infrastructure, institutional and technological support has constrained the growth of efficient industries in the private sector. This is evident from the frequent power failures, inadequate legal protection, and lack of financial facilities such as credit and insurance and technological information.20 The future of Bangladesh looks uncertain. It is yet to be seen whether the positive forces such as the democratization process in Bangladesh and the changing policy environment in India will help the government to overcome the resistance of powerful interest groups to undertake bold reforms. The priorities reflected in the Fourth Five Year Plan 1990–95, the New Industrial Policy 1991 and the Trade Policies for 1991–93 indicate that government intervention is still directed at trade and investment rather than infrastructure, technology and institutions. ACKNOWLEDGEMENTS The author gratefully acknowledges with thanks the valuable comments made by Yves Bourdet, Anders Danielson, Ronald Findlay, Göte Hansson, Mats Lundahl, Laszlo Somogyi and Bo Södersten.

223

ASIA

NOTES 1 In the wake of independence from British rule (1947), Bangladesh was the largest producer and exporter of jute in the world. 2 Although the British pound and Indian rupee were devalued in 1949, Pakistan did not follow suit. Devaluation did take place in 1955, but the Pakistani rupee remained overvalued. 3 It should be mentioned that the share of the public sector in terms of the number of establishments has always been low because public ownership has been limited to large capital-intensive enterprises. 4 During the Second Five Year Plan 1980–85, DFIs provided 78 per cent of the industrial investment in the private sector and 60 per cent of this institutional finance was in foreign aid. (Third Five Year Plan 1985–90, p. 236). 5 Nearly one-third of total exports are directed to the USA. 6 The inflation rate runs between 10 and 12 per cent. 7 Ninety per cent of the workforce in the garment industry are female (Gain 1990). 8 The government provides infrastructural facilities in the EPZ for enterprises producing exclusively for export. 9 Under the Wage Earners’ Scheme, several categories of imports are allowed without the formal permission which facilitates export activity and suppresses inflationary pressures. 10 Bangladesh failed to devalue the currency during 1988–9 to meet the goal of a flexible exchange rate policy. 11 According to Anderson et al.’s study, only 16 per cent of the 591 sample firms had trade unions. 12 Fieldwork carried out by the author in connection with an ongoing study of the urban labour market in Bangladesh. 13 According to the concepts used by the HIID study, while value added may be affected by both ‘subsidies’ (from trade and domestic policies) and market forces, the ‘statutory ERA’ assumes away variations on prices resulting from ‘market forces’. 14 This is reflected in the rapid growth of the service sector; see Figure 11.3. 15 Assuming constant costs of smuggling at a rate less than the tariff and competitive smuggling activities (Bhagwati and Hansen 1973). 16 According to the owners of domestic weaving mills, the major threat to them comes from the garment industry’s selling fabrics which have been imported duty free as inputs for readymade garments. (Author’s interview with industrialists in connection with an ongoing project on ‘Urban poverty and the informal sector: A study of the labour market in Bangladesh’ sponsored by SAREC.) 17 Corruption with respect to garment exports is even recognized at government level as is evident from the Finance Minister’s Budget Speech, July 1991. 18 One study calculated rents from import licensing in Bangladesh during 1983–4 and found them to be as large as 3.1 per cent of GNP (Abdullah 1991). 19 The public enterprises were bought by private entrepreneurs at a highly reduced price with borrowed capital from the Development Finance Institutions (Sen 1991, Abdullah 1991). 20 The poor performance of the non-financial public enterprises such as railways, telecommunications and the energy sector has not only caused a drain on the government budget, but also hampered growth in the productive sectors.

224

THE EXPERIENCE OF BANGLADESH

BIBLIOGRAPHY Abdullah, A. (ed.) (1991) Modernization at Bay: Structure and Change in Bangladesh, Dhaka: Dhaka University Press. Ahmad, A. (1993) ‘Urban poverty and the informal sector in Bangladesh: A study of the urban labour market’, Stencil Dept. of Economics, University of Lund. Anderson, K.H., Hossain, N. and Sahota, G. (1991) ‘The effect of labour laws and labour practices on employment and industrialization in Bangladesh’, in Z.Bakht, (ed.) The Bangladesh Development Studies—Special Issue on Industrial Policies and the State of Industrialization in Bangladesh XIX:(1) and (2). Bhagwati, J. (1982) ‘Directly unproductive profit-seeking (DUP) activities’, Journal of Political Economy 90(5): 988–1002. Bhagwati, J. and Hansen, B. (1973) ‘A theoretical analysis of smuggling’, Quarterly Journal of Economics 87:172–87. Bhattacharya, D. (1989) ‘Credit market and metal fabrication sector in Bangladesh’, Research Report, Bangladesh Institute of Development Studies (BIDS), No. 99, Dhaka. ––– (1991) ‘Role of policies, regulations, and institutions in self-employment promotion in Bangladesh’, Research Report, BIDS, No. 130, Dhaka. Chowdhury, F.A. and Rahman, A. (1989) ‘Urban informal financial markets in Bangladesh’, Research Report, BIDS, No. 103, Dhaka. Chowdhury, N. (ed.) (1989) ‘The handloom economy of Bangladesh in transition’, The Bangladesh Development Studies, special issue, XVII: March-June. Chowdhury, O.H. and Hossain, M. (1988) ‘Elasticity and buoyancy of Bangladesh tax structure’, Research Report, BIDS, No. 80, Dhaka. Gain, P. (1990) Women Workers in Bangladesh Garment Industry (in Bangla), Bangladesh. Government of Bangladesh, Bureau of Statistics (different years) Statistical Yearbook of Bangladesh, Dhaka. Government of Bangladesh, Ministry of Finance (1991) Annual Budget 1991–92: Budget Speech of the Finance and Planning Minister, Dhaka. Government of Bangladesh, Ministry of Planning (1990a) Bangladesh Economic Survey 1989/90, Dhaka. ——(1990b) The Fourth Five Year Plan 1990–95, Dhaka. ——Industrial Policy 1991, Dhaka. ——Import Policy 1991–93, Dhaka. ——Export Policy 1991–93, Dhaka. Government of Bangladesh, Planning Commission (1990) ‘An assessment of the impact of industrial policies in Bangladesh’, Working paper no. 16, HIID/ESEPP Project. Hossain, M. (1988) Green Revolution in Bangladesh, Dhaka: Dhaka University Press. Hossain, N. and Jamal, A.M.M. (1989) ‘The earnings of industrial labour in an LDC: The evidence from Bangladesh’ , The Bangladesh Development Studies XVII: (1). IFRI and BIDS (International Food Policy Research Institute and the Bangladesh Institute of Development Studies) (1985) Development Impact of Food-For-Work Program in Bangladesh Summary, Washington, DC and Dhaka. Islam, N. (1981) Foreign Trade and Economic Controls in Development: The Case of United Pakistan, New Haven, CT: Yale University Press. Islam, R. (1989) ‘Informal financial market and the handloom and small scale cotton textile sector in Bangladesh’, Research Report, BIDS, No. 102, Dhaka. Krueger, A. (1978) Foreign Trade Regimes and Economic Development: Liberalization Attempts and Consequences, New York: NBER. Lal, S. (1991) Building Industrial Competitiveness, Paris: Development Centre of the Organization for Economic Co-operation and Development. Lewis, S.R. (1970) Pakistan: Industrialization and Trade Policies, New York: Oxford University Press.

225

ASIA

Michaely, M., Papageorgiu, D. and Choksi, A.M. (eds) (1991) Liberalising Foreign Trade: Lessons of Experience in the Developing World, Oxford: Basil Blackwell. North, D. (1991) Institutions, Institutional Change and Economic Performance, Cambridge: Cambridge University Press. Pitt, M.M. (1981) ‘Smuggling and price disparity’, Journal of International Economics 11: 447–58. Sahota, G. (1991) ‘An assessment of the impact of industrial policies in Bangladesh’, in Z. Bakht (ed.) The Bangladesh Development Studies—Special Issue on Industrial Policies and the State of Industrialization in Bangladesh, XIX:(1) and (2). Sahota, G. and Huq, M. (1991) ‘Effective rates of assistance in Bangladesh economy’, in Z. Bakht, (ed.) The Bangladesh Development Studies—Special Issue on Industrial Policies and the State of Industrialization in Bangladesh, XIX:(1) and (2). Sen, B. (1991) ‘Privatization in Bangladesh: Process, dynamics and implications’, in V. Kanesaligam, (ed.) Privatization Trends and Experiences in South Asia, New Delhi: Macmillan India. Sobhan, R. and Mahmood, S.A. (1991) ‘The economic performance of denationalised industries in Bangladesh: The case of the jute and cotton textile industries’, Research Report, BIDS, No. 129, Dhaka. Sobhan, R. and Sen, B. (1989) ‘Trends in the repayment performance in the DFPIs: Implications for the development of entrepreneurship in Bangladesh’, The Bangladesh Development Studies XVII (3):21–66. The Economist Intelligence Unit (1990) Bangladesh: Country Report, London, no. 2. ——(1991) Bangladesh Country Report, London, no. 1. World Bank (1990) Bangladesh: Strategies for Enhancing the Role of Women in Economic Development, Washington, DC. ——(1992) World Development Report 1992, New York and Oxford: Oxford University Press.

226

Part IV AFRICA

12 EXCHANGE RATE POLICY AND MANAGEMENT IN THE CONTEXT OF ECONOMIC REFORMS IN SUBSAHARAN AFRICA Tanzania as an illustrative case Benno Ndulu

INTRODUCTION The economic crisis of the 1980s in sub-Saharan Africa has set into motion major economic reforms in an effort to reverse it and embark on a path of renewed growth. The first half of the 1980s saw the response to the crisis dominated by recessionary stabilization measures (Ndulu 1990). Resource gaps were largely closed via autonomous cuts in aggregate demand and reduced economic activity. Being import dependent in both production and investment, reduced import capacity resulting from a continued decline in exports and cuts in net foreign resource inflows led to import compression and strangulation of economic activity. In a large number of countries, including Tanzania, forced import compression was implemented via quantitative restrictions. Exchange controls and trade policy rather than the exchange rate were the main instruments for managing the balance of payments. In spite of these measures and losses in growth, however, both internal and external imbalances persisted. The second half of the 1980s saw the beginning of a major shift in policy away from recessionary stabilization and towards what has been popularly referred to as ‘adjustment with growth’. Both the concerned governments and international financial institutions seem to be converging to a position that growth is probably the best guarantor of sustainable economic stability. Pro-growth measures have thus gained momentum in spite in the widely appreciated difficulties of achieving simultaneously growth and sustainable internal and external macroeconomic balances. A pro-growth exchange rate policy has become a major component of the ‘new’ thinking and the ongoing economic reforms in sub-Saharan Africa. Over the last eight years, a large number of sub-Saharan African countries have pursued a rather active exchange rate policy. More than fifteen countries out of about thirty countries not operating under a fixed-peg arrangement depreciated

229

AFRICA Table 12.1 Effectiveness of devaluation in selected African countries

Sources: Guillaumont and Guillaumont Jeanneney (1991) Edwards, S. (1988) Exchange Rate Misalignment in Developing Countries, Baltimore, MD: Johns Hopkins University Press Notes: Ratio of the variation of REER to that of NEER corresponds to what Edwards (1988) calls ‘effectiveness of devaluation’

their nominal effective exchange rates (measured in foreign currency terms) by more than 50 per cent during 1982–8 in an effort to regain external viability (see Table 12.1). All except one achieved a very high rate of effective real depreciation. Such effectiveness measured as a ratio of real rate of depreciation to nominal effective rate of depreciation (Edwards 1988, Guillaumont and Guillaumont Jeanneney 1991) was in excess of 70 per cent for the majority of them (Table 12.1). Serious reconsideration of the role of exchange rate policy has been informed by the previous experience in dealing with the crisis as well as by empirical studies that have provided some estimates of the efficacy of exchange rate policy.

230

EXCHANGE RATE POLICY: TANZANIA

It has become increasingly clear that dealing with the external deficits through controls on imports alone is not sustainable and entails major losses in growth. On the other hand, rather sobering conclusions from empirical studies point towards a less than traditionally presumed level of efficacy of the exchange rate policy in regaining external viability in import-dependent economies. A number of recent studies on developing economies, including Khan and Knight (1988), Moran (1990) and Jebuni et al. (1991), have shown that the import expenditure reduction effects of the exchange rate adjustments are quite weak. Owing to imperfect substitution between imports and domestic goods in these economies, aggregate import spending is highly price inelastic downwards. Moreover, given that export production is also import dependent, increased exports can only come on stream with increased imported inputs (Khan and Knight 1988). The expenditure switching effects of exchange rate adjustment, though significant, are not as large as presumed in the traditional analysis. A longer time span and perseverance in action to realize benefits from persistent exchange rate adjustment are implied. It appears also from the above that the export-promoting role of exchange rate policy under the situation of import compression is relatively more important than import reduction for achieving external viability. Evidence of the responsiveness of exports to real exchange rate movements in sub-Saharan Africa is abundant (see e.g. Balassa 1990, Jebuni et al. 1990, Fosu 1992, Ndulu and Lipumba 1991 among others). The essay discusses next the major concerns and issues in the conduct of an appropriate exchange rate policy in the context of reforms. This is followed by an illustrative example of Tanzania focusing on the determinants of the appropriate exchange rate level consistent with its efforts in economic reforms. The last section concludes with issues and lessons for the future. MAJOR ISSUES IN THE CONDUCT OF EXCHANGE RATE POLICY IN A REFORMING AFRICAN ECONOMY Four major issues arise with respect to the conduct of an appropriate exchange rate policy. First is the choice of an appropriate exchange rate regime. The range here spans from a fixed-peg to a freely floating exchange rate regime. The advantages and disadvantages of each have been discussed at length (see e.g. Collier and Joshi 1989). Within sub-Saharan Africa, a variety of regimes are operational. These include the fixed peg used in the West and Central African franc zones, Ethiopia and Liberia, adjustable basket-peg widely operational in Anglophone Africa, and managed float more recently adopted in countries such as Ghana, Nigeria, Zambia and Uganda. The key distinguishing characteristics across the typologies entail fixity and variability of the nominal exchange rate. The advantages of a stable trading environment and anti-inflationary discipline are associated with the variants of fixed exchange rate regimes. On the other hand, the main advantages of the more flexible regimes are associated with flexibility in responding to changing

231

AFRICA

economic conditions. Defending against slides in the external competitiveness and profitability of the tradable sector internally is widely considered to be more easily achievable under these types of regimes. A further consideration in the choice of regimes, particularly in view of experiences in sub-Saharan Africa, is the extent of depoliticization of this instrument. While legitimate income distributional concerns are important to take into account, inertia in required action has often been promoted by excessive accommodation to political pressure that ultimately proved unsustainable. Rulebased exchange rate regimes reduce the latitude for discretion and political pressure on the conduct of exchange rate policy. The second issue concerns the determination of an ‘appropriate’ level of the exchange rate. Here again two aspects are further considered. First is the required depreciation to compensate for the past accumulated appreciation of the real exchange rate. Second is the determination of the transitional exchange rate levels consistent with the reductions in distortions envisaged in the economic reforms. Exchange and trade liberalization are particularly noteworthy here. Appropriate sequencing and consistency with other policy changes are fundamental. The difficulty of proper implementation of both these tasks cannot be overestimated. Determining whether a country’s real exchange rate (RER) level at a point in time is at a sustainable ‘equilibrium’ is at best indicative. Indeed, as Edwards (1990) remarks, ‘it is one of the most difficult analytical tasks in macroeconomics’. Conclusions from studies on the experiences of major floating currencies as reviewed by Manzur (1990) and Mark (1990) as well as their own results attest to this. Using both absolute and relative definitions of purchasing power parity (PPP) doctrine to characterize equilibria, Dornbusch (1976), Frenkel (1981), Flood (1981), Mussa (1982), Miller (1984), Hakkio (1984) and Manzur (1990) among others show that although PPP holds quite well in the long run, it exhibits a random walk in the short run as a result of persistent real shocks and the differential speed of adjustment between prices and nominal exchange rates. In any case, the need to obtain some reasonably appropriate exchange rate levels in the medium term, particularly starting from a scenario of disequilibrium RER, cannot be overemphasized if the exchange rate policy has to be supportive of economic reforms. This partly explains our relative emphasis on the determinants of the real exchange rate in the illustrative case in order to enable estimation of past accumulated slide of parity related to real shocks, project necessary exchange rate depreciation for compensation and approximate required changes consistent with projected changes in fundamentals. The third issue relates to the link between exchange rate policy and other complementary measures in the process of economic reforms. Starting with a situation of fundamental disequilibria, the exchange rate is viewed here as a policy instrument, and not just a price, to be used in conjunction with other policy measures for corrective purposes. Over- or undershooting of the ‘appropriate’

232

EXCHANGE RATE POLICY: TANZANIA

level may thus be permissible at a point in time to force adjustments in other complementary policy areas mentioned earlier. The fourth issue relates to the differential initial conditions obtaining in the countries concerned, regarding the nature of the foreign exchange markets. With the exception of the CFA franc zone countries, most countries in varying degrees are saddled with the problem of local currency inconvertibility. Buttressed by extensive exchange controls, inconvertibility has given rise to dual foreign exchange markets of varying significance. The parallel foreign exchange market where the marginal cost of foreign exchange is determined has tended to respond more flexibly to changes in the fundamentals affecting the exchange rate (Kaufmann and O’Connell 1991, Pinto 1989). Wide gaps between the official and the much more depreciated parallel rates have arisen as a result of real appreciation of the official exchange rates. However, depending on the relative size of the parallel market, the effective exchange rate applicable lies in between the two. A policy concern in the reforms is whether the two exchange markets should be unified and what implications would such unification have on the determination of the appropriate exchange rate. THE DETERMINANTS OF APPROPRIATE EXCHANGE RATE: TANZANIA AS AN ILLUSTRATIVE CASE Exchange rate policy experience in Tanzania Until 1986, the conduct of exchange rate policy in Tanzania could be characterized as passive. With the advent of exchange controls in 1965 and a further intensification of its use for balance of payments management since 1972, Tanzania’s exchange rate was used mainly as an accounting instrument for external transactions. The nominal rate remained fixed for long periods with occasional adjustments induced by programme requirements (Table A12.1). This passiveness of the exchange rate policy at the time when other policy instruments—including monetary and fiscal policies—were unable to check domestic inflation, and when the economy experienced real shocks affecting external competitiveness, accounts for the rapid real appreciation of the Tanzanian shilling between 1973 and 1985. This is shown in Table A12.1 in the appendix as a rapid appreciation of the real exchange rate (RER) during the period. RER is measured here as the ratio of Tanzanian CPI to a weighted average of foreign wholesale prices converted to Tanzanian shillings. A rise in the index signifies real appreciation and a decline real depreciation. The index used here moved closely together with five other computed multilateral real effective rate indices with cross-correlations ranging from 80 per cent to 98 per cent (Bank of Tanzania 1990). Looking at the pattern of movements of RER over the period 1966–90, one observes three distinct periods. Compared with the rest of the period, 1966–73 was characterized by relative stability of RER. Both external and internal

233

AFRICA

balances were reasonably in equilibrium in spite of a more or less fixed nominal exchange rate being in force. Real activity continued to grow. Exchange controls were largely for capital account transactions. The years 1973–85 saw a strong trend of RER appreciation, mainly as a result of expansionary macroeconomic policy, a rapid increase in the taxation of exports and intensified use of exchange controls. This trend was much more pronounced during 1978–85 as recourse to the monetization of the fiscal deficit was sought partially to fill resource gaps and a wide ranging and intensified use of quantitative restrictions to manage balance of payments. During 1986–90, a steep depreciation of the RER was effected, spearheaded by nominal maximum devaluations as part of stabilization and adjustment efforts. It is important to note here that nominal devaluation during the four years translated into a real depreciation of approximately 88 per cent, more than offsetting the accumulated historical appreciation on PPP terms alone (see Table A12.1 and also Krumm (1990)). Whether this compensates for changes in the fundamentals affecting external competitiveness is an issue we take up later below. These movements in the real exchange rate are further vindicated by movements in the parallel market premium. As shown in Table A12.1, the premium tended to increase with appreciation of RER and fall with depreciation. Kaufmann and O’Connell (1990) provide an excellent conceptual and empirical analysis of this phenomenon in the case of Tanzania. The parallel market for foreign exchange is quite large in Tanzania. Conservative estimates based on own-funded imports alone put it at $400–500 million out of the total import bill of $1.2 billion (Ndulu and Hyuha 1989). Since foreign exchange is rationed its marginal cost is determined in the parallel market. Domestic prices adjust fully to the parallel market rate. This has particularly been the case since July 1984 with the introduction of wide ranging price decontrols and legal sanctioning of imports financed by private foreign exchange sources. The official exchange rate in this situation as far as imports are concerned is inframarginal (Pinto 1989). Fundamentals and the appropriate exchange rate level From an economic reform perspective, two categories of fundamentals affecting the real exchange rate can be distinguished. One group is exogenous and the other endogenous to policy reform. Fundamentals exogenous to policy include exogenous factors such as external terms of trade, autonomous net foreign resource inflows and irreversible effects of past policy mistakes. Not all components of foreign resource inflows are truly exogenous to policy, however. Stabilization and adjustment programme funding depends to a large extent on donor perception of a policy stance conducive to sustainable growth and economic stability. Foreign credit, debt relief and private foreign investment are particularly sensitive to this perception. The irreversible effects of past policy mistakes mainly include accumulated debt overhang, both foreign and domestic, which have caused a somewhat long-term upward shift in

234

EXCHANGE RATE POLICY: TANZANIA

the size of the fiscal gap that needs to be accommodated since most of the debt (more than 95 per cent) is officially contracted in Tanzania. Fundamentals endogenous to policy reform (policy-induced real shocks) can be further broken down into two categories. The first refer to those related to reversible institutional structures and commercial policy that affect the effectiveness of the exchange rate as an incentive instrument. These include inefficiency of the marketing system, direct and indirect taxation of external trade which affect the pass-through of incentive signals and militate against the effectiveness of exchange rate policy. Improvements in these factors will contribute towards defending against the slide in the domestic competitiveness of tradables and will foster exchange rate stability by reducing the pressure on the nominal exchange rate in the medium term to accommodate real shocks. The second category of fundamentals endogenous to policy reform refer to those that affect the co-ordination between the exchange rate and macroeconomic policies. This coordination affects mainly the shorter-term management for maintenance of parity with trading partners. The relationship between the nominal exchange rate and inflation differentials with trading patterns is emphasized (Cavallo and Cottani 1985). The co-ordination of monetary and exchange rate policy is at the heart of this, since the other factors affecting inflation differentials, real growth and inflation rates of the trading partners can be considered as predetermined in this regard. In the medium term, prudent coordination requires that domestic credit grows at a rate consistent with a reasonable target inflation rate and real output growth while the nominal exchange rate changes at a rate equal to the inflation differential. This puts the burden of defending against parity slide on both the exchange rate policy and monetary policy. The link of this co-ordination to fiscal policy is strong in Tanzania in view of the previous high reliance on the monetization of fiscal deficit to bridge resource gaps and a rather undeveloped financial market constraining other potential financing sources. The relationship between fundamentals and the real exchange rate Recent literature on the determinants of equilibrium exchange rate motivates the analysis in this section. This literature includes Edwards (1987a,b, 1988,1989a, 1990), Dornbusch (1974), Khan and Ostry (1991), Cavallo and Cottani (1985), Krumm (1990), Elbadawi (1989a,b). The analysis typically starts with a general equilibrium model that emphasizes the determination of the equilibrium exchange rate which is consistent with both internal and external equilibria. An adapted simple basic model is presented here with this literature serving as a background. Exchange control and currency inconvertibility is an important feature to keep in mind for our subsequent discussions. Two sets of factors determining the real exchange rate are distinguished: real and monetary. The simplified model consists of two basic equations (see also Edwards 1990):

235

AFRICA (12.1) (12.2)

The reduced form is obtained by substituting (12.2) and (12.1). (12.3)

The first equation postulates that movement in the official real exchange rate (ORER) in the short run around its appropriate level (ARER) is caused purely by excess aggregate demand, measured as the excess of domestic expansion over nominal depreciation of exchange rate, real income growth and a weighted inflation rate of the trading partners (xt). The second equation explains the determination of the unobservable appropriate real exchange rate (ARER) using its observable determinants—fundamentals (Ft). These include external terms of trade, commercial policy, quantitative restrictions, net foreign resource inflows and the parallel market exchange premium. Details on the effects of fundamentals on ARER are provided later below. The dynamic version of (12.3) allows for non-instantaneous adjustment of changes in ORER to changes in ARER. Incorporating the full specification of xt into (12.3) the dynamic version of (12.3) or partial adjustment model can be written as (12.4)

which yields (12.5)

or (12.6)

ORERt=ORERt-ORERt-1

ß=adjustment coefficient D=domestic credit growth g=real net output growth e=rate of change of nominal exchange rate p*=weighted average inflation rate of the trading partners The coefficient ß measures the speed of adjustment of ORER to exogenously induced changes in ARER in the absence of accompanying macroeconomic measures to counter them (Edwards 1990:29). Now we turn to the details of the relationships between fundamentals and ARER. Classical trade theory postulates unambiguously that for a ‘small’ country, an improvement in the external terms of trade (TOT) will lead to an appreciation of the equilibrium real exchange rate in order to clear a surplus and vice versa in the

