Resource Abundance and Economic Development (W I D E R Studies in Development Economics)

TESOURCE ABUNDANCE AND ECONOMIC DEVELOPMENT

UNU WORLD INSTITUTE FOR DEVELOPMENT ECONOMICS RESEARCH ( U N U / W I D E R ) was established by the United Nations University as its first research and training centre and started work in Helsinki, Finland in 1985. The purpose of the Institute is to undertake applied research and policy analysis on structural changes affecting the developing and transitional economies, to provide a forum for the advocacy of policies leading to robust, equitable, and environmentally sustainable growth, and to promote capacity strengthening and training in the field of economic and social policy-making. Its work is carried out by staff researchers and visiting scholars in Helsinki and through networks of collaborating scholars and institutions around the world.

Resource Abundance and Economic Development

Edited by R. M . AUT Y

A study prepared for the World Institute for Development Economics Research of the United Nations University (UNU/ WIDER)

OXFORD UNIVERSITY PRES S

This book has been printed digitally and produced in a standard specification in order to ensure its continuing availability

OXFORD UNIVERSITY PRESS

Great Clarendon Street, Oxford OX2 6DP Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide in Oxford New York Auckland Bangkok Buenos Aires Cape Town Chennai Dar es Salaam Delhi Hong Kong Istanbul Karachi Kolkata Kuala Lumpur Madrid Melbourne Mexico City Mumbai Nairobi Sao Paulo Shanghai Singapore Taipei Tokyo Toronto Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries Published in the United States by Oxford University Press Inc., New York © The United Nations University 2001 World Institute for Development Economics Research of the United Nations University (UNU/WIDER) Katajanokanlaituri 6B, 00160 Helsinki, Finland The moral rights of the author have been asserted Database right Oxford University Press (maker) Reprinted 2002 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this book in any other binding or cover and you must impose this same condition on any acquirer ISBN 0-19-924688-2

Foreword Until recently, most economists have neglected the role that natural resources play in economic development. They have reported sharp differences in the performance of differing countries and regions, but the connection with the natural resource endowment has either been ignored or, at best, underestimated. This book corrects this neglect by synthesizing research from numerous subfields within the economics literature that have hitherto tended to be isolated from each other. The synthesis builds a strong case for the fact that since the 1960s the natural resource endowment of a developing country has strongly affected the efficiency with which it uses capital and the nature of its long-term developmental trajectory. The book distils the basic findings into two distinct models and then increases their explanatory power by introducing permutations to allow for differences in the specific character of the natural resource endowment. The first model is the staple trap model and it is strongly associated with resource-abundant development. It shows how trade policy closure leads to the misallocation of inputs that depress investment efficiency and retard (and even reverse) the competitive diversification of the economy so that the economy becomes vulnerable to a growth collapse. The staple trap model is also associated with social tensions that stem from heightened income inequality, a retarded passage through the demographic cycle, an unfavourable dependency/worker ratio, the slow acquisition of skills and the tardy accumulation of the public social capital that an increasingly sophisticated economy requires in order to lower transaction costs. Recovery from a growth collapse is protracted because all forms of capital (natural, physical, human and social) tend to be eroded. Policies are required that go beyond the 'prudent macro and micro' measures of the Washington Consensus. Governments must also address more deep-seated issues that include bolstering the more positive features of the political state, the accumulation of public institutional capital and recognition of the various impacts that different forms of natural capital introduce into the economic trajectory. The second model of competitive industrialization is associated with resource-poor countries, albeit not exclusively so. Appropriate policies can avert a growth collapse in resource-abundant countries as a handful of countries that includes Malaysia shows. The competitive industrialization model sustains a relatively open trade policy from a low level of per capita income that promotes labour-intensive manufactured exports that quickly absorb surplus rural labour. This sets in train a virtuous circle of equitable income distribution, the rapid accumulation of human and social capital, an accelerated passage through the demographic cycle and the speedy emergence of a dependency/ worker ratio that is highly favourable to saving and investment. Rates of capital investment quickly rise above 25 per cent of GDP without significant loss in investment efficiency and the economy diversifies competitively so that growth is rapid and egalitarian as well as being environmentally and socially sustainable.

vi

Foreword

The models are not deterministic, however, because policy counts and the basic models can be extended to account for a range of outcomes that reflect fundamental differences in the type of political state and the specific characteristics of the natural resource endowment. The research findings have strong implications for policy reform in the developing market and transition economies alike. UNU/WIDER gratefully acknowledges the financial contribution to the project by the government of Sweden (Swedish International Development Cooperation Agency—Sida). Giovanni Andrea Cornia Director, UNU/WIDER

Acknowledgements The editor gratefully acknowledges the stimulating, efficient and friendly working environment furnished by UNU/WIDER. Helpful comments were received from Moshe Syrquin and two anonymous reviewers of the manuscript, but they bear no responsibility for any shortcomings. The UNU/WIDER support staff provide an outstanding example of 'enabling social capital'. Special thanks are due to Anne Ruohonen, the exceptionally able and eternally cheerful secretary to the project. R. M. Auty Lancaster, November 2000

This page intentionally left blank

Contents List of Tables

xi

List of Figures

xiv

List of Contributors

xv Part I . Introductio n

1. Introduction and Overview

3

Richard M. Auty Part II . Critica l Parameters in Resource-Based Development Models 2. Natural Resources, Capital Accumulation, Structural Change, and Welfare

19

Richard M. Auty and Sampsa Kiiski 3. The Sustainability of Extractive Economies

36

Kirk Hamilton 4. Natural Resources, Human Capital, and Growth

57

Nancy Birdsall, Thomas Pinckney and Richard Sabot 5. The Social Foundations of Poor Economic Growth in Resource-Rich Countries

76

Michael Woolcock, Lant Pritchett and Jonathan Isham Part III. Long-Ter m Perspective on, and Models of, Resource-Based Growt h 6. Natural Resources and Economic Development: The 1870-1914 Experience

95

Ronald Findlay and Mats Lundahl 1. Short-Run Models of Contrasting Natural Resource Endowments

113

S. Mansoob Murshed 8. Political Economy of Resource-Abundant States Richard M. Auty and Alan H. Gelb

126

x

Contents Part IV. Development Trajectories of Resource-Abundant Countries

9. Competitive Industrialization with Natural Resource Abundance: Malaysia 147 Mahani Zainal Abidin 10. A Growth Collapse with Diffuse Resources: Ghana

165

Robert Osei 11. A Growth Collapse with Point Resources: Bolivia

179

Richard M. Auty and J. L. Evica 12. A Growth Collapse with High Rent Point Resources: Saudi Arabia

193

Richard M. Auty 13. Large Resource-Abundant Countries Squander their Size Advantage: Mexico and Argentina

208

Richard M. Auty Part V. Lessons for Policy Reform 14. Reforming a Small Resource-Rich Developing Market Economy: Costa Rica

225

Gustavo Barboza and Jose Cordero 15. Growth, Capital Accumulation, and Economic Reform in South Africa

239

Carolyn Jenkins 16. Reforming Resource-Abundant Transition Economies: Kazakhstan and Uzbekistan

260

Richard M. Auty 17. Reforming a Large Resource-Abundant Transition Economy: Russia

277

Anil Markandya and Alina Averchenkova 18. A Nordic Perspective on Natural Resource Abundance

296

Thorvaldur Gylfason Part VI. Conclusions 19. Conclusions: Resource Abundance, Growth Collapses, and Policy

315

Richard M. Auty Index

329

List of Tables 1.1. 1.2. 1.3. 2.1.

Characteristics of six natural resource endowment categories 4 Real per capita GDP growth 1820-1992, by region 5 Per capita wealth, by major global regions 1994 11 Per capita income, structural change and domestic absorption, post-1973 (share of GDP) 22 2.2. Per capita GDP and economic growth, Syrquin/Chenery aggregates 23 2.3. Investment, GDP growth and investment efficiency, six natural resource endowment categories 1960—1997 24 2.4. Trends in the dependants/worker ratio, selected countries 1960—1995 27 3.1. Genuine saving as per cent of GDP, 1997 43 A3.1. Genuine saving and its components, per cent of GDP, 1997 49 4.1. Education and resource abundance 58 4.2. Income distribution, education, and growth in Latin America, Asia, and sub-Saharan Africa 63 5.1. Social and political institutions among different export classifications 86 5.2. Inequality among different export classifications 87 5.3. Growth performance of different export classifications (1957-1997), by civil liberties 89 5.4. Growth performance of different export classifications (1957-1997), by social development indicator 89 7.1. Summary of analytical results 122 8 .1. t ypology if politivsl dysyrd, sfterlalll 821 8.2. Lagging socio-political indicators in sub-Saharan Africa 130 8.3. Share of rents in GDP, six natural resource endowment categories 1994 (%) 131 8.4. Four stage model of competitive sequenced industrialization (based upon South Korea, 1953-1997) 137 8.5. Extended genuine saving rates, selected countries 1970-1994 (% GNP) 138 8.6. Stylized facts models of the political economy of the natural resource endowment and economic growth 139 9.1. Structure of production, absorption, and exports, Malaysia 1960-1995 (% GDP) 148 9.2. Structure of Malaysian exports, 1960-1995 149 9.3. Domestic investment and saving, Malaysia 1960-1998 (% GDP) 151 9.4. Social capital indices, by resource endowment and selected country 154 9.5. Consolidated public sector finance, Malaysia 1970-1990 (% GNP) 155 9.6. Unweighted nominal and effective protection rate, Malaysia 1963-1987 157 9.7. Planned and actual development expenditure by sector (% share) 161 9.8. Number enrolled in education, Malaysia 1960-1995 (% age group) 161

xii

List of Tables

10.1. Resource endowment of Ghana compared to other countries 10.2. Physical capital accumulation and external trade indicators, 1960-1994 10.3. Selected indicators for the structure of production and growth, 1960-1994 10.4. Selected demographic and natural resource rent indices, 1960-1994 11.1. Changing production structure, Bolivia 1970-1995 (% GDP) 11.2. Some economic indicators, Bolivia 1970-1994 11.3. Share of mineral rents in GNP 1970-1994, four countries (% GNP) 11.4. Fraction of Bolivian taxes from mining and petroleum 1970-1997 (%) 11.5. Composition of exports, Bolivia 1970-1994 (%) 11.6. Comibol production costs 1980-1986 (US cents/lb) 12.1. Oil and gas rents share in GNP 1970-1994, three countries (% GNP) 12.2. Changing production structure, economy of Saudi Arabia 1970-1996 12.3. Scale of Saudi oil revenues 1970-1996 12.4. Investment and saving rates, Saudi Arabia 1970-1995 (% GNP) 12.5. Government expenditure, twelve countries 1994 (% GDP) 12.6. Trends in government revenue, Saudi Arabia 1976-1997 (billion rivals) 12.7. Sectoral growth rates, Saudi Arabia 1971-1995 (%/Year) 13.1. Socioeconomic preconditions in Argentina, Mexico, and Korea 13.2. Per capita GDP, selected countries 1900-1992 (1990 US$) 13.3. Scale of external shocks, four countries (% GNP) 13.4. Policy responses to external shocks, four countries (% shock) 13.5. Consequences of a neutral policy regime, Argentina 1960-1985 13.6. Per capita GDP growth rate trends 1870-1994 (%/year) 13.7. Trends in real exchange rate, three countries 1961-1984 (1965 = 100) 13.8. Mexican oil revenues and public finances 1971-1990 (% GDP) 14.1. Natural capital estimates, Costa Rica 1994 14.2. Human development index 1995, selected developing countries 14.3. Per capita GDP growth and reform, selected countries 1960-1997 (%/yr) 14.4. Current account deficit financing, Costa Rica selected periods (%) 14.5. Sectoral composition of GDP, Costa Rica 1976-1997 (%) 14.6. Employment by sector, Costa Rica 1976-1997 (% total employment) 14.7. Factor accumulation and total factor productivity growth, Costa Rica 1962-1995 14.8. Costa Rica, land use 1970-1994 (million hectares) 15.1. Economic performance indicators (% p.a.) 15.2. The determinants of growth in GDP per capita in comparative perspective 15.3. Wealth estimates for upper middle-income developing countries, 1994 15.4. Extended genuine saving, South Africa and income group averages, 1993

166 169 170 174 181 183 184 185 186 187 195 195 196 198 199 201 203 209 210 210 211 213 215 217 217 226 226 227 230 232 233 234 235 241 245 249 251

List of Tables 16.1. Liberalization, growth, and inflation in transition economies 1989-1994 16.2. Sectoral composition of GDP, Kazakhstan and Uzbekistan 1993-1997 (% GDP) 16.3. Projected Kazakhstan oil rent stream, three scenarios 16.4. Social capital indices, by natural resource endowment group 16.5. Dutch disease effects 1990-1997 (actual c.f. expected employment share) 17.1. Comparison of transitional indicators and GNP per capita 1994-1998 17.2. Transition indicators for selected countries, 1998 17.3. Structure of the economy of the Russian Federation, 1992-1997 17.4. External trade for Russia, 1994-1997 17.5. Transition indicator sum for CEE countries to 1998 17.6. Real exchange rate appreciation for selected transition countries 17.7. Potential and actual rents from natural resources in Russia ($billion) 17.8. Taxation of natural resources in Russia (consolidated federal and regional) 18.1. Nordic living standards 18.2. Nordic wealth: human, physical, and natural, 1994 18.3. Nordic investment, education, trade, and growth: An overview

xiii 261 265 266 267 269 279 283 285 286 287 288 290 291 297 297 298

List of Figures 1.1. Median GDP per capita (constant 1995 US$) resource-rich and resource-poor developing countries 2.1. Structure of value added in GDP, by country resource endowment 3.1. Genuine saving rate vs. depletion share of GDP in resource-dependent economies, 1997 4.1. Government's positive impact on the virtuous circle 4.2. Impact of resource abundance on the virtuous circle 5.1. Smoothed median per capita growth rates in 90 developing economies, 1955-1997 5.2. Smoothed median growth rates for 90 developing economies, 1957-1997 7.1. Effects of resource booms, devaluation, and industrial policies 8.1. Competitive industrialization model 8.2. Staple trap model 9.1. Resource rents relative to gross fixed capital formation, Malaysia 9.2. Per capita investment, Malaysia 12.1. Oil production and revenues 12.2. Budget deficits and estimated financial reserves, Saudi Arabia 15.1. 15.2. 15.3. 15.4. 18.1. 18.2. 18.3. 18.4. 19.1.

5 21 44 66 68 83 84 120 136 140 151 152 194

1983-1996

202

Ratios of GDI and CDS to GNP (%), 1970-1994 Education/GNP (%), 1970-1994 Extended genuine saving II/GNP (%), 1970-1994 Resource rents/GNP (%), 1970-1994 Nordic exports, 1960-1997 Nordic primary exports, 1963-1997 Primary exports and domestic investment Primary exports and economic growth, 1960-1997 Median annual growth in GDP per capita (const 1995 US$) resource-rich and resource-poor

250 251 252 253 300 301 304 305 316

List of Contributors Richard M . Auty, University of Lancaster, Lancaster, UK Alina A . Averchenkova, University of Bath, Bath, UK Gustavo Barboza, Universidad de Costa Rica, San Jose, Costa Rica Nancy Birdsall, Carnegie Endowment for International Peace, Washington, DC,

USA Jose A. Cordero, Universidad de Costa Rica, San Jose, Costa Rica Jose Luis Evia, Universidad Cato lica Boliviana, La Paz, Bolivia Ronald Findlay , Columbia University, New York, USA Alan Gelb , The World Bank, Washington, DC , USA Thorvaldur Gylfason, University of Iceland, Reykjavik, Iceland Kirk Hamilton , The World Bank, Washington, DC, USA Jonathan Isharm, Middlebury College, Middlebury, VT, USA Carolyn Jenkins, Centre for the Study of African Economies, University of Oxford, Oxford, UK Sampsa Kiiski, UNU/WIDER, Helsinki, Finland Mats Lundahl, Stockholm School of Economics, Stockholm, Sweden Zainal Abidin Mahani , University of Malaya, Kuala Lumpur, Malaysia Anil Markandya , University of Bath, Bath, UK S. Mansoob Murshed, UNU/WIDER, Helsinki, Finland Robert Osei, University of Nottingham, Nottingham, UK Thomas C . Pinckney, Williams College, Williamston, MA, USA Lant Pritchett, The World Bank, Washington, DC, USA Richard Sabot , Williams College, Williamstown, MA, USA Michael Woolcock , The World Bank, Washington, DC, USA

This page intentionally left blank

PART I

INTRODUCTION

This page intentionally left blank

1 Introduction and Overview RICHARD M . AUTY

1 . 1 . TH E U N D E R P E R F O R M A N C E O F TH E R E S O U R C E - A B U N D A N T ECONOMIE S In recent decades the resource-abundant developing countries have underperformed when compared with the resource-deficient developing countries (Ranis 1991; Lal and Myint 1996; Sachs and Warner 1995, 1999). Table 1.1 shows that between 1960 and 1990 the per capita incomes of the resource-poor countries grew at rates two to three times faster than those of the resource-abundant countries and that the gap in the growth rates widened significantly since the 1970s. Although there is reason to expect that crop-led growth in resource-abundant countries is inherently slower than the manufacturing-led growth of resource-deficient countries (Mellor 1995), the difference in the growth rates is greater than would be expected. Moreover, the mineraldriven resource-abundant countries have been among the weakest performers. Yet the mineral economies have the potential for rapid growth because most of them also have ample cropland (Table 1.1, column 3) so that mineral exports further enhance their capacity both to invest and to import compared with the non-mineral economies. The disappointing performance of the resource-abundant countries appears to be robust with regard to differences in the classification of the natural resource endowment. This is just as well because there is as yet no consensus on the measurement of resource abundance. Some studies have used single indicators, such as dependence on primary product exports (Sachs and Warner 1995), per capita land area (Wood and Berge 1997) and labour force in the primary sector (Gylfason et al. 1999); whereas others have used dual indices such as export orientation, and population size (Syrquin and Chenery 1989). Gylfason, Wood and Berge, and Sachs and Warner all conclude that there is little evidence that the basic findings are sensitive to the classification system used. Table 1.1 uses two criteria to measure the resource endowment, namely, per capita cropland and country size, measured in terms of absolute GDP.1 The helpful comments of Moshe Syrquin and Thorvaldur Gylfason are gratefully acknowledged. 1

Per capita cropland affords an index of land endowment and agricultural potential, whereas aggregate GDP gives an index of the scope for domestic manufacturing. The basic four-fold classification produces a very large group of small resource-rich countries, so it is further modified. The first distinction is between economies that are crop-driven and those that are mineral-driven (where the share of minerals exceeds 40 per cent

4

R. M. Auty Table 1.1 . Characteristics of six natural resource endowment categories

Resource Numbe endowment countrie

r of Croplan d GD P PCGD P PCGD P PCGD P s (ha/hd ) 197 0 ($b) 197 0 ($) Growt h Growt h 1960-90 1970-93 (%/Yr) (%/Yr)

Resource poor a,bb

Large Small Resource rich Large

7C 13d

0.15 0.16

21.05 1.94

196 343

3.5 2.5

3.7 2.1

10e

0.56

22.99

574

1.6

1.3

31f

0.57 0.66 0.44 0.48

1.41 1.23 2.01 5.67

250 304 831 362

1.1 0.8 1.7 1.6

0.7 -0.2 0.8 1.1

Small

Non-mineral Hard mineral Oil exporter All countries

16g

8h 85

a

Resource-poor = 1970 cropland/head < 0.3 hectares. Large = 1970 GDP > $7 billion. c Bangladesh, China, Colombia, Egypt, Indonesia, Philippines, and S. Korea. d El Salvador, Haiti, Hong Kong, Jordan, Kenya, Mauritania, Mauritius, Nepal, Singapore, Somalia, Sri Lanka, Taiwan, and Tanzania. c Argentina, Brazil, Chile, India, Mexico, Nigeria, Pakistan, South Africa, Turkey, and Venezuela. f Benin, Burundi, Cameroon, Chad, Costa Rica, Cote d'lvoire, Ethiopia, Fiji, Gambia, Ghana, Guatemala, Guyana, Honduras, Lesotho, Madagascar, Malawi, Malaysia, Mali, Morocco, Nicaragua, Panama, Paraguay, Rwanda, Senegal, Sudan, Swaziland, Thailand, Tunisia, Uganda, Uruguay, Zimbabwe. g Bolivia, Botswana, Burkina Faso, Central African Republic, Dominican Republic, Jamaica, Liberia, Namibia, Niger, PNG, Peru, Sierra Leone, Suriname, Togo, Zaire, Zambia. h Algeria, Congo, Ecuador, Gabon, Kuwait, Saudi Arabia, Syria, and Trinidad and Tobago. b

Source: Auty (1997), based upon UNCTAD data.

If the tendency for resource-abundant countries to underperform is insensitive to the classification criteria, this is not the case with regard to the time period observed. For example, Table 1.2 shows that the resource-abundant countries of new settlement (West European offshoots) as well as those of Latin America grew strongly at the end of the nineteenth century. Moreover, the average per capita income of the resourceabundant countries was substantially higher than that of the average resource-deficient countries in the early 1960s (Fig. 1.1). This makes it possible that at some time prior to this period countries that were rich in natural resources grew faster than those that were not. It is therefore not obvious when comparing two alternative growth trajectories, which one is to be preferred: one that gives a rapid initial boost to incomes and then slows and a second where income stagnates for a long period and then takes off at of total exports). This distinction is made because the principal economic linkage from minerals tends to be fiscal linkage, that is taxation, so that the governments of mineral economies must play a more important role in channelling the economic stimulus from primary exports than in the case of the governments of cropexporting countries. The second step in the subdivision is to break the mineral economy category into oil exporters and hard mineral exporters. This is a useful distinction because the rents tend to comprise a significantly higher fraction of oil revenues than is the case with most hard mineral revenues.

Introduction and Overview

5

Table 1.2. Real per capita GDP growth 1820-1992, by region

W Europe W Europe offshoots E Asia SE Asia S Asia Latin America SS Africa

Cropland in 1990 (ha/hd)

1820-1870

1870-1913

1913-1950

1950-1973

1973-1992

0.35 3.28

0.9 1.4

1.3 1.5

1.2 1.3

3.8 2.4

1.8 1.2

0.23 1.04 0.70 2.62 0.83

0.1 n.a. 0.1 n.a. n.a.

0.8 n.a 0.4 1.5 n.a

0.1 -0.1 -0.6 1.9 1.5

5.6 2.5 0.9 2.4 1.9

5.3 3.6 2.4 0.4 -1.2

Source: Maddison (1995), 62-3. WRI (1998).

Figure 1.1. Median GDP per capita (constant 1995 US$) resource-rich and resource-poor developing countries Source: World Bank (1999). a very fast rate. This is certainly the case if the decision about which trajectory to pursue must be taken in the initial year. Moreover, a handful of resource-abundant countries have sustained rapid economic growth in recent decades and any inherent tendency towards slower growth in the resource-abundant countries has been exacerbated by the fact that during this period such countries have proved more prone than resource-deficient ones to growth collapses (Lal and Myint 1996). Figure 1.1 compares the post-war growth trajectories of the resource-abundant and resource-poor countries since the 1950s. It shows that the growth per capita in the GDP of the resource-abundant countries collapsed after the oil shocks of the 1970s whereas the growth in per capita GDP of the resource-deficient (manufacturing-led)

6

R. M. Auty

countries actually accelerated. The deterioration among the resource-abundant countries is more severe where the natural resource rents emanate from 'point' resources, such as mining, rather than from 'diffuse' source resources like land under peasant farms (Table 1.1). Point rents are associated with staples that are relatively capital-intensive and thereby concentrate ownership. They include not only mines but also plantations where the crop requires immediate processing as in the case of sugarcane. In contrast, where the staple poses more modest investment barriers to entry, as with rice and maize, and some tree crops such as coffee and cocoa, the rents are likely to be more widely dispersed throughout the population. Table 1.1 also suggests that small resource-abundant countries are more vulnerable to growth collapses than large ones are. Large resource-abundant countries have two main advantages over small ones for avoiding a growth collapse. First, the probability of depending on one or two primary exports is lower for large countries compared with small ones. This is because, all else being equal, the large country has a more diversified natural resource endowment than a small one while it can also diversify earlier into manufacturing because its large market affords scale economies. Second, trade accounts for a smaller share of GDP in large economies so that they are more self-contained and less vulnerable to external shocks (Perkins and Syrquin 1989).

1.2. EXPLANATION S FO R TH E UNDERPERFORMANC E O F RESOURCE-ABUNDANT COUNTRIE S 1.2.1. Exogenous explanations of performance divergence A controversy broke out among development economists in the post-war period concerning the contribution of primary commodity exports to growth. Mainstream economists argued that primary commodity exports are the only way that countries in the early stages of development can generate the foreign exchange necessary to pay for essential imports and to service external debt. In addition, an expansion of trade attracts foreign investment and modern technology. Growth is therefore maximized by maintaining free internal and external markets for goods, and by allocating capital in free capital markets. This view was opposed by the structuralist economists who favoured planning and the nonmarket allocation of resources through a variety of controls and government incentives for prices, domestic investment, and foreign trade. The structuralist economists like Prebisch (1950) argued that a long-run decline in prices for primary exports is an inevitable result of the increasing use of synthetics, shrinking raw material content of finished products, and low elasticity of demand for raw materials. Also, oligopolistic markets in the industrial countries meant that productivity increases there were captured in the form of higher income by workers and owners, whereas in the developing countries productivity gains were passed on to (northern) consumers in the form of lower prices. Prebisch projected a downward trend in the terms of trade for primary commodities in relation to manufactured goods imported by the developing countries. Aware also of the dislocation inflicted on

Introduction and Overview

1

Argentina by the collapse of commodity prices in the 1930s, Prebisch therefore urged reduced dependence on primary commodity exports in favour of heavy investment in manufactures and infrastructure financed by capital imports. He advocated that the developing countries should temporarily close their economies in order to encourage infant manufacturing aimed initially at supplying their domestic markets. The structuralist approach tended to dominate development theory during the 1950-80 period, and its prescriptions strongly influenced economic policy. However, the Achilles' heel of the policy of infant industry protection is policy capture and the slow maturation of the infant industry. Meanwhile, the more successful resourceabundant countries diversified out of low-growth commodity exports into highgrowth commodity exports, whereas the less successful countries exhibited no such flexibility, mainly on account of policy failure (Duncan 1993). The phenomenon of Dutch disease provides a second exogenous explanation for the disappointing performance of the resource-abundant countries. Corden and Neary (1982) explain the effects with a three-sector model that has a resource sector, such as oil or other primary product exporting industry, a sector of other tradables, which would include both manufacturing and agriculture, and non-tradables. A boom in the resource sector has three separate effects: a spending effect; a relative price effect; and a resource movement effect. First, spending the increased export revenues increases the demand for both tradables and non-tradables, but spending on tradables fails to raise their domestic prices because in an open economy their prices are determined in international markets. Therefore, any excess demand is met by imports. Second, unless the increase in foreign exchange income is sterilized, there will be an appreciation of the currency, which reduces the domestic prices of exports and of imports competing with domestic output. A currency appreciation will also reduce the rents of the booming sector, but it may not be sufficient to reduce that sector's output. The domestic prices of the non-tradables will rise with the increase in demand, and these prices will not be affected by either the currency appreciation or competitive imports. Hence, there is a rise in the prices of non-tradables relative to the prices of tradables, with a consequent shift of resources from tradables to non-tradables, and a reduction in exports and increase in imports. Third, the movement of resources between sectors will also affect capital accumulation. If the non-tradable sector is relatively labour-intensive while the tradable sector is capital-intensive, the movement in favour of the non-tradable sector will tend to raise wages and lower returns to capital, thereby reducing capital accumulation. In addition, if manufacturing is favourable to growth and resource booms cause it to shrink, the resource-abundant economy could experience slower long-term growth than if it had no resources. Consistent with this risk, Matsuyama (1992) argues that manufacturing is characterized by learning-by-doing that is external to the enterprise but internal to the manufacturing sector as a whole. Elsewhere, Krugman (1987) identifies the conditions under which temporary resource boom can lead to an enduring loss of competitiveness. Rodriguez and Sachs (1999) show empirically how windfalls can lead to a temporary rise in consumption that is subsequently lost. However, proponents of the dominant role of Dutch disease like Sachs (1999), neglect

8

R. M. Auty

the fact that an export boom may not have harmful consequences if the increased primary export revenue is sustainable and/or the adjustment process is not too rapid. Moreover, Neary and van Wijnbergen (1986: 40-1) point out that some de-industrialization may be a symptom of the economy's adjustment to a new equilibrium rather than that of a disease. Export base theory offers a third exogenous explanation for differential growth, albeit one that focuses on the performance of resource-abundant countries with different natural resource endowments. Baldwin (1956) attributes divergent economic growth to differences in the staple production function. Drawing on examples from the United States in the nineteenth century, staples with capital-intensive production functions like sugar and cotton plantations (or mining) in Baldwin's 'South' region create linkages2 that are not favourable to growth. The bulk of the workforce comprises cheap labour and the resulting skewed distribution of income provides little market for locally produced consumer goods so that final demand linkage is stunted. Nor is fiscal linkage (government expenditure of tax revenues) stimulated because such production units find little incentive to pay taxes to improve the domestic socioeconomic infrastructure because each plantation and mine tends to function as a state within a state. Moreover, both backward and forward production linkages are limited because nonlabour inputs to the production process tend to be imported, while further processing of the staple tends to occur at overseas markets. The economy is therefore likely to remain specialized. In fact, firms may retard diversification by seeking to repress new competing activity during price downswings, so that the economy becomes locked in a staple trap. The staple trap defines a dependence upon one or a handful of commodity exports with declining viability and shrinking size relative to GDP. Such economies are especially vulnerable to external shocks and a growth collapse. Basically, the Baldwin 'South' model postulates that the capitalintensive production function of plantations and mines causes them to become economic enclaves with a limited stimulus to the local economy. In contrast, staples in Baldwin's 'West' region such as food grains or tree crops have more beneficial linkages due to their low barriers to entry and tendency to respond to modest investments with increasing productivity. These staples favour smaller farms that furnish incrementally rising incomes that stimulate domestic demand for locally produced basic consumer goods and also for producer goods in the form of simple farm equipment. The children of the farmers provide a stream of entrepreneurs to run such businesses and the farmers have an incentive to pay taxes in order to improve economic and social infrastructure. These favourable diffuse economic linkages cause the economy of the West region to diversify and escape the staple trap. 2 Linkage analysis is closely associated with Hirschman (1977) but has its roots in export base theory (Innis 1920; Baldwin 1956; Watkins 1963). Export base theory recognizes four categories of linkages. There are two sets of productive linkages that arise out of activity established to supply inputs to produce the staple (backward linkage) mad activity stimulated in the processing of the staple prior to export (forward linkage). The third is fiscal linkage, which is the taxation of activity connected with the staple. Lastly, final demand linkage is the multiplier arising from activity created by the local spending of wages and profits earned from the staple. An economy is said to have reached maturity when it has become sufficiently diversified so that the dominance of any staple is no longer evident.

Introduction and Overview

9

However, although export base theory can account for economic growth with primary products in some temperate regions, it is less successful when applied to the tropics. One important reason for this is that the difference between development in Baldwin's South and West regions is not wholly attributable to the production function. The outcome is sensitive to the political economy in which production takes place. For example, the diffusion of political power to the Caribbean workforce in the 1930s ushered in a period of rapid change that saw sharp increases in wages met by steep rises in productivity. Prior to that, the plantation owners had been able to use their political dominance to mobilize a stream of cheap labour from the more densely settled tropical regions of Asia (Lewis 1978) and this diminished the need to boost productivity. This implies that the rigidity of the tropical staples was due, in part, to the structure of society rather than simply to the technology of production (Auty 1995: 215-19), an outcome that is central to the thesis of this book.

1.2.2. Internal explanations of performance divergence The exogenous explanations of divergent performance tend to be deterministic and to neglect the role of domestic policies. In marked contrast, Lal and Myint (1996) identify policy failure as the prime cause of the underperformance of the resourceabundant countries. They find that economic policy strongly affects the efficiency of investment, which is the most important of three parameters determining the rate of economic growth. The other two parameters are the level of investment and growth in the quality-adjusted workforce. The most critical component of the policy environment is the trade regime, according to Lal and Myint (1996). They argue that whereas the 'mild dirigisrne' pursued by many developing countries through the first two postwar decades did not greatly damage their growth performance, the increasingly ambitious interventions thereafter proved less benign. Lal and Myint (1996: 214—15) find that within a sample of twenty-three countries, most land-abundant countries experienced a growth collapse whereas none of their labour-abundant (land-scarce) countries did. They attribute this to the fact that the land-abundant country relies on primary product exports longer and so must industrialize with higher wages than the resource-deficient country. This poses greater problems for the governments of land-abundant countries in terms of infrastructure coordination and the provision of requisite skills. Moreover, cuts in real wages may be required as the resource declines in relative importance (Sachs 1999) but they are usually strongly resisted, especially in the weaker 'factional' political states that have characterized most developing countries. This results in chronic fiscal deficits, a precursor to a growth collapse. In contrast, the development trajectory of the labourabundant countries is politically easier to manage. Lal and Myint (1996) provide a solid foundation upon which to build. Recent advances in the literature, notably in social capital and environmental economics, open additional avenues for analysis. Social capital research has shown how natural resource rents can be a source of conflict, corruption, and policy failure, especially in the case of the factional political states that are associated with heterogeneous societies

10

R. M. Auty

(Easterly and Levine 1997). Comparative country case studies reveal how the mechanisms used to capture and transfer natural resource rents contribute to policy failure (Krueger et al. 1991). Recent progress with environmental accounting illuminates this line of analysis by furnishing not only comparative measures of natural resource rent (World Bank 1998) but also a rationale for its deployment (Auty and Mikesell 1998). Domestic contests for natural resource rents have profoundly affected the resourceabundant countries, but so have global contests, notably the oil-exporting countries' efforts to boost the rents through collusion over oil production in the 1970s. These efforts inflicted external shocks upon oil-importers and oil-exporters alike that sternly tested the soundness of their policies. Chenery (1981) estimates that the oil shocks of 1973—78 and 1979—81 each transferred around 2 per cent of GWP from the oil-importers to the oil-exporters. Kaletsky (1986) argues that the release of this rent stream is analogous to a radioactive cloud. It inflicted severe damage on those countries over which it lurked, whether they be low-income oil-importers unable to expand exports to cover the higher costs of imported oil, mid-income oil-importers that borrowed petro-dollars to finance their trade gaps, or the oil-exporters themselves whose governments grossly overestimated the capacity of the domestic economy to absorb the oil windfalls. The damaging chain reaction triggered by the flow of petro-dollars underlines how exogenous variables interact with internal ones in explaining diverging economic performance. Yet the literature favours mono-causal explanations. Consequently, a synthesis of factors may provide more insight. This study draws upon the literature from growth theory, trade theory, export base theory, environmental economics, human capital, demography, social capital, institutional economics, and political economy. The compartmentalization of this literature has hitherto masked growing evidence of the links between natural resources and many other variables. This study fuses these emergent links into short-run and medium-term models to explain how natural resources affect economic development.

1.2.3. Widening the growth accounting framework The growth accounting framework of the neo-classical growth model tends to produce a sizable residual. For example, the World Bank (1993) study of the East Asian miracle concludes that the model explains barely one-third of the difference in economic performance between East Asia and Latin America. Despite efforts to elaborate the neo-classical theory (Mankiw 1995; Young 1995) and also the development of endogenous growth theory (Romer 1987, 1994), both these basic growth models share the problem of being narrowly specified (Sachs and Warner 1996). Therefore, this study extends the growth accounting framework to include natural capital and institutional capital alongside the traditional factors of produced capital (physical or man-made), education, and technological change. Table 1.3 shows recent World Bank (1997, 1998) estimates of the importance of natural capital relative to produced capital and human capital in generating economic growth. The relative importance of natural capital is negatively correlated with per

Introduction and Overview

11

Table 1.3. Per capita wealth, by major global regions 1994

Region

Total wealth ($ per capita)

Human resources (%)

Produced assets (%)

Natural capital (%)

North America Pacific OECD Western Europe Middle East South America North Africa Central America Caribbean East Asia East + Southern Africa West Africa South Asia

326,000 302,000 237,000 150,000 95,000 55,000 52,000 48,000 47,000 30,000 22,000 22,000

76 68 74 43 74 69 79 69 77 66 60 65

19 30 23 18 17 26 15 21 15 25 18 19

5 2 2 39 9 5 6 11 8 10 21 16

Source: World Bank (1998).

capita income so that the lower the income the more important natural capital becomes. In the case of low-income countries, natural wealth makes up roughly the same share of total wealth as produced assets do (around 20 per cent) according to the World Bank (1997). The central thesis of this study is that during the closing decades of the twentieth century, natural capital interacted with socio-political factors to play a central role in conditioning the efficiency with which both produced and human capital are deployed, and also in guiding the long-term development trajectory. Table 1.3 also suggests that the explanatory power of growth models can be further improved if, in addition to adding natural capital, the dominant human capital component is subdivided. The poorest countries have relatively high values for the share of human resources in total wealth because the wealth estimates are derived by dismembering GDP, so that people are basically all that they have. Further insight into the nature of human capital could be achieved if, in addition to the traditional concern for skills, social capital can be explicitly recognized as a second component. Social capital refers to society's social and political institutions. It revolves around trust, backed by sanctions that provides information and security to facilitate economic transactions by reducing risk (Knack and Keefer 1997; World Bank 1996: 81). One subcomponent is civic social capital, which functions at the level of the local community. It includes the building of trust in order to lower transaction costs, to improve the knowledge externalities of social networks and to enhance the capacity for collective action (Collier and Gunning 1999). The second subcomponent is public social capital that embraces those functions of the public sector of interest to institutional economists. These functions include the maintenance of property rights, government accountability, and the legal framework. Their effective provision is governed by the political economy. The study of this form of social capital is still in its infancy in economics, so the steps taken in this book to build it into the analysis must of necessity be rudimentary.

12

R. M. Auty 1.3. STRUCTURE OF THE BOOK

The book is divided into four principal parts. The next part, Part II, examines the relationship between the natural resource endowment and the four main types of capital, namely: produced, natural, human, and social. Each of the four chapters in Part II summarizes the stylized facts of the relationship between its particular form of capital and economic development and goes on to analyse the most salient features further. More specifically, Chapter 2 examines changes in the level of investment and structure of production with rising per capita GDP. The changing importance of the manufacturing sector is an especially important facet of structural change because that sector is capable of higher productivity growth than either agriculture or services. Manufacturing is also strongly associated with the spillover of positive externalities. Chapter 2 establishes that the size of the manufacturing sector varies systematically not only with per capita GDP but also with the size of the economy and the nature of the natural resource endowment. The chapter also examines the implications of recent evidence concerning the impact on saving rates and economic growth of the demographic cycle as it affects the dependency/worker ratio (Bloom and Williamson 1998). Chapter 3 addresses the contribution of natural capital to sustainable development. It concentrates on the depletion of nonrenewable natural capital. This is partly because the sustainability issue is of special significance to countries that exploit finite resources like minerals, and partly because less progress has been made in measuring the rents from cropland and pasture. Environmental accounting demonstrates the need to convert natural capital into other forms of capital in order to sustain long-term growth, should production of the natural resource decline in absolute or relative terms. Chapter 3 provides a measure of the relative importance of natural resource rents and shows how the genuine saving rate provides an index of sustainability. The genuine saving rate is defined as conventional saving minus the depreciation of produced capital, natural capital, and environmental capital and adjusted for net foreign saving and any accumulation of human capital. Birdsall et al. (1997) have shown that skill accumulates at a slower rate in resourceabundant countries than in resource-deficient ones. Chapter 4 compares the demand for education from intensive agriculture and competitive industrialization on the one hand, with that from the more extensive commodity production and import substitution industry favoured by many resource-abundant countries during the post-war decades on the other. It also analyses the relationship between natural resources, income distribution, skill acquisition, and the rate of economic growth. Finally, Part II concludes with an analysis of social capital in Chapter 5 which first reviews the literature in this fast-developing field. It then focuses upon how public social capital interacts with the natural resource endowment. Part III integrates the findings from Part II. It tests the degree to which the relationship between natural resources and development is a deterministic one by first exploring this relationship historically and then by constructing short-run and medium-term models of resource-abundant and resource-poor growth. Chapter 6

Introduction and Overview

13

explains the impact of natural resources on economic performance during the first golden age of economic growth (1870—1913). It focuses upon two subsets of resourceabundant countries, namely temperate-zone countries of recent West European settlement (see Table 1.2) and the tropics. Chapter 7 draws upon Sachs (1999) to develop a formal short-run three-sector model to compare the responses to price changes of resource-poor countries and two types of resource-abundant countries (with point and diffuse natural resource endowments). Chapters 6 and 7 both suggest that the relationship between natural resources and economic performance is not a deterministic one, so that policy and the political economy that drives it, counts. Chapter 8 then builds the political economy into two stylized facts models of resource-driven development, the competitive industrialization model and the staple trap model. The basic models are elaborated to allow for variations in the natural resource endowment and political economy. Part IV applies the analytical framework developed in the two preceding parts, to case studies of the development trajectories since the 1960s, of resource-abundant countries with different permutations of natural resources and political economy. It explores how the growth trajectory varies with differences in the initial conditions, including the socioeconomic linkages that different natural resource endowments generate. Chapter 9 examines Malaysia, a rare example of a successful resourceabundant country, which moreover, was characterized in the early post-independence years by acute social tensions and potentially unfavourable socioeconomic linkages. The chapter explains how successive Malaysian governments managed to both avoid the staple trap by a 'growth-based-security' policy and diversify the economy into competitive manufacturing. Chapters 10-12 examine three examples of growth collapses in Ghana, Bolivia, and Saudi Arabia, countries with different socioeconomic linkages from those of Malaysia. Ghana provides a classic case of a growth collapse due to policy failure. Chapter 10 establishes that Ghana had relatively favourable prospects at independence, including the diffuse socioeconomic linkages associated with peasant farm cash crops and a relatively undistorted policy environment. Nevertheless, the country experienced an extremely rapid collapse from which recovery has been protracted. Chapter 11 explains why Bolivia, a country with point socioeconomic linkages experienced a debilitating growth collapse despite substantial improvements in its terms of trade that provided an opportunity to consolidate the economic reforms that had been introduced in the early 1970s. Finally, Saudi Arabia provides an example in Chapter 12 of a capital-surplus oil exporter that maintained an open economy, accumulated large financial assets and, consistent with environmental accounting, attempted to convert its natural capital into economic infrastructure and human capita, but still failed to diversify its economy competitively. Finally, Chapter 13 looks at two large resource-abundant countries, Mexico and Argentina, that began the industrial transition later than the core countries of the North Atlantic, but had diversified more than most tropical countries (as represented by the four earlier case studies) had done by the close of the first golden age of economic development in 1913. Large size favours diversification and the evasion of a

14

R. M. Auty

growth collapse for the reasons given earlier, yet neither Mexico nor Argentina achieved its potential and both experienced a growth collapse. Part V distils the policy lessons for the successful reform of resource-abundant countries that have experienced a growth collapse. The legacy of a growth collapse includes a chronic fiscal deficit with inflationary consequences, severe economic distortion, and sizable foreign debt that presents acute servicing problems. The three principal policy objectives are, therefore, stabilization, trade and price liberalization, and debt rescheduling. The international financial institutions (IFIs) have dominated the debate in recent years. They assume that the rate of recovery is inversely related to the speed and depth of reform (de Melo et al. 1996). Incremental reform is also regarded as more susceptible to policy fatigue. The IFIs have therefore favoured a 'Big Bang' reform rather than more gradual change. Yet, the faltering recovery of many resource-abundant market economies such as Ghana and Bolivia, and the failure of many transition economies outside eastern Europe to match the predictions of the rapid reform model, suggest that initial conditions have been neglected (Popov 2000). For example, evidence is emerging that an abundant natural resource endowment may amplify the problems of market reforms rather than reduce them (Rosenberg and Saavalainen 1998). One reason for this is that a rich natural resource endowment may encourage corruption and slow the response to reform, thereby amplifying loss of output. In addition, the ability to implement reform is linked to the endowment of social capital or what the EBRD (1999) terms 'social cohesion'. Chapters 14 and 15 examine reform in two market economies with very different socioeconomic linkages, neither of which has succeeded in kindling rapid economic growth. Chapter 14 explains why Costa Rica failed to avoid a growth collapse and has achieved lacklustre post-reform economic growth, despite its potentially favourable socioeconomic linkages and long-standing policies of redistribution and investment in social capital. Chapter 15 analyses South Africa as a mineral economy whose point socioeconomic linkages have been heightened by counter-productive discriminatory policies towards human and social capital accumulation. It explains why the South African economy remains constrained despite more than two decades of reform. Chapters 16 and 17 address the transition economies. Chapter 16 analyses the performance of two smaller transition economies, relatively fast-reforming Kazakhstan, a country with large mineral reserves and Uzbekistan, a slow-reformer which nevertheless outperformed Kazakhstan through the 1990s. The chapter suggests ways in which the rapid reform model might be improved. Chapter 17 examines Russia, a large mineral economy that shares many adverse features of the less successful oil-rich market economies, including inadequate attention to the institutions required in order to rationalize the deployment of the mineral rents. Finally, Chapter 18 examines the policy lessons for resource-driven growth in more prosperous countries with reference to the experience of five small Nordic countries with resource-driven growth. It notes how Norway has managed its sizable oil windfall. Chapter 19 summarizes the principal findings of the study and highlights the main policy implications.

Introduction and Overview

15

REFERENCES Auty, R. M. (1995), Patterns of Development: Resources, Policy and Economic Growth, London: Edward Arnold. Auty, R. M. (1997), 'Natural resources, the state and development strategy', Journal of International Development, 9, 651—63. Auty, R. M. and R. F. Mikesell (1998), Sustainable Development in Mineral Economies, Oxford: Clarendon Press. Baldwin, R. E. (1956), 'Patterns of settlement in newly settled regions', Manchester School of Social and Economic Studies, 24, 161-79. Birdsall, N., D. Ross and R. Sabot (1997), 'Education, growth and inequality', in N. Birdsall and F. Jasperson (eds.), Pathways to Growth: Comparing East Asia and Latin America, Washington DC: Inter-American Development Bank, 93-127. Bloom, D. E. and J. G. Williamson (1998), 'Demographic transitions and economic miracles in emerging Asia', The World Bank Economic Review, 12, 419-55. Chenery, H. (1981), 'Restructuring the world economy', Foreign Affairs, 59, 1102-20. Collier, P. and J. W. Gunning (1999), 'Explaining African economic performance', Journal of Economic Literature, 37, 64—111. Corden, M. and J. P. Neary (1982), 'Booming Sector and Dutch Disease Economics: A Survey', Economic Journal, 92, 826-44. de Melo, M., C. Denzier and A. Gelb (1996), 'Patterns of transition from plan to market', World Bank Economic Review, 10, 397-424. Duncan, R. C. (1993), 'Agricultural export prospects for sub-Saharan Africa', Development Policy Review, 11, 31-45. Easterly, W. and R. Levine (1997), 'Africa's growth tragedy: policies and ethnic divisions', Quarterly Journal of Economics, 112, 1203-50. EBRD (1999), Transition Report 1999: Ten Years of Transition, London: European Bank for Reconstruction and Development. Gylfason, T., T. T. Herbertsson and G. Zoega (1999), 'A mixed blessing: natural resources and economic growth', Macroeconomic Dynamics, 3, 204-25. Hirschman, A. O. (1997), 'A generalized linkage approach to development, with special reference to staples', in M. Nash (ed.), Essays on Economic Development in Honor of Bert F. Hoselitz, Chicago IL: Chicago University Press, 67-98. Innis, H. (1920), The Fur Trade in Canada, New Haven, CT: Yale University Press. Kaletsky, A. (1986), 'Oil and the world economy: a free lunch at OPEC's expense', Financial Times (7 February, Section E: 16, London edition). Knack, S. and P. Keefer (1997), 'Does social capital have an economic pay-off? A cross-country investigation', Quarterly Journal of Economics, 112, 1250-88. Krueger, A. O., M. Schiff and A. Valdes (1991), Political Economy of Agricultural Pricing Policies, Baltimore MD: Johns Hopkins University Press. Krugman, P. R. (1987), 'The narrow band, the Dutch disease and the competitive consequences of Mrs Thatcher', Journal of Development Economics, 27, 41-55. Lal, D. and H. Myint (1996), The Political Economy of Poverty, Equity and Growth, Oxford: Clarendon Press. Lewis, W. A. (1978), Growth and Fluctuations, London: George Allen and Unwin. Maddison, A. (1995), Monitoring the World Economy 1820-92, Paris: OECD. Mankiw, N. G. (1995), 'The growth of nations', Brookings Papers on Economic Activity, 275-326.

16

R. M. Auty

Matsuyama, K. (1992), 'Intercultural productivity, comparative advantage, and economic growth', Journal of Economic Theory, 58, 317-34. Mellor, J. W. (1995), Agriculture on the Road to Industrialization, Baltimore MD: Johns Hopkins University Press. Neary, P. J. and S. N. Van Wijnbergen (1986), Natural Resources and the Macro Economy, Cambridge MA: MIT Press. Perkins, D. H. and M. Syrquin (1989), 'Large countries: The influence of size', in H. Chenery and T. N. Srinivasan (eds.), Handbook of Development Economics. Volume 2, Amsterdam: North-Holland, 1691-753. Popov, V. (2000), 'Shock therapy versus gradualism', Comparative Economic Studies, 42, 1—57. Prebisch, R. (1950), The Economic Development of Latin America and its Principal Problems, Santiago: ECLA/UN. Reprinted in: Economic Bulletin for Latin America, 7, 1962, 1—22. Ranis, G. (1991), 'Towards a model of development', in L. B. Krause and K. Kihwan (eds.), Liberalization in the Process of Economic Development, Berkeley, CA: University of California Press, 59-101. Rodriguez, F. and J. D. Sachs (1999), 'Why do resource abundant economies grow more slowly? A new explanation and an application to Venezuela', Journal of Economic Growth, 4, 277-303. Romer, P. M. (1987), 'Growth based on increasing returns due to specialization', American Economic Review: Papers and Proceedings, 77, 56-62. Romer, P. M. (1994), 'The origins of endogenous growth', Journal of Economic Perspectives, 8, 3-22. Rosenberg, C. B. and T. O. Saavalainen (1998), 'How to deal with Azerbaijan's boom? Policy strategies in a resource-rich transition economy', IMF Working Paper, WP/98/6, Washington DC: IMF. Sachs, J. D. (1999), 'Resource endowments and the real exchange rate: A comparison of Latin America and East Asia', in T. Ito and A. O. Krueger (eds.), Changes in Exchange Rates in Rapidly Developing Countries, Chicago 111: University of Chicago Press, 133—53. Sachs, J. D. and A. M. Warner (1995), Economic Convergence and Economic Policy, NBER Working Paper, 5039, Cambridge, MA: NBER. Sachs, J. D. and A. Warner (1996), 'Fundamental sources of long-run growth', American Economic Association Papers and Proceedings, May, 184-8. Sachs, J. D. and A. Warner (1999), 'Natural resource intensity and economic growth', in J. Meyer, B. Chambers and A. Farooq (eds.), Development Policies in Natural Resource Economies, Cheltenham: Edward Elgar, 13-38. Syrquin, M. and H. B. Chenery (1989), 'Patterns of Development 1950 to 1983', World Bank Discussion Paper, 41, Washington DC: World Bank. Watkins, M. (1963), 'A staple theory of economic growth', The Canadian Journal of Economics and Political Science, 29, 141-58. Wood, A. and K. Berge (1997), 'Exporting manufactures: human resources, natural resources and trade policy', Journal of Development Studies, 34, 35-59. World Bank (1993), The East Asian Miracle, Oxford: Oxford University Press. World Bank (1996), World Development Report 1996, Oxford: Oxford University Press. World Bank (1997), Expanding the Measure of Wealth, Environmentally Sustainable Development Studies and Monograph Series 17, Washington DC: World Bank. World Bank (1998), 'Estimating national wealth: methodology and results', Environment Department Papers, 57, Washington DC: World Bank. World Bank (1999), World Development Indicators 1999, Washington DC: World Bank. WRI (1998), World Resources 1998-99, New York: Oxford University Press. Young, A. (1995), 'The tyranny of numbers: confronting the statistical realities of the East Asian growth experience', Quarterly Journal of Economics, 110, 641-80.

PART I I

CRITICAL PARAMETERS IN RESOURCE-BASED DEVELOPMENT MODELS

This page intentionally left blank

2 Natural Resources, Capital Accumulation, Structural Change, and Welfare R I C H A R D M. AUTY AND SAMPSA K I I S K I

2.1. OVERVIE W

This chapter draws upon the literature on economic growth and trade, along with the World Bank (1999) database, in order to establish the stylized facts regarding capital accumulation, structural change, and social welfare1 in developing countries with different natural resource endowments. It begins by establishing the basic trends in these parameters with rising per capita income for all developing countries in the next section, Section 2.2. The principal differences between the resource-abundant countries and the resource-poor countries are then set out in Section 2.3. It is shown that through the post-war years the resource-poor countries tended to experience an earlier expansion of competitive manufacturing than resource-abundant countries. This development trajectory of competitive industrialization contrasts with the staple trap trajectory of the resource-abundant countries as a group. The second half of the chapter concentrates upon differences among the resourceabundant countries. It first compares large and small countries in Section 2.4 and establishes that large countries pursued remarkably high levels of autarky (selfsufficiency). Section 2.5 analyses the smaller resource-abundant countries. It may be recalled from Table 1.1 that the smaller resource-abundant countries are much more numerous than the larger ones. They are therefore subdivided into three groups on the basis of the contrasting socioeconomic linkages of their leading primary product exports. Per capita GDP growth decelerates for all three export staples and turns negative for the mineral economies after the 1973-85 period of oil shocks. Section 2.6 summarizes the findings and links them to later chapters.

2.2. TH E STYLIZE D FACT S O F PE R CAPIT A INCOME AND STRUCTURA L CHANGE 2.2.1. Structural change This section establishes the stylized facts concerning the relationship between rising per capita income and changes in the investment rate, structure of production, The constructive criticism of Moshe Syrquin and Thorvaldur Gylfason is gratefully acknowledged. So1cial welfare is measured through income distribution, demographic change, and per capita income.

20

R. M. Auty and S. Kiiski

economic growth, and income distribution. It draws heavily on the exhaustive studies by Syrquin and Chenery (1975, 1989) that are based on data for more than one hundred countries and cover the period from 1950 to 1983. Syrquin and Chenery use a simpler classification than that in Table 1.1: the staple categorization is based on dependence on primary product exports alone. They recognize four basic categories (large and small primary-oriented and large and small manufacturing-oriented). They also measure country size in terms of population (with a 20 million threshold in 1965) rather than GDP. There are therefore some differences between the two typologies, but for the trends discussed here such differences should not matter. Figure 2.1 summarizes the pattern of structural change for the countries analysed by Syrquin and Chenery as their per capita income rises from below US$300 to more than $4,000, measured in 1980 dollars.2 It shows a steady decline in the initially dominant role of agriculture as the manufacturing sector rises in importance before it too begins to decline. More specifically, the share of agriculture in GDP declines from almost one-half to less than one-fifth by the time the per capita income reaches $1,500, at which level manufacturing overtakes it (Table 2.1). Within manufacturing, heavy and chemical industry (HCI) outstrips light industry in importance at that income level. Thereafter, according to more recent data for the mature economies (Brown and Julius 1993), further increases in income are associated with a contraction of agriculture to 2 per cent of GDP or less, while manufacturing falls below 20 per cent of GDP. The corollary is the dominance of the economy by services, which generate more than two-thirds of output, a trend that Brown and Julius (1993) expect to be intensified. The changing structure of production implies changes in the rate of economic growth because agriculture, manufacturing, and services have differing capacities to drive economic growth. Echevarria (1997) uses a dynamic general equilibrium model with which to explore the impact of the changing structure of production on GDP growth. She reports that productivity growth differs by sector, with manufacturing displaying the fastest rate of growth, and services the slowest rate.3 Echevarria estimates that changes in the sectoral composition of GDP explain more than one-fifth of the observed variation in economic growth among countries. The literature generally assumes that the benefits of manufacturing for economic growth include strong externalities associated with learning by doing (Matsuyama 1992; Sachs and Warner 1997). For example, Wood and Berge (1997) speculate that countries that export manufactured goods grow faster than countries that export primary products because manufacturing makes strong demands on skills and so accelerates human capital accumulation. Importantly, Chapter 4 finds that human capital accumulation is slower where manufacturing is protected than when a strategy of competitive industrialization is pursued. This implies that the choice of trade policy

Sinc2e 19802 the US GDP deflator has risen by 1.92, so the income range rises from less than $585 to more than $9,600 in 1998 dollars. 3 Although services may confer significant positive spillover effects on manufacturing.

Natural Resources and Welfarree

21

Figure 2.1 . Structure of value added in GDP, by country resource endowment Sour,ce: Syrquin and Chenery 1989: 56-7.

affects the economic stimulus from manufacturing, a theme that is explored more fully in Sections 2.3-2.5 below. Although the manufacturing sector has the potential to grow at least twice as fast as agriculture, the latter sector can still make a strong contribution to economic development if barriers to the productive deployment of resources are removed. Timmer (1988) uses stylized facts from empirical studies to break the contribution of

22

R. M. Auty and S. Kiiski Table 2.1. Per capita income, structural change, and domestic absorption, post-1973 (share of GDP) Income pe r capita (1980 US $ <300 30

Production Agriculture Mining Manufacturing Construction Utilities Services Final demand Private consumption Government consumption Investment Exports Imports

0 50

0 1,00

0 2,00

0 4,00

0>

5,000

0.48 0.01 0.10 0.04 0.06 0.31

0.39 0.05 0.12 0.04 0.07 0.32

0.32 0.07 0.15 0.05 0.07 0.35

0.23 0.08 0.18 0.06 0.08 0.38

0.15 0.08 0.21 0.06 0.09 0.41

0.10 0.06 0.24 0.07 0.09 0.45

0.07 0.01 0.28 0.07 0.10 0.47

0.79 0.12 0.14 0.16 0.21

0.73 0.14 0.18 0.19 0.25

0.70 0.14 0.21 0.21 0.25

0.66 0.14 0.23 0.23 0.26

0.63 0.14 0.25 0.23 0.27

0.60 0.15 0.26 0.26 0.28

0.60 0.14 0.26 0.23 0.23

a

Since 1980 the US GDP deflator has risen by 1.92, so the income range rises from less than $575 to more than $9,600 in 1998 dollars. Source: Syrquin and Chenery 1989: 20.

agriculture into four stages, of which the first two are of most interest here. The first stage is a period of agricultural enablement that might include land reforms, infrastructure upgrading, credit provision, and the correction of market failure. The second stage brings a strong positive contribution from agriculture to the economy through the provision of cheaper food and also a supply of cheap labour to the emerging urban sector. Agriculture also generates foreign exchange, tax revenues and inputs for manufacturing (Johnston and Mellor 1961), but it performs two additional functions that were neglected by policymakers until the 1980s. First, the sector is a market for locally manufactured farm inputs and household goods. Second, agriculture is an important absorber of labour, while nonfarm rural activity helps to promote the more efficient absorption of labour within the sector. Agriculture can therefore play a useful additional role by helping to curb rural poverty, especially if the transport network fosters rural trade and rural nonfarm jobs. However, the potential dynamism of the agricultural sector, like that of manufacturing noted earlier, is sensitive to policy. For example, Mellor (1995) draws on extensive empirical evidence to show that agriculture may expand in Timmer's most dynamic second stage at 4—6 per cent annually in a well-managed economy. He estimates from Asian experience that, even with such a modest pace of growth, a well-managed agricultural sector can propel total GDP at a rate of 7.5 per cent annually in the early stages of development. This growth potential was rarely achieved after the 1960s, however. This implies that misguided policies, rather than inherent structural weakness, explain slow crop-driven growth.

ural

resources and welfarrtlfsree

323

Table 2.2. Per capita GDP and economic growth, Syrquin /Chenery aggregate? PC GDP Net ICO R Populatio n P C GDP Valu e added Labou r Lengt h (US$ 1980 b ) investmen t growt h growt h growt h productivit y (Years ) (% GDP) (%/Year) (%/Year) (%/Year) growt h 100-140 140-280 280-560 560-1120 1120-2100 2100-3360 3360-5040

8 10 13 15 16 16 14

2.2 2.3 2.4 2.5 2.4 2.4 2.4

2.6 2.8 2.5 2.2 2.0 1.5 1.0

1.3 2.0 3.2 4.1 4.6 4.7 4.6

3.8 4.8 5.7 6.3 6.6 6.2 5.6

1.3 1.7 2.8 4.0 4.8 4.8 4.1

27 35 22 17 14 10 9

The numbers are synthetic ones designed to represent a likely pattern over the very long run (personal communication, M. Syrquin). See footnote a in Table 2.1.

b

Source: Syrquin 1986: 233.

2.2.2. Investment, saving, and the worker dependency ratio The different growth potential of agriculture and manufacturing, when linked to the aggregate pattern of structural change, partly explains why the rate of economic growth initially accelerates and then declines with rising per capita income. Table 2.2 suggests that the other factors at work include the changing rate of investment relative to GDP (column 3) and decelerating population growth (column 4). Table 2.2 is based on extrapolation from the Syrquin and Chenery data. It suggests that investment, net of depreciation, might be expected to double its share of GDP with rising per capita income and then begin to fall. This implies a rise in gross fixed investment to 20-25 per cent of GDP.4 Bloom and Williamson (1998) argue that the rate of investment may also be affected by the demographic transition as well as by structural change. Their case is based upon an analysis of the changing ratio between dependants and workers during the demographic cycle. This relationship has been neglected in the literature on economic growth, which has focused upon the overall rate of population increase. Bloom and Williamson argue that the dependency ratio traces three distinct phases with rising per capita income, with significant effects on the rate of saving and investment. First, a rise occurs in the number of dependants per active worker that lags the population cycle by about one generation. This retards economic growth because it raises consumption at the expense of saving and investment. Second, economic growth then accelerates as the birth rate declines and a surge in active workers improves the dependency ratio and boosts saving. Finally, the demographic cycle resumes a growth-depressing effect as it causes the number of elderly dependants to rise. Bloom and Williamson (1998) cite five reasons why the rate of per capita GDP growth accelerates during the middle stage of the dependency cycle. First, the saving Thine consequences of considering the depleting of natural and environmental capital are considered in the next chapter, Chapter 3.

Table 2.3. Investment, GDP growth and investment efficiency, six natural resource endowment categories 1960-1997 Resource endowment category

Investment GDP ICOR PC GDP (%GDP) growth growth (%/Year) (%/Year) (1) (2) (3) (4)

Population Number of growth countries (%/Year) (5) (6)

14.8 20.5 21.9

4.2 3.4 3.5

3.5 6.9 6.0

1.6 0.7 0.9

2.6 2.7 2.6

24" 29b 27C

24.5 31.0 23.9

6.6 6.5 1.9

3.7 5.7 12.4

4.0 2.3 -0.7

2.6 4.2 2.6

8d 8d 8d

17.5 21.8 17.1

4.9 3.0 2.3

5.7 7.3 7.5

2.2 0.1 -0.4

2.7 2.9 2.6

10e 10e 10e

20.3 21.8 20.1

5.4 3.1 4.0

4.0 7.1 5.0

2.7 0.7 1.9

2.7 2.4 2.1

8f 10 10

18.8 24.8 23.0

6.1 4.0 4.4

3.2 6.2 5.2

3.5 1.8 2.4

2.6 2.2 2.0

8g 8h 8h

17.7 25.5 26.3

5.0 5.8 6.0

4.2 4.4 4.4

2.4 3.7 4.7

2.6 2.1 1.3

7 7 7

Small non-mineral resource-rich

1960-73 1973-85 1985-97 Small oil-exporting resource-rich

1960-73 1973-85 1985-97 Small ore-exporting resource-rich

1960-73 1973-85 1985-97 Large resource-rich 1960-73 1973-85 1985-97 Small resource-poor 1960-73 1973-85 1985-97 Large resource-poor

1960-73 1973-85 1985-97

Note 1: Countries from classification of Auty (1997a), hut due to data deficiency it excludes: a Ethiopia, Gambia, Morocco, Panama, Swaziland, Tunisia, Uganda. bEthiopia and Uganda. c Ethiopia, Guyana, Nicaragua, Panama, Uganda. d Kuwait. e Bolivia, Liberia, Namibia, Peru, Sierra Leone, Suriname. India and Turkey. g Haiti, Jordan, Somalia, Taiwan, Tanzania. Jordan, Somalia, Taiwan, Tanzania. Note 2: Data points are missing for some years for some countries, but they tend to average out so that this does not inject an intolerable bias. Some discrepancy also arises from the use of constant 1995 US dollars because the more the inflation rate the greater the degree of the uncertainty that is added. Although there are large differences between countries in the starting level numbers, the trends within groups remain similar when these differences are allowed for, with reduced samples or country-specific fixed effects. The comparative analysis in this table between the different points and a fixed effects continuous growth regression both yield similar results for the growth differences between resource endowment categories. Overall, therefore, the figures lack some precision, but the magnitude of the changes in the growth rates is large enough to be robust in the face of the statistical discrepancies. Source: World Bank (1999).

Natural Resources and Welfare

25

rate increases as the level of consumption that each worker must support falls. Second, the economy-wide input of labour per capita rises. Third, the rising labour participation rate triggers an investment boom to meet the employment and housing needs of the new workers and their families. Fourth, the faster rate of economic growth brings higher tax revenues and raises public saving by reducing transfers. Finally, as the scale of transfers relative to the public finances falls, so too do distortionary taxes. If Bloom and Williamson are correct, then countries that embark upon the demographic transition earlier can significantly raise their rate of economic growth. In fact, resource-poor countries are likely to embark earlier on the demographic cycle than resource-abundant countries because they industrialize at a lower per capita income and this brings earlier urbanization that, as explained by the demographic transition model, advances the slowdown in population growth (column 5, Table 2.3). Welfare depends upon the distribution of income as well as the rate of per capita GDP growth. Kuznets (1955) suggests that income inequality is related to rising per capita income and that it describes an inverted U-shape in which it is first amplified but then declines at higher income levels. The initial rise in inequality is presumed to result from a shift of the workforce out of a relatively egalitarian agriculture and into sectors that are characterized by greater income differentials. Thereafter, the narrowing in income differences in the high-income countries reflects the converging inter-sectoral productivity of the workforce; a decline in the share of property income in household income; and an enhanced preference by governments for social security and full employment (Kindleberger and Herrick 1977). Early studies by Williamson (1965) and Adelman and Morris (1973) appeared to confirm, the Kuznets curve. However, these studies are based upon cross-sectional data rather than inter-temporal data for individual countries. Recent studies based on inter-temporal data (Fields 1989; Bowman 1997) reject the Kuznets curve, as do some recent cross-sectional studies (Anand and Kanbur 1993; Deininger and Squire 1996). Fields (1989) draws upon data for twenty-two countries that yield a total of seventy 'growth spells' and finds that inequality is about as likely to increase with rising income levels as it is to decrease. This is so whether the unit of observation is the set of countries or the set of growth spells. Moreover, income inequality is as likely to increase with growth, as often in low-income countries as in high-income countries. A more significant finding by Fields for this study, however, is that income inequality is associated with natural resource abundance. In fact the original Kuznets curve owes its shape to the fact that when it was constructed the mid-income countries were dominated by Latin America, a resource-abundant region with high-income inequality.