236

EXCHANGE RATE POLICY: TANZANIA

case of deterioration (Johnson 1966, Balassa 1982, Bacha and Taylor 1971). Recent developments in the literature, however, show that the impact of change of TOT on ARER cannot be signed a priori if the analysis is extended to incorporate second-round effects of TOT through other fundamentals (Edwards 1987a,b, Cavallo and Cottani 1985, Elbadawi 1989). Major among the second-round effects are the endogenous trade liberalization/restriction effects of TOT improvement/deterioration and a rise/fall in real income which may induce an offsetting effect on the standard substitution effect on the real exchange rate (Khan and Ostry 1991). In economies with exchange controls such as Tanzania, the endogenous trade liberalization effect is particularly noteworthy. Turning to explicit commercial policy, we consider, separately, import tariffs (tm) and export taxes (tx). Both of these are measured in their effective terms, the ratio of actual tax paid to the tax base. We postulate that a decrease in export or import (tariff liberalization) taxes will cause a depreciation in RER. The assumption here is that the substitution effect between tradables and nontradables dominates the real income effect resulting from tax cuts, leading to an increase in the demand for non-tradables and hence upward pressure on their prices (Elbadawi 1989a,b). In the Tanzanian case implicit tariffs on imports through quantitative restrictions more than swamped explicit taxation during the period of analysis. This significantly reduces the effect of explicit import taxes on ARER, a fact that we verified empirically. Only export taxes were therefore included in the final estimated equation. An increase in the net foreign resource inflows has two channels of effects on ARER. First it can be considered as a positive shock from the income side. If the increase in income is considered temporary (i.e. holding desired absorption constant) it would only cause a temporary aberration in the real exchange rate (ORER). However, if such an increase is considered permanent it would raise disposable income and hence desired absorption, raising the price of nontradables, and thus lead to the appreciation of ARER. The second channel is the endogenous trade liberalization effect such inflows generate as a result of relaxation of the foreign exchange constraint. This effect would prompt a depreciation of ARER. The net effect can only be determined empirically. Furthermore, to the extent that these inflows are used primarily to purchase traded goods, their impact on the prices of non-traded goods through domestic excess aggregate demand may be negligible. More recently in reforming African economies increased inflows of net foreign resources have also tended to reduce pressure on the monetization of the fiscal deficit. Quantitative restrictions as pointed out earlier have been used widely for the management of balance of payments in Tanzania. These constitute partly an endogenous trade policy via higher implicit tariffs and partly endogenous macroeconomic policy through their substituting for exchange rate policy as an instrument of balance of payments management (Collier and Joshi 1989). The inflationary effect of increased quantitative restrictions in a situation of binding import quotas leads to an appreciation of ARER. This effect is further

237

AFRICA

strengthened by the foregone government revenue and its partial inflationoffsetting impact compared with the situation where explicit tariffs would be used. On the macroeconomic side, unlike the exchange rate, the use of quantitative restrictions while producing the inflationary effect forgoes the partial income-exports increasing effect associated with real depreciation. An increase in the extent of trade restrictions will result in a fall in actual imports which will require an appreciation of equilibrium RER to clear a surplus, while a relaxation will require a real depreciation of equilibrium RER to clear the deficit. In this study we use import tightness, measured as a ratio of real imports to net output (GDP+real imports) to represent the effects of trade restrictions. The brunt of protecting reserve levels is borne by import compression rather than the reserve levels by design. Addition of real imports to GDP in the denominator for IMPTGT (import tightness) ratio is to correct for some trend of structural changes in import content, thus emphasizing marginal movements in import tightness associated with changes in the levels of import restrictions. The parallel market exchange rate premium on the one hand accrues to importers with access to official foreign exchange sources as import licence rents, but on the other hand it is a tax on exporters. Export earnings have to be surrendered at the official exchange rate while the domestic prices of imports fully adjust to the parallel rates. Thus the difference between the official exchange rate at which export earnings are surrendered and the parallel rate at which imports are valued domestically divided by the parallel rate represents the marginal tax rates on exports (Pinto 1989). As the premium rises, this additional (to explicit taxes) taxation on exports rises. Akin to the explicit trade taxes, an increase in the marginal taxes will cause an appreciation in RER and vice versa for a decrease. An excess of domestic credit growth over nominal exchange rate depreciation, foreign inflation and real growth (i.e. x>0 implying expansionary macroeconomic policy) will lead to an appreciation of the real exchange rate and vice versa for deflationary macroeconomic policy. A direct link between net credit expansion and domestic inflation in the medium term is assumed here. Real demand for money is assumed to be stable in the medium term and hence it grows with real income. The plausibility of this assumption arises from the dominance of transactions demand for money and the negligibility of portfolio effects across the very limited range of financial instruments (see Khan 1980). Empirically estimated inflation equations have validated this point for Tanzania (Kilindo 1990, Ndulu and Hyuha 1989, Hyuha and Ndulu 1991, among others). Since a close link between the fiscal deficit (broadly defined to include quasi-deficit) and domestic credit expansion via its monetization exists, the variable x incorporates fiscal pressure on the real exchange rate.

238

EXCHANGE RATE POLICY: TANZANIA

Empirical relationship between fundamentals and the real exchange rate in Tanzania The reduced form of the dynamic model was estimated using ordinary least squares. Data for the variables is presented in Table A12.1. The estimated equation is presented in (12.7) below and the period of estimation was 1966–88:

(12.7)

where TOT=external terms of trade LUOEXPT=marginal tax rate on exports IMPTGT=import tightness EXPTX=explicit export taxation measured as (1-tx) MACROPOL=(DC-g-ê-p*) Import tariffs and foreign resource inflows were excluded from the equation owing to their statistical insignificance after trial estimations and for reasons of indeterminacy of signs given earlier. Import tightness increases with the decline of IMPTGT. The trend variable was included to capture the effects of relative Table 12.2 Determinants of the real exchange rate in Tanzania (the dependent variable is ORER for (12.7) and (12.8) and ∆ORER for (12.9))

Notes: All variables except MACROPOL, MACROP, GNOER (expressed in growth terms) are in logarithms t-statistics in parentheses GNDER is the rate of change of the nominal exchange rate

239

AFRICA

productivity changes. A second equation (12.8) was estimated separating nominal exchange rate changes from the rest of the macroeconomic policy variables to obtain the impact of nominal devaluation on the real exchange rate and hence assess the efficacy of this corrective policy instrument. Adjusted MACROPOL appears as MACROP(DC-g-p*). Equation (12.9) is the direct dynamic version of (12.8) with the dependent variable being dlog (ORER) and enables us to verify the size of ß, the coefficient of ORERt-1. The results are presented in Table 12.2. The above regression results are highly satisfactory in spite of the brevity of the period of analysis and they compare reasonably with the results obtained by several other studies on the same. The joint explanatory power of the included variables is quite high and statistically significant. An instrumental variable version of the estimated equation yielded virtually no differences with ordinary least-squares estimates. All coefficients have the expected signs and most are statistically significant at conventional levels. The weak significance of TOT is probably explained by the offsetting second-round effects discussed earlier on. The results show that both real and monetary factors are important in the determination of the real exchange rate. Real shocks, whether exogenous or policy induced, affect external competitiveness. Trade policy (explicit and implicit) in particular appears to have a large influence on the appropriate RER, suggesting that a relaxation of quantitative restrictions and tariff liberalization will require a substantial depreciation of ARER to maintain external viability. A further relaxation of quantitative restrictions alone, for example (reduction of import tightness, say by 10 per cent), will require an accompanying real depreciation of about 5 per cent. The results also show that macroeconomic pressures, particularly excess credit growth if not countervailed by commensurate nominal exchange rate depreciation, will tend to generate exchange rate overvaluation and loss in external competitiveness. The coefficient on MACROPOL is not only statistically significant but also large. When the nominal exchange rate effect on RER is accounted for separately (equation (12.8)), the results show its high and dominant effectiveness among the macroeconomic policy instruments. Nominal depreciation has an 80 per cent effectiveness on real depreciation, making it an important instrument for protecting external competitiveness. It should be noted, however, that this high effectiveness is partly explained by the fact that active nominal depreciation was implemented following a long period of persistent overvaluation, and that the inflation rate following the maximum depreciations actually declined (from 34 per cent per annum in 1985 to about 19 per cent in 1990). Apart from maintaining draconian cuts in real wages, renewed real output growth and the fact that domestic prices of imports already reflected the much more depreciated parallel market exchange rate are the main reasons behind this apparently non-conventional observation. The estimated ‘automatic’ adjustment coefficient ß was 0.15 (equations (12.8) and (12.9)) suggesting that if the system was left to its own devices, only

240

EXCHANGE RATE POLICY: TANZANIA

15 per cent of a unitary discrepancy between actual and equilibrium real exchange rates would be self-corrected within one year. This emphasizes the importance of the active use of macroeconomic policy to counter the effects of exogenous shocks on the real exchange rate in order to protect external competitiveness. CONCLUDING REMARKS The foregoing analysis points to several important general insights with respect to the conduct of an appropriate exchange rate policy in the process of economic reforms in African economies. 1

2

3

4

A flexible exchange rate regime capable of responding to both exogenous and policy-induced shocks is a sine qua non, in order to protect the economy against loss in external competitiveness as well as to preserve competitiveness of tradables internally. The results above show that real shocks have a significant and large influence on the appropriate real exchange rate. Corrective action through changes in the nominal exchange rate should not only cover in the medium term the inflation differential to maintain parity with trading partners but also compensate for negative effects of real shocks on the real exchange rate. The effectiveness of nominal depreciation in achieving real depreciation in the short to medium term given the initial large disequilibria points to the important corrective role an active exchange rate policy can play in the transitional stage. Both continental data (Table 12.1) and the specific case of Tanzania point to this fact. The distinction between exogenous and policy-induced shocks affecting the real exchange rate needs to be emphasized. While purely exogenous shocks delineate the minimum corrective action to be absorbed via an active exchange rate policy, automatic absorption of policy-induced real shocks can create an irresponsible attitude of dumping all policy garbage on this instrument. The transitional period undoubtedly will require large nominal depreciations to correct for accumulated real appreciation as well as for alignment of the real exchange rate with restructured institutional set-ups and incentives. The emphasis for the future should be on the complementarity of policy measures that minimizes the stress on the nominal exchange rate and enhances the potential effectiveness of this instrument. More specifically, complementary macroeconomic policy emphasizing controls on excessive domestic credit expansion; reduction of exchange controls and trade restrictions; and institutional restructuring that ensures effective passthrough of incentive signals is of crucial importance. The links between the major policy reforms and the conduct of an appropriate exchange rate policy need to be highlighted. The most important and relevant among the measures being implemented/or contemplated in

241

AFRICA

5

6

reforming African economies are the following four: tariff liberalization (mainly compression of rates and structure), exchange controls relaxation, price decontrol, and financial sector restructuring. These measures imply a more active use of the exchange rate policy for balance of payments management; a move towards enhanced ‘tariffication’ of protection due to reduced reliance on implicit tariffs associated with exchange controls and a passive exchange rate policy; enhanced pass-through to generators and savers of foreign exchange via reduction in margins associated with monopsonistic marketing agencies and oligopolistic sellers; and reduced pressure on the exchange rate resulting from monetization of fiscal deficits as the banking sector frees itself from state directives. Important sequencing questions need to be resolved. Should policy authorities front-load the required exchange rate adjustment consistent with the reforms or move in tandem with the actual implementation of complementary reforms? The answer, though not definitive, depends to a large extent on the resolve to implement the various reforms in question. Where feet dragging is evident, front loading may be necessary to prevent loss in external competitiveness. While the unification of the dual foreign exchange markets (official and parallel) is strongly desired by authorities in the various countries, its achievement depends on a number of other factors in addition to the measures highlighted above. Most of the above measures relate to current account transactions. Capital flow controls are still in place in most African countries and are envisaged to stay for the foreseeable future. Current account viability seems to be the prime concern. The parallel foreign exchange market will thus persist, albeit on a much smaller scale, to meet the demands for asset flows and other illegal current transactions. This is evident from the experiences of Ghana, Zambia, Nigeria and Uganda which have tried to effect unification through adopting various forms of controlled foreign exchange auction systems (Pinto 1989, Chibber and Shafik 1990). It is expedient to reiterate at this closing point that maintaining an appropriate real exchange is a necessary but not sufficient condition for achieving external viability and supporting real growth. In African economies export performance is largely supply constrained (Svedberg 1990). While an appropriate incentive structure is a sine qua non to elicit desired supply responses, an enabling institutional structure and infrastructure that is supportive of efficient production and marketing is critical to achieve results. Rehabilitation of supportive infrastructure, investment in improved technologies and reduction in administrative bottlenecks to external trade are all critical for this purpose.

242

EXCHANGE RATE POLICY: TANZANIA

ACKNOWLEDGEMENTS I am indebted to the Bank of Tanzania for providing me with an opportunity to contribute to the ‘Tanzania: Exchange rate study’, from which many of the insights herein are drawn.

243

Sources: Krumm (1990)—appendix tables Kaufmann and O’Connell (1990)—appendix tables

Table A 12.1 Tanzania: real exchange rate and the fundamentals

APPENDIX

Economic Surveys, Tanzania (various issues) Bank of Tanzania, Economic and Operations Reports (various issues) Economic Research Bureau, Tanzania Economic Trends (various issues) IMF, International Financial Statistics World Currency Year Book Notes: ORER65=Official real exchange rate index defined as the ratio of Tanzanian CPI to a weighted average of foreign wholesale prices converted to Tanzanian shillings. The relevant weights for trading partners are USA (9.4%), Japan (12.1%), Belgium (5.1%), Germany (19.0%), Italy (8.4%), Netherlands (10.7%), Sweden (4.1%), UK (27%) and India (4.8%) TOTUNC=External terms of trade index (UNCTAD) RIMPGDP=Ratio of real imports to GDP (import tightness) UOEXPT=Marginal taxation of exports (parallel rate premium/parallel market exchange rate) EXPTXIX= (1-tx)×100 explicit export taxation; tx=actual export tax rate MACROPOL=[domestic credit growth (GDCR) – real GDP growth rate (GGDP) – foreign inflation (FORINF) – rate of change of the official nominal exchange rate (GNOER)] MACROP3=(GDCR–GGDP–FORINF) GNOER=Rate of change of the official nominal exchange rate (Bank of Tanzania 1990)

AFRICA

REFERENCES Bacha, E. and Taylor, L. (1971) ‘Foreign exchange shadow prices: A critical review of the current theories’, Quarterly Journal of Economics 85:197–224. Balassa, B. (1982) ‘Reforming the system of incentives in developing economies’, in B. Balassa (ed.) Development Strategies in Semi-Industrial Economies, Oxford: Oxford University Press. ——(1990) ‘Incentives policies and export performance in sub-Saharan Africa’, World Development 18:(3). Bank of Tanzania (1990) ‘Tanzania: Exchange rate study’, Dar-es-Salaam (mimeo.). Cavallo, D. and Cottani, J. (1985) ‘Economic performance and the real exchange rate in less developed countries’, Background paper for World Development Report 1986, IBRD. Chibber, A. and Shafik, N. (1990) ‘Exchange reform parallel markets, and inflation in Africa: The case of Ghana’, PRE Working Papers, WPS427, World Bank. Collier, P. and Joshi, V. (1989) ‘Exchange rate policy in developing countries’, Oxford Review of Economic Policy 5:(3). Dornbusch, R. (1974) ‘Tariffs and non-traded goods’, Journal of Political Economy vol. 84. ——(1976) ‘Expectations and exchange rate dynamics’, Journal of Political Economy, vol. 4. Edwards, S. (1987a) ‘Exchange rate misalignment in developing countries’, CPD Working Paper, The World Bank. ——(1987b) ‘Tariffs, terms of trade and real exchange rate in an intertemporal model of the current account’, NBER Working Paper. ——(1988) ‘Real and monetary determinants of real exchange rates: Theory and evidence from developing countries’, Journal of Development Economics 29:311–41. ——(1989) ‘Exchange rate misalignment in developing countries’, World Bank Research Observer 4:(1). ——(1990) ‘Structural adjustment, stabilization and real exchange rates: Morocco in the 1990s’, Mimeo. for IBRD, Washington, DC. Edwards, S. and Ostry, Jonathan D. (1990) ‘Anticipated protectionist policies, real exchange rates, and current account’, Journal of International Money and Finance 9: 206–19. Elbadawi, I. (1989) ‘Terms of trade commercial policy, and the black market for foreign exchange: An empirical model of real exchange rate determination’, Discussion Paper no. 570, Economic Growth Center, Yale University (January). Flood, R,P. (1981) ‘Explanation of exchange volatility and other empirical regularities in some popular models of foreign exchange market’, Carnegie–Rochester Conference Series on Public Policy, no. 15. Fosu, K.Y. (1992) ‘The real exchange rate and Ghana’s agricultural exports’, African Economic Research Consortium Research Paper no. 9 (Nairobi) Frenkel, J.A. (1981) ‘The collapse of PPP during the 1970’s’, European Economic Review vol. 16. Guillaumont, P. and Guillaumont Jeanneney, S. (1991) ‘Exchange rate policies and the social consequences of adjustment in Africa’, in Ajay Chibber and Stanley Fischer (eds) Economic Reform in Sub-Saharan Africa, Washington, DC: IBRD. Hakkio, C.S. (1986) ‘Does the exchange rate follow a random walk? A Monte Carlo study of four tests for a random walk’, Journal of International Money and Finance, vol. 5. Hyuha, M. and Ndulu, B.J. (1991) ‘Inflation and economic recovery in Tanzania: Further theoretical and empirical analysis’ , University of Dar es Salaam. Jebuni, C.D., Sowa, N.K. and Tutu, K.A. (1991) ‘Exchange rate policy and macroeconomic performance in Ghana’, African Economic Research Consortium, Research Paper no. 6. Johnson, H.G. (1966) ‘A model of protection and the exchange rate’, Review of Economics

246

EXCHANGE RATE POLICY: TANZANIA

and Statistics XXXIII (2):(94), 159–63. Kaufmann, D. and O’Connell S. (1991) ‘The macroeconomics of the unofficial foreign exchange market in Tanzania’, in Ajay Chibber and S. Fischer (eds) Economic Reform in Sub-Saharan Africa, Washington, DC: IBRD. Khan, M. (1980) ‘Monetary shocks and dynamics of inflation’, Staff Papers International Monetary Fund 27:250–284, Washington, DC. Khan, M. and Knight, M.D. (1988) ‘Import compression and export performance in developing countries’, The Review of Economics and Statistics LXXX:(2). Khan, M. and Ostry, J.D. (1991) ‘Response of the equilibrium real exchange rates to real disturbances in developing countries’, IMF WP/91, Washington, DC. Kilindo, A. (1990) ‘Inflationary finance and the dynamics of inflation in Tanzania’, University of Dar es Salaam. Krumm, K. (1990) ‘Medium-term framework for analysis of the real exchange rate: Applications to the Phillipines and Tanzania’ Mimeo., IBRD, Southern Africa Department, Washington, DC Manzur, M. (1990) ‘An international comparison of prices and exchange rates: a new test of PPP’, Journal of International Money and Finance 9:(1) Mark, N.C. (1990) ‘Real and nominal exchange rates in the long run: An empirical investigation’, Journal of International Economics no. 28. Miller, S. (1984) ‘Purchasing power parity and relative price variability: Evidence from the 1970s’, European Economic Review no. 26, December. Moran, C. (1990) ‘Imports under a foreign exchange constraint’, The World Bank Economic Review 3:(2). Mussa, M. (1982) ‘A model of exchange rate dynamics’, Journal of Political Economy no. 90, February. Ndulu, B.J. (1991) ‘Growth and adjustment in sub-Saharan Africa’, in A.Chibber and S. Fischer (eds) Economic Reform in Sub-Saharan Africa, ch. 28, Washington, DC: IBRD. Ndulu, B. and Hyuha, M. (1989) ‘Inflation and economic recovery in Tanzania: Some empirical evidence’, UCHUMI 2:(1), Dar es Salaam. Ndulu, B.J. and Lipumba, N. (1991) ‘International trade and economic development in Tanzania’, in J.H.Frimpong Ansah, S.M.Ravi Kanbur, and P.Svedberg (eds) Trade and Development in Sub-Saharan Africa, ch. 9, Manchester and New York: Manchester University Press. Pinto, B. (1989) ‘Black market premia, exchange rate unification, and inflation in subSaharan Africa’, The World Bank Economic Review, 3:(3). Svedberg, P. (1990) ‘The export performance of sub-Saharan Africa’, Economic Development and Cultural Change, vol. 38.

247

13 MACROECONOMIC ADJUSTMENT AND TRADE LIBERALIZATION IN KENYA, UGANDA AND ZAMBIA Arne Bigsten and Steve Kayizzi-Mugerwa

INTRODUCTION Kenya, Uganda and Zambia should, at a glance, have a lot in common. Former British colonies, they achieved independence within a year of each other between 1962 and 1964. All, Zambia most consistently, tried at one time to use comprehensive economic planning to transform their economies. Kenya’s modern agriculture and relatively sophisticated manufacturing sector, as well as Zambia’s mining and related industry, were foreign dominated. The modern sector in Uganda, otherwise a peasant economy, was also mainly in nonindigenous hands. Nationalization of the foreign firms and/or Africanization of their leadership became vital policy concerns in the 1960s and early 1970s. The three countries depended on the primary sector for the bulk of their export incomes. However, even though the post-independence decade saw, for all three countries, increased investment, high growth, expansion of formal employment and improved welfare, there was even then considerable variation in specific areas of performance. Zambia was highly dependent on the mining sector, and Uganda on cash crops, for foreign exchange and government revenue. Kenya, though dependent on commodities like coffee and tea, had substantially diversified into manufactures and, subsequently, tourism. With the economic shocks of the 1970s, resource constraints became pronounced. Governments then resorted to extensive controls. Invariably, the state assumed control of prices, interest rates and foreign exchange. The adjustment policies pursued in the 1980s can be seen as attempts at dismantling the control regimes of the earlier decade. But here also the structure of the economy has largely determined the level of success. The purpose of this essay is to analyse the efforts of three African economies, Kenya, Uganda and Zambia, at economic adjustment and liberalization. For most of the 1980s, our sample period, the three countries pursued largely similar

248

TRADE LIBERALIZATION: KENYA, UGANDA AND ZAMBIA

adjustment policies based on the IMF/World Bank mode. In spite of this, there has been a wide variation in results. In evaluating the response, focus is thus put on the role of the economic structure, and the nature of the domestic control regimes. The study proceeds as follows. In the next section we describe the events that led to the acute crises of around 1980, and that demanded concerted adjustment efforts. In the third section we set out the analytical framework that forms the basis of our discussion. The adjustment policies and the results obtained in the respective countries during the 1980s are outlined in the fourth section, while we compare experiences and draw conclusions from this in the fifth section. THE BUILD-UP TO CRISIS After a period of relatively rapid economic development in the 1960s, Kenya, for the first time, was faced with a balance of payments crisis in 1971. This was mainly due to a combination of budgets which were too expansionary and falling terms of trade. Regarding the economic setback as temporary, the government chose to counter it with administrative regulations rather than price adjustments. There were restrictions on both bank lending and foreign exchange transactions. Price controls were imposed on a variety of goods and services. With the oil crisis of 1973, there was yet another balance of payments crisis which led to even tighter import controls. In 1975 there was a modest adjustment package which included a devaluation, but the pressure for reform was deflated by the coffee boom, which led to dramatic improvements in Kenya’s external terms of trade. For a brief period the good economic performance led to both financial and trade liberalization, and the economy expanded at close to 7 per cent per year from 1976 to 1978. Government revenue grew rapidly, but this in turn meant that expenditure control slackened. The coffee boom translated into a public expenditure boom. The government thus behaved as if the coffee price increase was permanent. However, in 1978 the terms of trade fell by 20 per cent. The imbalances which had been overshadowed by the boom now surfaced. The government tightened monetary policy, but due to the excess liquidity which had built up during the boom, the measures were initially not very effective. For instance, even though the boom had ended investments went on to reach record levels between 1978 and 1981. The combination of falling terms of trade and nonsustainable levels of budgetary expenditures meant that the macroeconomic situation deteriorated rapidly. The onset of the second oil crisis in 1979 made the situation untenable and most observers realized that drastic measures were called for. It would take another few years, though, before the government took serious action. In Uganda, a promising post-independence decade, which saw the rapid expansion of the social and economic infrastructure, was followed, at the beginning of the 1970s, by a period of deepening crisis. The sudden deterioration

249

AFRICA

in fortunes was not accidental, however. The good economic performance of the 1960s had helped patch up the unresolved political conflicts that had emanated from the constitutional difficulties of the 1950s and that, on one or two occasions had erupted into armed conflict. The military regime that ruled for most of the 1970s was faced not only by the contradictions of its erratic economic policies, including the expulsion of Asian business families to make room for Africans, but also by adverse developments in the external sector. By the mid 1970s, the institutions of government had begun to crumble. Unlike Kenya, the country was not able to take full advantage of the boom in coffee prices of 1975–7. Part of the Ugandan crop was smuggled to neighbouring countries. Though the oil shocks of 1973–4 and 1979 were debilitating, the negative growth of the economy in the 1970s was also caused by a falling rate of capital formation. From 12 per cent of GDP in 1972, total investments were less than 3 per cent of GDP at the end of the decade. Many of Uganda’s traditional aid donors had left the country by 1975. The Asian businessmen, expelled earlier, had undertaken most of the large-scale investment. In the 1960s the country had possessed a fairly diversified export base, which included coffee, tea, cotton and tobacco. The increasing domestic conflicts cut off some commodity-producing regions, however, while poor producer prices discouraged the rest. At the end of the 1970s coffee accounted for 98 per cent of official export incomes and for the bulk of government revenue. The increasing budget deficit was financed via the Bank of Uganda. Money supply grew by 30 per cent per annum during 1972–80. Thus, while at the beginning of the 1980s Kenya and Zambia had a reasonably intact infrastructure on which to base their adjustment efforts, Uganda’s socioeconomic institutions were largely in ruins. The country would be faced with the double task of institutional rehabilitation and structural adjustment, but on the basis of very limited resources. Zambia’s economic well-being has traditionally been intertwined with that of its mining sector. When mining was booming in the years that followed independence, the country embarked on policies to redress its colonial legacy. In the post-independence decade, the country had one of the highest rates of capital formation in Africa. There were huge investments in social infrastructure and in the budding import-substitution industry. To reduce wage demands in the economy, wide-ranging subsidies were instituted for food and key social services. Health and education were to be provided free of charge. At the end of the 1960s, and with copper prices at unprecedentedly high levels, the government embarked on the nationalization of the ‘commanding heights’, notably the mining companies. When this exercise had just about been completed in the early 1970s, oil prices shot up and a recession hit the world economy. Copper prices then fell, in real terms, to their Second World War levels. Zambia tried to respond to the crisis in a number of ways. Initially, the government sought bridging finances while waiting for the upturn, which never quite materialized. In 1978, the government embarked on an Action Programme

250

TRADE LIBERALIZATION: KENYA, UGANDA AND ZAMBIA

with IMF support. The kwacha was devalued and restrictive monetary policies were put into place. Agriculture was now increasingly seen as the viable alternative to copper and an ‘Operation Food Production Programme’ was announced in 1980. At the end of the 1970s all our three economies were thus in serious macroeconomic imbalance, which they tried to manage by regulations and administrative controls. Over the subsequent decade they struggled to regain macroeconomic balance, while, at the same time, trying to dismantle the system of quantitative controls. It is these adjustment efforts and the constraints thereon that we analyse and compare in this study. In the next section we set out the framework within which we undertake the analysis. A MODEL FOR A SMALL, CONTROLLED OPEN ECONOMY The model used to compare the adjustment experiences of Kenya, Uganda and Zambia is based on Bevan et al. (1990). The three economies are sufficiently similar to justify a common analytical framework. Each of the countries studied could be described as a small, controlled open economy, that is a trading economy with numerous policy interventions and where financial markets are virtually absent. Though the interventions may have distortionary microeconomic effects, concern here will be primarily on the question of the feasibility of the control regime, with regard to exchange rate, trade, fiscal and monetary policies. If, between them, the policies are at least compatible, the economies could reach a sustainable equilibrium, though not necessarily an optimal one. In the sample economies, import protection was introduced largely for balance of payments reasons, although import-substitution considerations also played a part. There have also been examples of export taxation. These trade interventions imply that, even if world market prices were to remain unchanged, the ratio of domestic prices for exportables to importables would not be constant. It is thus not possible to define a composite Hicksian good ‘tradables’. The model has instead three sectors: exportables, importables and non-tradables. Domestic prices for importables, PM, are determined by the exogenous world market prices, PM*, the exchange rate, e, and the import protection at the tariff equivalent rate tq. Domestic prices for exportables, Px, are given by world market prices, Px*, the exchange rate and export taxes, tx. Prices for non-tradables, PN, are determined by domestic supply and demand. We assume that the only asset is money and that transactions demand is proportional to planned expenditures. Money supply, d, is endogenously determined by the fiscal deficit, which is the difference between taxes, t, and government expenditures, g, the balance of payments deficit, and the policy variable, cash-to-deposit ratio, ß. The equilibrium of the model can be illustrated with the help of a diagram (Figure 13.1) in the two-dimensional space defined by the two independent relative prices, exportable to importable prices, Px/PM, and non-tradable to

251

AFRICA

Figure 13.1

importables prices, PN/PM.1 The NN curve is the locus of notional equilibria in the non-tradables market. The DD curve shows the equilibria in the money market for given income and money supply (valued in foreign currency, i.e. d/e). On combining the two loci, we can define a compatible policy set that generates a sustainable equilibrium at E. The prices at E ensure that the non-tradable and money markets clear. Assuming the government budget to be balanced, the balance of payments will be zero, and there will thus be no change in money supply. Outside E, a number of disequilibrium zones can be derived. To be able to specify these we first derive the effective loci. When there is excess supply of money, expenditures on non-tradables increase, while in the case of excess demand for money the demand for non-tradables is reduced. This will flatten the NN curve to NeNe. Excess supply of non-tradables will reduce current income below planned income. Demand for money must then be reduced if the agent is to maintain planned expenditures. The consequences of excess demand for nontradables could shift the curve either way. We assume in the figure that agents choose to build up balances for future periods when markets are expected to clear. We thus get the effective money market equilibrium locus DeDe. We also consider disequilibrium dynamics in the model. Endogenous variables are non-tradable prices, money supply, and investment. Policy variables are e, tq, tx, ß, t and g. Endogenous price changes in the non-tradables sector depend on excess supply ESN or excess demand EDN relative to NeNe. The endogenous changes in money supply depend on the balance of payments and the fiscal deficit. Increases in money supply will shift the DD locus downwards, while a decrease in money supply shifts it upwards. Investments are endogenous in the long run and capital will reallocate towards the sector with excess demand. If we

252

TRADE LIBERALIZATION: KENYA, UGANDA AND ZAMBIA

Figure 13.2

are to the east of the NN locus with excess supply of non-tradables we will see a depreciation of the capital stock in the non-tradables sector and a build-up of the capital stock in the tradable sectors. This will shift the NN curve eastwards. Let us now use the model to analyse some important policy changes. We simplify the exposition by letting the policy parameters ß, t and g jointly determine money supply here denoted d, thus reducing the number of policy variables. We distinguish between the short- and long-run impacts of policy changes as illustrated in Figures 13.2–13.5.

Figure 13.3

253

AFRICA

Figure 13.4

Let us first look at the effect of a reduction in the level of protection, that is a lower tq, implying a lower domestic price for importables. Assuming the economy initially to be in equilibrium at E, the impact effect is to shift it to point L, while the NN and DD loci are unaffected in the short run (see Figure 13.2). This is not a sustainable position, since at L there is excess supply of both non-tradables and money, and there is a balance of payments deficit. The former lowers PN until the economy has been moved to P on the NeNe curve. However, the balance of payments deficit will reduce foreign exchange reserves and eventually shift DD to D’D’ and to a sustainable equilibrium at E’. This of course requires that there are

Figure 13.5

254

TRADE LIBERALIZATION: KENYA, UGANDA AND ZAMBIA

reserves of such magnitude that the government is willing to let this happen. Otherwise policy measures reducing domestic money supply are called for to sustain the liberalization effect. Second, we could consider imposing an export tax, tx. In terms of our model the impact effect is to shift the economy vertically down from E to Z (see Figure 13.3). At this point there is excess demand for money, that is a balance of payments surplus, and excess supply in the non-tradables market. This means that the economy will move westwards in our diagram owing to the excess supply of non-tradables. In the long run the balance of payments surplus will build up money supply so that the DD curve shifts down to intersect the NN curve at the new level of domestic relative prices for tradables. In this case the relative price of non-tradables to importables will fall, while the relative price of non-tradables to exportables will increase. The exportables sector will thus be squeezed. A third policy experiment we could consider is the reduction of money supply, that is a higher ß lowering d, which would shift DD to D’D’, but would have no impact effect on prices (see Figure 13.4). In the short run the economy stays at E. What is the endogenous response to this disequilibrium? Since E is now out of equilibrium, the relevant non-tradables locus is NeNe through E’ to the left of E. At E there is now an excess supply of non-tradables, which will tend to lower PN. There will, in this case, also be excess demand for money since E is now below the D’D’ locus. This will imply reduced demand for goods and a balance of payments surplus that builds up money supply and shifts D’D’ down so as to return the economy to a sustainable equilibrium. The fourth policy change is a devaluation, that is a higher e, and both relative prices and the DD locus will be affected immediately. Note that money supply in our diagram is measured as d/e. Tradable prices increase immediately relative to those for non-tradables shifting the economy to K (see Figure 13.5). Money supply measured in foreign currency is reduced shifting the monetary equilibrium locus to D’D’. At this point the economy is running a balance of payments surplus and thus gradually increasing money supply. We cannot say for certain whether we are in a situation with excess supply or excess demand in the non-tradables market, since NeNe could lie either to the right or to the left of K. In the figure we assume that we are in a position of excess demand which leads to an increase in PN. There will again be a balance of payments surplus continuously shifting D’D’ downwards until a sustainable equilibrium is reached. We may note here that some of the endogenous adjustments outlined require the reduction of reserves. To the extent that these are depleted or are already at the lowest acceptable level, we cannot rely on this mechanism. To liberalize in these circumstances one must choose a policy combination that takes the economy to the sustainable equilibrium without having to rely on the running down of foreign exchange reserves.

255

AFRICA

COUNTRY EXPERIENCES OF MACROECONOMIC ADJUSTMENT The adjustment efforts of our three countries during the last decade or so can be described as attempts at stabilization after the exposure to external shocks. Attempts at attaining a sustainable equilibrium have been accompanied by liberalization, that is to move to a position with lower trade protection and to reduce distortions which impede efficiency. The impact effect of the shocks was to shift the economies into disequilibrium positions with balance of payments deficits. In the analysis we first trace the movements in our three economies during the 1980s, using a relative price diagram. Second, we use our model framework to study the shifts of the equilibrium loci. Both endogenous and policy impacts are included in the analysis. Kenya In Figure 13.6 we have plotted the movements in a price diagram of the actual positions for the Kenyan economy between 1979 and 1989. The nature of these movements is analysed below (see Tables 13.1–13.4 for background statistics).

Figure 13.6

256

TRADE LIBERALIZATION: KENYA, UGANDA AND ZAMBIA Table 13.1 Price indices for Kenya 1979–89 (1979=100)

Sources : Statistical Abstracts, Economic Surveys, various years Notes: PX, exportable prices. The GDP deflator for agriculture is used as a proxy PM, importable prices. The GDP deflator for domestic manufacturing is used as a proxy for the prices of import substitutes PN, non-tradable prices. The GDP deflator for building and construction, wholesale and retail trade, transport, storage and communication, hotels and restaurants is used as a proxy for non-tradable prices TOT, international terms of trade. Derived from price indices from total exports and imports of products

In 1979 the Kenyan economy was still growing (at about 5 per cent) in spite of the impact of the fall in coffee prices, but the economic policy was much too expansionary and the current account deficit was increasing. During the year there was a second oil price shock, which definitely shifted the Kenyan economy to a disequilibrium position with an unsustainable balance of payments deficit. There was thus a leakage of money out of the economy, but this was offset by an enormous budget deficit and an expansionary monetary policy, which fed more money into the system. This meant that the equilibrium locus, DD, continued to shift downwards. The policy set thus became incompatible. The policy changes that could be considered to make policies compatible again were some combination of devaluation, more restrictive monetary policy or more restrictive trade policy. The government refused to consider devaluation, which would have lowered the money balances in terms of foreign exchange. Moreover, the fiscal budget was out of control, so it was impossible to control money supply. Instead the government chose to postpone real adjustment by making the trade policy more restrictive and by resorting to financing the external deficit by short-term commercial borrowing from abroad. (The political economy of policy making in Kenya is discussed in Bigsten (1993).) This unsustainable situation persisted during 1980 and 1981, in spite of extensive discussions with the IMF and the World Bank on how to achieve a compatible policy stance. While the external terms of trade continued to fall, the policy stance remained expansionary due to the public expenditure boom. Domestic credit continued to expand rapidly in 1980 and the import regime

257

AFRICA Table 13.2 Macroeconomic indicators for Kenya 1979–89 (% growth)

a

Consumer price index for urban middle-income groups. Kenyan shillings per US dollar. c Commercial bank lending rates. d In per cent of government revenue. Sources: Central Bank of Kenya, Annual Report 1990 Statistical Abstract, various issues Economic Survey, various issues b

continued to be restrictive. It was not until 1981 that a first devaluation was undertaken under pressure from the international institutions. By 1982 fiscal policy was somewhat less expansionary, but bank credit was still expanding too rapidly. In spite of agreements with the IMF and the World Bank, unwillingness to devalue and pursue a sufficiently restrictive monetary policy meant that it was going to be virtually impossible to liberalize trade. Thus, although reduced from the year before, the balance of payments deficit in 1982 was unsustainable. During the period 1979 to 1982 the domestic prices of exportables fell relative to those of importables. One reason was, of course, the deterioration of the international terms of trade. The figures in Table 13.1 suggest that the fall in the latter was larger than that in domestic relative prices. This would seem to suggest that there was a measure of liberalization during this period, which partly compensated for the deterioration in the international terms of trade. However, the two relative prices are not strictly comparable, since the international index contains all trade while the domestic index is a proxy covering a limited number of sectors. To make a fair comparison one would need to compare PX/PM with international prices for the same categories, which we have not been able to do here. The DD curve may thus be assumed to have shifted downwards, particularly in 1979 and 1980. The downward shift in 1979 was due to the too expansionary monetary policy, which in 1980 was offset by the large balance of payments deficit. In 1981 and 1982 devaluations helped halt the drift. Over the period there was hardly any shift of the NN locus, which may seem surprising given the expansionary policies pursued. Still, the effect of these was counteracted by the

258

Sources: IMF, International Financial Statistics, various years

Table 13.3 Kenya: balance of payments statistics 1980–8 ($US millions)

AFRICA Table 13.4 Real sectorial and GDP growth rates in Kenya, 1979–89 (1979=100)

Source: See Table 13.1

effects of the investments in non-tradables during and immediately after the coffee boom. The outputs from these investments now came on to the market and held prices down. The drastic fall in the terms of trade also affected real incomes negatively, which reduced demand for non-tradables. All in all, the policy stance was unsustainable throughout the period 1979 to 1982. Major changes in economic policy were not initiated until after the attempted coup in 1982. The exchange rate regime was now based on a crawling-peg system which gradually moved the exchange rate towards more realistic levels. Trade policy reform was slow, although moves towards less restrictive import controls were being made. In 1983 the domestic relative prices actually shifted in favour of exportables in spite of the continued fall in the external terms of trade. The monetary and fiscal policy was now under reasonable control. This, together with a depreciation of the currency, led to a fall in the d/e ratio. The DD curve thus shifted upwards. At the same time the NN curve shifted eastwards. The reason for this was thus not expansionary policies, but rather slow growth of supply. The contraction since the early 1980s had significantly affected investments, which fell by four percentage points of GDP between 1980 and 1982. Thus, we would argue that by 1983 policy makers had managed to bring about stabilization and create a compatible policy stance. Now the focus could shift to the restructuring of the economy which among other things implied a liberalization of the trade regime. Another dramatic year was 1984 with the worst drought for half a century coupled with a mini-boom in tea. This meant that there was a combination of terms of trade improvements and stagnation in production. The monetary policy again became more expansive and fiscal control slackened once more, partly due to the expenditure needs created by the drought. The expansive policies tended to increase money balances and to shift the DD curve downwards, but at the same time the gradual devaluations due to the crawling peg (higher e reducing d/e) counteracted this and the net effect was that the locus changed little in 1984. The government thus still managed to sustain a compatible policy set in 1984. The following year terms of trade came down again, but the process of trade

260

TRADE LIBERALIZATION: KENYA, UGANDA AND ZAMBIA

liberalization continued so that all in all the domestic relative price PX/PN actually increased. The macroeconomic policy was somewhat looser than earlier, but the depreciation was still enough to compensate for this and to prevent a downward drift of the DD schedule. The situation also in this year may be considered to have been a sustainable equilibrium. Then the year 1986 saw a positive shock, in the form of a new boom in coffee prices combined with falling oil prices. This meant that the terms of trade improved. Fiscal and monetary policy again became expansive, but even this time the depreciation of the currency was enough to prevent a downward shift of the DD curve. The boom conditions and the expansionary policies, however, shifted the NN curve to the east. The position for 1986 in Figure 13.6 may still be considered to have been a sustainable equilibrium in the sense defined above, although this might be debated, since the long-run sustainability of the policy stance depended on the coffee prices remaining. Since the coffee price increases even this time were temporary, the government repeated the errors of the 1970s although on a smaller scale. Thus when the terms of trade started to come down in 1987 the economy again shifted to a position with a non-sustainable balance of payments. The fiscal deficit increased rapidly and the monetary policy was too expansionary at the same time as there was a continued liberalization of trade. Still, the continued depreciation of the currency hindered a downward shift of the DD schedule and checked the extent of the disequilibrium. Still, for the first time since 1982 the economy was back in an unsustainable position, although less serious than in the early 1980s. In 1988 there was a move towards a sustainable equilibrium with a reduced budget deficit and a reasonable credit expansion. The fall in real income due to the terms of trade loss may have been the cause of the westward shift of the NN locus. Between 1983 and 1988 producers saw a gradual increase in relative exportable prices largely due to domestic liberalization. At the same time, relative non-tradable prices increased due to low investment. The period can be characterized as one in which some successful attempts at dismantling the control regime set up during the 1970s were made. Reductions of trade interventions were under way and the introduction of the crawling peg had improved the competitive position of exporters. In other areas, the liberalization or structural adjustment efforts had been somewhat less successful, but the distortions in the economy were less crippling than in the early 1980s. In 1989, the international terms of trade fell particularly due to lower coffee prices. In the face of increasing balance of payments problems, the restrictiveness of the trade regime increased once again. The shift towards the south-west in Figure 13.6 was again a result of lower relative prices for the exportables and nontradables, the latter being caused by a restrained economic policy and possibly increased supply following high investments during the recovery since the mid 1980s. It should be noted that Kenya possesses other sources of export revenues such as tourism, which are not fully captured in this diagrammatic analysis. Still,

261

AFRICA

if export prices remain at their present levels, the demands on the stabilization policy will be very difficult to meet. Fortunately for Kenya there was a major increase in foreign aid inflows which helped keep the balance of payments under control. Given the level of foreign transfers, we could argue that the situation in 1989 was sustainable. However, while a sustainable policy stance is necessary, it is not sufficient for the removal of distortions and imperfections that constrain development. Much more is required for the country to make a breakthrough, for example in export markets for nontraditional products. The changes required touch on many areas of a mainly microeconomic nature. To be able to get on with these changes the minimum requirement is that the macro regime is compatible. Kenya has, after all, been more successful in this respect than most other countries in the region, including the other two sample countries. Uganda Uganda’s adjustment experience in the 1980s can be divided into three phases: 1981 to mid 1984 was a period of optimism as the economy seemed to respond favourably to the economic incentives in the adjustment package introduced in June 1981. The period mid 1984 to 1986 was, in contrast, marked by a degree of policy regression as domestic politics, notably the intensified civil conflict, replaced macroeconomic concerns at the centre stage. The third phase, beginning in 1987, marked the hitherto most concerted effort at getting the economy off the ground. Unfortunately for the new government, it coincided with a drastic worsening of the external environment, as coffee prices fell. With regard to the specific goals of the adjustment programme in the 1980s, there was, at the end of the decade, still considerable doubt as to the level of success achieved. We use Figure 13.7 and Tables 13.5–13.8 to discuss macroeconomic changes in Uganda during the three periods. With advice from the IMF/World Bank, the new government (Obote II) floated the Uganda shilling in June 1981 and introduced restrictive fiscal and credit policies. The float led to a depreciation of the shilling. Table 13.6 shows that there were sharp decreases in the growth of domestic credit and money (M1) in 1982. Money supply grew by 5 per cent (103 per cent in 1981). The restrictive macroeconomic policy stance reduced demand for non-tradables, which meant that the relative price of non-tradables fell. Though the international terms of trade rose by about 27 per cent between 1981 and 1982, the domestic equivalent, PX/PM, fell by 15 per cent (Table 13.5). The implicit level of protection of domestic producers of importables thus increased so that initially the currency depreciation was not accompanied by a trade liberalization. In the configuration of Figure 13.7, the economy moved south-west between 1981 and 1982, a movement towards a situation with continued excess supply of money and a balance of payments deficit. In August 1982, the government introduced a dual exchange rate regime

262

TRADE LIBERALIZATION: KENYA, UGANDA AND ZAMBIA

Figure 13.7

(window 1 and window 2). Window 1, the favourable rate, was reserved for government business, traditional exports and debt servicing. The exchange rate at window 2, for all other importers, was determined by weekly auctions. In real and relative terms, the years 1982–4 saw vast improvements in producer prices for all major export commodities, notably coffee. This was in spite of the fact that the international terms of trade for the country had fallen sharply in 1983. Figure 13.7 shows that from 1982 to 1984 the economy moved in a northeasterly direction. Exports were improving (see Table 13.7), mainly due to increased coffee production, or reduced smuggling of the crop, in response to better prices for farmers. The overall balance was still negative, however. On the whole, the size and pace of aid and loan disbursement fell short of government expectations. However, the increased inflow of resources from abroad helped pay for a larger volume of imports. Though in retrospect the adjustment challenge was still formidable, the improvement in economic performance achieved between 1981 and 1983 seemed remarkable. The galloping inflation of the earlier years was checked, the growth of domestic credit was not excessive, manufacturing output (see Table 13.8) was responding strongly to increased availability of imported inputs and, above all,

263

AFRICA Table 13.5 Price indices for Uganda 1981–9 (1981=100)

Source: Statistical Bulletin No. GDP/2, Gross Domestic Product of Uganda 1981–1989, July 1990 Definitions: Note that export prices are for products, while the remaining ones are GDP deflators and thus for value added PX, exportable prices. Producer coffee prices (robusta) is used as a proxy for exportables prices. In recent years coffee has constituted 97 per cent of commodity exports PM, importables prices. The GDP deflator for domestic manufacturing is used as a proxy for import substitutes prices. PN, non-tradables prices. The GDP deflator for utilities, construction, retail and wholesale, transport and community services is used as out proxy for non-tradables prices TOT, international terms of trade

Table 13.6 Uganda: macroeconomic indicators 1980–9 (% growth)

a

Kampala cost-of-living index: low-income group 1980–2, then middle-income group. Figures are commercial bank lending rates. c Derived from annual wage bill for selected manufacturing industries. d Uganda, shillings per US dollar. e In per cent of government revenue. Sources: Bank of Uganda, Annual Report 1985 Ministry of Planning and Economic Development, Background to the Budget (various years) IMF, International Financial Statistics (various years) World Bank, World Tables, 1988/89 edition Statistical Bulletin No. GDP/2, Gross Domestic Product of Uganda 1981–1989, July 1990 Statistical Bulletin No. CPI/1, Consumer Price Index Kampala, October 1990 Background to the Budget, 1982–3 b

264

TRADE LIBERALIZATION: KENYA, UGANDA AND ZAMBIA Table 13.7 Uganda: balance of payments statistics 1980–9 ($US millions)

Source: Key Economic Indicators, 3rd issue, October 1990 Notes: Arrears are included in the overall balance. Increases in net arrears (+), decreases (–). The net overall balance is financed by changes in gross reserves, the net position vis-à-vis the IMF or other (including errors and omissions).

and from the point of view of the government, the foreign donors voiced satisfaction with the level of progress. Policy makers were optimistic. The year 1984 was something of a turning point in the liberalization effort, for in spite of the increase in both the domestic and international terms of trade (Table 13.5), the optimistic trend of the past two years was reversed. Three factors explain this setback: first, the escalating guerilla insurgence pushed government expenditure beyond planned limits. The earlier discipline shown in meeting IMF performance criteria was now clearly lacking. Second, the improvement in the economy was impressive only in relation to the earlier and very poor performance. In fact, the economy was still extremely fragile. Third, the ‘fourth parliament’ was drawing to an end and government was already gearing for the coming elections. Among the measures taken during this time, definitely outside the IMF/World Bank mode, was a fourfold increase in public sector wages. Credit to government increased by 70 per cent in 1984 in comparison to 25 per cent in 1983. Money supply expanded by 127 per cent. This expansionary stance led to an increase in PN/PM. At the same time the government let the relative prices of exportables increase, although this was largely due to an improvement in terms of trade. The record increase in export revenue was probably at least partly due to favourable terms of trade (TOT) movements, while the fall in imports was partly a response to increased regulation of the ‘windows’. There was also the need to reduce arrears, by $US77 million in 1984, as part of the conditionality attached to IMF/World Bank funds. GDP fell by 16 per cent in 1984, as both manufacturing and agriculture showed considerable contraction (see Table 13.8). The policy stance was thus still not compatible, as was reflected by the large balance of payments deficit. The latter half of 1985 saw a jump in inflation, due mainly to shortages caused by increased insecurity and the looting that followed the fall of Obote II. There

265

AFRICA Table 13.8 Sectorial and GDP growth rates for Uganda, 1981–8 (1981=100)

Source: Statistical Bulletin No. GDP/2, Gross Domestic Product in Uganda 1981–1989 Note: Government includes general government, health and education. In the latter two categories some private provision is included.

was also disruption in the export trade, as part of the coffee produce was smuggled out of the country. The regime which replaced Obote II, in July 1985, continued with the auction but without the earlier enthusiasm. The relative price changes in favour of exportables of the past few years were now reversed (see Table 13.5). The domestic terms of trade fell by 9 per cent, mainly reflecting unfavourable shifts in the international terms of trade. Money supply (Table 13.6) rose by 140 per cent and credit to government by 115 per cent. The latter in part was in response to the new government’s need for finances to secure its delicate military and political position. The military claimed 30 per cent of government expenditure in 1985 (Bank of Uganda 1985). Inflation went up to 125 per cent in 1985. The rapid price increases in the importable sector were not fully compensated by commodity (coffee) producer price increases. Part of the increase in import prices was due to the shortages caused by the administrative and transport disruption that followed the military coup. Imports fell to $US 200 million in 1985, and the implicit tariff increased. With exports at $US 380 million, the overall balance was positive. Though somewhat paradoxical, the causes were fairly obvious: imports of items like petrol had simply ceased and the bulk of manufacturing was left to fend for itself with regard to imported intermediates. There was a fall in non-tradable prices owing to the general contraction of the economy, and the slippage in liberalization policy caused by political difficulties led to lower domestic relative prices for exportables. The sharp reduction in imports still implied that there was an overall surplus. In 1986, yet another new government took power in Kampala. In the course of the year, the international terms of trade for Uganda improved due to a mini coffee boom. However, the domestic terms of trade actually fell in 1986. The degree of protection of the importable sector thus increased during the first year of the NRM government. The new government favoured a ‘strong’ shilling and united the dual exchange rate at the much appreciated rate of 14 shillings (new