2.3. COMPARIN G RESOURCE-POO R AN D RESOURCE-ABUNDANT DEVELOPMEN T 2.3.1. Contrasts in structural change, investment, and welfare Syrquin and Chenery (1989: 58-9) demonstrate that during the post-war decades the evolving manufacturing sector has accounted for a greater share of GDP in large

26

R. M. Auty and S. Kiiski

countries than in small ones. Extrapolating an income-related trend from their data, the manufacturing sector of the large resource-/xw countries expanded earlier than that of the large resource-abundant countries and peaked at a level that is 4 per cent of GDP higher (Fig. 2.1). Within the subset of small countries, manufacturing expanded its share of GDP in the resource-poor countries by 2-3 per cent more than in the case of the small resource-abundant countries. Consequently, all else being equal, the pattern of structural change in the resource-abundant countries, especially the smaller crop-driven ones, confers an inherent tendency towards slower growth. In fact, the magnitude of the underperformance of the resource-abundant countries is greater than would be produced by structural change alone, because all else was not equal and the level and efficiency of investment deteriorated more in the resource-abundant countries through the 1970s and 1980s (Table 2.3, column 3). Gylfason (1999) shows that the level of domestic investment is inversely related to dependence on primary product exports, although there is significant variation about the mean. Moreover, Table 2.3 shows that whereas the more successful resourcedeficient countries raised their gross fixed investment above 25 per cent of GDP and sustained the efficiency of that investment (after allowing for some slippage due to a shift towards more capital-intensive investment), very few resource-abundant countries achieved this double feat. For example, Edwards (1997) reports that the Latin American countries sustained an investment rate of barely 20 per cent of GDP 196080, and even after adding in foreign saving, the level of investment remained at 23 per cent of GDP. More importantly for economic growth, their investment efficiency collapsed in the 1970s, a characteristic of resource-abundant countries generally (Table 2.3). As noted earlier, the demographic cycle plays a role in differentiating GDP growth rates and Bloom and Williamson (1998) estimate that compression of the demographic transition may account for up to half of the above trend growth that the East Asian countries experienced after the 1960s. They trace the onset of the benevolent middle stage of the demographic transition to 1970 for East Asia as a whole, but the resourcedeficient countries of Northeast Asia commenced that stage even earlier (compare the dependency ratios of South Korea and Malaysia in Table 2.4). The welfare gains from early industrialization are compatible with the Boserupian thesis of the beneficial stimulus from population pressure on natural resources. It may be recalled that Table 1.2 shows that the per capita cropland ratio of the resource-poor East Asian countries was already low in 1900, and by the early post-war years per capita cropland was lower still, at 0.1 hectare or less in China, Hong Kong, South Korea, Singapore, and Taiwan (Auty 1997a) compared with more than three times that level in Bangladesh. Thereafter, the faster rates of industrialization, urbanization, and income growth in the resource-poor East Asian countries helped to advance the deceleration in population growth. The changing dependency ratio carries implications for income distribution as well as for investment and economic growth. In post-war Asia, income inequality has been higher in the resource-abundant countries of Southeast Asia than in the resource-poor Northeast Asian countries. For example, the Gini coefficients of South Korea and

Natural Resources and Welfaree

27

Table 2.4. Trends in the dependants/worker ratio, selected countries 1960-1995 Country

1960

1965

1970

1975

1980

1985

1990

1995

Argentina Bangladesh Bolivia Costa Rica Finland Ghana Kazakhstan Malaysia Mexico Norway Russia Saudi Arabia South Africa South Korea Uzbekistan

0.57 0.81 0.85 1.02 0.60 0.93 0.72 0.95 0.98 0.59 0.57 0.87 0.81 0.83 0.79

0.57 0.89 0.86 1.04 0.54 0.94 0.79 0.98 1.02 0.58 0.58 0.90 0.84 0.87 1.00

0.57 0.96 0.87 0.97 0.51 0.93 0.75 0.92 1.03 0.60 0.52 0.91 0.81 0.83 1.04

0.58 0.98 0.87 0.84 0.49 0.93 0.68 0.85 1.02 0.60 0.47 0.90 0.81 0.71 0.96

0.63 0.98 0.85 0.74 0.48 0.91 n.a. 0.74 0.96 0.59 0.47 0.89 0.79 0.61 0.85

0.65 0.92 0.84 0.69 0.47 0.93 n.a. 0.74 0.85 0.56 0.48 0.84 0.79 0.52 0.81

0.65 0.95 0.81 0.69 0.49 0.92, 0.60 0.72 0.74 0.54 0.49 0.81 0.74 0.45 0.82

0.62 0.88 0.80 0.65 0.50 0.91 0.58 0.66 0.66 0.55 0.50 0.79 0.66 0.41 0.79

Source: World Bank (1999).

Taiwan were below 0.36 compared with more than 0.46 for Malaysia and Thailand (World Bank 1997, 1999). The respective per capita cropland endowments in 1991 for the contrasting pairs of countries was around 0.05 hectares compared with more than 0.27 (WRI 1994). The Philippines occupies an intermediate position, with a Gini coefficient of 0.43 and a per capita cropland endowment of 0.13 hectares. The equitable distribution of income in resource-poor countries also reflects the early redistribution of land and the rapid elimination of surplus labour as labourintensive manufactured exports boomed (Auty 1997b) and removed the drag of low rural wages. The subsequent rapid accumulation of human capital (described in Chapter 4) narrowed the wage premium for skilled labour (Londono 1996). In contrast, land reform lags in resource-abundant countries while later urbanization delays the elimination of surplus labour and retards skill accumulation (see Chapter 4) so that cheap labour depresses unskilled wages compared with the modern urban sector. A further cause of higher income inequality in the resource-abundant countries is a more capital-intensive industrialization path. The World Bank (1993) shows that the resource-abundant countries of Southeast Asia, like their counterparts elsewhere, leap-frogged the labour-intensive manufacturing stage that helped to reduce income inequality in the resource-deficient East Asian countries.

2.3.2. The role of economic policy A significant feature of the resource-abundant countries is that the efficiency of their investment collapsed through the 1970s (Lal and Myint 1996). This suggests that systematic differences in economic policy may be at work. In the post-war debate on the role of government, mainstream economists saw it as creating an enabling

28

R. M. Auty and S. Kiiski

environment within which resources could be effectively deployed in line with a country's evolving comparative advantage. However, structuralists advocated a more active role for the state in the promotion of structural change. The structuralist views dominated through the 1960s and 1970s so that many countries closed their trade policy and/or relaxed market discipline. Sachs and Warner (1995) examine the relationship between trade policy and natural resource dependence for the period 1970-89. They find a U-shaped relationship so that, as resource dependence increases, trade policy first closes but then opens again with increasing dependence on primary product exports. The trough of the curve occurs where primary exports reach 33 per cent of GDP, with most developing countries below this level. The left arm of the U-shaped curve implies that resourcepoor countries are less prone to policy closure (for reasons that are developed at length in Chapter 8, which deals with the political economy of resource-driven growth). Sachs and Warner attribute the trough of the U-shape to fear of Dutch disease by the governments of most resource-abundant countries. That fear engenders protectionist policies in order to sustain the fledgling manufacturing sector. Finally, Sachs and Warner explain the right arm of the U-shape by the dominance of this section of the curve by oil-exporters whose large oil reserves lower the incentive to diversify out of dependence on the depleting mineral asset. The closure of trade policy diminishes resilience to external shocks (Balassa 1985; Sachs 1985). However, oil-exporting countries like Saudi Arabia that had relatively open trade policies also experienced a growth collapse (Table 2.3). Chapter 12 demonstrates that the critical condition for a growth collapse is not a closed trade policy but rather the relaxation of market discipline and associated accumulation of economic distortions. Lax policy retards competitive diversification and lies at the heart of the underperformance of the resource-abundant countries. 2.4. R E S O U R C E - A B U N D A N T DEVELOPMEN T I N L A R G E COUNTRIE S The successful resource-poor countries provide a counter-factual to the growth trajectory of the resource-abundant countries. Previous sections have established that resource-poor countries pursue a virtuous economic circle: they embark on manufacturing-driven growth earlier, pass through the demographic cycle faster and adopt more open trade policies. Table 2.3 shows that since the 1960s they expanded the share of investment in GDP faster and maintained higher investment efficiency. The result was the competitive diversification of the economy that proved resilient to the trade shocks of the 1970s and 1980s and sustained rapid per capita income growth with low levels of income inequality. The rest of the chapter compares the virtuous development trajectory of competitive industrialization with the trajectories of the principal subgroups of resource-abundant countries, beginning with the larger countries in this section. The large resource-abundant countries have two advantages over the small ones for sustaining economic growth. First, the probability of depending on one or two primary

Natural Resources and Welfaree

292

exports is lower for large countries than for small ones. This is because the large country has a higher probability of embracing a more diversified natural resource endowment and it can also diversify earlier into manufacturing because its large market affords scale economies. Second, trade accounts for a smaller share of GDP in large economies so that they are more self-contained and therefore less vulnerable to external shocks. The Syrquin and Chenery (1989) data suggest that the share of agriculture in GDP declined more slowly in the large resource-abundant countries than in any other endowment category. The corollary is that manufacturing expanded its share of GDP more slowly in the large resource-abundant countries, albeit from a higher initial level than in the smaller countries, so that its share of GDP grew by only one-half across the income range analysed by Syrquin and Chenery (1989). The structure of manufacturing in the large resource-abundant countries also shows a skew towards capital-intensive industry. Early industry (which includes labour-intensive textiles and also agro-processing) and middle industry both expanded their share of GDP slowly in the large resource-abundant countries whereas the rise in capitalintensive late industry was larger in terms of GDP. This more capital-intensive manufacturing sector employed a smaller share of the total workforce than the manufacturing sectors of either of the two resource-poor categories. This singular pattern of structural change partly reflects a strong policy bias towards autarky (self-sufficiency) in the large resource-abundant countries. The Syrquin and Chenery data suggest that even as global trade re-expanded through the post-war years, the share of trade in the GDP of the large resource-abundant countries shrank with rising per capita income, and by one-quarter over the per capita income range they cover. This contrasts with a modest rise (one-quarter) of the share of trade in the GDP of the small resource-abundant countries, and one-half in both the large and small resource-poor countries. Not surprisingly, the increase in the share of manufacturing in total exports with rising per capita income for the large resource-abundant countries is also modest, amounting at 5 per cent of GDP over the income range covered. Although this is slightly more than the relative increase in the small resourceabundant countries, it is less than half that of the large resource-poor countries and one-quarter that of the small resource-poor countries. This modest performance in regard to the expansion of manufactured exports reflects the slow maturation (i.e., the achievement of international levels of competitiveness and ability to manipulate technology) of the manufacturing sector in the large resource-abundant countries. Sectoral maturation took decades in such countries (Bell et al. 1984; Auty 1994), rather than the five to eight years that mainstream economists consider to be the maximum allowable if the discounted costs of protecting the infant industry are to be recouped by the benefits (Krueger and Tuncer 1982). This does not appear to be an inevitable outcome of an abundant endowment of natural resources, but rather it reflects policy choices, as illustrated in Part IV. The modest share of trade in their GDP might be expected to insulate the large resource-abundant countries from the adverse impacts of trade shocks, but as Balassa (1985) and Sachs (1985) show in general, and Chapter 13 shows with reference to two case studies, this was not the case. In common with the smaller resource-abundant

30

R. M. Auty and S. Kiiski

countries, the per capita GDP growth rate of the large countries collapsed in the face of the shocks of the 1970s (see Table 2.3). Nevertheless, the larger countries recovered faster from the collapse. Table 2.3 also shows that over the entire period 1960-97, the large resource-abundant countries still grew significantly faster than the smaller resource-rich countries, to which attention now turns. 2.5. RESOURCE-ABUNDANT DEVELOPMENT IN SMALL COUNTRIES: THE STAPLE TRAP A high level of autarky is less feasible for the smaller resource-rich countries than for the larger ones, given the greater likelihood of them needing to import raw materials and their more limited opportunities for capturing both the external and internal economies of scale in manufacturing. Nevertheless, over the years 1970-89 an increased dependence on primary product exports is positively associated with a closed trade policy for most resource-abundant countries, excluding countries with unusually high resource rents like Saudi Arabia and Botswana (Sachs and Warner 1995). The scale of expansion of trade in the GDP of the smaller resource-abundant countries over the income range covered by Syrquin and Chenery (1989) was barely one-fifth compared with one-half in both the resource-poor categories. Compared with large resource-abundant countries, the small resource-abundant countries initially tended to be slower in diversifying into both the production and export of manufactured goods. For example, the Syrquin and Chenery data show that even at a per capita income level of $4,000, manufactured exports tended to remain below 4 per cent of GDP and perhaps to be only one-quarter of all exports. In addition, the smaller resource-abundant countries usually relied on one or a handful of primary product exports and this raised their vulnerability to external shocks. Imposed upon these shared features of the small resource-abundant countries are differences linked to the degree of their dependence on minerals. The remainder of this section contrasts the diffusee linkaages5 that are associatedted with cropss withh thhe pointt linkages of mining where ownership and production tends to be highly concentrated.

2.5.1. Crop-driven growth The beneficial linkages from a well-managed agricultural sector lie at the heart of Baldwin's (1956) export base model of successful resource-driven growth in the temperate regions of North America during the nineteenth century. Baldwin attributes that success to a production function that exhibits low barriers to entry, encourages production on small farms and responds positively to modest increments in inputs that generate steadily rising and equitably distributed incomes. The small farmers provide an increasing market for simple domestically-produced goods such as clothing and household furnishing (final demand linkage) rather than leaking expenditure abroad on imported luxuries. Such farmers are also more prepared to pay taxes to improve the 5

See footnote 2 in Chapter 1 for a definition of linkages.

Natural Resources and Welfsree

31

local human and economic infrastructure (fiscal linkage) like a transport network that facilitates local specialization and trade or educating the workforce to take advantage of the emerging commercial opportunities. Meanwhile, demand also grows for locally manufactured farm goods (backward linkage), such as simple agricultural machinery. The growing output also encourages investment in further processing (forward linkage), especially if the crop is perishable or experiences a weight loss. Overall, these linkages promote competitive manufacturing within a diversified and resilient economy. Baldwin contrasts these favourable diffuse linkages with those emanating from the more capital-intensive production function that he associates with many tropical plantation crops. Such crops require large investments in capital-intensive processing equipment. Mining exhibits many of these characteristics in a more extreme form, as discussed in Section 2.5.2. According to Baldwin, the production function relies upon cheap labour that generates a highly skewed income distribution and stunted fiscal, final demand, and productive linkages. This pattern is in line with dependency theory models (whose utility improves when applied to historical cases). Such enterprises function as socioeconomic enclaves with few backward and forward linkages to the local economy. Their owners exhibit little interest in paying taxes to educate the workforce and their activity promotes a dendritic transportation system that extracts the local surplus for export but does little to facilitate local trade. Final demand linkage is also stunted because, historically, workers were frequently remunerated with plots of land for subsistence food production and have little purchasing power while the elite consumes imported luxury goods. However, as noted in the previous chapter, Baldwin's focus is too narrow. The historical underperformance of tropical staples is not wholly attributable to the nature of their production function because the political economy is also important. Elitist colonial societies were able to mobilize surplus labour in the Asian tropics and so lacked the incentive to invest in enhancing labour productivity. Subsequently, the urban/industrial bias of the new post-colonial elite in developing countries often neglected agricultural productivity, especially in the small farm sector. Rather, agriculture was viewed by many such governments as a convenient source of rents that could be extracted at low political cost and deployed to accumulate political support within the emerging urban economy. The resulting disincentive to peasant farmers discouraged agricultural investment and accelerated rural-urban migration, demonstrating that the potentially beneficial linkages of peasant farming are easily repressed. In contrast, the resource-poor countries tended to take greater care to nurture their farming sectors by land reforms that widened access to land, by investments in rural infrastructure and by the dissemination of productivity-enhancing inputs. The resulting rise in farm productivity furnished the government with rents that could be invested in the emerging non-primary sector without depressing farm incentives and incomes, so that farmers provided an important market for domestically produced manufactured goods. For example, Tomich et al. (1995) estimate that one-fifth of farm income in Taiwan was extracted by state intervention during the 1950s when agriculture remained above 30 per cent of GDP. Interestingly, however, in land-abundant Thailand, the economic linkages from agricultural growth were somewhat less

32

R. M. Auty and S. Kiiski

beneficial to long-term development. Siamwalla (1995) explains this outcome in terms of the low demand that extensive agricultural expansion (via land colonization) provides for both locally produced manufactured inputs and for skill acquisition.

2.5.2. Mineral-driven growth As discussed in Chapter 1, the small resource-rich mineral economies might be expected to out-perform the crop-driven economies, given that in most cases, the mineral resource provides a source of foreign exchange and investment in addition to crops. However, the mineral economies have the slowest per capita GDP growth rate among the natural resource endowment categories (Table 2.3), although Davis (1995) finds that their income distribution and rate of human capital accumulation do not differ significantly from those of other LDCs. An important distinguishing feature of the mineral economies is a markedly skewed set of linkages that is dominated by fiscal linkage (i.e., taxation). This reflects the capital-intensive nature of mineral production and it confers a pronounced enclave character upon mining. The mining sector usually employs much foreign capital but only a small, albeit well-paid workforce, so that final demand linkage (i.e., domestic spending by capital and labour) is modest. Productive linkages are also limited because the mine inputs are usually imported due to their specialized nature, while the higher value added stages of mineral processing are usually market-orientated in their location. This leaves normal corporate taxes and any resource rent that governments secure through royalties or special taxation, as mining's principal economic stimulus. These point linkages therefore require the government of a mineral economy to play an especially critical role in the capture and domestic deployment of the mining sector's fiscal linkage. Moreover, the problems of effective management of the revenue stream are compounded by its volatility. This arises from the fact that the rent stream fluctuates more than total mineral revenues do because the rent is, by definition, the residual after all other costs are met (Auty and Mikesell 1998). Faced with this uncertainty, the governments of mineral economies have tended to be over-optimistic about the revenue stream and to absorb it too rapidly into the domestic economy. The over-rapid domestic absorption of the revenues intensifies the Dutch disease effects and establishes patterns of investment and consumption that are difficult to sustain when revenues level off or decline. Falling mineral revenue calls for a sharp depreciation of the real exchange rate and a contraction in output in order to stabilize the economy and restore growth, but this is unpopular and likely to elicit strong political opposition. Both Gelb and Associates (1988) and Hill (1991) find that the economic benefits from a mineral boom may be more than offset by the costs of adjustment during the downswing. Although some mineral exporters, notably Indonesia and, more recently, Chile, were able to sustain mineral-driven growth, most mineral economies were less successful. There is evidence that the adjustment problems are greater for the smaller mineral economies like Jamaica or Trinidad and Tobago. This is because they lack the opportunity to diversify into the production of non-mining resources compared with mid-sized countries like Chile and Peru, or into

Natural Resources and Welfaree

33

labour-intensive manufacturing that a large low-income country like Indonesia could take advantage of (Auty and Mikesell 1998). 2.6. CONCLUSION S The stylized facts suggest that, compared with resource-abundant countries, the resource-poor ones pursue a virtuous economic circle. They diversify into manufactured production and exports earlier and are less prone to trade policy closure so that their manufacturing sectors tend to be competitive. They also experience an earlier passage through the favourable middle stage of the dependency cycle. As a result, resource-poor countries with unusually low levels of cultivated land per capita (notably Northeast Asia) raised their investment rates earlier and also sustained investment efficiency. They achieved rapid GDP growth with remarkably low levels of income inequality. They completed the industrial transition within two generations as opposed to the five generations that Syrquin (1986) suggests as more typical (Table 2.2). All else being equal, crop-driven growth is likely to be slower than competitive manufacturing-driven growth, but not as slow as it became during the 1970s and 1980s. The stylized facts suggest that policy has played a central role. The experience of some countries in Southeast Asia confirms that many tropical crops retained considerable potential for growth, but that the outcome depended upon policy choice. Unfortunately, most resource-abundant countries tended to close their trade policy and/or to relax market discipline. This retarded the competitive diversification of the economy and increased its vulnerability to a growth collapse. The development trajectory of the resource-abundant countries varies with the specific character of their natural resource endowment. The large resource-abundant countries have greater scope to absorb external shocks, but they still failed to avoid growth collapses, although they subsequently recovered faster than small ones (Table 2.3). The mineral economies are potentially more vulnerable to policy error than countries with more diffuse economic linkages are, but Chapter 10 shows that the potential benefits of diffuse linkages have proved easy to repress. One might be forgiven for concluding from this discussion that the faster resourceabundant countries deplete their resource rents, especially minerals, the sooner they can embark on the advantageous path of competitive industrialization. However, Chapter 3 draws upon recent advances in environmental and natural resource accounting to demonstrate the conditions under which rapid and sustainable economic growth can accompany the exploitation of even finite (depleting) natural resources. It confirms that policy counts. REFERENCES Adelman, I. and C. T. Morris (1973), Economic Growth and Social Equity in Developing Countries, Stanford CA: Stanford University Press. Anand, S. and S. M. R. Kanbur (1993), 'Inequality and development: a critique', Journal of Development Economics, 41, 19-43.

34

R. M. Auty and S. Kiiski

Auty, R. M. (1994), 'Industrial policy reform in six large newly industrialising countries: the resource curse thesis', World Development, 22, 11-26. Auty, R. M. (1997a), 'Natural resources, the state and development strategy', Journal of International Development, 9, 651-63. Auty, R. M. (1997b), 'Competitive industrial policy and macro performance: has South Korea outperformed Taiwan?', Journal of Development Studies, 33, 445-63. Auty, R. M. and R. F. Mikesell (1998), Sustainable Development in Mineral Economies, Oxford University Press, Oxford. Balassa, B. (1985), 'Adjusting to external shocks: the newly-industrialising developing countries', Weltwirtschaftsliches Archiv, 122, 1141-61. Baldwin, R. E. (1956), 'Patterns of development in newly settled regions', Manchester School of Social and Economic Studies, 24, 161-79. Bell, M. et al. (1984), 'Assessing the performance of infant industries', Journal of Development Economics, 16, 101-28. Bloom, D. E. and J. G. Williamson (1998), 'Demographic transitions and economic miracles in emerging Asia', The World Bank Economic Review, 12, 419-55. Bowman, K. S. (1997), 'Should the Kuznets effect be relied on to induce equalising growth: evidence from post-1950 development', World Development, 25, 127-43. Brown, R. and D. Julius (1993), 'Is manufacturing still special in the new world order?', The Amex Economic Review 6-17. Reprinted in J. Drew (ed.), Readings In International Enterprise, London: Routledge, 1995: 275-85. Davis, G. (1995), 'Learning to love the Dutch disease: evidence from the mineral economies', World Development, 23, 1765-79. Deininger, K. and L. Squire (1996), 'A new data set for measuring income inequality', World Bank Economic Review, 10/3, 365—91. Echevarria, C. (1997), 'Changes in sectoral composition associated with economic growth', International Economic Review, 38, 431-52. Edwards, S. (1997), 'Why are Latin American saving rates so low?', in N. Birdsall and F. Jasperson (eds.), Pathways to Growth: Comparing East Asia and Latin America, Baltimore MD: Johns Hopkins University Press: 131-58. Fields, G. (1989), 'Change in poverty and inequality in the developing countries', World Bank Research Observer, 4/2, 167-85. Gelb, A. H. and Associates (1988), Oil Windfalls: Blessing or Curse? New York: Oxford University Press. Gylfason, T. (1999), 'Natural resources and economic growth: a Nordic perspective on Dutch disease', UNU/WIDER Working Papers, 167, Helsinki: UNU/WIDER. Hill, C. B. (1991), 'Managing commodity booms in Botswana', World Development, 19, 1185-96. Johnston, B. F. and J. W. Mellor (1961), 'The role of agriculture in economic development', American Economic Review, 51, 566—93. Kindleberger, C. P. and B. Herrick (1977), Economic Development, London: McGraw Hill. Krueger, A. O. and B. Tuncer (1982), 'An empirical test of the infant industry argument', American Economic Review, 72, 1142-52. Kuznets, S. (1955), 'Economic growth and income inequality', American Economic Review, 45, 1-28. Lal, D. and H. Myint (1996), The Political Economy of Poverty, Equity and Growth, Oxford: Clarendon Press.

aturalressourcesandwelfarwelfaree

35

Londono, J. L. (1996), Poverty, Inequality and Human Capital Development in Latin America, 1950-2025, Washington DC: World Bank. Matsuyama, K. (1992), 'Agricultural productivity, comparative advantage and economic growth', Journal of Economic Theory, 58, 317-34. Mellor, J. W. (1995), Agriculture on the Road to Industrialisation, Baltimore MD: Johns Hopkins University Press. Sachs, J. D. (1985), 'External debt and macroeconomic performance in Latin America and Asia', Brookings Papers on Economic Activity, 2, 523—75. Sachs, J. D. and A. M. Warner (1995), 'Natural resources and economic growth', Cambridge MA: HIID (mimeo). Sachs, J. D. and A. M. Warner (1997), 'Natural resource abundance and economic growth', Cambridge MA: HIID (mimeo). Siamwalla, A. (1995), 'Land-abundant agricultural growth and some of its consequences: Thailand', in J. W. Mellor (ed.), Agriculture on the Road to Industrialisation, Baltimore MD: Johns Hopkins University Press: 23-66. Syrquin, M. (1986), 'Productivity growth and factor reallocation', in H. B. Chenery, S. Robinson, and M. Syrquin (eds.), Industrialization and Growth: A Comparative Study, New York: Oxford University Press, 229-62. Syrquin, M. and H. B. Chenery (1975), Patterns of Development, Oxford: Oxford University Press. Syrquin, M. and H. B. Chenery (1989), 'Patterns of Development, 1950 to 1983', World Bank Discussion Paper 41, Washington DC: World Bank. Timmer, C. P. (1988), 'The agricultural transformation', in H. B. Chenery and T. N. Srinivasan (eds.), Handbook of Development Economics, Vol. 1, Amsterdam: North-Holland: 275-331. Tomich, T. P., P. Kilby and B. F. Johnston (1995), Transforming Agrarian Economies: Opportunities Seized and Opportunities Missed, Ithaca: Cornell University Press. Williamson, J. G. (1965), 'Regional inequality and the process of national development: a description of the patterns', Economic Development and Cultural Change, 13, 3-45. Wood, A. and K. Berge (1997), 'Exporting manufactures: human resources, natural resources, and trade policy', Journal of Development Studies, 34, 35—59. World Bank (1993), The East Asian Miracle, Oxford: Oxford University Press. World Bank (1997), World Development Indicators 1997, Washington DC: World Bank. World Bank (1999), World Development Indicators 1999, Washington DC: World Bank. World Resources Institute (1994), World Resources 1994-95, Oxford: Oxford University Press.

3 The Sustainability of Extractive Economies KIRK HAMILTO N

3 . 1 . INTRODUCTION This study is broadly concerned with the nature of the development path in resourceabundant economies, and in particular, the risk that these economies may fall into a staple trap. This chapter draws upon recent progress in environmental and natural resource accounting to pose the question: do these resource-abundant economies actually consume wealth along the development path? The answer to this question has clear implications on whether their development can be sustained. The publication of the Brundtland Commission report in 1987 introduced a critical new dimension to our conception of economic development by raising the issue of the suslainability of development. While definitions of sustainable development abound (Pezzey 1989 is a good summary), economists have settled on a simple formulation that can capture a very rich set of phenomena: a development path is sustainable if total welfare does not decline along the path. As long as the welfare function is sufficiently expansive in what it measures (consumption, environmental quality, social equity, and other factors contributing to the quality of life), this definition permits a rigorous characterization of sustainable development. Pearce et al. (1989) take the sustainability argument one step further by positing the existence of critical natural capital (the ozone layer, for instance) for which no substitute exists. This conception of strong sustainability therefore requires the preservation of critical natural capital in order for development to be sustainable. Weak sustainability assumes that there are substitutes for all assets. As should be obvious, opting for sustainability is an ethical decision. The Utilitarian maximand assumed in most models of economic growth, the present value of welfare along the optimal path, can be shown to lead to unsustainable outcomes under simple assumptions (fixed technology and pure rate of time preference combined with an exhaustible resource that is essential for production (see Dasgupta and Heal 1979)). Choosing sustainability implies a concern with the welfare of future generations that is not captured by the Utilitarian maximand. The fact that most countries and international institutions have adopted sustainable development as an explicit goal suggests that a powerful impulse is at work.

Sustainability of Extractive Economies

37

While the question of sustainability is relevant for all resource-dependent: economies, it is particularly stark for extractive economies. Economists posed this question in somewhat different words at the time of the first oil crisis of the 1970s: would economies that depend upon an exhaustible resource inevitably decline as the resource is depleted? An early paper on this topic by Solow (1974) spurred Hartwick (1977) to provide a simple answer to this question. If the 'Hartwick rule' is followed, so that investment in produced capital just equals current scarcity rents on the exhaustible resource at each point in time, then the resulting path for the economy is one where welfare equals a maximal constant value ad infinitum—in other words, this is a sustainable economy. This holds even for a fixed technology. Hamilton (1995) explores the critical role that the elasticity of substitution between produced assets and natural resources plays in the feasibility of the Hartwick path. This chapter develops a theoretical approach to measuring whether an economy is on a sustainable path, and then presents empirical evidence on the sustainability of a range of extractive economies. While sustainability is not the only issue in managing an economy that is dependent on natural resources, as other chapters in this study attest, it is a critical issue if societies do in fact adopt an ethical stance that is concerned with the welfare of future generations. Expanded national accounting is the tool that will be used below to shed light upon the question of sustainable development. Valuing depletion and degradation within a national accounting framework is an increasingly viable proposition, both as a result of the significant progress made in the techniques of valuation of environmental resources (see, for example, Freeman 1994) and as a result of the expanding foundation that theoretical developments are placing under the methods of'green' national accounting (Weitzman 1976; Hartwick 1990; M a ler 1991; Hamilton 1994, 1996). The first cross-country application of these greener accounting methods to the measurement of net savings appeared in Pearce and Atkinson (1993), who combined published estimates of depletion and degradation for 20 countries with standard national accounting data to examine true savings behaviour. By this measure many countries appear to be unsustainable because their gross savings are less than the combined sum of conventional capital depreciation and natural resource depletion. Enlarging the concept of net saving to include the depletion of natural resources is a natural extension of traditional savings concepts. This is because the depletion of a natural resource is, in effect, the liquidation of an asset and so should not appear as a positive contribution to net income or net savings. While minor technical issues remain, the methods of valuing the depletion, discovery and growth of commercial natural resources in the context of the SNA are by now well developed (Hamilton 1994; Hill and Harrison 1994). The indicator of sustainability developed below is termed 'genuine saving', to distinguish it from the usual national accounts definitions of saving. Not surprisingly, the definition of this measure hinges on a fundamental question: what is income?

38

K. Hamilton 3.2. GENUINE SAVING AS A SUSTAINABILITY INDICATOR

3.2.1. The theory of genuine saving However it may be defined in detail, achieving sustainable development necessarily entails creating and maintaining wealth. Given the centrality of savings and investment in economic theory, it is perhaps surprising that the effects of depleting natural resources and degrading the environment have not, until recently, been considered in the measurement of national savings. Augmented measures of savings and wealth in the national accounts offer promise therefore, as indicators of sustainable development, which was a prime motivation for the publication of Expanding the Measure of Wealth (World Bank 1997). For this to be a useful framework for the analysis of sustainability, however, a suitably expansive definition of wealth is required. At a minimum, wealth should include produced assets, human capital, natural resources, and liabilities in the form of stocks of pollution—this goes far beyond the usual confines of national accounting, and introduces new challenges in both theory and measurement. Some of the theoretical underpinnings of expanded national accounting are presented in Appendix 3.1, while measurement issues are considered below. Wealth represents the potential to generate income or, more broadly, to generate welfare. Changes in wealth are therefore intimately linked to the question of sustainability, and the notion of 'genuine' saving can most simply be conceived as the net change in wealth over an accounting period as new assets are invested, old assets depreciate, and human and natural resources are augmented or depleted. The accounting relationships can be made more precise as follows. First, net national saving NNS is equal to the amount of gross national product GNP that is not consumed C, less the value of depreciation of produced assets S K:

Genuine saving G goes beyond this definition by deducting the depletion of subsoil resources R, valued at the unit resource rental rate ns, and the net depletion of living resources (measured as harvest h minus growth g) valued at rental rate nl, then by deducting net pollution accumulation e valued at marginal social costs cr , and finally by adding current educational expenditures E as a measure of investment in human capital:

The economic principles underlying this expression derive from growth theory. In the model of depletion and discovery of subsoil resources presented in Appendix 3.1, a more general and all-encompassing expression for wealth is employed: wealth is equal to the present value of welfare (assumed to include, for example, the amenity value of the environment as well as consumption) along the optimal development path for the

Sustainability of Extractive Economies

39

economy. From this can be derived the formal definitions of the national accounting aggregates of interest: • NNP (net national product) is the maximum amount of produced output that could be consumed at a point in time while leaving wealth instantaneously constant. « Genuine savings is the difference between NNP and consumption. Note that the definition of NNP is a straightforward generalization of Hicks' (1946) definition of income, where Hicks referred to income as the maximum amount that could be consumed while still leaving oneself as well off at the end of the accounting period as at the beginning. Being 'as well off in the current context implies maintaining the level of the present value of welfare along the optimal path. This is equivalent to the definition employed in Pemberton and Ulph (1996). One of the key results from theory, shown in Appendix 3.1, is that negative genuine saving implies that welfare must decline at some point in the future on the optimal development path, that is, negative genuine saving implies unsustainability. In practical terms, this result can be taken to imply that continuing the policy mix that leads to negative genuine savings will eventually lead to an unsustainable outcome. This framework is general enough to handle strong as well as weak sustainability. As long as it is assumed that there are unbounded losses of welfare associated with some stock of critical natural capital falling below a threshold, measuring negative genuine savings will indicate that the optimal development path is not sustainable.

3.2.2. Practical genuine saving measures While theory gives clear answers on the definition of expanded accounting aggregates such as genuine saving, it is important to consider some of the practical issues involved in actually measuring these aggregates. Several points about the expression for genuine saving are worth noting. First, current educational expenditures clearly include a consumption element as well as investment. Second, human capital investment should be valued on the basis of marginal rather than average costs (these are, of course, equal to marginal benefits at the optimum), and so current expenditures may represent an undervaluation of human capital investment. Third, this approach ignores human capital depreciation. Next, although in principle, forest growth that exceeds harvest should be added to genuine saving, in practice it is likely that many of the trees in 'net growth' countries are not economic. Finally, turning to pollution, care must be taken to exclude any damages that are already reflected in GDP, such as loss of current output in the forestry or agriculture sectors as a result of pollution emissions. The economic principles underlying mineral accounting are well understood: (1) asset values should be measured as the present value of economic profits over the life of the resource; and (2) asset depreciation should be measured as the user cost, the change in asset value associated with extraction of the mineral over the accounting period. If there were a private market for subsoil resource deposits, these values for assets and depletion could be readily observed in market transactions and the accounts

40

K. Hamilton

of resource owners. In practice, virtually all resource deposits around the world are owned by states, which in turn contract out the extraction of resources in return for the payment of royalties. Under these circumstances the wealth accounting problem lies in the estimation of values for resource assets and their depletion, with the aim of arriving at an approximation of the values that would be observed in a free market. A key practical problem with the above expression for genuine saving, derived from growth theory, is that it assumes full optimality: for a given resource, scarcity rents are assumed to obey the Hotelling rule, and the optimal extraction path maximizes the present value of total resource rents. There are a wide variety of reasons why real mines and forests may not behave the same as textbook mines and forests, including the influence of distortionary taxes, oligopolistic pricing, and incomplete property rights regimes. There are two polar methodologies for estimating mineral depletion in the natural resource accounting literature: the simple present value approach (El Serafy 1989), which assumes constant total rents and no optimization, and the total rent approach (Repetto et al. 1989), which is consistent with optimality only if resource prices can be assumed to increase at a near-exponential rate. For low social discount rates (which feed into the calculation of resource stock values in the simple present value approach) and moderate resource asset lives (less than 20 years, which is typical for many minerals) the divergence in estimates between the competing methodologies is not huge. There are many possible methods for estimating mineral depletion and no obvious 'best' method given the divergences between the real world and textbook optimality. As the most parsimonious approach in terms of data requirements, total rents are used to derive the empirical results below.

3.2.3. Coverage and calculation issues The list of data sources for the resource rental estimates are given in Hamilton and Clemens (1999). The basic approach to calculating resource rents for nonrenewable resources is to subtract country- or region-specific average costs of extraction from the world price for the resource in question, all expressed in current US dollars. Many world prices were derived from World Bank commodity data—where multiple markets, for example, London and New York, are reported, a simple average of these market prices serves as the world price. For minerals the levels of total resource rents are thus calculated as: Rent = world price — mining cost — milling and benefication costs — smelting costs — transport to port — 'normal' return to capital. For crude oil, unit rents are calculated as the world price less lifting costs. Natural gas, though its international trade has soared in recent years, does not have a single world price. A world price was estimated by averaging free-on-board prices from

Sustainability of Extractive Economies

41

several points of export world-wide, following which the unit rents were calculated, as for oil. For hard coal, a world price was calculated by combining data on steam and coking coals after adjusting for differences in heat content and quality. A world price for lignite was obtained by analysis of national-level differences in prices between hard coal and lignite in various countries and estimating a similar proportion of values with respect to the world price for hard coal. Unit rents for both hard coal and lignite were then calculated as for oil. For forest resources, a price for the wood of each country was calculated based on the proportions of fuelwood, coniferous softwood, non-coniferous softwood, and tropical hardwood, found in its total annual production. Representative world prices were used for each type of wood, and a price for fuelwood was estimated using World Bank data. Unit rents were calculated by subtracting average unit production costs from the world price. There are several further points to note about this methodology: (1) Countries may or may not be selling their natural resources for internal consumption at the world market price, although one would expect that they have every incentive to do so. Moreover, the use of uniform world prices will tend to overstate rents for countries with lower-grade resources. (2) Extraction costs are measured at a fixed point in time, a point that differs from country to country and resource to resource according to data availability, and held constant in real terms. World prices vary over time, leading to corresponding variations in calculated rental rates. (3) Where the extraction cost data were region- rather than country-specific, the regional cost structure was applied to all of the producing countries in the region. (4) Rents on minerals are generally viewed as accruing to the resource owner for the production of the crude form of the material in question, typically an ore. In practice, most mineral operations are vertically integrated to a considerable extent and so the only price and cost data are for refined forms of the materials. Measuring resource rents as described above for these vertically integrated mineral operations therefore implicitly ascribes any excess returns to capital for the milling and refining stages to the resource rent. The total rent estimates used in this chapter are therefore fairly crude. In compensation, the estimates are calculated using a uniform methodology and the coverage is quite wide. In addition to timber, coal, oil and natural gas, the minerals covered include zinc, iron ore, phosphate rock, bauxite, copper, tin, lead, nickel, gold, and silver. Data problems led to the exclusion of diamonds from these estimates. Turning to pollution damages, these should ideally reflect emissions and exposure data for the full range of local, regional, and global pollutants. In practice, there are no comprehensive data on local and regional pollutants. As a 'place-holder' for other pollutants, therefore, damages from carbon dioxide emissions are included in the genuine saving calculation, using a figure of $20 per ton of carbon derived from Fankhauser (1995) and widely available data on CO2 emissions from industrial

42

K. Hamilton

sources. Making a deduction for emissions of a global pollutant is conceptually correct if certain property rights are assumed, in particular the right not to be damaged by your neighbour's pollution emissions. Note that local air and water pollution damages may be 5-10 times greater than carbon dioxide damages in most countries. Education expenditure data are from Unesco (1998). 3.3. RESULTS Country-level results for genuine saving and its components in 1997 are presented in Table A3.1. As this table indicates, negative genuine saving is more than a theoretical possibility. It is important to note several issues with regard to these figures. First, a point measure of genuine saving does not necessarily imply that the country in question is fated for an unsustainable development path; it does imply, however, that continuing the current policy mix is unsustainable. Second, it may be perfectly rational for either extremely poor or extremely rich countries to consume wealth in the short run, in the former case to hold off starvation, in the latter because consuming a very small proportion of wealth entails a low loss of welfare over time. Finally, negative genuine savings rates represent an opportunity not taken: resource endowments represent a type of stored development finance, and some countries choose not to benefit from this natural advantage. Broad trends in the savings figures can be seen in Table 3.1, which summarizes genuine savings for countries aggregated by income level and natural resource endowment. The income level figures are weighted according to the size of the economy, whereas those for the natural resource endowments are straight averages. This means that the smaller countries exert greater influence on the results for the natural resource endowment categories and, as this study demonstrates, the smaller countries have tended to be less successful in their development efforts than the larger countries have. Comparing low and middle income countries in Table 3.1, the 10.2 per cent difference in genuine savings is largely explained by a 9.2 per cent difference in the gross saving rate. However, depletion is significantly higher in low income countries, at 6.6 per cent of GDP as compared with 4.5 per cent in middle income. Much of this difference in turn is a function of the 1.8 per cent of GDP that net forest depletion represents in low income countries. In high income countries, depletion is only 0.5 per cent of GDP and education expenditures are 2 per cent higher than in low and middle income countries. Turning to the comparisons by natural resource endowment, the large resourcepoor countries exhibit relatively high gross and genuine savings rates, with natural capital depletion amounting to only 2.3 per cent of GDP. In contrast, the relatively high gross saving of the large resource-abundant countries is significantly offset by their consumption of natural capital. This results in their genuine savings rate being little higher than those of the smaller resource-poor countries and also the small crop-driven and ore-driven resource-rich countries, whose net domestic saving effort is considerably lower. This largely reflects the importance of oil in the economies of the large resource-rich countries and the high depletion rate that this high-rent

Table 3.1. Genuine saving as per cent of GDP, 1997 Gross Consumptio n of Net Educatio n Energ y Minera l Ne t Carbo n dioxid e Genuin e domestic fixe d capita l domesti c expenditur e depletio n depletio n fores t damag e domesti c savings saving s depletio n saving s World Low income Middle income High income Large resource-poor Small resource-poor Large resource-rich Small resource-rich crop Small resource-rich

22.2 17.0 26.2 21.4 23.6 14.7 22.0 13.5

11.7 8.0 9.2 12.4 7.4 6.4 8.0 6.6

10.5 9.1 17.0 9.0 16.3 8.3 14.0 6.8

5.0 3.4 3.5 5.3 2.7 3.2 3.8 3.5

1.2 4.2 3.8 0.5 1.7 0.0 6.7 0.7

0.1 0.6 0.5 0.0 0.2 1.4 1.1 0.6

0.1 1.8 0.2 0.0 0.4 2.5 0.4 0.9

0.4 1.2 1.1 0.3 0.8 0.5 0.9 0.3

13.6 4.8 15.0 13.5 15.9 7.1 8.7 7.9

15.2

7.5

7.7

3.5

0.6

2.4

0.0

0.4

7.7

30.8

9.2

21.6

4.2

21.9

0.0

0.0

0.9

2.9

ore

Small resource-rich oil

Note: The income figures are weighted by the size of the economy whereas the natural resource endowment figures are unweighted averages and are therefore more heavily influenced by the smallest countries compared with the figures for income groups, Source: Derived from World Bank (1999).

44

K. Hamilton

commodity entails. Not surprisingly therefore, the small oil-rich countries tend to have the lowest genuine saving rate (barely positive) despite very high rates of gross saving. This outcome points towards a relatively inefficient conversion of natural capital into alternative forms of capital, an issue that is examined in Chapter 12 with reference to Saudi Arabia. More insight into 1997 genuine savings rates is provided by Fig. 3.1, which scatters the genuine saving rate for the most resource dependent countries (those where depletion exceeds 5 per cent of GDP) against the share of depletion in GDP. As the regression equation next to the fitted line shows, there is a weak tendency (R2 = 0.389) for increasing depletion shares of GDP to be associated with lower genuine savings rates. The regression coefficients are significant (/ = —4.44 and 3.75 for the slope and the constant respectively). The slope of the regression line can be interpreted as an elasticity, and therefore says that each 1 per cent rise in the depletion share of GDP is associated with a 0.82 per cent decline in the genuine saving rate. This slope can also be interpreted as the marginal propensity to consume resource rents, although it must be kept in mind that the slope is derived from cross-sectional data. It says that there is a tendency, looking across countries, for 82 per cent of each increment in resource rents as a share of GDP to be consumed. If in each country all resource rents were invested rather than consumed then no statistically significant tendency would be visible in Fig. 3.1—an extra unit of resource rent in GDP would have no effect on genuine savings, other things being equal. This interpretation of the regression results must be tempered by recalling that the countries with largest depletion share of GDP (those in the lower right of Fig. 3.1)

Figure 3.1. Genuine saving rate vs. depletion share of GDP in resource-dependent economies, 1997 Source: World Bank (1999).

Sustainability of Extractive Economies

45

tend also to be those with the largest resource endowments, and so the total rent approach to measuring depletion may be influencing the results derived. 3.4. POLICY ISSUES It. is abundantly clear that monetary and fiscal policies are the biggest levers for boosting savings rates. The first policy issue is therefore: to what extent do monetary and fiscal policies encourage strong domestic savings? While natural resource exports boost foreign savings and therefore the overall savings effort, the analysis of genuine savings suggests a further question: to what extent do exports of exhaustible resources boost the genuine rate of saving?1 The answer to this lies in netting out the value of resource depletion from gross export values. More optimal natural resource extraction paths will, other things being equal, boost the value of genuine savings. The policy question for natural resource management is therefore: to what extent can stronger resource policies (royalty regimes, tenure) boost the genuine rate of saving? Similarly, reducing pollution emissions to socially optimal levels will boost the value of genuine savings. The policy issue with respect to pollution is: to what extent can more optimal pollution control policies increase the rate of genuine saving? Note that the policy prescriptions for boosting genuine savings should never be to stop extracting resources or emitting pollutants altogether. Rather, pricing resources and pollutants correctly and enforcing property rights will lead to efficient levels of exploitation of the environment, reducing incentives to high-grade resources or pollute indiscriminately. Optimal resource and environmental policies will maximize genuine savings, subject to the macroeconomic policy regime in place. The sorts of issues raised by Gelb and Associates (1988) about the nature and effects of oil windfalls in developing countries are particularly relevant to the policy issues just raised: without sound macroeconomic policies and prudent allocation of public resources, the effects of reliance upon large resource endowments can be negative for many countries. The other chapters in this study explore and clarify the ways in which this can happen. 3.5. CONCLUSIONS Growth theory provides the intellectual underpinning for expanded national accounting and, through the measure of genuine saving, an indicator of when economies are on an unsustainable development path. This theory points in useful directions for countries concerned with sustainable development. Far from being a mere theoretical possibility, there is abundant evidence that there are countries whose policy mix results in negative genuine savings rates. While the latest World Bank estimates for 1997 are emphasized here, previous studies such as The 11question is also germane for unsustainable forest harvest programs.

46

K. Hamilton

Hamilton and Clemens (1999) and Atkinson et al. (1997) have shown this to be true over decades as well. The evidence suggests that, while extractive economies are potentially sustainable if resource rents are invested in other productive assets (including human capital), many of these economies have not chosen this path. The results presented here show distinctive patterns of genuine savings across country income and natural resource endowment groups. However, even robust genuine savings do not necessarily lead to a smooth development path if the quality of investment is low, as the financial crisis in Southeast Asia attests. The genuine savings analysis raises an important set of policy questions that transcend the traditional concern with the macro and microeconomic determinants of savings effort. The questions of rent capture, public investments of resource revenues, resource tenure policies, and the social costs of pollution emissions are equally germane in determining the overall level of saving, although it is clear that monetary and fiscal policy remain the big levers. Finally, human capital plays an increasingly dominant role as development proceeds, as Table 1.2 demonstrates. The next chapter shows how the rate of accumulation of human capital is affected by the natural resource endowment.

Appendix 3.1 . Deriving Net Incom e an d Genuine Savin g What follows is an outline of a general result with regard to savings and sustainability. We assume a simple economy in which a composite good can be consumed, invested, or used to abate pollution. Welfare U for the representative individual in this economy is a function of consumption and some number of stocks of living and nonliving natural resources and pollutants, with the stocks of natural resources (e.g., forests and other green spaces) generally adding to welfare, and the stocks of pollutants decreasing it. The measure of wealth W for this society is defined to be the present value of welfare on the optimal path over an infinite time horizon,

where p is the pure rate of time preference (assumed to be constant). Production in this simple economy is defined by an aggregate production function that combines produced assets and natural resources (labour and population are assumed to be constant, and so are factored out of the model) to yield the composite good. Production leads to pollution emissions, which may be abated by some input of the composite good; pollution emissions accumulate in stocks that dissipate as a result of natural processes. Genuine saving G for this economy is defined to be the sum of the net investment in produced assets and the physical changes in the various stocks of natural resources and pollutants, valued at the shadow prices supporting the optimal path—scarcity rents in the case of natural resources, marginal damages in the case of pollutants. For this simple model it follows

Sustainability of Extractive Economies

47

that at each point on the optimal path,

and therefore that,

If genuine saving is negative at a point in time t on the optimal path, therefore, wealth is declining and welfare at some time in the future must be less than it is at t. Measuring negative genuine savings is an indication that the optimal path is not sustainable. Expressions (1) and (2) are quite general, requiring only that the pure rate of time preference be fixed. The general model of genuine saving just presented can be adapted to highlight the accounting issues for a small exporter of exhaustible resources. To achieve this, a simplifying assumption needs to be made: assume N deposits of a homogeneous resource, where the Mb. deposit is specialized for exports, the (N— l)th is on the frontier and expands as a result of exploration expenditures, and the remaining N—2 neither export nor expand, serving to meet domestic demand. Although the resource is homogeneous, the deposits are not and have specific extraction cost functions fi(Rt)2 (where Ri is the quantity extracted from the ith deposit) and individual stock sizes Si. For the (N— l)th deposit, new discoveries of quantity D cost an amount v(Q, D) in exploration costs. These exploration costs are an increasing function of both the quantity discovered and cumulative discoveries Q—that is, VD > 0 and VQ < 0. Total domestic supply of the resource is defined by R,

Depletion and accumulation for each of the resource deposits is given by,

Cumulative discoveries are given by

For foreign assets A, repatriation of assets M, fixed foreign interest rate r, and international resource price path p, the accounting identity for foreign trade is,

Domestic production is defined by a production function F = F(K, R), where K represents produced capital. Labour and population are assumed to be fixed, and so can be factored out of the production function. The overall accounting identity for supply and disposition in the 2 Unless otherwise stated, all variables are assumed to vary over time (and so have an implicit time subscript).

48

K. Hamilton

economy is defined by

This says that the sum of domestic production and repatriated foreign assets just equals consumption C plus investment plus resource extraction costs plus resource discovery costs. The usual efficiency condition for foreign trade is assumed: the domestic economy is indifferent between exporting and domestic use of the resource commodity, implying that the domestic resource price FR (the marginal product of the resource) just equals the international price,

The optimal growth problem for this economy is to maximize the present value of welfare U, assumed to depend only on consumption C, subject to conditions (1)-(10),

by choosing optimal time paths for C, Ri,D, and M. By solving this growth problem and deriving the shadow prices of all assets, net national product (NNP) for this economy, using the extended Hicksian definition of income, is given by

Genuine saving is just NNP — C. Note that the first five terms in this expression are simply GNP: consumption plus investment plus net factor income from abroad, plus exports, less imports. The next term represents the value of depletion for each of the resource deposits, while the next implies that resource discoveries should be included in national income valued at marginal discovery costs (which is necessarily less than the marginal rental value of these discoveries). The final term embodies the effects of changing terms of trade as international resource prices vary over time. It equals the present value of resource exports (discounted at the fixed international interest rate) weighted by the percentage rate of change of international resource prices. If resource prices are constant in the future then this term equals zero. If international resource prices fall in the future at an average rate of 1 per cent per year (a good approximation to what has happened to crude petroleum prices over the long term), then Hicksian income will be less than national income, as ordinarily measured, by 1 per cent of the present value of resource exports. Rising resource prices would have the effect of boosting Hicksian income. Expression (12) can be interpreted to mean that depletion should be measured resource deposit by resource deposit, assuming that sufficient data exist. The treatment of resource discoveries can be traced to Hartwick (1992)—as long as there is not a wide divergence between average and marginal discovery costs, this suggests that standard national accounts already provide an adequate treatment of discoveries, since the SNA treats most exploration expenditures as investment. The effects of changing terms of trade were derived in Vincent et al. (1997).

Table A3.1. Genuine saving and its components, per cent of GDP, 1997 Gross Consumption Net Education Energy Mineral Net Carbon dioxide Genuine domestic of fixed domestic expenditure depletion depletion forest damage domestic savings capital savings depletion savings Albania Algeria Angola Argentina Armenia Australia Austria Azerbaijan Bangladesh Belarus Belgium Benin Bolivia Bosnia and Herzegovina Botswana Brazil Bulgaria Burkina Faso Burundi Cambodia Cameroon Canada Central African Republic

-13.2 34.5 27.3 18.4 -28.8 20.7 23.5

9.5 14.7 21.6 22.3 10.8 10.1

12.4

9.3 6.0 10.5

— 14.6 12.9 14.0

7.2

-25.6 25.2 21.2

7.9 — 6.1 10.5 -4.5

7.5 4.4

17.2 10.1

12.2

5.4 8.1 —

5.4 2.0 —

44.7 18.6 17.4

13.3

31.4 11.1

9.2 2.6 4.2

4.6 4.4 5.0 7.5

—

7.5 10.1

20.6 21.5

12.5

6.7

5.2

7.3 4.6 -1.8 -0.8 13.1

9.0 1.5

2.8 6.3 2.6 2.4 — 4.7 4.9 — 2.1 4.7 4.9 — 2.6 — 6.9 4.2 4.0 2.8 3.0 — 2.3 6.1 3.8

0.0 2.4 20.7

0.5 0.0 1.2 0.1 21.8

0.2 0.0 0.0 0.0 0.9 0.0 0.0 0.6 0.5 0.0 0.0 0.0 7.4 1.5 0.0

0.1 0.1 0.0 0.0 0.0 1.5 0.0 0.0 0.0 0.0 0.0 0.0

0.5

-23.5 27.9

0.4 0.2 1.3 0.4 0.1 5.1 0.3 1.7 0.2 0.2 0.7 —

2.7 9.6 — 7.6

0.0

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

0.8 0.7 1.3 0.0 0.0 0.0 0.0 0.4 0.0

0.0 0.0 0.0 0.0 8.5 0.0 0.0 0.0 0.0

0.3 0.2 2.7 0.2 0.1 0.1 0.3 0.4 0.1

1.1

1.1

15.2 -31.4

9.1 7.4 16.9

5.2 1.8 — 37.2 13.9

6.7 7.1 -7.4

-0.9

7.7 12.8

5.1

Table A3.1. (Continued) Gross Consumption Net Education Energy Mineral Net Carbon dioxide Genuine domestic o fixed domestic expentditure depletion depletion forest damage e domestic savings depletion savings savings capital Chad 1.2 Chile 24.5 China 42.7 Hong Kong, China 30.6 Colombia 15.8 Congo Democratic Republic 9.0 Congo, Republic 34.8 Costa Rica 25.0 Co ted'Ivoire 23.1 Croatia 3.3 Cuba — Czech Republic 28.4 Denmark — Dominican Republic 22.1 Ecuador 21.2 Egypt Arab Republic 13.0 El Salvador 4.5 Eritrea -17.4 Estonia 18.4 Ethiopia 8.7 Finland 24.6 France 19.7 Gabon 48.3 Gambia 3.8

4.6 6.8 6.2 — 6.5 5.0 9.2 2.5 7.0 9.2 — 17.2 15.3 5.9 6.9 7.9 6.1 4.1 10.5 — 16.7 12.9 15.2 12.3

-3.4 17.7 36.5 — 9.4 4.0 25.6 22.5 16.0 -5.9 — 11.2 — 16.1 14.2 5.1 -1.6 -21.5 7.9 — 7.9 6.8 33.1 -8.6

— 3.2 1.9 2.0 2.8 0.7 4.3 4.6 5.7 — — 5.3 7.7 1.2 2.7 4.8 2.2 — 4.3 2.9 7.2 5.0 3.1 3.3

0.0 0.1 0.0 0.0 4.4 0.0 23.9 0.0 1.5 0.9 0.0 0.4 0.4 0.0 12.0 3.2 0.0 0.0 1.7 0.0 0.0 0.0 15.8 0.0

0.0 6.4 0.5 0.0 0.1 0.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.5 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

0.0 0.0 0.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

0.0 0.4 2.4 0.1 0.4 0.2 0.4 0.3 0.7 0.5 — 1.3 0.2 0.8 0.7 0.7 0.3 — 2.3 0.3 0.2 0.1 0.4 0.3

-3.5 14.1 34.9 — 7.3 3.8 5.6 26.8 19.6 -7.3 — 14.8 — 16.0 4.2 5.9 -1.6 — 8.2 — 14.8 11.6 20.0 -5.6

Georgia -3.7 Germany 22.4 Ghana 9.8 Greece — Guatemala 8.1 Guinea 18.7 Guinea-Bissau 5.0 Haiti -4.5 Honduras 21.8 Hungary 26.9 India 20.0 Indonesia 30.6 Iran Islamic Republic — Iraq — Ireland 33.1 Israel 8.7 Italy 22.3 Jamaica 21.6 Japan 30.5 Jordan 5.5 Kazakhstan 13.5 Kenya 11.4 Korea Democratic Republic — Korea Republic 34.2 Kuwait 25.2 Kyrgyz Republic 13.8 Lao PDR 11.4 Latvia 9.6 Lebanon -16.7 Lesotho -9.8

— 13.2 4.3 8.5 6.0 6.1 4.8 1.5 6.0 8.0 10.0 5.0 15.3 9.0 9.2 13.8 12.4 6.4 15.8 9.7 7.4 6.7 — 10.0 8.6 10.2 5.5 11.7 8.7 8.3

_ _ _ _ _ 9.2 4.4 5.5 2.4 — — 2.1 2.1 12.6 2.3 0.3 1.8 -6.0 2.3 15.8 3.7 18.9 5.2 10.0 4.3 25.6 0.9 — — — — 23.9 5.1 -5.1 6.7 9.9 4.2 15.2 4.4 14.6 5.8 -4.2 3.4 6.1 — 4.7 5.9 _ _ _ _ _ 24.3 3.0 16.6 4.3 3.6 4.3 5.8 1.8 -2.1 6.5 -25.5 1.5 -18.1 4.8

0.0 0.1 0.0 0.0 0.6 0.0 0.0 0.0 0.0 0.4 2.6 3.8 0.0 0.0 0.0 0.0 0.1 0.0 0.0 0.0 18.5 0.0 0.0 0.0 44.5 0.6 0.0 0.0 0.0 0.0

0.0 0.0 2.5 0.0 0.0 18.8 0.0 0.0 0.2 0.1 0.5 0.8 0.0 0.0 0.1 0.1 0.0 12.5 0.0 1.2 0.0 0.0 0.0 0.0 0.0 0.0 1.4 0.0 0.0 0.0

0.0 0.0 0.0 0.0 2.0 0.0 0.0 7.4 0.0 0.0 2.3 0.7 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 8.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0

0.9 0.2 0.4 — 0.2 0.2 0.5 0.1 0.4 0.7 1.6 0.9 —— 0.3 0.3 0.2 1.1 0.1 1.1 5.5 0.4 — 0.5 — 2.3 0.1 1.2 0.5 —

— 13.4 5.0 — 1.4 -4.1 1.5 -11.3 18.8 23.0 7.4 20.5 — — 28.7 1.3 13.9 6.0 20.3 -3.1 -17.9 2.1 — 26.7 -23.6 5.0 6.2 3.1 -24.5 -13.3

Table A3.1. (Continued) Gross Consumptio n Net Educatio n Energ y Minera l Ne t Carbo n dioxide Genuin e domestic expenditur e depletio n depletio n fores t damag e domesti c domestic o f fixed savings capita l saving s depletio n saving s Libya Lithuania Macedonia, FYR Madagascar Malawi Malaysia Mali Mauritania Mauritius Mexico Moldova Mongolia Morocco Mozambique Myanmar Namibia Nepal Netherlands New Zealand Nicaragua Niger Nigeria Norway Oman

— 16.0 3.5 3.6 2.1 44.4 13.6 8.5 24.1 26.4 0.3 17.5 16.8 13.6 — 14.2 10.0 26.3 22.5 2.8 3.3 21.9 — —

— 7.1 — 4.9 6.4 9.3 5.8 8.6 7.7 10.4 — 7.6 7.7 3.6 2.8 13.8 3.0 11.7 9.4 6.3 4.5 2.4 16.4 —

— 8.9 — -1.3 -4.3 35.1 7.8 -0.1 16.4 16.0 — 9.9 9.1 10.0 — 0.4 7.0 14.6 13.1 -3.5 -1.2 19.5 — —

— 4.4 5.7 2.3 3.2 4.8 2.8 4.9 3.1 3.7 5.0 5.9 4.7 3.9 — 1.7 3.4 6.0 5.0 3.9 1.9 0.8 6.7 —

0.0 0.0 0.0 0.0 0.0 4.1 0.0 0.0 0.0 4.8 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.3 0.0 0.1 30.7 5.9 0.0

0.0 0.0 0.0 0.0 0.0 0.1 0.0 14.6 0.0 0.2 0.0 9.6 0.4 0.0 0.0 0.6 0.0 0.0 0.1 0.1 0.0 0.0 0.0 0.0

0.0 0.0 0.0 0.0 5.4 2.1 0.0 0.0 0.0 0.2 0.0 0.0. 0.0 3.7 0.0 0.0 10.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0

— 1.0 — 0.2 0.2 0.7 0.1 1.7 0.2 0.5 2.8 6.2 0.5 0.2 — — 0.2 0.2 0.2 0.9 0.4 1.5 0.2 —

— 12.3 — 0.8 -6.7 32.9 10.5 -11.5 19.3 14.2 — 0.1 13.0 9.9 — 1.5 0.0 20.3 17.4 -0.6 0.2 -12.0 — —

Pakistan Panama Papua New Guinea Paraguay Peru Philippines Poland Portugal Puerto Rico Romania Russian Federation Rwanda Saudi Arabia Senegal Sierra Leone Singapore Slovak Republic Slovenia South Africa Spain Sri Lanka Sudan Sweden Switzerland Syrian Arab Republic Tajikistan Tanzania Thailand Togo Trinidad and Tobago

10.4 32.0 33.2 20.3 20.8 14.5 18.1 — — 14.5 24.7 -7.5 34.6 13.2 -8.0 51.2 28.4 23.1 17.0 21.4 17.3 — 21.3 — 19.0 — 3.4 35.7 9.8 29.1

6.4 7.2 11.0 7.8 4.3 9.0 8.9 4.5 6.6 7.6 19.3 5.6 10.0 5.3 5.7 13.2 15.4 16.9 13.8 11.4 5.0 5.5 13.3 10.5 3.5 5.3 2.8 10.9 5.1 11.2

4.0 24.8 22.2 12.5 16.6 5.5 9.3 — — 6.9 5.3 -13.1 24.6 7.9 -13.8 38.1 12.9 6.2 3.2 10.0 12.3 — 8.0 — 15.5 — 0.6 24.8 4.7 17.9

1.9 5.0 5.8 1.5 3.1 3.1 5.7 5.0 — 3.5 4.1 3.2 5.8 4.1 2.5 2.2 5.0 7.1 6.6 4.8 2.5 2.6 6.6 5.2 2.8 — 2.9 2.9 5.3 4.1

1.1 0.0 6.7 0.0 0.6 0.0 0.6 0.0 0.0 3.3 9.3 0.0 43.6 0.0 0.0 0.0 0.0 0.0 2.1 0.0 0.0 0.0 0.0 0.0 22.5 0.0 0.0 0.2 0.0 10.8

0.0 0.0 8.2 0.0 0.8 0.2 0.3 0.1 0.0 0.1 0.0 0.1 0.0 0.4 3.6 0.0 0.0 0.0 1.9 0.1 0.0 0.0 0.1 0.0 0.1 0.0 0.0 0.0 2.4 0.0

1.5 0.0 0.0 0.0 0.0 1.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

0.8 0.4 0.3 0.2 0.3 0.4 1.7 0.3 — 2.1 1.8 0.2 1.0 0.4 0.2 0.5 1.3 0.4 1.4 0.2 0.2 0.2 0.1 0.1 1.6 — 0.2 0.6 0.3 1.8

2.5 29.5 12.8 13.8 18.0 6.7 12.5 — — 4.9 —1.6 —10.2 -14.2 11.1 -15.1 39.8 16.6 12.9 4.4 14.5 14.6 — 14.4 — -5.9 — 3.2 26.9 7.4 9.4

Table A3.1. (Continued) Gross Consumption Net Education Energy Mineral Net Carbon dioxide Genuine domestic offixed di domestic expernditure devpletion depletion forest damage domestic savings capital savings depletion savings Tunisia Turkey Turkmenistan Uganda Ukraine United Arab Emirates United Kingdom United States Uruguay Uzbekistan Venezuela Vietnam West Bank and Gaza Yemen Republic Zambia Zimbabwe

24.2 19.3 — 7.5 16.3 — 15.1 16.0 12.5 18.6 26.9 21.1 — 12.8 9.8 11.9

8.7 6.5 7.0 5.0 18.4 14.6 10.4 10.7 7.4 4.4 7.1 5.0 — 7.7 9.9 6.0

15.4 12.8 — 2.5 -2.1 — 4.8 5.3 5.1 14.2 19.7 16.2 — 5.1 -0.1 5.9

6.0 3.0 — 2.6 4.6 — 4.5 5.8 2.6 7.7 4.1 1.4 — 3.9 3.8 8.2

2.4 0.3 0.0 0.0 3.0 0.0 0.8 0.7 0.0 8.2 22.5 4.0 0.0 34.7 0.1 0.8

0.7 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.7 0.1 0.0 0.0 1.3 9.9

0.4 0.0 0.0 3.4 0.0 0.0 0.0 0.0 0.4 0.0 0.0 3.4 0.0 0.0 0.0 0.4

0.5 0.5 — 0.1 3.0 — 0.3 0.4 0.1 2.4 1.1 0.8 — — 0.4 1.0

17.3 14.9 — 1.6 —3.4 — 8.2 9.9 7.1 11.4 -0.4 9.2 — -25.7 1.9 2.0

Notes: In certain instances (Benin, for example) there are missing data on educational expenditures, but a genuine saving rate has been calculated. In these instances the genuine saving rate is understated. Source: World Bank (1999) (plus author's estimates derived from this publication for energy exporters and mineral exporters).

Sustainability of Extractive Economies

55

REFERENCES Atkinson, G., R. Dubourg, K. Hamilton, M. Munasinghe, D. Pearce and C. Young (1997), Measuring Sustainable Development: Macroeconomics and the Environment, Chichester: Edward Elgar. Dasgupta, P. and G. Heal (1979), Economic Theory and Exhaustible Resources, Cambridge: Cambridge University Press. El Serafy, S. (1989), 'The proper calculation of income from depletable natural resources', in Y. J. Ahmad, S. El Serafy and E. Lutz (eds.), Environmental Accounting for Sustainable Development, Washington: The World Bank. Fankhauser, S. (1995), Valuing Climate Change: The Economics of the Greenhouse, London: Earthscan. Freeman, A. M. (1994), The Measurement of Environmental and Resource Values. Theory and Methods, Washington, DC: Resources for the Future. Gelb, A. H. and Associates (1988), Oil Windfalls: Blessing or Curse? New York: Oxford University Press. Hamilton, K. (1994), 'Green adjustments to GDP', Resources Policy, 20/3, 155-68. Hamilton, K. (1995), 'Sustainable development and green national accounts', doctoral dissertation, University of London. Hamilton, K. (1996), 'Pollution and pollution abatement in the national accounts', Review of Income and Wealth, Series 42 Vol. 1 (March). Hamilton, K. and M. Clemens (1999), 'Genuine savings in developing countries', World Bank Economic Review, 13/2, 333—56. Hartwick, J. M. (1977), 'Intergenerational equity and the investing of rents from exhaustible resources', American Economic Review, 67, 972-4. Hartwick, J. M. (1990), 'Natural resources, national accounting and economic depreciation', Journal of Public Economics, 43, 291-304. Hartwick, J. M. (1992), 'Deforestation and national accounting', Environmental and Resource Economics, 2, 513—21. Hicks, J. R. (1946), Value and Capital (2nd edn.), Oxford: Oxford University Press. Hill, P. and A. Harrison (1994), 'Accounting for subsoil assets in the 1993 SNA', presented to the London Group on National Accounts and the Environment, London, (15-18 March). M aler, K.-G. (1991), 'National accounts and environmental resources', Environmental and Resource Economics, 1, 1-15. Pearce, D. W. and G. Atkinson (1993), 'Capital theory and the measurement of sustainable development: an indicator of weak sustainability', Ecological Economics, 8, 103-8. Pearce, D., E. Barbier and A. Markandya (1989), Sustainable Development: Economics and Environment in the Third World, London: Earthscan. Pemberton, M. and D. Ulph (1996), Measuring National Income and Measuring Sustainability, Dept. of Economics, University College, London (mimeo). Pezzey, J. (1989), 'Economic analysis of sustainable growth and sustainable development', World Bank Environmental Department Working Paper 15, Washington DC: World Bank. Repetto, R., W. Magrath, M. Wells, C. Beer and F. Rossini (1989), Wasting Assets: Natural Resources in the National Accounts, World Resources Institute, Washington. Solow, R. M. (1974), 'Intergenerational equity and exhaustible resources', Review of Economic Studies, 41, 29-4-5.