266

TRADE LIBERALIZATION: KENYA, UGANDA AND ZAMBIA

currency) to the US dollar (the parallel rate was more than six times that). Thanks to substantial net transfers, unrequited transfers amounted to over $US 200 million in 1986, and the deficit on the current account was small (see Table 13.7). Money supply increased rapidly in 1986 and the DD curve thus shifted downwards. At the same time there was a modest increase in the demand for nontradables, while the prices of importables were allowed to increase rapidly, which shifted the NN curve westwards. The level of import protection thus increased in conjunction with the unwillingness to let the exchange rate depreciate. In the configuration of Figure 13.7, there was thus a south-west shift to a position characterized by a higher level of protection. An Economic Recovery Programme was launched in 1987. The package included measures to reduce external imbalance, rehabilitate domestic productive capacity and restructure the public sector. Supported by the IMF, the reform saw the currency depreciate from 14 to 60 shillings to the US dollar. The gap between the official and parallel rates was substantially reduced. To rationalize foreign exchange allocation, and thereby achieve a degree of import liberalization, a system of open general licensing (OGL) was started to provide foreign exchange to industries where a quick output of consumer goods, thereby reducing domestic excess demand and contributing to government revenue via excise duties and sales tax, could be realized. This was later supplemented by Special Import Programmes (SIPs) that made foreign exchange available to all importers on a first-come, first-served basis. From a low base, economic growth increased to 7 per cent in 1987. In spite of these, at least on paper, major reforms, the overall picture remained precarious. On the external front, coffee prices had fallen and Uganda was allotted an export quota which was much smaller than her coffee stocks. Thus, in spite of trade liberalization measures, the terms of trade decline implied that domestic relative prices for exportables fell. Ironically, there was now talk of the need to restrain coffee production and to concentrate on quality. Exports (see Table 13.7) were much smaller than imports, leading to a huge trade deficit for both years. There were also substantial lags between commitment from foreign donors to supply loans and credits and actual disbursement. In November 1987, the Uganda-Kenya border was closed after armed clashes. On the domestic front, the government needed to contain armed insurgency in parts of the country. These factors led to increased military expenditure. As Table 13.6 shows, money supply continued to expand in response to a net increase in credit to government, with unfavourable impacts on inflation, which went up to 237.1 per cent in 1987. In spite of the currency depreciation the DD curve was shifted downwards by the expansionary policies, and there was again a small deficit in the overall balance. The July 1988 budget announced a large devaluation, from 60 to 150 shillings to the US dollar, and a further liberalization of prices. The growth of domestic credit was somewhat reduced and, owing to increased disbursement from donors, there was now more foreign exchange available. The low international coffee prices and thus lower government revenue coupled with tight credit policies of the

267

AFRICA

Bank of Uganda sparked off a liquidity crisis for the government. The net effect of these forces was to push the DD curve further down, while the NN. curve was shifted slightly to the east by the higher domestic demand due to the recovery of the economy. The relative price of exportables to importables fell owing to higher price increases in the importables sector than in coffee. All in all the economy was in a position with an equilibrium that one could say was sustainable given the aid inflow, but one which was increasingly characterized by a high degree of protection. Further attempts were made to buttress the adjustment effort in 1989. Two devaluations were undertaken which saw the exchange rate depreciate from 165 to 340 shillings to the US dollar. With the continuing downward slide of coffee prices, after the collapse of the International Coffee Agreement in July 1989, the incentives for coffee farmers were still not improved. The government was trying to protect its dwindling export tax revenue (see Table 13.6). Since domestic coffee prices were not adjusted, the relative price of exportables instead deteriorated drastically. The financing of coffee purchases, managed by the Bank of Uganda in a bid to speed up payments to farmers and to curb corruption at the co-operative union level, proved fiscally impossible to sustain and would be abandoned to commercial banks. The latter were supposedly in a better position, than the central bank, to assess the creditworthiness of the coffee-purchasing co-operative unions and that of any other middlemen. In spite of the devaluations, the policy stance was expansionary enough to continue, together with aid inflows, to shift DD downwards. At the same time the NN curve continued to shift eastwards due to the economic expansion. We thus end up in an equilibrium that may be characterized as sustainable, but where the tariff equivalent is very high. In summary, the policy makers in Uganda have in recent years managed to achieve a sustainable policy stance, but one which is tremendously precarious. It is highly dependent on continued and substantial inflows of foreign aid, and one in which the public sector is extremely weak and inefficient. The latter provides poor services, while a very large part of the small budget is allocated to the military. The stance is also characterized by a restrictive trade regime, with negative allocative consequences which will tend to slow down growth. Zambia In May 1980, Zambia’s president presented the ‘Operation Food Production Programme’. As a blueprint for the country’s transformation from a copper to an agricultural economy, it demanded vast investments in machinery and infrastructure. Unfortunately, the country had not yet recovered from the ravages of the second oil shock (1979) nor had the Action Programme of 1978, put together with IMF help, had any appreciable impact on the fiscal and balance of payments disequilibria. All efforts at introducing ‘economic prices’, as outlined in the Action Programme, had failed.

268

TRADE LIBERALIZATION: KENYA, UGANDA AND ZAMBIA

Figure 13.8

The brief improvement in copper prices in 1980 led to a substantial trade balance surplus (see Table 13.11). The impact on money supply via reserves was neutralized by the large negative service payments (net), which exceeded the capital inflow. Money supply fell by 0.7 per cent in 1980. Inflation was at 11.5 per cent. During 1981 and 1982 the government pursued a fairly expansionary policy at a time when the exchange rate changed very little. As a consequence, the DD curve shifted down. Copper prices weakened markedly after 1980 and the external terms of trade deteriorated. This was reflected in the fall of the domestic relative price of exportables (see Figure 13.8). However, the country successfully negotiated an extended fund facility with the IMF worth SDR800 million in 1982. This ‘growth oriented’ financial package, it was hoped, would put the country back on the road to recovery. It emphasized reductions in aggregate demand and advocated a freeze on wages and government employment. Prices were decontrolled and interest rates were increased in a bid to encourage savings and efficient use of capital. The kwacha was devalued by 20 per cent (see Table 13.10). However, by 1983 the large programme was in trouble as the country failed to meet the IMF performance

269

AFRICA Table 13.9 Price indices for Zambia 1980–9 (1980=100)

PX , exportable prices. The GDP deflator for mining is used as a proxy PM, importable prices. The GDP deflator for domestic manufacturing is used as a proxy for the prices of import substitutes PN, non-tradable prices. The GDP deflator for utilities, construction, retail and wholesale trade, transport, hotels and restaurants, financial institutions and real estate is used as a proxy for non-tradable prices Sources: CSO, National Accounts Statistics Bulletin No. 1, 1987 National Commission for Development Planning, Economic Report 1989, 1989 World Development Report, various issues

criteria. Domestic inflation rose by the unprecedented rate of about 20 per cent in 1983 and 1984. The pace of currency depreciation was not enough to shift the DD curve significantly upwards. Arrears were building up. The economic decline reduced demand for non-tradables shifting the NN curve somewhat to the west. The situation was one of a balance of payments deficit (see Table 13.11), and the policy stance was not sustainable. Another difficult year for Zambia was 1985. The rising inflation blamed on the floating of the kwacha had eaten away at public sector wages. The year had begun with a go-slow by junior doctors demanding higher pay. In June miners went on strike. In September the prices for mealie, the staple food, were increased, on average, by 50 per cent. With regard to economic policy, far-reaching steps were taken in October 1985. The allocation of foreign exchange by FEMAC was discontinued. Foreign currency would henceforth be auctioned. There would, however, still be exceptional foreign exchange allocations: party and government, debt servicing, use by ZCCM (under the retention scheme), oil and medical supplies, educational supplies and IATA bills for Zambian Airways. The reasons given for the drastic moves were as follows: high oil prices, copper prices were at their lowest levels since the Second World War; a general increase in prices for imported intermediates; debt service arrears had risen to $US 500 million; the copper industry was marginalized; and the country was said to be facing the most serious unemployment situation since independence. It was thus necessary to ‘reexamine the system and to see in what way it can be made less subjective and less dependent on the human factor’. In the first week of the auction, the kwacha depreciated from K2.35 to K5.01

270

b

Consumer price index for urban higher income groups. Actual central bank rate. c In per cent of government revenue. Source: Monthly Bulletin of Statistics, CSO, Lusaka, various issues

a

Table 13.10 Macroeconomic indicators for Zambia 1980–9 (% growth)

Sources: UNCTAD, Handbook of International Trade and Development Statistics, Geneva IMF, Balance of Payments Statistics, 1989

Table 13.11 Zambia: balance of payments statistics 1980–9 ($US millions)

TRADE LIBERALIZATION: KENYA, UGANDA AND ZAMBIA Table 13.12 Zambia: real sectorial and GDP growth rates, 1980–9 (1980=100)

Source: See Table 13.9

to the US dollar. After reaching K7 to the US dollar at the end of October, it gained some strength in November but was slipping again by Christmas. The contractionary policies pursued during 1985 reduced demand for nontradables. At the same time the depreciation of the kwacha together with restrictive monetary policy shifted the DD curve upwards. The devaluation led to an increase in the domestic relative price of exportables, and thus to a measure of liberalization. The 1986 budget, in the spirit of the earlier moves, declared its ambition ‘to force the economy to live within its means and to set the stage for further growth’. The budget deficit was nevertheless a record Kl,789 million. This was reflected in a sharp increase in domestic credit (Table 13.10). As the pressure on the kwacha continued unabated, the general goal of ‘a more acceptable and stable exchange rate’ was still elusive. To restrain credit and boost the kwacha on the auction market, commercial bank deposits with the central bank were increased. Firms bidding for foreign exchange would have to show that goods ordered three-four months earlier had indeed arrived in the country and that their income taxes were cleared. These measures did not seem to be sufficient and in July it was decided to publish the names of all foreign exchange bidders, with the hope of bringing some social pressure to bear. As a further insurance against excessive bidding, a ‘Dutch auction system’ was introduced whereby bidders obtained foreign exchange not at the marginal bid but at the level of their bid. Though the 1986 harvest saw a bumper maize crop, there was a total of 12 million bags of which 8.5 were marketed. The undercurrent of dissatisfaction at the way the economy was performing was unmistakable. By the end of October 1986 there were calls for the scrapping of the auction from several quarters. The General Secretary of the Zambia National Council of Commerce and Industry (ZNCCI) noted that the auction had done ‘untold misery to consumers and squeezed business against the wall’. At the end of 1986, another effort was made to stabilize the currency. On top of the earlier stipulations, bidders were now supposed to make interest-free advance payments of 30 per cent of the kwacha value of their bids. This, it was hoped,

273

AFRICA

would help siphon off excess liquidity. Bank interest rates were raised from 25 to 30 per cent as well as minimum reserve requirements against demand and time deposits from 20 to 25 and 12 to 17 per cent, respectively. Preferential interest rates for agriculture would continue. Following these moves, and confirming the speculative mood in the economy, there was a heavy run on the kwacha. By 30 November the US dollar was selling at K15.25. Furthermore, payments arrears were building up. The Paris Club estimated that the country needed $US 550 million in foreign funding to make ends meet. The benefits of the reforms were thus slow in coming, although agriculture was showing some positive signs (Table 13.12). Thanks to the auction domestic relative prices of exportables increased substantially in spite of falling external terms of trade. The policy stance was still expansionary and relative non-tradable prices were starting to increase (see Table 13.9). The balance of payments for 1986 was very negative (Table 13.11). At the beginning of 1987, the government, increasingly disillusioned with the workings of its venture into market forces, decided to control the dollar-kwacha rate letting it fluctuate between 9 and 12.5 kwacha. Interest rates would be raised from 20 to 30 per cent. Subsequently, the auction was suspended for two months and resurrected in the form of a two-tier system with a priority rate and the auction rate. The kwacha slid further and was K18.75 to the US dollar by April 1987. When it came, Zambia’s break with the IMF was hardly surprising. Everything had been pointing to some dramatic climax. Academics, students and civil servants had become increasingly restive and tension was mounting on the Copperbelt. In an indignant speech, President Kaunda recounted the close to ‘two decades of fruitless cooperation with the IMF’, during which period there had been ‘mounting social and political tensions and a drastic fall in the living standards of the people’. The auction was suspended and the kwacha was pegged at 8 to the US dollar, a large appreciation from the K21.01 reached at an earlier auction. There was a general price freeze and all the earlier credit and foreign exchange controls were reintroduced. Bank interest rates were reduced. A New Economic Recovery Programme (NERP) with its own alternative strategies was launched. Its main ideas were to be expanded in the ambitious Fourth National Development Plan (FNDP) (see Kayizzi-Mugerwa 1990). It is noteworthy that the army, through its commander, pledged total support for the economic measures that had been adopted. In 1987, therefore, there was a radical turn around from the policies of demand restraint and currency depreciation. The emphasis shifted to expansionary policies in an attempt to revitalize the economy. The combination of currency overvaluation and monetary expansion led to a drastic shift downwards of the DD curve (see Figure 13.8). Although the international terms of trade were favourable (see Table 13.9), there was increased restrictiveness in the import regime, as shown by the fall in PX/PM. This could, however, not return the country to a

274

TRADE LIBERALIZATION: KENYA, UGANDA AND ZAMBIA

compatible policy stance. The external deficit was out of control, and foreign indebtedness increased. The NERP provided little respite to policy-makers. Smuggling and black marketeering and police raids became visible expressions of the disequilibria that engulfed the economy. There were now accusations that the allocation of foreign exchange was shrouded in secrecy. The hospitals and health centres lacked drugs and bribes were now necessary in most hospitals. The ‘cash first, pill later’ syndrome had set in. Inflation at 54 per cent meant that interest rates, at 17.5 per cent, were negative in real terms (see Table 13.10). Food subsidies were still a thorny problem. Food coupons were devised to target subsidies to the poor and to stop ‘other nations eating subsidised food from Zambia’. On the whole, the situation in 1988 was better than 1987, partly due to improvements in the external terms of trade. For the first time since 1981 per capita incomes actually increased. Money supply growth slowed down somewhat compared with the previous year, but with the appreciation of the currency the DD curve continued to shift downwards (Figure 13.8). The implicit tariff equivalent continued to increase, though, and this contributed to the improvement in the balance of payments (see Table 13.11). With the sense of stabilization, the government felt confident enough to invite the IMF to look at its economic recovery programme and ‘if o.k. to join us’. In early 1989, Zambia was hit by heavy rains and most of Lusaka was flooded. The Western Province was declared a disaster area. By March, policy makers began preparing people for new policy initiatives. The youth were asked to ‘go back home and farm’. University students were to start paying school fees. The economic climate, it was said, was not conducive to the provision of free services. Short of financial resources, there was now increasing realization that Zambia would have to come back to the IMF/World Bank fold if her adjustment efforts were to continue. This was also impressed on the government by the donor community that insisted that it undertake a well-defined programme that had the sanction of the IMF before new credits could be agreed. At the beginning of July 1989, prices of all commodities, except maize, were decontrolled. The role of the Prices and Incomes Commission (PIC) was changed from control to market research. It was with some elation that the government announced that the IMF had found the measures formulated in the FNDP to be adequate. But no funds were as yet forthcoming from the multilateral organs. The change of currency notes in 1989 gave no respite. Inflation was at over 112 per cent and black marketeering persisted, including the smuggling of copper and cobalt. Although international copper prices continued to be favourable (Table 13.9) domestic economic performance was mediocre. The Kafue Gorge power disaster led to the rationing of electricity. Lacking a firm agreement with the IMF on economic policy meant that aid flows would be low. The poor domestic transport net reduced the flow of imports and the marketing of agricultural goods.

275

AFRICA

COMPARISONS OF ADJUSTMENT EXPERIENCES In the previous section, we undertook detailed descriptions of the macroeconomic experiences of Kenya, Uganda and Zambia in the 1980s. Let us now undertake a comparative analysis of the factors that have determined the differences in crisis response. From Figures 13.6–13.8 we may note that when the exchange rate is overvalued and, at the same time, money supply is expanding governments are forced to adapt a less liberal trade policy stance. This implies a move southwards in our diagrams. Kenya stands out as having been able to achieve a degree of macroeconomic stability combined with a move towards a more liberal regime. Except for 1981–2, the rate of inflation in Kenya was consistently below 12 per cent in the 1980s. This, and the crawling-peg regime, prevented an excessive overvaluation of the exchange rate, enabled modest increases in per capita incomes and prevented the wide swings evident in the other two sample countries’ performance. The configurations of Uganda and Zambia show, on the other hand, movements in the opposite direction. Probably as a reflection of the seriousness of the economic crisis there, relative price movements in Uganda were extreme. Uganda and Zambia faced many hurdles in their adjustment efforts, which have put the brakes on their liberalization policies. To balance their external accounts it was necessary to tighten import controls. This has driven up the tariff equivalent. An important observation is that to delay economic adjustment, until the economic imbalances have become large, is hardly good policy. For eventually the country would have to adjust to internal and external disequilibria bringing about very drastic changes in relative prices and in public sector expenditure programmes. This inevitably implies drastic reductions in the income levels of groups which often are politically influential. For instance, urban residents may revolt and further delay adjustments or could make measures ineffective. The socio-economic structure of the economy is an important determinant of the impact and character of adjustment policies. In all three countries, currency depreciation via devaluation has been controversial. It has been used with much reluctance. The main reason for this is because devaluation has impacts on relative prices and by implication the welfare of certain groups. A currency depreciation tends to increase prices for exportable producers. In Kenya and Uganda these are mainly to be found in the rural areas, while in Zambia they are in the mining industry. In political terms, Uganda’s peasant farmers, the main producers of coffee, do not constitute a very influential pressure group. In Kenya, where agriculture is more diversified and sophisticated and where land ownership and land shortage are politically sensitive issues, agriculture has been central to economic policy. A number of key political figures have a stake in agriculture. In Zambia, the bulk of the mining industry is owned by the government. While miners might not necessarily belong to the government base support group, it has often been deemed expedient not to overlook their demands. Those that are hurt by higher import prices, on the other hand, belong to the urban population, which

276

TRADE LIBERALIZATION: KENYA, UGANDA AND ZAMBIA

forms the basis of support for the government. When these react negatively, the government may have to retreat from its policies as has been the case in Zambia. The import-substituting industrialists are hurt by a devaluation since it drives up the costs for imported intermediates and raw material, although as long as quota regulations or tariffs remain in place they still are partly protected. However, once major liberalization is being considered these groups become more vocal in their protest. This has taken place in various degrees in Kenya, Uganda and Zambia. In all three countries foreign born traders, often better able to cope with the changed circumstances, have been blamed for ruining the nation’s enterprises via economic sabotage. Since a liberalization policy almost always will require devaluation plus a restrictive monetary policy, the government will have a hard time selling this policy to import-competing industrialists. As primary exporters, our sample countries have, to some extent, depended on the rent accruing from the exploitation of immobile resources, mineral-laden land in Zambia and land and geographical conditions suitable for coffee growing in Kenya and Uganda. The presence of these resources has over the years generated substantial rents which have accrued to the owners of the resources, to the state or to other economic actors. During a period of economic expansion, based on a boom in the prices of the domestic resource, governments, via taxes, royalties and share ownership, accumulate large amounts of revenue. How to manage rents once they have been captured has always been a problem, but only in retrospect. In the course of the boom itself, governments embark on spending sprees and on the expansion of political patronage. In the case of the coffee boom in Kenya it turned out to be impossible to manage rents effectively when they came in the form of a temporary boom. In a country where the needs are vast and obvious it has turned out to be extremely difficult to control government expenditures and to let expenditure levels be lower than revenues during the boom period. Governments have used several methods to extract the rent. In Kenya, coffee producers have been able to retain most of the land rent since export taxes have been very modest or nil. In the case of Uganda taxes on coffee exports have, until the recent slump, been the main domestic generator of government revenue. This has been one reason for the reluctance to increase coffee prices. In Zambia, finally, taxes on copper exports have been the basis for a whole array of economic activities, while the miners have been the wage leaders in the modern sector. The governments of Zambia and Uganda have thus been very rent dependent, while that of Kenya has been less so. Stability in government policy requires a sustainable underlying growth process which is not wholly dependent on the primary resource, otherwise parts of the modern sector could be wiped out by a bust, as has been the case in Zambia. Kenya was more diversified in exports (tourism) and could manage the decline in coffee revenue much better. Uganda, on the other hand, has been totally dependent on coffee for export revenue (99 per cent in the 1980s) and had to rely on foreign loans and aid to try and counter the impact of the decline in coffee prices.

277

AFRICA

In terms of economic adjustment, governments which are highly rent dependent are faced with a dilemma. The sustenance of government requires the capture of a large rent. On the other hand, the rent captured should not be so large as to ruin the rent-generating sector. The policy conflict may become particularly acute in cases where negative exogenous shocks reduce export tax revenue. The government could then be inclined to increase export taxes in order to protect the revenue levels, while, on the other hand, it needs to reduce export taxes to compensate producers and to preserve the export base for the future. There may thus be a sharp conflict between the political necessity of sustaining government activities in the short run and the economic one of preserving the export sector. ACKNOWLEDGEMENTS Kayizzi-Mugerwa gratefully acknowledges financial support from the Swedish Institute, Stockholm. NOTES 1 See Bevan et al. (1990) for a detailed derivation.

BIBLIOGRAPHY Andersson, P.-A. and Kayizzi-Mugerwa, S. (1993) ‘External shocks and the search for diversification in Zambia’, in M. Blomstrom and M. Lundahl (eds) Economic Crisis in Africa. Perspectives on policy responses, London: Routledge. Bank of Uganda, Annual Report 1985. Bevan, D., Collier, P., Gunning, J. with Bigsten, A. and Horsnell, P. (1990) Controlled Open Economies. A Neoclassical Approach to Structuralism, Oxford: Clarendon Press. Bigsten, A. (1993) ‘Regulations versus price reforms in crisis management: The case of Kenya’, in M. Blomstrom and M. Lundahl (eds) Economic Crisis in Africa. Perspectives on policy responses, London: Routledge. Bigsten, A. and Kayizzi-Mugerwa, S. (1992) ‘Adaption and distress in the urban economy: A study of Kampala households’, World Development 20: (3). ILO (1981) Planning for Basic Needs in an Economy Under Pressure, Addis Ababa: Jobs and Skills Programme for Africa. IMF, International Financial Statistics, various years. Kayizzi-Mugerwa, S. (1990) ‘Growth from own resources: Zambia’s Fourth National Development Plan in perspective’, Development Policy Review 8:(1). ——(1991) ‘External shocks and adjustment in a mineral dependent economy’, World Development 19:(7). ——(1991b) ‘Urban bustle rural slumber: Dilemmas and uneven recovery in Uganda’ in M.Blomstrom and M.Lundahl (eds) Economic Crisis in Africa. Perspectives on policy responses, London: Routledge. Kenya, Development Plan, 1979–84,1984–8. ——Budget Speech, annual. ——Economic Survey, annual. ——Statistical Abstract, annual.

278

TRADE LIBERALIZATION: KENYA, UGANDA AND ZAMBIA

——Sessional Paper no. 4 of 1982, On Development Prospects and Policies, The Government Printer, Nairobi. ——Sessional Paper no. 1 of 1986, On Economic Management for Renewed Growth, The Government Printer, Nairobi. —— Policy Framework Paper, 1990–1992,1990. Killick, T. and Mwega, P.M. (1990) ‘Monetary policy in Kenya, 1967–88’, mimeo., ODI, London. Mosley, P. (1990) ‘Kenya case study in ESCOR research project on World Bank and policy based lending in the 1980s’, mimeo. Ssemogerere, G. (1990) ‘Structural adjustment programmes and the coffee sector in Uganda, 1981–1987’, Research Paper no. 1, AERC, Nairobi. Tumusiime-Mutebile, E. (1990) ‘Selected issues in stabilization and adjustment policy in Uganda’, Discussion Paper no. 3, Ministry of Planning and Economic Development, Kampala. Uganda, Ministry of Planning and Economic Development, Background to the Budget, various years. ——Statistical Bulletin No. CP1/1, Consumer Price Index Kampala (to September 1990), October 1990. ——Statistical Bulletin No. GDP/2, Gross Domestic Product of Uganda 1981–1989, July 1990. ——Statistical Bulletin No. IP/6, Index of Industrial Production (to June 1990), October 1990. ——Key Economic Indicators, 3rd issue, October 1990. ——The Way Forward. The Macroeconomic Strategy: 1990–1995, Kampala, 1990. World Bank (1988a) Kenya: Recent Economic Development and Selected Policy Issues, Washington, DC. ——(1988b) Uganda: Towards Stabilization and Economic Recovery, Washington, DC. ——World Tables, 1988/89 edition, Washington, DC.

279

14 INTERNATIONAL TRADE AND ECONOMIC GROWTH The experience of Ethiopia Göte Hansson INTRODUCTION Now, at the outset of the 1990s, Ethiopia is one of the poorest countries in the world with an economy in deep distress and a per capita GNP of around 120 US dollars, one of the lowest in the world and around 35 per cent of the average for low-income economies and sub-Saharan Africa (World Development Report 1991, Table 1). The Ethiopian economy is a typical dual economy, in the sense that it has a dominating and highly traditional agricultural sector and only a minor commercialized industrial and service sector. The agricultural sector accounts for 40–45 per cent of GDP and 85 per cent of total exports, and employs some 90 per cent of the labour force. The core problem in Ethiopia can be stated by referring to Gustav Ranis’ words concerning dual economies in general: The heart of the development problem in the dual economy is thus the ability of the agricultural sector to yield sufficiently large agricultural surpluses and to preserve a sufficiently large part of such surpluses for productive investment in non-agriculture. Simultaneously, the nonagricultural sector, financed by this agricultural surplus plus the reinvestment of industrial profits, must grow fast enough to absorb the labor force being reallocated. (Ranis 1988:82) The surplus from agriculture can go into non-agriculture directly but also through exchange in the international market, where agricultural products are traded against various types of investment goods and other non-agricultural inputs and thereby contribute to economic growth. In international and development economics trade is often considered as an engine of growth that can transform a highly traditional subsistence economy into a modern economy. The main purpose of this study is to analyse the past seventeen years of economic development in Ethiopia and ask what went wrong and how the negative economic trends can be reversed.