56

K. Hamilton

Unesco (1998), Statistical Yearbook 1998, Paris: UNESCO Publishing and Bernan Press. Vincent, J., T. Panayotou and J. Hartwick (1997), 'Resource depletion and sustainability in small open economies', Journal of Environmental Economics and Management, 33, 274-86. Weitzman, M. L. (1976), 'On the welfare significance of national product in a dynamic economy', Quarterly Journal of Economics, 90. World Bank (1997), Expanding the Measure of Wealth, Washington DC: World Bank. World Bank (1999), World Development Indicators 1999, Washington DC: The World Bank.

4 Natural Resources, Human Capital, and Growth N A N C Y B I R D S A L L , THOMAS P I N C K N E Y A N D R I C H A R D SABOT

4 . 1 . INTRODUCTIO N What role does human capital investment play in lowering the growth rates of resource-abundant countries? Most governments around the world extol the benefits of education; investment in this sector is limited, so they say, because of a lack of money. But if limits on human capital investment primarily result from binding government budget constraints, resource abundance should induce additional investment, ceteris paribus. Statistics tell a different story: resource-abundant countries, on average, invest less in education than other countries; this holds true for different indicators, at different times, whether or not one controls for per capita income (Table 4.1). What is the explanation for this negative link between resource abundance and human capital investment? And how important is this link in explaining the lower growth rates of resource-abundant countries? We have argued elsewhere (Birdsall et al. 1999) that countries can achieve rapid, equitable growth by inducing the poor to invest in the assets they control, including their human capital. Raising the rates of return to human capital and other assets leads to increased work effort, higher savings, and higher investment immediately, while generating higher productivity and lower inequality in the future. This lower inequality itself fosters further growth, providing the incentive for yet more savings and investment by the poor. This virtuous circle of human capital accumulation with equitable growth is one important explanation for the rapid development achieved by many East Asian countries. Resource abundance, however, tempts governments to move away from the policies that generate this virtuous circle. Dutch disease and related effects tend to lower the rate of return to the agricultural and human capital investments that are available to the poor. Resource rents tend to accumulate in the hands of the government, and/or a small number of businessmen, leading to high inequality and thus lower growth. Staple trap effects frequently lead to the subsidization of capital, thereby taxing labour. The authors gratefully acknowledge the helpful remarks of Richard Auty and other project participants. The usual disclaimer applies.

58

N. Birdsall, T. Pinckney and R. Sabot Table 4.1. Education and resource abundance

Secondary enrolment, %

Adult literacy, %

Mean

Median

Mean

Median

1975

Resource poor 3.2Resource r Difference Controlled difference

28.5 ch2 2.8 5.7

26.0 19.5 6.5

56.4 .52. 4.2 6.2

61.3 . 8.1

1985

Resouce poor Resource rich Difference Controlled difference

39.5 35.7 3.8 7.4

40.5 34.0 6.5

64.7 60.8 3.9 5.9

72.5 63.4 9.1

Note: 'Controlled difference' takes into account the average impact of income on secondary enrolment and illiteracy by means of regression analysis. Categorization of countries taken from Table 1.1. Source: World Bank (1999) and authors' calculations.

The government may then try to assuage the poor by directing some proportion of resource rents to populist programmes, including the expansion of education. However, schooling in this environment is more of a consumption good than an investment good so that the quality of education is likely to suffer. Moreover, the labour market in a capital-intensive economy offers little benefit for moderate levels of education. The likely result is little investment and dynamism outside of the natural resource sector along with high inequality, poor quality schools, and little demand for education. The rest of this chapter proceeds as follows: First, we relate human capital accumulation to the equitable growth development trajectory. Choices concerning investment in human capital are central to this development path. As part of this section, we lay out a conceptual framework, a set of eyeglasses, as it were, that allows us to understand the decisions millions of individuals make that lead to equitable growth. We then examine the role of human capital accumulation in spurring productivity and in reducing inequality. The reduction in inequality has a further indirect and beneficial impact on growth. This highlights the importance of the government's development strategy in fostering the virtuous circle of human capital accumulation with equitable growth. Second, we investigate the impact of resource abundance on the virtuous circle. Resource abundance tends to break the virtuous circle in several places: the choice of development strategy, the level of inequality, the lack of incentives for investment in education, and the creation of a welfare state. We illustrate this breakdown by contrasting the cases of Korea and Brazil. Finally, we draw out policy lessons for resource-abundant economies. Resource abundance is not destiny because most of the negative impact of resource abundance is mediated through government policies. By gaining an understanding of the virtuous circle of savings, investment, and equitable growth, resource-abundant governments can identify those policies that disrupt the virtuous circle, and seek to avoid them, while pursuing policies that spur equitable growth.

Natural Resources, Human Capital, and Growth

59

4.2. TH E VIRTUOU S CIRCL E O F EDUCATIO N AND EOJUITABLE GROWT H 4.2.2. A conceptual framework Considering the poor as potential investors is the key to understanding the virtuous circle of human capital accumulation and equitable growth. The poor have assets under their control: labour in all cases, with the potential for increased skill that would come about through investment in human capital, and agricultural land in many cases. Investment in these resources, however, may be limited because of either low rates of return or lack of investible capital. Here we lay out a simple conceptual framework for analysing investment decisions by poor households. Consider a poor household with a utility function containing two arguments, consumption and leisure. The household uses its labour to produce a single good. Since time is limited, an increase in work implies a decrease in leisure. How much will the household produce? If the household exists for a single period, the solution is simple: it is to keep working until the marginal utility of an hour of leisure equals the marginal utility of consuming what would be produced in an hour of work. To make this relevant to savings and investment, we must add a second time period. But adding a time period also forces us to take account of the discount rate, the rate of time preference of the household. Given a choice, consumers would prefer to consume today rather than tomorrow. Call the discount rate d. The amount of the good produced in the first year is then subject to a choice: the household can consume all that it produces, or consume less and save/invest the remainder. If the household saves amount s, in the next period it will have (1 + r)s available, where r is the rate of return to investment. Now, assume further that the household is credit-constrained; that is, it cannot borrow today on the basis of expected future earnings. (This assumption together with the absence of shocks in the model implies that savings equal investment in each year.) When will such a household save? Only if r is greater than d. Otherwise, the household is better off consuming all that it produces in the first period. So imagine an economy in which r is less than d. This economy has: • No savings (or, in a stochastic world, only precautionary savings); the household consumes whatever it produces each year; • No increase in income across years; • No growth. Note that we are not assuming any cultural predisposition to laziness. This low level of work effort and no progress over time is a rational response to low returns to investment. Now, assume something changes, making poor households believe that r is greater than d. This could result, for example, from an improvement in local schools, a change 1

See Birdsall et al. (1998, 1999) for a fuller exposition of this framework.

Birpinckneyyandr.sabotsabotdsall, T. pinckney and R. Sabot in agricultural pricing policy, or the introduction of a new seed variety. Now there is an incentive to invest: How can the household finance the investment? Recall that the household cannot borrow. The household has only two ways to raise money: cut back on consumption in the present year, thereby freeing resources for investment, or decrease leisure, allowing more time for work and thus creating more income. As it turns out, a rational household will do both, and invest all of the increase in income. Therefore, when r increases from below d to above d, the rational actor will reduce consumption and increase work effort, and invest all of the savings and increased income. The increase in investment divided by the increase in income is greater than 100 per cent. In these circumstances, the poor contribute to an increase in present GDP by providing more labour, and then take advantage of high-payoff investment opportunities, thereby contributing to an increase in future GDP, all the while helping the country achieve growth with equity. Under what circumstances will this effect be large? For the very poor, the consumption basket is scanty; they are not able to reduce consumption substantially without inducing considerable suffering, or at least limiting the ability of household members to work hard. So for this impact to be substantial, they must be able to get significant returns from working harder. More technically, the marginal product of labour must not fall rapidly as labour supply increases. This provides an indication of the role of government in getting this process moving and allowing it to continue. Governments can encourage this response by achieving two objectives: raising the rate of return to assets held by the poor, and continually improving returns both to labour and investment in skills. Governments that succeed in accomplishing both objectives, as occurred in many East Asian countries, can over time provide the poor with considerably more investible assets, thereby allowing them to take advantage of more high-payoff investment opportunities.

4.2.2. Human capital in the virtuous circle There is a high, positive correlation between income per capita and the accumulation of human capital. For example, expected years of schooling per person in 1995 for low, lower middle, upper middle, and upper income countries were 6.2,10.8, 11.8, and 14.8 years, respectively. But the direction of causality implied by this positive correlation is not clear a priori; education may be an investment good, directly or indirectly improving output in the economy, or it may be a luxury consumer good, increasingly demanded as incomes rise. 4.2.2.1. The direct impact of education on output Both country and firm-level studies provide evidence that education can be an investment good. Barro-style cross-country growth regressions generally show a large and significant impact of initial level of educational attainment on subsequent growth rates, regardless of the specification of the equation or method of measuring education 2

Authors' calculations from data in World Bank (1999).

Natural Resources, Human Capital, and Growth

61

(Levine and Renelt 1991; but see Pritchett 1996 which challenges these studies). Birdsall el al. (1995), for example, estimate that countries with primary and secondary enrolment rates one-half standard deviation above the mean had annual growth rates 1.4 per cent higher than countries with enrolment rates one-half standard deviation below the mean. Over 25 years, such a difference in growth rates leads to a 40 per cent difference in per capita incomes. Human capital theory predicts that education augments cognitive and other skills which in turn, augment the productivity of labour (Becker 1964; Schultz 1961). Endogenous growth theory predicts the same: a larger stock of human capital facilitates technological progress or, for a country not on the technological frontier, relatively rapid acquisition of technological capability (Nelson and Phelps 1966; Romer 1990). Empirical microeconomic studies verify the theory: both in agriculture (Pinckney 1997; Jamison and Lau 1982) and nonagriculture (Knight and Sabot 1990; Maluccio 1998), numerous studies show that increases in human capital have a substantial direct impact on returns to labour. For example, Maluccio's study of a poor area in the rural Philippines finds that each year of education raises wages approximately 10 per cent. For urban labour markets in East Africa, Knight and Sabot find that the human capital accumulation that takes place during secondary school increases wages by approximately 25 per cent, even after taking account of the credential effect of schooling as well as the impact of innate ability. The impact of human capital on agricultural productivity is especially large in environments where new agricultural technology is available. For example, Jamison and Lau (1982) find that four years of schooling on average raise farm output by 7.2 per cent, but in dynamic agricultural environments the increment increases to 9.5 per cent. Thus, strong evidence supports the view that education improves productivity, both in agriculture and nonagriculture. 4.2.2.2. The indirect impact of education on output In addition to improving productivity directly, human capital investment also has an indirect impact on national per capita income. There are two paths: first, educational expansion can reduce inequality, which then improves growth (Birdsall et al. 1995); second, education of women has a negative impact on fertility and a positive impact on child health (Summers 1992), providing an important catalyst for demographic changes associated with increased savings and investment, as well as more rapid economic growth.3 We focus here only on the first of these paths. Education for the poor can reduce inequality most obviously by raising the productivity of the substantial number of poor who manage their own farms or informal-sector firms. But education also has an impact on the inequality of pay for As3a discussed in chapter 2, recent analyses contend that a change in age composition towards a proportionately larger working-age population resulting from the combination of declining mortality followed (with some lag) by declining fertility has contributed to higher rates of economic growth. See Bloom and Williamson (1998) as well as the summary of these analyses found in Birdsall and Sinding (1999); the latter refers to Bongaarts (1998), Kelley and Schmidt (1998), and Williamson (1998b), among others.

62

N. Birdsall, T. Pinckney and R. Sabot

formal-sector employees. Inequality of pay accounts for two-thirds of total income inequality in rich countries, and an increasing share of total inequality over time in poorer countries. The impact of education on inequality of pay is ambiguous theoretically. There can be a disequalizing composition effect as an initially small portion of the population that is better educated and better paid increases in size. To take a simple example, suppose all workers are either educated or uneducated, and educated workers receive five times the pay of uneducated workers. If the percentage of educated workers increases from 10 to 20 per cent of the labour force holding the education wage premium constant, the Gini coefficient in income inequality increases from 0.26 to 0.36. As more of the labour force is educated, however, the scarcity rents (signalled by the wage premium) paid to such workers erode. If the increase in supply of educated workers causes the pay of educated labour to fall to only 2.5 times that of uneducated labour, the Gini coefficient falls from 0.26 to 0.18; income is more equal than before. The resulting compression of the educational structure of wages tends to reduce inequality of pay. In other words, if the spread of education to members of the labour force is rapid enough, and the scarcity rents of those with relatively more education rise little, then the equalizing compression effect is more likely to offset the disequalizing composition effect. Whether the compression or composition effect dominates in any particular country depends on the nature of the educational system and the country's development strategy. When the country pursues a labour and skill-intensive growth path in conjunction with expanding opportunities for high-quality education, demand and supply of skilled labour can both expand rapidly, as seen in East Asia. In such cases, the compression effect will often outweigh the composition effect. On the other hand, if the country pursues a capital-intensive growth path which pays a small number of workers high salaries, and allows the quality of the bulk of its education sector to deteriorate, there may not be any compression effect at all. Brazil and Korea exemplify these two possibilities. From 1975 to 1985 secondary enrolment in Brazil increased from 26 to 35 per cent, and in Korea from 56 to 92 per cent. However, for reasons we explore later in this chapter, there was no compression effect on wages of educational expansion in Brazil; the composition effect dominated, thus increasing wage inequality by four per cent. In contrast, during the same years in Korea, the premium paid to workers with a high school education fell from 60 to 34 per cent over a primary graduate; the premium paid to those with higher education fell from 164 to 93 per cent. This large compression effect dominated the composition effect of educational expansion, leading to decrease in wage inequality of 22 per cent. Therefore, expanding opportunities for high-quality education coupled with a labour-demanding growth path can lead to reductions in inequality. But does inequality have an impact on growth? Our conceptual framework suggests that, holding per capita income constant, the poor will be able to afford more high-payoff investments if inequality is lower. Clearly if the lowest income quintile receives only 2 per cent of GDP, as in Brazil, they have considerably less opportunity to invest than if they receive 7 per cent of GDP, as in Indonesia (Table 4.2). Given that investments in human and other forms of capital

Table 4.2. Income distribution, education, and growth in Latin America, Asia, and sub-Saharan Africa Country

Percentage of national income

Secondary Per capita GNP enrolment (growth rate (1980) 1965-97)

Poorest Richest 20o/o(l()2) 20%(2)

Quotient (2/1)

Asia Resource poor Bangladesh China Hong Kong Philippines Singapore South Korea Sri Lanka Taiwan Unweighted average

9.4 5.5 5.4 5.9 6.5 5.7 8.9 7.1 6.8

37.9 47.5 47.0 49.6 49.2 45.3 39.3 39.6 44.4

4.0 8.6 8.7 8.4 7.6 7.9 4.4 5.6 6.9

18 46 64 64 58 78 55 80 58

1.4 6.8 5.7a 0.9 6.3 6.7 3.0 8.5 4.9

Resource rich Indonesia Malaysia Thailand Unweighted average

6.6 3.5 5.6 5.2

49.4 56.1 49.8 51.7

7.4 16.0 8.9 9.9

29 48 29 35

4.8 4.1 5.1 4.7

Latin America Resource poor Colombia El Salvador Haiti Unweighted average

3.1 3.7 n.a. 3.4

61.5 54.4 n.a. 58.0

19.8 14.7 n.a. 17.3

39 24 14 26

2.1 -0.5 —0.7 0.3

Resource rich Argentina Bolivia Brazil Chile Costa Rica Mexico Peru Venezuela Unweighted average

4.4 5.6 2.0 4.2 4.0 2.9 1.9 3.0 3.5

50.3 48.2 66.6 60.4 51.8 57.7 61.0 54.0 56.3

11.4 8.7 33.3 12.2 13.0 19.9 32.1 18.0 16.1

22 37 33 53 48 49 59 21 40

0.3 -0.5b 2.3 1.7 1.1 1.5 -0.4 -0.9 0.6

Sub-Saharan Africa Resource poor Kenya 5.0 Tanzania 6.8 Unweighted average 5.9

50.2 45.5 47.9

10.0 6.7 8.4

20 3 11.5

1.3 n.a. n.a.

64

N. Birdsall, T. Pinckney and R. Sabot Table 4.2. (Continued)

Country

Percentage of national income Poorest 20% (1)

Richest 20% (2)

Quotient (2/1)

6.8 7.9 3.0 5.8 4.0 3.5 3.3 6.8 3.9 4.0 4.9

44.1 42.2 50.2 50.0 49.4 58.6 63.3 48.8 54.5 62.2 52.3

6.5 5.3 16.7 8.6 12.4 16.7 19.2 7.2 14.0 15.6 12.2

Secondary Per capita GNP enrolment (growth rate (1980) 1965-97)

Resource rich C o te d'lvoire Ghana Guinea Madagascar Nigeria Senegal South Africa Uganda Zambia Zimbabwe Unweighted average

19 41 17 n.a. 16 11 n.a. 5 16 8 17

-0.9 -0.9 n.a -1.9 0.0 -0.5 0.1 n.a -2.0 0.5 -0.7

a

Indicates datum is GDP per capita. Indicates years are 1965-96. No adjustment is made here for the impact of income on secondary enrolment; see Table 4.1 for those adjustments.

b

Source: World Bank (1999), except Taiwan, from Deininger and Squire (1996) and CEPD (1998). Categorization of countries taken from Auty (1997).

controlled by the poor have the potential for very high returns, countries with greater income inequality miss out on this source of growth. But there are other possible ways in which low income inequality can spur growth. First, low inequality of income can accelerate growth by enhancing political and macroeconomic stability. Lower inequality reduces the tendency for fiscal prudence to be sacrificed to political expediency,4 while simultaneously minimizing the risk that a large proportion of the population becomes politically alienated. A more stable political and macroeconomic environment is conducive to economic growth: investment will be higher where the risks of economically disruptive political upheaval and runaway inflation are lower. These benefits accrue when reduced inequality results from expanded education and increasing returns to assets held by the poor. They may not accrue when low inequality has been purchased through transfers, subsidies, and public sector employment policies, for three reasons. First, expenditures such as these tend to grow over time, taking an increasing proportion of the government budget, thereby increasing the likelihood of economic instability (Gelb et al. 1991). Second, at some point the increase in transfers will become unsustainable; their reduction or removal 4 Persson and Tabellini (1995) argue that where distributional conflict is greater, the incentives for the accumulation of knowledge by individuals are weakened. Alesina and Rodrik (1994) argue that democracies with a more unequal distribution of capital ownership will grow less rapidly than more egalitarian democracies because the median voter has a relatively small endowment of capital when wealth is unequally distributed and thus favours high taxes on capital.

Natural Resources, Human Capital, and Growth

65

increases the likelihood of instability. Third, educational expansion in this context can lead to the destabilizing force of frustrated young people, educated and having a subsistence income, but having no outlet for their training. Second, low inequality can raise growth by increasing the x-efficiency of low-income workers. Children from low-income households are unlikely to work hard if, no matter how great their effort, they have no chance of gaining scarce places in the university system. By contrast, poor children in high quality schools who see tangible rewards for effort are more likely to make such an effort. Similarly, extra effort is unlikely to be forthcoming from low-income workers or farmers who, because of policy biases, face economic incentives that do not reward effort. They are more likely to shirk and to become resentful and alienated. By contrast, if economic incentives do reward effort, they are likely to respond. Although difficult to quantify, the increases in productivity associated with low inequality may, nevertheless, be large. Third, low inequality spurs growth by expanding the multiplier effects of a given increase in per capita income. Many of the poor reside in rural areas and participate in the agricultural sector. Higher incomes for the poor generate higher demand for non-tradable agricultural inputs and consumer goods, stimulating production of these items. In general, investment and consumption goods demanded by the poor have a much higher non-tradable component than those demanded by the rich. This has led, for example, to the strong positive correlation in Asia between the rate of growth of the agricultural sector and the rate of growth of the nonagricultural sector (Mellor 1995). The estimated multiplier effects of agricultural growth on manufacturing, construction and services are large: a 1 per cent increase in agricultural growth is associated with a 1.5 percentage increase in the growth rate of the nonagricultural sector.5 Because the relatively simple manufactured inputs and consumer goods demanded by rural residents are generally more efficiently produced with labour intensive techniques, the employment effects of these increases in demand are amplified.6 By contrast, when the incomes of the urban elite increase, the tradable goods on which they spend their increased income tend to be capital intensive goods.

4.2.3. The human capital/equitable The role of government

growth virtuous circle:

Figure 4.1 displays key parts of the virtuous circle we have described. If the rate of return to human capital investment rises above the discount rate, the poor will increase work intensity to finance these investments, leading to growth in the short run. Even this short run growth will have some impact on the rate of return to 5 The relationship also implies that the faster agriculture grows, the faster its share of total output declines. Again with the exception of the East Asian city states, those Asian countries with the fastest rates of growth of agricultural output over the last 30 years have tended to experience the biggest declines in the share of agricultural output in GNP. 6 See Ranis and Stewart (1987). A detailed study of these backward and forward linkages in the Muda River region of Malaysia provides microeconomic confirmation of the magnitude of the intersectoral multiplier suggested by the cross-country relationship. See also Bell et al. (1982).

66

N. Birdsall, T. Pinckney and R. Sabot

Figure 4.1 . Government's positive impact on the virtuous circle

additional human capital investment. Demand for schooling will rise, and—if schools are available—human capital accumulation will take place.7 This leads to reductions in inequality given proper government policy, which both spurs growth in the long run and increases further the demand for schooling (by providing more resources for the poor). The growth itself once again feeds into higher rates of return to human capital investment, continuing the circle. The lower circuit in Fig. 4.1 traces the government's impact on this virtuous circl8e. It will be necessary for the government to supply schooling to the poor in most circumstances. The poor will find investment in education difficult if there are no schools, or if existing schools are full. But classrooms and teachers are not enough. It is vital that high quality educational services are provided to the poor if human capital accumulation is to take place. In some environments, education is seen as a plum awarded by the government, an enhancement of consumer (and voter) satisfaction. If that is the only role played by education, if education is almost exclusively a consumer good, then quality may matter little. But if the poor view education as a form of investment in human capital, then, as our diagram shows, perceived quality is an important determinant of educational demand9. 7 In some conditions, as in Kenya in the 1960s and 1970s, the demand for education will be so great that the poor will create their own schools when the state fails to supply educational services. See Thomas (1981) and Knight and Sabot (1990). 8 We have included in the figure the impact of government on human capital accumulation; clearly government affects growth and inequality in many additional ways. 9 Behrman and Birdsall (1983) and Behrman et al. (1994) find that the social rate of return to improving primary school quality exceeds the rate of return to increasing the available quantity of either primary or middle school places. With regard to improving school quality, the production function results suggest that larger gains can be made by investments that improve teacher quality (by increasing their cognitive skills) and increase student exposure to teachers (by reducing class size) than by investments that improve physical infrastructure and equipment.

Natural Resources, Human Capital, and Growth

67

Governments can build schools and maintain their quality, but the poor will not invest in education unless the perceived payoff warrants the financial and opportunity costs of sending a child to school. The conceptual framework in Fig. 4.1 highlights the need for the return to investment to exceed the discount rate. Government has a major impact on this return through its choice of development strategy. A strategy that focuses on increasing demand for labour and for skills in an internationally competitive economy will raise the rate of return to human capital investment; whereas a set of policies that biases incentives against manufactured exports and agriculture will have the opposite effect. We thus have outlined a human capital/equitable growth virtuous circle that can be spurred and maintained by appropriate government policies. Government works to expand efficient demand for skilled labour, to provide opportunities for high quality universal education, and to generate other profitable investment opportunities for the poor. The poor then respond by working harder at present to finance these investments—including the investment in education—thereby raising growth rates in the short and long run. Over time this leads to higher incomes for the poor, and more investment by them. Growth is generated by high-payoff investments by the poor. This virtuous circle is related to the competitive industrialization model of resource-poor countries outlined in Chapter 8. Industrialization with a rapidly expanding labour force and rising real wages took place in resource-poor East Asian countries in part because of successful land reforms and agricultural research that induced productivity-enhancing investments in agriculture. With no labour 'sink', and with the drive for international competitiveness keeping wages in line with productivity, there were large incentives for everyone involved to invest in high quality education. The poor were active participants in this process.

4.3., RISK S POSE D T O TH E VIRTUOU S CIRCL E B Y RESOURCE ABUNDANC E Resource-rich countries face several challenges in generating this virtuous circle. We here examine those challenges, using the contrasting examples of Korea and Brazil to highlight the problems.

4.3.1. The risks to the virtuous circle Figure 4.2 details some of the impacts of resource abundance on the virtuous circle. First, consider inequality, which as noted in Chapter 2 tends to be higher in resourceabundant economies. This stylized fact held even 100 years ago, as noted by Williamson (1998a). In 'point' resource-abundant countries, resource rents are controlled by a small number of persons or firms. Owners of the firms or the government officials concerned can capture the rents. Furthermore, Dutch disease effects also tend to concentrate income in the hands of those with investments in the booming sector, as incomes fall for producers of other tradable goods. Both of these impacts are most obviously the case in some mineral exporters, such as Nigeria. With initial high

68

N. Birdsall, T. Pinckney and R. Sabot

Figure 4.2. Impact of resource abundance on the virtuous circle

inequality, the effects of the virtuous human capital/equitable growth circle are necessarily of less importance. Resource constraints are binding at lower levels, limiting the ability of the poor to invest even if rates of return exceed their discount rates. Other impacts of resource abundance are mediated through government policy. First, as discussed more fully in Chapter 8, resource abundance tends to push countries towards a staple trap, often leading to the closure of trade policy and the adoption of a capital-intensive, inward-looking development strategy. Dutch disease impacts are important; when the price booms for the major export, unsterilized foreign exchange increases will cause the real exchange rate to appreciate, with negative (and possibly devastating) effects on all other producers of tradable goods. Thus, in Fig. 4.2 the Dutch disease effects have a negative impact on the choice of a labour-demanding development strategy. The story is more complex than can be shown in the diagram, however. The appreciated exchange rate makes imported capital cheaper while biasing incentives against exports. This encourages the government to attempt import substitution based industrialization via the importation of capital, rather than following an export oriented strategy based on many small domestic investments, or through reliance on local unskilled or semi-skilled labour. Governments frequently exacerbate the problem by making imported capital even cheaper through duty-free importation of capital equipment, investment tax breaks and tax holidays for foreign investors.

Natural Resources, Human Capital, and Growth

69

The result is a high capital/labour ratio, inefficient production, an expanding production base, but little employment. The appreciated exchange rate also hurts what is generally the largest employer in the country, the agricultural sector. Agriculture tends to be more highly tradable than other sectors, and thus is hurt most by appreciation. In many countries, agricultural land and labour are the two most important assets of the poor, in which the virtuous human capital/equitable growth circle induces investment. Resource abundance, in these circumstances, tends to lower the rate of return to investment in both assets, thereby impoverishing rather than enriching the poor. The appreciation also hurts nonagricultural producers of exports and importcompeting goods. Countries sometimes respond to the resulting outcry from local producers by raising tariff walls and encouraging capital-intensive, inefficient production of tradable goods for the domestic market. But the resulting industrialization employs few people, tends to exacerbate inequality, and offers little scope for the development of an efficient entrepreneurial class. Manufactured exports, one of the most important driving forces behind the labour-demanding growth path in the East Asian economies, have a difficult time even getting started in this environment. Resource abundance thus tends to concentrate income and assets, while reducing efficient employment and investment opportunities for the poor. This leads to a political problem: how does the government assuage the majority of the people, who see much money made from their country's resources, but benefit little themselves? The logical response is to provide benefits to the people from the resources gained. While in and of itself this is not problematic, the role of government becomes much different from that seen in the conceptual framework of Fig. 4.1. There, the government is concerned with creating an environment in which there are large returns to the assets of the poor. The poor then invest, and through their hard work and improvements in productivity achieve growth with equity. In resource-abundant countries, however, the role of the government can easily become exclusively a provider of welfare goods, with the poor—indeed with much of the populace—as beneficiaries. The government's role is then to extract resources from those involved in tapping the natural resources, and turning over what is extracted to the people, frequently through make-work jobs or inefficient development projects.10 The government becomes Santa Claus, dispensing its largesse on the people, who in consequence demand more and more. The impact of the Santa Claus method of allocation of rents is most important in the education sector, as shown in Fig. 4.2. We have already seen that the predominant development strategies adopted in resource-abundant countries do not raise the rate of return to human capital investment. The most highly-skilled workers in these economies may receive a hefty wage premium, as they compete with foreign workers for the best salaries in the resource-based sector, or for management positions in protected, import-competing firms. But the child from a poor family who completes 10 Gelb et al. (1991), in an article originally entitled 'Lewis Through the Looking Glass', show how a large, unproductive labour sink in the formal sector can have devastating impacts on growth.

70

N. Birdsall, T. Pinckney and R. Sabot

primary or lower secondary school reaps few benefits; agriculture is moribund, and the manufactured exports sector, so important in many countries for hiring semi-skilled workers, is nonexistent. Therefore, even if schools are of high quality, there may be little return to investment in human capital. Furthermore, if a Santa Claus government has created schools in effect as a palliative, as a consumption good to assuage the disgruntled poor, these schools are unlikely to be effective at imparting knowledge. Thus, the quality of schooling suffers, as Fig. 4.2 indicates. At least for a while, even poor schools can serve their purpose simply by dispensing school-leaving certificates, giving a credential but not producing human capital. So low quality schools abound in poor regions, lowering yet further the expected rate of return to investment in human capital. Ironically, when the resource sector boom ends, this poor quality education leaves the country without the welltrained workers or entrepreneurs required for developing efficient, non-resource based firms (see Chapter 12 on Saudi Arabia for a good example). Not all resource-abundant countries face the same risks. Land rich countries in which land is inequitably distributed—those that pursue a bimodal agricultural strategy, to use Johnston's (Johnston and Kilby 1975) term—tend to behave much like mineral-rich countries. Thus, both such economies are characterized by resources with 'point' socioeconomic linkages. The resource rents are concentrated, leading to high inequality; governments are unlikely to pursue a labour-demanding growth path, leading to low returns to investment in human capital; school quality is unlikely to be a government priority. Unless government resists these temptations, the virtuous human capital/equitable growth circle cannot begin moving11. But, on the other hand, if land is equitably distributed—Johnston's (Johnston and Kilby 1975) unimodal agricultural strategy, as in 'diffuse' resource-abundant economies—the country faces fewer of the same temptations as mineral-rich economies. In particular, incomes are likely to be more equitably distributed and if the agricultural sector is dynamic, investment in education may have high returns. While the country still faces the temptation not to diversify its export base, and thus can be subject to Dutch disease, generating the virtuous human capital/equitable growth circle we describe, is considerably easier for 'diffuse' compared to 'point' resource-abundant economies.

4.3.2. The examples of Brazil and Korea Let us now turn to the examples of Brazil, which is land- and resource-abundant, but with an unequal distribution of land and other assets, and Korea, which is resource-poor. While a single product did not dominate Brazil's export base, nevertheless for reasons elaborated in Chapter 13 with regard to Argentina and Mexico, as a large country it pursued an inward-looking development strategy. This reinforced the country's unequal initial distribution of income because it generated little demand for 11

Birdsall and Londono (1998) show the effects of this cycle for education in Latin America. See also Spilimbergo et al. (1997) on income distribution and factor endowments in Latin America.

Natural Resources, Human Capital, and Growth

71

moderately-skilled workers, reducing the return to, and thus the demand for, schooling (Table 4.2). By contrast, East Asia's outward-oriented competitive industrialization strategy provided strong incentives for the growth of manufactured exports. These industries initially generated demand for unskilled labour, but as wages rose they became increasingly skill intensive. Korea and Brazil had roughly the same shares of exports to GDP in 1960. By the mid-1980s, Korea's share of exports to GDP had increased from 9 to 37 per cent while Brazil's remained at 8 per cent. High tariffs and quotas protected the Brazilian market from international competition while an overvalued exchange rate weakened incentives to export. Export orientation increases the demand for skilled workers—and thus the household demand for education—in two ways. First, the higher growth rate of aggregate output in an export-oriented economy increases the growth rate of the overall demand for labour. Second, an export-oriented development strategy leads, as wage levels rise, to more skill-intensive production. In East Asia, high returns to schooling, driven by the demand for skills from the export push, kept the incentives for acquiring education strong. This led to a positive interaction between improved educational performance and faster export-oriented growth (Birdsall and Sabot 1996). In Brazil, the limited demand for educated workers in the protected labour market interacted with the poor quality of schooling and the high level of income inequality to weaken the household demand for education. Brazil's import-substitution strategy initially appeared to be a success, creating a temporary boom in the 1960s and 1970s that masked the strategy's longer-term implications. The strategy's negative effects on the accumulation of human capital were therefore protracted (Pastore and Zylberstajn 1996). Besides stunting the demand for labour, Brazil's inward-looking economy used human capital less efficiently than did the East Asian economies. For human capital accumulation to pay off, those leaving the school system must find jobs where their skills are used most productively. Because Brazil did not need to compete internationally, employers and government put up less resistance to rent-seeking by workers. Organized demands for high wages and excessive levels of employment led to the inefficient use of human capital and ultimately became a drag on growth (Robbins and Minowa 1996). Labour extracted rents by raising wages to levels above the supply price, or by raising employment to levels above those justified by the derived demand for labour. To recoup revenue to offset the higher costs, private and public firms could raise the prices they charged in their protected domestic market. Alternatively, public enterprises could hold the line on prices by receiving increased subsidies from the government, thereby raising the tax burden. Both phenomena tended to lower rates of saving and investment and to reduce the international competitiveness of Brazil's manufacturing enterprises (Birdsall and Sabot 1995). With lack of dynamism in Brazil's labour market in the late 1970s and the 1980s, workers in the organized sector tended to resort to rent-seeking, thereby maintaining high wage premiums for educated workers. In East Asia, however, labour was much less successful in organizing in the 1960s and 1970s. In Korea and Singapore, there had been political suppression of labour movements. More importantly, throughout

Birdsall, T. PinckneN. birdsall, ti pinckeney and r. sabottt East Asia there were regular and large increases both in real wages and in alternative employment opportunities. The dynamism of the labour market, stimulated by an export-oriented economy, reduced incentives for rent-seeking by workers. All this serves to explain the large difference in enrolment, dropout, and repetition rates between Brazil and Korea, even in the early 1970s when the Brazilian economy was booming. Korea's enrolment ratio at this time was almost double Brazil's. Korea's export-oriented, labour-demanding growth strategy raised the marginal return of the labour of the poor, making it attractive to increase the time allocated to work to finance high-return investments in the workers' children. Public policy also ensured high quality schooling, even in poor districts, thereby contributing to the high rates of return to investment in schooling. The labour-demanding growth path became increasingly skill-intensive over time—yet another factor contributing to high expected rates of return to schooling and hence strong household demand. In sum, in Korea there were strong incentives for the poor to invest in their children and to work more to finance that investment. It is reasonable to suppose that marginal savings rates among the poor were exceptionally high. By contrast, in Brazil the inward-looking growth strategy was not labourdemanding and thus the returns to additional work effort were quite low. In addition, school quality for the poor tended to be abysmal. Although the average returns to investment in schooling were high, for the poor, returns to investment in schooling were low. Lack of dynamism in the demand for labour and skills also held down expected returns to investment in schooling. In sum, Brazilian public policy created incentives for high levels of leisure and low levels of saving among the poor.

4.4. POLIC Y IMPLICATION S FO R RESOURCE-ABUNDANT COUNTRIES Resource abundance generally leads to high initial levels of income inequality, making it more difficult to generate a virtuous circle of human capital/equitable growth. Yet, as Fig. 4.2 shows, the other impacts of resource abundance are mediated through government policy. Consequently, if governments can resist the temptation to ignore the quality of schooling and the temptation to pursue a capital intensive growth strategy, the virtuous circle is achievable. Indonesia demonstrates this. Despite substantial windfalls from the 1973-78 and 1979-81 oil booms, Indonesia managed to avoid the policy traps we describe, at least in part by trying to follow the development examples of Taiwan, Korea, and Japan. Indonesia successfully avoided most Dutch disease effects by sterilization and nominal devaluations. It also pursued a unimodal agricultural strategy, together with considerable investment in agricultural research; and it spurred the development of a relatively low-capital, manufactured export sector. Education expanded rapidly, while poverty fell. The problems the country experienced in the late-1990s resulted, in part, from departures from this successful development strategy. Other countries can also use resource abundance to their advantage by carefully managing the macroeconomic effects of commodity booms. They can use revenue

Natural Resources, Human Capital, and Growth

73

raised from resources to improve the rate of return to the investment options available to the poor—especially in human capital—instead of increasing the availability of consumption goods directly. They can invest in high quality education, viewing schooling as an investment good. In this way, the virtuous circle of rapid human capital accumulation with equitable growth is achievable by all countries. Certainly, resource abundance creates more challenges in getting the virtuous circle moving, but if the governments of those countries can overcome the challenges, the poor can become an engine of growth, and not only its beneficiaries.

REFERENCES Alesina, A. and D. Rodrik (1994), 'Distributive policies and economic growth', Quarterly Journal of Economics, 109/2, 465-90. Auty, R. M. (1997), 'Natural resources, the state and development strategy', Journal of International Development, 9, 651-63. Becker, Gary S. (1964), Human Capital: A Theoretical and Empirical Analysis, Princeton, NJ: Princeton University Press. Behrman, J. and N. Birdsall (1983), 'Quality of schooling: quantity alone is misleading', American Economic Review, 73, 5 (December). Behrman, J. R., D. Ross, R. Sabot and M. Tropp (1994), 'Improving the quality versus increasing the quantity of schooling', Williamstown, MA: Williams College Center for Development Economics Research Memoranda, RM-140 (May). Bell, C., P. Hazell and R. Slade (1982), Project Evaluation in Regional Perspective, Baltimore, MD: Johns Hopkins University Press. Birdsall, N. and J. L. Londono (1998), 'No tradeoff: efficient growth via more equal human capital accumulation', Chapter 5 in Nancy Birdsall, Carol Graham and Richard H. Sabot (eds.), Beyond Tradeoffs: Market Reform and Equitable Growth in Latin America, Washington: Brookings Institution and Inter-American Development Bank. Birdsall, N. and R. H. Sabot (1995), Virtuous Circles, Human Capital Growth and Equity in East Asia, (mimeo), Washington DC: Inter-American Development Bank. Birdsall, N. and R. Sabot (eds.) (1996), Opportunity Foregone: Education in Brazil, Washington: Johns Hopkins University Press and Inter-American Development Bank. Birdsall, N. and Steven Sinding (1999), 'How and why population matters: new findings, new issues', in Nancy Birdsall, Allen Kelley and Steven Sinding (eds.), Demography Matters: Population, Economic Growth, and Poverty in the Developing World, processed, Washington: Carnegie Endowment for International Peace. Birdsall, N., D. Ross and R. Sabot (1995), 'Inequality and growth reconsidered', World Bank Economic Review, 9, 3 (September), 477—508. Birdsall, N., T. Pinckney and R. Sabot (1998), 'Why low inequality spurs growth: savings and investment by the poor', Chapter 5 in A. Solimano (ed.), Social Inequality: Values, Growth, and the State, Ann Arbor: University of Michigan Press, 68-81. Birdsall, N., T. Pinckney and R. Sabot (1999), 'Equity, savings, and growth', Brookings Institution Center on Social and Economic Dynamics Working Paper 5, Washington DC. Bloom, D. E. and J. G. Williamson (1998), 'Demographic transitions and economic miracles in emerging Asia', The World Bank Economic Review, 12, 419-55.

Birdsalland R. SaboNt birdsall,t. tinckney and r. sabot Bongaarts, John (1998), 'Dependency burdens in the developing world', presented at the Symposium on Population Change and Economic Development, Bellagio, Italy (November 2-6). CEPD (1998), Taiwan Statistical Data Book 1998, Taipei: Council for Economic Planning and Development. Deininger, K. and L. Squire (1996) 'A new data set for measuring income inequality', World Bank Economic Review, 10/3, 365-91. Gelb, Alan, John B. Knight and Richard H. Sabot (1991), 'Public sector employment, rent seeking and economic growth', Economic Journal, 101/408 (September), 1186-99. Jamison, Dean T. and Lawrence J. Lau (1982), Farmer Education and Farm Efficiency, Baltimore: Johns Hopkins University Press. Johnston, Bruce F. and Peter Kilby (1975), Agriculture and Structural Transformation, New York: Oxford University Press. Kelley, Allen C. and Robert M. Schmidt (1998), 'Economic and demographic change: a synthesis of models, findings, and perspectives', presented at the Symposium on Population Change and Economic Development, Bellagio, Italy (2-6 November). Knight, John B. and Richard H. Sabot (1990), Education, Productivity, and Inequality: The East African Natural Experiment, New York: Oxford University Press. Levine, R. and D. Renelt (1991), 'A sensitivity analysis of cross-country growth regressions', World Bank Working Paper Series 609, Washington DC: World Bank. Maluccio, John (1998), 'Endogeneity of schooling in the wage function: evidence from the rural Philippines', Food Consumption and Nutrition Division Discussion Paper 54, Washington: International Food Policy Research Institute. Mellor, John (ed.) (1995), Agriculture on the Road to Industrialization, Baltimore: Johns Hopkins University Press. Nelson, Richard and Edmund Phelps (1966), 'Investment in humans, technological diffusion, and economic growth', American Economic Review (Proceedings), 56, 69-75. Pastore, J. and H. Zylberstajn (1996), 'Social mobility: the role of education in determining status', Chapter 10 in N. Birdsall, and R. Sabot (eds.), Opportunity Foregone: Education in Brazil, Washington: Johns Hopkins University Press and Inter-American Development Bank. Persson, T. and G. Tabellini (1995), 'Is inequality harmful to growth?', American Economic Review, 84/3, 600-21. Pinckney, Thomas C. (1997), 'Does education increase agricultural productivity in Africa?', in Issues in Agricultural Competitiveness: Markets and Policies, IAAE Occasional Paper 7, edited by R. Rose, C. Tanner and M. Bellamy. Brookfield, VT: Dartmouth Publishing Co. Pritchett, L. (1996), 'Where has all the education gone?' Policy Research Department Working Paper 1581, Washington: World Bank. Ranis, G. and F. Stewart (1987), 'Rural linkages in the Philippines and Taiwan', in F. Stewart (ed.) Macro-Policies for Appropriate Technology in Developing Countries, Boulder, CO: Westview Press. Robbins, D. and M. Minowa (1996), 'Do returns to schooling vary across industries?' Chapter 7 in N. Birdsall and R. Sabot (eds.), Opportunity Foregone: Education in Brazil, Washington: Johns Hopkins University Press and Inter-American Development Bank. Romer, Paul (1990), 'Endogenous technological change', Journal of Political Economy, 98, S71S102. Schultz, T. W. (1961), 'Investment in human capital', American Economic Review, 51/1, 1-17.

Natural Resources, Human Capital, and Growth

75

Spilimbergo, A., J. Londono and M. Szekely (1997), 'Income distribution, factor endowments and trade openness', Office of the Chief Economist Working Paper 356, Washington: Inter-American Development Bank. Summers, Lawrence H. (1992), 'Investing in all the people', Pakistan Development Review, 31/4, 367-93. Thomas, Barbara P. (1981), 'Local organization, politics and participation in rural development: a study of harambee self-help in Kenya', Ph.D. Dissertation, Brandeis University. Williamson, Jeffrey G. (1998a), 'Growth, distribution, and demography: some lessons from history', Explorations in Economic History, 34/3, 241-71. Williamson, Jeffrey G. (1998b), 'Demographic change, economic growth and inequality', presented at the Symposium on Population Change and Economic Development. Bellagio, Italy (2-6 November). World Bank (1999), World Development Indicators 1999, Washington DC: World Bank.

5 The Social Foundations of Poor Economic Growth in Resource-Rich Countries MICHAEL WOOLCOCK, LANT PRITCHETT AND J O N A T H A N ISHAM

The poor performance of countries with abundant natural resources can be explained using orthodox economic tools, but a central contention of this study is that a more comprehensive picture emerges when social and political factors are also taken into consideration. Since World War II, the economies of countries rich in natural resources have typically displayed a growth trajectory resembling a 'mountain' (Pritchett 1998)—that is, solid performance during times of general global prosperity, followed by a spectacular downturn in the wake of exogenous shocks. Weak public institutions, political instability, inequitable distributions of wealth, and high levels of corruption appear to be common features of these economies (Karl 1997). Not surprisingly, these factors typically manifest themselves in slow overall growth rates, as described in earlier chapters, or at worst, spill over into a plundering of national wealth, and gruesome episodes of violent civil war (Collier 1999a). But what is the nature of the mechanisms linking resources abundance with weak institutions, divided societies, and hence slow growth? Why do some countries (e.g., Malaysia) appear to largely avoid these problems, while others (Zambia) get overwhelmed by them? What comfort can be given to Sheik Ahmed Yamani, the one-time oil minister of Saudi Arabia, who surmised that 'All in all, I wish we had discovered water'?1 This chapter provides some preliminary answers to these questions by examining the social foundations of developing economies with different types of natural resource endowments. We begin by providing a brief overview of two recent literatures in development studies—namely, economic research on 'social' issues, and social research on 'economic' issues—with a view towards forging a set of propositions that draws on the insights of both. Next, we contrast the primary arguments and evidence that economists have submitted on how social variables help to explain the disappointing performance of the resource-abundant countries with those offered by The views expressed in this paper are entirely those of the authors, and should not be attributed to their respective organizations. The research assistance of Maya Tudor is gratefully acknowledged. ' As cited in Ross (1999: 297).

Social Foundations of Poor Economic Growth

77

other social scientists. In Section 5.3 we propose a methodological strategy for integrating the findings of both literatures. We gauge the veracity of our theoretical propositions and methodological strategy in Section 5.4 by examining empirical evidence from ninety developing economies. We conclude with a discussion of the policy implications of this analysis. Our central contention is that the nature of communities, institutions, and statesociety relations has a vital impact on economic growth trajectories in general, and the management of shocks in particular. Moreover, we show that in developing countries the efficacy of these three dimensions is powerfully influenced by export structures— that is, whether foreign exchange is derived primarily from manufactures, or diffuse or point-source natural resources. These different types of export structure create differential capacities to reduce social and economic divisions—that is, to produce a 'middle class consensus' (Easterly 1999b)—and viable institutions of conflict management. We find that the adverse social effects are particularly strong in countries abundant in point-source natural resources (such as oil and minerals), but that a rigorous commitment to sound policy can generate more equitable growth-enhancing outcomes. 5.1. SOCIAL V A R I A B L E S , SOCIAL THEORY, AND ECONOMIC D E V E L O P M E N T Recent efforts to account for, enhance, and even predict divergent development prospects have sought to include a social dimension, encompassing issues ranging from civil liberties and ethnic diversity to trust and community participation. This has opened up new terrain for scholars and policymakers from a variety of backgrounds,2 but we restrict our coverage of these endeavours to two literatures, first that of economists studying the effects of social variables on growth, and second, that of sociologists and political scientists investigating how communities and institutions shape development outcomes.3

5.1.1. The 'social variables' literature in economics Contemporary efforts to incorporate social variables into economic development have their origins in the seminal work of Adelman and Morris (1967), but at that time a data deficit and an ideology surplus combined to give this research programme a short life. However, innovative efforts have been made in recent years to tease out causal relationships between different social variables and economic growth. Most of these " To be sure, there have always been dissenting views within the development community. A key feature of the renewed interest in 'the social dimension' and 'civil society' issues, however, is not so much the vindication of these dissenting views (which were, and remain, largely driven by ideological commitments). Rather, it is that innovative attempts are now being made to expand the substantive scope of orthodox and revisionist theories (in the case of economics) and/or to build constructive, even complementary, alternatives (in the case of sociology and political science). 3 The economic literature reviewed here is restricted to work on cross-national studies. There is an equally interesting and important literature on the individual and household effects of social variables (e.g., Narayan and Pritchett 1999), but we shall only incorporate that material as and where necessary.

78

M. Woolcock, L. Pritchett and J. Isham

results still rest on admittedly crude growth regressions,4 but the broad results obtained to date are highly suggestive, and a more or less coherent picture appears to be emerging. These studies attribute a range of effects to social variables, running from the mildly influential to the highly significant. The pioneering papers in this field take their cue from the endogenous growth literature, but among the first to explicitly incorporate and test social variables were Knack and Keefer (1995, 1997) and Keefer and Knack (1997), who took data from the World Values Survey on trust and institutional credibility, and used it to explain economic growth rates and patterns of conditional convergence (or divergence).5 Knack and Keefer's results provide moderate support for so-called 'Olson effects'—that is, that social groups can in fact have constraining effects on growth— but (consistent with Putnam 1993) they also argue for the importance of trustworthy, credible political institutions. In more recent work, La Porta et al. (1997), Zak and Knack (1999) and Knack (1999) use data sets from more countries over longer time periods to reaffirm the importance of social trust to the growth of large firms and economies. Mauro (1995) and La Porta et al. (1998) use similar strategies to show how corruption and lax government institutions undermine growth, while Hall and Jones (1999) argue that the quality of a nation's 'social infrastructure' is a major determinant of its productivity levels. Africa has provided the most fertile ground for economists seeking to integrate social variables into analyses of economic growth patterns. Easterly and Levine (1997) argue that high levels of ethnic and linguistic fractionalization in Africa, coupled with the high spillover effects (externalities) of one country's poor economic policies on its neighbours, can explain up to 45 per cent of that region's slow growth rates.6 Temple (1998) takes a different approach but reaches broadly similar conclusions, showing that low levels of communications infrastructure and other measures of social capability7 account for slow African economic growth rates. Collier and Gunning (1999), echoing Rodrik (1999), argue that sub-Saharan Africa is not 'different', as popularly imagined, but that it merely has an unhappy confluence of growth-reducing factors: isolationist trade policies, deficient public services, unfavourable geography, political instability, lack of financial depth, high aid dependence, and low social capital. These authors distinguish between civic social capital, by which they mean cohesion and collective action at the community level, and government (or public) social capital, namely the 4 As Aron (1997: 2) correctly observes, 'while various studies establish statistically significant relationships, they frequently do not test the sensitivity of their results to model specification, outliers, measurement error, endogeneity of regressors and omitted variables bias. Much of the cross-sectional empirical literature, while loosely related to the augmented Solow growth model, consists of reduced form regressions which present correlations amongst largely endogenous variables, but with no clear indication of the specific channels by which particular determinants affect growth.' 3 For an early (though less explicitly 'social') analysis in this tradition, see Kormendi and Meguire (1985). 6 But see Collier (1999b), who maintains that high levels of ethnic fractionalization only have a negative effect on growth in countries that also deny political and civil liberties. Posner (1999) argues that a more accurate indicator is the number of 'politically relevant' ethnic groups, not their mere demographic size or diversity. (Groups may be small numerically, for example, but nonetheless control key industries.) 7isSee Koo and Perkins (1995) and Temple and Johnson (1998) for a more extensive coverage of this theme as it applies to economic growth.

Social Foundations of Poor Economic Growth

79

quality of the political and legal infrastructure. On both counts they argue, social capital in Africa is in short supply, with civic social capital undermined by ethnic fractionalization and high inequality, and public social capital eroded by unelected and unaccountable rent-seeking political elites. Of these various factors, Collier and Gunning rate lack of openness to trade and low levels of social capital as being the most damaging to Africa's growth. Rodrik (1998, 1999), using similar themes if not words, extends this 'lack of social capital' thesis, arguing that countries of all stripes and hues showed remarkably similar growth rates until the shocks of the early 1970s. Accordingly, the key issue to be explained is why some countries recovered rapidly from these shocks while others sank into near-permanent decline. For Rodrik, the combined effect of divided societies and weak institutions of conflict management explains why the series of shocks in the 1970s were unable to be absorbed. As such, he argues that openness to trade should be a component of, not a substitute for, a national development strategy centred on forging broad domestic social coalitions and constructing effective institutions for managing conflict. This approach, as we shall see, is particularly important for explaining Pritchett's 'mountains' in the evaluation of output in resource-abundant economies, which are frequently subjected to severe, and potentially destabilizing economic shocks. Economic research on social variables is important in its own right and also because it shows that economists are finally beginning to take the social dimensions of growth and development seriously. As Aron (1997: 3) astutely notes however, '[t]he typical collective classification and treatment of all these proxies as "socio-political measures" has tended to obscure the different channels through which institutions operate, and has impoverished the interpretation of the role of institutions in growth. This is serious for Third World countries, where weak institutions are centrally implicated in low growth.' To get a more detailed insight into the different mechanisms and channels by which institutions shape development outcomes, it is instructive to review recent work in the sociology of economic development.

5.1.2. The social theory literature on communities, institutions, and state—society relations Research by sociologists and political scientists studying economic development issues has made a similar distinction between civic and public social capital, but has disaggregated both realms while simultaneously exploring more closely the dynamic interaction between the two levels. In discussions of civic social capital, for example, a distinction has been made between intra-community ties ('bonding') and inter-community networks ('bridging'), with different combinations of these dimensions having profoundly diverse consequences (see Woolcock 1998; Woolcock and Narayan 2000 for detailed reviews). Social scientists have also concerned themselves with public social capital, the principal concern of this chapter, examining the manner in which institutional structures contribute (or not) to economic development. Drawing on detailed case

80

M. Woolcock, L. Pritchett and J. Mam

studies, this body of research has given rise to three broad propositions: (1) States, firms, and societies are neither good nor bad in and of themselves; governments, corporations, and civic groups are 'variables' in terms of the impact they can have on the attainment of collective goals. (2) States, firms, and communities alone do not possess the resources needed to promote broad-based, sustainable, welfare-enhancing outcomes; complementarities and partnerships forged both within and across these different sectors are therefore required. Identifying the conditions under which these synergies emerge is thus a central task of development theory and practice. (3) Of these different sectors, the state's role in facilitating positive developmental outcomes is at once the most problematic and important, because the state is the ultimate provider of public goods, the final arbiter and enforcer of the rule of law, and the actor best situated to transcend the potentially parochial interests of a particular class, sex, ethnic group, political party, business lobby, or religious community. If these propositions are true, equitable and effective, economic development requires the difficult task of assembling and maintaining coherent, credible, and competent public institutions that are simultaneously actively engaged with the day-to-day affairs of their constituents (Evans 1996; Tendlcr 1997). These twin features ensure that public officials are at once skilled employees who can relate to their colleagues, yet are also openly accountable for their actions, able to make timely responses to crises and opportunities alike, and are in touch with the ongoing concerns of those they feign to serve. This result is generalizable across a broad range of formal organizations: positive developmental outcomes are most likely when institutions—public, private, and civic alike-—are characterized by both high organizational integrity and synergy. Importantly, where one or more of these features is absent or weak, institutional performance is likely to be inefficient and ineffective at best, highly destructive ('predatory') at worst. Upon closer inspection, the arguments from both the civic and public schools of social capital research have a remarkable degree of congruence (Woolcock 1998). Whether one is discussing formal organizations or local communities, the nature and extent of the social relationships within and between component sectors or groups is a key variable shaping development performance. Furthermore, it is vital to understand these relationships in a dynamic rather than static sense; that is, to appreciate that success or failure at one moment in time is likely to have a profound effect on subsequent efforts. Paradoxically, success today may not necessarily result in success tomorrow, but may undermine the very conditions that made it possible in the first place; conversely, what appears as failure today may in fact sow the seeds for happier outcomes further down the line.

5.1.3. Implications for the study of resource-abundant economies The 'social variables' and 'social theory' literatures have a number of common and complementary findings with important implications for the study of For8. a representative account of work in this tradition, see World Development (1996; Vol. 24, No. 6).

Social Foundations of Poor Economic Growth

81

resource-abundant economies. These can be expressed as three broad propositions: (1) A social structure that is diverse is good, but to a point. Different communities within society, especially marginalized groups, need to have a legal and legitimate voice; they need forums in which different views are aired and new ideas shared. Where dissemination of information is suppressed or narrowly confined, there are high transaction costs to exchange of all sorts, and the creation of a social consensus that can be called upon to mediate times of crisis is undermined. (2) Competent, credible, and coherent public institutions are vital for providing public goods of all types, and importantly, for creating, implementing, and monitoring sound policies. Prosperity, while potentially the product of such a regime, can also mask flaws within it; but crises like severe economic shocks put it to the test. Economies with solid institutions of conflict management will weather these shocks better than those without because their governments can call upon the support and cooperation of a majority of their citizens. (3) The capacity of groups within society to organize in their collective interest turns crucially on the nature of the groups' relationships with the state. Where some groups secure grossly disproportionate access to the resources of the state, they will distort policies to the point of eroding the credibility and capacity of the state. At best, democracy curbs these tendencies by providing mechanisms of democracy public accountability and peaceful opportunities for marginalized groups to displace dominant ones. Other mechanisms for ensuring transparency and the two-way flow of information within and between the public, private, and civic sectors (e.g., free markets, free press, free association) serve similar purposes. We contend that these propositions help to advance our understanding of the political economy of the disappointing growth of many resource-abundant countries. 5 . 2 . TYPE S O F NATURA L RESOURC E ENDOWMENT S AND ECONOMI C DEVELOPMEN T The central thrust of social capital research to date, as the above examples show, has been to treat social capital as an independent variable, that is, as something that has an impact on policy outcomes such as education, institutional performance, and economic growth. A smaller number have considered social capital formation as a dependent variable, the independent variable of interest in this instance being the state and the nature of its engagement with civil society. Our approach takes two novel methodological steps. First, we treat social capital as an intervening variable, so that the input of social capital, in addition to whatever direct effect it may have, also has an interactive effect in mediating the impact of other variables, such as exogenous shocks, on growth. Second, we treat social capital as influenced by resource endowments, and explore whether 'natural capital' shapes 'social capital' in ways that have identifiable consequences for growth. We argue that the best way to address this question is to look at the effects of different types of natural resource endowments on social capital formation, since

82

M. Woolcock, L. Pritchett and J. Isham

certain types of natural resource economies appear to experience more volatile growth trajectories than others. Our analysis builds on export base theory and the consequences of different staples not only for economic linkages, but also for social relationships. Baldwin (1956) ascribes too much importance to the role of the production function in generating economic linkages and largely ignores the social milieu with which they interact (Auty 1995). Plantation crops (cotton, sugar processing, tobacco) as well as oil and hard minerals are typically associated with highly concentrated ownership. This renders the state heavily dependent on a small number of owners (rentier capitalists) for generating its income, and the rentier capitalists, in return, are able to extract generous tax breaks, subsidies, and levels of import protection. Moreover, since these types of point-source resources tend to be more capital than labour intensive, they tend to polarize society between haves and have-nots, so that it is difficult to generate a socially agreed-upon consensus regarding the management of conflicts. During downswings, these vested interests resist reforms that would diversify the economy because this would create 'rivals' competing for labour and government influence. Prosperity may mask these fragile social arrangements, but shocks expose and exacerbate them. Extended over several decades, the trajectory is likely to be one of short-term booms and long-term busts that lead to flawed policy responses and very slow, even negative, overall growth rates. The transition from an agricultural to an industrial economy is at best painful and may not occur at all if the economy becomes trapped into a dependence on slow-growth staples. Two variations can be identified. In the first, the state owns the rentier capitalists so that a regime of rentier autocrats emerges, as with Algeria and Nigeria; in the second, the rentier capitalists effectively own the state, as in Angola and El Salvador, and oligarchic regimes emerge. Small holder export crops and staples (wheat and rice, and in some cases coffee and cocoa), on the other hand, display a different pattern of 'diffuse' socioeconomic linkages. Here, low barriers to entry mean that wealth is more evenly dispersed since the resource is less amenable to capture by an elite (as in the United States and Australia at the end of the nineteenth century). The state has a greater degree of autonomy because it is not beholden to a small economic group and must instead appeal to and appease a more diverse constituency. Such a wider constituency tends to favour the mobilization of tax revenues for investments in human capital (education and health care) as well as for economic infrastructure. In addition the wider diffusion of wealth is more conducive to democratic institutions so that state—society relations in small holder economies generate a more sophisticated social consensus regarding conflict management. Consequently, when shocks occur, capital, labour, and the state have a broader array of social (human and institutional) resources to call upon to help mediate the crisis. This, in turn, fosters greater economic flexibility to adjust from slow-growth to high-growth commodities and escape the staple trap. Extended over several decades, a trajectory of longer-term booms and shorter-term busts emerges, generating modest but substantial overall growth rates. Diversification into a competitive industrial economy, if not always smooth, nonetheless occurs, and does so more or less evenly.

Social Foundations of Poor Economic Growth

83

5.3. DAT A A N A L Y S I S The data set we use to test these arguments has been compiled from a number of sources. Using UNCTAD's Handbook of International Trade and Development Statistics (1988), we assembled data on the leading export staples of every country in 1985 that had a GNP per capita under $10,000 and a population greater than one million. UNCTAD classifies export structures into five categories, namely foods, agricultural raw materials, fuels, minerals, and manufacturing. Within the leading export category, we listed the two most important commodities, enabling us to make the subsequent assignment of different countries into point-source and diffuse natural resources, and manufacturing. (A separate category was also created for coffee and cocoa producers, since these commodities resist easy classification because they may be grown on peasant farms and plantations.) Judgements from country and commodity experts were used where there was some ambiguity. Growth rate data for the period 1957-97 was compiled from the Penn World Tables and the World Development Indicators 99. Dropped from the sample were those countries listed by UNCTAD as 'socialist' (with the exception of China) and 'developed' (Greece, Ireland, and Portugal), and those with growth rate data spanning less than twenty-five years, leaving a final sample size of 90. Single measures—obtained by calculating averages across decades—of social and political data were adapted from Temple and Johnson (1998), Easterly (1999a), World Bank (1999), Deininger (1999), and Filmer and Pritchett (1999). In order to examine the comparative performance of different resource economies, it is instructive to begin by looking at the overall pattern of growth in developing countries over the 1957—97 period. As shown in Fig. 5.1, all developing countries performed relatively well from the mid-1950s to the mid-1970s; they enjoyed a median annual growth rate of 2.3 per cent over this period. As Rodrik (1999) points out, the 1950s, 60s, and early 70s were a time of vibrant debate among development theorists, with the strong performance of many different strategies making it difficult to adjudicate between them.

Figure 5.1 . Smoothed median per capita growth rates in 90 developing economies, 1955—1997 Source: World Bank (1999).

84

M. Woolcock, L. Pritchett and J. Isham

Figure 5.2. Smoothed median growth rates for 90 developing economies, 1957-1997 Source: World Bank (1999).

From the mid-1970s until the mid-1990s, developing economies endured a growth collapse of 'Grand Canyon' proportions, setting back their development agenda by at least a decade. Indeed, many countries remain lost in the canyon, with the global financial crises of the late 1990s threatening to further impede their recovery. The use of medians ensures that outliers do not have an undue effect on the results, but they also (by definition) tell us that, in any given year, half the sample performed much better, while the other did even worse. Figure 5.2 shows that there is a big difference in the performance of manufacturing, point-source, diffuse, and coffee/cocoa economies in terms of how they responded to the growth collapse from the mid-1970s to the mid1990s. Manufacturers had always done well (note: this sample includes India and Bangladesh, not just the East Asian NICs), and while they experienced a mild downturn in growth rates in the mid-1970s they responded quickly and effectively. Diffuse resource economies were also adversely affected, but they show a steady rate of recovery. The point-source and especially the coffee/cocoa economies, however, experienced a protracted growth collapse. Why? This differential capacity to handle growth collapses, we maintain, is largely a function of the strength of socioeconomic divisions that, in turn, affect the quality of a country's political institutions. Both the socioeconomic divisions and quality of institutions can be measured, albeit imperfectly, and to this end we have compiled various indicators that attempt to evaluate them. We classified these into indicators of social and political institutions and indicators of inequality. The indicators of social and political institutions, with original sources are: • Social Development (Adelman and Morris 1967): an indicator constructed by factor analysis to measure socioeconomic development;9 9 On this indicator, countries with the lowest score 'are primarily tribal and are characterized by a preponderant nonmarket sector' (Adelman and Morris 1967, as cited in Temple arid Johnson 1998).

Social Foundations of Poor Economic Growth

85

• Civil liberties (Freedom House 1994): a subjective indicator based on a checklist of 14 civil liberties, including media free of censorship, open public discussion, and freedom of assembly and demonstration; • Political liberties (Freedom House 1994): a subjective indicator based on a checklist of 11 indicators of political rights, including chief authority recently elected by a meaningful process and fair election laws; e Bureaucratic quality (Easterly (1999a) drawing on data from the International Country Risk Guide); « Rule of law (Easterly (1999a) drawing on data from the ICRG).10 Tables 5.1-5.4 confirm that items such as 'Political liberties', 'Civil liberties', 'Bureaucratic quality' and 'Rule of law' are all higher in the resource-poor economies than they are in the resource-rich economies, while levels of land inequality are lower (see Learner et al. 1999). Income and education inequality, however, present a more ambiguous picture. The first two rows of Table 5.1 present the means and subsample sizes of these five indicators among all ninety countries in the sample as well as their minima and maxima; the other rows present the respective means and subsample sizes among each of the four export classifications.11 These results yield two empirical results that are consistent with our thesis on social and political institutions. First, social development and civil liberties are higher in the resource-poor economies than in the three types of resource-rich economies. For example, the mean of the social development indicators for the manufacturing economies is 0.23, compared with 0.17 among the diffuse economies, —0.14 among the point-source economies, and —0.25 among the coffee/ cocoa economies. Second, political liberties and government quality are lowest in the coffee/cocoa economies (although they are somewhat higher in the point-source economies than in the diffuse economies).12 The indicators of inequality, with original sources are: • Income Gini (Deininger and Squire 1996; World Bank 1999): a Gini coefficient based on income; » Income share to mid-60 per cent (Easterly 1999a): a related indicator of income inequality; • Land Gini (Deininger 1999): a Gini coefficient based on land holdings; 10

In a large number of recent studies, these indicators (separately and in various combinations) have provided evidence on the institutional determinants of different economic outcomes, including of per capita growth rates (Temple and Johnson 1998; Barro 1997; Knack and Keefer 1997) and of the performance of public investment projects (Isham et al. 1997). 11 In all cases, higher values indicate a more desirable social or political outcome. Thus, the original scale of the Freedom House variables—where higher values represent lower civil and political liberties—has been reversed. 12 Subsequent econometric research will explore the statistical significance of these associations and the possible link between these indicators and economic performance in these four export classifications. A finding that social and civil institutions, as opposed to political institutions, are significantly associated with the improved performance of the diffuse economies would be consistent with the findings of Isham et al. (1997) on civil liberties and project performance.

Table 5.1 . Social and political institutions among different Total Socia l No . Civi development libertie index All countries (minimum, maximum) Diffuse Point source Coffee and cocoa Manufacturing

90

-0.07 (-1.86, 1.59)

l No . Politica s libertie

export classifications

l No . Bureaucrati s qualit y

c No

. Rul

e of law No

.

62

3.19(1,7)

88

3.06(1,7)

88

2.26(0,6)

80

2.19(0,6)

80

18 45 18

0.17 -0.14 -0.25

12 31 14

3.31 3.17 3.14

18 44 18

2.94 3.09 2.83

18 44 18

2.06 2.24 1.73

17 40 15

1.94 2.08 1.87

17 40 15

9

0.23

5

3.19

8

3.63

8

3.75

8

3.88

8

Notes'. The first row lists the means of five indicators of social and political institutions and their available subsample sizes (among the larger sample of 90 countries); the second row lists the minima and maxima or the indicators. The other rows are the means and subsample sizes of the indicators among four export classifications. See text for descriptions of variables. In all cases, higher values indicate a more desirable social or political outcome. Source: See text.

Table 5.2. Inequality among different export classifications Total Lan All countries (minimum, maximum) Diffuse Point source Coffee and cocoa Manufacturing

d gini No

. Incom

e gin i No

. Incom

e share to No

. Educatio

mid 60%

n gini No

.

90

0.65 (0.32,0.92)

59

0.46 (0.29,0.62)

78

0.43 (0.22,0.54)

66

4.46 (1.1,18.6)

33

18 45 18

0.60 0.66 0.73

13 28 12

0.48 0.46 0.47

16 37 16

0.41 0.43 0.42

14 29 14

5.57 5.56 3.71

5 12 12

9

0.50

6

0.38

9

0.49

9

1.99

4

Notes: The first row lists the means of four indicators of inequality and their available subsample sizes (among the larger sample of 90 countries); the second row lists the minima and maxima of the indicators. The other rows are the means and subsample sizes of the indicators among four export classifications. See text for descriptions of variables. In all cases except 'income share to mid 60%', higher values indicate more inequality. Source: See text.

88

M. Woolcock, L. Pritchett and J. Isham

• Education Ratio (Filmer and Pritchett 1999): the ratio of rich to poor who finish primary schoo13l. Table 5.2 (with the same format as Table 5.1) yields only one empirical result that is consistent with our thesis on the absence of sociocconomic divisions. Land inequality is higher among the coffee/cocoa economies (0.73) and the point-source economies (0.66) than among the diffuse economies. And admittedly, this result, unlike the others, is directly tied to the classification of the economies: coffee and cocoa economies are by definition characterized by very high land holdings among the few. By contrast, the two measures of income inequality show no difference among the classifications; the indicator of education inequality shows a better result (less inequality) among the coffee/cocoa economics. The previous two tables suggest that better social and political institutions, not more equality, most characterize diffuse resource-rich economies. What is the empirical evidence that better social and political institutions are in turn associated with better growth performance? May the overall growth rate of the diffuse economies—and in particular their steady rate of recovery since the oil shock (as shown in Fig. 5.2)—be attributed to this difference? Tables 5.3 and 5.4 use the two indicators of social institutions, 'Social Development' and 'Civil Liberties', to address these questions. The first three rows of Table 5.3 show that during the entire time period (1957—97) and during the two subperiods demarcated by the first oil shock, growth is higher among countries with more civil liberties. The contrast is most striking after the oil shock: annual growth of 1.78 per cent among countries with high civil liberties, compared with 0.60 where civil liberties are low. A similar pattern is seen in the first three rows of Table 5.4.14 The next nine rows of Table 5.3, which compare the performance of diffuse, pointsource, and coffee and cocoa economies, show: (1) that diffuse economies have the highest share of 'high civil liberties' countries (12 countries among a total of 18); (2) that diffuse economies have the best growth performance over the entire time period (1.94 per cent versus 1.37 per cent and 0.93 per cent); and (3) that this difference can be attributed to the vastly superior performance of the diffuse economies since the oil shock (1.91 per cent). Finally, even within each of the export classifications, economies with higher civil liberties outperform those with lower civil liberties. Again, a similar pattern is seen in the same nine rows of Table 5.4. As for the best performing manufacturers, only the classifications of 'social development' show an association between strong social institutions and high growth performance (compare the last three rows of Tables 5.3 and 5.4, respectively; the smaller sample sizes prevent us from reaching firmer conclusions about the role of social institutions in the growth performance among manufacturers). 13 Separately and in various combinations, these indicators have provided evidence of the institutional determinants of different economic outcomes in a large number of recent studies, including per capita growth rates (Temple and Johnson 1998; Barro 1997; Knack and Keefer 1997) and the performance of public investment projects (Isham et at. 1997). 14 For comparative purposes, each of these tables uses in the first row only those countries with available data on civil liberties and social development, respectively.

Table 5.3 . Growth performance of different export classifications (1957-1997), by civil liberties

Civil libertie s rank No

All countries

Diffuse

Point source

Coffee and cocoa

Manufacturing

— High Low — High Low — High Low — High Low — High Low

. Growt

88 47 41 18 12 6 44 20 24 18 10 8 8 5 3

h rat e

1957-97

1957-74

1.66 2.00 1.27 1.94 2.38 1.08 1.37 1.84 0.97 0.93 1.34 0.42 4.27 3.08 6.27

2.38 2.55 2.19 2.25 2.99 0.79 2.45 2.46 2.45 1.94 2.43 1.34 3.27 2.10 5.24

1975-97 1.23 1.78 0.60 1.91 2.28 1.17 0.65 1.51 -0.06 0.49 0.99 -0.14 4.55 3.26 6.71

Notes: Growth rates are medians over the designated time periods and means among all countries and among the four export classifications. Source: See text. Table 5.4. Growth performance of different export classifications (1957-1997), by social development indicator

Social development No

. Growt

rank All countries

Diffuse

Point source

Coffee and cocoa

Manufacturing

— High Low — High Low — High Low — High Low — High Low

62 31 31 12 8 4 31 14 17 14 6 8 5 3 2

h rate

1957-97

1957-74

1.74 2.25 1.24 2.01 2.31 1.42 1.54 1.98 1.18 1.29 1.69 0.99 3.62 4.50 2.30

2.43 3.09 1.76 2.13 2.68 1.03 2.51 2.94 2.16 2.39 3.42 1.61 2.76 4.26 0.50

1975-97 1.33 1.98 0.68 1.96 2.28 1.32 0.90 1.45 0.45 0.79 1.33 0.38 4.02 4.97 2.60

Notes: Growth rates are medians over the designated time periods and means among the four export classifications. Source: See text.

90

M. Wookock, L. Pritchett and J. Isham 5.4. CONCLUSION S AND POLIC Y IMPLICATION S

A common finding from both economics and sociology is that community structures, institutional capacity, and state-society relations play a key role in shaping economic policy and economic performance. In many resource-abundant economies, communities tend to be weak (in terms of human resources) and divided, with little sense of a shared commitment to common goals. Institutions in general, and those of conflict management in particular, lack capacity, credibility, and coherence; and state—society relations are marked by distrust, grossly disproportionate access to resources, and lack of transparency. In this chapter we have endeavoured to show that natural resourcerich economies in general, and different types of resource abundance in particular, place different types of pressures on these three junctures, with those resources more easily captured and controlled by a narrow elite being most likely to undermine them. This is not a deterministic model, however; countries that, by historical accident or good management, are able to sustain sound and open policy environments will be more likely to reap any potential windfall. Consistent with the World Bank (1998) and Rodrik (1999), we argue that trade openness should be seen as an important part of the development strategy, but not a substitute for one. Institutions are also required that bring a measure of transparency and competition necessary for supplying information on mismanaged economies and societies. Capable governments must actively build the social capacity and political consensus needed to moderate the shocks to which natural resources economies are repeatedly exposed. This is no easy task, but it can and has been achieved both in the past and present. A sustained effort to replicate these accomplishments is crucial. Nowhere is this commitment needed more strongly today than in the impoverished energy-rich transition economies like Azerbaijan, Kazakhstan, and Turkmenistan (see Chapter 16). The results for the cocoa and coffee producing countries are puzzling. One explanation is that the offsetting features arising from a mix of landholding types (estate and smallholding) blurs the overall features of the group. A second is that, as Chapter 10 suggests, the potentially favourable socioeconomic linkages of diffuse smallholders are easily corrupted into point linkages by the operations of crop marketing boards. The sunk costs that the planting of tree crops involves (in contrast to perennial crops) renders them especially attractive to rapacious taxation regimes. In any event, a full accounting for the poor performance of resource-rich economies is only enhanced by giving serious attention to their social foundations.

REFERENCES Adelman, Irma and Cynthia Taft Morris (1967), Society, Politics and Economic Development, Baltimore: Johns Hopkins University Press. Aron, Janine (1997), 'Political, economic and social institutions: a review of the growth evidence', University of Oxford, Centre for the Study of African Economies: Working Paper Series, 98-4.

Social Foundations of Poor Economic Growth

91

Auty, Richard (1995), Patterns of Development: Resources, Policy and Economic Growth, London: Edward Arnold. Baldwin, Robert (1956), 'Patterns of development in newly settled regions', Manchester School of Economic and Social Studies, 24, 161-79. Barro, Robert (1997), Determinants of Economic Growth: A Cross-Country Empirical Study, Cambridge, MA: MIT Press. Collier, Paul (1999a), 'The economic causes of civil war', Oxford Economic Papers and Bulletin, 51/1, 168-83. Collier, Paul (1999b), 'The political economy of ethnicity', Annual Bank Conference on Development Economics 1998, Washington, DC: The World Bank. Collier, Paul and Jan Willem Gunning (1999), 'Explaining African economic performance', Journal of Economic Literature, 37 (March), 64-111. Deininger, Klaus (1999), 'A new data set for measuring land inequality', The World Bank, (mimeo). Deininger, K. and L. Squire (1996), 'A new data set for measuring income inequality', World Bank Economic Review, 10/3, 365-91. Easterly, William (1999a), 'Life during growth: international evidence on quality of life and per capita income', Policy Research Paper 2110, Washington, DC: The World Bank. Easterly, William (1999b), 'Happy societies: the middle class consensus and economic development', The World Bank (mimeo). Easterly, William and Ross Levine (1997), 'Africa's growth tragedy: policies and ethnic divisions', Quarterly Journal of Economics, 62 (November), 1203-50. Evans, Peter (1996), 'Government action, social capital and development: reviewing the evidence on synergy', World Development, 24/6, 1119-32. Filmer, Deon and Lant Pritchett (1999), 'The effect of household wealth on educational attainment: evidence from 35 countries', Population and Development Review, 25/1, 85-120. Freedom House (1994), Freedom in the World: The Annual Survey of Political Rights and Civil Liberties 1993-94, New Brunswick, NY: Freedom House. Hall, Robert and Charles Jones (1999), 'Why do some countries produce so much more output per worker than others?', Quarterly Journal of Economics, 114/1, 83-116. Isham, Jonathan, Daniel Kaufmann and Lant Pritchett (1997), 'Civil liberties, democracy, and the performance of government projects', World Bank Economic Review, 11/2, 219—42. Karl, Terry (1997), The Paradox of Plenty: Oil Booms and Petro-States, Berkeley, CA: University of California Press. Keefer, Philip and Stephen Knack (1997), 'Why don't poor countries catch up? A cross-country test of an institutional explanation', Economic Inquiry, 35/3, 590-602. Knack, Stephen (1999), 'Social capital, growth and poverty: a survey and extensions', College Park, MD: IRIS, University of Maryland (mimeo). Knack, Stephen and Philip Keefer (1995), 'Institutions and economic performance: crosscountry tests using alternative institutional measures', Economics and Politics, 1 (November), 207-27. Knack, Stephen and Philip Keefer (1997), 'Does social capital have an economic payoff? Acrosscountry investigation', Quarterly Journal of Economics, 112, 1251-88. Koo, Bon Ho and Dwight Perkins (eds.) (1995), Social Capability and Long-Term Economic Growth, New York: St. Martins. Kormendi, Roger and Philip Meguire (1985), 'Macroeconomic determinants of growth', Journal of Monetary Economics, 16, 141-63.

92

M. Woolcock, L. Pritchett and f. Isham

La Porta, Rafael, Florencio Lopez-de-Silanes, Andrei Shleifer and Robert Vishney (1997), 'Trust in large organizations', American Economic Review, Papers and Proceedings, 87/2, 333-8. La Porta, Rafael, Florencio Lopez-de-Silanes, Andrei Shleifer and Robert Vishney (1998), 'The quality of government', NBER Working Paper 6727, Cambridge, MA: NBER, Learner, Edward, Hugo Maul, Sergio Rodriguez and Peter Schott (1999), 'Does natural resource abundance increase Latin American income inequality?', Journal of Development Economics, 59, 3-42. Mauro, Paolo (1995), 'Corruption and growth', Quarterly Journal of Economics, 110/3, 681-712. Narayan, Deepa and Lant Pritchett (1999), 'Cents and sociability: household income and social capital in rural Tanzania', Economic Development and Cultural Change, 47/4, 871-97. Posner, Daniel (1999), 'Ethnic fractionalization: How (not) to measure it? What does (and doesn't) it explain?' Paper presented at the annual meeting of the American Political Science Association, Atlanta. Pritchett, Lant (1998), 'Patterns of economic growth: hills, plateaus, mountains, plains', Policy Research Working Paper 1947, Washington, DC: The World Bank. Putnam, Robert (1993), Making Democracy Work: Civic Traditions in Modern Italy, Princeton, NJ: Princeton University Press. Rodrik, Dani (1998), 'Where did all the growth go? External shocks, social conflict, and growth collapses', NBER Working Paper 6350, Cambridge, MA: NBER. Rodrik, Dani (1999), The New Global Economy and Developing Countries: Making Openness Work, Baltimore, MD: Johns Hopkins University Press. Ross, Michael (1999), 'The political economy of the resource curse', World Politics, 51/1, 297322. Temple, Jonathan (1998), 'Initial conditions, social capital, and growth in Africa', Journal of African Economies, 7/3, 309-47. Temple, Jonathan and Paul Johnson (1998), 'Social capability and economic growth', Quarterly Journal of Economics, 113/3, 965-90. Tendler, Judith (1997), Good Government in the Tropics, Baltimore: Johns Hopkins University Press. UNCTAD (1988), Handbook of International Trade and Development Statistics, New York: United Nations. Woolcock, Michael (1998), 'Social capital and economic development: toward a theoretical synthesis and policy framework', Theory and Society, 27/2, 151—208. Woolcock, Michael and Deepa Narayan (2000), 'Social capital: implications for development theory, research, and policy', World Bank Research Observer, 15/2, 225-49. World Bank (1998), Assessing Aid: What Works, What Doesn't, and Why, New York: Oxford University Press. World Bank (1999), World Development Indicators 99, Washington, DC: The World Bank. World Development (1996), Special issue on social capital and economic development 24/6. Zak, Paul and Stephen Knack (1999), 'Trust and growth', Department of Economics, Claremont Graduate University (mimeo).

PART II I

LONG-TERM PERSPECTIVE ON, AND MODELS OF, RESOURCE-BASED GROWTH

This page intentionally left blank

6 Natural Resources and Economic Development: The 1870-1914 Experience R O N A L D FINDLAY AND MATS L U N D A H L

This chapter examines the record of two types of resource-abundant economies during the 'golden age' of the world economy from 1870 to 1914. It provides a check on whether the underperformance of the resource-abundant countries during the last three decades of the twentieth century is peculiar to that time period or a more lasting characteristic. The chapter is divided into four parts. The first section outlines a 'North—South' or 'core-periphery' framework within which the 1870—1914 experiences can be evaluated. In this framework, the Northern core is the 'engine of growth' for the world economy. The core is the autonomous source of technological changes and capital accumulation that not only raises per capita incomes in the countries that have already initiated or undertaken the Industrial Revolution but also transmits their growth to the rest of the world by the induced demand for raw materials and other primary staples, as well as the export of capital and migrants in substantial amounts. The periphery is divided into two categories. One is what the League of Nations called the 'Regions of Recent Settlement' and the other is a set of plantation and peasant economies that we group together under the rubric of 'Tropical Countries'. The Regions of Recent Settlement are located in the temperate zones of North and South America, southern Africa, and Australia. These were areas of substantial European settlement even prior to 1870 and received further large inflows of European emigration. The particular countries that we examine in this group are the United States, Canada, Australia, Argentina, Bolivia, Chile, and South Africa. In the Tropical Countries category we have economies such as Burma, Siam, and the Gold Coast, dominated by peasant family farms, and the richer plantation economies of Brazil, Colombia, Costa Rica, Ceylon, and Malaya. We thus have fifteen countries in all, seven in the Regions of Recent Settlement and eight Tropical Countries. The second and third sections of the chapter summarize the experience of each of these groups in the 1870-1914 period. The fourth and final section links the theoretical and empirical This is a condensed version of Findlay and Lundahl (1999). We thank Richard Auty, S. Mansoob Murshed and Moshe Syrquin for generous and constructive comments.

96

R. Findlay and M. Lundahl

analysis and discusses the conclusions about natural resources and economic development that can be drawn from both sets of outcomes.

6.1. THE T H E O R E T I C A L F R A M E W O R K The links between the European core and each of the two sets of peripheral regions differed considerably. The geographical factor of climate, temperate in some cases and tropical in the other, clearly made the former zone more favourable for European penetration and settlement. Furthermore, the native populations were less densely settled and politically organized than in the tropical countries. This made these regions overseas extensions of Europe to a considerable extent. In terms of the traditional three-factor classical model, these regions were endowed abundantly with land (see Table 1.2), while labour and capital were both essentially imported from Europe. The factor 'land' however, cannot be regarded as simply given. It had to be created by a process of investment that extended the 'frontier' further and further at increasing marginal cost.

6.1.1. The Christopher Columbus model This whole process can be explained in stylized terms by means of the 'Christopher Columbus' general equilibrium model of international trade and factor mobility (Findlay 1993, 1995: Chapter 5) and the integrated 'staples' and 'vent for surplus' model (Findlay and Lundahl 1994). In the Christopher Columbus model, 'Europe' has a fixed endowment of land and population with which she can produce two goods, food and manufactures. The supply of capital is perfectly elastic at a fixed interest rate in the long run. Manufactures require a raw material (e.g., cotton) that can only be supplied from overseas, say 'America'. Land in America can be obtained at increasing marginal cost by the investment of capital (e.g., railroads to the West). Food and the raw material can be produced with land and labour at different factor intensities, the labour being provided by migration from 'Europe'. In equilibrium, the value of the marginal product of land in food production is equal to the marginal cost (in terms of capital) of clearing land. The model thus has three factors—capital, land, and labour—and three goods—food and manufactures, which are final consumer goods, and the raw material that is an input to manufactures. Given constant returns to scale technology for the three goods and consumer preferences for food and manufactures, the model determines the distribution of population between 'Europe' and 'America', the output of the three goods, the prices of all the goods and factors, the extent of the land frontier and the endogenous supply of capital. Finally, the implications of population growth, technological change in manufactures or an exogenous shift in the cost of frontier extension can all be ascertained by the appropriate comparative statics exercise for the model. The nineteenth-century experiences of the United States, Canada, Australia, and Argentina can all be accounted for in terms of the predictions of this model. The Christopher Columbus model does not allow for industrial production in the regions of recent settlement. This feature can, however, be introduced in a model of

Economic Development: 1870-1914

97

staples production (Findlay and Lundahl 1994). The (agricultural) staple is produced with the aid of labour and land. The latter has to be developed with the aid of capital before it can be used. Manufactures are in turn produced with the aid of labour and a capital stock consisting of the very manufactured good itself. As before, manufactures can be both consumed and invested and capital is in perfectly elastic supply, at a given interest rate. The given rate of interest determines the marginal product of capital and hence the capital-labour ratio in manufacturing as well as the wage rate. The marginal product of labour in agriculture depends on the land-labour ratio. Using manufactures as the numeraire, the wage rate then provides us with a unique ratio for each relative price of agricultural goods and hence also with unique marginal products of labour and land in the sector. Exactly as in the Christopher Columbus model, the value of the marginal product of land in agriculture must be equal to the marginal cost (in terms of capital) of developing land in equilibrium. When the relative price of agricultural goods increases, labour is pulled into agriculture from manufacturing. The land—labour ratio falls and the marginal product of land increases. If equilibrium is to be preserved the marginal cost of land must then increase, that is, the price rise must lead to increased development of land (an extension of the frontier). This, in turn, leads to further migration of labour from industry to agriculture. Trade is opened by an increase in the relative price of the agricultural staple, above the autarky price. Then the agricultural frontier will be extended, labour moves into agriculture from manufacturing and agricultural production increases, whereas in manufacturing, the opposite sequence takes place, with a reduction of the capital stock and falling output as labour leaves. Staples production also attracts inflows of capital and labour. Assuming that the rate of interest above was the one that prevailed in a closed economy, the opening up of the capital market to international conditions can be expected to make the interest rate fall. The capital intensity of the manufacturing sector will increase and labour will be pulled into that sector from agriculture. Possibly the land frontier will also be extended. This happens if (as a result of the loss of labour) the value of the marginal product of land in agriculture falls proportionally less than the rate of interest, so that the marginal cost of land will have to rise again, to preserve the equality of marginal product value and marginal cost of land. Thus, an inflow of capital will lead to an expansion of industrial production and possibly—if the rate of interest falls enough— also to increased, agricultural output, that is, to a general expansion of the economy. The opening of staples trade may also be connected with an inflow of labour. As stated above, the initial effect of a rise in the relative price of agricultural goods is that labour will leave manufacturing for agriculture. In sparsely populated countries the manufacturing labour force may be very small, so small that the increased demand emanating from agricultural expansion may empty the sector completely at the given wage rate. This is likely to result in immigration. Labour will then flow into agriculture until output has risen there to the extent required by the relative world market price of agricultural goods. Any inflow in excess of this level will go into manufacturing, where it will trigger a corresponding increase of the capital stock, so as

98

R. Findlay and M. Lundahl

to maintain the factor proportions in that sector that are compatible with the given rate of interest and wage rate. This, in turn, increases the national income and with that the domestic demand for the agricultural staple so that the exportable excess supply of the latter will shrink.

6.1.2. The tropical open dual economy model For our tropical case studies the most useful theoretical framework will be the classical dual economy formulation of Arthur Lewis (1954), but extended and formalized to explicitly account for international trade in primary products and manufactures in Findlay (1973: Chapter 5). The model has a 'peasant hinterland' that plays the role of a source for a perfectly elastic supply of labour at a fixed real wage. The 'modern' sector consists of profit-maximizing entities such as plantations, mines, or commercial farms that use capital and labour to produce an output that can be sold in the world market. The economy is too 'small' to affect the relative price of imported manufactured consumer or capital goods but is assumed to be a significant supplier of its primary export (e.g., Brazilian coffee or Malayan rubber). There is thus a world demand curve for the export staple that is assumed to shift over time at some exogenous rateλ. Given the initial capital stock profits are maximized when the marginal value (or revenue) product of labour is equal to the given real wage. Profits are the sole source of capital accumulation. Denoting the profit rate by and the propensity to save out of profits by s, the growth rate of capital, g, will be equal tosπThe profit rate π is endogenous and depends upon the terms of trade, the relative price of the primary exportable in terms of manufactures. It is readily demonstrated that if g is greater than A, that is, supply of the primary export grows more rapidly than world demand, its relative prices will fall, thus pulling downπand hence g until it is equal to A. Increasing g through a higher rate of saving orπthrough technical progress does no good in the long run since the economy is permanently constrained by the growth of world demand. When this is brisk, as it was for most tropical products in the 1870—1914 period, trade is truly an 'engine of growth'. When it sputters, however, as in the inter-war period, the prospects are bleak. The possibility of labour-intensive manufactured exports has, however, to be introduced. Here any developing country by itself would be 'small', so the growth prospects of an open dual economy are still given by g =sπas before, butπis now constant (given the technology and the real wage) since the relative price of the labour-intensive exports and capital goods imports is given. Higher rates of saving or technical progress can now permanently raise g since they no longer induce declining terms of trade. If we denote the rate of growth of a 'pure' manufacturing dual economy (i.e., an economy where the modern sector produces only manufactures)byμ, an interesting case to consider is when the economy has both a profitable primary sector and the possibility of manufactured exports. Thus suppose that g is greater thanμwhich is in turn greater than A. Primary production may initially be very profitable if demand is very high. These high initial primary sector profits make g greater thanμ,and so the economy can initially grow even faster than if it were completely specialized on

Economic Development: 1870-1914

99

manufactured exports. Even though initially high, the world demand for the primary export grows relatively slowly at the rate A (less thanμ).Manufacturing production however, now starts to become increasingly important and the primary exports, though growing in absolute volume at the rate A, becomes less and less important relative to manufactured exports. The limiting growth rate isμ(greater than A) and so the economy can successfully escape the 'trap' of slowly growing world demand for primary exports. The foregoing assumes that the staple economy has at least some degree of monopoly power in world trade. If the economy is truly 'small', however, the relative price of the staple would be exogenous and so the growth rate would depend only on supply-side factors such as capital accumulation and the rate at which the frontier can be extended. Here, the staples model (Findlay and Lundahl 1994) is also of relevance. Tropical staples production in the context of a peasant economy, however, differs from staples production in the regions of recent settlement. In the tropics, there was no specialized manufacturing sector but just subsistence agriculture combined with handicrafts during the dead season. Assuming that the economy has abundant land, we may model this sector simply as one where the peasants can obtain a given wage (real income) along the lines of Bent Hansen (1979) and Arthur Lewis (1954). When international trade is opened, this economy can export a staple in return for imports of manufactures and besides get access to capital at a given rate of interest. Again, the staple has to be grown on land that must be cleared, and for the latter, the peasants have to be fed out of the subsistence sector, that is, the cost of clearing land is measured in terms of subsistence goods (a wage fund or a kind of circulating capital). The value of the marginal product of labour in staples production must then be equal to the given subsistence wage, and the value of the marginal product of land must equal the marginal cost (in terms of the wage fund that has to be borrowed) of clearing it. This determines the agricultural frontier and, hence, also the number of workers producing staples, since once the world market price of staples is known so is the landlabour ratio. In the peasant setting, the rent created by clearing land for staples production will simply be equal to the interest cost of the wage fund that has to be borrowed in order to make it possible for the peasants to clear the land, since land is in infinitely elastic supply. Thus, the real income of a cash crop cultivator and a subsistence cultivator will be equal. If instead, the farm producing staples is a plantation, there is no real difference in the mechanism through which the land is cleared. However, instead of peasants we will have agricultural workers recruited from the subsistence sector earning the fixed wage, and instead of the wage fund we have the plantation owner borrowing the funds necessary for the clearing. 6.2. REGION S O F RECEN T SETTLEMEN T

6.2.1. North America The traditional view of the growth process in the Regions of Recent Settlement is that: countries like the United States and Canada managed to extend growth across a wider

100

R. Findlay and M. Lundahl

spectrum of sectors than just the primary export ones, in a manner reminiscent of Walt Rostow's (1960) takcoff stage while, for example, Australia and Argentina failed to do so. Beginning with Canada, the economy for which the staples theory of growth was originally developed (Innis 1930, 1940), linkages were created from the numerous staples, notably from wheat production which made the westward extension of railroads across the prairies necessary and furthermore stimulated the manufacture of agricultural inputs. Warehousing, processing, construction, housing, and public utilities received strong impulses, staples production provided raw materials for flour mills, saw mills, and paper mills, and the incomes generated in the West were spent on manufactures in the East. The growth rate of the Canadian manufacturing industry was surpassed only by that of the Japanese (Meier 1953: 5-6). Possibly, tariff protection had something to do with this, but it does not seem to have been responsible for the establishment of any of the major branches (Stovel 1959: 96, 102). The impetus from staples production carried over into later periods as well. Not only between 1900 and 1913, but also 1920-29 and 1946—56, the Canadian manufacturing sector grew as domestic markets developed around the export raw material base, replacing imports (Watkins 1963: 158). Pulp and paper production, as well as mining and natural gas, led to the development of forward linkage and in addition turned out to stimulate each other (Caves and Holton 1959: 46). The United States constitutes an even better example of how trade in staples created extensive linkages across a wide territory (North 1966). The first impulses were generated in the cotton kingdom of the South before the Civil War. A growing international demand for cotton raised incomes and also created a demand for inputs not produced in the region. The South provided neither the food it needed nor the manufactured consumer goods, so linkages were established both with the wheatgrowing West and the industrialized Northeast. The western frontier could, however, not be developed without previous extension of the canal and railroad network, pulling migrants as well as international capital (North 1956) along in the process. Stapleinduced growth was at work. The process continued after the Civil War, however, with slightly different protagonists. Now, wheat itself, together with livestock and minerals, notably coal, was a staple, with migrants pushing the frontier westwards, along with the transportation network, in a movement that would not cease until the West had been settled, around 1890. Simultaneously a widening of the market for north-eastern manufactures took place (Perloff et al. 1960: Chapters 12-14).

6.2.2. Larger southern temperate economies To what extent the staples theory applies to the case of Australia has been subject to some controversy. As a minimum however, it appears to apply to the 1850s, the 1870s, part of the 1880s and the years immediately before World War I (Lougheed 1968). Yet, even though growth took place in rushes and booms, the entire period beginning in the 1830s and extending up to World War I was characterized by a substantial extension of the resource base. Gold and wool provided the stimulus (McLean 1989),

Economic Development: 1870-1914

101

and immigrants (Pope and Withers 1994; Taylor 1994) and capital (Pope 1990) flowed in to help with the extension of the frontier on the outback and the extraction of the precious metal. The overwhelming majority of the immigrants were British. (Australia maintained a 'whites only' policy.) Most of these ended up in urban-based jobs (Pope and Withers 1994: 243). A diversified industrial sector was built up, under the natural protection of distance. At certain points this sector came under strain. Thus, the gold rush in the 1850s produced Dutch disease effects for a full decade. It stimulated imports of a number of manufactured consumer goods, to the detriment of domestic producers, but spurred the output of industries related to construction that were producing nontradables (Maddock and McLean 1984: 1059-60). At any rate, between 1860 and 1890 the industrial sector picked up again. In 1891, one out of every three Australians lived in an urban centre and less than one-third of the wage earners worked in mining or rural production (Denoon 1983: 147). This trend was to continue and in 1910 almost one-fifth of the labour force was employed in manufacturing (Denoon 1983: 153). It was, however, still the primary sector that was most important economically, being the only exporter. When World War I broke out, the Australian economy continued to be dominated by pastoral, agricultural, and mining production, although a diversification had taken place, with mutton, butter, wheat, and sugar having been added to wool and gold (and other minerals) on the export list (Denoon 1983: 100-4). Argentina attracted both capital (Taylor 1992) and large numbers of immigrants (Taylor 1994) who peopled the Pampas (Jefferson 1926) as the agricultural frontier was extended. Between 1890 and 1913 half the population growth of the country was due to immigration (Taylor 1994: 97). From the 1860s a number of staples appeared on the scene: hides, wool, salted meat, wheat (notably in the 1890s), corn, linseed, and frozen meat. Linkages were not missing. Extending the frontier required railroads (Cuccorese 1969), and exportation required frigorificos, packing facilities etc. Forward and backward linkages were developed from livestock and agricultural products (Denoon 1983: 157—9), and domestic manufacturers began to compete with imports in the market for consumer goods (Diaz-Alejandro 1970, 1984). The conventional wisdom in the Argentine case is that the industrialization process dates from the 1930s (e.g., Ferrer 1963). This has, however, been questioned by a number of authors (see the references in Cortes Conde 1992) who have argued that an import-substitution process was at work already before World War I. High growth rates of manufacturing output were obtained as domestic production of consumer goods replaced imports, in the face of increasing and diversifying demand (Cortes Conde 1985: 349-51) and linkages to primary production were exploited (Denoon 1983: 157-9). While the industrial sector remained small in absolute terms in 1913, a beginning had been made.

6.2.3. The mineral exporters We now turn to the three mineral exporters. In the case of Bolivia the initial expansion came from silver mining during the early days of the Spanish Empire. A revitalization

102

R. Findlay and M. Lundahl

took place after 1850 due to cheaper steam engines making it possible to reopen closed and flooded mines, while a reduction in the cost of mercury permitted cheaper extraction of refined silver from the ore. Foreign capital and expertise (from Chile and Peru) and reinvestment of domestic agricultural profits led to further expansion of this sector, with new mines being opened on the Pacific coast. The output of silver increased nine-fold from the 1850s to the 1890s (Klein 1986: 554-7). This expansion created linkages through increased demand for domestic agriculture and, more importantly, to railroad construction, which further widened agricultural markets (Klein 1986: 557-9). The reduction of high internal transport costs in a mountainous country stimulated the production and export of a new mineral, tin, to meet the rapidly growing world demand for canning. The labour skills and technological knowledge acquired in silver mining were capable of being transplanted to the new mineral. Exports of tin grew at almost 6 per cent per annum in the first two decades of the twentieth century (Klein 1986: 566). In the end, however, silver and tin exports did not transform the Bolivian economy because mining failed to employ sufficient workers so that the domestic market was too small to permit any industrialization to speak of, even as late as 1950 (Klein 1986: 585-6). Bolivia's neighbour Chile gained control over nitrate deposits belonging to Bolivia and Peru in the so-called War of the Pacific from 1879 to 1883. Chile thereby acquired a virtual world monopoly, with exports expanding twelve-fold between 1880 and 1913 and employment growing to over 50,000 (Mamalakis 1971: 184; Cariola and Sunkel 1985: 186). The expansion of nitrate production and exports raised per capita incomes at above 1 per cent per annum, stimulating domestic agricultural and industrial production, which almost doubled between 1880 and 1913 (Kirsch 1973: 49). In contrast to Bolivia, Chilean mining did have a significant impact on the rest of the economy through a variety of linkage effects. Finally, South Africa became a mining economy during the last quarter of the nineteenth century after the discoveries of diamonds in Kimberley in 1867 and gold on the Witwatersrand in 1886. Mining employment rose from virtually nil in 1870 to 325,000 in 1912 (Hobart Houghton 1971: 19). The sources of the labour force were internal migration, neighbouring African regions and Europe for skilled personnel. Foreign capital played a major role, especially in gold mining (Frankel 1938). The diamond and gold fields required a railroad network to exploit them successfully. Road construction was also necessary to provide food supplies to the mining regions. Manufacturing production was also stimulated, not only by demand but also by the supplies of technical skills and entrepreneurial experience developed in the mining sector (Nattrass 1981: 164).

6.3. THE TROPICAL COUNTRIES 6.3.1. Latin American commodity exporters The experience of the tropical countries was less uniform than that of the regions of recent settlement. This is not surprising since the structure of their economies and the

Economic Development: 1870-1914

103

nature of their export crops differed widely from country to country. We begin with three Latin American countries that all relied initially on coffee as their primary staple. In Brazil the volume of coffee exports almost tripled over the 1870-1914 period. Since she supplied more than half of world exports Brazil could exert considerable monopoly power. The locus of production shifted from Rio to Sao Paulo and from dominance by large estates to considerable smallholder participation. The coffee sector attracted around 1.5 million immigrants between 1880 and 1913 (Coes 1970:102—3). The state supported the expansion of coffee exports by currency devaluations, price stabilization schemes, and railway extension into the coffee-growing areas. Coffee exports were supplemented by rubber exports as well, production and employment in this sector growing rapidly in the Amazonas region until Asian competition led to a precipitous fall in prices in 1910 (Goes 1970: 119-20). The coffee boom did bring incipient industrialization to Brazil. Sacks for packing the coffee and clothing for the workers stimulated the growth of a local textile industry that supplied over 75 per cent of the country's consumption of cotton textiles in 1913. Food processing, metalworking, and chemical industries were also established quite early. Profits from coffee growing were also reinvested to some extent in manufacturing, with foreign exchange earned by the coffee exports providing the necessary imported capital goods. On the whole, the results of the coffee and rubber booms in terms of overall growth was modest, not exceeding 1 per cent of GDP per capita (Cat227o 1992: 1). In the corresponding regions, however, coffee was responsible for a growth rate of 1.5 per cent and rubber of 4 per cent, a figure that could not be sustained (Coes 1970: 123-5). Coffee, supplemented by bananas was also the main export of Costa Rica. The two export staples jointly benefited from the extension of railroads and banking services but the demand generated by the growth of staples was not sufficient to lead to domestic industrialization in such a small country as Costa Rica. As shown by the work of Buhner-Thomas (1987) and Cardoso (1977), coffee and bananas in Costa Rica were classic enclaves that did little to transform the rest of the economy. Elsewhere, in Colombia, coffee also dominated exports. It was a smallholder crop, fitting into peasant agriculture along with food crops (Harbison 1970: 79). Exports expanded eighteen-fold between 1878 and 1914 and the widespread diffusion of export earnings among the farmers stimulated handicrafts and small industry. But only the Medellin area developed any significant manufacturing activity before 1914, so that growth was not impressive in the aggregate.

6.3.2. New peasant and plantation crops in Asia In Ceylon, the rise of three new staples: tea, coconut products, and rubber, displaced the ailing coffee sector in the 1880s. Rising demand stimulated the extension of tea cultivation, notably between 1884 and 1906 (Corea 1975: 71-5). Rubber, benefiting from strongly rising prices, as the automobile industry expanded, joined tea as an important export product in the 1890s and the cultivated area grew rapidly up to 1910. Together with coconut products, tea and rubber made for a rapid growth

104

R. Findlay and M. Lundahl

of exports whose dollar value rose at 5.4 per cent annually between 1880 and 1913 (Craig 1970: 221). Two of the new staples were excellently suited to smallholder production. Capital requirements were low in both rubber and coconut production, and relatively little labour was needed. Tea required both more labour and more capital per unit of land (Craig 1970: 228). Capital had to come from abroad and the plantations also absorbed large numbers of Indian (Tamil) immigrants who numbered 366,000 by 1911 (Craig 1970: 230). The native Sinhalese avoided plantation work because they had better options remaining in their villages. As elsewhere, the government was active in the extension of railroads, roads, irrigation facilities and port facilities (Reynolds 1985: 137; Bruton el al. 1992: 21). The Ceylonese plantation sector has often been pointed to as a classic example of enclave development, importing even the food for its workers. However, as Reynolds (1985: 137—8) has emphasized, the export sector as a whole is likely to have had important spread effects, since it made it possible for smallholders to increase their incomes, and it led to the creation of a domestic planter class in the coconut and rubber sectors. This, in turn, may have widened the domestic market, and some industrial development in fact took place: employment in manufacturing grew at 2.8 per cent annually between 1881 and 1911 to reach 170,000, 10 per cent of the workforce (Craig 1970: 245). Altogether, GDP per capita may have increased at some 2 per cent per annum (Reynolds 1985: 137). Malayan rubber exports began in 1906, and expanded extremely fast until the beginning of the 1920s (Gunnarsson 1983: 6), initially on plantations. British capital combined with cheap Tamil labour from India to make use of virgin lands, mainly on the West Coast where a good infrastructure, in the form of roads, railroads, and ports, had already been established to serve tin exports (Bruton et al. 1992: 178). With time, Chinese and Indian growers became involved as well, both on plantations and on small farms, where rubber could easily be fitted into the existing growing pattern. After 1910, smallholder output, which by then also included Malays, grew exceedingly fast until World War II, despite government measures to the contrary (Gunnarsson 1983:21). Malaya had a second staple, tin, which had been mined mainly by Chinese entrepreneurs and workers. In 1911, when Malaya accounted for half the world production, the Chinese composed over 95 per cent of the tin mining community (Bruton et al. 1992: 177). After 1912, however, capital-using, technological change that required skilled workers caused the British to begin to take over. Yet, overall, neither rubber production nor tin mining did much to establish linkages. Chinese and Indian workers ate imported rice (which differed from the domestic variety consumed by the Malays) and manufacturing failed to develop. As late as 1947, less than 7 per cent of the labour force was employed in 'industry', which actually consisted mainly of handicraft operations and small establishments (Reynolds 1985: 152).

6.3.3. Peasant crops Some of the most successful Asian staple exporters grew by simply expanding the output of their main food crop, rather than through imported staples. Such was the

Economic Development: 1870-1914

105

case for example with Burma (Furnivall 1957; Hlaing 1964; Adas 1974). In the midnineteenth century, Lower Burma was a sparsely populated, potentially fertile area. The British occupied the territory in 1852 and, together with the opening of the Suez Canal in 1869, triggered change. The availability of European manufactures increased. An upward pressure on the price of rice turned it into a commercial crop. Cultivators from Central and Upper Burma were induced to move southwards, and labourers (both permanent and temporary) were imported from southern India (from the 1880s) with the aid of subsidies. Transportation facilities (railroads, canals, etc.) were improved (Hlaing 1964: 32-6). The expansion process quickly reached the point where it fed on itself without any further need for intervention. The area under rice cultivation grew eight-fold between 1852 and 1915 (Hlaing 1964: 6) and Burma became the leading rice exporter in the world. Yet, with the exception of rice milling and sawing of logs, rice production failed to produce any substantial linkages. Handicrafts and cottage industries failed to meet the competition from imported manufactures (Furnivall 1957: 161—2). Although per capita income may have grown at 0.8 per cent annually during 1900—30 (Hlaing 1964: 51, Appendix A), the benefits may not have been widely shared because a process of land concentration was operating (Furnivall 1957: Chapter 6) and real wages may have stagnated or even fallen (Hlaing 1964: Chapter 2). Elsewhere, in Siam in 1851 the ascendance of a modernizing king to the throne opened the country up for international trade. Rice exports grew fifteen-fold from the late 1850s to World War I in response to the opening of the Suez Canal and the concomitant cheapening of European manufactures. The share of exports in total output rose from about 5 per cent in 1850 to 50 per cent at the turn of the century (Reynolds 1985: 158). Tin and teak wood supplemented rice on the export list, both with upward trends that were as steep as the one for rice exports. Siam could draw on unused land in the expansion process, first in the central plain and thereafter east of Bangkok. The Crown freed the agrarian population from forced labour services for the nobility and facilitated access to these lands for smallholders, refusing to create a landed aristocracy (Phongpaichit and Baker 1995: 3, 23-5). Canals, mainly for drainage, were dug and railroads constructed, financed by bond issues in London. Migration played a comparatively smaller role in Siam than in Burma. Capital needs were comparatively small, in the main limited to a plough and a bullock, which could usually be financed by a local merchant (Phongpaichit and Baker 1995: 32). Once again, a rice producing economy failed to develop any linkages into manufacturing (Akira 1996: 33-9). On the contrary, hand in hand with the expansion of exports went an expansion of imports of European manufactures that sent domestic industries, like sugar refining and cotton weaving, into decline. In 1919, Bangkok could boast of a mere seven modern factories. As in the Burmese case, the only offshoots of rice production were saw milling and rice milling. The Siamese economy grew by absorbing the growing population into rice cultivation, not by diversifying into manufacturing (Ingram 1971). Peasant economies, however, also managed to develop export crops that were not originally part of their agricultural tradition. One of the best examples of this is cocoa

106

R. Findlay and M. Lundahl

in the Gold Coast. Cocoa production rapidly took off in the 1880s and by 1913 the Gold Coast was the world's leading cocoa exporter (Hill 1963). Between 1891 and 1913 real income per head may have grown at 1.8-1.9 per cent annually (Szereszewski 1965: 111, 127), driven by cocoa (and to some extent gold, the production of which, however, was very enclave-like) (Holmes 1970: 151-2). Cocoa was easy to integrate into the already existing agricultural systems. Traditional food crops and the new export products could coexist without problems (Szereszewski 1965). Labour was secured by a migration to the southern cocoaproducing region where forest land that could be cleared was plentiful (Hill 1963). In the beginning, little capital was needed, but as the industry expanded, European merchant houses gradually moved in and provided credit for investment and consumption during the months preceding the harvest (Gunnarsson 1978: 51-3). Transportation was no problem, since the production areas remained within about eighty kilometres from the coast (Holmes 1970: 152). By the same token, however, few linkages developed. Railroads did not arrive in the Gold Coast until after 1900 and then also, only to serve the needs of gold mining (Holmes 1970: 163—5).

6.4. CONCLUSION S We now attempt to draw the lessons of the 1870-1914 experience that we have surveyed and relate it to the theoretical framework. The four countries of recent settlement present a relatively uniform pattern. All based their export activities on the existence and extension of an unexploited land frontier. All received significant capital imports and labour inflows from Europe and all were beginning the process of industrialization by the end of the period, most notably of course in the case of the United States, where exports of manufactures increased dramatically from about the turn of the century. Primary exports provided an impulse that was not confined to that sector itself but spread to the rest of the economy through linkage effects of various kinds. The three mineral economies also grew by extending their relevant 'frontiers', though these were in mining instead of agriculture. Here again reliance was placed on foreign capital inflows but immigration from Europe was on a relatively smaller scale, with internal migration playing a more significant role. South Africa drew on labour supplies from other parts of Africa to dig the minerals, with European migrants confined to the more skilled occupations. The linkage from the primary export to manufacturing was operative, except for Bolivia, but not as strong as in the case of the four agricultural exporters. The experience of the regions of recent settlement can easily be linked both to the Christopher Columbus model and the staples model. In the former (Findlay 1993), the initial equilibrium established after the 'discovery' is perturbed by three further exogenous shocks—an increase in the total labour force of the system, an 'industrial revolution' in European manufacturing and an extension of the 'frontier' made possible by a reduction in the cost of clearing new land due, say, to a transport innovation such as the introduction of the railway. The combined effects of these shocks is an

Economic Development: 1870-1914

107

increase in the supply of land in 'America', an increase in migration to America and an inward flow of capital to provide the infrastructure underlying the extension of the frontier. The supplies of food and raw materials to Europe both increase. The effects on product and factor prices of these shocks are not always in the same direction. Real wages, for example, are reduced by the first shock but raised by the other two. Since we know that real wages rose on both sides of the Atlantic, this must be because the positive effects of the last two shocks outweighed the negative effects of the first. Specific quantitative results could only be obtained by a numerical 'calibration' exercise, that is not undertaken here but which would be both possible and desirable in further research. Findlay (1995: Chapter 5), and the staples model of Findlay and Lundahl (1994), extend the model to allow for manufacturing production in the New World. In the staples model agricultural production also expands when transport costs are reduced, trade is opened and the relative price of the staple increases. At the same time, this tends to reduce manufacturing output. However, when the international capital market is opened up, at given international commodity prices, manufacturing production receives an impulse as well, and if the rate of interest falls more than the value of the marginal product of land, as labour leaves agriculture, output in the latter sector may expand as well, since it then pays to extend the frontier. Immigration, in turn, as pointed out in the theoretical section, is predicted to be directed primarily towards agriculture but could spill over into manufacturing once the demand for labour in agriculture has been satisfied. Thus, in the staples model as well, it is easy to envisage a sequence where growth in the regions of recent settlement comes from an interplay between staples production and manufacturing. (The same argument can be made with the aid of the 'extended' Christopher Columbus model.) The patterns characterizing the tropical countries are not as easy to sum up because of the greater variety of economic structures. Nevertheless, a few common themes emerge. The most salient one is that despite a significant growth of primary exports in all cases, the growth impulse was not strong enough to lead to industrialization, except in the case of Brazil, which is a country that in this respect cornes close to the patterns of the Christopher Columbus or staples models. However, the open dual economy model is of relevance here as well, since economic expansion in Brazil was obviously constrained by the world demand for coffee. As far as the rest of our tropical countries are concerned, they are better modelled in terms of the variety of the staples model that does not incorporate any manufacturing sector. (Colombia is a bit of an 'intermediate' case.) There, the expansion of staples production draws on 'unlimited' supplies of both land (the agricultural frontier) and labour (notably through international or internal migration). The only thing that happens is either that the peasant sector starts to develop cash crop production or that a plantation sector springs up. Obviously the crucial difference between the two groups of countries is whether a manufacturing sector emerges or not. Several reasons for this disparity in performance come to mind. Perhaps the most obvious is that the per capita resource endowment of the regions of recent settlement, despite large inflows of migrants, was much higher on average than in the tropical countries, where most had already fairly dense

108

R. Findlay and M. Lundahl

populations. The necessary internal market for domestic manufacturing was thus available much earlier in the temperate rather than the tropical primary exporters. Significantly, real wages did not grow substantially either, with tropical cash crops drawing their labour supplies from the peasant sector on very favourable terms, and Indian and Chinese immigration providing a further check on upward wage pressure. Given that the initial per capita incomes were substantially lower in the tropics it would be wrong to conclude that their continued dependence on staples and failure to industrialize by 1914, reveals any defects in the strategy of development through specialization and comparative advantage that the liberal mainstream approach in economics recommends. Why then are they still so far behind today, despite the remarkable growth of the last few decades, particularly in Asia as exemplified byMalaysia and Thailand? Arthur Lewis (1970: 44) attributes the long lag to the impact of the two world wars and the Great Depression in between. Given another opportunity during the 'second golden age' of the world economy of the last few decades it would seem that several of the tropical countries have achieved the transformation to manufactured exports and industrialization that had eluded them before 1914. Although several tropical countries such as Burma, Sri Lanka, and Ghana continue to lag substantially in their transitions, the fault lies more with sustained policy errors and internal problems rather than with natural resource dependence (Myint 1967; Lal and Myint 1996). Natural resources as such, however, may delay the emergence of labour-intensive manufactured exports, for some of the 'political economy' reasons discussed in Chapter 8 and also by Findlay and Wellisz (1993). The rents generated by the natural resource sectors divert policymaking into competition between interest groups for the spoils whereas Hong Kong and Singapore, for example, had no alternative but to concentrate on manufactured exports from the beginning. The three Latin American tropical countries do not present a common picture. Brazil has succeeded in industrializing through the post-war years, despite many inefficiencies and imbalances linked to trade policy closure and an unstable macroeconomic framework. Costa Rica has had a high initial level of per capita income but import substitution based on the small Central American Common Market had its inherent limitations and a rapid expansion of the public sector (in combination with an unfavourable external environment) led to a crisis in the 1980s that retarded growth (Chapter 14). Colombia, finally, has suffered from chronic internal violence and instability caused by the drug trade but has still managed a fairly impressive growth performance. In the southern regions of recent settlement, Argentina wasted her early golden years as a result of the distortionary policies of the Per96n era and the aftermath of macroeconomic instability (Chapter 13). In the 1990s however, growth has been rapid and sustained in the context of monetary and exchange rate stability. Once again, we see the importance of the right policy choices. Elsewhere, among the mineral economies Chile has clearly been the biggest success, once again as a consequence of the market-oriented policies undertaken by the military regime and continued by its democratic successors. Bolivia has had spurts of growth and reform but does not seem to be on a sustained path to stability and prosperity (Chapter 11). The future of

Economic Development: 1870-1914

109

post-apartheid South Africa remains uncertain, with strong pressures for redistribution towards the poor black majority—a redistribution that requires a faster rate of economic growth (Chapter 15). Despite many disappointments in the performance of resource-abundant countries outside of the 'regions of recent settlement' we believe it is clear that pessimism about resource-led growth is firmly refuted by the historical record that we have outlined. Even if growth is 'dependent', the stable growth of the 'North' has meant that the 'South' has had the opportunity to grow as a consequence. The performance of the laggards is clearly due more to internal problems and policy failures than to lack of opportunities in the world market. We therefore feel that the mainstream liberal framework, with a dose of interventionism, has on the whole been vindicated. International trade and factor movements were instrumental when it came to raising the rate of growth in resource-abundant economies during the late nineteenth and early twentieth century, albeit under circumstances that may have been fairly unique to that period. That some countries performed badly later is a different story. They made the wrong policy choices and either failed to industrialize, or industrialized in an inefficient manner. That is, they failed to make the transition that has been critical for making the best use of trade in the second half of the twentieth century when international exchange of goods and services has indeed been an engine of growth again—perhaps even more than it was during the 'golden age' of 1870—1914. The next chapter further explores the nature of the relationship between natural resources, policy and economic growth with reference to a short-run macroeconomic model.

REFERENCES Adas, Michael (1974), The Burma Delta: Economic Development and Social Change on an Asian Rice Frontier, 1852-1941, Madison, WI: University of Wisconsin Press. Akira, Suehiro (1996), Capital Accumulation in Thailand, 1855-1985, Bangkok: Silkworm Books. Bruton, Henry J. (1992), in collaboration with Gamini Abeysekera, Nimal Sanderatne and Zainal Aznam Yusof, The Political Economy of Poverty, Equity, and Growth: Sri Lanka and Malaysia, New York: Oxford University Press. Buhner-Thomas, Victor (1987), The Political Economy of Central America since 1920, Cambridge: Cambridge University Press. Cardoso, Ciro F. S. (1977), 'The formation of the coffee estate in nineteenth-century Costa Rica', in Kenneth Duncan and Ian Rutledge (eds.) with the collaboration of Colin Hardin, Land and Labour in Lain America. Essays on the Development of Agrarian Capitalism in the Nineteenth and Twentieth Centuries, Cambridge: Cambridge University Press, 165-202. Cariola, Carmen and Osvaldo Sunkel (1985), 'The growth of the nitrate industry and socioeconomic change in Chile, 1880-1930', in Roberto Cortes Conde and Shane J. Hunt (eds.), The Latin American Economies: Growth and the Export Sector 1880—1930, New York and London: Holmes & Meier, 137-254. Cat227o, Luis (1992), 'The failure of export-led growth in Brazil and Mexico, c. 1870-1930', Research Paper 31, London: University of London, Institute of Latin American Studies.

110

R. Findlay and M. Lundahl

Caves, Richard E. and Richard H. Holton (1959), The Canadian Economy: Prospect and Retrospect, Cambridge, MA: Harvard University Press. Goes, Donald (1970), 'Brazil', in W. Arthur Lewis (ed.), Tropical Development 1880-1913, London: George Allen and Unwin, 100-27. Corea, Gamani (197S), The Instability of an Export Economy, Colombo: Marga Institute. Cortes Conde, Roberto (1985), 'The export economy of Argentina 1880-1920', in Roberto Cortes Conde and Shane J. Hunt (eds.), The Latin American Economies: Growth and the Export Sector 1880-1930, New York and London: Holmes & Meier, 319-91. Cort223s Conde, Roberto (1992), 'Export-led growth in Latin America: 1870-1930', Journal of Latin American Studies, 24, Quincentenary Supplement, 163-79. Craig, J. Edwin (1970), 'Ceylon', in W. Arthur Lewis (ed.), Tropical Development 1880-1930: Studies in Economic Progress, London: George Allen and Unwin, 221-49. Cuccorese, Horacio Juan (1969), Historia de los ferrocarriles en la Argentina, Cordoba: Ediciones Macchi. Denoon, Donald (1983), Settler Capitalism: The Dynamics of Dependent Development in the Southern Hemisphere, Oxford: Clarendon Press. Diaz-Alejandro, Carlos F. (1970), Essays on the Economic History of the Argentine Republic, New Haven, CT and London: Yale University Press. Diaz-Alejandro, Carlos F. (1984), 'No less than one hundred years of Argentine economic history plus some comparisons', in Gustav Ranis et al. (eds.), Comparative Development Perspectives: Essays in Honor of Lloyd Reynolds, Boulder, CO: Westview Press: 328-61. Ferrer, Aldo (1963), La economia argentina: Las etapas de su desarrollo y problemas actuales, Mexico and Buenos Aires: Fondo de Cultura Economica. Findlay, Ronald (1973), International Trade and Development Theory, New York and London: Columbia University Press. Findlay, Ronald (1993), 'International trade and factor mobility with an endogenous land frontier: some general equilibrium implications of Christopher Columbus', in Wilfred J. Ethier, Elhanan Helpman and J. Peter Neary (eds.), Theory, Policy and Dynamics in International Trade, Cambridge University Press: 38-54. Findlay, Ronald (1995), Factor Proportions, Trade, and Growth, Cambridge, MA: MIT Press. Findlay, Ronald and Mats Lundahl (1994), 'Natural resources, "Vent-for-Surplus", and the staples theory', in Gerald M. Meier (ed.), From Classical Economics to Development Economics, New York: St. Martin's Press: 68-93. Findlay, Ronald and Mats Lundahl (1999), 'Resource-led growth—a long-term perspective: the relevance of the 1870-1913 experience for today's developing economies', UNU/ WIDER Working Paper 162, Helsinki: UNU/WIDER. Findlay, Ronald and Stanislaw Wellisz (eds.) (1993), The Political Economy of Poverty, Equity, and Growth: Five Small Open Economies, Oxford: Oxford University Press. Frankel, S. Herbert (1938), Capital Investment in Africa, London: Oxford University Press. Furnivall, J. S. (1957), An Introduction to the Political Economy of Burma, Third Edition, Rangoon: People's Literature Committee & House. Gunnarsson, Christer (1978), The Gold Coast Cocoa Industry, 1900-1939. Production, Prices and Structural Change (Ph.D. thesis), Lund: Department of Economic History, Lund University. Gunnarsson, Christer (1983), 'Malaysian rubber production: patterns of growth 1900-1975' (mimeo), Lund: Department of Economic History, Lund University. Hansen, Bent (1979), 'Colonial economic development with unlimited supply of land: a Ricardian case', Economic Development and Cultural Change, 27, 611—27.

Economic Development: 1870-1914

111

Harbison, Ralph W. (1970), 'Colombia', in W. Arthur Lewis (ed.), Tropical Development 18801913. Studies in Economic Progress, London: George Allen and Unwin, 64-99. Hill, Polly (1963), Migrant Cocoa-Farmers of Southern Ghana: A Study in Rural Capitalism, London and New York: Cambridge University Press. Hlaing, U. Aye (1964), 'A study of economic development of Burma, 1870-1940', (mimeo), Rangoon: Department of Economics, University of Rangoon. Hobart Houghton, Dennis (1971), 'Economic development, 1865-1965', in Monica Wilson and Leonard Thompson (eds.), The Oxford History of South Africa, Vol. II: South Africa 18701966, Oxford: Oxford University Press: 1-48. Holmes, A. Baron (1970), 'The Gold Coast and Nigeria', in W. Arthur Lewis (ed.), Tropical Development 1880-1930: Studies in Economic Progress, London: George Allen and Unwin: 147-77. Ingram, James C. (1971), Economic Change in Thailand, 1850-1970, Stanford, CA: Stanford University Press. Innis, Harold A. (1930), The Fur Trade in Canada: An Introduction to Canadian Economic History, New Haven, CT: Yale University Press. Innis, Harold A. (1940), The Cod Fisheries: The History of an International Economy. New Haven, CT: Yale University Press. Jefferson, Mark (1926), Peopling the Argentine Pampa, Port Washington, NY and London: Kennikat Press. Kirsch, Henry William (1973), The Industrialization of Chile, 1880-1930 (Ph.D. thesis), Gainsville: University of Florida. Klein, Herbert S. (1986), 'Bolivia from the war of the Pacific to the Chaco war, 1880-1932', in Leslie Bethell (ed.), The Cambridge History of Latin America. Volume 5, c. 1870 to 1930, Cambridge: Cambridge University Press: 553-86. Lai, Deepak and Hla Myint (1996), The Political Economy of Poverty, Equity, and Growth: A Comparative Study, Oxford: Clarendon Press/Oxford University Press. Lewis, W. Arthur (1954), 'Economic development with unlimited supplies of labour', Manchester School of Economic and Social Studies, 22, 139—91. Lewis, W. Arthur (ed.) (1970), Tropical Development 1880-1913: Studies in Economic Progress, London: George Allen and Unwin. Lougheed, A. L. (1968), 'International trade theory and economic growth', Australian Economic History Review, 8, 99-109. Maddock, Rodney and Ian McLean (1984), 'Supply side shocks: the case of Australian gold', Journal of Economic History, 44, 1047-67. Mamalakis, Markos J. (1971), 'The role of government in the resource transfer and resource allocation processes: the Chilean nitrate sector, 1880-1930' in Gustav Ranis (ed.), Government and Economic Development, New Haven, CT and London: Yale University Press: 181-215. McLean, Ian W. (1989), 'Growth in a small open economy: an historical view', in Chapman, Bruce (ed.), Australian Economic Growth, Melbourne: Macmillan, 7-33. Meier, Gerald M. (1953), 'Economic development and the transfer mechanism: Canada, 18951913', Canadian Journal of Economics and Political Science, 19, 1—19. Myint, Hla (1967), 'The inward and outward-looking countries of Southeast Asia', The Malayan Economic Review, 12, 1—13. Nattrass, Jill (1981), The South African Economy: Its Growth and Change, Cape Town: Oxford University Press.

112

R. Findlay and M. Lundahl

North, Douglass C. (1956), 'International capital flows and the development of the American west', Journal of Economic History, 16, 493-505. North, Douglass C. (1966), The Economic Growth of the United States 1790-1860, New York: Norton. Perloff, Harvey S., Edgar S. Dunn Jr., Eric E. Lampard, and Richard F. Muth (1960), Regions, Resources, and Economic Growth, Baltimore, MD: Johns Hopkins Press. Phongpaichit, Pasuk and Chris Baker (1995), Thailand: Economy and Politics, Kuala Lumpur and Oxford: Oxford University Press. Pope, David H. (1990), 'Australia's payments adjustment and capital flows under the international gold standard, 1870-1913', Working Papers in Economic History, 141, Canberra: Australian National University. Pope, David and Glenn Withers (1994), 'Wage effects of immigration in late-nineteenth-century Australia', in Timothy J. Hatton and Jeffrey G. Williamson (eds.), Migration and the International Market, 1850-1939, London and New York: Routledge, 240-82. Reynolds, Lloyd G. (1985), Economic Growth in the Third World, 1850-1980, New Haven, CT and London: Yale University Press. Rostow, Walt W. (1960), The Stages of Economic Growth: A Non-Communist Manifesto, Cambridge: Cambridge University Press. Stovel, John A. (1959), Canada in the World Economy, Cambridge, MA: Harvard University Press. Szereszewski, R. (1965), Structural Changes in the Economy of Ghana, 1891—1911, London: Weidenfeld and Nicolson. Taylor, Alan M. (1992), 'External dependence, demographic burdens, and Argentine economic decline after the Belle Epoque', The Journal of Economic History, 52, 907-36. Taylor, Alan M. (1994), 'Mass migration to distant southern shores. Argentina and Australia, 1870-1939', in Timothy J. Hatton and Jeffrey G. Williamson, Migration and the International Market, 1850-1939, London and New York: Routledge: 91-115. Watkins, Melville H. (1963), 'A staple theory of economic growth', Canadian Journal of Economics and Political Science, 29, 141—58.

7 Short-Run Models of Contrasting Natural Resource Endowments S. M A N S O O B MURSHED

7 . 1 . INTRODUCTIO N This chapter presents a short-run theoretical macroeconomic model of a small open economy endowed with a natural resource exporting sector to explore the links between natural resources, economic growth, and economic policy.1 It follows the sectoral delineation between traded and non-traded goods outlined in Sachs (1999). That paper attempts to differentiate between the economic development experience of East Asia and Latin America. The most important feature of Sachs' dichotomy is that East Asian economies rely more heavily on manufactured exports, whereas Latin America has a relatively greater share of natural resource intensive exports. According to Sachs (1999), Latin America, when compared to East Asia, exhibits a pattern of growth associated with natural-resource abundance. This in turn implies that its longrun growth rates will be less impressive than in East Asia. By way of general motivation, let us start from the premise that a country with a greater endowment of natural resources relative to the rest of the world may be blessed for the present, but will eventually be subjected to a 'winners curse'. In other words, its economic performance, in the long run, will be worse than that of countries that have a relatively lower resource endowment. Also countries placing a heavier reliance on natural resource based goods in their production structure will experience greater cyclical fluctuations in their trend national income, compared to nations that have a more diversified production base. The model in this chapter involves three sectors: a resource based commodity, a non-traded good and a traded goods sector. It examines short-run effects on the economy of resource booms, devaluation, and a consumption tax on the non-traded good. Within a single basic analytical framework, typologies are developed, distinguishing between what could be the East Asian and Latin American experience. The model is not a growth model involving the accumulation of produced, human and social capital. The short-run results contained in the chapter 1

A fuller and more technical version of the model in this chapter can be found in Murshed (1999).