280

TRADE AND ECONOMIC GROWTH: ETHIOPIA

Ethiopia is a country that has suffered from civil war for the last thirty years. This has had a very negative impact on economic growth and development. This is not the whole story, however. Even without the war, the political socialist system that developed after the revolution in September 1974, and the economic policy under that regime, would have led to a very problematic economic situation. The civil war has contributed to the magnitude of the economic crisis rather than created the crisis per se. In March 1990 a change of policy was announced and partially introduced. The dramatic changes in the political situation in May 1991, when President Mengistu left the country, meant a major move away from socialism and dictatorship towards a democratic political system and an economic system where markets play a crucial role. This study will only deal briefly with the intended new economic system.1 Instead the focus will be on the incentive system and the trade and exchange rate regimes during the socialist period. It will be shown that the socialist mode of economic organization has undermined the incentives in the agricultural sector. In particular it is the smallscale family farmers, the peasants, who have suffered from the policy. The production of agricultural surplus that could have been exported and thus make up a basis for imports of foreign goods, in particular capital goods for the nonpeasant sectors, has been heavily reduced over time. The Ethiopian trade and exchange rate regimes have worsened the situation. Trade has not been allowed to work as an engine of growth. Because of the domestic incentive system and the trade and exchange rate misalignment, the engine (trade) has not been provided with enough fuel (exports) to promote economic growth. In the next section an analytical model for a dual economy will be presented. This model will be used in the two following sections to analyse how the economic policy in socialist Ethiopia has affected economic growth. The first of these sections will focus on domestic economic policy, whereas the second will analyse trade and exchange rate policy. Then follows an analysis of the economic situation that has arisen from the economic policy in Ethiopia during the past years of socialism. Thereafter, economic aspects of the intended reforms and reform programmes that have been presented and also partially implemented during the last few years will be analysed and, in the penultimate section, some additional suggestions for reform will be developed. Finally, in the last section, some concluding remarks are made. SOCIALIST ETHIOPIA: A DUAL ECONOMY APPROACH As pointed out in the introduction, Ethiopia can be characterized as a dual economy. A modified version of the model presented by Gustav Ranis (1988) can be used to highlight the problems faced by Ethiopia and to identify their origin. Figure 14.1 describes the model. The labour force (= population, or a constant share of the population for simplicity) and employment in peasant agriculture (in

281

AFRICA

Figure 14.1

the following labelled ‘agriculture’) are measured along the horizontal left-hand axis (L, LA), whereas the downward axis (L, LN) measures the same labour force and employment outside agriculture. Initially it is assumed that labour that cannot find employment in non-agriculture will be employed in agriculture. The vertical upward axis measures the per capita agricultural consumption and production, whereas the horizontal right-hand axis measures the per capita non-agricultural production and consumption. The relative price of agricultural and nonagricultural goods is measured in the upper right-hand quadrant and is defined here as the terms of trade between the wage in terms of agricultural goods and the wage in terms of non-agricultural goods. For simplicity it is assumed that there is no wage differential between work in traditional agriculture and work in nonagriculture. The maximum agricultural per capita consumption (i.e. the per capita income measured in agricultural goods, wA) is assumed to be given by the per capita production possibilities in agriculture (described by the F0 curve in Figure 14.1) and is set equal to the maximum amount, WA0, that can be produced per capita if the whole labour force is allocated to the agricultural sector. We assume a nonnegative relationship to exist between the labour input in agriculture and per capita agricultural production. Furthermore, per capita agricultural production is

282

TRADE AND ECONOMIC GROWTH: ETHIOPIA

assumed to depend on the organization of agricultural production, the technology, and the size of the cultivated land area. The relative price between agricultural and non-agricultural goods (i.e. between wA and wN) is assumed to be exogenously given by the domestic government or by the international market. Thus, the domestic relative price line, or domestic terms of trade line, can be interpreted as a budget line for the average citizen. It is assumed that as long as the agricultural sector is not commercialized, i.e. labour is not employed according to its marginal revenue product, the wage level measured in non-agricultural goods is given by the wage level measured in agricultural goods, wA, and the relative price of agricultural and non-agricultural goods. The demand for labour in non-agricultural sectors is given by the marginal revenue product (MRP) curve in the lower right-hand quadrant. To show how the model works, assume that the per capita production possibility in agriculture is given by the F0-curve in Figure 14.1. Furthermore, assume that the typical worker (in agriculture or non-agricultural production) chooses to consume agricultural and non-agricultural goods according to the price-consumption curve, PC, a curve that describes the utility-maximizing consumption mix at various relative prices and a given income. Given the initial relative commodity price and per capita income, the representative worker will choose the consumption mix c0. Initially, the demand for labour in the non-agricultural sector is assumed to be given by MRP0. At the wage wm this means that LN0 workers will be employed in the non-agricultural sector. Consequently, LA0 workers will be employed in agriculture. Given the per capita consumption mix c0, this allocation of labour will produce a per capita agricultural surplus equal to X0. This surplus will be exported. Initially, the domestic relative price between agricultural goods and nonagricultural products is assumed to equal the international relative price. Then the agricultural exports give rise to a per capita import of non-agricultural goods equal to M0 that can be invested in non-agricultural production. The producer surplus generated from employment in non-agricultural production can also be reinvested in this sector. Then, both the value of the marginal productivity of labour and the demand for labour in non-agriculture will increase, i.e. the MRP curve will shift outwards and more workers can be employed in non-agriculture. Thereby, the total production possibilities of the economy and thus the national income and the per capita income will grow, ceteris paribus. Thus, trade based on agricultural surplus production will work as an engine of growth.

283

AFRICA

INSTITUTIONAL CONDITIONS AND DOMESTIC ECONOMIC POLICY IN SOCIALIST ETHIOPIA—IMPLICATIONS FOR AGRICULTURE In the introduction it was argued that the current economic problems in Ethiopia are man-made. The socialist political system has produced a highly inoptimal economic system with inoptimal domestic incentives and policies related to international commercial relations. This section concentrates on the domestic incentive system and overlooks the fact that Ethiopia’s international economic relations are heavily distorted by various trade barriers and an inoptimal exchange rate regime, factors that will be analysed in the next section. Ethiopian agriculture can be characterized as a peasant sector, i.e. the farmers derive their subsistence and cash incomes mainly from raising crops, the capital intensity in production is low, small farms dominate, the main objective of production is consumption and not the creation of a basis for investment and the expansion of activities, the individual peasant grows both cash crops and subsistence crops, individual farms have links to the outside world (the urban areas as well as the international world) through production of agricultural surplus and ‘imports’ of consumer goods and inputs of production but the peasants have no power to affect the conditions of these links nor to affect the rules governing rural life (Lundahl 1979:13f). In Ethiopia the peasants cultivate around 95 per cent of the cropped area and produce 90–94 per cent of the cereals, pulses and oilseeds. This production is mainly a subsistence production; only 20–30 per cent is marketed. The peasant sector also produces the major share, 98 per cent, of the dominating export crop, coffee. Up to 1974 Ethiopia was a feudal society. Less than one year after the revolution, in March 1975, a Land Reform was launched and all private land was nationalized.’ Peasant associations were organized to implement the land reform. These peasant associations covered around 800 hectares each and individual farmers got usufructuary rights to up to 10 hectares per peasant family (Goyder and Goyder 1988:83). Through the land reform all peasants were guaranteed land within their peasant association area. As a result the pressure on land increased. Thus, the size of plots decreased and yields declined (Clapham 1988:164–5). It should be noted also that large scale commercial farms were not distributed to the peasants. Instead these farms were taken over directly by the government and transformed into state farms. In post-revolution Ethiopia agriculture has been transformed not only through the nationalization of land and an emphasis on state farms but also through resettlement, villagization, and collectivization programmes. The aim of the resettlement programmes, which took an organized form in 1976, was to move people from one region or area to another in order to improve the balance between food potential and population. However, the results of the

284

TRADE AND ECONOMIC GROWTH: ETHIOPIA

programme seem to have been of doubtful value. One reason was the very high costs involved in resettling a family and the lack of adequate planning. In the case of the villagization programme, which was initiated during the Ethiopia-Somalia War of 1977–8 and was extended in 1985, 35 to 40 per cent of the rural population were moved into villages. Important official motives behind this programme were that it would be easier to improve social services such as health care and education. Moreover, through the programme it would be easier to increase the standard of living in terms of access to water and sanitary systems. Thereby the programme was assumed not only to contribute to a more decent standard of living but also to improve agricultural production through higher labour productivity. However, other and, from the point of view of the Ethiopian government, even more important motives can be expected to have been behind the villagization programme. At the 9th Plenum of the Workers’ Party of Ethiopia in 1988, the government explicitly classified the villagization programme as a prerequisite to collectivization and the establishment of producer co-operatives (The Ethiopian Herald 1988:November 12). President Mengistu saw the individualistic peasant structure of agriculture as an obstacle to efficiency in Ethiopian agriculture (Harbeson 1988:172f). In addition to collectivization through land reform, collectivization in Ethiopian rural areas has taken two different forms:service co-operatives and producer co-operatives. The main economic function of most service cooperatives has been to supply the peasants with consumer goods and to provide health services, education and grain mills. These objectives have been achieved only to a very small extent. The political objective of agricultural service co-operatives has been no more successful. According to the Ten Year Perspective Plan, issued in 1984, around 50 per cent of the total cultivated area in Ethiopia was to be farmed by producer cooperatives by the end of the plan period. At the outset of the plan only 1.4 per cent of the cultivated area was under the control of producer co-operatives. From embracing around 1.7 per cent of all peasant households in 1983–4, these cooperatives had increased to around 4 per cent in 1986–7. Thus, the peasants were very reluctant to join producer co-operatives and as a result, the transformation of the agricultural sector progressed much more slowly than was planned. In the 1990 economic reform programme, the heavy emphasis on the collectivization of Ethiopian agriculture was not only reduced but it was clearly stated that existing producer co-operatives could be dismantled if it was the will of their members (see also p. 296). As a consequence, in the autumn 1991 there were no producer co-operatives left. Even if progress of collectivization was very slow, the political emphasis on socialist modes of production did have a negative effect on agricultural incentives and performance. As a direct consequence of the collectivization efforts, the individual peasants suffered from economic discrimination in many ways. When a producer co-operative was formed (or when the number of members

285

AFRICA

increased), its members were given priority to the best land and to irrigation facilities. Private peasants were allocated land of lower quality. This of course created great uncertainty for private peasants for whom conservation and other forms of improvements on the individual holdings became a highly risky investment. The uncertainties related to the Ethiopian land tenure system up to March 1990 had a determinate negative influence on the land use of private peasants and thereby on the overall agricultural performance. The land tenure system giving priority to members of producer co-operatives meant that land was not taken care of properly; it was only used and its quality and productivity were reduced over time. In Figure 14.2, the negative impact of the socialist land tenure system is represented by a downward shift of the F curve (the per capita production possibilities in agriculture) to F1. This downward shift affects the economy in various ways. The shift of the F curve means a reduction in the real wage measured in terms of agricultural goods to wA1 and thereby, at the given relative price of agricultural and non-agricultural goods, a fall in the real wage measured in terms of non-agricultural goods to wN1, also. If there are no restrictions placed on the non-agricultural wage (e.g. a minimum wage law), employment increases in non-agriculture and decreases in agriculture. The producer surplus generated from employment in non-agriculture expands, which in turn increases the potential for investment with domestic capital goods in the sector. In agriculture, the downward shift of the F curve to F1 reduces the agricultural surplus per se. When employment in non-agriculture increases, the labour force in agriculture decreases, leading to a further reduction of the agricultural surplus and in the export of agricultural products. The per capita exports of agricultural produce will fall, in Figure 14.2, from X0 to X1. As a consequence, the room for imports of investment goods, i.e. of foreign capital goods, and inputs to the nonagricultural sector will be reduced from M0 to M1 units per capita. The socialist ambitions in Ethiopian agriculture were manifested not only in the allocation of land but also in the allocation of inputs, e.g. fertilizers and capital equipment, where priority was given to state farms and producer co-operatives over individual peasants. The more favourable incentives for socialist modes of production also included such items as interest-free loans, lower land tax and priority allocation of improved livestock and disproportionate extension staff support from the Ministry of Agriculture. One consequence of the policy is that the use of improved seed and fertilizers is still very rare in Ethiopian peasant agriculture. According to estimates in Addis Ababa only around 2 per cent of farmers use improved seed. Another example can be taken from the credit market. Only 4 per cent of the official credit in the mid 1980s went directly to the peasant farmers, whereas the state farms absorbed no less than 40 per cent. The peasants were also subjected to discrimination compared with state farms and producer cooperatives by obtaining lower prices (up to 20 per cent) on grain sold to the Agricultural Marketing Corporation (AMC). Undoubtedly there have been very strong economic incentives for peasant

286

TRADE AND ECONOMIC GROWTH: ETHIOPIA

Figure 14.2

families to join producer co-operatives. However, the incentives notwithstanding, resistance to joining co-operatives has been strong. One reason behind this is the fact that the co-operatives exhibited a poor performance compared with private peasants. The 4 per cent of the peasants who by the end of the 1980s were members of producer co-operatives have been estimated to contribute around 2 per cent of the total annual agricultural output. As a rule, the emphasis on state farms has been a failure also, even if the yield per hectare in these farms is quite high. In 1986–7 the individual peasants produced 10.9 quintals per hectare whereas the corresponding figures for state farms were 23.4 quintals.4 However, by looking at the use of other inputs than land we note that state farms have often been very inefficient production units. Take the production of wheat for instance: in 1983–4 production in state farms resulted in a social loss of 646 birr per hectare, whereas the same production in the peasant sector resulted in a social profit of 159 birr per hectare.5 In terms of the analytical model, the more resources allocated to producer cooperatives and state farms, the lower will be the per capita agricultural production, i.e. the F curve shifts downwards and the effects described in Figure 14.2 will be strengthened, ceteris paribus. Taxation of cash incomes from agricultural sales has also contributed to

287

AFRICA

distorting agricultural production. The compulsory quota delivery system together with land-use tax, income tax and local ad-hoc levies place implicit and explicit taxes at a very high level (estimates point to 80–90 per cent of cash income from grain sales). It is obvious that taxes of this magnitude impose a heavy burden on the peasants and constitute a very severe disincentive for agricultural production for cash sale. In terms of Figures 14.1 and 14.2, this means that even if there is a technological possibility of producing an agricultural surplus, the incentives to produce this surplus are heavily reduced by the tax system. The lack of foreign currency has contributed to the scarcity of consumer goods. It is important that the service co-operatives can satisfy the demand for consumer goods when peasant incomes increase, otherwise full advantage cannot be taken of changes in the agricultural incentives, and peasant production risks being limited to production for subsistence. From the point of view of the national economy, this is a wasteful use of resources. Agricultural surplus production and exports are reduced which limit the import possibilities, making non-agriculture grow more slowly than it would otherwise. This in turn reduces the availability of consumer goods and thereby the incentives for peasant surplus production. To summarize, the Ethiopian domestic incentive system has emphasized collectivization and state farms and discriminated against individual peasants. This has led to a reduction of cash sales production per capita. From the point of view of our analytical model, by undermining the production and export of agricultural products through the incentive system, the government has undermined the possibilities for trade to work as an engine of growth also. Instead the policy has worked towards impoverishment of the country. TRADE AND EXCHANGE RATE POLICY IN SOCIALIST ETHIOPIA International economic relations, mainly in terms of international trade, are of great importance for growth potential. Up to now the model analysis has assumed that the domestic relative price equals the international price. This section will take the Ethiopian trade and exchange rate policy into explicit consideration. In Ethiopia, foreign trade has been completely controlled through licensing and exchange rate allocation. In relation to exchange rate policy, Ethiopia has continued to have an exchange rate regime where the Ethiopian birr has been in a fixed relationship to the dollar at a value of 2.07 birr per US dollar since the beginning of the 1970s. The development of the Ethiopian economy since the revolution has resulted in an obvious need for devaluing the birr (see the next section). The effects of the Ethiopian trade and exchange rate policy can be analysed by means of Figure 14.3. The per capita wage measured in non-agricultural goods under the current trade and exchange rate policy is wN0 and is lower than the per capita wage measured in non-agricultural goods at the international prices wNi.

288

TRADE AND ECONOMIC GROWTH: ETHIOPIA

Figure 14.3

Note that the per capita wage measured in terms of agricultural goods is the same under the two trade and exchange rate regimes, since the per capita income measured in agricultural products has been defined as the maximum per capita production in agriculture when all available labour is allocated to agriculture and the change in policy is assumed not to affect the production possibilities in the agricultural sector. The consumption pattern is given by c0 instead of ci. Thus, the per capita agricultural surplus for a given agricultural labour force will be lower under the current policy compared with what would have been the case if domestic prices were set equal to international prices. The marginal revenue product in non-agricultural production is also affected by the trade and exchange rate policy. With the current policy the MRP curve is located further out from the origin and employment outside agriculture is higher compared with the situation with international prices. As a result, agricultural employment is lower and so is agricultural surplus production, and thereby the per capita exports of agricultural produce also; X0 units will be exported per capita instead of Xi. This means that per capita imports of capital and other inputs are also lower, M0 instead of Mi units. However, the generation of domestic capital through generating a surplus from employment in non-agriculture is greater under the current trade and exchange rate policy than under free trade and proper exchange rates. To summarize, our model analysis suggests that the Ethiopian trade and exchange rate policy has tended to reduce the relative share of foreign produced capital goods in Ethiopian investments. Given the current problems in obtaining

289

AFRICA

foreign exchange for imports of important foreign capital goods and other inputs, this policy can be expected to have generated productivity losses both in agriculture and in non-agriculture. Thus, the trade and exchange rate policies in socialist Ethiopia seem to have tended to reduce the role of trade to work as an engine of growth. ECONOMIC PERFORMANCE IN SOCIALIST ETHIOPIA The analysis in the two preceding sections has shown and explained how the policy in socialist Ethiopia can be expected to have affected GDP per capita, exports and growth. It was found that we should expect a negative effect on all these variables. This section will illustrate the actual economic development in Ethiopia during the post-revolutionary period. The low per capita income in Ethiopia is the result of poor economic growth combined with high population growth. The growth of production has been very low since the revolution, even if the average annual growth during the past decade has been above the average for sub-Saharan Africa. The better growth performance in Ethiopia during the 1980s is due to increases in industrial production and services. Investments in Ethiopia have taken 10–15 per cent of GDP since 1974–5 but domestic saving has decreased from around 8 per cent of GDP in the mid 1970s to 2–5 per cent during the second half of the 1980s. Consequently, Ethiopia has had to borrow for its investments because of the insufficient mobilization of domestic resources. The analysis in the previous sections predicted a negative per capita development in agriculture, and hence also in the per capita exports.. Figure 14.4 shows the development of per capita GDP and per capita agricultural production. As expected from the model analysis in the two preceding sections, the overall economic performance as well as the performance in agriculture have been poor during the period when the various socialistic programmes were emphasized. As a consequence, the availability of food (including cereal imports and exports of pulses) also decreased and at the end of the 1980s the per capita food availability was approximately 20 per cent below the level at the end of the 1970s. The poor development of the agricultural sector is crucial in explaining the current economic situation in Ethiopia. The problematic situation in agriculture is due partly to the civil war and droughts, but, as has been argued in the previous sections, the crisis is to a high degree man-made through the design of the incentive system and the Ethiopian trade and exchange rate policies. The poor economic performance in agriculture, combined with high population growth, particularly in urban areas, has made the food situation very problematic. The rapidly falling per capita agricultural production since the 1960s has given rise to increased imports of food to the country even during non-drought years. The poor economic performance in Ethiopia has resulted in severe balance problems both in public finance and in external trade. Even though tax revenues have increased their share of GDP since the mid 1970s (from 13 to 30 per cent),

290

TRADE AND ECONOMIC GROWTH: ETHIOPIA

Figure 14.4 Development of GDP and agricultural production (per capita figures, 1980 Ethiopian birr) Source: World Bank (1991:240f)

the growth of public spending has been even greater, owing to rapidly increasing defence expenditures and payments of interest and charges on public debt.7 Figure 14.5 shows that the budget deficit in Ethiopia has increased from around 6 per cent of GDP in 1974–5 to around 17 per cent in 1989–90. The model analysis in the two preceding sections predicted a negative development in Ethiopian exports. In reality, Ethiopia has experienced increasing trade deficits ever since the mid 1970s. At the beginning of the 1990s the deficit amounted to around 10 per cent of GDP. To explain this development by reference to the oil crises of the 1970s is not convincing considering the length of the period and the composition of Ethiopian imports (see Figure 14.6). Furthermore, as is clear from Figure 14.7, the trade deficit cannot be explained by referring to the development of terms of trade. The major problem in Ethiopia is that while the export volume (i.e. export measured at constant prices) increased by around 40 per cent during the decade after the revolution, the import volume increased by around 90 per cent. Ethiopian export per capita has decreased to one of the lowest levels in the world. In 1989 Ethiopia had one of the lowest official per capita export values among the countries reported in the World Bank’s World Development Report 1991: around 20 per cent of the average for low income countries and 15 per cent of the average for sub-Saharan Africa. One explanation for this situation, advanced in the model analysis in the previous section, is the domestic incentives in agriculture and the Ethiopian trade and exchange rate policy. As shown in the section on institutional conditions and domestic economic policy, the agricultural incentives in socialist Ethiopia have undermined the basis for economic growth based on a high and balanced per capita foreign trade.

291

AFRICA

Figure 14.5 Trade and budget deficits (per cent of GDP) Source: National Accounts Division, NCCP

Figure 14.6 Import composition Sources: Customs Office, National Bank of Ethiopia

Figure 14.8 shows that Ethiopian exports are heavily concentrated. The dominating export commodity is coffee which has accounted for around 60 per cent of all export revenues ever since the mid 1970s, followed by hides and skins for which the share of total export value has varied between 8 and 16 per cent. By the end of the 1980s these two sets of commodities accounted for more than 75 per cent of total export revenues. On the export side pulses have decreased in importance: from 10–15 per cent in the mid 1970s to around 2 per cent by the end of the 1980s.

292

TRADE AND ECONOMIC GROWTH: ETHIOPIA

Figure 14.7 Net barter terms of trade (1980=100) Source: World Bank (1990:Annex Table 3.13)

Figure 14.8 Export composition Sources: Customs Office, National Bank of Ethiopia

Figure 14.9 shows the development of the per capita export volume of crops, hides and skins, and live animals. It can be seen that there was a dramatic fall for the major export products during the 1970s. This fall has not been reversed during the 1980s. In Ethiopian imports, the shares of raw materials and semi-finished goods have decreased, whereas the shares of capital and consumer goods have increased. Among the capital goods imported, transport equipment and industrial goods dominate. To a large extent the imports have been inputs to the civil war, e.g. fuel, trucks, weapons and ammunition, and not, for example, capital goods for investment in production. It should also be noted that imports of food and live animals make up a significant share of total imports. From 1976 up to 1983 food and live animal

293

AFRICA

Figure 14.9 (a) Per capita crop export and (b) per capita exports of hides and skins and live animals (volumes, 1978–9=100 for 1974–5 to 1978–9; 1979–80=100 for 1979–80 to 1987–8) Source: Customs Office

imports varied between 4 and 8 per cent of the total import value. As a consequence of the droughts, in 1983 and 1985 these imports amounted to no less than 18 and 25 per cent of total imports, respectively. In per capita terms, food imports have increased to compensate for the falling per capita domestic agricultural production. As a consequence the room for imports of capital goods, inputs and consumer goods has not expanded as would have been possible with a more proper incentives system. Thus, the possibilities for trade to work as an engine of growth have been negatively affected also.

294

TRADE AND ECONOMIC GROWTH: ETHIOPIA

Figure 14.10 Debt service ratio (per cent of exports) and international reserves (months of import cover) Sources: Debt service ratios: (a) Ministry of Finance (1991), Ethiopia’s Economic Policy During the Transitional Period, p. 8; (b) World Bank (1987:5,1989:Table 21,1990:4), IMF (1988:iv). International reserves: National Bank of Ethiopia

The increasing deficit in public finance, the low savings rate (in particular compared with the investment rate in the economy), together with the rising trade deficit have resulted in a more or less completely depleted stock of international reserves and an increasingly problematic external debt situation. The debt service ratio (debt expenditure, i.e. interest and amortization, as a share of export revenues) has increased from around 6 per cent to around 76 per cent during the past decade (see Figure 14.10). Thus, in present Ethiopia a major role for export is to service public debt, rather than to provide for economic growth through imports of investment goods and inputs for production. THE ECONOMICS OF THE ANNOUNCED ECONOMIC REFORM PROGRAMMES The previous sections have described and analysed the economic development in Ethiopia during the period 1974–90. Because of the dual nature of the Ethiopian economy, the most urgent economic problem is to reverse the negative trend in agriculture. This problem can be subdivided into how to increase the area under cultivation from the present 10 per cent of cultivable area under crop production, and how to improve the performance in agricultural production. In the third section it was argued that the agricultural sector in socialist Ethiopia has been characterized by bad management in terms of taxation, price incentives and institutional organization. As should be clear both from the model analysis and the description of the economic performance in socialist Ethiopia, these factors have negatively affected both the performance of the agricultural sector and the willingness to cultivate increased land areas or to improve technology in

295

AFRICA

agricultural production. Thus, a more or less complete economic reform programme must be introduced to break the downward economic trends. During the 9th Regular Plenum of the Central Committee of the Working People of Ethiopia in November 1988, a number of resolutions related to the economic system of Ethiopia were adopted (see e.g. Hansson 1989:26ff). These resolutions indicated signs of changes towards a more market-oriented economic system. After the close of the Plenum very little was done in terms of implementation and one and a half years after the adoption of the resolutions the major impression was that they had not been followed by appropriate changes in policy. However, the economic system was an important topic again at the 11th Regular Plenum of the Central Committee of the Workers Party of Ethiopia in March 1990. On 5 March, President Mengistu Haile Mariam announced the introduction of a completely new Ethiopian economic system. The new system was to be based on the principles of what is known as a ‘mixed economy’.8 According to the adopted six-point resolution, economic actors and the market will be guiding the allocation of resources in the new economic system. The main points in the new economic strategy can be summarized as follows: • • • •

• • • • • •

The Ethiopian economic strategy will be that of a mixed economy. The state enterprises should be managed on the basis of competition, profitability and productivity. The private sector will from now on be encouraged and strengthened in all ways. The land will still be owned by the state but the individual peasant’s right to use a specific area of land is indefinite in time and can be transferred to his legal heirs who also derive their livelihood from farming. Trees and other perennial plants grown on the land will be owned by the respective peasants. Individual peasants may hire workers to work on their farms. Producer co-operatives may be dismantled if this is the democratic will of the members. Private investors may establish modern large farms. Private traders are allowed to compete without any restrictions with the staterun trade enterprises in all sectors of the economy. The grain control stations and the quota system will cease to exist.