114

S. M. Murshed

have implications, however, for long-run growth and development. Also, it is not a model involving game theoretic strategic interaction between various interested parties following a resource boom. The rest of the paper is organized as follows. Section 7.2 contains a review of the literature on macroeconomic models involving resource booms. Section 7.3 presents a sketch of the model, Section 7.4 discusses variations in the model's parameters. Finally, Section 7.5 includes a summary and some policy implications. 7.2. LITERATUR E REVIE W The debate on the superiority of manufacturing compared to agriculture and other natural resource based economic activities has gone on since the dawn of the history of economic thought. During the early days of development economics the combination of three reasons led to the belief that manufacturing should be encouraged at the expense of agriculture as part of a sound development strategy. First, the notion of surplus labour in agriculture (Lewis 1954) suggested that a costless shift of labour from agriculture to manufacturing could be engineered. Second, the work of Hirschman (1958) pointed out that manufactures generated beneficial forward and backward linkages and therefore that sector should be encouraged. Finally, and most influentially, Prebisch (1950) and Singer (1950), pointed to a secular tendency for primary goods prices to decline vis-a-vis manufactures, thus relying on primary goods exports was like backing the wrong horse. A dissenting voice could be found in Myint (1958), who asserted that agricultural exports provided the necessary spur for growth via a 'vent for surplus'. Present-day theorists emphasizing the growth retarding effects of relative natural resource abundance are, however, deeply opposed to the activist industrial policy and protectionism that followed from the suggestions of the earlier school advocating manufacturing led development. Their reasoning follows a different course. One such strand is associated with the effects of resource booms: the 'Dutch disease' literature, see Neary and Wijnbergen (1986) for example. The problem is not associated with a steady reliance on agriculture or resource based exports, but a sudden windfall gain. This may be due to a sudden, but temporary, increase in the price of oil and other primary commodities, as in the 1970s. Alternatively, it can also be associated with natural resource discoveries, increases in worker remittances or other unrequited international transfers. Irrespective of the cause, a resource boom crowds out the leading sector of the economy. In an open economy a massive current account surplus appears, leading to currency appreciation under a regime of flexible exchange rates. This renders existing (nonresource boom) exports even more uncompetitive in world markets. A shift in domestic output from tradables to non-tradables takes place. Why should this be a problem over time? As the model in Krugman (1987) illustrates, if there are learning by doing effects, a country whose manufacturing base is eroded during a resource boom can irreversibly lose competitiveness, even when the exchange rate reverts to normal after the boom has subsided. Thus, temporary resource booms cause hysteresis, a permanent loss of competitiveness. For developing countries, this means that their future potential for exporting manufactured goods and

Contrasting Natural Resource Endowments

115

diversifying the production base is stunted. If there are positive externalities from human-capital accumulation in manufacturing only, as in Matsuyama (1992), and resource booms retard the development of the more dynamic manufacturing sector, the growth path of the economy under free trade is lower than that of more resource poor countries. A related, but different economic literature concerns the so-called 'staple trap'. A staple trap implies reliance on a narrow range of 'staples' in national income and for export, or more simply an undiversified production base. The origins of the staple trap lie in a high natural resource endowment, including cropland. It is perfectly possible to have countries that are caught in a staple trap and therefore very poor at present, but that were at one time relatively well endowed with land or mineral resources. Since these staples are primary commodities, their prices are subject to considerable fluctuations and along with it the economic fortunes of the nations concerned. A staple trap also promotes policy errors, as discussed more fully in Chapter 8. A second strand of the modern literature on the 'winners curse' emphasizes the political economy aspects of corruption and inefficiency that follows a resource boom, see Auty (1997), as well as Sachs and Warner (1999a) on this. Revenues from a resource boom lead to rent seeking behaviour and directly unproductive activities. Ham-fisted attempts could also be made to reverse Dutch disease, via the pursuit of counterproductive and flawed industrial policies under a protective umbrella. In the paper by Sachs and Warner (1999a) a role for human capital is incorporated into a model with a non-traded sector, a traded good, and a purely exportable natural resource sector. Human capital accumulation, in the form of an externality, takes place as a result of traded/manufacturing production only. In contrast, one of the earliest treatments of human capital accumulation is in Shell (1966). In that model, knowledge is a public good, the accumulation of which benefits all sectors in the economy through cost savings. Acquiring knowledge is not a costless process as in Sachs and Warner (1999a), but being a public good, is entirely financed by the state via taxation of the productive sector in Shell (1966). Resource booms, in the Sachs and Warner (1999a) model retard the growth of the economy via the crowding out of production in the traded (manufactures) sector. The stock of human capital is diminished as employment in tradables declines; this in turn hampers future production of all goods, and hence the growth of the economy. Growth is, however, studied within the limited framework of a two-period model. The Sachs and Warner (1999b) paper permits increasing returns to scale in either of the two sectors of the economy (traded or non-traded), but not both. Increasing returns characterize the production of a range of intermediate inputs that could be employed in final production. This feature is similar to the Grossman and Helpman (1991) model, where, in a setting of monopolistic competition, varieties of new products are either innovated or imitated from abroad. The model then addresses whether resource booms can contribute towards 'big-push' type industrialization. A resource boom unambiguously expands the non-tradable sector, while at the same time shrinking the traded sector. If it is the expanding (non-traded) sector that uses these intermediate inputs, it may contribute to a successful big push. If the

116

S. M. Murshed

opposite is the case, and it is the traded sector which uses the intermediate inputs, big-pushes are less likely. Unless expectations about the future are optimistic, even the most propitious circumstances may not trigger accelerated industrialization or the big-push. Implicitly, these expectations are related to the political system and social capital. Rodriguez and Sachs (1999) present a Ramsey growth model where an economy undergoing a resource boom, which eventually peters out, enjoys higher consumption temporarily. The key assumption driving this result is that the country (for example Venezuela) does not accumulate assets abroad leading to a net creditor position in the international economy. If it did (Gulf States, say), it could permanently raise its consumption, based on asset income from abroad. 7.3. SKETC H O F TH E MODE L The economy is comprised of three sectors on the real side, two of which are traded goods and one a non-traded commodity. A monetary sector is also incorporated. The first traded good is denoted by R, for the natural resource based sector whose output is entirely exogenous and purely for export. R represents the value of exports from this sector in domestic currency units. The value of output in the resource-based sector is treated exogenously, following Sachs (1999), Sachs and Warner (1999a) and other treatments of 'Dutch disease' models, as in Neary and Wijnbergen (1986). R could also include foreign aid and other forms of unrequited transfers such as worker remittances from abroad. This view of the R sector, however, makes it less appropriate for modelling or analyzing agriculture, as output is exogenously given, and no supply factors such as land are taken into account. M indexes the other traded sector, which is both consumed domestically and exported. It is basically a labour-intensive manufactured good. In addition, there are consumption imports, C\; which compete with M in domestic consumption. M is produced utilizing labour only, following Sachs (1999), in order to capture the part played by labour intensive manufactured goods for export and domestic consumption. The price of M, PT, is normalized at unity, and this price is in any case given in a small open economy. Following Sachs (1999) the supply of M is described by:

LM represents labour employed in the M sector and 9 stands for the marginal value product of labour in that sector. The non-traded goods sector is represented by N, the production of which requires capital, labour, and an imported intermediate input (T), as in Sachs (1999). It therefore needs some foreign technological input, and is the capital-intensive sector by definition. In a sense the output of the N sector is more 'sophisticated' than in the other sectors, but perhaps that is precisely why it is non-traded. Note that manufacturing could lie within both the M and N sectors, and the 'real-life' counterpart of the non-traded sector is not restricted to public and private services. In summary, the output of the N sector could include government services, private services, as

Contrasting Natural Resource Endowments

117

well as some shielded but sophisticated manufacturing. For the sake of analytical convenience, fixed proportions characterize the use of the intermediate input from abroad in the N sector. See Findlay and Rodriguez (1977) for a discussion of production functions where an imported input enters in this fashion. The output of the N sector, in general reduced form, can be represented as:

Pn represents the price of the non-tradable good. The supply of N increases with PN but declines as the nominal exchange rate depreciates (E increases) as this makes the intermediate input more expensive. As far as the domestic value added of the N sector is concerned, this is obtained by subtracting the value of the intermediate input:

whereλ= ET, as PT = 1, PDN measures domestic value added in the TV sector. Turning to consumption, or the demand side, in the traded manufactured goods sector this is composed of domestic demand (CM) and foreign or export demand (XM):

Domestic demand for the output of the M sector depends positively on the price of the non-traded good,PN as well as income, Y. It is also positively related to the exchange rate, a rise in E represents devaluation, an increase in the cost of obtaining imported substitutes. Export demand is positively related to the nominal exchange rate. Equation (4), therefore, represents equilibrium in the M sector. It can be interpreted as demand, on the left-hand side equalling supply on the right-hand side, excess demand causes output to rise. In the non-traded goods sector, equilibrium supply equals demand is represented by:

Excess demand causes the relative price of TV, PN to be bid up to restore equilibrium. Domestic consumption of non-tradables is negatively related to its own price and positively linked to income. IN stands for investment, that is the savings leading to capital formation in that sector, negatively related to the interest rate (r). We need to specify the concept of national income, Y or GDP. This consists of domestic value-added in all three productive sectors, TV, M and R, less imports. Thus:

It is useful at this stage to define an overall price index (P), representing the aggregate cost of consumption of all three goods: imported consumption goods (priced by E), non-traded goods and domestic nonresource based traded goods prices. This price index is a cost of living or consumer price index. It represents the cost of purchasing a basket of goods comprising imported consumer goods, non-tradables and

118

S. M. Murshed

traded goods. The prices of the three consumption goods are represented by E, PN and PT respectively.2 The consumer price index is of use in measuring the real consumption wage and arriving at an appropriate definition of real money balances. The consumer price index takes the form:

The monetary sector for this economy takes the following form:

Equation (8) is exactly the same as the LM function for the economy. It represents equilibrium on the monetary side of the economy. Excess demand for money will push up equilibrium interest rates. Money demand, on the left-hand side of (8), is negatively related to interest rates and positively linked to Y. The latter also incorporates the wealth effect of resource booms, a rise in the value of R will not only impact on goods demand, but also raise asset (money) demand. H denotes the value of nominal balances. When it is deflated by the consumer price index P, we obtain the value of real balances. Note also that changes in the exchange rate will impact on real balances, for example nominal exchange rate depreciation or devaluation (rise in E) lowers real money supply. Finally we come to the balance of trade:

The left-hand side represents the trade balance or exports minus imports. There are two exports: the natural resource based exports as well as the manufactured traded sector's exports. The two imports are consumption and intermediate inputs respectively. F stands for the trade balance, which is positive if there is a trade surplus, negative if there is a deficit. We postulate a fixed exchange rate regime. This corresponds to the stylized facts for the vast majority of developing countries. Under a system of fixed exchange rates the balance of payments is a residual in the short run, improvements in the trade balance cause an increase in the stock of foreign exchange reserves, F and vice versa. Flexible exchange rates can, however, be easily incorporated. This will add an extra endogenous variable, E, into the system. E will rise (depreciate) with balance of payments deficits and vice versa. For the sake of tractability, we have omitted capital flows from Equation (9); this makes the trade balance equivalent to the balance of payments. Since the model that follows is not a growth model, we cannot meaningfully incorporate human capital If the accumulation of human capital, however, takes place only as a result of increased employment in the M sector (Sachs and Warner 1999a), we can say that an increase in the output of M increases the stock of human capital. 2 The exponents in P (β,αand 1 —α—β)represent the weights or shares of the three goods in the representative consumers consumption basket. They sum to unity.

Contrasting Natural Resource Endowments

119

It ought to be emphasized at this juncture that the distinction between point and diffuse natural resource based economies is similar to the Latin American and East Asian characterization contained in Sachs (1999).3 In Latin America there is a greater tendency to rely on exports of the resource sector. In the East Asian case the more important exportable is labour-intensive manufactures (M). Sachs (1999) also asserts that non-traded and imported consumption goods figure more largely in the Latin American consumption basket, relative to East Asia. Sachs (1999) states that in the typical East Asian economy it is M that is the labour intensive sector, whereas it is N in Latin America.

7.4. VARIATIONS IN PARAMETERS We now discuss the variations in parameters of a macroeconomic model.'

7.4.1 An increase in R This may come about as a result of natural resource discoveries, an increase in the world price of mineral exports, or increased transfers. The resource boom raises asset or money demand that pushes up domestic interest rates, r, which in turn dampens investment(capital accumulation) in the A*" sector. This possibility is not considered in Sachs and Warner (1999a, b), but can be found in Neary and Wijnbergen (1984). When we examine the combined effect of the rise in R on both the N and M sectors, four possibilities occur. First, both M and N could expand. This will happen if there is a high propensity to consume non-tradable goods, a low wealth effect of the resource boom on money demand, and a low interest rate response to investment in nontraded goods. The second possibility is for the non-traded sector to expand while the traded sector, M contracts; the Latin American case. This requires that the price elasticity of non-traded goods demand (C N1 ) is low or inelastic. In other words, the rise in PN as that sector expands does not reduce demand by very much. This is, of course the classic Dutch disease outcome. Third, there is the stylized East Asian example. Here it is the M sector that expands, whereas the N sector contracts. We need to have a high propensity to consume the domestically produced traded good, M. Finally, both the M and N sectors could contract. This is the most extreme form of Dutch disease and associated with strong wealth effects impacting on money demand, as in Neary and Wijnbergen (1984). It also has affinities to 'growth collapses' following resource booms (see Chapter 8). 3 It has to be remembered that many North East Asian economies, such as Taiwan never had much in terms of natural resources. The position is different in South East Asia, where several countries have natural resources that are 'diffused'. This means we have two East Asian types of economy Diffused natural resource rich and resource poor economies, both of which are relatively egalitarian (see Table 4.2). 4 In East Asia, over time, technologically and human capital intensive manufactured exports take over from unskilled labour intensive manufactured goods exports. See Tables 8.4 and 8.6. 5 Full technical details related to the results of this section can be found in Murshed (1999).

120

S. M. Murshed

The results above with respect to a rise in 7? on the N and M sectors can be depicted in terms of a diagram, in M and PN space. In Fig. 7.1, the AW and MM schedules represent equilibrium (supply equals demand) in the non-traded and traded goods markets respectively. They are both positively sloped, as an increase in either M or PN raises income and thus the demand for the other good goes up. The initial equilibrium in both markets occurs at the intersection point A. An upward movement in AW represents an expansionary effect on output in the N sector, NN0 moves to NN1. It reflects the fact that more N is demanded for each level of M. In the M sector, a rightward movement signals expansion from MM0 to MM 1 . This indicates that a greater quantity of M is demanded for each level of N produced. If the resource boom produces an expansion in both sectors we arrive at point C in the new equilibrium following the rise in R. Contractionary effects on output are depicted by leftward movements in the MM to MM2, and downward movements in NN to NN2- If both sectors contract, point B indicates final equilibrium. The Latin American or Dutch disease outcome is shown at point E, with the non-traded sector expanding but the traded sector contracting. The diametrically opposed East Asian case is depicted at point D in Fig. 7.1. Finally, we come to the impact on the trade balance (also the balance of payments) of the resource boom. If the impact of the rise in R on total income, Yis both positive and very large, and the propensity to import consumer goods (CF2) is substantial, then there could be a reversal of initial improvements in the balance of payments. In the East Asian case, the trade balance may improve as long as CF2 is not very high.

Figure 7.1. Effects of resource booms, devaluation, and industrial policies

Contrasting Natural Resource Endowments

121

7.4.2. Devaluation Policy based exchange rate depreciation can be motivated by a variety of reasons, including balance of payments crises, the desire to improve competitiveness and attempts to cope with debt servicing. It could also be part of a programme of structural adjustment or efforts to counteract Dutch disease. Devaluation, which is an increase in E, will from Equation (8) lower the value of real money balances, and hence put upward pressure on the interest rate r. Note that devaluation, at least upon impact, lowers the real wage as the price of imported consumption goods increases. It will also make the intermediate import more expensive in terms of domestic currency. There are two opposing effects of devaluation upon imports: one negative on the supply-side as imported inputs cost more domestically; the other is the positive impact devaluation has by reducing consumption imports, which become more expensive in terms of the home currency. If the negative impact of devaluation on the non-traded sector dominates its positive effect via reducing consumption imports, we have, in the words of Krugman and Taylor (1978), 'contractionary' devaluation. It is also the classic Latin American 'structuralist' outcome. The converse is more likely in East Asia, as the nontraded sector is less significant there, and consumption imports are likely to be highly price-elastic. In the East Asian case both sectors are likely to expand, as indicated by a shift from point A to C in Fig. 7.1. The increase in the traded goods sector will, however, be the greater of the two. In the Latin American case there could be a negative impact in one or both sectors of the economy. If the contractionary effect is only in the non-traded commodities, point D will be the new equilibrium in Fig. 7.1. This is the opposite of the Dutch disease effect of a rise in R in the stylized Latin American case. If both sectors decline, the new equilibrium is at point B. The impact of devaluation on the balance of payments is unclear. If imports are inelastically demanded then the balance of payments worsens, however if they are elastically demanded, the trade balance improves. There is the additional payment, which is needed to finance intermediate inputs for N production: the smaller is the N sector, the lesser is this adverse supply-side effect of devaluation on the balance of payments. 7.4.3.1. Tax on non-traded goods In addition to devaluation more directly interventionist policies could be pursued to foster the output of the tradable labour-intensive sector. One form of such policies could be an ad-valorem tax,Τ, on the price of the non-traded good, PN. Note that even when the tax is levied on the supplier, it ends up being borne by the consumer. The object is to discourage consumption of the N sector's output after a resource boom and sustain domestic demand for the traded good, M. The proceeds of the tax on the consumption of non-traded goods are redistributed back to the population in a lump sum fashion. The tax does not alter the distribution of income, and national income Y 6 Alternatively, the tax revenues may be utilized to subsidize the production of M. The algebraic effects of doing this are very similar to the case when the revenue is given back to consumers.

122

S. M. Murshed

is unchanged. In the technical sensedy/dΤ= 0 in Equation (6), because the public receives the tax revenue back as an income supplement. This policy is akin to industrial policy favouring labour-intensive tradables production vis-a-vis more capital-intensive non-traded commodities. If successful, the policy will expand M sector production at the expense of N sector output in the context of a constant national income, Y. Ultimately, the aim is to avoid some of the pitfalls of the Dutch disease which shifts the production base towards non-tradable goods and away from tradable goods that are also for export. The policy instrument chosen, however, amounts to a consumption tax, which is essentially an 'expenditure-switching' policy with a view to making consumers spend more on M relative to N. In the more virtuous East Asian case, the policy works neatly and unambiguously, relying on a high price elasticity of non-traded goods consumption (see Murshed 1999 for details). In Fig. 7.1, point D illustrates the 'successful' post-tax intervention, with a decline in the N sector accompanied by an expansion in the M sector. Starting from an initial position at A where NN0 and MM0 intersect, the tax will shift the MM schedule to the right, to MM\ say, indicating expansion. The NN schedule moves downward to NN2 depicting contraction.

7.5. S U M M A R Y A N D P O L I C Y I M P L I C A T I O N S The preceding short-run macroeconomic model analyses the effects of natural resource endowment, and policies designed to cope with the associated winner's curse. The effects on the economy are dichotomized along the lines of Latin America/East Asia, suggested in Sachs (1999); or alternatively, the point/diffuse natural resource delineation (see Table 7.1). These two sets of differences are not mutually exclusive and can often be paired to form an analytical correspondence. We will then potentially possess two distinct cases and sets of results; one reflecting the Latin America (point-source linkages) stylization, and the other the East Asia (diffuse resourcerich) case. The first analytical result in Section 6.4 refers to a resource boom, which can occur for a variety of reasons. A major innovation of our analytical model is that

Table 7.1. Summary of analytical results Type of economy

Resource boom

Devaluation

Industrial policy (for M against N)

Latin America/ point source East Asia/ diffuse resource-rich East Asia/diffuse resource-poor

M falls, N rises; or, both decline Both rise; or, M rises, N falls n.a.

N contracts. M may also fall Both M and N expand Both M and N expand

May work Works unambiguously Works unambiguously

Contrasting Natural Resource Endowments

123

resource booms do not automatically cause the traded sector to contract and the nontraded sector to expand. This is partially a result of the existence of excess capacity in the economy. But relative parameter sizes do matter. A full range of possibilities emerges. Under East Asian (diffuse linkage) conditions of a high propensity to consume the domestically produced traded good, M, this sector could expand. Also, both the traded and non-traded sectors could contract. This is the most extreme form of Dutch disease, and is associated with strong wealth effects impacting on money demand. It is connected with extreme configurations of point resource/Latin American conditions. The other two results in Section 7.4 are concerned with policies to tackle resource booms. First, devaluation is the opposite of a resource boom, in a certain sense, and may be a policy initiated to avoid the adverse effects of resource booms. There is the possibility of contractionary devaluation, particularly for the non-traded sector. This likelihood is strongly associated with Latin American or point socioeconomic linkages. When devaluation is expansionary, it is so because the non-traded sector is less important to domestic consumers and there is a sharp reduction in imported consumer goods. These are more likely in the East Asian or diffuse resource-rich case. Second, a policy to tax non-traded goods consumption will be akin to an industrial policy favouring the production of traded goods. Such a policy would be initiated because of the view that labour-intensive manufactured traded goods are superior to non-traded goods production; the latter incorporating manufacturing 'dinosaurs' from the past. This policy is most likely to succeed when non-traded goods are quite price-elastic in demand, and the propensity to consume them out of income is small. Arguably, these are features of the more successful East Asian economies with their diffuse resource production structure. Country size would also be an important consideration. Without a critical mass of consumers geared to the domestic consumption of labour-intensive traded manufactures, industrial policies of this type cannot succeed. In considering further extensions to the model, one would want to extend the short-run flow equilibrium analysis above into a more dynamic stock-flow equilibrium. Incorporating a role for human capital would be important, as this factor plays an important role in the development of the more dynamic manufacturing sector. In Sachs and Warner (1999a), human capital accumulation only takes place as a result of manufacturing production, the counterpart of the M sector in our model. It takes the form of a production externality, in the context of perfect competition, and there is no cost involved. A proper formulation of human capital accumulation would take into account the public goods aspect, as well as firm-specific features of this process, as in Murshed (1997). A further analytical refinement would be to augment human capital with the innovation-imitation notions of brand creation under monopolistic competition of the type considered in Grossman and Helpman (1991). This paradigm, when contrasted to a purely human-capital approach, can offer richer insights, as it points to the importance of infrastructure, social capital and the ability to adapt to a changing environment. These factors, of course, are the product of the non-traded sector, therefore, that side of the economy may have a

124

S. M. Murshed

positive part to play after all. Other extensions would involve very different gametheoretic models of endogenous policy formation and rent seeking. The problem of underdevelopment is closely related to the notion of the staple trap. In other words, low incomes and poor long-term growth rates are ultimately the result of an undiversified production structure, and the inability to adapt to change. Policies matter, and these often need to be adapted to the size of the economy. The right policies include macroeconomic stability, openness (as it promotes competitiveness), as well as the accumulation of human and social capital. Here, issues of political economy, which determine policy, can be crucial. Also, one cannot discount the value of the appropriate timing of policy changes, which are often the product of pure chance. These issues are discussed further in Chapter 8.

REFERENCES Auty, Richard M. (1997), 'Natural resources, the state and development strategy', Journal of International Development, 9, 651-63. Findlay, Ronald and Carlos A. Rodriguez (1977), 'Intermediate inputs and macroeconomic policy under flexible exchange rates', Canadian Journal of Economics, 10, 209-17. Grossman, Gene and Elhannan Helpman (1991), Innovation and Growth in the Global Economy, Cambridge, Mass.: MIT Press. Hirschman, Albert O. (1958), The Strategy of Economic Development, New Haven: Yale University Press. Krugman, Paul (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, 27, 41-55. Krugman, Paul and Lance Taylor (1978), 'Contractionary effects of devaluation', Journal of International Economics, 8, 445-56. Lewis, W. Arthur (1954), 'Economic development with unlimited supplies of labour', Manchester School, 22, 139-91. Matsuyama, Kiminori (1992), 'Agricultural productivity, comparative advantage and economic growth', Journal of Economic Theory, 58, 317-34. Murshed, S. Mansoob (1997), 'Does trade with the South disadvantage unskilled workers in the North?, in S. Dev Gupta (ed.), The Political Economy of Globalization, London: Kluwer Academic Publishers: 129-46. Murshed, S. Mansoob (1999), 'A macroeconomic model of a developing country endowed with Working Papers 165, Helsinki: working papaers 165,helsinkiunuwiderr../WID Myint, Hla M. (1958), 'The "classical theory" of international trade and the underdeveloped countries', Economic Journal, 68/270, 317—37. Neary, J. P. and S. van Wijnbergen (1984), 'Can an oil discovery lead to a recession? a comment on Eastwood and Venables', Economic Journal, 94, 390—95. Neary, J. Peter and Sweder van Wijnbergen (eds.) (1986), Natural Resources and the Macroeconomy, Oxford: Blackwell. Prebisch, Raul (1950), The Economic Development of Latin America and its Principal Problems, New York: United Nations. Rodriguez, Francisco and Jeffrey Sachs (1999), 'Why do resource abundant economies grow more slowly? A new explanation and an application to Venezuela', Journal of Economic Growth, 4, 277-303.

Contrasting Natural Resource Endowments

125

Sachs, Jeffrey (1999), 'Resource endowments and the real exchange rate: a comparison of Latin America and East Asia, in T. Ito and A. O. Krueger (eds.), Changes, in Exchange Rales in Rapidly Developing Countries, University of Chicago Press, Chicago 111.: 133-53. Sachs, Jeffrey and Andrew Warner (1999a), 'Natural resource intensity and economic growth', in J246rg Mayer, Brian Chambers and Ayisha Farooq (eds.), Development Policies in Natural Resource Economies, Cheltenham: Edward Elgar: 13-38. Sachs, Jeffrey and Andrew Warner (1999b), 'The big push, natural resource booms and growth', Journal of Development Economics, 59, 43-76. Shell, Karl (1966), 'Towards a theory of inventive activity and capital accumulation', American Economic Review, 56, 62-8. Singer, Hans (1950), 'The distribution of gains between borrowing and investing nations', American Economic Review, 40, 473-85.

8 Political Economy of Resource-Abundant States R I C H A R D M . AUTY A N D A L A N H . GEL B

8 . 1 . POLITICA L E C O N O M Y A N D N A T U R A L RESOURCE ENDOWMEN T Natural resource endowments define and shape social groups as well as determine sectors of comparative advantage; they also can imply a greater or lesser 'natural' share of revenue in the hands of a central authority. They therefore can have a major influence on political economy, but not in a fully deterministic way. Resource rents can, for example, help to consolidate tendencies towards a strongly centralized state (as in Algeria); or alternatively, competition over rents can spark civil war (as in Nigeria) and the disintegration of the state. Rents can support an authoritarian regime; but by providing the financial basis for the organization of middle and working class groups (as in the mining camps of Venezuela), they can also underpin democracy. Political economy, in turn, will have a major influence on how agents compete for access to resource rents and on how these are allocated. The relationship between political economy and natural resource endowment is therefore a complex one. Nevertheless, some generalization is possible. The next section of the chapter extends Lai's (1995) typology of political states and uses it to examine the relationship between natural resources and political economy. The extended typology helps to provide a plausible explanation for why the governments of resource-deficient countries are more likely to align their interests with those of the majority of the populace than the governments of resource-abundant countries are. Section 8.3 explains why the political economy of the deployment of natural resource rents tends to be indirect and lacks transparency. Section 8.4 presents a set of stylized facts models of the political economy of resource-driven growth. The five resource-abundant development trajectories identified are explored in depth through case studies in Part IV of the book. 8 . 2 . T Y P O L O G Y O F P O L I T I C A L STATE S Political science has yet to produce a classification of political states that identifies in a convincing manner systematic variations in their capacity to secure the benefits of

Political Economy of Resource-Abundant States

127

'rational' economic policies. The classifications developed by political scientists tend to lack criteria that are easily operationalized and they feature lengthy and idiosyncratic description. In fact, some political scientists are sceptical about the utility of such typologies on the grounds that each country experiences a unique outcome that reflects the interaction of complex social, political, personal and chance factors. For example, Haggard (1990: 3-4) questions the feasibility of devising a viable typology: 'Theories seeking to predict policy choice from the configuration of social forces—whether conceived in class, interest group, or sectoral terms—have fundamental empirical and theoretical limitations ... I see political coalitions not as given by social structure or international constraints, but as the constructs of political leadership and organisation.' Haggard considers country-specific circumstances as the main determinants of policy outcomes. He sees attempts to generalize as disappointing at best, and fruitless at worst. Nevertheless, other political scientists such as Andrain (1988) and Leftwich (1995) are prepared to generalize, as are some economists (Findlay 1990; Lal 1995). However, Andrain's criteria are very diffuse. They contrast with the succinct summary of the 'developmental state' provided by Leftwich (1995) who bases his characterization on seven successful developers; South Korea, Taiwan, China, Indonesia, Malaysia, Thailand, and Botswana. He identifies six key features that are: • • • • • •

A determined developmental elite, in: A weak and subordinated civil society, which confers: Relative autonomy, that is deployed by: A powerful, competent, insulated economic bureaucracy, in: The effective management of non-state economic interests, while: Political legitimacy is conferred first by repression, and then, by performance.1

The developmental state can be incorporated into a typology of political states that is elaborated from Lal (1995) and based on two criteria. The two basic criteria are: the restraint upon the state's action, and the aim of the state. With regard to the first criterion, Lal distinguishes between the autonomous state and the factional state. The autonomous state can formulate and pursue its own objectives so that it can implement a coherent economic policy. The factional state must appease political groupings and thereby risks compromising the coherence of its economic policy. With regard to the second criterion, the aims of the state, Lal distinguishes between long-run welfare maximization and the service of sectional interests. Table 8.1 extends Lai's typology to provide links to the natural resource endowment and to the role assigned to markets. A resource-poor endowment tends to engender autonomous benevolent states that place a premium on investment efficiency whereas resource-abundance tends to foster factional and predatory states that relax market constraints and depress investment efficiency. B1isotswana calls this last characteristic into question, because it is a factional state. However, Botswana is a consensual democracy, a variant of the factional state (see Table 8.1) that has the potential autonomy and motive to pursue coherent economics policies that raise social welfare.

128

R. M. Auty and A. H. Gelb Table 8.1. Typology of political states, after Lal

Autonomy

Aims

Sub-type

Markets role

Country examples

Autonomous benevolent

Maximize social welfare

Developmental

Hard constraint

Paternalistic monarchy

Relaxed constraint

Indonesia, Hong Kong, Korea, Singapore, Taiwan Brunei, Kuwait, Saudi Arabia, UAE

Military elite

Soft constraint

Ghana, Nigeria

Autonomous predator Factional democracy

Factional oligarchy

Maximize rent siphoning Maximize social welfare

Maximize rent siphoning

Central planning Soft constraint

Myanmar, North Korea

Consensual

Hard constraint

Polarized

Relaxed constraint

Malaysia, Botswana, post-Pinochet Chile Costa Rica, Sri Lanka, Jamaica

Urban/industry captures policy Public officials capture policy Ethnic alliance captures policy

Soft constraint Soft constraint Soft constraint

Argentina, Bolivia, Brazil, India, Mexico Azerbaijan, Kazakhstan, Russia, Uzbekistan Kenya, Sudan, South Africa

8.2.1. Natural resources, autonomous states and welfare Consideration of the aims of the autonomous state distinguishes two basic sub-types, the benevolent state and the predatory state. The benevolent autonomous state seeks to maximize long-run social welfare. Within this sub-group, one variant equates with the developmental state of Leftwich and is strongly associated with resource-poor East Asian countries, albeit not exclusively. A second variant is the paternalistic monarchy that is associated with oil-rich states like Saudi Arabia and Brunei (Table 8.1). However, whereas the paternalistic monarchy relaxes market constraints in deploying resource rents to maintain support, the developmental variant initially maintains power by force but subsequently consolidates support by policies that rapidly reduce poverty. The link between limited natural resources, a developmental benevolent autonomous state and efficient investment is consistent with Lal and Myint (1996). There are four main reasons why the developmental benevolent autonomous state is associated with extremely resource-poor countries. First, intense population pressure on scarce land resources creates low tolerance by the majority for inequitable asset distribution and predatory rent extraction. The skewed distribution of assets becomes socially unacceptable in the face of chronic land scarcity and creates political tensions that enhance the probability that the political state will realign its interests with those of the low-income majority and engage in some asset redistribution. Such realignment occurred in Northeast Asia shortly after the Second World War, when the per capita cropland of China, Taiwan, and Korea was barely 0.1 hectares, a level that Bangladesh and Kenya did not reach until forty years later (Auty 1997a).

Political Economy of Resource-Abundant States

129

Second, resource-poor countries are less prone than resource-abundant countries are to trade policy closure because they experience minimal Dutch disease effects (Sachs and Warner 1995) and if they experiment with closed trade policies they abandon them earlier. Third, manufacturing tends to be competitive because it is transparent at a relatively low level of per capita income that the small size of the resource sector vis-a-vis the rest of the economy means that it cannot support slowmaturing infant industry or a bloated government bureaucracy (Auty 1994). Nor can the emerging competitive manufacturing sector support such activity because the fraction of rent within that sector's revenue stream is less than in the case of natural resources. Furthermore, competitive manufacturing cannot afford the higher cost of inputs from protected infant industry. For example, Taiwan and South Korea abandoned infant industry policies much earlier (i.e., at a much lower income level) than their more resource-abundant counterparts because it was soon very clear that the transfers would depress competitive export subsectors (Auty 1997b). Hong Kong and Singapore faced even more pressure to promote competitive activity. Elsewhere, China and Vietnam opted for central planning but both abandoned it when the diminishing returns of that system set in. Finally, a poor natural resource endowment reduces the scope for cumulative policy error that abundant resource rents afford and thereby places a premium upon the efficient use of scarce resources and so promotes market discipline.

8.2.2. Natural resources and departure from welfare maximization Autonomous states may also become predatory, especially where resource rents can be captured and pressure to realign the interests of the state with the majority is less urgent. Predatory states tend to pursue a narrow range of goals that reflect a concern for the welfare of a powerful elite. Public officials are not easily insulated from political interference in such states (Lal 1995; Krueger 1993), In siphoning away the rents, predatory states relax market constraints and misallocate resources.3 Two basic subcategories of political state may be recognized: one that diverts revenue to the ruler and his entourage, often military, and one that creates patronage through public employment (Table 8.1). However, the factional state is the most common type of political state among the resource-abundant developing countries. It may be either democratic or oligarchic (in which a powerful elite may manipulate a token democracy, as in Mexico under the PRI). Within the democratic state, tension arises over the trade-off between economic growth and redistribution. The consequences of assigning differing priorities to these goals are illustrated by comparing Sri Lanka and Malaysia. In each case, political 2

Personal communication from A. Warner, based on the Sachs and Warner model. However, in line with the limits on generalization about political states at the outset of the chapter, North Korea provides an example of a resource-poor predatory state that extracted political rents first as a client state of the Soviet Union and when that country collapsed, by blackmailing the international community with its bizarre behaviour. 3

130

R. M. Auty and A. H. Gelb

independence conferred control over a hitherto plantation-dominated economy upon a peasant farmer majority (Bruton 1992). Malaysia redistributed assets without impairing long-term economic growth as a result of a political compact between the two largest racial groups (see Chapter 9). In effect, Malaysia functioned as a consensual democracy, which is a second form of developmental state (of which Botswana is also an example). Elsewhere, consensual democracies emerged out of severe political and economic stress in Mauritius and Bolivia through the 1980s. Sri Lanka, however, remained a polarized democracy whose pursuit of redistribution repressed incentives not only for the plantations but also within the peasant subsector. Factional oligarchic states have been common in resource-abundant Latin America, where a landed elite tended to sustain power long after the achievement of independence from the colonial government. When the elite was eventually displaced, it was often by new factions. Engerman and Sokoloff (1997) attribute this to a natural resource endowment whose commercial crops initially favoured large estates whose inequality in income and wealth resembles Baldwin's South region, discussed in Chapter 2. The initial estates shape institutions that perpetuate inequality. The oligarchic factions manage social tensions through clientelistic relations that, however, depress long-run economic growth when they undermine the effectiveness of government (Kurer 1996). The voluntary promotion of infant industry after 1945 degenerated into a patronage system that sustained the political support of an urban oligarchy. The predominance of peasant society in sub-Saharan Africa, with its potentially favourable diffuse socioeconomic linkages (discussed in Chapters 2 and 5) might be expected to generate developmental consensual democracies. Yet factional and predatory states predominate. This is because with the removal of the colonial power, political energy was usually deflected into policy capture by individual tribes or tribal coalitions. For example, Easterly and Levine (1997) show that ethnic diversity fosters rent-seeking behaviour that engenders growth-retarding policies and inhibits consensus on investment in growth-promoting infrastructure. They also show that ethnic diversity is positively associated with low schooling, underdeveloped financial systems, distorted exchange rates and deficient infrastructure. Such rent-seeking corrodes both investment efficiency and social capital (Table 8.2). Table 8.2 Lagging socio-political indicators in sub-Saharan Africa Sub-Saharan Africa Othe r LDCs Inde x range Corruption Bureaucracy Enforceability Civil war Fractionalism Social development Inequality

4.97 1.38 1.95 1.27 67.60 1.10 31.0

Source: Collier and Gunning 1999: 67.

6.03 1.72 2.09 0.72 32.70 —0.43 31.0

0-10 high-low 0-6 poor-high 0-4 weak-strong months per year 0-100 low-high 1.86 least and—1.91 most % income of 3rd + 4th quintiles

Political Economy of Resource-Abundant States

131

Many countries in sub-Saharan Africa became trapped in the first stage of Williamson's dependency cycle (Table 2.4) which retards economic growth, so that some countries passed from resource-abundance to resource-deficiency within two generations. The impact of increasing land-scarcity in factional states is sharply illustrated by countries of the East African rift valley, the Caucasus and Central America, the most resource-poor region of Latin America. It may take social strife and bloodshed before the government interest is realigned with that of the majority poor. More optimistically, however, Mauritius shows that rising land scarcity can be peacefully managed. Tension was defused in Mauritius by the spillover of investment from plantations into labour-intensive industry that created nonfarm employment, competitively diversified the economy, and spawned a consensual democracy. Costa Rica in Chapter 14 gives a second example of peaceful transition.

8.3. TH E POLITICA L ECONOMY O F RESOURC E RENT DEPLOYMEN T Table 8.3 provides estimates of the size of the resource rents relative to GDP for six categories of natural resource endowment. The figures are for 1994 because this is the only year for which rents from cropland and pastureland are available to complement those from minerals and forests. This is due to the onerousness and consequent high cost of calculating the rents from agriculture for each country. The resource rents for the small oil-exporting countries exceed one-fifth of GDP and are more than double those of the resource-poor countries while the rents of the two other small resourcerich categories are between two-thirds and one-half higher. It should be noted that the ratios for individual countries are likely to fluctuate about the mean and to be higher during commodity booms when substantial long-term economic damage can be done in the absence of appropriate policies. It may well be significant therefore that among the handful of small resource-abundant countries that sustained high growth, Table 8.3 . Share of rents in GDP, six natural resource endowment categories 1994 (%) Resource Endowmen t Pastur Resource Poora,b Large Small Resource Rich Large Small, non-mineral Small, hard mineral Small, oil exporter All Countries a

e an d cropland Mineral

l

7.34 5.41

3.22 4.45

10.56 9.86

5.83 12.89 9.62 2.18

6.86 2.53 7.89 19.04

12.65 15.42 17.51 21.22

8.78

6.25

15.03

Resource-poor= 1970 cropland/head <0.3 hectares. Large = 1970 GDP > S7 billion. Source: Derived from World Bank (1999). b

s Tota

132

R. M. Auty and A. H. Gelb

Botswana and Malaysia enjoyed unusually stable rent streams over several decades while Chile joined them only after it adopted a rent sterilization fund in the mid-1980s (Auty and Mikesell 1998). The rents for the large resource-abundant countries are only marginally greater than those of the resource-poor countries, an outcome that reflects the greater relative significance of manufacturing in the large resource-abundant countries by the 1990s. Consistent with this interpretation, the post-shock performance of the large resourcerich countries has been more resilient than the performance of the small resourceabundant countries has (Table 2.3). The reasons for this are discussed in depth in Chapter 13. Finally, the average rents for the resource-poor countries are higher than might be expected, but they also mask variations, being especially low for the most successful resource-poor East Asian countries. This suggests that the potentially damaging effects of resource abundance attenuate sharply below some basic relative size threshold that is less than 10 per cent of GDP. Summarizing, the developing countries display a continuum in regard to the scale of resource-rents and the impacts are most clearly evident at either extreme. Rents, particularly in concentrated form, offer a natural magnet for political competition. To stay in power, governments need to find a way to redistribute rents to favoured groups. Only rarely is it politically expedient (or technically possible) to use transparent mechanisms such as direct distribution to households through vouchers. Use of this method is usually confined to high-income regions with a strongly individualistic tradition (as in Alaska and Alberta). One difficulty with such a mechanism is that it opens up the prospect of an open national debate on the appropriate distribution of income and wealth that few states may be willing to contemplate. Indirect mechanisms for distributing rents can invoke other goals (such as the 'need' to build an industrial sector or the 'urgency' of creating jobs) to validate a distribution strategy that is appealing on political grounds. Direct distribution also requires that the government is willing to renounce the use of rent as an instrument for boosting its own power relative to that of its citizens—and few governments are eager to do this.4 One characteristic of resource-abundant countries is therefore the use of indirect redistribution mechanisms to allocate natural rent. Another—especially relevant to mineral exporters in which a large share of rent goes directly to the government—is a chronic tendency for the state to become overextended, especially in the face of large fluctuations in the value of natural resources. This section focuses on the implications of the need to use indirect redistribution mechanisms and on the political economy of chronic over-extension. First, it considers one main avenue of redistribution—its use to underpin the development of import-competing sectors through extended periods of protection. It then considers a second, but related, avenue—the implications of using rents to create employment through growth of the public sector. Taking into account models of migration and the endogenized response of government, it is then shown how such a deployment of rent for political purposes can so depress the economy-wide efficiency of investment as to cause economic growth to virtually cease. 4

Some transition economies provide the exception.

Political Economy of Resource-Abundant States

133

Yet both of these mechanisms have the potential to create concentrated and vocal pressure groups and may therefore be more appealing to a government than direct distribution. Finally, the section considers the political economy of overspending and how this reduces the benefits of an abundant natural resource endowment.

8.3.1. Rent re allocation through protection The industrialization that many developing countries were forced to undertake during the 1930s and 1940s was increasingly pursued on a voluntary basis thereafter. The policy of protecting infant industry was strongly supported by structuralist economists such as Prebisch (1950) and Singer (1950). Yet if the initial justification of this industrial policy was to promote economic development, the industrial policies of most countries, especially the resource-abundant countries, degenerated into mechanisms to transfer resources from exporters to more favoured groups. The exporters were usually in the primary sector and the transfers were to politically influential urban groups that comprised capitalists and workers together with the politicians who extracted part of the rents in return for sustaining the policy. This process of policy capture (Auty 1995) resulted in the 'infant' industries either failing to mature or taking several decades to do so (Auty 1994). Yet Krueger and Tuncer (1982) demonstrate that maturation rates must not exceed five to eight years if the discounted benefits of the mature industries are to compensate for the discounted costs of infant industry support. Bergsman (1970) estimates that such transfers reached 9 per cent of GDP annually in Brazil after reforms in the mid-1960s had reduced protection. Such transfers reflect real rates of effective protection that are often orders of magnitude higher than the domestic value added. Moreover, the protection tends to be ad hoc and counter-intuitive (see Chapter 11 on Bolivia). Revenue extraction from resource-intensive sectors has been especially high in subSaharan Africa. Krueger (1993) finds that direct agricultural taxation in sub-Saharan Africa averaged 25 per cent of revenues, some four times that of Latin America and ten times Asian rates. The figure for sub-Saharan Africa doubles when indirect taxes, such as overvalued exchange rates are added. Rates of taxation were at levels above those required to maximize the tax take in the long run. Part of the reason has been imperfect information on the outcome of myriad distortions within the economy. Another reason may lie in the short time horizons of governments and their recognition that farmers (and mining companies even more so) have high fixed costs. Officials can then set tax rates to cover the variable costs of production and, in effect, capture the returns to the investments of the farmers or mining companies. Consistent with this, McMillan (1997) finds that crops with a relatively high ratio of sunk costs to total costs such as cocoa and coffee have been much more heavily taxed than those with relatively low sunk costs, such as groundnuts, cotton, and tobacco that afford immediate scope to curb production. Tree crop farmers (or mining companies) respond by withdrawing from planting (or from investment) the next time round, so that the government revenue base shrinks, creating even greater pressure to squeeze revenues from elsewhere in the system.

134

R. M. Auty and A. H. Gelb

8.3.2. Rent reallocation through job creation Public employment can be a politically appealing way to redistribute rents. Gelb et al. (1991) develop a model of a resource-rich country whose government creates unproductive jobs in public administration and in state-owned enterprises in order to alleviate urban unemployment. Labour flows between the rural and urban sectors are determined by a Harris-Todaro migration model. A single urban wage is assumed to hold in the three urban subsectors (which comprise a private sector, a productive public sector, and a nonproductive public sector). The latter could include the growth of a permanent bureaucracy or the use of rents to create unproductive employment through special employment programmes (as in Trinidad and Tobago, see Gelb and Associates (1988)). An exogenous rise in the urban wage (e.g., due to spending the rents from a windfall or resource bonus) creates a wage gap that raises the premium on rural out-migration. Unemployment then increases in the modern urban sector. The government responds to additional urban unemployment by increasing taxation (whose burden falls disproportionately on the private sector) in order to invest capital in the creation of extra urban jobs. But, in addition to depressing the return on investment, this process is self-defeating because it renders work in the unproductive public sector preferable to farming. More people migrate to the city, where their unemployed presence intimidates the government from which the unemployed rural migrants extract still more rent. A CGE model provides an estimate of the potential scale of the resulting rent misallocation under widely differing savings functions. Simulations using empirically plausible data suggest that the consumption losses grow over time and that within a decade the efficiency of capital can be depressed below the level required to sustain economic growth. Gelb et al. (1991) conclude that the role of direct government misallocation has been neglected in explaining the differences in economic growth between countries with similar endowments and saving rates, and that too much attention has been given to the role of trade regimes. In fact, however, trade policy and resource misallocation are two sides of the same coin. A closed trade regime increases the scope for the public sector to misallocate resources; one way of doing this is to create unproductive jobs.

8.3.3. The political economy of overspending In any economic model that recognizes capacity constraints and less-than immediate clearing of markets by prices (including the labour market), the effects of chronic excess demand and deficient demand are not symmetric. Excess demand causes inflation and high import leakages; deficient demand causes real activity to contract, unemployment to rise and real incomes to fall. There is therefore a premium on maintaining spending levels relatively stable, or growing at moderate rates. This prescription is especially important for government spending, because booms tend to lock in powerful hysterisis effects that prolong high spending levels and set the stage for catastrophic macroeconomic imbalances marked by high inflation, abrupt demand

Political Economy of Resource-Abundant States

135

cuts and sharp falls in output and growth. Over a cycle, the economy progressively moves towards stagflation. Gelb and Associates (1988) analyse the effects for oil exporters using a dynamic general-equilibrium model with imperfect price adjustment. Simulations of the effects of 'optimal' spending profiles based on imperfect prediction confirm the very high costs associated with overoptimistic projections. Locked in to unsustainable spending levels (and also to high debt caused by borrowing against projected oil income) countries face a catastrophic decline in incomes and output when reality diverges from projections. In contrast, slowing spending when resource prices are high simply preserves the option of increasing spending later. Especially for mineral-rich economies, the implication is that macroeconomic programmes and government spending should be planned on the basis of conservative projections of natural resource prices and incomes—but few are. Political competition for rents, combined with non-transparent mechanisms of redistributing them (and in some cases of accounting for them), makes it more difficult for governments to moderate spending levels in response to fluctuations. With lucrative procurement contracts at stake, many projects undertaken by resource-rich governments are valued for their immediate income-generating effects rather than their growth-enhancing potential. Moreover, with high current and potential incomes, governments face limited restraint in spending. Nevertheless, such practices distort the economy and once the distortions have been created, reform is protracted.

8.4. STYLIZE D FACT S MODELS O F TH E POLITICA L ECONOMY O F RESOURCE-DRIVE N GROWT H This section draws upon earlier chapters to distil medium-term stylized facts political economy models to explain the divergence in the growth trajectories of the resourcepoor and resource-abundant countries. The models are not deterministic, in line with the evidence from the historical overview and short-run model in Chapters 6 and 7. Nevertheless, the last three decades the resource-poor countries have tended to diversify their economies competitively whereas the resource-abundant countries tended to fall into a staple trap of dependence on a weakening primary sector. The competitive industrialization model describes the first development trajectory whereas the staple trap model captures the second trajectory.

8.4.1. Competitive industrialization model Chapter 2 establishes that resource-poor countries tend to industrialize earlier than the resource-abundant countries so that they urbanize earlier and thereby move faster into the growth-enhancing phase of the demographic cycle. This singular pattern of structural change triggers a virtuous cumulative economic circle that is also environmentally more sustainable, as indicated by the genuine savings measure in Chapter 3. This circle also encourages greater demand for skills, especially among the poor, and also distributes income more equitably for the reasons given in Chapters 2 and 4. It also builds social capital that, as shown in Chapter 5, strengthens the capacity

136

R. M. Auty and A. H. Gelb

Figure 8.1 . Competitive industrialization model

to cope with shocks and is part of a virtuous cumulative social circle that complements the economic circle (Fig. 8.1). However, the developmental benevolent state that is so strongly associated with the successful resource-poor countries is fostered by extreme pressure on natural resources. Among the smaller resource-poor countries there are a few, usually with rents above the group average, that lack the policy coherence to sustain these virtuous circles. The competitive industrialization model is applied to South Korea in Table 8.4 where a coup in the early-1960s installed a developmental benevolent autonomous state that reformed trade policy in 1963, when the per capita income was still relatively low. This was the catalyst for the virtuous economic circle. It ushered in a rapid expansion of labour-intensive manufactured exports (Stage 2 in Table 8.4) that sharply increased domestic investment and triggered the expansion of some heavy industry as domestic demand passed minimum size thresholds. This second stage also rapidly absorbs surplus rural labour (within a decade of the initial reform in both Korea and Taiwan), pushing the labour market to its turning point and creating pressure for wage increases that must be accommodated by upgrading workforce skills and moving into heavy and chemical industry, HCI (Stage 3 in Table 8.4). In Stage 3, the share of investment rises towards 25 per cent of GDP, helped by the boost to domestic savings from a steeply falling dependency ratio which, in the case of South Korea dropped from 0.83 to 0.61 between 1970 and 1980 and continued to plummet (see Table 2.4). The efficiency of investment remains relatively high and HCI that is competitive comes to dominate the manufacturing sector, perhaps more than is prudent in Table 8.4, if the aggressive policy of Korea is compared with the

Political Economy of Resource-Abundant States

137

Table 8.4. Four stage model of competitive sequenced industrialization (based upon South Korea, 1953-1997) Index

Stage 1 1953-62a

Stage 2 1963-73

Stage 3 1974-82

Stage 4 1983-97

Growth GDP growth Investment/GDP ICOR

4.1 11.6 2.8

10.4 19.9 1.9

7.1 29.1 4.1

8.9 33.1 3.7

Structure Agriculture Manufacturing Light Heavy

1960 45.5 11.9 9.5 2.4

1970 28.7 23.0 14.0 9.0

1980 15.4 33.2 14.5 18.7

1990 9.0 28.9 9.7 19.2

Note: aDas 1992: 24-5. Source: Auty (1997b).

more cautious approach of Taiwan (Auty 1997b). Although assistance may be given to infant industry, it is targeted at sectors of emerging comparative advantage and tapered to force rapid maturation. The diversified economy proves resilient to external shocks, like the 1979 oil shock and this reduces the risk of a growth collapse. However, the growing complexity of the economy penalizes excessive intervention (like the over-ambitious Korean HCI Big Push) and encourages further liberalization (Stage 4 of Table 8.4). Ranis and Mahmood (1992) argue that this 'linear liberalization' of the economy effectively depoliticizes the growth trajectory, but perhaps the weakness of the chaebol-dominated banking system during the 1997 financial crisis renders this too optimistic an interpretation. These features of the virtuous economic circle of the competitive industrialization model are linked to the virtuous social circle, which has three main characteristics. First, income equality is maintained because the rapid absorption of surplus rural labour removes the drag of underemployed workers on the wages of the poor. In addition, the upgrading of workforce skills in Stage 3 rapidly reduces the premium on skilled labour (Londono 1996). Second, relatively equitable income distribution deepens social capital, lowering transaction costs and strengthening the capacity of the political economy to adjust to shocks. Third, these trends in human and social capital accumulation combine to push the political state towards a democracy that is consensual, rather than polarized so that the risk of government capture is reduced. The genuine savings estimates in Table 8.5 confirm that competitive industrialization strengthened the sustainability of the Korean development trajectory compared to examples from any other natural resource endowment category. However, the impact of the natural resource endowment is not a deterministic one and a handful of resource-abundant countries including Botswana, Malaysia, and Chile have evolved into consensual democracies, a form of developmental state. Where this occurs, the longer dependence upon primary products is not associated with

138

R. M. Auty and A. H. Gelb Table 8.5. Extended genuine saving rates, selected countries 1970—1994 (% GNP)

Country Argentina Bangladesh Bolivia Costa Rica Ghana India Malaysia Mexico Saudi Arabia South Africa South Korea

1970-74 16.85 -1.80 6.10 11.99 6.46 7.96 17.31 10.43 -32.84 12.03 11.13

1975-79 18.07 -2.37 -13.70 13.20 1.63 8.78 17.61 7.33 -23.51 8.82 20.60

1980-84 4.35 -1.86 -43.77 9.10 -10.94 5.91 9.08 -4.22 -29.34 8.14 20.69

1985-89 6.39 -0.16 -27.52 15.30 -1.61 7.75 10.43 -1.83 -21.74 3.13 30.33

1990-94a 5.02 0.64 -13.97 14.20 -1.94 8.31 15.21 2.25 -30.57 5.53 30.53

Note: aCosta Rica 1990-92, Saudi Arabia 1990-93 only. Source: World Bank (1999).

policy failure but it does retard diversification into the labour-intensive competitive industrialization stage of the model (compare columns 1 and 2 in Table 8.6). This has four important consequences. First, diversification must initially occur into other primary products and this may prove difficult for the smallest resource-rich economies whose natural resource endowment may be skewed towards the production of one or two commodities. Second, slower industrialization retards urbanization and delays the favourable middle stage of the worker/dependency cycle so that the accumulation of capital (including human capital) is slower. Third, the labour market turning point is postponed, so that surplus rural labour persists longer and amplifies income inequality. Fourth, the industrialization that does occur tends to be capital-intensive so that the benefits of the labour-intensive manufacturing stage of the competitive industrialization model are lost and governments may feel obliged to intervene to create employment (Lal and Myint 1996). However,these outcomes can all be ameliorated to some extent by the promotion of competitive manufacturing through an export processing zone (EPZ).

8.4.2. The staple trap model In the more typical resource-abundant case, however, the preconditions include a predatory or factional government so that the distribution of income and assets may already be skewed (Fig. 8.2). There may also be inertia from protected industry where, as in much of Latin America, enforced import substitution during the period 1930-45 was substantial. Yet even where infant industry protection is adopted later, as in subSaharan Africa, it can become entrenched within a decade, as Ghana demonstrates (Chapter 10). Once again, the second stage of the competitive industrialization model is leapfrogged and the advantages of speedy accumulation of produced and human capital, maintenance of income equality and economic diversification are lost (compare

Table 8,6. Stylized facts models of the political economy of the natural resource endowment and economic growth Model

Competitive industrialization

Natural Resource Endowment Poor

Staple trap

Abundant diffuse

Abundant diffuse

Abundant point ore

Abundant point oil

Autonomous benevolent 4.0-9.0

Factional consensual democ. 7.0-16.0

Factional oligarchy/ Predatory 5.0-33.0

Factional oligarchy/ Predatory 9.0-32.0

Autonomous paternalistic

Low

Low

Low/High

Low/High

Low/High Partial Open Neutral/Public sector

Pre-conditions

Political state Income rich 1/5: poor 1/5a Obsolete capital stock

12.0-20.0

Policy

Reform response Trade policy Incentive bias

Early Early Open Open Comparative advantage Comparative advantage

Postponed Closed Infant industry/ Public sector

Postponed Closed Infant industry/ Public sector

Capital build-up Genuine saving Produced Human Social

Strongly positive Rapid rise > 24% GDP Virtuous circle Transaction improving

Moderately positive Slower rise > 20% GDP Slower virtuous circle Transaction improving

Low/Negative Slow rise > 20% GDP Lagged + skewed Transaction repressing

Low/Negative Negative Slower rise + < 20% GDP Rapid rise > 25% GDP Lagged + skewed Rapid + ineffective Transaction repressing Feudalistic

4.8 Negligible Competitive widening Robust Sustained: 2.5-4.0

5.5 Small Slow competitive widening Robust Sustained: 1.0-2.5

6.0 Strong Retarded/Regressive High vulnerability Erratic: 0.5-1.5

7.5 Strong Regressing High vulnerability Boom + bust: (0.5)-2.0

12.4 Strong Mono-product Financial cushion Boom + decline: (0.4)-4.0

South Korea Singapore Bangladesh

Malaysia Thailand Myanmar

Ghana Argentina Costa Rica

Bolivia Zambia Botswana

Saudi Arabia Nigeria Oman

Economic outcome

ICOR 1985-97 Dutch disease effect Diversification Resilience to shocks PCGDP growth (%) Examples

Classic Classic Anomalous

Note: aData from Tables 2.3 and 4.2.

140

R. M. Auty and A. H. Gelb

Figure 8.2. Staple trap model

columns 2 and 3 in Table 8.6). However, in contrast to the resource-rich consensual democracy that seeks to maximize long-term welfare, the political economy now seeks to increase the patronage of politicians and this leads to trade policy closure in all but the countries with very large rents in relation to GDP, mainly the oil-exporters. The more usual outcome is trade policy closure, which may initially occur in response to the Dutch disease effects arising from a resource boom, as discussed in Chapter 7, but is more likely to arise from the conscious espousal of forced industrialization. Industrialization by infant industry protection has three principal flaws. First, it yields discretionary rents that are deployed with minimal transparency so that they easily degenerate into a corrupt rent-dispensing mechanism that distorts the economy in an ad hoc and cumulative fashion. Second, such industry is capital-intensive and creates few jobs so that governments may feel obliged to provide nonproductive employment to ease social tension. This intensifies the relaxation of market discipline and expands still further the protected sector. Third, it is slow-maturing and imposes increasingly onerous demands on the primary sector for transfers and foreign exchange that undermine that sector's competitiveness. The economy-wide efficiency of investment declines steeply and the level of investment flattens off or falls (Table 2.3). Within this political economy, the rents are transferred from the potentially efficient primary sector (whose share of GDP is shrinking as structural change proceeds), into a burgeoning non-tradable sector that includes slow-maturing infant industry and nonproductive public sector employment. As the protected clientele expands, the

Political Economy of Resource-Abundant States

141

demand for transfers outstrips the rents and depresses incentives in the primary sector below the levels required to sustain production, let alone to expand it. The distortion of the economy is greater under extreme resource abundance because the transfers can be sustained far longer. The economy is locked into a staple trap as economic diversification regresses so that growth is increasingly dependent on staples with declining competitiveness. As the demands of the non-tradable sector for transfers and foreign exchange outstrip the capacity of the primary sector to supply them, fiscal and external deficits recur and economic growth becomes erratic and slows. The staple trap is also associated with the erosion of social capital. Income inequality deteriorates as the failure of the labour market to absorb surplus rural labour depresses the wages of the poorest while a unionized labour aristocracy boosts the wages of the favoured, but relatively few workers in the protected capital-intensive manufacturing sector. Political contests over the rent stream also skew the distribution of income and assets in favour of a politically powerful minority. Decelerating economic growth exacerbates social tensions and further saps the capacity of society to cope with crises even as the severity of economic difficulty increases. The distortion of the economy and the deterioration in social capital are potentially greater with 'point' resources, such as minerals, than with resources with more diffuse socioeconomic linkages such as crops (compare columns 4 and 5 in Table 8.6). In the case of mining, revenues are typically concentrated on one or two large firms, a small labour aristocracy of mineworkers and the government. Taxation therefore dominates the domestic economic linkages so that the role of the government is increased so that both the probability of policy failure and the magnitude of its effects are enhanced. However, even the potentially favourable 'diffuse' socioeconomic linkages of peasant farm economies (see Baldwin 1956) may be repressed in factional political states (Table 8.6, column 3). In order to restore growth following a collapse, a period of declining real wages is required, but that elicits strong political opposition so that reform is often counterproductively postponed. Efforts at piecemeal reform may only aggravate the situation due to the fact that where there are many distortions, the removal of a few may be welfare depressing. Rodrik (1998) argues that the political costs of economic reform are high in relation to the rewards because reform may entail redistribution among ethnic groups or along class lines. Yet the more protracted the growth collapse and the longer the postponement of reform, the more the deterioration in economic infrastructure, human capital, and social capital. Not surprisingly, the revival of growth after a collapse may take several decades. Rodrik (1998) suggests that the political economy may demand a two-track approach to reform in which the privileges of the beneficiaries are sustained while the desired new efficiency-oriented constituency is established. 8.5. CONCLUSIO N The natural resource endowment differentiates structural change and it affects the likelihood of the political state being a developmental one in ways that have proved advantageous to the most resource-poor countries. The economy diversifies early in the

142

R. M. Auty and A. H. Gelb

case of resource-poor countries into competitive manufacturing that accelerates the accumulation of produced, human and social capital whose efficient deployment sustains rapid and egalitarian economic growth. However, this sequence requires a political state that has sufficient autonomy to sustain a coherent economic policy and the objective of raising social welfare. Although these features are strongly associated with extreme resource deficiency, the relationship between the natural resource endowment and the political state is not wholly deterministic. Nevertheless, the government of a resource-poor country is more likely to align its interests with the majority and to redistribute assets while still maintaining efficiency incentives. Such a state resists pressures to close the economy, which adheres to its emerging comparative advantage. Although resource-abundant countries may pursue competitive industrialization where they engender a developmental political state, usually a consensual democracy, their progress will initially be slower due to their longer reliance on primary exports and their later and more capital-intensive industrialization, which challenges governance. More often, however, the natural resource rents feed conflicts so that the resource-abundant country has a factional or predatory government that relaxes market discipline in capturing and redistributing the rents. The economy is thereby deflected from its comparative advantage and cumulates economic distortions that retard diversification and/or cause the economy to regress into a staple trap of dependence on a weakening primary sector. The governments of predatory and factional oligarchic states prefer non-transparent methods for deploying the rents in order to maximize the scope for political manoeuvring. The favoured channels for deploying rents are trade protection, job creation and over-extended public expenditure. Such policies cut economic growth by reducing investment efficiency. It is ironic that the growth collapses of the late-1970s and early1980s resulted from the backfiring of the resource-abundant countries' efforts to reduce their commodity dependence. Domestic policies to promote infant industry (which resource abundance could sustain for longer than resource-poor countries) weakened their economies through the 1960s. Thereafter, international efforts to reverse the longrun decline in real commodity prices by establishing producer agreements triggered trade shocks in the 1970s that damaged their weakened economies. The next section of this book, Part IV, draws upon case studies to explore in depth the five resource-abundant development trajectories, starting with the competitive industrialization of Malaysia in Chapter 9 (summarized in column 2 of Table 8.6). Chapters 10-13 examine growth collapses under different natural resource endowments. REFERENCES Andrain, C. F. (1988), Political Change in the Third World, Boston MA: Unwin Hyman. Auty, R. M. (1994), Economic Development and Industrial Policy: Korea, Brazil, Mexico, India and China, London: Mansell. Auty, R. M. (1995), 'Industrial policy capture in Taiwan and South Korea', Development Policy Review, 195-217.

Political Economy of Resource-Abundant States

143

Auty, R. M. (1997a), 'Natural resources, the state and development strategy', Journal of International Development, 9, 651—63. Auty, R. M. (1997b), 'Competitive industrial policy and macro performance: has South Korea outperformed Taiwan?', Journal of Development Studies, 33, 445-63. Auty, R. M. and Mikesell, R. F. (1998), Sustainable Development in Mineral Economies, Oxford: Clarendon Press. Baldwin, R. E. (1956), 'Patterns of settled regions', Manchester School of Social and Economic Studies, 24, 161-79. Bergsman, J. (1970), Brazil: Industrialization and Trade Policies, Oxford: Oxford University Press. Bruton, H. J. (1992), The Political Economy of Poverty, Equity and Growth: Sri Lanka and Malaysia, Oxford: Oxford University Press. Collier, P. and J. W. Gunning (1999), 'Explaining African economic performance', Journal of Economic Literature, 37, 64—111. Das, D. K. (1992), Korean Economic Dynamism, Basingstoke: Macmillan. Easterly, W. and R. Levine (1997), 'Africa's growth tragedy: policies and ethnic divisions', Quarterly Journal of Economics, 112, 1203-50. Engerman, S. L. and K. L. Sokoloff (1997), 'Factor endowments, institutions, and differential paths of growth among new world economies', in S. Haggard (ed.), How Latin America Fell Behind, Stanford CA: Stanford University Press: 260-304. Findlay, R. (1990), 'The new political economy: its explanatory power for LDCs', Economics and Politics, 2, 193-221. Gelb, A. H, and Associates (1988), Oil Windfalls: Blessing or Curse? Oxford University Press: New York. Gelb, A. H., J. Knight and R. Sabot (1991), 'Public sector employment, rent seeking and economic growth', The Economic Journal, 101, 1186-99. Haggard, S. (1990), Pathways from the Periphery: The Politics of Growth in the Newly Industrialising Countries, Ithaca NY: Cornell University Press. Krueger, (1993), Political Economy of Policy Reform in Developing Countries, Cambridge MA: MIT Press. Krueger, A. O. and B. Tuncer (1982), 'An empirical test of the infant industry argument', American Economic Review, 72, 1142-52. Kurer, O. (1996), 'The political foundations of economic development policies', Journal of Development Studies, 32, 645-68. Lal, D. (1995), 'Why growth rates differ. The political economy of social capability in 21 developing countries', in B. H. Koo and D. H. Perkins (eds.), Social Capability and Long-Run Economic Growth, Basingstoke: Macmillan: 288-309. Lal, D. and H. Myint (1996), The Political Economy of Poverty, Equity and Growth, Oxford: Clarendon Press. Leftwich, A. (1995), 'Bringing politics back in: towards a model of the developmental state', Journal of Development Studies, 31: 400-27. Londono, J. L. (1996), Poverty, Inequality and Human Capital Development in Latin America, 1950-2025, Washington DC: World Bank. Mcmillan, M. (1997), 'A dynamic theory of primary export taxation: evidence from sub-Saharan Africa', Working Paper, New York: Columbia University. Prebisch, R. (1950), The Economic Development of Latin America and its Principal Problems, New York: United Nations.

144

R. M. Auty and A. H. Gelb

Ranis, G. and S. Mahmood (1992), The Political Economy of Development Policy Change, Oxford: Blackwell. Rodrik, D. (1998), 'Trade policy and economic performance in sub-Saharan Africa', NBER Working Paper 6562, Cambridge MA: National Bureau of Economic Research. Sachs, J. D. and A. Warner (1995), 'Economic reform and the process of global integration', Brookings Papers on Economic Activity, 1, 1—118. Singer, H. (1950), 'The Distribution of Gains between Borrowing and Investing Nations', American Economic Review, 40, 473—85. World Bank (1999), World Development Indicators, World Bank, Washington DC.

PART I V

DEVELOPMENT TRAJECTORIES OF RESOURCE-ABUNDANT COUNTRIES

This page intentionally left blank

9 Competitive Industrialization with Natural Resource Abundance: Malaysia MAHANI ZAINA L ABIDI N

9 . 1 . INITIA L CONDITIONS Malaysia is classified in this study as a small resource-abundant country (see Table 1.1). As such, it is one of only a handful of resource-abundant countries (the others comprise Botswana, Chile, Oman, and Thailand) that escaped the low growth staple trap that such countries have commonly experienced in recent decades. Rather than falling into a staple trap, Malaysia systematically diversified its economy from slow-growth commodities to high-growth commodities and then into competitive manufacturing (Table 9.1). The natural resource endowment of Malaysia is favourably more diversified than the endowments of many small resource-abundant countries. It includes cropland, forests, and minerals whose relative importance has changed over time. The tin industry started in the early nineteenth century when the Dutch commercialized production. Rubber was first cultivated in Malaysia at the end of the nineteenth century and by 1930 it occupied two-thirds of the cultivated area. By the time Malaysia gained independence in 1957 tin and rubber were the two main contributors to GDP, and the combined share of natural resources in GDP was 45 per cent. However, a diversification began into palm oil in the late 1960s in order to reduce dependence on rubber when prices were falling. About the same time, forest resources in the form of sawlogs and sawn timber began to be leading primary commodities. Subsequently, the discovery of crude petroleum in the mid-1970s and then liquefied natural gas (LNG) offset a decline in tin production (that was caused by depletion of deposits, rising costs, and plummeting prices). In this way, the structure of primary exports steadilydiversified and helped cushion the impact of individual commodity price shocks (Table 9.2). It thereby sustained an unusually stable and large stream of resource rents whose favourable consequences are discussed in later sections. If the nature of the natural resource endowment was relatively favourable, however, the presence of ethnic tensions was not, at least on first sight. In 1980 the Bumiputeras1 1 The Bumiputeras are the 'sons of the soil', the early inhabitants of Malaysia. They are the Malays, the natives of East Malaysia (such as the Ibans and Dayaks) and the aboriginal.

148

M. Z. Abidm

Table 9.1. Structure of production, absorption, and exports, Malaysia 1960-1995 (% GDP)

1960

1970

1980

1990

1995

Agriculture Manufacturing

33 8

29 12

22 21

19 26

13 32

Private consumption Public consumption Investment

61 11 15

58 16 22

51 17 30

50 14 34

48 13 44

Total absorption

87

96

97

98

105

Gross domestic saving Gross genuine saving

28 n.a.

27 19

33 12

36 18

39 26

Exports Imports

51 39

42 38

58 55

76 74

95 99

Source: World Bank (1999).

constituted about 55 per cent of the population compared with 33 per cent Chinese and 10 per cent Indians. Economic function and wealth varied with these ethnic groups. The Bumiputeras had a lower income level, lived in the rural areas and worked in agriculture and the public sector. Conversely, the Chinese had control of the economy, lived in the urban areas and were in commerce and trade. Finally, the Indians were mainly on the rubber plantations and were part of the lower income group. These characteristics were, however, turned to advantage because they gave an impetus to successive governments to select economic policies that maximized social welfare by improving income distribution and raising employment. As a result, Malaysia became, and continues to be, a consensual democracy governed through a coalition of ethnic-based political parties. The Bumiputra majority has the stronger political power and dominates the ruling coalition party, but the interests of the traditionally more entrepreneurial Chinese are also safeguarded. This political compromise led to the growth-based-sccurity development strategy that has underpinned the efficient deployment of resources within Malaysia. This chapter identifies and explains the principal differences between the development of Malaysia and that of most other small resource-abundant countries that did not achieve competitive industrialization. Section 9.2 examines the relationship between savings and investment to explain why Malaysian investment sustained high growth. Section 9.3 discusses in greater detail how the public sector made a remarkably high contribution to total investment without damaging overall investment

Since independence, Malaysia has been ruled by the National Front Coalition, which consists of threeee core parties (United Malay National Organization, the Malaysian Chinese Association and the Malaysian Indian Congress) and other smaller political parties. The component parties will negotiate on the allocation of parliamentary seats, representation in the cabinet, and other political positions. The coalition is dominated by the Malay (Bumiputera) party.

Table 9.2. Structure of Malaysian exports, 1960-1995 Share (%) of total primary commodity exports Crude petroleum share (%)

1960 1965

1970 1975 1980 1985 1990 1995

5 3

Palm oil share (%)

Rubber share (%)

Sawlogs share (%)

2 4

69 51

4 9

Sawn timber share (%)

Tin share (%)

LNG share (%)

3 3

17 30

n. n.

n. n. n. 10 9 10

5

7

43

16

5

24

13 32

20 13

31 23

11 13

7 7

18 12

37 37 22

17 15 33

12 11 13

12 14 8

5 11 12

7 3 2

Share primary commodity of total exports (%)

Share of manufactured exports (%)

80 76

a. a.

79

10

71 72

18 22

62 37 17

32 56 79

Source: Data for reports of rubber, tin, sawlogs, palm oil, and crude petroleum are from Bank Negara Malaysia Quarterly Economic Bulletin, various issues. Data for LNG and manufactured exports are from Government of Malaysia, Treasury Economic Report, various issues.

150

M. Z. Abidm

efficiency. Section 9.4 explains why Malaysia adhered to an open trade regime while Section 9.5 analyses the related issue of the industrial policy that fostered competitive industrialization. The accumulation of human capital is studied in Section 9.6. Finally, Section 9.7 summarizes the argument by reviewing Malaysia's capacity to manage external and internal shocks without experiencing a growth collapse.

9.2. INVESTMEN T AN D SAVING S If a country is to maintain its growth, it must ensure that its net investment (which includes produced capital, human capital and natural resources) is non-negative. The most basic indicator of the accumulation of inputs for development is the investment rate for produced assets. The Malaysian investment ratio was only 15 per cent of GDP in 1960 (Table 9.1) but it increased rapidly over the next two decades to exceed 25 per cent of GDP in the 1970s, the level that characterizes the high-growth developing countries. One important reason for this achievement, which occurred despite the lag in passing through the demographic cycle (see Table 2.4), is a saving rate that was relatively high for a developing country at an early stage of development. Table 9.3 shows that in 1960, when Malaysia was a low-income country with a per capita GDP of US$281, the gross domestic saving rate was already at 27 per cent of GDP and that it continued to increase and reached 39 per cent by the mid-1990s. Thrifty households were largely responsible for the high savings rate (Kharas 1991), but in addition to voluntary savings, workers were required to contribute to a compulsory savings scheme called the Employee Provident Fund (employers made roughly matching contributions). In addition, fiscal surpluses from exports and also savings from private sector commodity exporters contributed. The savings proved more than sufficient to meet Malaysia's investment needs until the late 1980s (Table 9.3). When investment subsequently exceeded domestic savings in the 1990s the shortfall was met by a large capital inflow.3 A second reason for the high savings rate is a stream of natural resource rent that was large and relatively stable, despite changes in the composition of exports and fluctuations in the prices of individual commodities. Malaysia shares this characteristic of a relatively high and stable rent stream with Botswana, in contrast to most mineral and crop exporters. Vincent and Ali (1997) show that both gross investment and total resource rents were rising during the period 1970 to 1990 (Fig. 9.1). During that period, and subsequently, resource rent has been an important source of investment. On average the rent has been about one-third of gross investment and this ratio has been fairly constant throughout this period. Vincent and Ali (1997) calculate net 3 During the 1990s, the high rate of investment was financed by both short- and long-term capital inflow. For example, on average the annual saving-investment gap for the 1993-96 period was RM25 billion. During this period, the annual inflow of foreign capital was more than RM15 billion annually, except for 1994 where the inflow was smaller.

Competitive Industrialization: Malaysia

151

Table 9.3. Domestic investment and saving, Malaysia 1960-1998 (% GDP)

Gross domestic investment Gross domestic saving

1960

1965

1976

1982

1987

1991

1995

1998

14 27

20 24

22 32

34 25

23 37

36 30

41 37

34 39

Source: Lee (1996).

Figure 9.1. Resource rents relative to gross fixed capital formation, Malaysia Source: Vincent and Ali (1997). investment by deducting the resource depletion allowance from gross investment.4 Figure 9.2 shows that conventional net investment was positive in every year from 1970 to 1990, indicating an increment in per capita stocks of produced capital. In addition, the per capita adjusted net investment—conventional net investment minus 4 The resource consumption allowances were estimated for two natural resource sectors, minerals and timber. Vincent and Ali (1997) used the Stauffer (1986) procedure to estimate the total resource rents and then converting them into Hotelling rents. The estimation procedure was also an improvement of previous studies because it circumvented the use of average costs when calculating net prices to overcome the problem of overestimation of the value of natural resource depletion. The 1970 to 1990 data was utilized but it was expressed in real terms of 1978 price-level. Resource consumption allowances for minerals rose significantly during the 1980s caused by both expanding production and the rising share of Hotelling rent. Timber allowance rose until 1979-80 but declined thereafter caused by decreasing prices.

152

M. Z. Abidin

Figure 9.2 . Per capita investment, Malaysia Source: Vincent and Ali (1997).

the resource depletion allowance—is also non-negative in every year but one (Fig. 9.2). Therefore, the total stock of capital per capita increased in Malaysia during the period 1970 to 1990, despite the depletion of the country's mineral and timber resources. This outcome is consistent with the increasing trend in genuine savings for Malaysia shown in Table 8.5. In the 1970s, the genuine savings rate was, for most years, more than 15 per cent of GDP. It then declined through the oil boom in the first half of the 1980s, before rising again after 1987. The rise in the genuine saving rate after the mid-1980s reflects the contraction of the oil rents and also strong efforts to upgrade labour skills and productivity. Table 8.5 shows that Malaysia's genuine saving rate compares favourably with that of other resource-abundant countries. While Malaysia's rate is not as high as that of South Korea, it is much higher than those of Costa Rica, Ghana, and Mexico. Finally, Vincent and Ali (1997) show that the average annual growth rate of the per capita adjusted EDPj (2.9 per cent) for Malaysia was lower than per capita GDP growth (3.7 per cent). The difference is neither enormous nor unexpected: the net growth after deducting resource depletion will naturally be lower. Therefore, Malaysia

Competitive Industrialization: Malaysia

153

has invested not only income from the resource rents but it has also added more investment from other sources so that its growth has been sustainable. 9 . 3 . PUBLI C INVESTMEN T AN D COMPETITIV E DIVERSIFICATION The investment streams described in the previous section were made by both the public and private sectors. This section focuses on public investment, and upon development expenditure in particular, and the way that the Federal Government managed and distributed it. This is because public investment was central in providing not only basic capital accumulation but also in bringing about the distribution of wealth with economic growth. Table 9.4 shows the social capital ratios introduced in Chapter 5 for Malaysia and the other case study countries. It suggests that the quality of the Malaysian bureaucracy has been fairly high, reflecting the steady accumulation of social capital more generally. During the 1950s and 1960s, a central aim of the Malaysian development strategy was to reduce the dependence of the economy on rubber and tin. It was achieved through the diversification and modernization of the agriculture sector and the promotion of industry (Ismail 1990). Programmes were developed to launch new commercial crops (as in the case of palm oil) and to cope with declining cultivated areas (rubber). A two-pronged approach was adopted. First, new land schemes were introduced to increase and modernize the production of rubber and palm oil. These new schemes also alleviated the plight of landless and unemployed Malays. The second initiative was to sustain production through replanting schemes, especially for the rubber industry, which became less profitable due to declining prices. Rubber smallholdings tend to be less productive than cash-crop estates whose large size permits efficient management and maximization of resources. Since smallness was identified as one of the causes of low productivity, the solution was to establish miniestates among smallholders. A similar programme was designed for forest resources. Government expenditures increased the stock of social overhead capital (SOC), primarily in the form of economic infrastructure, as a basis for the expansion of economic activity. This objective is reflected in the development expenditure allocation whose share in total public sector finance tripled from 8.3 per cent in 1970 to peak over the oil boom years 1981-84 (Table 9.5). From 1978 to 1987, economic infrastructure accounted for at least half of total investment, with energy, communications, and transport predominating. Investment in social infrastructure (education, housing, and health) made up 10-17 per cent of total public investment from 1979. However, that rate of public sector investment outstripped domestic absorptive capacity and depressed investment efficiency. Using the incremental capital output ratio (ICOR) as a measurement of efficiency, the ICOR deteriorated from 3.7 in 1971—80 to 7.2 in 1981-85. It then recovered to 4.8 in 1986-90 as the government scaled down its role in the economy. The deterioration in the ICOR is partly due to the capital-intensive heavy and chemical industry (HCI) Big Push that was undertaken by state-owned enterprises in the first half of the 1980s (Ismail 1990).

Table 9.4 . Social capital indices, by resource endowment and selected country

Category Socia

l Incom development gin

e Lan i gin

d Incom i shar

e Politica l Civi e right s libert

l Civi y qualit

l servic e Rul y la

w

e of

to 65%

Range (low-high)

-1.9-1.6

0.3-0.6

0.3-0.9

0.2-0.5

7-1

7-1

0-6

0-6

-0.07 0.23 0.17 -0.14 -0.25

0.46 0.38 0.48 0.46 0.47

0.65 0.50 0.60 0.66 0.73

0.43 0.49 0.41 0.43 0.42

3.06 3.63 2.94 3.09 2.83

3.19 3.19 3.31 3.17 3.14

2.26 3.75 2.06 2.24 1.73

2.19 3.88 1.94 2.08 1.87

0.18 -0.35 0.78 -0.01 n.a. 0.75 n.a. 0.62

0.42 0.46 0.47 0.35 0.51 0.52 n.a. 0.55

0.86 0.77 0.80 n.a. 0.67 0.61 0.83 0.70

0.41 n.a. 0.40 0.50 0.41 0.41 n.a. 0.37

6.0 6.0 1.0 4.0 2.5 4.0 6.0 4.5

4.0 4.5 1.0 4.5 3.5 3.5 6.0 5.5

3.0 1.0 2.0 1.0 5.0 4.0 1.0 6.0

2.0 1.0 4.0 1.0 4.0 4.0 1.0 1.0

n.a. 0.85

0.34 0.35

0.42 0.34

0.50 0.51

2.5 5.0

3.5 6.0

1.0 3.0

1.0 3.0

Countries in Table 5.1

All countries Resource-poor Diffuse linkages Point linkages Cocoa-coffee group Case study countries

Argentina Bolivia Costa Rica Ghana Malaysia Mexico Saudi Arabia South Africa Resource-poor comparators

Bangladesh South Korea Source: See Tables 5.1-5.4.

Table 9.5. Consolidated public sector finance, Malaysia 1970-1990 (% GNP)

Revenue Operating expenditure Current surplus/deficit Non-financial public enterprise surplus Total public sector current surplus/deficit Development expenditure General government Non-financial public enterprises Total expenditure Overall surplus/deficit GDP growth rate Source: Ismail (1990).

1970

1979

1980

1981

1982

1983

1984

1985

1986

1987

1988

1989

1990

24.6 20.9 3.7 0.8

30.7 26.8 3.9 1.2

32.7 30.1 2.6 0.6

32.9 31.1 1.9 5.1

33.1 30.5 2.6 5.7

33.2 30.6 2.6 6.6

32.5 29.5 2.9 6.7

36.5 30.8 5.6 7.8

37.1 34.8 2.4 4.2

32.61 31.15 1.45 4.78

32.99 29.65 3.34 4.21

31.49 28.05 3.45 3.91

30.40 27.72 2.68 3.86

4.5

5.1

3.2

6.9

8.3

9.2

9.7

13.5

6.6

6.23

7.55

7.36

6.54

8.3 7.6 0.7

13.1 11.0 2.1

19.4 16.3 3.1

27.3 22.6 4.7

27.2 20.5 6.7

26.2 16.8 9.4

22.9 12.4 10.5

18.3 9.6 8.6

16.6 10.9 5.7

11.29 7.43 3.86

11.91 7.14 4.35

13.36 8.67 4.69

14.84 10.05 4.79

29.2 -3.7 6.2

39.9 -8.1 6.2

49.5 -16.2 7.4

58.4 -20.4 6.9

57.7 -18.9 5.9

56.8 -17.0 6.3

52.4 -13.2 7.8

52.4 4.7 -1.0

55.4 -10.0 1.2

42.44 -5.05 5.3

41.14 -3.94 8.7

41.40

-6.00 7.6

42.56 -8.29 6.5

156

M. Z. Abidin

A geographically balanced distribution of produced capital among the various regions of Malaysia was achieved because of the structure of federal-state finances. Such investment in produced capital helps to balance income distribution and development patterns between those states that are rich in natural resources and those that are deficient. The Federal Government adopted a centralized revenue and expenditure system, collecting directly most state revenues including resource rents, and then assigning state expenditure allocations according to individual development needs. Thus, states that generated large resource rents might only receive a portion of them because the Federal Government reallocated much of the rents. For example, in states with large timber resources, the rent from timber constituted less than a quarter of the state revenue (Vincent and Ali 1997). Nevertheless, this resource outflow from the states is compensated by the development expenditure, shared operating expenses, national defence, and other responsibilities financed by the Federal Government. In addition, loans arc provided at concessional rates (Umikalsum 1991: 186). The ratio of development expenditure to resource-based revenue was greater than unity in all years so that the Federal Government had invested more than the value of the resource rent collected. This is a key step in optimizing the utilization of resource rents. 9.4. TRAD E POLIC Y AN D E X P O R T - O R I E N T E D INDUSTRIALIZATION As noted in earlier chapters, during the post-war years most resource-abundant developing countries sought to force the pace of industrialization by protecting infant industry rather than by competitive industrialization. Although Malaysia did start on the path of infant industry protection in the 1960s and also used the oil rents to launch an over-ambitious state-led HCI Big Push 1980-85, it maintained a relatively open economy. Rather, successive governments encouraged export-oriented industrialization, taking advantage of low cost unskilled labour, and they also bolstered trade liberalization. According to Lim (1973: 116), prior to independence, West Malaysia5 followed 'a classical export economy in the colonial period in which she produced and exported rubber and tin in exchange for the import of essential foodstuffs and manufactured goods'. The British favoured an open trade policy: export taxes were kept low to help British rubber and tin companies to be profitable while import duties were minimal to make West Malaysia a market for British manufactured goods. Consequently, import tariffs were minimal under colonialism and they were mainly for revenue purposes, as there were no industries to protect. Import taxes were imposed only on tobacco, liquor, petrol, sugar, and textiles and apparel where the first three items accounted for 69 per cent of total tax collection (Lim 1967: 269). Exports contributed 53 per cent of GDP in 1958. 5 In 1957, Malaya, which constituted the Malay Peninsular, obtained its independence from the British. In 1963, the nation of Malaysia was born through the joining together of Malaya with the British Borneo states of Sabah and Sarawak. Malaya was then known as West or Peninsular Malaysia.

Competitive Industrialization: Malaysia

157

Table 9.6. Unweighted nominal and effective protection rate, Malaysia 1963-1987

Year NR

P (%) ER

P (%)

1963 1965 1970 1973

10 12 15 19

21 24 44 40

1978 1979 1982 1987

20 22 17 13

38 43 46 41

Note: The NRP and ERP estimates were based on products. Source: Mahani (1998).

After independence, tariffs were regarded as a tool to promote import substitution industrialization. However, levels of tariff protection were not excessive by developing country standards: as indicated in Table 9.6, the average Nominal Rate of Protection (NRP) and the Effective Rate of Protection (ERP) were moderate. In 1963 the NRP was only 10 per cent, and it peaked at 22 per cent in 1979 before falling again.6 The ERP, a more accurate measurement of protection was 21 per cent in 1963 and at its peak in 1982 it was 46 per cent. The 1987 estimates show a decline for both the NRP and ERP that reflects further liberalization in the wake of the HCI Drive. The tendency towards stronger protection during the 1960s eased when Malaysia accelerated its export-oriented industrialization strategy in the early-1970s. Foreign direct investment (FDI) was critical to this strategy and a key precondition for such investment is a low or free tariff regime for export industries. The establishment of free trade zones allows exporting companies to import their intermediate goods duty free (in other words, they buy inputs at world prices) and they are also exempt from export and excise duties. The open trade policy for export industries had a demonstration effect and the government created a more open economic environment for the rest of the economy, increasing competitiveness and encouraging efficient resource allocation. Despite the export promotion strategy, industrial policy regressed when Malaysia pursued an HCI Big Push in the first half of the 1980s. The policy was modelled on the ill-fated South Korean HCI Big Push and drew upon the surge in oil rents generated by the 1979 oil shock. Public sector companies spearheaded the HCI Big Push and they helped to lift the size of the public sector to 48 per cent of GDP by 1981, twice the level of fifteen years earlier. The sectors that were promoted included vehicles, steel, and cement. The structure of protection was extended from consumer goods to In6d comparison, most of the developing countries had imposed a high protection rate in the 1960s and 1970s. For example, the NRP for consumer goods in Pakistan was 266 per cent in 1970/71 while the ERP was 585 per cent. Similarly, in Brazil, the ERP for consumer goods in 1963 was 360 per cent. See Krueger et al. (1981).

158

M. Z. Abidin

include heavy industry and also consumer durables (Mahani 1998); the ERP for consumer durables reached 165 per cent in 1982, compared with — 11 per cent in 1963. However, the policy was dualistic, in as much as non-targeted sectors were left to operate in a competitive environment.7 The dualistic industrial policy contributed to the dichotomous nature of the Malaysian manufacturing sector in which export industries have a large FDI presence while domestic firms concentrate on producing for the local market. Fortunately, the level of protection was not so excessive as to distort resource allocation and damage export potential. In this regard, resource-based industries such as wood products and palm oil have always operated in an environment with low protection, thus making them strong exporters. The renewed tilt towards inward-looking policies did not last very long, however. By 1985 the Malaysian economy was in recession due to large fiscal deficits and the collapse of primary commodity prices. In order to revive the economy, the Government stimulated exports, particularly of manufactures, to offset the adverse impact of fluctuating primary commodity prices (Table 9.2). Elsewhere, Indonesia successfully adopted a similar strategy (Auty and Mikescll 1998). Priority was given to attracting new FDI into manufacturing, and tariffs were further liberalized to support this move. The recession of the mid-1980s also spurred the Malaysian Government to launch a privatization policy and the private sector was given the lead role in economic growth. The weighted average import tariff rate decreased from 14.7 per cent for the 1984-87 period to 11.2 per cent in 1991-93 (Mahani 1998). Trade policy was further liberalized in response to regional and multilateral trade agreements so that the post-Uruguay Round trade weighted average nominal tariff rate for Malaysia fell to only 9 per cent. Overall, the presence of a large and buoyant commodity export sector facilitated the management of social tensions and the pursuit of the growth-based-security development strategy. It encouraged Malaysia to retain an open trade policy and to support an export-oriented industrialization strategy over infant industry protection. Consequently, unlike most resource-abundant countries, Malaysia did not dissipate the rents generated by the dynamic resource sector in expanding a slow-maturing manufacturing sector at the expense of incentives in the primary sector. Instead, the rents were channelled into strengthening output in the primary sector and into the accumulation of produced and human capital. Therefore, the decision to promote export-oriented manufacturing from the late-1960s helped Malaysia to avoid the staple trap and achieve competitive industrialization (see Table 8.6) 9 . 5 . C O M P E T I T I V E I N D U S T R I A L I Z A T I O N WIT H RESOURCE-ABUNDANCE Malaysia espoused competitive industrialization despite its sizable natural resource revenue in order to accelerate the economic growth rate, lessen the negative 7 There are wide variations of NRP and ERP among industries; for example the lowest NRP is consistently the rubber industry while the highest is tobacco products. The range between the lowest and the highest NRP was 93 per cent in 1963 and it was 210 per cent by 1978 (Mahani 1998).

Competitive Industrialization: Malaysia

159

impact of export price instability, and restructure the racial composition of economic participation under the objectives of the New Economic Policy.8 At the early stage of development, Malaysian economic growth, which was mainly generated by the agricultural sector, was unable to provide sufficient employment opportunities for the rapidly expanding population. According to Lim (1973: 73) 'another consequence of a persistently high rate of increase in the population is a high rate oj increase in the labour force

the need to create employment opportunities is therefore

clear' .9 The solution was to introduce labour-intensive industrialization that could absorb excess labour from rural areas and generate high growth. The rural population was the basis of the political power of the government so that it was logical for the government to want to improve the welfare of this constituency. The most effective means was to create employment in order to raise rural income. This shift in development strategy coincided with the decision by multinational electronics firms to relocate their manufacturing from developed countries to cheaper locations. Com petitive manufactured exports provide rates of growth in GDP and also employment that the more capital-intensive strategy of import substitution industry cannot match. Thus, an important reason why Malaysia embarked on a non-resource-based indus trialization path was to create employment without burdening the primary sector. Export instability also pushed Malaysia into industrialization. In the period 1954-66, West Malaysia registered a higher export instability than other developed and developing countries (Lim 1973). West Malaysia's instability index of 16.5 per cent far exceeded that of developed countries (6.2 per cent) and other developing countries (13.4 per cent).10 Ariff (1972) attributes high export instability largely to the dominance of rubber in the export commodity mix. The instability in rubber revenues was greater for smallholders than for estate workers because the workers were cushioned through partial absorption by the estates. Therefore, potential social discontent could be reduced by allowing smallholder commercial production of palm oil and other cash crops like cocoa and pepper, as well as opening up opportunities in manufacturing. A third reason to industrialize was to attain the social restructuring objective. In the late 1960s, wealth was concentrated in the hands of MNCs that controlled primary commodity production and also the Chinese who dominated the commercial sector. The Bumiputeras were active mainly in agriculture and were economically weak. Th8toe New Economic Policy (NEP) is a social and economic restructuring programme designed to create better socioeconomic conditions and to improve communal relations. It had two main objectives: the eradication of poverty regardless of race and the restructuring of Malaysian society to eliminate the identification of race with economic function. The Plan period was from 1970 to 1990. The NEP marked a new departure for a more active and direct public sector role in resource allocation, production, and distribution. For e9respectiveluyrestpectivelyxample, the annual increase in population in 1967 and 1969 rcen pe trespectively 58an 2. wa 3. respectively ds and during the 1967-68 period, the rate of unemployment for the youngest age-group (31.9 per cent) was very much higher than that experienced by any other age group. 10 The instability is caused by the demand and supply conditions of rubber and tin. Low price elasticity of supply combined with uncontrolled variability in demand, supply, or both could explain the sharp instability in both prices and proceeds of primary commodities (Lim 1973).

160

M. Z. Abidm

Under the New Economic Policy (NEP) the goal of social restructuring required that the prevailing composition of economic influence should be changed to avoid the identification of race with a particular economic activity. A target was set to increase the Malays' corporate wealth from 2 per cent of the total in 1970 to 30 per cent by 1990. Industrialization was expected to accelerate growth, the critical requirement for successful restructuring of the economy. Rapid economic growth would foster a Malay entrepreneurial class and Malay commercial wealth without constraining other communities, thus preserving social harmony and stability. Industrialization was also regarded as a channel to increase the pool of professional and skilled Malays. In 1970 when most Malays were in the rural areas, the mean rural household income of RM200 per month was only half of that of urban households (Ishak and Ragayah 1997). Nationally, the Malay mean household income was also about half that of the Chinese. Moreover, although the production of primary commodities expanded, it did not significantly improve the Malays' income level because they were mostly smallholders. Since the resource rents were chiefly in the hands of foreigners and nonMalays, industrialization was one means of increasing the income of Malay households. It would draw Malays to urban areas where they could receive a higher income and thereby improve the national income distribution.11 From a broader perspective, higher income and reduced income disparity would sustain national savings, a critical need for maintaining economic growth. Yet, compared with the resource-poor competitive industrialization model (column 2 in Table 8.6), reliance on primary exports was longer (Table 9.2) and income inequality remained relatively high in Malaysia (Table 9.4). 9.6. A C C U M U L A T I O N OF H U M A N SKILLS Although the Malaysian genuine saving rate traces a rising trend since the 1960s, the fact that this rate is very much lower than the traditional savings rate is of concern. It implies that either the investment in skills is insufficient or the resource depletion is too large. One reason for this outcome is that Malaysia invested heavily in human resource development, whose contribution to economic growth is lagged. The rapid expansion of the manufacturing sector required a large pool of skilled workers and thus Malaysia has devoted a substantial portion of its expenditure to education. In addition, the expansion in education was driven by the goals of the NEP to reduce poverty and to increase the Malays' participation in the corporate sector. Education attacks poverty and expands the professionally qualified Malays in areas previously dominated by non-Malays. The total expenditure on education by 1991 was about 5.5 per cent of GNP, a higher rate than that of some developed countries such as Japan (4.7 per cent) and the United States (5.3 per cent) (Lee 1996). In all the Malaysian development plans, T11hrough labour-intensive, export-oriented industries, the growth of manufacturing employment was impressive. Between 1970 and 1982, the manufacturing employment growth rate was double that of the whole economy and its share of total employment rose from 11.4 to 15.5 per cent in that period (Ishak and Ragayah 1997).

161

Competitive Industrialization: Malaysia Table 9.7. Planned and actual development expenditure by sector (% share) Sector

IMP

2MP

3MP

4MP

5MP

6MP

7MP

Economic Social Education Administration and defence Total expenditure

60.1 17.7

72.3 13.7

64.0 17.1

60.5 21.5 10.1 17.9 46,320

64.8 24.8 16.1 10.5 35,300

50.6 24.8 13.4 24.6 54,705

50.0 29.3 15.1 20.7 67,500

7.8

6.9

7.3

22.1 4,242

14.0 9,821

18.8 24,937

Source: Table 9.5, Lee (1996).

Table 9.8. Number enrolled in education, Malaysia 1960-1995 (% age group)

Primary Secondary Higher education

1960

1965

1975

1981

1986

1990

1995

96 19 1

90 28 2

93 41 3

92 53 5

101 54 6

93 56 7

91 — —

Source: Lee (1996).

education took up about half of the social development allocation (Table 9.7). The increased importance of education after 1986 is shown by its rising share of development expenditure in successive development plans. As the first step towards increasing human capital, the government increased the accessibility of education by providing free primary education. The literacy rate rose from 53 per cent in 1960 to 73 per cent in 1985 (Ismail and Meyanathan 1993: 24). Similarly, secondary education was expanded in line with the achievements at the primary level. Thereafter tertiary, technical, and vocational education were the priority. As indicated by Table 9.8, a notable improvement occurred in the secondary level where the enrolment rate tripled from 1960 to reach 56 per cent in 1990. Enrolment in higher education jumped seven-fold during the 1960-90 period. The composition of skills has also changed with more emphasis given to producing scientific and technical manpower. In comparing Asian educational attainment, Booth (1995) concludes that although Malaysia has lagged the achievements of the resource-poor East Asian economies, it has made more progress than other Asian countries, both in terms of enrolment ratios and the share of educational expenditure in GDP.

9.7. ENHANCE D CAPACIT Y T O MANAG E SHOCK S External and internal economic shocks provide a test of whether a country is managing its economy well. Malaysia has dealt with shocks effectively and has simultaneously alleviated ethnic inequality by means of income redistribution. The external shocks came mainly from deterioration in terms of trade, and supply fluctuations were a secondary cause. Racially-motivated riots in 1969 provided the major internal shock.

162

M. Z. Abidn

Export instability during the 1953-67 period was attributed to the performance of the rubber and tin industries. The response to this was to diversify into the production of new primary products and manufactured goods. In addition, there were price support measures such as the formation of a buffer stock for rubber. Together, these responses stabilized the income flows of tin and rubber producers, particularly rubber smallholders. The 1979 oil shock boosted Malaysia's oil revenue. The challenge was to sterilize this income surge in order to avoid problems associated with the Dutch disease. However, the HCI Big Push of the early-1980s proved over-ambitious and it required prompt macroeconomic adjustment when the negative external shock of 1985 occurred. The 1985 shock reflected an abrupt fall in the prices of Malaysia's export commodities. The terms of trade declined by 4.5 per cent in 1985 and 15.6 per cent in 1986 (Ismail 1990). By 1986, as revenue shrank, a deficit in the operating account emerged, the first such deficit since 1972. The government drastically reduced fiscal expenditure and encouraged the private sector to lead economic growth through the privatization policy. It further encouraged the private sector by liberalizing FDI and trade policy and through selective deregulation of the financial sector. However, the most serious threat to Malay stability came from racial riots in 1969. The NEP was the radical solution to this problem. Its two objectives were to restructure the society and to reduce poverty. Implicit in this policy is the belief that restructuring could only take place when there was growth—in other words, the redistributed wealth has to be newly created wealth. This policy was acceptable because existing wealth was not taken from one ethnic group to be given to another. Paradoxically, therefore, the sociopolitical challenge of ethnic diversity has helped Malaysia avoid the staple trap for two important reasons. First, the social contract agreed at independence demands that the economic status of the Bumiputeras be improved while allowing other races to expand economically. For that reason, exportoriented industrialization was introduced relatively early to create employment and raise the income of the rural sector. Consequently, trade policy was closed far less than in other resource-abundant developing countries. The presence of foreign capital and the need to ensure the success of export-oriented industries constrained any divergence from market discipline. Second, the rural sector is the political base of the government so that it could not squeeze commodity rents in the way that many other resource-abundant countries did. In fact, the agriculture sector was expanded and its productivity improved to keep pace with industrialization. Simultaneously, investments were made that enhanced welfare and narrowed income disparity between the rural and urban areas and this met the political needs. Provision of education, skill improvement and rural infrastructure development contributed to the general economic development. In this way, Malaysia made income distribution a cornerstone of the development process. It should also be recalled, however, that relative to many small resource-abundant developing countries, Malaysia benefited from a stream of natural resource rents that was not only sizable but also stable.

Competitive Industrialization: Malaysia

163

The coalition of parties that reflects the consensual democracy of Malaysia and has ruled since independence has served as a stabilizing force, allowing each race a role in government. The policy of growth-based-security avoided the development pitfalls of many resource-rich countries and shows the importance of sociopolitical management in the development process.

REFERENCES Ariff, K. A. M. (1972), 'Export trade and the Malayan economy: an inquiry into the economic implications of export instability', Monograph Series on the Malaysian Economic Affairs, Kuala Lumpur: Faculty of Economics and Administration, University of Malaya. Auty, R. M. and R. F. Mikesell (1998), Sustainable Development in Mineral Economies, Oxford: Clarendon Press. Bank Negara Malaysia Quarterly Economic Bulletin, various issues. Booth, A. (1995), 'Southeast Asian economic growth: can the momentum be maintained?', Southeast Asian Affairs 1995, Singapore: Institute of Southeast Asian Studies. Government of Malaysia, Treasury Economic Report, various issues. Ishak, S. and M. Z. Ragayah (1997), 'Economic growth and equity in Malaysia: performance and prospects' in A. Mahdzan (ed.), Southeast Asia on the Growth Path: Book of Readings, Universiti Pertanian Malaysia Press. Ismail, M. S. and S. D. Meyanathan (1993), Malaysia: Growth, Equity and Structural Transformation, Washington, DC: World Bank. Ismail, M. S. (1990), 'Fiscal policy and structural adjustment', in H. Yokoyama (ed.), Malaysian Economy; Policy and Structural Change, Tokyo: Institute of Developing Economies. Kharas, H. (1991), 'Malaysian savings in the 1990s: problems and prospects', in K. H. Lee, and S. Nagaraj (eds.), The Malaysian Economy Beyond 1990: International and Domestic Perspectives, Kuala Lumpur: Persatuan Ekonomi Malaysia. Krueger, A. O., H. B. Lary, T. Monson and N. Akrasanee (eds.) (1981), Trade and Employment in Developing Countries, Vol. 1, Chicago: the University of Chicago Press for the National Bureau of Economic Research. Lee Kiong Hock (1996), 'Labour markets issues: skills training and labour productivity', paper presented in The Challenge of Human Capital Formation, 5-7 August 1996, Kuala Lumpur. Lim, D. (1973), Economic Growth and Development in West Malaysia 1947-1970, Kuala Lumpur: Oxford University Press. Lim, C. Y. (1967), Economic Development of Modern Malaya, Kuala Lumpur: Oxford University Press. Mahani, Z. A. (1998), 'Malaysian exports: seeking new sources of growth', report submitted for the Asian Development Bank for the Study on Asian Exports: Past Trends and Future Prospects. Stauffer, T. R. (1986), 'Accounting for "Wasting Assets": measurements of income and dependency in oil-rentier states', Journal of Energy and Development, 11/1, 69-93. Umikalsum, H. M. N. (1991), 'Fiscal federalism in Malaysia, 1971-1987', thesis submitted for the Degree of Doctor of Philosophy, Faculty of Economics and Administration, University of Malaya.

164

M. Z. Abidm

Vincent. J. R. and R. M. Ali (1997), Environment and Development in a Resource-Rich Economy: Malaysia Under the New Economic Policy, Harvard Institute for International Development and Institute of Strategic and International Studies, Malaysia, Harvard University Press. World Bank (1999), World Development Indicators 1999, Washington, DC: World Bank. Young, K., W. C. F. Bussink and P. Hasan (1980), Malaysia: Growth and Equity in a Multiracial Society, Baltimore MD: Johns Hopkins University Press.

10 A Growth Collapse with Diffuse Resources: Ghana R O B E R T OSEI

1 0 . 1 . INITIA L CONDITIONS Ghana is endowed with a broad range of natural resources, including arable land, forests, sizable deposits of gold and diamonds and a considerable potential for hydroelectric power. This study classifies Ghana as a small resource-abundant country. Its per capita cropland in 1970 was around 0.34 hectares, a figure below the average for resource-abundant countries (Table 10.1), but when it is combined with minerals, forests, and hydro, the status of Ghana as a resource-abundant country on the eve of the oil shocks is confirmed. In addition, Ghana's absolute GDP of US$2.2 billion at that time was above the average of the small resource-rich countries, an advantage for industrialization. Ghana also derived potential benefits for economic development from the fact that its dominant export crop cocoa, generates diffuse socioeconomic linkages. Other initial conditions were also propitious. Like Malaysia, Ghana achieved independence from Britain in 1957 with institutions that included the rule of law, democracy, an effective civil service and a level of education that was high relative to those of its regional peer group. Ghana was also one of the wealthiest countries in subSaharan Africa with a per capita income close to that of Malaysia and South Korea. Finally, the economy was open and relatively undistorted, although economic growth during the 1950s, at around 4 per cent per annum, barely kept pace with population growth. Killick (1999: 54) writing on the immediate post-independence period notes; 'In terms of average incomes, absence of a balance of payments constraint, a sound budgetary situation and a well-functioning public administration, Ghana compared well with many other of the nations starting on the road of economic modernization and development'. Ghana was the first black African state to receive independence and it was seen by many as a pilot case for Africa's readiness for independence. Ghana may well have possessed more favourable preconditions for competitive industrialization than Malaysia, but its development trajectory conformed to the staple trap model of resource-abundant countries during the first twenty-five years I am indebted to Oliver Morrissey, Richard Auty, Tony Addison, Tony Killick and all the project participants for very helpful comments. Views expressed in this paper are solely mine.

166

R. Osei Table 10.1 . Resource endowment of Ghana compared to other countries

Resource Croplan endowment (ha/hd

d GD ) ($b

P PCGD ) 197 0

P PCGD ($b) 197 0 1970-9

P Growt h 3

Resource poor

Large

0.15

21.05

196

3.7

Large Small non-mineral Small hard mineral

0.56 0.57 0.66

22.99 1.41 1.23

574 250 304

1.3 0.7 -0.2

Ghana

0.34

2.2

257

0.8

Resource rich

Source: Auty (1998) and World Bank (1997).

of post-independence development.1 The staple trap model predicts that in the absence of either a developmental autonomous benevolent state or a consensual democracy (see Table 8.1), a resource-abundant country will mismanage the challenges posed by its longer dependence on primary product exports. It will drift towards an inward looking trade policy and an over-reliance on primary products whose competitiveness is waning so that the accumulation of all forms of capital, including social capital, will decelerate and even regress. An external shock to such an economy is likely to cause a growth collapse. Certainly, the economic performance of Ghana over the post-independence years has been disappointing to say the least. Growth collapsed through the 1960s and annual per capita GDP growth between 1970—93 averaged 0.8 per cent (Table 10.1). Moreover, this figure reflects a significant improvement in performance after gradual reforms commenced in 1983: the annual GDP growth rate has averaged 5 per cent since the mid-1980s. Recent attempts to incorporate political and social factors into the explanation of sub-Saharan African development emphasize low levels of both social capital and trade openness (Collier and Gunning 1999). A critical difference between Ghana and Malaysia is the political economy, which profoundly affected the balance of advantage between the predominantly rural commodity exporters and more vocal urban groups that favoured protection. The consequences are summarized by comparing columns 2 and 3 in Table 8.6. This chapter assesses the extent to which Ghana conforms to the staple trap model and why it does so. It is organized as follows. Section 10.2 analyses the political economy and its impact on the economic policies pursued by successive governments of Ghana since independence. Section 10.3 then examines the impact of the external environment on the domestic economy. Section 10.4 establishes the growth trajectory and links it to saving and investment and the evolving structure of production. The Th1ye discussion is restricted to the period from independence to the start of the Economic Recovery Programme (ERP) although occasional mention is made of the reforms for comparative purposes. 2 With regard to trade openness, Edwards (1997) for instance argues that increased openness can lead to resource reallocation and consequently to growth.

Diffuse Resources: Ghana

167

sustainability of economic development in Ghana is evaluated in Section 10.5. The conclusions and some policy lessons are set out in Section 10.6. 10.2. POLITICA L ECONOM Y AN D ECONOMI C POLICIE S Frimpong-Ansah's (1991) analysis on the nature of Ghana's efforts at development rightly concludes that the political economy played a major role in shaping the country's development trajectory.3 He fails, however, to explore the fact that the natural resource endowment helped to promote the 'vampire state' (or autonomous predatory state in Table 8.1) after independence, although this is implicit in his analysis. Even though the political ideology of Nkrumah, the first Prime Minister of independent Ghana, was also a factor, natural resource rents provided the opportunity for him to pursue a dogmatic 'pan-African' ideology. This section explains how these factors contributed to a growth collapse that began in the mid-1960s and lasted into the mid-1980s. Although Ghana achieved independence in 1957, it became self-governing in 1951. Its economic policies became distinctively different from those of the colonialists after 1960 as Nkrumah aimed to reduce economic dependence by accelerating industrialization and expanding the role of the state. On top of this, his pan-African objectives were to be vigorously pursued, even when they conflicted with economic policies. Inspired by the dominant ideas of development economics at the time, Ghana espoused a 'big-push' approach to accelerate industrialization. The strategy emphasized a less open economy and import substitution. It also entailed radical changes in macroeconomic policy, including foreign exchange restrictions and import quotas. Under this system, imports which had all been on open general licence were given specific licences. A Seven-Year Development Plan was drafted that was to operate from 1963-70. The plan aimed to modernize agriculture as well as to accelerate industrialization. This was to be achieved by shifting the emphasis of investment from the creation of social overhead capital into directly productive activities in which the government was to be a major player. The bulk of the plan was to be financed from domestic resources and mainly from an implicit taxation of cocoa exports (Kusi 1991). The government generated rents for itself by using the cocoa marketing board to create a wedge between the producer prices and prices obtained on the international market. This of course did not mean an end to taxing cocoa exports either directly through export duties or indirectly through an overvaluation of the exchange rate. Krueger (1993: 61-73) finds that out of a sample of eighteen developing countries, Ghana had the highest level of discrimination against agriculture through the 1960s and 1970s. The lost export earnings during 1963-85 were estimated to be 80 per cent of what Ghana's actual exports would have been under a neutral trade policy. In his defence, Nkrumah probably shared the scepticism of many economists at the time about the efficacy of the 3 He is not alone in making that assertion. Other authors like Killick (1978) and Rimmer (1992) reach similar conclusions. 4 Frimpong-Ansah (1991) argues that the pursuit of this approach to development was also because it satisfied Nkrumah's ideological inclinations and commitments.

168

R. Osei

market mechanism in inducing economic development. He therefore sought to break from the legacy of colonial dependency and reliance on primary product exports, and the vicious cycle of poverty that he associated with it. Nkrumah identified development with industrialization and viewed agriculture as an inferior activity.5 His idea of modernizing agriculture was the setting up of mechanized state farms. Nkrumah was overthrown in February 1966 by some army and police officers calling themselves the National Liberation Council (NLC) who in 1969 handed over power to an elected civilian government, the Progress Party (PP). The NLC had no ideology or desire to rule indefinitely. Its main ambition, as well as that of the elected PP government that succeeded it, can be seen as economic recovery and political revival. The NLC sought to reverse the 'vampire' political state that Nkrumah had created. Uneconomic and irrelevant state commercial and industrial activities were quickly dismantled. The NLC and PP governments both built credibility among the aid agencies so that medium-term debts were re-scheduled along with the grace periods in repayment of long-term loans. Pressure from external donors resulted in the introduction of an IMF supported reform. This involved two devaluations of the currency (30 per cent in 1967 and 44 per cent in 1971), liberalization of the external sector, and monetary reforms. The 1971 devaluation led directly to the overthrow of the PP government, an event that made subsequent governments reluctant to consider such a move even after the real exchange rate had become grossly overvalued in the late-1970s. Political instability also retarded reform: in all, five changes in government occurred between 1971 to 1982. First, the National Redemption Council (NRC) under the leadership of Acheampong (1972-78) took power on the pretext that the two budgets of the previous government had reduced the real incomes of the military and public servants. The NRC opposed the excessive rate of devaluation with its beneficial redistributive effects on the export sector. As Frimpong-Ansah (1991) argues, the 1971 coup can be interpreted as a rejection of the policy of the previous government that was beginning to recognize the inherent weaknesses of the economy and the legacy of the Nkrumah government. Consequently, within less than a decade, the Big Push strategy had entrenched urban-based political interests that depended upon transfers from the primary sector and therefore opposed trade policy and other economic reform. Adverse economic pressures led to the replacement of NRC by the SMC II. This government was also to be short-lived as there was a revolt when it introduced tight monetary policy and currency demonitization. Rawlings was then installed in office by 5 In one of his writings on neo-colonialism, Nkrumah states ' . . . industry rather than agriculture is the means by which rapid improvement in Africa's living standards is possible. There are however, imperial specialists and apologists who urge the less developed countries to concentrate on agriculture and leave industrialization for some later time when their populations shall be well fed. The world's economic development, however, shows that it is only with advanced industrialization that it has been possible to raise the nutritional level of the people by raising their levels of income.' (Killick 1978: 46). 6 The SMC I was formed in the mid-1970s when Acheampong realized that the members of his coalition were becoming the subject of public resentment. Members of the SMC I were entirely service commanders.

Diffuse Resources: Ghana

169

the ranks and junior officers of the military. After a so-called 'house-cleaning' exercise the Armed Forces Revolutionary Council (AFRC) handed over power to a civilian government but only for a brief period. In December 1981 Rawlings headed a second coup and the PNDC (Provisional National Defence Council) was installed. In its first year in power, PNDC policies aimed to protect the 'vampire elements' that had been created by the predatory state through the early 1960s (Frimpong-Ansah 1991). These capricious military interventions meant that no coherent or sustainable economic policy could have been pursued especially during the late-1970s and the early-1980s. 10.3. EXTERNAL TRADE AND ECONOMIC GROWTH

10.3.1. Terms of trade, exports and imports Some commentators, especially politicians, have not hesitated to lay blame for the poor performance of the economy on external factors. Looking at the commodity terms of trade, however, it is difficult to conclude that external factors alone account for the growth collapse. True, the terms of trade for Ghana over the period 1960-71 averaged about 87 (1960 = 100). Moreover, import prices as well as export prices rose during the period of 'economic despair' (1972—82). However, the rise in the export prices outstripped that of import prices, resulting in an improvement in the commodity terms of trade averaging about 114 (1960 — 100) for these years (Table 10.2). This is 31 per cent higher than the pre-1972 period average and the highest in the country's post-independence history. Such a period of relatively high export prices would not have caused the abysmal economic performance of Ghana in the 1970s, if prudent policies had been pursued. Perhaps the most significant drop in the terms of trade in the 1960s occurred over 1965/66 when the index fell to 70 (1960= 100), coinciding with the first: negative growth rates since independence. This was associated with a decline in the ratio of Table 10.2. Physical capital accumulation and external tra.de indicators, 1960-1994

Gross investments: GDP (%) Gross savings: GDP (%) Terms of trade (1960= 100) NER (cedi/$)* EER (cedi/$)* Exports: GDP (%) Imports: GDP (%) Cocoa: GDP (%)

Cocoa: Exports (%)

1960

1966

1971

1977

1982

1994

24.38 17.14 100 0.71 0.71 28.2 34 13.9 49.29

12.86 7.86 70.64 0.71 1.14 14.6 19.6

14.12 9.64 85.65 1.03 1.10 15.76 20.24 8.1 51.39

11.06 10.01 154.5 1.15 3.72 10.49 11.55 6.1 58.13

3.38 3.73 79.64 2.75 24.10 3.34 2.98 1.2 35.95

15.9 4.54 n.a. n.a. n.a. 25.52 36.88

7.7 52.74

n.a. n.a.

*NER and EER are the nominal and equilibrium exchange rate respectively. Data for both series were obtained from Frimpong-Ansah (1991). Source: Data obtained from the World Bank (1997) and the World Data (1995).

170

R. Om Table 10.3. Selected indicators for the structure oj production and growth, 1960-1994

Real per capita GDP (US $) GDP growth (%) Inflation, CPI (annual %) Agriculture contribution to GDP (%) Industry contribution to GDP (%) Services contribution to GDP (%) Agriculture value added: GDP (%) Manufacturing value added: GDP (%) Industry value added: GDP (%) Services value added: GDP (%) Domestic absorption: GDP (%) Private consumption: GDP (%)

1960

1966

1971

497.48 n.a. n.a. n.a. n.a. n.a. 40.79 n.a. n.a. n.a. n.a.

467.44 -4.99 n.a. 0.80 -0.70 -5.08 43.27 10.21 18.37 38.36 n.a. 79

529.1 421.04 1.82 5.45 9.70 116.46 2.57 -2.83 1.02 0.31 2.57 3.63 44.15 56.21 11.00 10.78 15.84 18.28 27.95 37.57 104.48 101.06

73

77

1977

77

1982

1994

371.61 -6.49 22.28 -3.09 -2.76 -0.65 57.34 3.61 6.25 36.41 99.64

412.38 3.71 25.31 n.a. n.a. n.a. 46.39 6.49 15.95 37.65 111.36 n.a.

90

Source: Data obtained from the World Bank (1997) and the World Data (1995).

exports to GDP from 28 per cent in 1960 to 15 per cent in 1966, but the trend continued. The import: GDP ratio also fell from 34 to 19.6 per cent over the same period, 1960-66. Rimmer (1992) writing on the years 1959-65 points out that the deterioration in the terms of trade was not entirely an external problem as the supply of cocoa exports had increased over the period. Given that the world demand for cocoa was growing very slowly, and Ghana produced about two-fifths of the world output, it was to be expected that the average realized price of cocoa would fall. The fall in prices more than offset the rise in export volumes, however, to decrease the net barter terms of trade.

10.3.2. The importance of cocoa exports The economy of Ghana has been dependent on primary exports through most of the post-colonial era. Although the share of cocoa in GDP contracted through the 1960s and 1970s, it still constituted about 50 per cent of total exports (Table 10.3) with the exception of 1982. Over the period 1972-82 the share of cocoa in GDP averaged 7 per cent which compares with the pre-1972 average of about 11 per cent. This is so, even though the terms of trade over the 1971-82 period were about one-third higher than the pre-1972 average. The initial decline in the contribution of cocoa to GDP in the 1960s has been attributed to: variations in the weather, reductions in both the realized and producer prices of cocoa and the diminishing official support for cocoa production. The latter may be the key factor. For instance, by 1965 the UGFC (United Ghana Farmers' Council) agreed to the withdrawal of subsidies on insecticides and spraying machines, which in any case had become constrained by foreign exchange shortages.7 Cocoa has a gestation period of 5-10 years and so the continued fall in volume through the 1970s is partly a response to the discrimination in policy against the sector 7 The UGCF had been formed to protect the interest of farmers, but by 1965 had become a political wing of the UGCC (Nkrumah's Party).

Diffuse Resources: Ghana

171

in the 1960s. This discrimination, partly through excessive taxation and partly through the overvaluation of the cedi, depressed not only cocoa production, but also economic growth, exports and the import capacity upon which much of the protected manufacturing sector depended. The overvaluation was quite large as shown by the divergence between the nominal official exchange rate and the equilibrium exchange rate. For instance, by 1982, the equilibrium exchange rate was 776 per cent higher than the official rate (Table 10.2). Farmers resorted to smuggling cocoa to neighbouring Cote d'lvoire where producer prices were higher (Rimmer 1992). There were also reports in the early 1980s of farmers switching from cocoa cultivation to other food crops such as maize. Not surprisingly therefore, the proportion of cocoa in the total exports of the country had fallen from about 50 per cent in 1960 to about 36 per cent in 1982. The poor export performance brought about by erroneous policies, rather than external conditions, was a major feature in Ghana's growth collapse. Meanwhile, consistent with the staple trap model, total exports contracted from 28 per cent of GDP to 8 per cent of GDP between 1960 and 1980. Cocoa in Ghana shows very clearly how easily the potentially benevolent diffuse socioeconomic linkages can be converted into damaging 'point' linkages by a predatory state.

10.4. TH E NEGATIV E ECONOMI C CIRCL E 10.4.1. Economic growth in Ghana Ghana's growth trajectory since independence has displayed occasional blips (mostly in response to exogenous factors like the weather and external conditions) but the general trend from about 1960 to the early 1980s has been downwards. GDP growth in Ghana averaged just over 3 per cent between 1961—71 with an annual rate of inflation averaging a little over 8 per cent (Table 10.3). A substantial portion of that growth occurred in the 1970-71 period when it averaged 7.6 per cent. In fact, by 1966 (the year of Nkrumah's overthrow) Ghana was recording negative growth rates of — 5 per cent and real per capita income fell from $497 in 1960 to $467 in 1966 but it recovered to $529 in 1971. After the 1970-71 blip, a distinct feature of the economy was its inability to grow. The decline in GDP averaged —0.77 per cent per annum over the 1972-82 period. In 1975 GDP in real terms fell by over 13 per cent. The average growth in population over 1972-82 was about 2.4 per cent, so that real per capita GDP growth was —3.1 per cent. However, from 1983, the implementation of a gradual reform strategy under IMF supervision, the Economic Recovery Programme (ERP), is associated with some improvement, although with some concerns about the distributional impact. Some politicians (especially those in the opposition) and some economic commentators claim that the incomes of the poor are no better off than before the ERP (Weissman 1990). Yet Sarris (1993) argues that it is not likely that the livingstandards of the poor have deteriorated with the onset of the ERP. He uses data from the Ghana Living Standards Survey (GLSS) to compute the real incomes of various households for both the pre- and post-ERP period.

172

R. Osei

10.4.2. Production structure The economy on the eve of independence was thought to have a weak and fragile structure (Killick 1999: 53). Agriculture was the main economic activity in 1960 contributing over 40 per cent to GDP compared with just 2 per cent of GDP for manufacturing. Yet the Big Push industrialization strategy failed to establish competitive manufacturing so that the dependence on agriculture persisted. Rimmer (1992) notes that the consequence of an inflated domestic demand at the end of the 1960s was an increase in inflation and a balance of payments crisis. It could be argued that the Big Push strategy had not yet had sufficient time to prove itself. A counter argument however may be that some of the projects undertaken were 'white-elephants'. It is clear that the structure of the economy could not have been changed overnight. Agriculture, despite being discriminated against (both explicitly and implicitly) still out-performed both the manufacturing and services sectors, a condition that endured through the 1970s and early-1980s. For instance, over the 1971—82 period, the average for agriculture's share of GDP was about 54 per cent. This compares with about 10 per cent for manufacturing and 32 per cent for services. The industrial sector was hampered by shortages of raw materials and parts due to foreign exchange shortages. Added to this was the dominance of SOEs (stateowned enterprises), which were inefficient. The government barred private sector involvement in industry. An often stated but less documented fact is that people came to view the 'government' as just a group of people; consequently there was a lack of patriotism and a lacklustre attitude by workers in government employment. A moral hazard problem had been engendered because the cost of shirking in the public sector was very low. Table 9.4 reveals low levels of both social capital and quality of government for Ghana. This moribund state subsector together with the closed trade policy, discouraged competition and also the efficiency gains that normally come with it. Not only did the Big Push fail to create competitive industry, it also damaged agriculture and dissipated a significant fraction of the natural resource rents. 8 Considering that agriculture (and the cocoa sector in particular) contributed the most to GDP and absorbed the highest percentage of the labour force, policies that discriminated against this sector meant an increase in inequality. Increased inequality can have growth retarding effects for two reasons. The first is that the poor, who are more likely to invest in labour-intensive productive activities in which the country has a comparative advantage, cannot do so. Second, an increase in rural incomes through more pro-agriculture policies would lead to increased effective demand for manufactured goods, mostly with domestic comparative advantage and consumed mainly by the poor. 8 It is questionable whether even state gold mining companies were truly solvent. An unofficial source tells me that in most of the state gold mining companies, recurrent expenditures far exceeded revenues but the government still kept these mines running by borrowing from the central bank. 9 Although low inequality may not increase investment and consequently growth, if it is purchased through transfers, subsidies, and public sector employment policies (see Chapter 4).

Diffuse Resources: Ghana

173

The service sector fared better than either agriculture or industry. Over the period 1971-82, the contributions to GDP growth of agriculture, industry, and services were respectively —0.1, —0.88, and 0.23 per cent. The importance of services in GDP growth was due to the fact that the economy had basically become a commercial one. The huge rents that had been created by a chronically overvalued exchange rate made buying in the formal sector and selling in the parallel market a very attractive venture. People therefore shifted their entrepreneurial skills and resources into commerce, a move that also entailed less fixed investment and therefore less risk. The increasing importance of services in GDP is normally associated with development and/or modernization, but this is not so in Ghana. Killick (1978: 170) notes: 'Investment without growth cannot adequately describe the paradox of the Ghanaian economy, for it was also a case of modernization without growth'. In fact the structure of Ghana's economy in the early 1980s could best be described as 'a buy and sell one with a little bit of agriculture activity'.

10.4.3. Savings and investment Ghana's poor growth performance also reflected its capital mobilization and investment. In 1960, saving and investment as a percentage of GDP were a respectable 17 and 24 per cent, respectively. Over the next decade, however, these ratios both fell by about 2.2 per cent annually, although the initial high levels still implied a rapid growth in the capital stock. From 1972-82 the average saving rate decreased to 8 per cent of GDP from 11 per cent before 1972, and collapsed to 3 per cent in 1982. The investment ratio of about 19 per cent during the early 1960s was seen by many as a massive investment effort. Yet the efficiency of investment was low and the unimpressive growth performance led to the period 1960-65 being termed the 'period of investment without growth' (Killick 1978). A significant proportion of public investmafter the overthrow of oif nkrumah thhe er the the overthrow of Nkrumae investment ratio fell to 11 per cent and remained there until 1970 when it increased to 14 per cent. This improvement was not sustained, however, and the downward trend in investment resumed through the 1970s, primarily because economic life had become devoted to satisfying immediate private consumption needs (especially in the late-1970s and early-1980s). The increasing overvaluation of the nominal exchange rate over the 1972-82 period exacerbated the situation. In addition to the depressing effect of policies on private investment there are two further reasons why investment decreased. First, public investment was checked by the exhaustion of the country's external reserves and the collapse of foreign aid, which fell from $69 million in 1966 to —$1 million by 1971 (Killick 1978). For example, efforts to revive the SOEs during the years 1972-74 were abandoned. Second, government policy also discouraged foreign direct investment.

This i10es an unproductive public sector employment that is created by the government because of the high urban unemployment. This high urban unemployment comes about when there is a migration of rural labour to the urban areas in response to the wage gap between the modern urban and rural areas. See Gelb et al. (1991) for a fuller explanation.

174

R. Osei

Yet, successive governments were not able to undertake viable investments, while private sector investor confidence was lost.

1 0 . 5 . NEGATIV E CIRCLE S O F HUMA N AN D NATURAL CAPITA L 10.5.1. Demographic trends Resource-abundant countries like Ghana pass through the initial growth depressing stage of the worker/dependency cycle more slowly than resource-poor countries do. Sub-Saharan African countries in particular, lag most other developing regions in this respect (Auty 1998: 8-10). Whereas Malaysia offset the potential growth-depressing effect by a high rate of saving (as described in Chapter 9), Ghana did not, and as the country slid into the staple trap, its passage through the demographic cycle was arrested so that high population growth combined with low or negative economic growth to intensify the pressure on the natural resource base. Ghana's average population growth in the 1960s was 2.6 per cent and its worker/ dependency ratio was a relatively high 0.9 (Table 10.4). The dependency ratio remained high through the 1970s (Table 2.4), even though the population growth rate eased as Ghana became less economically attractive and the net immigration from neighbouring countries of the 1950s gave way to net emigration. The population growth rate slowed between 1972-82 to 2.4 per cent as deteriorating economic conditions led to a massive exodus from Ghana. The 1979 oil price shock in particular stimulated emigration of Ghanaians into Nigeria. Rado is quoted as saying that 'Ghana had lost over a half of its experienced, top level professional manpower' (Rimmer 1992: 177). The sector that suffered the most in terms of loss of labour was agriculture. In 1966, more than 60 per cent of Ghana's labour force were employed in agriculture, but this ratio fell to 56 per cent by 1980. Some reasons for the loss of agriculture labour are the migration to the urban areas in search of public sector jobs, extension of schooling, and the loss of foreign residents. The discriminatory agricultural policies coupled with the fact that the dependency ratio still remained high, meant that the decline in savings Table 10.4. Selected demographic and natural resource rent indices, 1960-1994

Total population (million) Population growth Primary school enrolment rate (%) Secondary school enrolment rate (%) Education expenditure: GDP (%) Genuine savings rate (%) Natural resource rents: GDP (%) Labour force in industry (% of total) Labour force in agriculture (% of total)

1960

1966

1971

1977

n.a. 3.3 n.a. n.a. n.a. n.a. n.a. n.a. n.a.

n.a. 1.93 n.a. n.a. n.a. n.a. n.a. 16 60

8.88 3.05 n.a. n.a. 3.81 5.08 0.44 n.a. n.a.

10.18 11.37 1.53 3.05 71 n.a. 36 n.a. 3.04 1.7 2.37 -12.01 0.33 2.87 17 n.a. 58 n.a.

Source: Data obtained from the World Bank (1997) and the World Data (1995).

1982

1994

16.64 2.71 n.a. n.a. 3.12 -4.06 3.77 n.a. n.a.

Diffuse Resources: Ghana

175

and therefore investment was inevitable. Meanwhile, the proportion of the labour force employed in industry actually increased slightly from the pre-1970 average of 16 to 17 per cent in the 1970s. Yet the industrial sector's share of GDP was declining (and contributing negatively to growth) during the 1970s. Most of the increased employment in the sector was in SOEs because urban rent-seeking activity continued. The implication is that agriculture was not only starved of capital investment but also of human resources. Even if the SOEs were absorbing surplus labour from agriculture, this did not translate into increased capital accumulation because the SOEs were highly inefficient.

10.5.2. Human capital accumulation The accumulation of human capital, a widely accepted ingredient of sustainable growth, has been lacking in Ghana. The primary school enrolment rate in 1970 was about 64 per cent. Even though this was considered high by the standards of West Africa, most people did not continue beyond primary education. Secondary school enrolment in the same period was just about 14 per cent. Both primary and secondary enrolment seems to have increased during the 1970s. However in the case of Ghana, this did not translate into effective human capital accumulation that leads to economic growth. A World Bank study found that the quality of education in Ghana had actually deteriorated (Killick 1999: 62). Most of those who did attend either a primary or secondary school during the 1970s would agree that educational standards were poor.11 Educational expenditures decreased during the 1970s as school enrolments were increasing. Yet it is the quality and not quantity of education that matters for human capital accumulation (see Chapter 4). It seems that human capital accumulation was deficient as economic growth collapsed. There is a sense in which the natural resource endowment shaped the nature and pace of capital accumulation in Ghana over the years. This was mediated through government policy. Dutch disease effects triggered by favourable world cocoa prices in the immediate years of independence may have lulled the Nkrumah government into thinking that it could embark upon a capital-demanding Big Push development strategy. The effect was to worsen income inequality and retard capital accumulation, especially in the agricultural sector. The ability of the majority of Ghanaians to invest in human capital was severely constrained because of government policy.

10.5.3. Sustainabiiity of Ghana's development path A prediction of the staple trap model is that resource-rich countries acquire an excessive dependence on primary exports with declining competitiveness so that the development trajectory is environmentally unsustainable. By 1971, Ghana's genuine savings rate was a little over 5 per cent suggesting that although the economy was 11

The author was in primary and secondary school between 1975 and 1989. Although both were reputed to be among the best schools in Ghana, books, teachers, and other infrastructure were inadequate. It was not uncommon for some teachers to leave midway in an academic year for 'Agege' as Nigeria came to be termed.

176

R. Osei

heavily dependent on primary product exports, it was still on a sustainable course. This, in turn, implies that despite the adverse policies towards the primary sector, the potential to create, maintain and use the primary sector output as an engine of growth had still not been eroded when Acheampong assumed power in 1972. In fact, mineral and forest rents as a percentage of GDP amounted to 4.6 per cent in 1980 and crop rents probably added a further 7 per cent (based on estimates that are available only for 1994, as explained in Chapter 8). The net level of rents is somewhat less in relation to GDP than the rent stream of Malaysia (see Chapter 9) and it was also more volatile. For example, mineral and forest rents were only 0.4 per cent of GDP in 1970, one-tenth of the 1980 level, while the boom of 1977-80 added the equivalent of 35 per cent to GDP from cocoa alone, some 9 per cent of GDP annually (Wetzel 1999: 98). Ghana was recording negative genuine savings by 1980 (see Table 8.4), bottoming out at -17 per cent of GDP in 1981. This increased reliance upon natural capital for the country's development can be attributed not only to irresponsible government officials but also to ordinary rural folk whose only goal was that of survival. The forest, in particular, was the main focus of exploitation by rural dwellers because it not only supplied cheap energy, but also income from the sale of firewood to other urban dwellers was used to supplement whatever farm income was received. Even in the urban areas, very few people could afford to rely solely on electricity for all their energy requirements. A higher demand for firewood and charcoal was created because these were cheaper supplements and/or substitutes to electric energy. Yet again, the evidence suggests that despite the erroneous policies begun after independence the economy still had the potential to revive in 1970. However, economic reform held little appeal to the government of a predatory state so that discrimination against the primary sector continued, causing the economy to decline. Ghana invested significant portions of its natural resource rents inefficiently in sectors that could not sustain economic growth. The majority of Ghanaians therefore resorted to the extraction of any natural resource that they had access to. This deepened the environmental problems of the country and created negative feed back loops in natural and social capital that proved difficult to break. 1 0 . 6 . CONCLUSIO N Ghana's growth collapse follows the staple trap model and reflects policy failure to a large extent. Although its rent stream was slightly smaller and more volatile than that of Malaysia, the principal difference between the two countries is the type of political state. The economic policies pursued in Ghana through the 1960s and 1970s derived from the predatory ('vampire') political state that evolved after independence. Nkrumah launched an over-ambitious development strategy in the early-1960s that was very much nurtured by the natural resource rents. These policies quickly entrenched vested urban interests and exhibited strong inertia so that successive governments not only recognized the 'vampire' elements sown by Nkrumah, but they also nurtured and enlarged them.

Diffuse

Resources: Ghana

177

The instability of the international economy in the 1970s adversely affected most primary product exporters. Yet, cocoa, the main export of Ghana, did well on the world market so that the net effect was to improve the terms of trade. Cocoa functioned as the milch cow for sustaining political support, nurturing development plans and building private fortunes and, as a result, it was continuously discriminated against both directly and indirectly. A major source of growth in total factor productivity results from the reallocation of resources from low- to high-productivity sectors. In the case of Ghana, however, resources were diverted from higher productivity agriculture into industry, which was relatively inefficient. In the process, the potentially beneficial diffuse socioeconomic linkages from peasant farming were degraded into damaging point linkages that benefited a relatively small urban elite. It took a combination of severe drought and sharpty higher foreign debt service in 1982 to finally force change in the political state, in return for assistance from the IMF. Ghana moved hesitantly towards a consensual democracy. Yet, even though some significant strides were made towards economic recovery from the mid-eighties, Ghana was still recording negative genuine savings in 1994 so that its development was still not sustainable. More stringent policies are required to halt the rate of land and environmental degradation, especially in the mining areas, but the critical deficiency appears to lie in social capital. The rectification of this deficiency requires stronger links to tie rural areas into the political and economic decision-making process. It also calls for recognition of the role of the growing bureaucratic-service elite in the democratic process if the threat of violence and social disruption is to be avoided (Mikell 1989).

REFERENCES Auty, R. M. (1998), 'Resource abundance and economic development: improving the performance of resource-rich countries', Research for Action 44, Helsinki: UNU/WIDER. Collier, P. and J. W. Gunning (1999), 'Explaining African economic performance', Journal of Economic Literature, 37, 64—111. Edwards, S. (1997), 'Openness, productivity and growth: what do we really know?' Paper presented at the CREDIT session of the Royal Economic Society Meeting in March 1997.

Frimpong-Ansah, J. H. (1991), The Vampire State in Africa: The Political Economy of Decline in Ghana, Trenton, African World Press. Gelb, A., J. B. Knight and R. H. Sabot (1991), 'Public sector employment, rent seeking and economic growth', The Economic Journal, 101, 1186-99. Killick, T. (1978), Development Economics in Action: A Study of Economic Policies in Ghana, London: Heinemann. Killick, T. (1999), 'Fragile still? The structure of Ghana's economy, 1960-94', in E. Aryeetey, J. Harrigan and M. Nissanke (eds.), Economic Reforms in Ghana: The Reality and Mirage, London: James Currey: 51-67. Krueger, A. O. (1993), The Political Economy of Agricultural Pricing Policy: A Synthesis of the Political Economy in Developing Countries, Baltimore MD: Johns Hopkins University Press.

178

R. Oset

Kusi, N. K. (1991), 'The management and performance of the Ghanaian economy, 1960-1982', Occasional Papers in Economic Development, Faculty of Economic Studies, University of New England, Australia. Mikell, G. (1989), 'Peasant politicisation and economic recuperation in Ghana: local and national dilemmas', The Journal of Modern African Studies, 27/3, 455-78. Rimmer, D. (1992), Staying Poor: Ghana's Political Economy, 1950—1990, Pergamon Press. Sarris, A. H. (1993), 'Household welfare during crisis and adjustment in Ghana', Journal of African Economies, 2/2, 195-237. Weissman, R. S. (1990), 'Structural adjustment in Africa: insights from the experiences of Ghana and Senegal', World Development, 18/12, 1621—32. Wetzel, D. (1999) Ghana's management of a temporary windfall: the cocoa boom of 1976—77, in: P. Collier, J. W. Gunning and Associates, Trade Shocks in Developing Countries: Vol. 1, Oxford: Oxford University Press: 94-141. World Bank (1997), World Development Indicators 1997, Washington DC: World Bank. World Data (1995), 'World Bank indicators on CD-ROM', Washington DC: The World Bank.