After the fall of the Mengistu government the Transitional government adopted a transitional economic policy programme that means a continuation and deepening of the reform process indicated and also partially introduced by President Mengistu in March 1990.9 If completed, the changes in the Ethiopian economic system will have far-reaching effects on the performance of the Ethiopian economy. In terms of the dual model, several of the points of the reform in the above list will shift the per capita agricultural production curve in Figure 14.2 upwards and

296

TRADE AND ECONOMIC GROWTH: ETHIOPIA

thus create increased possibilities for increased imports of capital goods and other inputs, i.e. fuel and fertilizers, and consumer goods. Thereby, the per capita agricultural production can be expected to expand even more and this can also be expected to be the case for the modern sector. Thus, the completion of the Transitional Economic Policy will be a very important task for policy makers in Ethiopia in the 1990s. POLICY RECOMMENDATIONS As can be concluded from the analysis above, trade liberalization and a devaluation can be expected to give rise to important positive effects and increase the possibilities for trade to work as an engine of growth in Ethiopia. However, in the announced economic reform programmes there are few signs that the government will introduce a liberal trade regime, even if there are some changes indicated in the Transitional government’s economic policy programme. One frequently advanced reason for the reluctance to liberalize trade is the lack of foreign currency. This is, however, a short-sighted way of reasoning. In reality the current trade and exchange rate policy has placed the country in a situation where lack of foreign currency undermines foreign trade which in turn worsens the situation in terms of foreign reserves (see Figures 14.5 and 14.10). A trade policy reform is urgently needed, otherwise there is an obvious risk that the current trade procedures, in particular the licensing procedures and the rationing of foreign exchange, will continue to be a severe obstacle to changes in the domestic incentive system towards a more market-oriented system. Thus there is also a risk that central planning will be kept at the centre of the allocation of resources. For a small country’s performance the exchange rate regime is as important as the trade policy regime. In the case of Ethiopia, since the early 1970s the domestic currency, the birr, has been in a completely fixed relationship with the US dollar. From the descriptions of the development of the trade deficit, international reserves and the debt service ratio it is obvious that the Ethiopian birr is overvalued. Thus, a devaluation is needed. In a dual economy framework, the effects of trade liberalization and an exchange rate reform in terms of a devaluation can be analysed by means of Figure 14.3, where the effects will go in the opposite direction compared with the case analysed above, i.e. the starting point is c0. If trade is liberalized and the Ethiopian birr is devalued the relative prices in Ethiopia will change. The per capita wage measured in non-agricultural goods will increase, ceteris paribus. The new consumption pattern will be given by ci. Thus, the agricultural surplus will increase at a given agricultural labour force. However, the change in international prices affects the MRP curve in non-agricultural production also; thus the MRP curve will shift inwards and employment outside agriculture will decrease. As a result agricultural employment will increase and so will agricultural surplus production and thereby also the per capita export of agricultural produce, from X0 to Xi. This means that per capita imports of capital and other

297

AFRICA

Figure 14.11 (a) Ethiopian exports: country composition (per cent). (b) Ethiopian imports: country composition (per cent) Source: IMF, Direction of Trade Statistics Yearbook, various issues

inputs can increase from M0 to Mi. Furthermore, the surplus from employment in non-agriculture will decrease. Thus, the reform makes it possible to increase the relative share of foreign-produced capital goods in Ethiopian investments. This in turn can be expected to give rise to productivity gains both in agriculture and in non-agriculture and also improve the role of trade as an engine of growth. In addition to a devaluation, a change in exchange rate management, reducing the central role given to the US dollar in the Ethiopian exchange rate policy, is also needed (see Hansson 1989, 1990, 1992). The fixed relationship

298

TRADE AND ECONOMIC GROWTH: ETHIOPIA

Figure 14.12 Development of real effective exchange rate and trade deficit (index, 1980=100) Sources: World Bank (1990:Annex Table 3.12), National Accounts Division, NCCP

between the Ethiopian birr and the US dollar is probably due to the pre-1973 international monetary system, where the US dollar was the principle reserve currency whose value was defined and guaranteed in gold. However, since 1973, when the Bretton Woods monetary system with fixed exchange rates came to an end, there are no a priori motives for a country to relate its currency completely to the US dollar. Figure 14.11 shows that in the case of Ethiopia, only a minor share of the trade is with the USA, whereas the trade with the EEC countries is much greater. Exports to the USA in the period from 1975 to 1990 have varied between 8 and 33 per cent (on average 19 per cent) of the total Ethiopian exports. The corresponding variation of exports to the EEC countries is between 24 and 49 per cent (on average 37 per cent). On the import side the corresponding figures for imports from the USA are 4 to 19 per cent (on average 11 per cent) whereas the share of imports from the EEC countries has varied between 32 and 46 per cent (on average 37 per cent) during the period considered. It should be noted that even if at first sight Ethiopia has a fixed exchange rate system, this is definitely not the case in the commercial relations with her major trading partners, with the exception of the USA. In fact, the case is rather the opposite: Ethiopia has a system with very flexible exchange rates towards all her major trading partners but the USA. The relationship between the birr and the European currencies, for example, is determined by the relationship between the US dollar and these respective currencies. Therefore, if, for instance, the US dollar appreciates due to an improvement in American economic performance or speculative movements, the Ethiopian birr will appreciate as well. This will take place independently of the economic development in Ethiopia. In Figure 14.12 the indices for the real effective exchange rate and the deficit in Ethiopian foreign trade are presented. From this figure we conclude that since the revolution in 1974

299

AFRICA

up to 1985 the real effective exchange rate for Ethiopia increased at the same time as the trade deficit increased. This appreciation of the birr, relative to the currencies of Ethiopian trading partners, is a direct consequence of the 100 per cent link between the birr and the US dollar. By the same reasoning, the real effective exchange rate decreased during 1987 and 1988 (but not enough to compensate for the development up to 1985) when the US dollar depreciated. Even if the price of traded goods, like coffee, is set in dollars on the international market there is no reason to have such a strong link to the dollar. The reason is that the dollar is convertible with other currencies, for instance the European currencies. Therefore, the currency basket to which the Ethiopian birr should be linked should be decided on the basis of the country composition of Ethiopian trade and the composition of competitors to Ethiopia in her major export markets. A preliminary analysis points to a very heavy weight to the European currencies or the ECU (Hansson 1992). CONCLUDING REMARKS When the socialist period was ended by the fall of President Mengistu Haile Mariam in May 1991 the Transitional government soon declared a change in the economic system away from socialism. It is not unreasonable to believe that the new leaders seek economic support from the international donor society in their efforts to steer the economy on to a more promising track. To achieve such support it is likewise reasonable to expect conditionality to be introduced in terms of demands for completion of an economic reform programme following the major lines of the reform programmes that have been announced since 1988. However, the programmes must be accompanied by reforms of the trade and exchange rate policies if it is to prove really possible to talk about a true economic reform towards a mixed or market economic system. Changing the economic system also requires a change in the institutions in the country. Property rights must be defined and legal institutions must be created to protect these property rights. Furthermore, knowledge must be accumulated so that the economic actors and the new institutions can operate in the new economic environment that the economic reforms bring about. To assist Ethiopia in reforming or creating proper institutions and accumulating knowledge are perhaps the most urgent tasks for foreign donors working with Ethiopia. If the reforms are successfully completed and proper institutions are created, Ethiopia will have a clear possibility to develop in accordance with her comparative advantages. Then, the dominating peasant sector will improve its performance and produce not only for the Ethiopian population but also substantial cash crops that can be exported and used as fuel to enable trade to work as an engine of growth.

300

TRADE AND ECONOMIC GROWTH: ETHIOPIA

ACKNOWLEDGEMENTS The research for this study has been financed by SIDA and the Institute for Economic Research at Lund University and is gratefully acknowledged. Thanks are also due to Yves Bourdet, Hans Falck and Johan Torstensson for valuable comments on a preliminary version of this essay. NOTES 1 2 3 4 5 6 7 8 9

For an analysis of the Transitional government’s intended economic policy, see e.g. Hansson (1993). This section is based on Hansson (1989,1990) and references given therein. See e.g. Dessalegn (1984) for a description of the land reform. Calculated from IMF (1988:table IX). World Bank (1987a: Table 7.12). This section is based on Hansson (1992, 1993) and references given therein. According to a quotation from a speech by President Mengistu in The Ethiopian Herald (1988:4), defence expenditure has increased annually by on average around 20 per cent in nominal terms since the revolution in September 1974. The following description of the reform of March 1990 is based on Mengistu (1990). For a description and critical analysis of the Transitional government’s economic policy programme, see Hansson (1992, 1993).

REFERENCES Clapham, C. (1988) Transformation and Continuity in Revolutionary Ethiopia, African Studies Series no. 61, Cambridge: Cambridge University Press. Dessalegn, R. (1984) Agrarian Reform in Ethiopia, Uppsala: Scandinavian Institute of African Studies. Ethiopia’s Economic Policy During the Transitional Period (1991), Addis Ababa. Goyder, H. and Goyder, C. (1988) ‘Case studies of famine: Ethiopia’, in D.Curtis, M. Hubbar and A.Shepherd (eds) Preventing Famine, Policies and Prospects for Africa, London: Routledge. Hansson, G. (1989) Ethiopia: Macroeconomic Performance, Economic Policy and Swedish Aid, Macroeconomic Studies no. 1/1989, Stockholm: SIDA. ——(1990) Ethiopia 1990, A Change of Economic System? A critical assessment of the economic reform programme of March 5th, 1990, Macroeconomic Studies no. 9/1990, Stockholm: SIDA. ——(1992) ‘Ethiopia towards the year 2000—Prospects of economic development’, paper presented at the 7th International Conference on the Horn of Africa, ‘Changing Realities in the Horn of Africa: The Challenge of Economic Reconstruction and Political Pluralism’, Center for the Study of the Horn of Africa, New York, 23–24 May 1992. ——(1993) ‘Ethiopia away from socialism’, in M.Blomstrom and M.Lundahl (eds) Economic Crisis in Africa. Perspectives on policy responses, London: Routledge. Harbeson, J.W. (1988) The Ethiopian Transformation: The Quest for the Post-imperial State, Boulder, Co: Westview Press. IMF (1988) Ethiopia—Recent Economic Developments, SM/88/239. IMF Direction of Trade Statistics Yearbook, 1982, 1988, 1989. Lundahl, M. (1979) Peasants and Poverty: A Study of Haiti, London: Croom Helm. Mengistu, H.M. (1990) Report by Mengistu Haile Mariam, Resolutions Adopted by the Plenum, Documents of the 11th Plenum of the CC WPE, Addis Ababa, March 1990.

301

AFRICA

Ranis, G. (1988) ‘Analytics of development: Dualism’, in H.Chenery, and T.N. Srinivasan (eds) Handbook of Development Economics, Volume 1, Amsterdam: North Holland. The Ethiopian Herald (1988), 8–12 November 1988,17 June 1989, 8 March 1990, Addis Ababa. World Bank (1987) Ethiopia: Recent Economic Developments and Prospects for Recovery and Growth, Report no. 5929-ET. ——(1989) Sub-Saharan Africa: From Crisis to Sustainable Growth, Washington, DC: The World Bank. ——(1990) Ethiopia’s Economy in the 1980s and Framework for Accelerated Growth, Report no. 8062-ET. ——(1991) World Tables 1991, Baltimore, MD: The Johns Hopkins University Press.

302

15 REGIONAL ECONOMIC INTEGRATION AND INTRA-AFRICAN TRADE Issues for development Daniel B.Ndlela

INTRODUCTION Three decades have passed since the independence movement of African countries first took root. The economic ills of the 1960s still persist: a large number of fragmented mini-states, lack of an industrial base, strong trade links with developed market economies characterized by African exports of primary commodities and imports of industrial manufactures. These problems have not abated; on the contrary new ones have been added: the debt service problem, agricultural food shortages, and the worsening of Africa’s terms of trade over a long period. Between 1970 and 1988 Africa’s share in world trade remained insignificant and dropped from about 3.7 per cent to about 2.3 per cent (see Table 15.1). From 1970 to 1985 African exports to the developed market economies rose from $9,757 million to $48,023 million averaging 80.8 per cent (see Table 15.2). In 1985 African imports from the developed market economies were $45,701 million or 72.3 per cent of its total imports of $63,177 million.1 Meanwhile, in percentage terms, shares of intra-African trade (exports and imports) declined from 6.11 per cent in 1970 to 3.67 per cent in 1980, rising slightly to 4.87 per cent in 1985 (see Table 15.3). Thus Africa’s regional trade has remained below 5 per cent of total trade, as it was twenty years ago. It has instead become more dependent on foreign trade, foreign aid and even foreign food. Concern with Africa’s economic dependence on industrialized countries in general and the slow or even declining trend in intra-African trade links has led to African policy makers reasserting their determination to expand regional economic integration and intra-African trade. The Lagos Plan of Action (LPA) adopted by African Heads of Government in 1980 inter alia called for the development of intra-African co-operation and integration in order to attain collective self-reliance.2 More recently the International Conference on Africa: The Challenge for Economic Recovery and Accelerated Development held in June 1987, under the

303

c

b

Developing Africa is all African countries excluding South Africa. Current values. Data is not strictly comparable with data on other tables due to adjustments. Sources: UN (1987) Monthly Bulletin of Statistics May UN (1986) 1984 International Trade Statistics Yearbook, vol. I UN (1980) Monthly Bulletin of Statistics December UN (1990) Monthly Bulletin of Statistics May

a

Table 15.1 World trade and share of developing Africaa in world trade (values in millions of US dollars)b

Sources: UN (1987) Monthly Bulletin of Statistics May UN (1986) Monthly Bulletin of Statistics May 1983 Yearbook of International Trade Statistics, vol. 1, 1985 UN (1981) 1980 Yearbook of International Trade Statistics, vol. 1

Table 15.2 Developing Africa’s direction of trade (exports; values in millions of US dollars; percentages are shown in parentheses)

b

Excludes Zimbabwe. The intra- African import values were estimated from the export values by adding 10 per cent of the total value of exports to reflect the cost of insurance and transport. Sources: UN (1987) Monthly Bulletin of Statistics May UN (1986) Monthly Bulletin of Statistics Yearbook UN (1986) 1984 International Trade Statistics Yearbook, vol. I UN (1985) 1983 International Trade Statistics Yearbook, vol. I UN (1984) 1982 International Trade Statistics Yearbook, vol. I UN (1981) 1980 International Trade Statistics Yearbook, vol. I

a

Table 15.3 Share of intra- African trade in total trade of developing African countries (values in millions of US dollars)

ISSUES FOR DEVELOPMENT

auspices of the United Nations Economic Commission for Africa (UNECA), emphasized the need to move out of the present structure of an export-oriented primary commodity system (UNECA—Abuja Statement, 1987). Stress was put on increasing the role of domestic and intra-African production relationships until they become the dominant feature of African development. Resort to endogenous growth factors by African policy makers came about as a result of their disappointment with the fact that external shocks had not produced just adequate responses, they had not produced responses at all. African countries thus established regional institutions to encourage economic integration, inter alia, through trade liberalization and facilitation. Notwithstanding setbacks and a disappointing record, especially that of the low level of intra-African trade, political commitment and hope in regional integration among African leaders remains firm. A major reason for this continued hope in regional integration as a developmental mechanism is the difficulty, for so many African countries, in seeing much in the way of an alternative (Hazlewood 1987:1). Africa has most of the world’s mini-states: nine countries with a population of less than 1 million and thirty-five with a population of less than 10 million. Only three sub-Saharan African countries, Ethiopia, Nigeria and Zaire, have a population of more than 30 million (UN Population Statistics 1989). Domestic production for such small markets will be at extremely high cost given the factor of economies of scale. Also, given the fast changing international environment, at least in terms of offering less and less real and ‘imagined’ opportunities, Africa must tackle its obstacles to regional integration: tariff and non-tariff barriers, lack of transportation and communication, inadequate financing arrangements and a distorted industrial structure and distribution system. There is already a significant spread of unrecorded transborder trade sprawling a large number of African borders. Instead of recognizing and harnessing this trend as giving some market signals that could guide production, African states have responded by either ignoring it or by harassing the traders as smugglers or perpetrators of ‘bad trade’. We return to this discussion below. REGIONAL ECONOMIC INTEGRATION IN SUB-SAHARAN AFRICA Since the 1960s African governments have established a variety of regional economic integration and co-operation arrangements. Most of these schemes are based on pre-independence links and institutions. Their membership was usually confined to countries previously administered by the same metropolitan power. In eastern and southern Africa the Federation of Rhodesia and Nyasaland consisting of the then Southern Rhodesia (Zimbabwe), Northern Rhodesia (Zambia) and Nyasaland (Malawi) was formed under the auspices of the British colonial administration in 1953, and the East African Community was formed by Kenya, Uganda and Tanzania in 1967. The post-independence-established subregional economic integration and co-operation groupings include: UDEAC (1964) and

307

AFRICA

ECCAS (1983) in Central Africa; CEAO (1973), MRU (1973), and ECOWAS (1975) in West Africa; SADCC (1980) and PTA (1982) and the IOC (1982) in Eastern and Southern Africa; and UMA (1989) in North Africa. The economic community model The most complete and comprehensive example of economic integration and cooperation between African countries is the economic community model whose designs and intentions are embraced in the various customs unions and preferential trade areas. The East African Community, until its collapse in 1977, was perhaps the most developed of the economic community arrangements. In 1980 the Lagos Plan of Action (LPA) called for the establishment of five Regional Economic Communities (RECs) in Africa: North Africa, West Africa, Central Africa, eastern Africa and southern Africa. The major RECs established according to this criterion are ECOWAS in West Africa, ECCAS in Central Africa, UMA in North Africa, PTA and SADCC in eastern and southern Africa. Other Economic Communities and African intergovernmental organizations dealing with regional economic integration are intended to be turned into specialized agencies of the main RECs in each of the five regions. The main objectives of the RECs are broad in both coverage of areas and scope. They promote social and economic development through close cooperation among member states in all fields of social and economic activity such as transport and communications, energy, agriculture, natural resources, trade, customs, industry, money and finance, human resources development, tourism, education, culture, science and technology. They are also meant to ensure the movement of persons within each respective region. Their other objectives include organization of trade fairs, collection, analysis and dissemination of trade, production and technological information among member states. They also aim to assist in the establishment of the Pan-African Economic Community (PAEC). With the signing of the treaty to establish PAEC in June 1991, one of the objectives of the latter is to strengthen the RECs. The activities of the RECs will be rationalized, co-ordinated and harmonized so that they constitute strong pillars for the PAEC (Organization of African Unity 1991). Customs unions Customs unions in Africa, though falling under the broad definition of the economic community, are among the oldest institutions which were established mainly to promote intra-African trade. A customs union is an agreement between two or more member states to facilitate trade between them through the removal of tariffs levied on imports from other members of the union (see Viner 1950:41, Meade 1956). UDEAC and SACU3 are typical customs unions. The UDEAC Treaty aims at developing the economic integration of member states through the gradual establishment of a common market and a common external tariff. The treaty also

308

ISSUES FOR DEVELOPMENT

provides for the removal or reduction of tariffs and non-tariff barriers to interstate trade and harmonization of customs legislation, fiscal policies, industrial and planning policies (see Lipsey 1956–7:11–33). Like an economic community UDEAC also has the objective of developing co-operation in transport and communications, tourism and rural development, investment laws and regulations and science and technology. Since member states of UDEAC are also members of ECCAS, there is a case of duplication and overlapping of activities between the two institutions. SACU, on the other hand, has a somewhat limited objective in the area of abolition of tariffs and non-tariffs obstacles and ensuring the free movement of goods, services and persons among the member states of the Union. Preferential trade area In its proper definition a preferential trade area (PTA) is an agreement by which a group of countries agree to provide preferential treatment to selected goods originating from member states by according a gradual reduction of tariffs over a given period of time until a zero tariff level is attained.4 In the African experience, the PTA tends to embrace all major economic sectors approximating an economic community. Thus, for example, the PTA for eastern and southern Africa whose membership comprises eighteen countries5 covers a wide range of activities envisaged by its eleven protocols.6 The overall objective of the PTA is to promote co-operation and development in all fields of economic activity, particularly in the fields of trade, customs, industry, transport, communications, agriculture, natural resources and monetary affairs with the aim of raising the standard of living of its people, of fostering closer relations among its member states, and contribute to the problems and development of the African continent. (Treaty for the PTA: Article 3)

PRODUCTION APPROACH AS A BASIS FOR REGIONAL ECONOMIC INTEGRATION AND INTRA-AFRICAN TRADE The early African integration schemes were focused more on trade facilitation measures almost to the exclusion of other mechanisms, especially the ‘production approach’ to economic integration and co-operation. These schemes did not in principle attempt to go beyond a passive or negative integration scheme for the liberalization of trade (Hazlewood 1987:5). It was often forgotten that access to a larger regional market should be used as a means to break these countries’ tie-in with the world economy as producers of primary commodities and importers of manufactures. The focus on trade to the neglect of production issues is based on the

309

AFRICA

assumption that the benefits of integration normally accrue from a given industrial infrastructure which will be further enhanced by the enlargement of the market. The industrial enterprises will respond in such a way that those industries where technological factors make large-scale production necessary will become efficient. On the other hand, those industries which cater for limited markets and are frequently obliged to manufacture a variety of items in relatively small quantities will more efficiently supply the enlarged market using small runs. Such an argument considers only the effect of integration from the efficiency of allocation of given resources. There is little or no direct involvement of the integration process in creating productive activities. Integration theory was developed as an aspect of international trade theory and production is an implicit component of the argument: trade theory demonstrated how trade as against self-sufficiency, could result in more productive allocation of resources, resulting in an increase in world production and all trading partners being better off. Integration theory demonstrated the circumstances in which partial free trade among a group of countries—free trade between some trading partners with barriers maintained against the rest—brought an improvement over general protection. This was the purpose of the ‘trade-creating’ and ‘tradediverting’ effects of a customs union (Balassa 1961,1967:1–21). Economic integration benefits are thus assessed on the basis of the degree to which ‘trade creation’ outweighs ‘trade diversion’. Trade creation refers to a situation where as a result of economic integration a country moves from an expensive source outside the economic grouping to a less expensive source within the economic grouping. This will have the effect of increasing trade within members of the economic grouping. Thus where trade creation is predominant, one of the members of the integration group at least must benefit, all of them may benefit, and the world at large benefits, though the outside world loses at least in the short run and can gain in the long run only as a result of the positive externalities of the increased prosperity of the customs union area (Viner 1972:35). Trade diversion refers to movements from a cheap source outside the economic grouping to a more expensive source within the integrating countries. Where trade diversion is predominant, at least one of the members is bound to be injured, all combined will suffer a net injury, and there will be injury to the outside world and to the world at large (Viner 1972:35). With the objective of expanding trade and linking the subject of trade with regional integration, the central problem of African countries is the need to transform their economic structures. It is understandable that because of the predominance of external trade in the monetary economies of African countries, their first priority was to break the mould of external trade links, i.e. to escape from the colonial pattern of exports and imports. Intra-African trade was seen as one way to achieve this change. This explains why the emphasis on trade and particularly on trade in manufactures has become the dominant characteristic of the African integration schemes, especially following their adoption of the Lagos Plan of Action.

310

ISSUES FOR DEVELOPMENT

A basic requirement was to change the existing structure of production in each African country, a structure that has from colonial times provided primary exports to the developed countries. Primary exports were the major source of the foreign exchange needed to finance imports of intermediate and capital goods. It was, therefore, logical that the development of new forms of trade should require the setting up of a new pattern of the dominant structure of production. This is because changes in the structure of production and in the pattern of trade are interrelated and reinforcing features of economic development. The intra-African trade would provide a market for African manufactures and the manufactures would provide the products for intra-African trade. Thus, the establishment of economic integration and co-operation schemes (customs unions, common markets, etc.) should be seen as a mechanism to bring about this orientation of trade and production. It was therefore an aberration rather than a norm that, instead of emphasizing a production approach that would close the historically ever widening gap of the divergence between supply and demand of the African economies, a bias towards a pure common market or customs union approach became the more accepted regional integration mechanism. From the supply side, the growth and diversification of intra-African trade is impeded by well known obstacles. These include scarcity of foreign exchange and the inadequacy of the domestic output of the various manufactures which are indispensable for African countries’ investment and inputs requirements. Manufactures, which include virtually all capital goods and a large proportion of intermediate goods, have to be imported. The demand for these goods tends to increase rapidly during periods of economic expansion and, given the typically slow growth of exports receipts from primary products, the external balances of these countries deteriorate. This is further aggravated by growing shortages of all sorts of wage goods including food products which have to be imported. On the demand side, the narrowness of the markets of many small countries constitutes an insurmountable obstacle to growth, and especially industrial development. For example, in a number of industries, notably those producing investment goods, certain industrial semi-manufactures, and durable consumer goods, existing production techniques make large-scale production necessary for economic operation. Unless these countries form themselves into an economic integration grouping and enable themselves to specialize to a certain degree, as individual countries they run the risk of producing too wide a range of manufactures at high cost which is tantamount to production costs under protectionism.

311

AFRICA

INTRA-AFRICAN TRADE: FEATURES, PROBLEMS AND CONSTRAINTS Factors influencing the reordering of priorities in intra-African trade The long-term growth forecasts of industrial countries indicate a continued slowdown of 2.5 per cent per annum over the next ten to fifteen years. Compared with the 5 per cent per annum growth rates of the 1960s and 1970s, this slowdown will greatly constrain Africa’s development efforts. Even if one were to assume higher growth rates in developed countries and higher world trade, the pattern of commodity use in the developed countries and the protectionist policies of these countries would continue to work against the expanding markets for the exports of the African countries to the developed countries. In the short term, Africa’s exports are made more bleak by growing protectionist measures in the industrialized countries. They are also threatened by the technological displacement of some commodities often as a result of product substitution. For example, mutual recognition of products within the European Community will allow British standards for the manufacture of chocolate which permit the partial substitution of cocoa by vegetable fat. This will affect many African countries within the Africa-Caribbean-Pacific (ACP) group in which cocoa is a strategic product (Ivory Coast, Ghana, Cameroon, Nigeria, Togo, Sierra Leone, Equatorial Guinea, Liberia and Zaire). The introduction of this process will cause the price of cocoa to fall dramatically. The African countries producing karite or palm oil, which are substituted for cocoa butter, are unlikely to benefit as these products are already produced in increasing quantities by South Asian countries at lower prices. On the contrary, it is probable that exports of cocoa will experience a declining trend and producer prices will slump (ACP– EEC 1989). Changes in the global macroeconomic framework have important implications for Africa. The decline in trade associated with Africa’s shrinking share of commodity markets and financial flows mean that new sources of growth will have to be found in intra-African regional markets. There is little or no choice for Africa because the structure of relative prices the world faced for most of the postwar period has drastically changed. The sustained world output was helped by an expanding network of trade and finance, cheap and declining energy prices and low costs of food and raw materials. Real interest rates were stable at about 2–3 per cent per annum and exchange rates were fixed. Since the mid 1970s Africa failed to adjust to increases in the relative prices of her importables: food, raw materials, capital and borrowed money. Features and characteristics of African trade Africa’s share of world trade has been declining in recent years. During the period 1960–80 its share of total world exports and of world imports averaged 4 per cent.

312

ISSUES FOR DEVELOPMENT

By 1980–5, this share had dropped to 3.6 per cent (Mulat 1990:155). From 1970 to 1988 Africa’s total trade with the world averaged around 3.5 per cent. Its share in world trade remained insignificant and in fact dropped from about 3.7 per cent to about 2.3 per cent.7 In percentage terms the share of developing Africa’s exports in total world exports similarly showed a decline from 4.06 per cent in 1970 to 1.94 per cent in 1988. The fall in Africa’s total trade in percentage terms during this period indicates that world trade grew faster than Africa’s trade, thus seriously curtailing her development efforts. Africa’s exports are concentrated into primary commodities. In twenty-three non-oil-exporting African countries 50 per cent or more of total exports comprise food, beverages, tobacco and agricultural raw materials (Mulat 1990:156). Intra-African imports exhibited similar trends, declining from 6.68 per cent in 1970 to 3.85 per cent in 1981 and thereafter rising slowly to 4.90 per cent in 1985 (cf. Table 15.3). The main African imports are manufactured goods, machinery and transport equipment (SITC 6–8) which account for about 60 per cent of Africa’s total import bill (Mulat 1990:156). The share of intra-African trade in total trade of African countries also remained low over the last two decades. In spite of the establishment of institutional support systems for intra-African trade, trade among African countries accounts for no more than 4 per cent of total African trade with the world. An important caveat is necessary here. These are official estimates which exclude almost a similar percentage or even more of intra-African trade falling under unrecorded transborder trade discussed below. The percentage share of intra-African exports declined from 5.60 per cent in 1970 to 3.13 per cent in 1980 and thereafter increased slightly to 4.83 per cent in 1985. The intra-African trade in each country’s total trade and each region’s total trade is shown in Tables 15.4 and 15.5. The intra-regional exports as a percentage of total intra-African trade for 1980 were 3.12, 10.34, 0.81 and 4.24 per cent respectively for Central Africa, Eastern and Southern Africa, North Africa and West Africa. The intra-African exports as a percentage of Africa’s total exports remained at 4.6 per cent between 1980 and 1985 (Table 15.4). By 1985 the Central African, Eastern and Southern African and West African regions had registered declines in their intra-African trade (exports). Only North Africa, which is the region with the lowest intra-export trade, had slightly increased its percentage share from 0.81 to 1.94 per cent between 1980 and 1985. In relative terms the African industrial related groups (chemicals (SITC 5), machinery and transport equipment (SITC 7) and other manufactured goods (SITC 6+8)) have lagged behind the primary sector extractive industries (SITC sections 0+1, 2+4, and 3) (see Table 15.7). Within the primary sector exports, food production has grown relatively less than mineral fuels. Taking into account that only a few countries have an oil sector extractive industry (Nigeria, Cameroon, Congo, Gabon and Angola) there has been a relative decline of both the industrial and food sectors. From 1980 to 1985 intra-African imports as a share of her total imports

313

AFRICA Table 15.4 Intra-African trade (exportsa; values in millions of US dollars)

314

ISSUES FOR DEVELOPMENT Table 15.4 continued

a

Subregional totals relate only to countries for which data was available Source: IMF (1986) Direction of Trade Statistics Yearbook

remained at around 5 per cent. There was a decline in intra-African imports in Central Africa from 8.49 per cent in 1980 to 6.41 per cent in 1985 and slight percentage increases in other regions (Table 15.5). Total trade between African regions is relatively low. There is a tendency for trade to be concentrated between countries of the same region. Thus, for example, in 1982 the intra-Central African total trade accounted for 36.3 per cent of the region’s total intra-African trade (Table 15.6). In 1985 it had declined to 34 per cent but by then the volume of trade within Central Africa had declined by 258 per cent and its total trade with the rest of the African regions dropped from $423.6 million in 1982 to $175.4 million (241.5 per cent) which is a steep decline. The trend is more or less the same in other regions. The intra-regional trade of the Eastern and Southern African region dropped from $840.2 million in 1982 to $496.5 million in 1985 (69 per cent) which represented 73 and 82.8 per cent respectively of intra-African total trade (Table 15.6). This rather high percentage share in total intra-African trade in 1985 is explained by a large fall in total intra-African trade (191 per cent) rather than by an increase in actual trade volumes of the Eastern and Southern African region (Table 15.6). The commodity composition of intra-African trade in 1985 was led by the

315

AFRICA

Table 15.5 Intra-African trade (importsa; values in millions of US dollars)

316

ISSUES FOR DEVELOPMENT Table 15.5 continued

a

Subregional totals relate only to countries for which data was available. Source: IMF (1986) Direction of Trade Statistics Yearbook

following: mineral fuels and related materials (SITC section 3), 34.8 per cent; food, beverages, tobacco (SITC section 0+1), 26 per cent; other manufactured goods (SITC 6+8), 14.8 per cent (Table 15.7). The category mineral fuels and related materials became important between 1970 and 1985 and continued to increase its share from 11.2 per cent in 1970 to 34.8 per cent in 1985 (Table 15.7). The food, beverages and tobacco group which occupied first place until 1970 fell into second place, and its share remained almost constant from 1970 to 1985. On one hand, mineral fuels and food together constituted some 60 per cent of the total intra-African trade during this period. On the other hand, manufactured goods, which was the most important group in 1970 constituting 43.6 per cent, fell gradually to 29.8 per cent in 1975 and to around 15 per cent in 1985. Key manufactures such as chemicals (SITC section 5) and machinery and transport equipment (SITC section 7) stagnated at very low levels during this period. Some characteristics of unrecorded transborder trade Africa’s emphasis on establishing supranational organizations for regional economic integration and to orchestrate regional trade liberalization through

317

AFRICA Table 15.6 Intra-subregional trade matrix (total exports+imports; values in millions of US dollars)

Source: ECA Statistics Database

phased tariff reductions took the form of official institution building. But as shown above, in spite of the continuing proliferation of multilateral treaties, protocols and resolutions concerned with promoting regional trade, intra-African trade has stagnated. The share of intra-African trade in the 1980s is lower than it was before 1975. In contrast, unrecorded transborder trade (UTT) appears to be on the increase in much of Africa.8 Despite the constraints of poor transportation, difficult communications, and government restrictions, African traders seem to respond to market signals transmitted through UTT. Because of the difficulties connected with quantifying the magnitude of unrecorded trade across African borders an impressionistic picture of UTT’s general composition has only been derived through case studies. The main actors involved in UTT are small-scale quasi-professional traders and organized (often armed) professionals (Barad 1990:103). The former use various means to move small quantities of goods across borders relying mostly on

318

ISSUES FOR DEVELOPMENT Table 15.7 Value of intra-African trade by commodity classesa (in millions of US dollars)

a

Calculation based on export figures. Excludes Zimbabwe. c SITC=Standard International Trade Classification. Sources: UN (1987) Monthly Bulletin of Statistics May UN (1986) International Trade Statistics Yearbook, vol. I b

personal relationships with local residents on both sides of the border. The latter involve the participation of government officials and other politically powerful actors which guarantee ‘unnoticed’ entries across borders. Traded goods include luxury goods, food, clothing, tools, spare parts and construction materials. Much of the trade in manufactures appears to involve products originating from outside Africa, and local products like agricultural products, wild animal products (ivory, skins) and woven textile materials. The size of UTT flows is estimated to be at least equal to and in some cases much larger than the levels of intra-African trade officially recorded by African states (Barad 1990:104). The actual amount of UTT between any two countries is likely to vary in relation to relative prices and government policies, especially exchange control and high tariffs and quantitative restrictions. Often UTT provides economic linkages between raw material producers and states functioning as market poles within their subregions. For example, Benin, Togo, Djibouti, Kenya and Gambia provide outlets for extraregional imported goods for the countries surrounding them. African governments’ response to UTT has been that of ignoring its existence, or deeming it too small scale to be considered significant, or tightening up border controls through policing them. In reality, there is increasing UTT across African borders constituting unofficial means to achieve de facto liberalized trade. There are

319

AFRICA

many ways through which African governments can respond to UTT. Price adjustments at the regional level would greatly facilitate harmonization of monetary and fiscal policies. This would help eliminate policy-induced distortions that have been responsible for UTT. Government-organized regional economic integration could adopt policies that encourage producers to specialize in activities that respond to consumer demands within the larger regional markets. By adopting processes that facilitate regional trade liberalization, African regional institutions could integrate the bottom-up UTT activities into official trade. In this way, entrepreneurship gained during years of smuggling, often referred to as ‘bad trade’, could be harnessed to provide more employment and access to scarce resources. Problems and constraints of intra-African trade As shown above, at the global level intra-African trade has remained below 5 per cent of Africa’s total trade with the rest of the world. Though there are considerable variations between countries and regions the trends and structure of African trade have remained the same since the colonial days. It continues to depend very heavily on the exports of primary commodities to and imports of industrial products from developed market economies. There are numerous obstacles to the expansion of intra-African trade. The often cited ones relate to: 1 2 3

4 5 6

the lack of appropriate technology for the production of mass consumer goods; the paucity of technical expertise and trained personnel in management and entrepreneurial skills; the lack of policy measures focused on the removal of constraints, such as lack of information, supply and demand divergence in national economies, tariff and non-tariff barriers; the lack of banking and other financial facilities; the inadequacy of transport and communications and marketing structures; the relative and sometimes absolute neglect of rural African population leading to extremely low levels of income.

It can be argued further that the main reason for the low level of intra-African trade relates to the structure of production and distribution. Africa’s industrial sector was allocated about 40 per cent of total investment in the 1960s and 1970s compared with 5 per cent allocated to agriculture. After the first initial successes, Africa’s manufacturing value added (MVA) stagnated at 10 per cent of GDP and thereafter industrial capacity began to decline due, inter alia, to the deterioration of plant and equipment. The capabilities needed to run industrial plants efficiently were neglected. Instead there was simply adding to physical industrial capacity in the static sense of matching given factor endowments. The management and technical skill factors were not given due consideration. Most often new factories were

320

ISSUES FOR DEVELOPMENT

de-linked from the local economy and thus rendered incapable of generating exports, skills or growing employment. For example, in Cameroon attempts to expand heavy industry were constrained significantly by poor technical design and inadequate infrastructure, especially power supplies (Riddell 1989), while in Kenya shortages of engineers and technicians have been a cause of the lack of deepening of the industrial base (Coughlin and Ikiara 1988). For both Kenya and Zimbabwe, good production engineers in the textile subsector have been a major factor in creating international competitiveness (Riddell 1989). These factors have largely contributed to the present tiny industrial base in Africa. The total MVA in thirty-nine African countries came to $12.4 billion in 1983, some 46 per cent of that in India, 57 per cent that of Korea and 22 per cent of Brazil (Lall 1989:6). Only two countries in sub-Saharan Africa, Nigeria and Zimbabwe, had an MVA of over $1 billion each, while twenty countries had MVAs of under $100 million each (Lall 1989). Of the middle-income non-oil countries, Zimbabwe had the largest MVA followed by Ivory Coast and Zambia; of the oil-rich countries, Nigeria followed by Cameroon; of the low income countries, Kenya had the highest MVA. African manufacturing production has not been adequately geared to supply inputs and equipment for the economic sectors: agriculture, construction, industry, transport and communication, etc. Neither has it been able to provide the much-needed basic wage goods for the growing urban populations, let alone the rural populations. These domestic advantages have to be exploited before moving into the more competitive stage of exploiting existing and potential regional complementarities. Another main constraint in the improvement of intra-African trade is the question of closing the gap of divergence between demand and supply in African economies which is partly a result of the lack of information. This applies to all sorts of products including food commodities and consumer goods, in which imports from industrialized countries abound while current and potential capacity exists for production to meet the demand of these products.9 The rather long digression into production is meant to dispel the illusion found in some of the designs of integration schemes that the lowering of tariff and nontariff barriers and the establishment of preferences are enough. Intra-African trade has to be based on, and provide the market for, production of manufactures that is geared to the current demand of development. All other supportive policy measures—removal of tariff and non-tariff barriers, provision of banking facilities including trade credit and guarantee schemes, creation of institutions such as clearing houses and subregional trade and development banks—are meant to reinforce trade-flows arising from a newly created production base. Arthur Hazlewood observed over twenty years ago: integration is not simply a matter of lowering tariffs. The existence of tariffs is not the sole, or even the primary impediment to trade between the countries of Africa. The main reason for the low level of trade is to be

321

AFRICA

found in the economic structure of the countries. The economies of virtually all of them are directed to the export of primary produce and the import of manufactures. The import requirement of an African country cannot to a large extent (except in a few cases) be satisfied by other countries of Africa [and]… There is the fact that the ‘infrastructure’ for intra-African trade is generally lacking. (Hazlewood 1967:10). The problem just described is still with us. However, these problems do not in any way reduce the prospects and great potentials for developing and expanding intraAfrican trade. For instance, in terms of absolute market size, the African regional economic groupings currently have vast and untapped market potential. ECOWAS covers sixteen countries of the West African region with a total population of around 170.5 million. The PTA in the Eastern and Southern African region has a membership of eighteen countries forming a market of some 175.5 million people. ECCAS in Central Africa has ten member countries with a total population of 68 million people. In North Africa, the Maghreb Union has a membership of five countries comprising Algeria, Libya, Mauritania, Morocco and Tunisia with a population of 60.6 million (African Development Bank 1991). An integration scheme that will include Egypt with a population of 55 million people has not been worked out. ISSUES FOR REGIONAL ECONOMIC INTEGRATION AND INTRA-AFRICAN TRADE The general discussion on economic integration and intra-African trade has focused on the small size of the domestic markets for African products, and on non-complementary production, distribution and infrastructural structures of African economies. The market size issue alone is, however, not an issue as shown by the fact that large African countries (e.g. Nigeria, Ethiopia and Zaire with populations larger than 30 million) are not necessarily doing well. A central role in solving the main problem facing regional economic integration seems to lie in the twin issues of (a) enlargement and consolidation of the domestic markets of each country and (b) building up of capacities, particularly for the correction of the massive global imbalance between the demand for and supply of modern factor inputs. A partial list of factors which deeply affect the socio-economic policies and programmes of African countries include: 1 2

the long-term decline in the share of primary products in international trade and in their real purchasing power; the long-term decline of the share of developing countries in international trade and the insignificant and steeper declining share of Africa’s total trade with the rest of the world;

322

ISSUES FOR DEVELOPMENT

3

4

the irreversible weakening of North/South relations, reinforced by growing protectionist measures in the industrial countries and the establishment of regional trading blocs (Asia-Pacific Rim, EEC 1992, USA/Canada/Mexico); technological displacement of some basic primary commodities exported by Africa as a result of product substitution and changes in the standardization of products.

The building of relevant capacities at both national and regional levels is probably the most effective way of consolidating domestic markets as engines for growth. African countries must aim to reduce the excessive global imbalances between the demand for and supply of modern factor inputs (entrepreneurship, management, skilled manpower for procurement, production and marketing; natural resources, raw and intermediate materials; equipment, spare parts, implements and tools; physical infrastructure; institutional infrastructure and services). The majority of these factor inputs are at present supplied from developed market economies and paid for by foreign exchange earned from Africa’s dwindling basket-priced primary products. It would seem that, to policy makers, the economic theory’s great workings of the doctrine of dynamic comparative advantage and the concomitant results of ‘international division of labour’ were accepted without recognizing that the bulk of comparative advantage was neither accidental nor natural but man-made, consisting of modern factor inputs. It can be argued that comparative advantage, instead of being given, can be acquired and that one can change the nature of its dynamic paths. Yet another powerful element of the doctrine of international division of labour inhibiting intra-African co-operation is the understanding that international trade, i.e. the existing pattern of North/South trade, is itself the engine of growth. The preconditions for this dynamism seem to be lost in the careless reading of the economic history of western civilization which conceal the existence of fundamental changes (socio-economic, political, institutional, etc.) that took place within Britain, France, Germany and the United States. These changes set the stage for the emergency or deliberate creation of internal engines for growth with the attendant implications for the domestic supply of factor inputs. It was on the basis of these domestic factor inputs as the first engine of growth that fuelled the later engine of growth emanating from the functioning of international trade. It is, therefore, an important requirement for intra-African cooperation and regional economic integration that engines of growth have to be deliberately built up with each country’s domestic market playing a key role in economic growth in general and in economic co-operation in particular. This deliberate development of the internal market as a precondition for subregional and regional markets cannot be left to chance or entirely to the benevolence of foreign factor inputs. The most important function of imported (as well as domestic) factor inputs in the early stages of planned and engineered development and economic growth is to generate new additional and relevant domestic factor inputs as fast as possible. By its very nature economic co-operation is a long-term political economy

323

AFRICA

question. It deals with the long-term visualized future of what the co-operating group of countries and their peoples should achieve in, say, twenty to twenty-five years’ time or more. It is this normative or visualized future that should constitute the basis of planning and programming, of periodic decisions on priorities; of the acquisition or generation of scarce factor inputs and their allocation and use; of the designs of strategies and choices of instruments of national engines of growth. The lack of such long-term vision has led to the majority of integration schemes continuing with the absurdity of attempting to enlarge markets by adding together several tiny national markets without regard to their character and dynamics, the disintegrative pressures to which these markets are subjected and the need for their deliberate consolidation and development. With the similarity of African economies as producers and exporters of primary products and importers of manufactured goods, trade expansion resulting from market integration would likely remain modest for more than the short run. Even market size alone, as measured by population, is unlikely to play a positive role. Thus the economic benefits of specialization and economies of scale necessary to sustain efficient and competitive production cannot be achieved in markets as small and poor as those described above. This is worse still under conditions of no dynamic growth of intrafirm and intraindustry trade involving factor inputs. In other words, without the complementary growth of industrialization at the national and regional levels, there is no prospect for rapid economic growth in Africa and rapid expansion of intra-African trade, particularly in industrial raw and intermediate materials. Natural resources or industrial raw and intermediate materials and the setting up of physical infrastructure including national or multinational corporate instruments of production, transport and trade must, therefore, feature more prominently in negotiations over intra-African economic co-operation. Also in the long term, regional economic integration will depend on the political will of the member states, the existence of which is a precondition for its success. In the African experience the lack of or insufficient political will has seriously hindered the ability of integration schemes to implement the decisions taken by the member states. This has inevitably meant that their concrete achievements in terms of promoting investment, regional specialization and intra-African trade have been limited and in some cases non-existent (Robson 1990:142). Lack of political will manifests itself in various forms, including: (a) misdirected member states’ economic relations which become more outward than inward looking during the very process of economic integration; (b) perception of economic integration as an exclusive business of governments to the neglect of the private sector, chambers of industry and commerce, farming groups, etc.; (c) indecisiveness by member states in rationalizing, harmonizing and even reducing the number of intergovernmental organizations, where they have simply proliferated beyond their usefulness.10 But equally, the strength of political will to integrate is not unrelated to the expectations of material benefits. The member states’ performance will be influenced by the ability of the institutions ‘to adapt

324

ISSUES FOR DEVELOPMENT

their priorities and practices in the light of experience to identify and achieve not only long-term gains but also short-term benefits in a time perspective that is politically relevant’ (Robson 1990:143). Thus strengthening of the technical and administrative capacities of the integration institutions in turn strengthens the political will of the member states. The other issue that hinges on political will is the harmonization of member states’ macroeconomic policies, especially currencies, taxation and custom procedures. For example, the setting up of a future unified monetary zone can only start at the regional level. The establishment of a monetary zone and a single currency within each African region is a necessary first step in enhancing production and trade. In its endeavours to achieve a unified monetary zone, the PTA in 1984 established a clearing house which was followed later on by the introduction of travellers’ cheques in 1988 for use by residents of member states travelling in the region. These are necessary steps towards eventually ensuring the convertibility of subregional currencies.11 The fundamental role of the Regional Economic Communities is to implement a joint policy framework that will facilitate structural changes on a harmonized basis that are required to enable integration to yield its benefits. This will inevitably require that steps are taken to ensure that regional dimensions of economic policy are dealt with at the national level. Modern private sector integration, collective trade liberalization, consonant harmonization and rationalization of tariff structures, systems of internal taxation and investment incentives are policies that will improve the enabling environment for regional economic integration. CONCLUSION The external environment facing Africa will continue to be hostile. The growth forecasts of the industrialized countries indicate a continued slowdown, at least during the 1990s. This is compounded by growing protectionist policies, changes in the pattern of commodity use, shifts of the growing sector from manufacturing to services and establishment of regional trading blocs in industrialized countries. With the single market of the European Community from 1992 and the consolidation of other integration blocs in both North America and Asia, selling of African products in the markets of Europe, North America and Japan will get tougher and tougher. The success of regional economic integration in Africa will essentially depend on the political will of the member states in enabling the Regional Economic Communities to implement their programmes timely. A pragmatic approach in the exercise of political will will be fruitful and rewarding if it emphasizes development based on linkages between national and regional programmes and also creates an environment that enables maximum participation of nongovernmental actors. The latter include the private sector, chambers of commerce and industry, mass media, intellectuals and professional and other organizations.

325

AFRICA

The institutional mechanism has been set in the form of the African economic groupings, but half-hearted attempts in implementing decisions taken jointly have caused apathy rather than active promotion of regional economic integration and intra-African trade. It is probable that individually, African policy makers have been looking over their shoulders to erstwhile and traditional partners in Europe, North America and elsewhere. The message is, however, becoming clearer. The traditional trading partners are getting busy in their own areas and concentrating more and more in regional trading blocs near home which to them are more efficient and more competitive. Whilst Africa should work tirelessly to penetrate those markets with competitive products, raw materials and cheap labour, the values of African products are no longer as attractive as before. There has seldom been a systematic analysis of the economic and political actors who promote and whose interests are served by regional economic integration and growth of intra-African production and trade. Largely because of the superficial analysis of the problem of trade and development in Africa, it is not understood why member countries of an integration grouping often agree on policies that turn out to be independent of the conditions and political environment within which they are supposed to operate. It is, therefore, necessary that key local social forces whose interests help to promote intra-African trade come to play a central part in our analysis. In addition to governments there are industrial groups, farming and other economic actors. At present the ‘national’ positions of the African economic and social actors are not well developed and often they are inconsistent and thus more liable to be manipulated by foreign interests. The most immediate challenge in African trade and development lies in improving the production, infrastructural and trade liberalization mechanisms at the regional level. Rhetoric and countless resolutions by African leaders have to be followed through by practical measures and means, such as ratifying protocols and conventions on intra-African trade, and providing popular support for regional integration. NOTES 1 Data is derived from the UN Monthly Bulletin of Statistics, May 1987. 2 The LPA is an African development policy document adopted by the African Heads of State and Government under the auspices of the Organization of African Unity (OAU) in April 1980, Lagos Plan of Action for economic development of Africa 1980–2000, Organization of African Unity, 1985. 3 The UDEAC (L’Union Douanière et Economique de l’Afrique Centrale) was established in 1964 by six Central African countries: Gabon, Congo, Central African Republic, Cameroun, Chad and Equatorial Guinea. The Southern African Customs Union (SACU) was established in 1910 by South Africa, Botswana, Lesotho and Swaziland. Namibia joined SACU after its independence in 1990. 4 This compares well with the Latin American Free Trade Association (LAFTA) in Sidney Dell (1963), Trade Blocks and Common Markets, New York, pp. 198–227.