11 A Growth Collapse with Point Resources: Bolivia R I C H A R D M. A U T Y AND J. L. EVIA

11.1. M A N A G I N G M I N E R A L W E A L T H Bolivia is a relatively small mineral-driven economy whose political state was transformed into a consensual democracy by the trauma of hyperinflation and economic collapse in the mid-1980s. Prior to that, the political state was factional and governments swung between oligarchy and polarized democracy. This characteristic, when combined with the point socioeconomic linkages generated by its natural resource endowment, rendered Bolivia especially vulnerable to the staple trap. The principal features of the ore-driven variant of the staple trap model are summarized in column 4 of Table 8.6. The capital-intensive production function of mining confers a distinctive enclave character upon its socioeconomic linkages. The mining sector usually employs much foreign capital but only a small, albeit well-paid workforce, so that final demand linkage (i.e., domestic spending by capital and labour) is modest. Productive linkages are also limited because the mine inputs are usually imported due to their specialized nature, while the higher value added stages of mineral processing are usually marketorientated in their location. This leaves corporate taxes plus any resource rent that the government secures through royalties or special taxation, as the sector's principal economic stimulus. This concentrated income stream requires the government to play an expanded role in maximizing the mining sector's stimulus to the rest of the economy. This heightens the risk of policy failure if the government lacks the autonomy and motive to pursue an effective welfare-maximizing strategy. These inherent disadvantages are accentuated in the case of post-war Bolivia by two additional factors. First, a state-led growth model was established in 1952 that reinforced the dominant role of the government in the economy. Second, the country lacks a sound transport system. It has been landlocked since 1883 when war deprived it of direct access to the Pacific coast. It is also sparsely settled so that per capita transport costs are high and trade tax evasion (i.e., smuggling) is rampant. A recent study by Gallup et al. (1998) shows that the per capita GDP of landlocked countries outside of western and central Europe (where such countries are integrated into a modern

180

R. M. Auty and J. L. Evia

international transportation network) is one-third that of countries that are nonlandlocked. That study also shows that being landlocked may reduce a country's economic growth rate by 0.7 per cent. Gallup et al. (1998) go on to suggest that the disadvantage of being landlocked may be compounded by the shared problems of developing countries. These shared problems include long distances from large dynamic markets and policies that suppress trade and raise the cost of the imports required to build the infrastructure of a modern economy. Yet a growth collapse is not inevitable, Chapter 3 shows how environmental and natural resource accounting (EARA) furnishes a rationale for the management of mineral windfalls to secure sustained rapid economic growth. The government of a mineral economy needs to do three basic things in order to maintain the income component of the mineral stream (as opposed to the capital component) in perpetuity and to minimize the adverse impacts of short-run Dutch disease effects. It needs to sterilize the mineral windfalls, buffer public expenditure from a possible rent collapse, and invest a sufficient fraction of the natural capital in alternative wealth-producing assets. The effective implementation of this strategy requires, in turn, appropriate institutional mechanisms to manage the natural resource windfalls that include: • a capital development fund (CDF) to sterilize the capital inflows and mute the Dutch disease effects, • a mineral revenue stabilization fund to smooth the fluctuating levels of government revenue, and • a project evaluation unit to ensure the efficiency of public sector investment. There is also a strong case, based on empirical evidence, for erring on the side of caution in the domestic absorption of mineral windfalls (Gelb and Associates 1988; Auty and Mikesell 1998). In addition, and perhaps most fundamentally, adequate public social capital is required in order to provide a matrix within which such policies can be effective. The next section, Section 11.2, examines the initial conditions in Bolivia from 1952 to the early-1970s on the eve of the period of intensified external shocks. Section 11.3 then measures the scale and impact of the external shocks during the period 1970-95 and the resulting fluctuations in the rent stream from natural resources. Section 11.4 analyses the deployment of the natural resource rent and explains the growth collapse. Section 11.5 tackles the issue of the country's landlocked condition. 1 1 . 2 . TH E I N I T I A L CONDITION S

11.2.1. The natural resource endowment On the eve of the price shocks in 1970, Bolivia enjoyed a relatively high per capita cropland endowment of 0.55 hectares. Some twenty-five years later, the ratio was still just above the 0.3 hectares per capita boundary level that is used for classification as a resource abundant country (WRI 1998: 299). However, in the early-1970s, mining still dominated the Bolivian economy when it accounted for 77 per cent of exports, 44 per cent of government revenues, and 20 per cent of GDP (Table 11.1). The

Point Resources: Bolivia

181

Table 11.1. Changing production structure, Bolivia 1970-1995 (% GDP)

Agriculture Industry Mining Manufacturing Services

1970-72

1980-82

19.9 32.2 20.1 13.1 47.8

18.7 34.6 17.1 12.6 46.8

1993-95 16.5 33.0 6.5 18.3 50.5

Source: World Bank (1998) for 1970-72 and 1980-82; IMF (1998) for 1993-95; ECLA (1995) for mining share of GDP.

Bolivian mining sector was three times the size expected for a country of similar size and level of development (Syrquin and Chenery 1989). Yet, despite its abundant natural resources, Bolivia was still a poor country. Moreover, Dutch disease effects had already significantly weakened the structure of the non-mining economy (Auty 1993: 41). For example, the share of agriculture in non-mining GDP was little more than half that expected and although the share of manufacturing was close to the norm, it was a protected sector and little of its output was internationally competitive.

11.2.2. Human and social capital Given its relatively low per capita income, it is not surprising that Bolivian indices of human capital in the early-1970s lagged most other Latin American countries. The rate of population growth was still high at around 2.5 per cent per annum so that the country remained in the initial growth-repressing stage of Williamson's dependencycycle, with a dependency ratio of 0.87 (Table 2.3). Bolivian life expectancy was still only 46 years compared with 62 years in Mexico and the infant mortality rate at 153 per thousand was twice that of Mexico (World Bank 1992). Meanwhile, although public expenditure on education in Bolivia was similar to that of Mexico at around 2.2 per cent of GDP, primary school enrolment was relatively low at three-quarters of the Mexican level. Nor were the stocks of institutional and social capital well-developed (see Table 9.4). The system of state-led growth established by the 1952 revolution had failed to remove the social tensions within Bolivian society. It displaced a system of internal colonialism (wherein a local elite allied with external capital exploits the country) but income inequality remained moderately high. The ratio of the income share of the richest percentile was fifteen times that of the poorest percentile in the early-1970s. This is twice the gap in resource-poor South Korea and four times that of Taiwan. Although workers pressed the government to expand social expenditure by increasing taxes (levied on others), those with wealth successfully resisted such calls (Pastor 1991). This unresolved social tension created inflationary potential because the government was not able to expand domestic taxes to a level commensurate with the expenditure commitment that was assumed after the National Revolution in 1952. The 1952 revolution reacted against a pattern of mineral exploitation that had not fostered economic or social development, nor integrated the country. The current

182

R. M. Auty and J. L. Evia

account was in deficit and the government had to resort to international lending, for the development of infrastructure, the principal concern at the time. The Bolivian capitalists who controlled the mines all lived outside the country and invested the profits from Bolivian mining in the industrialized countries. The bulk of the fiscal income at that time was the taxes on international trade, mainly mineral exports, and the imports that were made possible by the resources that mining generated. 'The state had to depend basically on the taxes on international trade and external borrowing, as a consequence of the resistance of the local elite and the incapacity of the rest of the population to pay more taxes' (Drake 1995). Eder (1968) estimates that in the late1940s between 65 and 75 per cent of government revenues were derived directly or indirectly from mining. The revolution of 1952 nationalized the mines. It planned to use the resources from mining to invest in diversifying the economy through the domestic production of hitherto imported foodstuff like sugar, cattle, milk products, timber, comestible fats, vegetal oils, and rice (Guevara 1955). The diversification strategy also promoted oil exploration and drilling,which was seen as the best chance for the diversification of Bolivian exports. The strategy was based on colonizing the eastern lowlands, where the gas fields and best cropland were to be found. In 1955 the road to Santa Cruz was finished and migration was promoted to the lowlands where the agricultural bank provided credit and a sugar mill was built. The diversification strategy could only end the dependence on minerals, if the rents from mining were invested in projects with a proper rate of return. However, nationalization was associated with a slow decline of the mining sector. The mines were mismanaged because the workers tried to exploit them for their own immediate benefit. The miners had played a central part in the revolution so that the workers confederation enjoyed enormous power in the government. For example, the workers appointed the minister of mines and the mines were co-administered with the workers. Comibol, the state mining company, was expressly established to maintain and create new employment and also to provide welfare to the workers and their families (CEMYD 1990). This undermined the viability of Comibol as it paid higher wages, re-hired sick and retired workers, and tolerated 'social mines' (mines that were in deficit but kept in operation to avoid dismissing miners). 'According to the Comibol report, in 1952, under private operations, the nationalized mines employed 28,973 miners. In 1956, under Comibol operation, the mines employed 36,558 workers to produce one-half as many tons of ore. The 7,500 excess workers, according to the same report, were miners who had been dismissed by the private operators, but it was "indispensable" for Comibol to rehire them as they have played an important part in the 1952 revolution.' (Eder 1968: 542-3). Comibol increased the wages, employment, and social benefits to the miners at the expense of lower investment, notably in exploration and development of new reserves. This only exacerbated the neglect that had commenced before the revolution, when the mining companies attempted to withdraw an ever-increasing percentage of their profits from the country. The companies had invested minimally in exploring and developing new sources of ore due to political instability and the threat of

Point Resources: Bolivia

183

nationalization. After the revolution, mining equipment wore out due to inadequate maintenance and also due to failure to shut operations to prevent abuse to equipment. In many mines the replacement costs exceeded the value of the output and existing veins were quickly exhausted (Eder 1968: 505). The resources extracted from mining went to the government, which tried to use them to foster economic development. The public investment was inefficient, with the possible exception of the state oil company, YPFB. The government also increased employment, chiefly in state enterprises. The civil service approximately doubled in size (Eder 1968). A significant expansion of public investment took place in the 1960s and foreign aid began to finance the recurrent fiscal deficits. This investment created a series of industries dominated by a national smelter that commenced operations in 1971 but was, however, not economically viable due to the lack of adequate ore reserves. Almost all the projects undertaken by the Bolivian Development Corporation (BDC) were badly designed and poorly implemented so that they were uneconomical and wasted resources (Eder 1968: 63). Meanwhile, in 1969 the government nationalized the MNC, Gulf, which had been operating in Bolivia since the mid-1950s, and had discovered promising fields. The access of the SOE, YPFB, to these fields allowed it to sustain an important expansion in output until 1973. YPFB began to export petroleum and in 1974, oil exports represented around 25 per cent of total Bolivian exports. Nevertheless, these exports proved to be short-lived, disappearing by the end of the decade, although the start-up of the pipeline to Argentina in 1972 caused gas exports to assume some importance. In summary, although the production of commodities in Bolivia was mainly stateowned, the state lacked the capacity to execute its lead role in the economy. Public investment was typically half or more of total investment in the 1960s and 1970s, but the overall efficiency of investment was low. The rate of investment relative to GDP was 20 per cent (Table 11.2), close to the Syrquin and Chenery norm for a country of Bolivia's size and level of development, but the IGOR was disappointing. In the early-1970s the IGOR was around 5, half the efficiency of the resource-deficient Asian Table 11.2. Some economic indicators, Bolivia 1970-1994 1970-74

1975-79

Per capita GNP growth

1.8

0.9

Domestic absorption" Private consumption" Public consumption" Gross investment" Domestic saving" Exports0 Debt/GDP ratio Debt service/exports

99.4 64.0

104.3 64.8 12.2 22.1 23.0 25.6 77.8 32.9b

9.7 19.4 26.2 25.9 58.4 n.a.

1980-84 -4.0 97.6 69.4 16.6 11.2 14.0 15.6 115.7 48.9

Notes: "Expressed as a percentage of GDP; bData for 1990-93 only. Source: World Bank (1998).

1985-89

1990-94

-0.6 103.8 80.7 12.5 10.3

106.7 79.9 12.3 14.6

2.2

6.8

7.8

16.5 28.7 41.2

18.5 85.6 35.1

184

R. M. Auty and J. L. Evia

countries at a similar level of per capita income. Growth was extensive rather than being based upon rising productivity (IMF 1998: 18). Bolivia in the early-1970s was therefore a mineral economy that was already well advanced along the staple trap trajectory. Its economy was not diversified and exhibited clear signs of Dutch disease while its social indicators lagged and the factional political state had a low capacity to manage the state-led economy and social tensions alike. 1 1 . 3 . E X T E R N A L SHOCK S A N D N A T U R A L RESOURC E RENT FLUCTUATION S For Bolivia, the 1973 oil shock was strongly positive so that the gain in the terms of trade comparing 1974-78 with 1970-73 was 40 per cent. This was the result of favourable price shifts that saw tin prices double and oil prices quadruple during the period 1973—74. Given the ratio of Bolivian exports to GDP, this improvement in external conditions translated into a gain to the economy equivalent to an additional 12.1 per cent of GDP annually 1974—78. Such a positive shock contrasts with negative ones for neighbouring hard mineral economies over that same period, namely 4.3 per cent for Peru and 10.6 per cent for Chile (Balassa and McCarthy 1984). Thereafter, the 1979 shock was also initially positive for Bolivia, with an additional annual impact about one-quarter the size of the 1973 shock. The World Bank (1999) provides estimates for the years 1970—94 of the scale of the rent stream from Bolivian natural resources. The figure shown in Table 11.3 excludes the rents from pastureland and cropland. Preliminary data from the World Bank for 1994 suggest that these land uses each generated rents that were equivalent to around 4 per cent of GNP, slightly more for land and slightly less for pasture. If, as seems likely, this ratio has been relatively stable over time, the importance of rents from rural land use may have exceeded that from minerals in the early-1970s, but then declined to one-half through the rest of that decade, and to one-third in the early-1980s. Fluctuating mineral prices caused the contribution of natural resource rents to GNP to rise sharply over the years 1974-85, reaching one-third of GNP if land rents are included. The price increases transformed Comibol into a potentially profitable company. The earnings of YPFB also increased overnight so that the government Table 11.3. Share of mineral rentsa in GNP 1970-1994, four countries (% GNP) Country

1970-73

1974-78

1979-81

1982-85

1986-90

1991-94

Bolivia Indonesia Mexico Saudi Arabia

6.47 6.70 1.77 63.7

15.95 14.78 4.54 68.26

18.18 27.87 15.58 84.63

25.54 19.45 19.51 45.73

7.47 10.24 9.53 37.16

6.00 6.44 4.72 42.93

Note: aOil and gas only, except in the case of Bolivia that also includes rents from tin at 2.83% GNP 1970-73; 4.90% GNP 1974-78; 6.79% GNP 1979-81; 1.87% 1982-85; 0.57% 1986-94. Source: World Bank (1999).

Point Resources: Bolivia

185

Table 11.4. Fraction of Bolivian taxes from mining and petroleum 1970-1997 (%) Year

Share

Year

Share

Year

Share

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979

16.6 9.2 18.7 37.8 49.8 34.0 39.7 35.5 34.4 26.5

1980 1981 1982 1983 1984 1985 1986 1987 1988 1989

31.8 36.4 33.4 16.8 40.0 73.6 62.0 56.2 57.6 57.4

1990 1991 1992 1993 1994 1995 1996 1997

59.6 58.1 47.3 52.0 44.2 39.4 37.8 30.6

Source: Morales et al. (1992), elaborated with data of UDAPE.

found itself with a flow of resources from mining and petroleum companies. In addition, higher mineral prices improved the access of Comibol and YPFB to international credits. A big part of the rents generated in the commodities sector accrued to the central government in the form of taxes. Table 11.4 shows the percentage of the government income derived from taxing the mining and petroleum sector. In the early-1970s the taxes came from the petroleum sector and private mining. Higher prices then raised tax receipts from the public companies and between 1973 and 1983, taxation on commodities accounted for around 35 per cent of the government tax revenue. However, the mineral rents subsequently collapsed dramatically. Table 11.3 shows that, relative to GNP, the fluctuations in the Bolivian mineral rent stream were on a similar scale to those that impacted Indonesia and Mexico. Yet, whereas the government of Indonesia was able to convert the rents into sustained rapid economic growth, that of Bolivia like its counter-part in Mexico (see Chapter 13), mismanaged the rent stream and experienced a growth collapse.

1 1 . 4 . TH E BOLIVIA N GROWT H COLLAPS E 11.4.1. The windfall relaxes market discipline Bolivia therefore provides a very clear example of a country that used a resource boom to relax market discipline. The institutions required to manage the mineral windfalls sustainably that were outlined in Section 11.1 were not established. Instead, the government used the windfall to backtrack on the reform policies that it had adopted under external pressure in the early 1970s. Those policies aimed to stabilize the economy, diversify taxes and reallocate resources in line with the country's comparative advantage. The government failed to sterilize the additional foreign exchange inflows by accumulating overseas assets in a CDF or by retiring the country's sizable external debt (Table 11.2). Worse, it used the hydrocarbon reserves as collateral to increase foreign borrowing. The government expected mineral production to double

186

R. M. Auty and J. L. Evia

and oil production to quadruple 1976-80 (Cariaga 1982) and, in effect, brought forward the expected revenue stream by borrowing from abroad. This amplified the mineral rent windfall and the associated Dutch disease effects, the opposite of the effect required. The bulk of the mineral windfall went into higher consumption. One way it did this was through substitution of windfall taxes for direct taxes. The second way of increasing consumption was through higher import purchases, made cheaper by the appreciation of the real exchange rate.

11.4.2. Dutch disease effects increase the specialization of the economy The Bolivian government failed to translate the windfall into investments that would reverse the long-run fossilization of the economy or diminish its dependence on minerals. Although public investment was poured into expanding export crops in the lowlands in line with the diversification strategy of the revolution, much of it proved to be ill-engineered and not justified economically. Therefore, a modest increase in the rate of investment 1975-79 was associated with decelerating economic growth (Table 11.2). It would have been preferable to have saved the revenue so invested in a CDF, rather than absorb it inefficiently in the domestic economy. Instead, the competitiveness of the non-mining tradables sector was sapped as the rapid domestic absorption of the windfall rents caused the real exchange rate to appreciate (strengthen) by 17 per cent 1973-74, and then to double between 1979 to 1983 (Wood 1988). The trade deficit widened, despite the sharp real increase in export earnings brought about by higher mineral prices. As the growth in imports outstripped that of exports, the trade deficit ballooned to average 4.8 per cent of GDP 1974-78. Faced with growing trade and fiscal deficits, the Bolivian government expanded its foreign debt to 78 per cent of GDP through 1975-79. The debt/export ratio reached one-third. By 1982, very little non-mining activity was competitive (Table 11.5) and its share of exports slumped to 5.2 per cent of the total in 1985. The average rate of effective protection for the non-mining tradables at that time is estimated at 44 per cent, but the range of protection was from 17.5 per cent for forestry, through 33 per cent for agriculture to 94 per cent for manufacturing (Morales 1987). Agriculture was barely two-thirds of the Syrquin and Chenery norm for a country of Bolivia's size and level of development and generated only one-sixth of GDP although it still employed half the Table 11.5. Composition of exports, Bolivia 1970-1994 (%)

Mining Oil + Gas Other primary Manufacturing

1970-74

1975-79

1980-84

1985-89

1990-94

73.3 16.2 8.1 2.2

64.4 20.8 11.3 2.2

S0.8 39.3 6.4 2.1

40.2 43.1 13.8 2.7

41.3 18.1 26.0 12.9

Source: World Bank (1999).

Point Resources: Bolivia

187

workforce. Yet the 1952 reforms had transferred sizable amounts of land to peasants on the Altiplano and upper Andean valleys. However, the redistributed farm units were further subdivided through inheritance so that they became even less viable. Attempts to form cooperatives and to increase the capital available to the Altiplano farms proved inadequate. Agricultural growth therefore occurred primarily through extension of the cultivated area in the mid-altitude valleys and eastern lowlands, rather than through higher yields on the Altiplano. Within manufacturing, the import substitution policy had stimulated three main subsectors: two-fifths of sectoral output came from agro-industry, one-third from light industry (including some mining equipment) and one-sixth from textiles and leather (Ministerio de Planeamiento y Coordinacion 1989). However, indicative of strong distortion, there was substantial dispersal of effective protection rates around the mean. The levels of effective protection for tobacco, textiles, leather goods, and wood products were almost twice the mean. Yet these products should have required least protection because they are labour-intensive users of local raw materials targeted at a domestic market that is insulated from imports by high transport costs. Despite such high levels of protection, the Dutch disease effects caused manufacturing output to shrink by 7 per cent per annum 1978-86 compared with a 3 per cent per annum contraction for the economy as a whole (Morales 1987). The share of manufacturing in GDP fell over that period from 15.4 to 10.8 per cent. Consistent with the staple trap model, even as the Dutch disease effects intensified, the government decapitalized mining (Casanovas 1990). Higher prices tripled the rents from tin mining alone to almost 8 per cent of GDP 1977-80, but the windfall was taxed away from Comibol (Ayub and Hashimoto 1985) which stopped publishing accounts in 1977. It was a high-cost producer (Table 11.6) and its output declined by two-fifths in the late-1970s (CEPAL 1989; Suttill 1988) while its cumulative losses over the years 1981-85 are reported to have exceeded US$500 million. When tin prices fell at the beginning of the 1980s, Comibol tried to reduce its expenses, including its investments. However, the trade unions, especially the miners, had played an important role in the restoration of democracy in 1978 and they demanded Table 11.6 Comibol production costs 1980-1986 (US cents/lb)

Mining Royalties Other Cost cif LME price

1980

1981

1982

1983

1984

1985

1986

4.06 1.81 1.16 7.10 7.17

5.51 0.64 1.76 7.99 6.39

5.24 0.68 1.76 7.68 5.69

6.65 0.91 1.93 9.49 5.89

7.83 0.47 1.34 9.65 5.53

9.40 0.50 1.47 11.37 5.32

4.05 0.04 0.08 4.86 2.57

Note: Production (tonnes) Tin Zinc Lead 1982 17,493 23,726 7,233 1986 4,232 3,915 1,319 Source: Suttil (1988), 46.

188

R. M. Auty and J. L. Evia

higher wages and benefits. Relations between the government and workers collapsed with severe consequences for mineral production. Between 1984 and 1985 Comibol required more resources from the government, so that mining demanded resources from the rest of the economy, instead of providing them. The impact of the deterioration of the hard mineral sector on the rest of the economy was, however, partly offset by the higher rents from hydrocarbons, especially from gas exports to Argentina, and from forestry, which together averaged almost 15 per cent of GDP 1974—85. But hydrocarbon production had been adversely affected by nationalization. The government maintained low domestic prices that simultaneously discouraged exploration and encouraged wasteful consumption, so that reserves were depleted too quickly. YPFB over-produced in order to generate revenues faster, but its cash flow was still inadequate and it was forced to borrow heavily. Yet the investments in oil and gas exploration yielded negative results. Overall, the combination of the squeeze on mining revenues and Dutch disease effects meant that mineral rents were unable to sustain economic growth even when mining was booming, let alone when mineral prices fell. Table 8.5 shows that the combination of a modest rate of domestic saving and investment (Table 11.2), and high domestic consumption of natural capital caused the scale of negative true saving to exceed even that of Saudi Arabia in relative terms during the years immediately following the 1979 price shock.

11.4.3. Sociopolitical failure The benefits of the consumption boom had failed to trickle down to lower income groups. The US exerted pressure on the military government to stand down in 1978. A leftist coalition won the 1978 election but was prevented from assuming office by political manoeuvres and the ensuing struggle between left and right led to nine changes of government from 1978-82 and caused policy to drift. Political weakness hampered economic stabilization because that policy was expected to have adverse consequences for income redistribution (Morales 1988). Consequently, none of the Bolivian governments that briefly held power 1978-82 was able to enlist support to boost state revenues in line with expanding expenditures. Foreign finance was used to cover the fiscal deficit but the debt/GDP ratio jumped once again (Table 11.2). Negotiations on debt rescheduling with the commercial banks broke down in 1981 and both the World Bank and IMF ceased lending. Sachs (1987) calculates that this cut-off in foreign credit caused a net transfer of resources from Bolivia that was equivalent to the loss of 10 per cent of GDP over 1980-83: the cut-off accelerated the cumulative deterioration of the economy and triggered hyperinflation. The inflation rate, which had been low by Latin American standards at 25 per cent, jumped to 300 per cent by late-1982 and then accelerated to exceed 500 per cent and approach 20,000 per cent in 1985. Not surprisingly, the rate of investment sank through the early-1980s (Table 11.2) and private investment fell to a meagre 3 per cent of GDP. The government attempted to maintain foreign debt payments, but in 1984 it was forced to limit foreign debt repayments to one-quarter of export earnings.

Point Resources: Bolivia

189

The frightening and rapid descent of the Bolivian economy into hyperinflation welded a political consensus in 1985 that abruptly halted the spiral. It transformed the political state from a polarized democracy into a consensual democracy. The consensus found expression in the New Economic Policy that was implemented under extraordinarily adverse circumstances due to the dramatic collapse in both tin prices and hydrocarbon prices. The price collapse triggered a 45 per cent decline in Bolivia's terms of trade 1984-88. The loss of rents was equivalent to 18.1 per cent of GDP annually 1986-90 compared with 1981-85 (Table 11.3).

1 1 . 5 . TH E IMPAC T OF BOLIVIA' S LANDLOCKE D CONDITION There seems little doubt that poor transportation aggravated Bolivia's economic problems. Gallup et al. (1998) attribute the inferior economic performance of landlocked countries to three factors, namely: limits on cross-border labour migration, compared with internal migration; added difficulty in coordinating trans-border infrastructure projects; and deliberate attempts by coastal states to deny economic opportunities to their landlocked neighbours. None of these factors appears to have been significant in the case of Bolivia. Rather, the remoteness of much of Bolivian economic activity combines with difficult terrain and the long-term neglect of infrastructure to require that the country exports products with a high ratio of value to weight (Morales and Sachs 1988). High transport costs also absorb a significant fraction of the potential rents from gasbased exports in the east. Plans to use the gas in combination with adjacent high-grade iron ore for heavy industry languished for lack of markets and of capital. Instead, a pipeline was constructed to Argentina that generated meagre revenues of $1.12 per thousand cubic feet (IMF 1998). In addition, high transport costs act like a natural barrier to trade and give protection to industry which, being shielded from competition, tends to be inefficient. This in turn creates an anti-export bias that precludes domestic manufacturing from expanding into foreign markets and securing the economies of scale. With high transport costs, many products are effectively nontradable and this may exacerbate the Dutch disease problem in the event of a rise in the price of natural resource exports. High transport costs boost not only the cost of domestic exports but also the cost of domestically manufactured goods that rely on imported inputs. Gallup et al. (1998) point out that reliance on imported inputs ma3r reduce the share of value added on manufactured goods to one-fifth or one-tenth of the product price. This means for example that, if poor communications raise the cost of transport by a modest 5 per cent of the product price, it translates into 25 per cent of even 50 per cent of the local value added component. Even more problematic than lack of access to the sea, is the geography of Bolivia that makes it difficult and expensive to construct and to maintain communications within the country. Bolivia is crossed by two mountain ranges which divide the country into three geographical areas with different altitudes. This has hampered the development of a road or rail system and thereby maintained the high costs of migration and

190

R. M. Auty and J. L. Evia

commerce within the country. Only with the conclusion of the road to the lowlands in 1955 was it possible to travel by road across the three geographical areas. The high transport costs preclude the formation of a 'national market', which in turn precludes taking advantage of the agglomeration and specialization economies. Andersen et al. (1999) provide some evidence on the lack of geographical specialization in the industrial sector in Bolivia. The high transportation costs in Bolivia are therefore as much the product of deficient internal transportation system due to inadequate public sector investment exacerbated by difficult terrain, as a result of the country being landlocked. Moreover, a landlocked country faces very different prospects if it is adjacent to a large and potentially dynamic economy like Brazil, rather than small faltering economies like those in much of sub-Saharan Africa, with which Ghana must contend. The Bolivian economy is being reoriented towards the eastern lowlands and the country became an associate member of Mercosur in March 1997 (EIU 1997). A 3,150km gas pipeline to Brazil was completed in 1999 that is expected to double Bolivian exports to the Mercosur market and add an extra percentage point to the economic growth rate (Economist 1997). Plans are also being considered for a $1.3 billion project to construct a waterway in the eastern lowlands to link Bolivia with the four Mercosur countries along which the Paraguay and Parana rivers flow. Possibly even more significantly, as free trade is phased in, the revenues lost as a result of informal trading across Bolivia's extensive and porous borders, will also be much reduced. For example, according to official statistics, Bolivia's combined imports from Argentina, Brazil, Chile, and Peru totalled $517 million in 1996, yet the exporting countries claimed that $1.1 billion was exported (Economist 1998). The overall cost to government revenues throughout the economy from lost import duties and value added taxes is estimated at almost $450 million in 1997, a sum equivalent to 5.6 per cent of GDP or to 24 per cent of total government revenue (IMF 1998).

1 1 . 6 . CONCLUSION S In contrast to the newly-independent states of Malaysia and Ghana, by the early1960s, a series of factional governments had pushed the mineral-driven Bolivian economy well along the staple trap trajectory. Dutch disease effects had weakened agriculture and manufacturing while the mainly state-owned mining sector was fossilizing through neglect of investment. Consistent with the staple trap model, the sizable increase in resource rents that the commodity price shocks of 1973 and 1979 triggered, caused economic policy to regress. Instead of sterilizing the foreign exchange inflow by repaying debt and/or investing in a CDF, the Bolivian government relaxed market discipline further and backtracked on an economic stabilization and reform package. Nor were the increased resource rents applied to diversify the economy, eliminate the legacy of poor transportation, or improve social and institutional capital. Instead, the government amplified the boom effect of the windfall by increasing foreign borrowing and expanding consumption.

Point Resources: Bolivia

191

The resulting over-rapid domestic absorption of the mineral windfall caused the real exchange rate to appreciate through the 1970s and early-1980s. This intensified the country's reliance on minerals, even as the mining sector was being squeezed in order to transfer resources to the non-tradable sector. Price repression led to an overrapid depletion of oil reserves, and underinvestment in the state-owned hard mineral sector corroded its competitiveness still further. When foreign capital ceased to bridge the fiscal and trade deficits, the economy imploded into hyperinflation. Although the trauma of hyperinflation acted as a catalyst for transforming the political state into a consensual democracy and also for economic reform, economic recovery was slower than policy advisers expected. The problems of economic development in Bolivia are compounded by inadequate public investment in internal transport and difficult terrain. But Bolivia also shows that being landlocked is less problematic if the economies of neighbouring states are soundly managed and the transport infrastructure is maintained so that the landlocked country can specialize in its comparative advantage. Geography is providing new economic opportunities through the revival of the neighbouring large economies of Brazil and Argentina. Bolivia will only realize these opportunities if its government adheres to the basic principles of EARA and does not relax market discipline. A consensual democracy should be able to achieve this. The next chapter shows with reference to Saudi Arabia what happens when the basic principles of EARA are espoused, but market discipline is relaxed by a paternalistic autonomous benevolent state.

REFERENCES Andersen, L., E. Antelo, J. L. Evia and O. Nina (1999), 'Geography and development in Bolivia: migration, urban and industrial concentration, welfare and convergence', LaPaz: Universidad Catolica Boliviana (mimeo). Auty, R. M. (1993), Sustaining Development in Mineral Economies: The Resource Curse Thesis, London: Routledge. Auty, R. M. and R. F. Mikesell (1998), Sustainable Development in Mineral Economies, Oxford: Clarendon Press. Ayub, M. A. and H. Hashimoto (1985), The Economics of Tin Mining in Bolivia, Washington: World Bank. Balassa, B. and F. D. McCarthy (1984), 'Adjustment policies in developing countries 1979-83: an update', World Bank Staff Working Paper 675, Washington DC: World Bank. Cariaga, J. (1982), 'The economic structure of Bolivia after 1964, in J. Ladman (ed.), ModernDay Bolivia: Legacy of the Revolution and Prospects for the Future, Centre for Latin American Studies, Arizona State University, Phoenix Az. Casanovas, W. M. (1990), 'Capital accumulation and evolutionary nationalism in Bolivia, 195285', in C. Anglade and C. Fortin (eds.), The State and Capital Accumulation in Latin America, Vol. 2, London: Macmillan: 32—53. Centro de Estudios Miner243a y Desarrollo (CEMYD) (1990), Desempe241o y Colapso de la Mineria Nacionalizada en Bolivia, La Paz: CEMYD.

192

R. M. Auty and J. L. Evia

CEPAL (1989), The Tin Industry in Latin America, Santiago (mimeo). Drake, Paul (1995), Kemmerer en los Andes, Quito: Banco Central del Ecuador. ECLA (1995), Statistical Yearbook for Latin America and the Caribbean, Santiago: Economic Commission for Latin America. Economist (1997), 'Bolivia: an example in the Andes', The Economist (9 August). Economist (1998), 'Old Bolivian customs', The Economist (11 April). Eder, G. (1968), 'Inflation and development in Latin America', Michigan International Business Studies, 9. EIU (1997), 'Bolivia trade', Economist Intelligence Unit (11 February). Gallup, J. L., J. D. Sachs and A. D. Mellingcr (1998), 'Geography and economic development', NBER Working Paper 6849, Cambridge MA: National Bureau of Economic Research. Gelb, A. H. and Associates (1988), Oil Windfalls: Blessing or Curse?, New York: OUP. Guevara Arze, Walter (1955), Plan de Politica Economica de la Revolution National, La Paz: Government Printery. IMF (1998), 'Bolivia: Selected Issues', IMF Staff Country Report 98/125, Washington DC: International Monetary Fund. Ministerio dc Planeamento y Coordinacion (1989), Estrategia de Desarrotto Economic y Social, La Paz. Morales, J. A. (1987), Estudio Integral Del Arancel de Imporlaciones: Implications Para La Estructura De Proteccion a La Industria Manufactuna National, La Paz (mimeo). Morales, J. A. (1988), 'Inflation stabilization in Bolivia', in M. Bruno et al. (eds.), Inflation Stabilization, Cambridge MA: MIT: 307-57. MoralesJ. A. and J. D. Sachs (1988), Bolivia: 1952-86, San Francisco: ICS Press. Morales, J. A., J. Espejo and G. Chavez (1992), 'Shocks Extcrnos y Politicas de Estabilizacion para Bolivia', Working Paper 03, La Paz: Instituto de Investigaciones Socio Economicas, Universidad Catolica Boliviana (mimeo). Pastor, M. (1991), 'Bolivia: hyperinflation, stabilization and beyond', Journal of Development Studies, 27, 211-37. Sachs, J. D. (1987), 'The Bolivian hyperinflation and stabilization', American Economic Review Papers and Proceedings, 77/2, 279-83. Suttill, K. R. (1988), 'Bolivia's silver lining', Engineering and Mining Journal, 44-8 (March). Syrquin, M. and H. B. Chenery (1989), 'Patterns of development, 1950 to 1989', World Bank Discussion Paper 41, Washington DC: World Bank. UDAPE (Unidad de Analisis de Politicas Economicas), Dossier de information de Estadisticas Economicas de Bolivia, Vols. 5, 6, 7 and 8, La Paz: Udape. Wood, A. (1988), 'Global trends in real exchange rates, 1960 to 1984', World Bank Discussion Paper 35, Washington: World Bank. World Bank (1992), World Bank Tables 1992, Washington DC: Johns Hopkins University Press. World Bank (1998), World Tables, Washington DC: World Bank. World Bank (1999), Wealth Estimates and Genuine Savings, Washington DC: Environment Department, World Bank. WRI (1998), World Resources 1998-99, New York: Oxford University Press.

12 A Growth Collapse with High Rent Point Resources: Saudi Arabia RICHARD M . AUTY

1 2 . 1 . PRINCIPA L FEATURE S O F A CAPITA L SURPLU S MINERAL ECONOM Y Saudi Arabia provides an example of the fourth basic variant of resource-abundant growth identified in Chapter 8 (see column 5 in Table 8.6). During the 1974-78 and 1979-81 oil booms, an exceptionally large rent stream was furnished to those oil producers like Saudi Arabia whose populations were small relative to their oil reserves. At that time the rent comprised more than 85 per cent of the oil price and the aggregate rent often exceeded the size of the non-oil GDP by several orders of magnitude. Nankani (1979) recognized this condition as a special case of the resourceabundant economy, namely the 'capital surplus' oil exporter. Such countries might evade the staple trap by investing sufficient financial assets offshore and becoming a rentier economy, like Brunei. Even if the rents are not on a sufficient scale to open up the rentier option, they permit the accumulation of financial reserves with which to lengthen the adjustment to an abrupt contraction of the rent stream, compared with a capital-deficient oil-driven economy like that of Indonesia. A second advantageous feature of the capital-surplus economy is that trade policy closure is eschewed. Sachs and Warner (1995) attribute this to a reduced concern on the part of the governments of capital-surplus economies with the need to diversify away from oil dependence. But, in addition, an open trade policy helps to reduce the Dutch disease effects during a resource boom. The expenditure of foreign exchange on imported goods helps to sterilize the foreign exchange from export earnings and so reinforces the sterilization role of the accumulation of financial reserves overseas. It is especially effective if the imported goods are for infrastructure projects because they are likely to be capital-intensive (Gelb and Associates 1988). Moreover, if trade openness extends to labour markets as well as goods, domestic inflationary pressure is further constrained because foreign workers and construction companies help to eliminate bottlenecks in domestic absorptive capacity. Saudi Arabia had a third advantage for development, namely that it was an autonomous benevolent political state. Its government did accumulate large financial

194

R. M. Auty

reserves, it also maintained an open economy and it invested a significant fraction of the oil rents in economic infrastructure and human capital. Yet, the economy still experienced a prolonged contraction after the 1979-81 oil boom. The principal reason for this is that the Saudi government functioned as a paternalistic autonomous benevolent state, rather than as the developmental variant (see Table 8.1). This suggests that the misallocation of resources in resource-abundant economies is rooted in the relaxation of market discipline that all but the developmental political states tolerate, rather than in trade policy closure per se. This chapter is structured as follows. The next section, Section 12.2, establishes the sizable fluctuations in the country's external terms of trade 1970-95. It measures both the windfalls during the oil booms and also the subsequent contraction in oil rents. Section 12.3 then evaluates the deployment of the oil windfalls during the 1974-78 and 1979—81 oil booms. Section 12.4 appraises the post-boom adjustment to lower oil rents. Finally, the implications of the findings are assessed for both theory and policy.

12.2. EARLY-1970s ' PRECONDITION S AND SCAL E O F FLUCTUATION I N EXTERNA L CONDITION S The Saudi economy grew rapidly in the early-1970s but the growth was almost entirely due to a doubling of oil production (Fig. 12.1) so that the Saudi economy was becoming ever more dependent upon a single commodity, in this case, oil. The oil rents already accounted for almost two-thirds of Saudi GNP in the early-1970s. Table 12.1 shows that this ratio is very much higher than in the case of two capital-deficient oil-exporting countries, Mexico and Indonesia.

Figure 12.1 . Oil production and revenues Source: 1970-88 SAMA (1998): 11; 1992-97 SAMA (1998): 21; 1989-91 not available due to data source changes.

High Rent Point Resources: Saudi Arabia

195

Table 12.1. Oil and gas rents share in GNP 1970-1994, three countries (% GNP) Country

1970-73

1974-78

1979-81

1982-85

1986-90

1991-94

Indonesia Mexico Saudi Arabia

6.70 1.77 63.7

14.78 4.5 68.26

27.87 15.5 84.63

19.4 19.5 45.73

10.24 9.53 37.16

6.44 4.72 42.93

Source: Worl d Ban k (1999).

Table 12.2. Changing production structure, economy of' Saudi Arabia 1970-1996

Oil (%GDP) a Non-oil (% GDP)a Share in non-oil GDP (%) Primary Manufacturing Construction + utilities Services

1970-72

1979-81

1994-96

66.6 32.5

66.1 33.1

36.7 61.9

11.8 22.2 16.9 49.2

4.7 14.4 31.9 48.5

11.4 14.6 15.0 58.8

Note: aOil and non-oil sum to less than 100 due to omission of import taxes. Source: SAMA (1998).

There is also evidence of Dutch disease effects in the early-1970s. Although the non-oil tradable sector generated 34 per cent of non-oil GDP, close to the Syrquin and Chenery (1989) norm for a country of Saudi Arabia's size and level of development ($1,550 per capita income in 1980 dollars), agriculture was experiencing a rapid relative decline. Agriculture generated only 12 per cent of non-oil GDP (Table 12.2), barely two-thirds of the Syrquin and Chenery norms, yet it still employed two-thirds of the workforce. Although manufacturing compensated for the modest size of the agricultural sector and was almost twice as large, it was dominated by oil refining which is a minimal diversification from oil extraction. Moreover, oil refining is capitalintensive and provides very little employment. Finally, the construction and utilities sector was some 4 per cent of non-oil GDP larger than would be expected. In summary, the Saudi economy was already heavily dependent on oil rents in the early1970s and therefore vulnerable to a price shock. The 1973 oil shock and subsequent oil boom coincided with a further expansion in Saudi oil production. Given the fact that oil prices quadrupled, oil revenues increased more than eight-fold in current dollars (Fig. 12.1). During the first oil boom, Saudi Arabia secured additional oil revenues that were equivalent to an extra 176 per cent of non-oil GDP annually 1974-78. In the case of the 1979-81 oil boom, oil production increased by a more modest amount, but when combined with a further doubling in prices, the stream of oil revenues more than doubled compared with the 1974-78 boom. This boosted revenues on top of the previous large increase by the equivalent of an extra 82 per cent of non-oil GNP annually 1979—81. A second index of the magnitude of the adjustment required by the Saudi economy to the oil shocks, in addition to the

196

R. M. Auty Table 12.3. Scale of Saudi oil revenues 1970-1996

Oil output (million barrels) Oil revenue (billion riyals) Non-oil GNP (billion rivals)

1970-73

1974-78

1979-81

1982-85

1986-90

1991-96

2.028 15.98 11.31

3.041 109.43 53.16

3.561 236.44 154.74

1.697 192.58 218.36

1.866 58.55a 212.32

2.964 114.06" 282.53

Note: "Oil revenue data unavailable for 1990-91. Covers 1986-89 and 1992-96 only. Source: Ministry of Planning (1986), SAMA (1998).

revenue stream, is provided by the terms of trade (World Bank 1998). Taking 1987 as the base year 100, the index jumped from 29.7 for 1970-73 to 108.4 during the 197478 oil boom and then leapt again to stand at 165.0 during the 1979-81 boom. When oil prices weakened after 1981 the Saudi government initially responded by halving its oil production in an effort to bolster global prices. This caused the Saudi oil revenue stream to contract. The annual loss was equivalent to one-fifth of non-oil GNP 1982-85 (Table 12.3). The strategy of curtailing output failed to restore prices, however, so the Saudi government altered its approach and sought instead to discipline other members of the OPEC price cartel by modest increases in its own oil production. Such increases also promised to ease the scale of cuts in both foreign exchange and government revenue that the country had experienced. The new strategy worked to the extent that international oil prices did recover somewhat and Saudi oil revenues began to increase (Fig. 12.1). However, in 1998, oil prices collapsed again in the wake of the Asian financial crisis. The terms of trade index plateaued at a level of 182.5 during the years 1982-85 before halving to 93.2 1986-90 and then sliding further to 86.4 1991-93. However, the impact on the domestic economy was reduced, compared with the 1970s, because of the shrinkage that had taken place in the share of trade in Saudi GDP. This reduction in the reliance of the economy on trade was not due to any shift in policy towards a more closed trade stance however, but rather it reflected the internalization of more activity as the oil windfalls were invested in the domestic economy. Nevertheless, exports still accounted for more than 40 per cent of GDP through the 1980s and 1990s, so that the negative terms of trade shock after 1985 (mostly compressed within the first few months of 1986) was equivalent to the loss of one-third of non-oil GDP.

12.3. REN T DEPLOYMENT DURIN G TH E 1974-1978 AND 1979-1981 OI L BOOM S 12.3.1. Three prudent features of the Saudi oil windfall absorption The Saudi strategy for the deployment of the oil windfalls exhibits three prudent characteristics. First, the government captured the bulk of the oil rents through the nationalization of production and appropriate taxation. In order to avoid over-rapid absorption, it accumulated a substantial fraction of the revenues as reserves overseas. The IMF calculates that the offshore financial reserves peaked at around $170 billion

High Rent Point Resources: Saudi Arabia

197

in the early-1980s. This is equivalent to just over two-fifths of the oil revenue stream from the 1974-78 and 1979-81 oil booms (SAMA 1998: 11). However, by no means all of the reserves that were accumulated could be drawn upon by the Saudi government should they be required. This is because some of the reserves had been loaned to countries with war-ravaged economies like Iraq, and were not likely to prove redeemable. Informed sources estimated that around $110 billion of the accumulated foreign reserves were potentially accessible. Yet even this smaller sum, if it were assumed to earn annual real interest of 4 per cent, would be sufficient to sustain a budget deficit that was equivalent to 10 per cent of the country's non-oil GDP 1979-81 indefinitely. In addition to accumulating overseas assets, a second sensible feature of the Saudi government's strategy was to reject the rentier option favoured by Kuwait at that time for its smaller (more constrained) economy, and to invest in the diversification of the domestic economy. Development expenditure was clearly identified as a category separate from routine public expenditure so that cutbacks in such expenditure could be made more easily than if no such distinction was made within public expenditure. In addition, the allocation of the windfalls to domestic investment was systematic. The initial emphasis during the Second Development Plan of 1976—80 was upon the expansion of domestic absorptive capacity through the construction of the physical infrastructure of a modern economy. It was expected that the public sector provision of infrastructure would trigger private investment in directly productive activity (Looney 1992). The investment priority shifted in the Third Development Plan of 1980-85 to the diversification of the non-oil tradables sector through a state-led expansion of resource-based industry (RBI), based on joint ventures between stateowned enterprises (SOEs) and MNCs. More emphasis was also given to education and health under the Third Plan: they were allotted 34.6 per cent of development expenditure, with the gains coming at the expense of physical infrastructure. The share of human capital (health and education) in public investment had increased to over 60 per cent in the Fifth Development Plan covering the years 1990-95 (SAMA 1998: 24). Meanwhile, when the oil revenues began to decline through the mid-1980s the private sector was assigned a greater role in diversifying the economy. The third prudent feature of the Saudi windfall deployment strategy was the maintenance of an open economy. As noted earlier, this policy was intended to reinforce the sterilization achieved through the accumulation of offshore reserves in constraining the Dutch disease effects associated with the massive domestic absorption of capital. The economy was kept open not only for imported goods but also for foreign workers and international construction companies in order to ease bottlenecks in the construction sector.

12.3.2. Errors in the windfall absorption strategy Unfortunately, four other characteristics of the Saudi windfall deployment strategy were less prudent and reflected the reliance of the government on paternalism to manage social tensions. First, the strategy overestimated the fraction of the windfall

198

R. M. Auty Table 12.4 . Investment and saving rales, Saudi Arabia 1970-1995 (% GNP)

Fixed investment Domestic saving National saving

1970-73

1974-78

1979-81

1982-85

1986-90

1991-95

11.8 65.9 55.5

23.7 61.1 58.0

23.0 55.7 53.8

27.3 23.5 28.6

21.3 17.9 21.3

19.8 28.3 19.2a

Note: aFor 1991-93 only. Source: World Bank (1998).

that could be absorbed domestically. Overall, more than half of the oil windfall was absorbed domestically during the 1974-78 and 1979-81 oil booms. Estimates made by Gelb and Associates (1988) suggest that a more appropriate level of domestic absorption might have been one-third of the resource windfall. Table 12.4 shows that trends in investment and saving were closely related to the oil booms. The rate of domestic investment quadrupled over the period 1976-79 to plateau at a level of $3 billion per month 1979-81 and it peaked at $4 billion per month in 1983 (Auty 1990). Despite the sterilization efforts, the inflation rate did accelerate so that consumer prices doubled 1974—78, before prices stabilized and the rate of inflation fell below 3 per cent (IMF 1996). This over-rapid absorption was reflected in a dramatic leap in the real exchange rate through the two oil booms. Wood(1988) estimates that, taking 1965 as a base, the real exchange rate almost tripled between 1972 and 1977. Moreover, after easing slightly through 1978—79 the real exchange rate strengthened further after the 1979 oil shock to peak at just over four times its 1972 value in 1982. The second and related flaw in the Saudi strategy was that market discipline was relaxed so that the windfall deployment did not trigger the competitive diversification of the non-oil tradables. By the end of the 1979-81 oil boom, few non-oil tradable activities were competitive in Saudi Arabia and the share of the tradable sectors in nonoil GDP had almost halved compared with 1970-72 (Table 12.2). Worse, agriculture was subsidized, partly in pursuit of the strategic goal of food self-sufficiency and partly to diffuse state largesse towards the largest employment sector. Unfortunately, large farms captured much of the rents, being attracted by wheat prices that were set at five times world levels. Wheat surpluses emerged within five years even as other large farms imported fodder crops to feed to dairy cattle raised in the desert around Riyadh. Wheat production peaked at 4 million tonnes in 1992, almost three times domestic consumption, and absorbed almost $2 billion in subsidies (Financial Times 1996a). Most of the transfers went to six large producers associated with the Saudi royal family. Not surprisingly, the diversion of agricultural subsidies into rent-seeking behaviour failed to stem the urban drift, and the fraction of the population in rural areas declined from 51 per cent in 1970 to 22 per cent within just two decades (World Bank 1998). A third negative feature of the Saudi deployment strategy was that the taxation system was not broadened so that the Saudi government still relied on the oil sector for more than 90 per cent of its revenues into the mid-1980s (Ministry of Planning 1986). For example, import taxes were set at low rates and personal income tax was

High Rent Point Resources: Saudi Arabia

199

not levied, although individuals were expected to contribute 2.5 per cent of their income to the community while value added tax was negligible. Moreover, little attempt was made to recover the costs of government services, many of which were provided free of charge to even the remotest communities despite the high costs that this sometimes entailed. Government subsidies for fuel, water, and electricity rose to 6.7 per cent of GDP by 1980 and doubled their share through the next decade (Askari et al 1998). The final weakness of the windfall deployment strategy is the scale of the expansion in public sector employment, a feature of the 'Santa Glaus' state identified in Chapter 4. Saudi workers displayed a marked preference for service sector jobs, notably in government, leaving both manufacturing jobs and the more menial service tasks to immigrant workers. The Saudi government fed this preference for employment that is shielded from market pressures. It expanded public sector employment on a scale that, relative to GDP, outstripped that of other capital-surplus oil-exporters. By the late1990s, 95 per cent of private sector jobs were held by non-Saudis, mainly other Arabs and South Asians, while within the total workforce, between one-half and two-thirds of the workers were estimated to be non-Saudi (IHT 1999). Eken et al. (1997: 19) observe that the government acted as an 'employer of first resort' for a large part of the workforce, paying wages and benefit packages substantially higher than those in the private sector. Table 12.5 shows that, as a group, the capital-surplus oil-exporters expended a fraction of GDP on public sector wages and salaries that was orders of magnitude more than the capital-deficient oil exporters or the developed economies. Sirageldin and Al-Ebraheem (1999) estimate that even in the late-1990s the oil revenue continued to supply more than 70 per cent of government revenue and that Table 12.5. Government expenditure , twelve countries 7994 (% GDP)

Countries

Current expenditure

Wages and salaries

Subsidies and transfers

Middle East Bahrain Iran Kuwait Oman Qatar Saudi Arabia

27.6 16.2 50.3 37.1 38.1 32.7

15.8 14.5 10.2 14.7 19.4

4.0 2.5 16.9 2.7 0.0 1.0

Other oil producers, Indonesia Mexico Nigeria Venezuela

14.8 13.5 16.3

2.7 4.5 2.5 4.2

2.6 5.4 n.a. 5.3

OECD countries Sweden United States

49.5 22.2

3.0 2.2

25.6 5.9

8.5

Source: Chalk et al. (1997): 13.

9.0

200

R. M. Auty

government spending still employed over 90 per cent of the national workforce. The expenditure by this workforce, in turn, created demand within the private sector that was met almost wholly by using expatriate labour whose own expenditures then provided additional demand for both oil revenue (public goods) and expatriate labour (private goods and services). These multipliers meant that if oil revenue declined and foreign borrowing was constrained, then contraction in public expenditure would have a negative multiplier effect that would magnify the impact of reduced public expenditure on the economy as a whole. The multipliers also indicate that much 'non-oil' revenue accrued from taxes levied on income from transfers (such as public sector wages) that were fed by the once-for-all depletion of the oil rents. The basic need for sustainable economic development is to increase revenues from 'a tax system levied in non-oil production that engages the national labour force.` (Sirageldin and Al-Ebraheem 1999: 10, italics added).

12.4. ADJUSTMENT TO REDUCED OIL RENTS 12.4.1. Tardy fiscal adjustments An abrupt collapse in economic growth creates political risks (see, for example, David 1999), but when oil revenues fell, the windfall deployment strategy had buffered the Saudi government in three ways. The three 'cushions' were to draw upon the overseas financial reserves, to decelerate the infrastructure expansion and to repatriate immigrant workers and companies. However, the scale of the real decline in oil revenues also required timely adjustments in the real exchange rate and in fiscal policy, the latter including expenditure cuts and measures to diversify taxation. The Saudi adjustment to the sharp fall in oil revenues is consistent with the thesis that resource rents relax economic discipline in the absence of a developmental state or consensual democracy (Fig. 12.1). It was tardy because it initially relied heavily on the financial reserves that had been accumulated during the oil booms to offset the impact. Nevertheless, public expenditure peaked at $84 billion in the fiscal year 1981-82 and it was cut by one-third over the next four years (Table 12.6). The bulk of the cuts were concentrated on public construction. Recurrent expenditure such as subsidies proved much more difficult to rein in, however. The annual operating and maintenance costs of the new infrastructure system absorbed one-eighth of total public expenditure in the mid-1980s and transfers accounted for a further one-third (Auty 1990). Although public expenditure fell significantly over the years 1982-86 (Table 12.6), the oil revenues fell faster so that they covered barely three-quarters of public expenditure, leaving a deficit of $14 billion in 1985, some 13 per cent of total GDP. Subsequent efforts to close the fiscal gap were incremental and experienced a substantial setback as a consequence of the 1990-91 Gulf War. The Saudi national accounts contain gaps and inconsistencies in the data that preclude the accurate calculation of the impact on the foreign reserves of the recurring budget deficits after 1981. This lack of transparency is part of the neglect by the Saudi government of the institutional capital of a competitive market economy. For example,

High Rent Point Resources: Saudi Arabia

201

Table 12.6. Trends in government revenue, Saudi Arabia 1976-1997 (billion riyals) Year

Oil revenue

Total revenue

Expenditure

1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1991a 1992 1993 1994 1995 1996 1997

121.2 114.0 115.1 189.3 319.8 328.6 186.0 145.1 121.3 88.4 42.5 67.4 48.4 75.9 n.a. 127.1 106.0 95.5 105.7 136.0 160.0

136.0 130.7 131.5 211.2 348.1 368.0 246.2 206.4 171.5 133.6 76.5 103.8 84.6 114.6 316.6 165.4 141.4 129.0 146.5 179.1 205.5

128.3 138.0 148.0 188.4 236.6 284.6 244.9 230.2 216.4 184.0 137.4 173.5 134.9 149.5 476.8 232.5 205.5 163.8 173.9 198.1 221.3

Gap 7.7 -7.4 -16.5 22.8 111.5 83.4

1.3 -23.8 -44.9 -50.4 -60.9 ^69.7 -50.3 -34.9 -160.2 -67.1 -64.1 -34.8 -27.4 -19.0 -15.8

Note: Two-year fiscal period. Source: SAMA (1998): 21.

the banks were owned by the princes and secured high returns (15 per cent in the late-1990s) on safe loans that included loans to the government. Or again, the privatization of subsidized utilities and RBI joint ventures was postponed, denying timely revenue to the government. Nevertheless, it is possible to simulate the impact of the budget deficits if two assumptions are made. First, as discussed earlier, it may be assumed that the size of the accessible reserves that the Saudi government could draw on through the mediumterm, totalled $110 billion by 1983. The second assumption is that these financial reserves could secure an annual average real rate of return of 8 per cent. Using SAMA (1998) data for the budget deficit, the simulation suggests that the deficits were on a sufficient scale to have exhausted the financial reserves by the mid-1990s (Fig. 12.2). In fact, the assumptions are too generous because informed sources suggest that although the reserves were $7 billion in late-1998 (Financial Times 1999), domestic public debt had reached more than 85 per cent of GDP. The domestic debt had become a sizable burden on the government's current expenditure (Financial Times 1998) as well as a threat to the weak domestic banking system. Oil still provided over 70 per cent of government revenues and current expenditure (including domestic debt service) and public salaries already absorbed in excess of three-quarters of the budget

R. M. Auty

202

Figure 12.2 . Budget deficits and estimated financial reserves, Saudi Arabia 1983-1996 Source: SAMA (1998).

(some estimates say it takes 90 per cent). The residual was insufficient to maintain the infrastructure, let alone to expand and modernize it.

12.4.2. Inadequate depreciation of the real exchange rate If the fiscal adjustment to lower oil rents was inadequate, government policy towards the exchange rate was also neither as swift nor as bold as the situation required. The Saudi currency had been officially linked to the SDR in 1975 but in practice it shadowed the dollar. When oil prices fell sharply and the riyal began to depreciate against the strengthening dollar the government pegged the riyal to the dollar at 3.745 riyals in June 1986. The net effect of these policy shifts was initially to maintain the real effective exchange rate at the level it had reached at the height of the 1979-81 boom, whereas some depreciation would have been more appropriate. Although an accelerating depreciation (linked to the weakening dollar) then commenced that caused the real exchange rate to lose 40 per cent of its value by 1987, the decline slowed. By 1992 the real effective exchange rate was only half of its peak 1981 level (IMF 1996), but this was still double the level of the early-1970s, however, when oil revenues had been creating budget surpluses as opposed to the deficits of the 1980s and 1990s. Although the non-oil tradable sector sustained positive growth through the second half of the 1980s and into the 1990s (Table 12.7), the rate was too slow. Its share of non-oil GDP, at 26 per cent in the mid-1990s, remained some two-thirds of its expected size (Table 12.3). Moreover, growth within the manufacturing sector slowed to less than 5 per cent per annum 1990—95 compared with a growth rate in excess of

High Rent Point Resources: Saudi Arabia

203

Table 12.7. Sectoral growth rates, Saudi Arabia 1971-1995 (%/Year) Category

1971-75

1976-80

1980-85

1986-90

1990-95

GDP Oil Non-oil Agriculture Other mining Oil refining Petrochemicals Other manufacture Utilities Construction Trade Transport Finance Other services Government

13.4 13.8 11.7 3.6 21.1 n.a. n.a. 11.4 10.9 18.6 13.9 17.0 8.2 7.1 7.8

9.3 4.9 15.0 7.0 11.0 n.a. n.a. 15.4 22.3 16.4 22.8 19.4 14.1 11.8 6.5

-5.8 -14.6 5.1 8.7 5.7 n.a. n.a. 14.1 24.0

0.1 1.0 -0.8 13.8 -1.4 5.3 55.5 -4.6 5.7 -6.7 -1.5 -1.9 -4.8 0.5 1.5

4.1 9.0 2.1 3.1 4.9 7.5 -5.6 4.4 4.5 0.0 1.3 1.6 1.8 0.7 2.8

—14 8.8 7.1 13.1 7.9 5.8

Source: SAMA (1998): 19.

7 per cent 1985-90 (UNIDO 1997). In fact, an index of intra-industry trade that is taken as a measure of the degree of diversification and technical sophistication within the manufacturing sector, is much lower in Saudi Arabia than would be expected for a country with its per capita income. Havrylyshyn and Kunzel (1997: 6) calculate the index for Saudi Arabia as 0.096 in the early-1990s. This is one-quarter the level that the regression line predicts for a country with its per capita income. It compares, for example, with levels of 0.250 for the Arab countries as a whole, 0.519 for Mercorsur, a region of slightly lower per capita income, and 0.886 for the EU. Nevertheless, the Saudi intra-industry trade index had doubled 1984-94, due largely to a rapid expansion in petrochemicals in the 1980s by SABIC, the 70 per cent state-owned petrochemical company. Petrochemicals accounted for 56 per cent of Saudi manufacturing output in 1995 compared with 12 per cent each for non-metallic minerals and metals. The country's share of global petrochemicals was above 5 per cent. However, input subsidies were pervasive and European firms complained to the WTO that the price of gas purchased by SABIC, at US$0.50 per mcf, was barely two-thirds the lowest international rate. The WTO also objected to subsidies to wheat producers, tariff increases (set at 12 per cent for most goods but rising to 20 per cent if imports compete with domestic products), and weak safeguards for intellectual property rights (Economist 1997). Domestic private investors preferred to expand into real estate rather than invest in the manufacturing sector (Financial Times 1998). Consequently, the strategy of diversifying the economy by creating opportunities for private sector investors to process the intermediate products of the state-led RBI intermediate products made little progress. Looney (1992) confirms that, contrary to the intention, the private sector response to the massive public sector investment in infrastructure was not

204

R. M. Auty

strongly positive. This is because when the infrastructure was expanded the real exchange rate appreciated so that investors found overseas returns more attractive. In fact, Looney finds that domestic private investment responds more strongly to shortterm measures, such as the provision of subsidies for wheat production. GDP growth in Saudi Arabia like that in most Arab countries was extensive, relying basically on the accumulation of inputs of capital and labour rather than upon productivity growth. Bisat et al. (1997) estimate that, assuming the capital share was 0.3, the total factor productivity (TFP) growth was —4.6 per cent per annum 1974-85 and —0.1 per cent per annum 1986-96. If the national accounts method is used, then the capital share more than doubles for Saudi Arabia so that the rate of TFP growth is even more disappointing. However, the dominance of the oil sector largely accounts for the negative TFP growth of the economy. If the economy is sub-divided into oil and non-oil subsectors, TFP was positive in the non-oil sector and grew at 5.9 per cent annually over the years 1974-85. However, it turned negative at —1.45 annually 1986-96. Expenditure on education by Saudi Arabia is one of the highest in the world and exceeded 10 per cent of GDP in the mid-1980s. Nevertheless, the results have been disappointing. For example, both primary and secondary enrolment are barely twothirds of what they should be for a country of its level of development. Moreover, much of the national workforce is ill-equipped to thrive in a market economy. One explanation for this disappointing outcome is the catch-up effect as human capital makes up for previous deficiencies in expenditure when incomes were much lower. A second reason is that the expenditure was biased in favour of tertiary education and against rural provision and provision for women. However, Shafik (1996) concurs with the findings in Chapter 4 and argues that the main reason for the disappointing performance is an inappropriate labour demand due to the structure of the economy. This arises because of the bias towards capital-intensive production that creates too few jobs and encourages the government to provide cosseted public sector jobs. Private sector employers complain that Saudi graduates lack analytical skills because of the emphasis on religious teaching in the school curriculum and that they lack flexibility in their work practices (Financial Times 1996b). The demand for labour is overwhelmingly determined by government expenditure. Sirageldin and El-Ebraheem (1999: 19) note: 'For the past three decades, almost all policy recommendations and program efforts have focused on the supply side of human resource development, that is, education and training, without adequate attention to the role of demand. These efforts have succeeded on the quantity side of education and training but apparently failed on the quality and motivation side. The latter are essential for competitiveness in the present global environment. The present structure of labour demand is to a large degree responsible for this apparent failure.' The bulk of the national workforce is insulated from market competition so that 'citizen commitment and motivation in the market place are not adequately tested.' Elsewhere, Sirageldin (1999) argues that a generation has grown up within a dependency culture that is ill-suited to cope with the expected increasing exposure to market pressures. That generation lacks, ' . . . achievement motivation, vision of opportunities, sense of

High Rent Point Resources: Saudi Arabia

205

discipline, work ethic commitment, and self-esteem...' (Ibid. 3) that one generation passes on to another in a motivated and progressive society. Environmental and natural resource accounting (EARA) provides a further test of the Saudi oil windfall deployment. The genuine saving rate for Saudi Arabia consistently shows a very high level of negative saving (see Table 8.5). In other words, the high rates of investment in produced and human capital made by the Saudi government were nowhere near sufficient to compensate for the annual depletion of oil and natural gas. Moreover, the Saudi figure excludes a charge for the unsustainable consumption of fossil water resources by subsidized wheat farmers that threatens to exhaust such reserves between 2010 and 2020 (Financial Times 1992).

12.5. CONCLUSION Although capital surplus oil-exporting countries like Saudi Arabia have tended to maintain more open economies, they too experienced growth collapses in the mid1980s. Therefore, although trade policy closure is strongly associated with growth collapses, it is not a necessary cause. Rather, more fundamentally, growth collapses occur because the resource rents encourage a relaxation of market discipline that compounds distortions within the economy that are then politically difficult to correct. Paternalistic governments like that of Saudi Arabia play Santa Claus with the rents so that even large investments in education leave most nationals ill-equipped to cope without state support. In the case of Saudi Arabia, the government allocated too high a fraction of the oil windfall to domestic absorption. The rapid domestic absorption of the oil rents caused the real exchange rate to appreciate four-fold in the 1970s so that very little non-oil activity was competitive. Although RBI was successfully established, that sub-sector created few jobs so that the government relied heavily for jobs upon infrastructure construction, subsidized farming and above all, the public sector. In doing so it failed to convert sufficient of the natural capital (oil) into alternative growth-sustaining investment: the private sector is stunted and, despite high expenditure on education, the national workforce is ill-equipped to thrive in a market economy. When oil prices fell (because the high oil rents of 1974—81 were ephemeral, being determined by politics rather than by market fundamentals) the government made cuts in public expenditure and a depreciation of the real exchange rate that were, however, too small in each case. Consequently, the Saudi economy diversified too slowly and remained over-dependent on oil in the late-1990s. By then, however, the government had exhausted its accessible overseas financial reserves and accumulated domestic debt that was double a prudent level. Further confirmation that growth was not sustainable comes from environmental accounting data for 1970-94. Meanwhile, the growth collapse damaged the mechanisms used by the paternalistic government to assuage social tensions. More specifically, real per capita GNP, which had risen eight-fold over the period 1970-80, more than halved from the peak level of the early-1980s (World Bank 1998). Although welfare gains were sustained, such as improved life expectancy (70 years in 1995 compared with 48 in 1970) and expanded

206

R. M. Auty

access to education, their continuation relied upon the state largesse that the high oil rents had provided. Data on income inequality are not available, but the windfall deployment opened a wide gap between the well-connected elite of 1 per cent of the population with wealth estimated at $400 billion and the small farmers of the Southeast along with those entering the workforce as public sector opportunities are reined back. The worker/dependency ratio remains very high at around 0.79 (see Table 2.4). As conditions deteriorate for the poorest in Saudi society, the rent deployment strategy may also prove to have been socially unsustainable (David 1999). Indeed,the principal flaw in the Saudi deployment of its oil windfalls may turn out to be the failure of paternalism to build public social capital (see Table 9.4).

REFERENCES Askari, H., M. Bazzari and W. Tyler (1998), Policies and economic potential in the countries of the Gulf Co-operation Council, in Shafik, N. (ed.), Economic Challenges facing Middle Eastern and North African Countries: Alternative Futures, Basingstoke: Macmillan, 225-55. Auty, R. M. (1990), Resource-Based Industrialization: Sowing the Oil in Eight Exporting Countries, Oxford: Clarendon Press. Bisat, A., M. A. El-Erian and T. Helbling (1997), 'Growth, investment and saving in the Arab economies', IMF Working Paper 97/85, Washington DC: International Monetary Fund. Chalk, N. A., M. A. El-Erian, S. J. Fennell, A. P. Kireyev and J. F. Wilson (1997), 'Kuwait: from reconstruction to accumulation for future generations', IMF Occasional Paper 150, Washington DC: International Monetary Fund. David, S. R. (1999), 'Saving America from the coming civil wars', Foreign Affairs, 78/1, 103-16. Economist (1997), 'WTO: Enter the Saudis?', The Economist (May 17). Eken, S., T. Helbling and A. Mazarei (1997), 'Fiscal policy and growth in the Middle East and North Africa', IMF Working Paper 97/101, Washington DC: International Monetary Fund. Financial Times (1992), 'Saudi Arabia: A Survey', Financial Times (30 January). Financial Times (1996a), 'Saudi debt move gets mixed response', Financial Times (4 April). Financial Times (1996b), 'Giving young Saudis a stake in the future', Financial Times (12 July). Financial Times (1998), 'Sinking oil price rattles Saudi Arabia's shaky structures', Financial Times (3 April). Financial Times (1999), 'Editorial: Saudi Troubles', Financial Times (12 March). Gelb, A. H. and Associates (1988), Oil Windfalls: Blessing or Curse? New York: Oxford University Press. Havrylyshyn, O. and P. Kunzel (1997), 'Intra-industry trade in Arab countries: an indicator of potential competitiveness', IMF Working Paper 97/47, Washington DC: International Monetary Fund. IHT (1999), 'Fallen from oil plenty, Saudis learn to scrimp', International Herald Tribune, (22 March). IMF (1996), International Financial Statistics Yearbook 1996, Washington DC: International Monetary Fund. Looney, R. E. (1992), 'Real or illusory growth in an oil-based economy: government expenditures and private sector investment in Saudi Arabia', World Development, 20, 1367-75.

High Rent Point Resources: Saudi Arabia

207

Ministry of Planning (1986), Kingdom of Saudi Arabia: Achievements of the Development Plans 1970-1986, Riyadh: Ministry of Planning Press. Nankani, G. T. (1979), 'Development problems of mineral exporting countries', World Bank Staff Working Paper 354, Washington DC: World Bank. Sachs, J. D. and A. Warner (1995), 'Economic reform and the process of global integration', Brookings Papers on Economic Activity, 1, 1—118. SAMA (1998), Kingdom of Saudi Arabia; Thirty-Fourth Annual Report, Research and Statistics Department, Riyadh: Saudi Arabian Monetary Agency. Shafik, N. (1996), 'Big spending, small returns: the paradox of human resource development in the Middle East', Working Paper 9602, Cairo: Egyptian Center for Economic Studies. Sirageldin, I. (1999), Toward an integrated population-development policy in MENA: with special reference to the Arab countries, paper presented at the Fourth Conference on Population Challenges and Economic Growth: MENA, 14 April, Washington DC. Sirageldin, I. and Y. Al-Ebraheem (1999), 'Budget deficit, resource gap and human resource development in oil economies', in I. Sirageldin (ed.), Population and Development in the Middle East and North Africa: Challenges for the Twenty-first Century, Working Paper, Baltimore, MD: Johns Hopkins University. Syrquin, M. and H. B. Chenery (1989), 'Patterns of development, 1950 to 1983', World Bank Discussion Paper 41, Washington DC: World Bank. UNIDO (1997), Industrial Development: Global Report 1997, New York: Oxford University Press. Wood, A. (1988), 'Global trends in real exchange rates, 1960 to 1984', World Bank Discussion Paper 35, Washington DC: World Bank. World Bank (1998), World Development Indicators 1998, Washington DC: World Bank. World Bank (1999), Wealth Estimates and Genuine Savings, Environment Department, Washington DC: World Bank.