326

ISSUES FOR DEVELOPMENT

5 The PTA member states are Burundi, Comoros, Djibouti, Ethiopia, Kenya, Lesotho, Madagascar, Malawi, Mauritius, Mozambique, Namibia, Rwanda, Somalia, Sudan, Tanzania, Uganda, Zambia and Zimbabwe. Zaire has observer status. 6 The PTA protocols cover (a) reduction and elimination of trade barriers on selected commodities to be traded within the PTA; (b) customs co-operation within the PTA; (c) rules of origin for products to be traded between the members of the PTA; (d) reexport of goods within the PTA; (e) transit trade and transit facilities; (f) clearing and payments arrangements; (g) transport and communications; (h) co-operation in the field of industrial development; (i) co-operation in the field of agricultural development; (j) simplification and harmonization of trade documents; (k) protocol and standardization and quality control. In addition, there are two special protocols in respect of Botswana, Lesotho and Swaziland and one on special provisions in respect of the Comoros and Djibouti. 7 African imports over the same period showed an inconsistent trend of first increasing from $11,229 million in 1970 to a peak of $86,112 million in 1981 and thereafter dropping steadily to $64,150 million in 1986 before rising moderately to $75,271 million which was the same level as in 1982. Exports more or less followed the same trend (Table 15.1). 8 A working definition of ‘unrecorded transborder trade’ includes any trade that takes place outside of official or formal economic structures subject to state regulation and control. 9 African food imports have risen from $485.6 million in 1970 to $5,514.3 in 1985. There are a considerable number of African agricultural products that are exported abroad whilst other African countries import similar products from Europe, Latin America, Australia and New Zealand. See United Nations Economic and Social Council (1987b). 10 There are currently more than 200 African organizations involved in regional cooperation, of which more than 160 are intergovernmental and the rest nongovernmental. This raises problems of duplication of effort, and dissipation of member states’ meagre financial, human and other resources that are spread thinly over a number of organizations, most of which must either be rationalized or dissolved. See UNDP/RBA/DRP data base, New York, 1990. 11 Intra-African trade has shrunk steadily in the past decade from its already low levels in step with the decline of foreign exchange resources of these countries.

APPENDIX: ABBREVIATIONS ACP ADB CEAO ECCAS ECOWAS IOC MRU PAEC PTA RECs SADCC UMA UTT

African-Caribbean-Pacific Group of Countries under the ACP-EEC Conventions African Development Bank Communaute Economique de L’Afrique de L’Ouest Economic Community of Central African States Economic Community of West African States Indian Ocean Commission Mano River Union Pan-African Economic Community Preferential Trade Area for eastern and southern African states Regional Economic Communities Southern African Development and Co-ordination Committee Union de Maghreb Arab Unrecorded transborder trade

327

AFRICA

REFERENCES ACP-EEC (1989) Joint Assembly Working Documents 1989–1990, Document ACP-EEC 4/ 89B/Fin., 5 September. African Development Bank (1991) Selected Statistics on Regional Member Countries, Abidjan. Balassa, B. (1961) The Theory of Economic Integration, London: Allen & Unwin. ——(1967) ‘Trade creation and trade diversion in the European Common Market’, Economic Journal vol. 77. Barad, R. (1990) ‘Unrecorded transborder trade in sub-Saharan Africa and its implications for regional economic integration’, in IBRD/World Bank, Background Papers, The Long-Term Perspective Study of Sub-Saharan Africa, Volume 4, Proceedings of a Workshop on Regional Integration and Cooperation. Coughlin, P. and Ikiara, Gerrishon K. (eds) (1988) Industrialization in Kenya: In Search Of A Strategy, Nairobi: Heinemann Kenya and London: James Currey. Hazlewood, A. (1967) ‘Problems of integration among African states’, in A.Hazlewood (ed.) African Integration and Disintegration: Case Studies in Economic and Political Union, Oxford: Oxford University Press. ——(1987) ‘Economic integration: Lessons for African recovery and development’, International Conference on Africa: The Challenge of Economic Recovery and Accelerated Development, 15–19 June 1987, Abuja, Nigeria, United Nations Economic Commission for Africa. Lall, S. (1989), ‘Structural problems of African industry’, paper presented at the Workshop on Alternative Development Strategies in Africa, Oxford, 11–13 December 1989. Lipsey, R.G. (1956–7) ‘The general theory of the second best’, Review of Economic Studies. Meade, J.E. (1956) The Theory of the Customs Unions, Amsterdam: North-Holland. Mulat, T. (1990) ‘Intra-African trade’, in J.Pickett and H.Singer (eds) Towards Economic Recovery in Sub-Saharan Africa: Essays in honour of Robert Gardner, London: Routledge. Organization of African Unity (1991) Treaty Establishing The African Economic Community, Abuja, Nigeria, June 1991 . Riddell, R. (1989) ‘Industry: An overview’, paper presented at the Workshop on Alternative Development Strategies in Africa, Oxford, 11–13 December 1989. Robson, P. (1990) ‘Economic integration in Africa: A new phase?’, in J.Pickett and H. Singer (eds) Towards Economic Recovery in Sub-Saharan Africa, Essays in honour of Robert Gardiner, London: Routledge. Treaty for the Establishment of the Preferential Trade Area for Eastern And Southern African States, Lusaka, June 1982. United Nations Economic Commission for Africa (UNECA) (1987) Abuja Statement on the Challenge of Economic Recovery and Accelerated Development, Abuja, Nigeria, 15–19 June 19 87. United Nations Economic and Social Council (1987a) ECA Doc. E/ECA/Trade/24/Rev.1 ‘Intra-African trade: Current flows and prospects’, Addis Ababa, August 1987. ——(1987b) ECA Doc. E/ECA/Trade/85 ‘The development and expansion of intraAfrican trade: A policy paper’, Addis Ababa, 20 August 1987. United Nations Population Statistics, 1989. Viner, J. (1950) The Customs Union Issue, Carnegie Endowment for International Peace, 1950. ——(1972) ‘The economics of Customs Union’, in P. Robson (ed.) International Economic Integration, Harmondsworth: Penguin Modern Economics Reading.

328

INDEX

Africa: adjustment with growth 229–31; currency devaluations 230; customs unions 308–9, 310, 311; domestic factor inputs 323–4; economic community model 308; economic integration 307–8, 324–5, 326; exchange rate policies 229–33; exports 303, 305, 313; imports 303; industrial sector 313, 321; institutional mechanism 326; manufacturing value added 320; political will for integration 324–5, 326; preferential trade areas 309; production approach 309–11; regional economic groupings 307–9, 322; regional trade 303, 306; socioeconomic policies 322–3; tariffs 322; world trade 303, 304, 312–13; see also individual countries agriculture: Bangladesh 218; Ethiopia 280, 282–3, 284–88, 290–1; Guatemala 95; modernized 167–8; Taiwan 167–9, 198 (n6) Ahmad, A. 220 America (New World): Dominicans 56, 57; empire building 49–58; indigenous population 54–5, 64; revenue from 49; and Spanish conquest 43–4, 58–64; see also Central America; Latin America; US Anderson, K.H. 215 Andersson, T. 140, 142 ANIEs (Asian newly industrialized economies) 137, 151–2 (n2); debt service ratio 141; economic growth 138–40; government intervention 139– 40 appropriability 17 appropriate technology, lack 320

ARER (appropriate real exchange rate) 236–7, 240, 241–2 ASEAN (Association of South-East Asian Nations) 137; debt service ratio 141; economic growth 138–40; government intervention 139–40 Asia: export propensities 118; export values 125; growth of exports 120; sales 122; see also ANIEs; ASEAN; East Asia; South Asia balance of payments: import protection 251, 254; Kenya 249, 259; Uganda 265;Zambia 272, 274 Balassa, Bela 139, 200 (n22), (n23), 310 bananas 93, 94–5 Bangladesh 203–4, 206–8; agriculture 218; consumer losses 218; corruption 219– 20; current account balance 209; DUP activities 219–20; exchange rates 207– 8; export-oriented policies 208; exports 210; GDP 210–11; import substitution 204–6, 212–23; income distribution 222; industrial policies 206, 223; industrial production 217–18; labour costs 213, 215; macroeconomic effect of reforms 209–11; natural disasters 221; privatization 216, 222; production losses 215–19; protectionism 212–15; rates of assistance 214–15; trade policies 207, 222; unionization 215; see also Pakistan Barad, R. 319 Batra, R.N. 151 Bevan, D. 251 Bhagwati, J. 212 Bhattacharya, D. 213, 217

329

INDEX

Bigsten, A. 257 Bound, J. 39 Braudel, Fernand 44, 45, 48 Brazil, US affiliates 111 bribery 213, 219–20, 222, 275 cacao 89, 90, 92, 312 capital intensive sector, technological progress 35–6, 38 capital stock 34 capitalism: dirigiste 154, 178; and KMT 170–2 Cavallo, D. 235 Central America: and Nordic countries, compared 74–83; predatory policies 81–2; see also Latin America Chang, Ching-hsi 186 Charles V, King of Spain 46–7, 50, 51 Chen, Ting-an 190 Chen Cheng 169 Chiang Kai-shek 166, 169, 171 Chile, US affiliates 111 China, economic growth 138 Chowdhury, O.H. 212, 213, 217, 222 cocoa 89, 90, 92, 312 coffee: Ethiopia 284, 292; Guatemala 93, 94–5; Kenya 249, 261; Uganda 262 Collier, P. 237 colonial heritage, institutions 102 Columbus, Christopher 45–6, 50 comparative advantage 13–14, 18, 20 conquistadores 50–1, 56, 59–61, 62 Contractor, F. 140 corruption 219–20; see also bribery Costa Rica: GDP per capita 75, 76; land distribution 83; trade changes 80 Cottani, J. 235 cotton 98–9, 106 (n41) Crane, G.T. 156 credit 99–100, 106 (n37) curb market 194–5 currency devaluation 255; Bangladesh 208; Burundi 230; Ethiopia 297, 300; Gambia 230; Ghana 230; Guinea Bissau 230; Kenya 230, 257–8; Madagascar 230; Nigeria 230; Sierra Leone 230; Somalia 230; Tanzania 230, 233, 241; Uganda 262, 266, 267–8; Zaire 230; Zambia 230, 251; Zimbabwe 230 currency inconvertibility 233 customs unions 308–9, 310, 311

Davis, Ralph 50, 51 debt crisis 2, 109, 142 debt peonage 90, 94, 97 debt service ratio 141, 295 Denmark, GDP per capita 75, 76 dependista theory, Taiwan 159–62 devaluation: see currency devaluation developing countries: economic history 43–4; exports 86, 110–11, 115–18, 120, 125; studies 231; terms of trade 71–3; and US affiliates 120–2 development economics 2, 71 direct investment 145–6; and debt service ratio 141; East Asia 140 dirigiste capitalism 154, 178 Domar, Evsey 92 East Asia: capital inflows 135; debt service ratio 141; direct investment 140, 145; economic development 135–40; inward investment 140–4; Japanese direct investment 144–9, 150–1 ECCAS (Economic Community of Central African States) 322 ECLA (Economic Commission for Latin America) 72 economic growth 3, 16–18 economic integration 307–11, 322–5 ECOWAS (Economic Community of West African States) 322 El Salvador, GDP per capita 75, 76 Elbadawi, I. 237 electrical machinery 120–1 Elliot, J.H. 46, 48 empire, theory and practice 49–58 employment, Ethiopia 282–3 encomienda 53–6, 66 (n26); contract costs 61–3; in Guatemala 90, 92, 104 (n4); Las Casas 55–8 equality of opportunity theory 159, 168 Ethiopia 280–1; agriculture 280, 282–3, 284–8, 290–1; budget deficits 292; civil unrest 281; co-operatives 285–6; coffee 292; currency devaluation 297, 300; debt service ratio 295; dual economy approach 281–3, 297–8; economic performance 290–5; economic reform programmes 295–7; employment 282– 3; exchange rates 299; exports 291–2, 293, 294, 295, 297–8; imports 292, 293–4, 297–8; institutions, and

330

INDEX

domestic economic policy 284–88; peasant associations 284; policy recommendations 297–300; state farms 287; taxation 287–8; trade and exchange rate policy 288–90; trade liberalization 297–8; villagization 285; wage rates 289, 290 Evans, P. 171, 173 exchange rates: crawling-peg system 260, 276; dual 262–3; and fiscal policy 235; flexible 241; parallel market premium 234, 238; policies 139, 229–33; regimes 190–1, 207–8, 209–10, 260, 262–3 export economies 87 export performance 109–11; by industry 118–22; manufactured 125; primary exports 88–93, 97–101, 277, 311, 313 export promotion 139, 170–1, 187–93, 208 export propensities 112–18, 121 expropriations 140 Farmers’ Associations 167–8 Ferdinand, King of Spain 44–6, 50 financial systems: Guatemala 96, 100; and investment 193–5; Korea 194; Taiwan 176, 193–4, 196, 200 (n19) Findlay, R. 37, 87 Finland, GDP per capita 75, 76 fiscal policy, and exchange rate policy 235 flexible exchange rates 241 foods, as exports 119, 120 foreign national income, increasing 36 free market system 156–9, 165 fundamentals and RER 234–41 Galenson, Walter 186 garment industries 213, 215, 217 GDP per capita 74–5, 76, 260 Ginsburg, N.S. 185 Glade, W. 86, 87 GNP per capita 1, 3, 4 gold 50–1 Gold, T. 157, 169, 172, 193 governments: intervention 139–40; price policies 99; rent-dependent 165, 277–8 growth theory: new 22–5; traditional 13– 14, 16, 19 growth with equity, Taiwan 158–9 Guatemala: adaptations 89–91, 94–5, 98– 9; banking 96, 100; cotton 98–9, 106 (n41); credit 99–100, 106 (n37);

currency reforms 95; exports 86–7, 88– 93, 97–101; GDP per capita 75; growth, export-led 93–7; institutions 91–3, 95–7, 99–101, 102–3; labour force 92–3, 100–1, 103; staples theory 87–8, 101–3; transportation costs 88–9, 93–4, 97–8 Gupta, K.L. 142 Haggart, S. 169, 170, 171, 174 Hanke, Lewis 59 Hansson, G. 296, 298, 300 Hazlewood, Arthur 307, 309, 321–2 hidalgos 53 Hispaniola 51, 54 Ho, S. 166 Honduras, GDP per capita 75 Hong, Wontack 190, 191, 193, 196 Hossain, M. 222 Hossain, N. 213 Hou, Chi-ming 186, 191–2 human capital formation 3, 4, 17, 29, 37– 40 Iceland, GDP per capita 75, 76, 78 immigration, and wage levels 39 import protection, and balance of payments 251, 254 import substitution: Bangladesh 204–6, 212–23; Korea 196; motives 220–3; Taiwan 170–1, 175 income distribution 4; Bangladesh 222; Taiwan 160, 166–7, 196 indigo 88–9, 90, 92 industrialization: Bangladesh 223; Pakistan 205–6; Taiwan 167 infant industry argument 175 institutions 14–16; Africa 326; analysed 87–8; colonial heritage 102; and development economics 71; Ethiopia 284–8; Guatemala 91–3, 95–7, 99–101, 102–3; Korea 188–93; Taiwan 188–93; transaction costs 88 insurance 15 integration theory, and international trade 310 International Conference of Africa (1987) 303, 307 international trade model 29–39, 310 intra-African exports 313, 314–15 intra-African imports 313, 315, 316–17 intra-African trade 303, 306;

331

INDEX

characteristics 312–18; commodity composition 315, 317, 319; priorities 312; problems 320–2; and regional economic integration 309–11, 322–5; unrecorded transborder trade (UTT) 313, 318–20, 327 (n8) investment: direct 140, 141, 145–6; financial systems 193–5; inward 140–4; Japanese in East Asia 135–7, 141, 142– 9, 150–1 Isabella, Queen of Spain 44–6, 51 Islam, M.A. 142 Islam, R. 205, 213 Jamal, A.M.M. 213 Japan: adaptations 143; colonial impact 184–5; currency alignments 143; current account imbalances 143; direct investment in East Asia 135–7, 141, 142–9, 150–1; financial deregulation 143; rising costs 143; and Taiwan 179 (n5), 185 Jebuni, C.D. 231 Johnson, G. 39 Joshi, V. 237 jute 216 Ka Chih-ming 170 Kamen, Henry 45, 46–7, 48–9 Kenya: balance of payments 249, 259; coffee 249, 261; currency devaluation 230, 257–8; economic background 248–9; exchange rate regime 260; foreign aid inflows 262; GDP growth rates 260; macroeconomic adjustments 256–62, 276–8; price indices 257, 258 Khan, M. 231 Kierzkowski, H. 37 KMT (Kuomintang) 166–70; and capitalism 170–2; as developmental state 172–7; and local Taiwanese 169; see also Taiwan Knight, M.D. 231 knowledge: and economic growth 16–18; and trade 18–19 Kojima, K. 136 Korea: economic performance 195–8; education 184–5, 186; exchange rate 190–1; export promotion 187–8; financial systems 194; government policies 140, 192, 197–8; income distribution 196; institutionalized 188–

93; labour-intensive manufacturing 184; manufacturing value added 185; military training 186–7; population density 188 Krueger, Anne 212, 220 Kuo, S.W.Y. 157, 190 Kuomintang: see KMT labour costs: Bangladesh 213, 215; Guatemala 92–3, 100–1, 103; see also wage rates labour disputes 197–8 labour economists 28, 29 labour-intensive manufacturing: Bangladesh 217; exports 184; Korea 184; protection 36–7; relative price 33– 4; Taiwan 160–1 Lagos Plan of Action (LPA) 303, 308, 310 Lal, S. 223 land reforms: Central America 81–3; Sweden 82; Taiwan 166–8 Las Casas, Bartolomé 55–6 Latin America: cost competitiveness 86; debt crisis 109; debt peonage 90, 94, 97; economic development 71–2; encomienda 53–8, 61–3, 66 (n26), 90, 92, 104 (n4); export propensities 115– 18; export values 125; growth of exports 120; minifundio-latifundio system 96–7, 100–1, 106 (n39); mita 57, 58, 63; repartimiento 54, 57, 63, 90, 92, 94; sales 122; see also Central America; US MNC affiliates learningby-doing 17–18, 19 legal system, and trade 15 Lewis, S.R. 206 Li, Kwoh-ting 187 Liang, Kuo-shu and Ching-ing Hou 186 Little, I.M.D. 175, 188, 191 LPA (Lagos Plan of Action) 303, 308, 310 Lucas, Robert 16 Lundahl, M. 87, 284 MacLeod, M. 88, 89, 90, 92 macroeconomic adjustments: Bangladesh 209–11; Kenya 256–62; Uganda 262–8; Zambia 268–75 Mahmood, S.A. 222 manufactured exports, value 125 manufacturing output, value 129 Mason, E.S. 191 Meade, J.E. 308

332

INDEX

Mengistu, Haile Mariam, President of Ethiopia 281, 285, 296 metals, as exports 119, 120 Mexico: manufacturing exports 116; and Spanish; US affiliates 111 Michaely, M. 221 minifundio-latifundio system 96–7, 100–1, 106 (n39) mita 57, 58, 63 MNCs (multinational corporations) 109– 10, 155 Moran, C. 231 Mulat,T. 313

product adaptability 18 product prices, capital-labour ratios 28–9, 30 productive capacity 18 property rights 16, 82, 88, 300 protectionism 140; and balance of payments 251, 254; Bangladesh 212– 15; by industrialized countries 312; labour-intensive manufacture 36–7; Taiwan 157–8, 175–6 PTA (preferential trade area) 309, 322, 327 (n5 and n6) purchasing power parity 232

Naya, S. 136 Ndulu, B.J. 229 Nester, W.D. 136 new growth theory 22–5 Nicaragua: GDP per capita 75, 76; predatory rulers 82; trade changes 80 Nordic countries: economic growth 74–5, 79, 81–3; land reforms 82; terms of trade 75–9; see also individual countries North, Douglas 71, 72, 81, 88, 102–3, 222 Norway, GDP per capita 75, 76 Nunnenkamp, P. 157

Rahman, A. 212, 213 Ranis, Gustav 196, 280, 281 real exchange rate: see RER regional economic integration 307–11; and intra-African trade 322–5 rent-seeking behaviour 165, 277–8 repartimiento 54, 57, 63, 90, 92, 94 RER (real exchange rate) 232; Ethiopia 299; and fundamentals 234–41; policy reforms 234–5, 237–8, 240; Tanzania 233–4; trade restrictions 238; see also ARER; ORER research, as good 18–19 Riddell, R. 321 Robson, P. 324, 325 Romer, Paul 16–17

open economy, model for 251–5 ORER (official real exchange rate) 236 Ortiz, A.D. 47 Ozawa, T. 136 Pakistan 204–6; see also Bangladesh Pan-African Economic Community 308 Pang Chien Ku 171, 173 Pang Chien-Kuo 169, 170, 171, 174 parallel market premium, exchange rates 234, 238 peasant export economies 87 per capita incomes 4, 13, 203 Peru 57–8, 64 Philip II, King of Spain 47–9, 51 Philippines 113, 115, 137–40 Pitt, M.M. 220 plantation economies 87 predator states 15–16 preferential trade area: see PTA primary exports 277, 311, 313; Guatemala 88–93, 97–101 primary products 73–4, 322, 323 privatization 216, 222

SACU customs union 308–9 Sauer, C.O. 50, 51 secondary market trading 195 Sen, B. 213, 220 services sector: export 31, 32–3; and international trade 29; technological progress 36 Shea Jia-Dong 176 Shinohara, M. 136, 137 silver 49, 63 skill supply 34–5, 37–9 slavery, prohibited 56, 61 Smith, Adam 2, 16 Smith, R. 90, 91, 92 smuggling 217, 219–20, 275 see also UTT Sobhan, R. 213, 220, 222 socio-economic structure 276 Södersten, Bo 2, 3, 28 South Asia: debt service ratio 141; economic growth 138 Spain: and America, commercial relations

333

INDEX

58–64; Catholic Kings 44–6; Crown/ conquistadores 59–61, 62; exploration and conquest 49–58; fiscal crisis 44–9; and Guatemala 91; missionary motives 51–4; mixed motives 52;pecuniary motives 49–51, 59 stagnation 71 staples theory 87–8, 101–3 statist approach, Taiwan 162–6 structural adjustment programmes 109–10 SunYat-sen 170 Sweden: economic history 20–1; GDP per capita 75, 76; property rights 82 Taiwan: agriculture 167–9, 198 (n6); as autonomous state 164–6, 173–5; comparative advantages 157, 160–1; dependista theory 159–62; development 155, 168; direct foreign investment 160; economic growth 154–6, 161, 195–8; education 186; equality of opportunity 159, 168; exchange rates 197; export promotion 187–8; financial systems 176, 193–4, 196, 200 (n19); free marketeers 156–9; government policies 155–6, 163–4, 191–2, 193, 195; growth with equity 158–9; import duties 157; income distribution 160, 194, 196; industrialization 167; institutionalized 188–93; interest rates 190; and Japanese colonialism 179 (n5), 185; Kuomintang (KMT) 166–77; labour-intensive manufacturing 184; middle classes 173, 174; as model for developing countries 178–9; nationalism 177; population density 188; protectionism 157–8, 175–6; reform policy (1950s) 172–3; ROC government 188–90, 195, 196, 197; small-scale industries 168, 174; stateowned enterprises 156–7; statists 162– 6; subsidies 190; trade liberalization 171–2; transitional political economy 177; unemployment rate 199 (n11); and US 159–60, 171–2, 174–5 Tanzania: domestic credit growth 238; liberalization of trade 237; RER and fundamentals 233–41, 244–5; terms of trade 236–7; trade liberalization 237 tariff liberalization 242, 322 taxation 46, 221 technological progress 19–20, 35–6, 38, 39

technology, appropriate 320 terms of trade (TOT) 71–3; international 261–2; Nordic countries 75–9; secular development 75–6; shocks 78–9; Tanzania 236–7; variability 76–8 trade: and autarky 43–4; creation/diversion 310; as engine of growth 2, 280, 283, 288, 290; and institutions 14–16; and knowledge 18–19; and legal systems 15; see also terms of trade trade liberalization 251–5; Ethiopia 297–8; Kenya 256–62, 276–8; Taiwan 171–2; Tanzania 237; Uganda 265; Zambia 268–75, 276–8 trade restrictions, and RER 238 trade theory, traditional 13–14, 19 transaction costs, and institutions 88 transborder trade, unrecorded (UTT) 307, 313, 318–20, 327 (n8) transport equipment 119–20 trickle-down theory 158–9 Tsiang, S.C. 193, 200 (n19) UDEAC customs union 308–9 Uganda: balance of payments 265; civil conflict 262; coffee 262; currency 262, 266, 267–8; dual exchange rate regime 262–3; economic background 248–50; Economic Recovery Programme 267; expulsion of Asian business families 250; foreign aid 230; GDP growth rates 266; inflation 265–6; liquidity crisis 268; macroeconomic adjustments 262– 8, 276–8; price indices 264; public sector wages 265; trade liberalization 265 UNCTAD (United Nations Conference on Tariffs and Trade) 73 UNECA (United Nations Economic Commission for Africa) 307 unionization, Bangladesh 215 US: grant-in-aid 187–8; human capital 39– 40; and Taiwan 159–60, 171–2, 174–5; wage differentials 39–40 US MNCs affiliates: export propensities 112–15, 121; exports 111–12, 114–15, 120, 123; sales 118–20, 122, 123, 126, 127, 128, 130, 131 UTT (unrecorded transborder trade) 313, 318–20, 327 (n8) Valladolid debate 56–8

334

INDEX

vent for surplus theory 87–8, 94, 102 Viner, J. 308, 310

Yu, Tzong-shian 190

wage rates: Bangladesh 213, 215; Ethiopia 289–90; Guatemala 92–3, 97, 100–1, 103; and immigration 39; skilled/ unskilled 28–9, 34–5, 37–9; Uganda 265; US 39–40 world trade 1, 20; Africa’s share 303, 304, 312–13 Wu Yuan-li 157 Yeh, Wan-an 187 Yin, K.Y. 172

Zambia: balance of payments 272, 274; copper 269, 270; currency devalued 230, 251, 269–70, 273–4; economic background 248–9, 250–1; economic policy 270; education 275; Fourth National Development Plan 274; GDP growth rate 273; inflation 270, 275; macroeconomic adjustment 268–75, 276–8; mining 250–1; natural disasters 275; New Economic Recovery Programme 274; Operation Food Production 268; price indices 270

335