13 Large Resource-Abundant Countries Squander their Size Advantage: Mexico and Argentina RICHARD M. AUTY

1 3 . 1 . RESOURC E A B U N D A N C E , COUNTR Y SIZE , A N D GROWTH COLLAPSE S The four countries selected so far to analyse the basic permutations of natural resource-abundance and political state are all relatively small. This chapter turns to the large resource-abundant countries. All else being equal, a large resource-rich country should be more able than a small resource-rich country to escape a growth collapse. Chapter 2 provided two basic reasons for this. First, large economies have more scope for economic diversification because their greater geographical extent provides a wider range of natural resource exports and their potentially large domestic market facilitates the capture of economies of scale so that manufacturing diversifies earlier. Second, large economies are more self-sufficient so that the domestic impact of an external shock is automatically reduced. In addition, Table 8.3 suggests that the natural resource rents of large resource-rich countries may be not much greater, relative to GDP, than those of the resource-poor countries. Yet, econometric analysis by Perkins and Syrquin (1989) covering the period 195083 shows that large countries did not perform markedly better than small ones. Perkins and Syrquin (1989) do report, however, that large countries exhibit relatively high levels of autarky (self-sufficiency), especially the resource-abundant ones. This characteristic is at the heart of the reason why the larger resource-abundant countries squandered their potential size advantage. This chapter explains why the large resource-abundant countries failed to avoid a growth collapse with reference to case studies of two of the largest resource-abundant countries, Argentina and Mexico (Table 13.1). Natural resources played a significant role in propelling both Argentina and Mexico to mid-income status by the 1970s. In fact, Argentina had already harnessed the resources of the Pampas to become one of the richer countries in the world during the first golden age of economic growth (see Chapter 6). A century later, Argentina remains amply endowed with fertile land relative to its population (Table 13.1) and

Country Size Advantage: Mexico and Argentina

209

Table 13.1. Socioeconomic preconditions in Argentina, Mexico, and Korea Index Resource endowment Land area (000 krn2) Population 1970 (millions) Cropland/capita (ha) GDP 1970 ($ billion) GDP/head ($US) Sociopolitical Literacy (%) Life expectancy (yrs) Population growth ( ) Rural population (%) InCOMERATIO,RICHFIFHPOO PoliticALSTA Economic Trade stance Growth of GDP/head 1970-73 (%) GDI 1970-73 (% GDP) Foreign debt 1970-73 (% GDP) Exports 1970-73 (% GDP) Agriculture 1970 (% GDP) Manufacturing 1970 (% GDP)

Argentina

Mexico

Korea

2,787 24.0 1.13 30.0 1,250

1,909 52.8 0.45 38.3 725

99 32.2 0.07 8.4 260

93 67 2 22 11.4 FO

74 62 3 41 15.1 FO

90 60 3 41 6.9 DAB

Closed -1.5 23.3 21.5 7 10 32

Closed 3.8 19.0 15.0 7 12 22

Open 6.9 23.7 27.3 16 29 23

Note: aFOS = Factional Oligopolistic; DAB = Developmental Autonomous Benevolent. Source: World Bank (1997); WRI (1994). it has considerable mineral potential that is only just being developed (Financial Times 1999). Mexico is rich in minerals, although it too was also well-endowed with cropland until rapid population growth eroded that advantage during the second golden age of economic growth (1950-73). However, the confirmation of vast reserves of oil and gas in the 1970s restored Mexico to natural resource abundance. Indeed, oil rents reached around one-fifth of Mexican GDP in the early-1980s, a figure that was several times the combined natural resource rents of Argentina by that time. Meanwhile, by the 1970s, both Argentina and Mexico had already established a wide range of manufacturing despite their late start compared with the North Atlantic core economies. Their prospects for coping with the negative shocks of the 1970s therefore do appear to have been superior to those of small, less diversified resource-abundant countries. Yet, neither country avoided a growth collapse (Table 13.2). The chapter is structured as follows. The next section, Section 13.2, establishes that external shocks do not explain the growth collapses. Section 13.3 then shows why the trade policies of the two large Latin American countries came to be so strongly inwardoriented by the early-1970s. Section 13.4 then demonstrates why both governments mismanaged the shocks and caused growth collapses. The conclusions are summarized in Section 13.5.

210

R. M. Auty Table 13.2. Per capita GDP, selected countries 1900-1992 (1990 US$)

Country

1900

1913

1950

1973

1980

1992

Argentina Mexico Ghana South Africa Indonesia South Korea Australia USA

2,756 1,157 462 n.a. 745 850 4,299 4,096

3,797 1,467 648 1,45 917 948 5,505 5,307

4,987 2,085 1,193 2,251 874 876 7,218 9,573

7,970 4,189 1,260 3,844 1,538 2,840 12,485 16,607

8,245 5,254 1041 4,114 1,870 4,103 13,805 18,270

7,616 5,112 1,007 3,451 2,749 10,010 16,229 21,366

Source: Maddison (1995): 23-24, 194-206. Table 13.3 . Scale of external shocks, four countries (% GNP) Country Year

Argentina Mexico Korea Taiwan

s Trad e Expor t Tota price effect volum e shoc effect shoc

1974-76 0.8 1979-81 1.4 1974-76 1.3 1979-81 -1.0 1974-76 10.8 1979-81 7.5 1974-76 6.9 1979-81 8.3

0.1 0.1 0.3 0.6 1.7 3.7 2.4 5.2

l trad e Interes t Tota l k rat e shock combine k

0.9 1.5 1.6 -0.5 12.5 11.2 9.3 13.5

— 1.4 — 2.3 — 2.6 — 0.1

d

0.9 2.9 1.6 1.9 12.5 13.8 9.3 13.6

Source:Balassa(1984).

1 3 . 2 . M O D E S T E X T E R N A L SHOCK S 1 9 7 3 - 1 9 8 1 The negative external shocks that impacted the economies of Mexico and Argentina were several orders of magnitude smaller than those experienced by the resource-poor economies of Taiwan and Korea. Yet the latter countries successfully rode out their shocks whereas Mexico and Argentina experienced growth collapses. Table 13.3 measures the scale of the shocks in all four countries. It compares the actual price and output trends through the 1973 and 1978 oil shocks with a counter-factual that is based upon the continuation of pre-shock trends in export volume (column 4) and also in export and import prices (column 3). Table 13.3 also measures the economic impact of the sharp rise in international interest rates that occurred in the early-1980s by comparing those rates to the 1977 levels (column 6). Argentina might have been expected to manage the oil shocks relatively easily because the country imported negligible amounts of oil and experienced only moderately unfavourable trends in the prices of its principal exports during the

Country Size Advantage: Mexico and Argentina

211

immediate post-shock years. Mexico experienced slightly worse external conditions 1974-76 but its 1979-81 trade shock was positive. The oil price shift of 1979-81 coincided with a rapid expansion in Mexican production to boost the hydrocarbon rents by an extra 7.7 per cent of GDP annually (World Bank 1998). A windfall on this scale had less potential to destabilize the economy than those of, say, Indonesia or Nigeria, that were three times larger (Gelb and Associates 1988). Moreover, as a late participant in an oil boom, the Mexican government could also learn from the mistakes of other countries, notably Venezuela whose windfall was similar in size. Consequently, external conditions were relatively propitious in both the large Latin American countries. Balassa (1984) decomposes the policy response to the external shocks into four components (foreign borrowing, export expansion, import compression, and slower GDP growth). Table 13.4 compares the contribution of each policy component as a percentage of the total shock. It shows that foreign borrowing dominated the Latin American responses to the first shock and also figured prominently in regard to the second shock. Nevertheless, the ratio of foreign debt to GDP in the Latin American countries ended up not much higher than that of Korea. This is because the HCI Big Push of Korea had lifted that country's foreign debt to 27.7 per cent of GDP by 1981, compared with 30.9 per cent for Mexico and 31.6 per cent for Argentina. However, the two large Latin American economies were unable to service their debt when real interest rates rose in 1981 because their tradable sectors were too small and lacked resilience. The resource-poor East Asian countries coped better with much larger negative external shocks. South Korea avoided a growth collapse, despite experiencing a strong negative shock, by rapidly expanding its exports to pay for its higher oil imports and higher debt service charges (Table 13.4). As a result, GDP growth quickly resumed after a sharp economic contraction in 1980. The government of Taiwan was still more prudent: its external debt ratio was barely one-third that of the other three economies. It drew only modestly on foreign borrowing in response to the first shock and compressed imports and slowed the economy in response to the second shock, so that Table 13.4 . Policy responses to external shocks, four countries (% shock) Country Year s Foreig Argentina 1974-76 1979-81 Mexico 1974-76 1979-91 Korea 1974-76 1979-81 Taiwan 1974-76 1979-81

154.8 322.0 177.6 315.4 -32.3 -15.9 27.7 -36.8

n loans Expor t growt h Impor t cut Slowe r GD P growth -141.6 -68.1 -59.0 562.4 55.7 -4.8 -12.0 26.0

40.5 -201.3 -19.3 -904.6 89.9 13.1 16.4 89.8

Note: Sign reversed for Mexico when it experienced a positive shock. Source: Balassa (1984).

46.2 47.3 0.8 -73.3 -13.3 107.6 68.0 21.0

212

R. M. Auty

economic growth resumed after slowing down briefly. The next section explains why the two Latin American countries had small export sectors and large foreign debts.

13.3. TH E R E A S O N S FO R T R A D E POLIC Y CLOSUR E Both Mexico and Argentina adopted closed trade policies after the Second World War, swiftly in the case of Argentina and more gradually in the case of Mexico. The Syrquin and Chenery (1989: 53) 'norms' provide an index of trade policy closure. They show that in countries with a per capita income of $2,500 (in 1980 US$), close to that of Mexico and Argentina, exports comprised 36 per cent of GDP for small resource-deficient countries, 26 per cent for small primary exporters, 18 per cent for the large resource-deficient countries but only 13 per cent of GDP for large primary product exporting countries. In fact, Argentina and Mexico had both reduced their exports to a mere 7 per cent of GDP by 1970 (Table 13.1). This is half the level of Australia, a resource-abundant economy of comparable size to Mexico and Argentina at that time, and it is a similar level to that of the US, an economy twenty to thirty times bigger. Why did Argentina and Mexico opt for such high levels of autarky, and why did they persist with the policy for so long after its adverse effects became apparent?

13.3.1. High autarky and the Argentinean growth trajectory The government of Argentina began to intervene more intensively in the economy in the 1930s in response to short-run macroeconomic imbalances. It doubled average tariffs on light manufactures to 20 per cent, tolerated modest fiscal deficits and installed a dual exchange rate. The apparent success of these measures, along with the wartime accumulation of international reserves (analogous to a natural resource rent windfall), encouraged politicians to adopt a much more autarkic trade policy in the late-1940s. However, the aims of the intervention now became reduced dependence on the global economy and to redistribute income away from the landed interests towards favoured urban groups (Diaz Alejandro 1971). These aims seemed feasible for a large developing country because it was believed that their domestic markets could capture the economies of scale and sustain diversification into heavy and chemical industry (HCI). As with Ghana (Chapter 10), impatience with resource-driven growth led to forced industrialization that consolidated a factional oligarchic or predatory political state. Post-war Argentine governments used the autarkic policy to transfer natural resource rents to protected 'infant' industries in which the favoured urban groups (business, unions, and the bureaucracy) were prominent. The government captured and redistributed the rents by setting the domestic and export prices of agricultural goods below world levels, and by raising the prices of manufactured goods by taxing imports (Mundlak and Domenech 1995). The transfers were therefore at the expense of consumers as well as of primary product exporters.

Country Size Advantage: Mexico and Argentina

213

The autarkic policy pushed Argentina into the staple trap. First, it retarded competitive industrialization, despite the fact that by 1930 'natural' import substitution under an open trade policy had raised already the share of industry to 29 per cent of GDP, with manufacturing alone to 19 per cent of GDP (Lai 1989; Veganzones and Winograd 1997). In contrast, the HCI that was encouraged by the autarkic policy took decades to mature. For example, twenty years after the autarkic policy was introduced, the prices of Argentina's slow-maturing manufactured goods were still one to two orders of magnitude above world prices (Diaz Alejandro 1970). One unexpected problem was that the thresholds of minimum viable factory size in HCI were constantly rising so that optimum capacity outstripped domestic demand. For this and other reasons, entry into HCI therefore proved premature and it evoked requests for the continuation of import protection and subsidies (Auty 1994). The incentives were invariably extended so that the 'infant' HCI had little incentive to grow up and become internationally competitive and the policy was effectively captured by the vested interests. The corollary of infant industry support was the stagnation of the hitherto dynamic agricultural sector. In the 1940s, farm prices wrere set below world levels and then the real exchange rate appreciated due to inflationary public expenditure under Peron. Efforts to revive agriculture were unsuccessful and in the early1950s a new government intensified import substitution. Sturzenegger (1991) estimates that the repression of the agricultural sector in the Pampas region alone transferred 40 per cent of its potential revenue (at world prices) to the government over the period 1960-85 (Table 13.5). This is equivalent to around 4 per cent of GDP annually. As might be expected from these developments, the tradable sector of the economy shrank and failed to diversify. The fall in the share of agriculture from 18 to 12 per cent of Argentina's GDP during the 1940s was not offset by an expansion in manufacturing. Instead, the decline wras compensated by services whose share of GDP increased to 58 per cent (Veganzones and Winograd 1997: 27). By 1970 the share of agriculture had fallen below 10 per cent of GDP and although manufacturing had risen by then to onethird of GDP, HCI accounted for more than half the sector's value added and much Table 13.5 . Consequences of a neutral policy regime, Argentina 1960-1985

Period

1961-65 1966-70 1971-75 1976-80 1981-85

Transfers to agriculture (% Agricultural GDP)

Extra exports ($ billion/year)

Crops

Beef

Total

8.5

32.7 30.7 12.3 12.2 30.1

41.2 42.2 27.7 31.0 61.2

11.5 15.4 18.8 31.1

Source: Sturzenegger (1991): 29-31.

1.32 1.59 2.07 3.06 6.94

214

R. M. Auty

manufacturing remained uncompetitive and, effectively non-tradable (Berlinski 1992). The slow maturation of manufacturing along with reduced farm output compressed exports into barely half the expected fraction of GDP. Interestingly, Sturzenegger (1991: 29) estimates that exports could have doubled their share in GDP if prices had not been distorted. The Pampas alone could have provided the additional exports required to ease the emerging foreign exchange constraint on economic growth. This distorted economic structure was associated with a rate of investment that failed to sustain the 25 per cent or more of GDP that was achieved by successful open economies, resource-abundant and resource-poor alike. Ironically, the Argentine investment rate had occasionally touched 25 per cent of GDP during the first golden age of economic growth 1870-1913. But three-fifths of this investment was foreign saving. This made for high investment volatility: for example, during the 1930s total investment collapsed and fluctuated between 10 and 15 per cent of GDP (Veganzones and Winograd 1997). Although the Argentinean rate of investment did briefly return to 25 per cent of GDP in the late-1970s, that level was not sustained. More critically, however, as the staple trap predicts, the efficiency of investment declined. In the context of Table 2.3, the incremental capital output ratio (ICOR) for Argentina averaged a disappointing 5.9 during the period 1950-70; it then deteriorated to a dismal 8.5 in the 1970s and turned negative during the 1980s (Chisari et al. 1996). Human capital accumulation also failed to sustain its momentum of the early twentieth century because, as argued in Chapter 4, protected industrialization exerts weaker demand for skilled labour than industrialization. The early adoption of universal primary education had played an important role in the rapid diffusion of skills in Argentina at the beginning of the twentieth century, but the wholesale development of secondary and tertiary education lagged until the 1950s and the 1960s, respectively. More specifically, the average years of schooling reached 3.3 by 1910. It then grew relatively slowly to reach 4.9 years in 1950, 6.3 years in 1970, and 8.4 in 1990. Moreover, the initial high quality of primary education was not maintained and by the late-1980s the quality of secondary and tertiary education was deficient (Veganzones and Winograd 1997: 90-95). Finally, HCI is capital-intensive so that employment in manufacturing grew very slowly. Successive Argentine governments regarded urban unemployment as a threat. They reacted by expanding public sector jobs and maintaining an expensive system of social security entitlements. Although these measures reduced income inequality, they did so at the expense of worker incentives and fiscal prudence. By the eve of the oil shocks, the economic distortions that had cumulated under twenty-five years of forced (as opposed to the pre-war voluntary) infant industry protection were causing trade and fiscal deficits that elicited stop-go macro policies that slowed economic growth (Table 13.6). Yet, despite the worsening adverse impacts on social welfare, reform was blocked for a further two decades. As in Ghana, so in Argentina, in the absence of a consensual democracy or developmental autonomous benevolent state, the asymmetry of political power pitted a large number of scattered farmers against a union-backed urban industrial lobby (Sturzenegger 1991).

Country Size Advantage: Mexico and Argentina

215

Table 13.6. Per capita GDP growth rate trends 1870-1994 (%/year) Economy

1870-1910

1910-50

1950-60

1960-70

1970-80

1980-90

1990-94

Latin America Argentina Mexico

1.5 2.7 0.5

1.9 0.7 1.0

2.0 2.0 2.9

2.1 2.8 3.1

1.5 1.4 3.3

-1.0 -2.3 -0.4

2.1 4.9 0.7

Source: Maddison (1995): 202-03.

13.3.2. The lagging Mexican growth trajectory In contrast to Argentina's sudden shift towards autarky with the election of Peron in the mid-1940s, Mexico drifted into that policy over the 1950s and 1960s. One consequence of the lag was that Mexico sustained a brisker rate of economic growth than Argentina did between 1950 and 1970 (Table 13.6). The Mexican rate of investment rose from 16 to 21 per cent of GDP between 1958 and 1970 and the efficiency of investment exceeded that of Argentina (Table 13.1). Some observers (Buffie and Krause 1989) attribute this creditable outcome to the policy of Stable Development that was adopted in 1956. Stable Development centred upon a fixed exchange rate with the US dollar that exerted fiscal and monetary discipline. This link meant that an expansion of the fiscal deficit needed to be funded by forced loans based on high commercial bank reserve ratios (31 per cent), instead of by money creation. However, Buffie and Krause ignore sectoral trends that reveal an underlying deterioration in the competitiveness of the economy. Reynolds (1970) correctly attributes Mexico's solid economic growth rate prior to the 1970s to the diffusion of green revolution techniques 1945-65 (Wellhausen 1976). However, the economy-wide impact flagged as, first, the number of large farms on which the new technology could be applied was exhausted and subsequently the impact of negative effective protection for agriculture under autarky began to bite. More specifically, agricultural growth slowed from 7.8 per cent 1947-56 to a still robust 4.8 per cent 1957-66, but it was barely positive 1967-76 (Fitzgerald 1985). By 1970, the share of agriculture in GDP had shrunk to half the Chenery and Syrquin norms for a country of Mexico's size and level of development (Auty 1994: 165). A large trade deficit occurred in 1970 despite a sharp improvement in Mexico's terms of trade (up 16 per cent 1970-75 compared with 1964-69). The deficit reflected a cumulative appreciation (strengthening) of the real exchange rate by 25 per cent because Mexican inflation had outstripped that of its dominant trading partner the United States (Cardoso and Levy 1988). Mexican investment remained well below 25 per cent of GDP and, as in Argentina, autarky failed to stimulate competitive manufacturing. This is so even though Mexican manufacturing expanded to 22 per cent of GDP by 1970, one-fifth above the Chenery and Syrquin norm, and manufactured goods provided one-third of exports compared with only one-seventh in Argentina. However, Mexican exports of manufactured goods were barely one-tenth of sectoral output (just 2.2 per cent of GDP) and they were subsidized (Jaspersen 1997: 79): few subsectors of Mexican manufacturing were

216

R. M. Auty

internationally competitive. The protected domestic market accounted for 86 per cent of the growth in manufacturing output in the 1960s, compared with 11 per cent for import substitution and only 4 per cent for exports (Kubo et al. 1986). By the early-1970s, effective rates of protection exceeded 77 per cent on durable and capital goods, and in the mid-1970s import restrictions rose sharply to cover more than 90 per cent of imports. Protected sectors like steel and auto assembly exhibited maturation rates of several decades. As with Argentina, the corollary of the assistance for manufacturing was negative effective protection for primary products. Mexico lagged even Argentina's disappointing record in skill accumulation: for example, secondary school enrolment was barely two-thirds that of Argentina. As elsewhere, the capital-intensive HCI generated limited employment and provoked government concern over social tension. However, unlike Argentina where the government expanded social entitlements and stabilized income inequality—albeit at the expense of fiscal imbalances—the income differential in Mexico was higher (Table 13.1) and rising. The income share of the poorest quintile fell steadily in Mexico from 4.4 per cent in 1957 to 1.9 per cent two decades later (Haggard 1990). Yet the political economy that had evolved in Mexico to manage these social tensions was a factional oligarchy like that in Argentina. In the case of Mexico, the long-serving PRI government balanced the interests of industrialists and urban unions (many in the expanding public sector). It largely ignored the rural poor, many of whom were of Indian descent. Summarizing, both large countries squandered their potential size advantages in the pursuit of autarky. Mexico moved gradually towards that policy through the 1960s whereas Argentina abruptly adopted it in 1945. As the staple trap model predicts, the closed trade policy nurtured a large non-competitive and capital-intensive manufacturing sector at the expense of the primary/resource sector. The process of competitive industrialization was retarded and exports were compressed into an exceptionally low fraction of GDP. Yet the vested interests created by infant industry protection prevented reform and also sustained the factional oligarchic state. The next section shows how both governments mismanaged external shocks that at worst were mildly negative.

1 3 . 4 . E X P L A I N I N G TH E GROWT H COLLAPSE S

13.4.1. The Mexican growth collapse: The role of oil resources Mexico had yet to confirm its large oil reserves when the 1973 oil shock occurred. More expensive oil imports came on top of the growing burden of imports of food as agricultural dynamism waned, and also of inputs for the protected and 'non-tradable' manufacturing sector. When oil import costs rose in 1973-74 the left-leaning Mexican government, conscious of social tensions linked to high and rising income inequality (see Table 9.4), turned to foreign loans in order to sustain economic growth. The foreign loans helped to finance a public sector populist boom that like all such booms quickly proved unsustainable (Sachs 1989). In 1975 an economic stabilization programme and a real depreciation of the exchange rate were unavoidable (Table 13.7).

Country Size Advantage: Mexico and Argentina

217

Table 13.7. Trends in real exchange rate, three countries 1961-1984 (1965= 100) Country

1961-63

1964-66 1967-69 1970-72 1973-75 1976-78

1979-81 1982-84

Argentina Mexico Korea

110.5 98.9 133.6

127.8 100.5 109.7

141.4 99.7 177.9

87.3 99.3 132.0

79.0 94.2 135.4

92.2 96.0 131.5

87.5 82.9 161.6

81.6 83.3 178.4

Source: Wood (1988).

Table 13.8. Mexican oil revenues and public finances 1971-1990 (% GDP) Category

1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990

Revenue

Oil

Expenditure

Capital

Current

Interest

18.4 18.7 20.2 21.1 23.2 23.8 24.8 25.9 26.7 26.9 26.3 27.8 31.7 31.3 30.4 29.4 29.5 28.7 27.6 27.5

3.0 2.8 2.6 3.4 3.3 3.3 3.8 4.5 5.6 7.3 8.1 11.1 16.1 15.1 13.3 11.4 11.9 10.0 8.7 9.0

20.5 22.9 25.8 27.0 31.9 32.0 30.0 31.4 33.0 33.5 38.7 41.8 39.0 37.7 37.5 42.5 43.7 38.4 34.1 33.4

4.3 5.7 7.0 7.2 8.6 8.0 7.6 8.7 9.8 9.6 11.9 8.9 6.3 5.7 5.1 4.6 4.8 3.6 3.2 4.0

14.6 15.4 17.0 17.9 21.0 20.7 19.3 19.5 19.5 19.8 21.9 24.9 20.6 20.3 21.0 21.4 19.4 19.3 18.1 19.5

1.6 1.8 1.8 1.9 2.3 3.3 3.1 3.2 3.7 4.1 4.9 8.0 12.1 11.7 11.3 16.4 19.6 16.5 12.9 9.9

Deficit 2.1 4.2 5.6 5.9 8.7 8.2 5.4 5.5 6.3 6.6 12.4 13.0 7.3 6.4 7.1 13.1 14.2 9.7 6.5 5.9

Source: 1971-80 Bazdresch and Levy (1991): 234; 1981-90 Loser and Kalter (1992): 5.

The incoming Mexican president was more conservative and under strong economic constraints he espoused orthodox policies in 1976, only to abandon them after the announcement of large hydrocarbon reserves in 1977, echoing Bolivian experience a few years earlier. The oil rents appeared to remove the constraints on sustained economic growth that had been imposed by recurrent trade and fiscal deficits under the autarkic policy. The Mexican government assumed that the 1979 oil shock heralded a permanent improvement in the country's terms of trade. Therefore, instead of prudently saving part of the oil windfall, the oil resources were used as collateral to attract additional foreign capital on a scale that amplified the oil windfall (Gelb and Associates 1988). An acceleration in public spending (Table 13.8) intensified the 197981 oil boom and, as in Saudi Arabia, Bolivia, and also Malaysia, it outstripped domestic

218

R. M. Auty

absorption capacity. But whereas Malaysia was quickly able to grow out of the subsequent recession, Mexico was trapped. Gelb and Associates (1988) estimate that four-fifths of the real increase in domestic revenue absorption into the Mexican economy 1979-81 went into investment, mostly into long-gestation public sector projects like oil exploration and HCI. This briefly lifted Mexican investment above 25 per cent of GDP. However, investment efficiencydeteriorated and the IGOR averaged 10 through the 1980s (World Bank 1998). Meanwhile, the remaining oil rents were used to boost consumption by the deferral of personal tax increases and through lower import prices that were caused by a sharp real appreciation of the exchange rate. Finally, some of the rent was used to subsidize producers in the face of an appreciation of the real exchange rate (Table 13.7). The oil rents thereby merely postponed still further the competitive industrialization of the Mexican economy. Pleas for protection by manufacturers were granted, mostly through import bans. The rate of effective protection turned negative for mining, but it became positive (18 per cent) for agriculture as well as intermediate goods (43 per cent) and durable consumer goods (128 per cent). Important industries like auto assembly and steel now had even less incentive to mature. The World Bank concluded that Mexican automobile assembly was a failing infant industry with a domestic resource cost index of 1.51 in 1982 (Auty 1994: 86). The relaxation of fiscal policy caused the public sector deficit to balloon during the 1979-81 oil boom (Table 13.8), while the external debt approached $100 billion and the debt service ratio reached 45 per cent of export earnings (Balassa 1985). A sharp rise in international interest rates in 1981 triggered a loss of confidence in the economy. The incoming government that took office in 1982 like many of its predecessors therefore found it necessary to adopt stabilization measures. Conditions did not improve, however, because although public expenditure was cut as subsidies were removed and public investment was slashed by two-thirds, rising debt service payments offset the savings (Table 13.8). High debt service payments compressed domestic absorption to 93 per cent of GDP 1983-86, even though the trade balance had turned strongly positive because non-oil exports expanded in response to a one-third depreciation of the real exchange rate 1981-83. Investment fell to barely 16 per cent of GDP and per capita incomes shrank by 10 per cent. The genuine saving coefficient in Table 8.5 shows that Mexican development was not sustainable through much of the 1980s. Environmental damage alone was estimated to depress true saving by 8 per cent of GDP in 1985 (Van Tongeren et al. 1991).

13.4.2. The Argentinean growth collapse: External shocks and policy response The Argentinean government also reacted to the 1973 oil shock with a populist boom that, as in Mexico, proved unsustainable. It bought an initial spurt in per capita GDP growth at the expense of a subsequent large contraction (Sachs 1989). The fiscal deficit doubled to 11.4 per cent of GNP 1974-76 and inflation exceeded 400 per cent in 1976. In 1977 a new government launched a 'big bang' stabilization and reform package that

Country Size Advantage: Mexico and Argentina

219

failed. It adopted high interest rates to curb inflation, but also opened the capital market, causing the exchange rate to become strongly overvalued (Table 13.7). This aborted an expected surge in resource-based exports. Although, as noted earlier, investment touched 25 per cent of GDP in Argentina in the late-1970s, that level was not sustained. The principal adjustment to the 1979 oil shock was through further foreign borrowing (Table 13.4) and, like Mexico, the external debt reached almost $100 billion by the early-1980s. Debt service absorbed half the merchandise export earnings, and took some 3.5 per cent of Argentinean savings annually over the years 1982-90 (Chisari et al. 1996). The failure of the 1977 reforms meant that orthodox policies had yielded no political benefits to compensate for the social costs of the deflation that they had imposed. Yet the autarkic policy was also discredited because the bloated public sector had become a drag on the economy. For example, the state-owned enterprises generated more than 20 per cent of GDP and absorbed subsidies up to 2.6 per cent of GDP annually (Cortes Conde 1992). A further 5 per cent of GDP was transferred to the slowmaturing manufacturing sector in the form of incentives. The apparent failure of both orthodox and structuralist strategies caused Argentinean economic policy to drift through the 1980s: budget deficits averaged 9 per cent of GDP, inflation averaged 450 per cent, and per capita GDP contracted by 2.3 per cent annually (Table 13.6).

13.5. CONCLUSIO N Argentina and Mexico confirm how protectionist policies designed to force industrialization sapped economic resilience and squandered the country size advantage. In contrast to many smaller resource-abundant countries, notably in sub-Saharan Africa, the large Latin American countries were deemed more credit-worthy and were therefore able to postpone a growth collapse (and also reform) by borrowing recycled petro-dollars. The forced industrialization policy benefited key urban supporters of the government and consolidated a factional oligarchic political state. Reform was blocked in both countries because union-backed urban-industrial interests captured economic policy and proved too formidable for the fragmented rural interests. Mexico especially, shows clearly how resource-abundance postpones reform. In 1977, when large hydrocarbon discoveries were announced, the economic reforms that were adopted during the crisis of 1975 to open the economy were reversed. Both countries fell into the staple trap and distorted their economies so that competitive industrialization stalled. The protected manufacturing sector failed to mature and generate exports while it was also capital-intensive and created few jobs. Meanwhile, the resource sector was repressed (Pampas farming in Argentina) and/or unable to expand fast enough (oil in Mexico) so that exports shrank to a meagre 7 per cent of GDP in both countries, about half what might be expected in large undistorted economies. Investment efficiency declined and the rate of economic growth became erratic and slowed. Recourse to foreign borrowing in order to sustain economic growth merely deferred reform and rendered recovery more protracted because the tradeable sector was too small to service the debt.

220

R. M. Auty

Yet even more extreme closure of the economies of India and China (with exports only 1—2 per cent of GDP) did prevent growth collapses in those countries. This is parth' because such extreme closure did reduce the sensitivity of the economy to tradeinduced instability. More importantly, the two large Asian economies adjusted to the external shocks by relying far less on foreign borrowing that they might be unable to service (Auty 1994). Even so, both China and India abandoned autarky as its growthretarding effects became apparent. Interestingly, such reform proved far easier under Deng's developmental autonomous benevolent state in China than under the more resource-abundant and polarized democracy of India, which moved only slowly towards a consensus for reform. Finally, the large resource-abundant countries do appear to have recovered faster from their growth collapses than small countries (see Table 2.3). This reflects their greater capacity to diversify in response to the new opportunities created by trade liberalization (Auty 1999). However, just how quickly the recovery occurs depends upon the legacy of the industrial structure from infant industry protection, being faster with higher levels of MNC and domestic conglomerate ownership, and also upon the capacity to sustain prudent economic policies. The type of political state conditions the latter: Argentina led Mexico towards consensual democracy during the 1990s, and its economy recovered the fastest.

REFERENCES Auty, R. M. (1994), Economic Development and Industrial Policy: Korea, Brazil, Mexico, India and China, London: Mansell. Auty, R. M. (1999), 'Large Resource-Abundant Countries Squander their Size Advantage for Rapid Economic Growth and Reform: Argentina and Mexico', UNU/ WIDER Working Papers, Helsinki: UNU/WIDER. Balassa, B. (1984), 'Adjustment policies in developing countries: a reassessment', World Development, 12, 955-72. Balassa, B. (1985), 'Adjusting to external shocks: the newly-industrializing developing countries 1974-76 and 1979-81', Weltwirtschafiliches Archiv, 121, 116-41. Bazdresch, C. and S. Levy (1991), 'Populism and economic policy in Mexico, 1970-82', in R. Dornbusch and S. Edwards (cds.), The Macroeconomics of Populism in Latin America, Chicago II: University of Chicago Press: 223-62. Berlinski, J. (1992), 'Trade policies and industrialization', in S. Teitel (ed.), Towards a New Development Strategy for Latin America, Baltimore: lADB/Johns Hopkins University Press: 323-36. Buffie, E. F. and A. S. Krause (1989), 'Mexico 1956—86: from stabilising development to the debt crisis', in J. D. Sachs (ed.), Developing Country Debt and the World Economy, Chicago, II: University of Chicago Press: 141—68. Cardoso, E. and S. Levy (1988), 'Mexico', in R. Dornbusch and F. L. C. H. Helmers (eds.), The Open Economy, New York: Oxford University Press, New York, NY: 348-69. Chisari, O. O., J. Fanelli and R. Frenkel (1996), 'Argentina: growth resumption, sustainability and environment', World Development, 24, 277-40.

Country Size Advantage: Mexico and Argentina

221

Cortes Conde, R. (1992), 'Growth and stagnation in Argentina', in S. Teitel (ed.), Towards a New Development Strategy for Latin America, Baltimore: lADB/Johns Hopkins University Press: 337-53. Diaz Alejandro, C. (1970), Essays in the Economic History of the Argentine Republic, New Haven CT: Yale University Press. Diaz Alejandro, C. (1971), 'The Argentine state and economic growth: a historical review', in G. Ranis (ed.), Government and Economic Development, New Haven CT: Yale University Press: 216-52. Financial Times (1999), 'Success of Argentine mining on test', Financial Times (8 January). Fitzgerald, E. V. K. (1985), 'The financial constraint on relative autonomy: the state and capital accumulation in Mexico 1940-82', in C. Anglade and C. Fortin (eds.), The State and Capital Accumulation in Latin America: Volume 1, London: Macmillan: 210-40. Gelb, A. H. and Associates (1988), Oil Windfalls: Blessing or Curse? New York: Oxford University Press. Haggard, S. (1990), Pathways from the Periphery, London: Cornell University Press. Jaspersen, F. (1997), 'The East Asian miracle and the Latin American consensus: can the twain ever meet?', in N. Birdsall and F. Jaspersen (eds.), Pathways to Growth: Comparing East Asia and Latin America, Washington DC: Inter-American Development Bank: 57-91. Kubo, Y., J. de Melo and S. Robinson (1986), 'Trade strategies and growth episodes', in H. B. Chenery, S. Robinson and M. Syrquin (eds.), Industrialization and Growth: A Comparative Study, New York: Oxford University Press: 148-87. Lal, D. (1989), 'The political economy of industrialisation in primary product exporting economies: some cautionary tales', in N. Islam (ed.), The Balance Between Industry and Agriculture in Economic Development, Vol. 5, London: IEA: 279—314. Loser, C. and E. Kalter (1992), 'Mexico: the strategy to achieve sustained economic growth', IMF Occasional Paper 99, Washington DC: IMF. Maddison, A. (1995), Monitoring the World Economy 1820-1992, Paris: OECD. Mundlak, Y. and R. Domenech (1995), 'Agricultural growth in Argentina', in J. W. Mellor (ed.), Agriculture on the Road to Industrialization, Baltimore MD: Johns Hopkins University Press: 175-95. Perkins, D. H. and M. Syrquin (1989), 'Large countries: the influence of size', in H. Chenery and T. N. Srinivasan (eds.), Handbook of Development Economics. Vol. 2, Amsterdam: NorthHolland: 1691-753. Reynolds, C. W. (1970), The Mexican Economy: Twentieth Century Structure and Growth, New Haven CT: Yale University Press. Sachs, J. D. (1989), 'Social conflict and populist policies in Latin America', NBER Working Paper 2697, Cambridge MA: NBER. Sturzenegger, A. C. (1991), 'Argentina', in A. O. Krueger, M. Schiff and A. Valdes (eds.), The Political Economy of Agricultural Pricing Policy: Vol. 1, Latin America, Baltimore MD: Johns Hopkins University Press: 15-51. Syrquin, M. and H. B. Chenery (1989), 'Patterns of development, 1950 to 1989', World Bank Discussion Paper 41, Washington DC: World Bank. Van Tongeren, J., S. Scheinfest, E. Lutz, M. Gomez and M. Guillen (1991), 'Integrated environmental and economic accounting: Mexico', Environment Working Paper 50, Washington DC: World Bank. Veganzones, M-A. and C. Winograd (1997), Argentina in the 20th Century: An Account of LongAwaited Growth, Paris: OECD.

222

R. M. Auty

Wellhausen, E. J. (1976), 'The agriculture of Mexico', Scientific American 235/3: 128-50. Wood, A. (1988), 'Global trends in real exchange rates, 1960 to 1984', World Bank Discussion Paper 35, Washington DC: World Bank. World Bank (1997), World Development Report 1997, Washington DC: World Bank. World Bank (1998), World Development Indicators 1998, Washington DC: World Bank. WRI (1994), World Resources 1994, New York: Oxford University Press.

PART V

LESSONS FOR POLICY REFORM

This page intentionally left blank

14 Reforming a Small Resource-Rich Developing Market Economy: Costa Rica GUSTAVO B A R B O Z A AND JOSE C O R D E R O

1 4 . 1 . INTRODUCTIO N The great majority of resource-abundant countries experienced a growth collapse in the 1970s or 1980s from which recovery has been protracted at best. Part V draws upon case studies to analyse the hesitant recovery and derive policy implications. This study suggests that although the economic policy advice of the international financial institutions is basically sound, it pays insufficient attention to initial conditions. Earlier chapters have demonstrated significant differences in the capacity of countries to handle shocks whether the shocks emanate from dramatic shifts in the terms of trade or from social unrest or abrupt policy change. Among the more important of these initial conditions are natural capital, social capital, and the political state. At first sight, Costa Rica seems relatively well-endowed in all three of these initial conditions. It is a resource-abundant country, albeit a small one. However, it lies in the mostly densely-settled region of Latin America and rapid population growth has propelled it from a condition of natural resource abundance to one of growing resource paucity within fifty years. By 1970 the ratio of cropland/ capita for Costa Rica was well below the average for small resource-abundant countries, whereas in 1920 it had been relatively high at 1.24 hectares (compare Table 1.1). Nevertheless, Costa Rica also has substantial forests that are particularly rich in biodiversity. Moreover, the country still has twice the natural capital per head of the next most well-endowed Central American republic (Table 14.1). Costa Rica also differs from its Central American neighbours and most other midincome resource-abundant countries because it is widely regarded as being unusually well-endowed with social and human capital. Table 14.2 shows that in 1997 Costa Rica ranked thirty-fourth on the United Nations Human Development Index. More specifically, some 93 per cent of the population had access to drinking water and The authors acknowledge the comments of Ariabelle Ulate and the participants in a workshop at WIDER, Helsinki.

226

G. Barboza and J. Cordero Table 14.1 . Natural capital estimates, Costa Rica 1994

Country Natura l % of total capital ($ per capita) fPasture Cro p Timbe r Non-timbe r Protecte d Subsoi l land lan d resource s fores t area s asset s Costa Rica El Salvador Guatemala Honduras Nicaragua

7,860 1,150 1,720 3,380 3,690

19 22 18 12 15

72 77 54 47 57

2 1 10 24 16

1 0 6 6 10

5 0 9 7 2

0 0 4 3 0

Source: Kunte et al. (1998).

Table 14.2 . Human development index 1995, selected developing countries Country HD

Singapore South Korea Chile Bahamas Costa Rica Argentina Uruguay Panama Mexico Brasil El Salvador Honduras Nicaragua

I Classificatio n Rea by HDI GD

0.896 0.894 0.893 0.893 0.889 0.888 0.885 0.868 0.855 0.809 0.604 0.573 0.547

28 30 31 32 34 36 38 45 49 62 114 119 126

l per capita Classificatio n by P (ppp in $) rea l per capita GD P minus classification by HDI 22,604 11,594 9,930 15,738 5,969 8,498 6,854 6,258 6,769 5,928 2,610 1,977 1,837

-21 6 9 -3 28 11 14 14 5 1 -2 7 3

Source: UN (1998).

electricity, life expectancy at birth was 77 years, and the literacy rate was 95 per cent. The distribution of income is relatively equitable for resource-rich Latin America (see Table 9.4). As argued in Chapter 4, a government that is sensitive to broader social welfare can create a virtuous cycle between equity and economic growth if its policies promote demand by the poor for education, along with the efficient accumulation of other forms of capital. Moreover, as elaborated below, Costa Rica evolved towards a consensual democracy in the mid-twentieth century. Yet, the per capita income of Costa Rica is lower than might be expected (Table 14.2, column 5). This is because Costa Rica distorted its economy and experienced a growth

Developing Market Economy: Costa Rica

227

(%/yr) Table 14.3. Per capita: GDP growth and reform, selected countries 1960-1997

Country

1960-73

1973-85

1985-97

Malaysia

4.10 0.17 3.07 n.a. 3.21 1.21 7.61 3.19 2.23 3.30

3.85 -2.77 -0.02

4.95 1.48 1.71 0.97 1.25 5.90 -1.01 -0.68 1.98 0.43

Ghana Costa Rica

Bolivia Jamaica Chile Saudi Arabia

South Africa Argentina Mexico

n.a. -2.99 0.95 -2.25 0.00 -0.85 1.94

Source: World Bank (1999).

collapse, despite the enhanced capacity for crisis management that its equitable income distribution and long-established consensual democracy confer. A mismanaged coffee boom in the late-1970s precipitated the growth collapse. The government first postponed reform and then pursued only modest reforms that led to a somewhat lacklustre economic recovery (Table 14.3). One explanation for the rather disappointing economic outcome of economic reform is that the social capital assembled by Costa Rican governments has generated Olson effects as cohesive social groups pursued their own interests at the expense of society as a whole (see Chapter 5). A second explanation, drawn from Chapters 2 and 8, is that structural change is more difficult for smaller resource-abundant countries to achieve than it is for larger ones because small countries possess fewer options for diversifying into alternative primary products and manufacturing. This chapter attempts to explain why Costa Rica's potentially favourable endowment of social capital led to tardy reform and hesitant economic recovery. The next section examines the preconditions to the growth collapse of the early-1980s in more detail. Section 14.3 then analyses the growth collapse and the political constraints upon the nature of the reform package that was adopted in response to it. Section 14.4 evaluates the impact of the economic reforms that were initiated in 1985 after the successful stabilization of the economy. Finally, Section 14.5 considers the implications of the Costa Rican experience for other small resource-abundant countries that have experienced a growth collapse.

14.2. INITIA L CONDITION S 14.2.1. The evolving political economy The natural resource endowment of Costa Rica initially proved less attractive to Iberian settlers than that of adjacent areas. Large-scale Spanish colonization was deterred and a small, homogeneous, and spatially concentrated pattern of settlement

228

G. Barboza and J. Cordero

was spawned that fostered one of the most equitable land distributions within Latin America. For three centuries the central valley of Costa Rica supported a small community of yeoman farmers that numbered barely 50,000 at independence and occupied a mere 2 per cent of the land area. Despite the prosperity that coffee exports brought in the nineteenth century, the population of Costa Rica was still only 300,000 in 1900. Thereafter, a rapid expansion of population commenced that lifted the total to 2 million by 1976 (Rottenberg 1993) and 3.65 million in 1998 (World Bank 1999). The population remains geographically concentrated, however, with three-fifths living in the Central Valley, which is the most fertile part of the country but accounts for only 8 per cent of the land area. According to Rottenberg (1993), this concentration of population, together with a moderately equitable income distribution, relatively homogeneous ethnicity and predominantly smallholder farming system assisted in consensus formation and the shaping of the distinctive political economy of Costa Rica. Although BulmerThomas (1987) expresses some scepticism about assigning the key role to yeoman coffee production, he offers little evidence to support his view. Rottenberg (1993), however, does marshal the evidence. Moreover, his conclusion is in line with that of Hirschman (1977) as well as with the findings of export base theory about the potential for smallholder production to generate favourable socioeconomic linkages (Baldwin 1956). The Costa Rican government encouraged export-led growth and ploughed taxes back into infrastructure. By the late-nineteenth century, the government was already distancing itself from the coffee interests, in contrast to the more oligarchic governments elsewhere in the region. Costa Rica experienced a steady shift towards a consensual democracy over the years 1911—48 that displayed a mounting concern for social welfare. Shafer (1994) notes that, in contrast to the elite elsewhere in Central America, such as El Salvador, the Costa Rican elite made little attempt to wrest control of land. The elite benefited from synergies with the peasant farmers who supplied coffee of a superior quality to their 'beneficios' (wealthy coffee merchants) as well as seasonal labour on their estates. This compact between estates and yeoman farmers was formalized in 1933 in the depths of the Depression by regulations to control the maximum profit that the beneficios could make from coffee bean purchases. In effect, the relationship between the beneficios and smaller farmers is a forerunner of the highly successful nucleus plantation of Malaysia (Graham and Floering 1984). In the 1940s a struggle emerged between the long-serving liberal tradition and a social democrat challenge. The struggle ended in compromise after a disputed presidential election and a two month civil war of 1948. Thereafter, the economic growth rate quickened and per capita GDP expanded at almost 3.2 per cent per annum 195073, more than twice the rate during the period 1920-50. The coffee boom between 1940—55 consolidated the country's prosperity and, perhaps, provided a false sense of security and bred the assumption that the state could successfully manage the industrial transition. The political consensus favoured the expansion of social entitlements and paid insufficient attention to how the economy would be able to fund

Developing Market Economy: Costa Rica

229

them. Although this policy led to marked improvements in welfare, the entitlements outstripped the capacity of the economy to sustain them. A consensual political state is therefore not a sufficient condition to prevent the relaxation of market discipline.

14.2.2. Adoption of inward-oriented economic policies Before the 1960s, agriculture and related activities dominated the structure of production. But fluctuations in international commodity prices caused recurrent crises in the external sector that led to a decision to force industrialization. Legislation was approved in the late-1950s for a policy of infant industry protection. This strategy was launched in 1962 and given a boost in 1963 when Costa Rica joined the Central American Common Market (CACM), a market protected by an external common tariff. The industrial policy aimed to accelerate economic diversification and thereby expand foreign exchange earnings and reduce dependence on volatile commodities. The Costa Rican government simultaneously promoted employment through an expansion of public services and the establishment of state-owned enterprises (SOEs) in aviation, cement, sugar milling, and the stock exchange. Initially, the industrial policy appeared to be very effective. The level of investment rose from 18 per cent of GDP in the early-1960s to reach 26 per cent of GDP during the 1977 coffee boom. Real GDP per capita grew at an average annual rate of 3.4 per cent during 1962-69 and at 3.6 per cent during the 1970-79 period. Costa Rica experienced an improving standard of living, more equitable income distribution and an extended social security network. These advances in welfare were complemented by achievements in politics, including the strengthening of the Supreme Electoral Tribune, the elimination of the army by constitutional mandate in 1949 and the forging of a consensual democracy. Rising incomes and social cohesion helped Costa Ricans to enjoy social and political tranquillity in marked contrast to neighbouring countries. However, consistent with the staple trap model, competitive industrialization was not achieved and the economy became more vulnerable to external shocks. Slowmaturing manufacturing did not generate the foreign exchange required to import the raw materials and capital goods that it needed. Therefore, the agricultural sector continued to be the principal source of foreign exchange revenue, with close to 45 per cent of the exports coming from coffee and bananas alone. The rents were squeezed from the sector to sustain the growing social entitlements, irrespective of the impact on rural incentives (Rottenberg 1993). The expanding role of the state distorted prices throughout the economy due to a range of interventions intended to ensure domestic food supplies and dampen inflation. One consequence was that capital was underpriced due to efforts to reduce interest rates to favoured borrowers. This combined with high labour costs to discourage job-creation. Social security charges were equivalent to 26 per cent of wages by the late-1970s. Overspending by the government and private agents created a widening gap between investment and domestic saving. By 1980, central government spending absorbed 19 per cent of GDP compared with 6 per cent in 1950.

230

G. Barboza and J. Cordero

As Chapter 8 notes, overspending on such a scale generates powerful hysteresis effects that prolong high spending levels and lead to macroeconomic imbalances and instability. Despite evidence of waning economic resilience in the 1970s, successive Costa Rican governments sought to extend the role of the state still further. Minimum wage levels were increased and welfare coverage was extended from salaried employees to all citizens. This implied the redistribution of 2 per cent of GDP to the poorest fifth of the population by 1978 (Gindling and Terrell 1995). A state holding agency, CODESA, was created in 1972 to promote economic expansion and diversification. The complexity of the welfare system meant that much expenditure went on maintaining the bureaucracy rather than on delivering state assistance. For example, CODESA recorded losses 1975-83 that in every one of those years exceeded 25 per cent of its sales (Rottenberg 1993). By 1980, public investment had reached two-fifths of total investment and was crowding out private sector borrowers. The coffee boom of the late-1970s and access to petro-dollars eased pressure for reform. The 1976-80 coffee boom generated additional rents equivalent to 4.75 per cent of GDP annually (Gonzalez-Vega 1999). Whereas the private sector responded to the boom as if it were temporary by saving the windfall, the public sector did not. In fact, the public sector captured much of the surplus by borrowing domestic savings to expand the role of the state and through inflation tax. Much of the windfall therefore went into consumption that was unsustainable when coffee prices declined. The government responded to declining export revenues by borrowing rather than risk a confrontation with the public sector workforce, which was highly unionized and difficult to restructure (Shafer 1994). The deterioration in the current account deficit is traced in Table 14.4. The external deficit was financed mostly by means of direct foreign investment (DFI) and long term official capital (LTOC) in the 1970s. In effect, the social reforms were not financed by domestic economic activity, but by foreign

Table 14.4 . Current account deficit financing, Costa Rica selected periods (%)

Years DFI

/ LTOC deficit defici

1962-73 1974-78 1978-80 1981-82 1983-85 1986-90 1991-94 1995-97

34.57 24.03 9.26 13.03 34.41 47.21 114.30 289.92

/ STOC / LTPC* / STPC t defici t defici t defici

35.32 54.80 56.55 -0.22 224.33 -90.92 75.63 -112.91

4.93 0.68 8.96 118.59 -103.64 135.77 -31.44 -24.62

14.02 20.95 7.91 -3.12 -14.49 -0.37 -6.23 16.99

/ RedRes / Deficit / t defici t GD P

13.57 9.43 2.12 -47.30 -34.99 -27.05 20.19 -42.13

-10.82 -10.92 18.44 -20.17 -70.00 -34.56 -151.87 -51.51

DFI: Direct Foreign Investment; LTOC: Long-Term Official Capital; STOC: Short-Term Official Capital; LTPC*: Long-Term Private Capital (Net of DFI); STPC: Short-Term Private Capital; RedRes: Reduction of International Monetary Reserves.

6.93 10.58 12.58 13.16 5.50 5.33 4.46 2.22

Developing Market Economy: Costa Rica

231

saving that proved impossible to service when real interest rates turned strongly positive in the early-1980s.

14.3. CRISIS AND REFORMS In 1981, a severe economic crisis was triggered in Costa Rica by a combination of deteriorating terms of trade, a grossly overvalued exchange rate, excess government spending, military conflict in the other Central American nations, and higher international interest rates. From 1979 to 1983, per capita GDP fell at an average annual rate of 3.3 per cent. In 1982 the unemployment rate reached 9.4 per cent and the inflation rate, as measured by the consumer price index, jumped to 90 per cent. Bythen, foreign debt exceeded 100 per cent of GDP and debt service absorbed one-half of export earnings (Nelson 1990). A change of government in 1982 ushered in stabilization policies that controlled inflation, the trade gap, and budget deficit. A start was subsequently made on economic restructuring, although the thoroughness of the reform and its speed were both compromised by the need to maintain the consensus (Saborio 1990). Ulate (1999) identifies the four objectives of reform as to: rationalize the public sector, increase domestic savings, diversify production towards non-traditional exports, and minimize the costs of the adjustment to low income groups. The initial priority was to stabilize the economy by balancing the budget and reducing public employment. However, although the fiscal deficit was brought under control, it remained high and absorbed one-third of government revenue in debt service. The reduction of employment was also slow because laying off public sector workers usually entailed paying large compensation. SOEs engaged in producing goods were mostly privatized in order to raise funds to balance the fiscal deficit, but other SOEs were not privatized. Moreover, although social reforms went deep, the state created new organizations to assist the poor during the restructuring of the economy. Under the infant industry policy, the share of domestic savings in investment declined sharply from around four-fifths of the total to less than one-third. Efforts to reverse the decline and to revive domestic saving centred on maintaining positive real interest rates for savers and relaxing regulations to allow the expansion of private banks that were virtually absent previously. Trade reform focused on dismantling the system of protection that had raised effective protection rates above 100 per cent in some cases and created wide diversity in rates due to their ad hoc implementation. A commitment was made to translate non-tariff trade barriers into tariff equivalent ad valorem barriers, and import tariffs were reduced. Meanwhile, a more flexible foreign exchange rate regime was introduced in place of the overvalued exchange rate of the late-1970s. Export incentives were also enhanced through tax breaks, export zones, credit lines, and streamlined paper work, while the Central American Integration Treaty was revised. The next section assesses the impact of these reforms on four main areas, namely the structure of the economy, employment, productivity growth, and natural resource use.

232

G. Barboza and J. Cordero 14.4. EVALUATIN G TH E REFORM S

14.4.1. Production structure by economic activity The Syrquin and Chenery norms suggest that for a country with the size and per capita income of Costa Rica, agriculture and industry would be of similar importance, as was indeed the case. However, the service sector was larger than expected, indicating Dutch disease effects. Table 14.5 suggests that structural change has been rather small throughout the period 1976-97, with agriculture around 20 per cent of GDP, industry at 21 per cent, and the commercial sector 30 per cent of GDP. But within the service sector, the public sector shrank whereas transportation, electricity, and communications increased their share from 8 per cent in 1976 to 14 per cent in 1997. The shrinkage resulted from freezing the number of public sector employees wherever possible. This adjustment was slow and did not completely resolve the problem of excess government size. Overall, structural change was modest so that despite the reforms Costa Rica remained heavily reliant upon agriculture for production and foreign exchange. Nevertheless, the trade reforms and a sustained depreciation of the exchange rate increased the share of exports in GDP, which almost doubled to 49 per cent between 1980 and 1998 (World Bank 1999). Exports of manufactured goods remained fairly steady at one-quarter of all exports, so that the overall dependence on agricultural exports remained, with the main change being some diversification of those exports. However, a rapid expansion of manufacturing exports began after 1995 that lifted their share to 56 per cent by 1998. This was associated with a sharp rise in foreign direct investment from the early-1990s that averaged more than 6 per cent of GDP 1995-97. Although food and beverages continued to account for half of manufactured exports, a sharp rise also began in textiles and, more importantly, more high-tech manufacturing. This long lag in industrial restructuring echoes that of Chile and other small economies (Auty and Mikesll 1998). Chile commenced reform in the mid-1970s and quickly lost around half its erstwhile protected manufacturing capacity. It took more than a decade for compensating export-competitive activity to emerge. Table 14.5. Sectoral composition of GDP, Costa Rica 1976-1997 (%)

Agriculture Industry Construction Basic services* Commerce** Personnel services + government

1976

1978

1980

1985

1990

1995

1996

1997

20.2 21.3 5.9 8.0 30.0 14.7

19.0 22.5 5.6 8.4 30.6 14.0

18.0 22.0 6.3 9.3 30.1 14.3

19.2 21.9 4.5 10.3 30.0 14.1

19.3 21.5 4.1 11.8 30.4 13.0

18.7 22.0 3.5 13.5 30.6 11.7

18.7 21.3 3.2 13.9 31.0 11.9

18.0 21.5 3.6 14.0 31.2 11.7

*Includes electricity, gas, water, transportation, storage and communications. **Includes commerce, finances and real estates. Source: Banco Central de Costa Rica.

Developing Market Economy: Costa Rica

233

14.4.2. Employment by economic activity Table 14.6 indicates the relative changes in the level of employment generated by the reforms. The largest change is the reduction of agricultural employment that fell from 35 per cent of the total labour force in 1976 to only 21 per cent by 1997. However, real wages rose by 10 per cent in the agricultural sector relative to real wages in the nontradable sector during the years 1980-89 and rural poverty declined from 17 to 3 per cent of the rural population (Morley 1995). Overall, the largest increase in employment occurred within commerce whose workforce rose from 16.5 per cent in 1976 to 24.4 per cent by 1997 (Table 14.6). Industry expanded its share of the labour force from 14.7 per cent to almost 16 per cent over the same period, but the key difference is that by the 1990s most jobs were in subsectors that were internationally competitive. By 1997 export processing zones (EPZs) had created 44,000 jobs, with two-thirds of these in textiles, one-seventh in electronics, and one-tenth in metalworking (Jenkins et al. 1998). Firms in the EPZs accounted for one-fifth of manufacturing jobs, having doubled their workforce during 1991-96.

14.4.3. Total factor productivity Table 14.7 traces trends in productivity since the mid-1970s. It shows that the highest growth in TFP occurred during the period 1962-80 and especially during the subperiod 1966-73 when initial access to the CACM permitted manufacturers to sell most of their output within the region. The increased capital and labour translated into more production that was sold easily within the CACM. However, during the second phase of import substitution, from 1974 to 1980, the regional market lost its dynamism so that producers encountered problems. Diversification into markets outside the CACM was difficult because heavy protection had blunted incentives to search for efficiency gains and the international competitiveness of Central American manufacturingremained low. Meanwhile, as demand stalled, subsidized interest rates stimulated excessive investment and the capital stock expanded by more than 8 per cent annually. The resulting deterioration in efficiency is indicated by negative TFP growth. Table 14.6 . Employment by sector, Costa Rica 1976-1997 (% total employment)

Agriculture Industry Construction Basic services* Commerce** Personnel services + government

1976

1978

1980

1985

1990

1995

1996

1997

34.8 14.7 6.6 5.7 16.5 21.8

30.3 15.2 7.4 6.1 17.8 23.1

27.5 16.3 7.8 6.6 18.2 23.7

27.5 16.0 5.1 6.2 19.1 26.1

26.1 18.3 6.5 5.2 19.1 24.7

21.8 16.9 6.3 6.4 23.8 24.8

21.9 16.8 5.7 6.2 24.3 25.2

20.8 15.8 6.9 6.6 24.4 25.7

*Includes electricity, gas, water, transportation, storage, and communications. **lncludes commerce, finances, and real estates. Source: Banco Central de Costa Rica.

234

G. Barboza and J. Garden

Table 14.7 . Factor accumulation and total factor productivity growth, Costa Rica 1962-1995 Period Growt h Growt h Shar e K Employmen t Shar e of Residua l Real GD P K Stock Incom e o n Growt h wage s o n Growt h %% GD P% GD P ofTF P % 1953-95 1962-95 1962-80 1962-73 1966-73 1974-80 1981-82 1983-95 1983-85 1986-95

4.80 4.70 6.10 6.86 6.98 4.81 -4.70 4.09 3.14 4.37

6.16 5.75 7.42 7.03 6.70 8.08 1.94 3.89 2.08 4.43

0.47 0.47 0.47 0.48 0.47 0.46 0.53 0.45 0.47 0.45

3.30 3.52 3.73 3.35 3.58 4.36 2.42 3.38 2.86 3.54

0.53 0.53 0.53 0.52 0.53 0.54 0.47 0.55 0.53 0.55

0.15 0.15 0.64 1.75 1.93 -1.24 -6.78 0.49 0.64 0.44

Source: Cordero, J. (1999).

When the economy went into recession in 1981 and 1982, TFP fell by 6.9 per cent, but it grew at a modest 0.64 per cent annually through the 1983-85 stabilization programme (Table 14.7). This largely reflects the use of excess capacity that had been idled during the economic crisis which effect, when combined with a low rate of capital accumulation (2.1 per cent annually), yielded a modest rate of growth in output (3.1 per cent). The reform period lasted from 1986-95 but TFP growth averaged only 0.4 per cent per year, according to Table 14.7. Thus, efficiency gains during the trade liberalization period are lower than they were during the phase of forced industrialization of 1962-80. For the entire period 1953-95, TFP grew only 0.15 per cent per year on average. Consequently, the process of economic growth in Costa Rica has been based mostly on factor accumulation, as opposed to efficiency gains. Such accumulation-based growth is also identified by Young (1995) and Collins and Bosworth (1996) in their studies of the East Asian economies. The East Asian countries were much more successful than Costa Rica in generating efficiency gains, however, because the TFP growth rates for Costa Rica are well below the normal range of estimates for East Asia.

14.4.4. Natural resources Sustainable development requires the efficient use of renewable and nonrenewable resources. Table 14.1 shows that of the six categories into which natural capital is subdivided, Costa Rica has a higher value in cropland, pastureland, and protected areas than four other Central American republics. Costa Rica managed to achieve a modest rate of genuine saving, with the exception of the period of economic crisis (Table 8.5). Nevertheless, substantial degradation of the natural resource endowment occurred through the 1960s and 1970s.

Developing Market Economy: Costa Rica

235

Table 14.8. Costa Rica, land use 1970-1994 (million hectares) Element

1970

1980

1990

1994

Total Agricultural Arable and Permanent Crops Arable Land Permanent Crops Permanent Pasture Forests and Woodland All other Land

5.11 1.86 0.49 0.29 0.21 1.36 2.57 0.68

5.ll 2.52 0.51 0.28 0.22 2.01 1.83 0.76

5.11 2.86 0.53 0.29 0.24 2.33 1.57 0.68

5.11 2.87 0.53 0.29 0.25 2.34 1.57 0.67

Source: FAO http://www.fao.org/databases.

The original vegetative cover of El Salvador was overwhelmingly forest (more than 90 per cent), but by 1980 barely one-third of the forest remained (Table 14.8). Costa Rican farmers rapidly expanded agriculture in the post-war years as they diversified into bananas and beef rearing, so that the area farmed rose from one-fifth of the total in the late-1940s to almost two-thirds some twenty-five years later. The rapid closure of the land frontier raised fears about environmental damage. But these fears arose not so much from cropping, since the export crops of coffee and bananas occupy only 7 per cent of the total farmed area and are well-suited to the terrain. Rather, the problems arose with regard to grazing, which accounts for three-quarters of the farm-land. Land under permanent pasture rose sharply through the 1970s to just over 2 million hectares. The removal of the forest was associated with severe erosion that affects 17 per cent of the land while a further 24 per cent has experienced moderate erosion. Overgrazing has been responsible for 80 per cent of the soil loss in places where ranching has extended onto unsuitable land. Such land must be abandoned after a few years. There has also been a substantial loss of biodiversity in a region that is considered one of the richest habitats in the world with 5 per cent of the known global species. Rather belatedly, nearly one-fifth of the land area of Costa Rica has been designated for wildlife park management. Finally, the conversion of forest into pasture required little labour and so its expansion did little to offset the low level of job creation in the capitalintensive manufacturing sector. The economic reforms should improve this situation by lowering the exchange rate and export taxes so that the export of tree crops (bananas and coffee) becomes more competitive and those crops are better suited to sloping ground than cattle are. Table 14.8 confirms that the reforms have been associated with some change in the pattern of land use in Costa Rica. In particular, the conversion of land from forest to pasture has slowed and checked the expansion of the agricultural area. Meanwhile, cropland cultivation has been remarkably stable throughout the past thirty years. 14.5. C O N C L U S I O N S Costa Rica is unusual among resource-abundant countries because of its relatively high level of human development. Its social capital index is the highest among the countries

236

G. Barboza and J. Cordero

studied here for which the statistic is available (Table 9.4). Certainly, this social capital endowment has helped Costa Rica to avoid the severe sociopolitical tensions elsewhere within Central America, where assets were less equitably distributed, as population pressure placed increasing strains on society. However, the social capital was built through a trade-off that sacrificed growth in mean per capita income and, ultimately it was not sustainable either economically or environmentally. The democratic consensus led Costa Rica to pursue a development strategy of 'support-led security' as opposed to the 'growth-based security' of Malaysia. According to Dreze and Sen (1989), countries in the first category use public support to improve the standard of living without waiting for economic growth to generate such improvement. In contrast, in the 'growth-based security' category, public support is provided using the resources generated from economic growth. The missing condition for sustainable growth is adherence to market discipline in resource allocation and a consensual democracy does not automatically ensure that in the presence of natural resource abundance. The accumulation of social capital in Costa Rica was therefore associated with Olson effects. The ambitious expansion of social entitlements outstripped the capacity of the economy to sustain it. Meanwhile, forced industrialization distorted the economy through the 1960s and 1970s and caused it to lose momentum. The accumulated social capital then retarded economic reform. Like resource-abundant countries elsewhere in Latin America, a commodity boom (coffee during 1976-80) was used to relax pressure for reform rather than to improve the execution of reform. Recourse was also initially made to recycled petro-dollars to sustain growth in the late-1970s. Yet, as elsewhere, this merely postponed the need for reform and made the required adjustment more painful as economic distortions accumulated longer and foreign debt outstripped total GDP. Finally, having been delayed, the economic reforms were also less comprehensive than required to maximize improvements in economic efficiency, because of the need to negotiate with powerful urban interests. Nevertheless, the social capital did allow the reforms to proceed more smoothly and with less hardship for the poor than in other parts of Central America. If Olson effects provide one important reason for the lagged response to economic reform in Costa Rica, the difficulty of restructuring an economy that is small, but not too small, is another. Whereas microeconomics may be able to adjust quickly by serving niche markets in large economies (Easterly and Kraay 1999), small resource-abundant countries are likely to have acquired considerable inertia due not only to the difficulty of modifying social entitlements but also to stimulating additional commodity exports and restructuring obsolete capacity, notably in manufacturing. Yet, although the overall structure of Costa Rican production responded slowly to the reforms, the competitive diversification of the economy slowly accelerated, first into a wider range of primary product exports and, after a substantial lag, into manufacturing. As with Malaysia, export-promotion measures such as the EPZs attracted a considerable inflow of DFI into the manufacturing sector, albeit belatedly in the case of Costa Rica. This investment was initially attracted by cheap labour but began using more skilled and productive labour that had been disproportionately concentrated in the public sector, prior to the reforms.

Developing Market Economy: Costa Rica

237

It therefore seems likely that Costa Rica is finally gravitating towards the growth trajectory of Malaysia. In order to remain on that path, however, the Costa Rican government needs to learn from Malaysia and espouse an entitlement strategy of 'growth-with security'. This requires the emergence of sufficient groups that benefit from competitive activity to balance the interests of those who do not. This political process takes time. Economic reform therefore requires more than a change in economic policy, it also requires the accumulation of produced and social capital and that takes time, the exact amount of which will vary with the initial conditions. Moreover, Costa Rica shows that a social capital endowment that appears favourable under one set of conditions may subsequently prove an obstacle to improvement. REFERENCES Auty, R. M. and R. F. Mikesell (1998), Sustainable Development in Mineral Economies, Oxford: Clarendon Press. Baldwin, R. E. (1956), 'Patterns of settlement in newly settled regions', Manchester School of Social and Economic Studies, 24, 161-79. Banco Central de Costa Rica (Various Years) Anuario Estadistico, San Jose: Banco Central de Costa Rica. Buhner-Thomas, V. (1987), The Political Economy of Central America Since 1920, Cambridge: Cambridge University Press. Collins, S. and G. Bosworth (1996), 'Economic growth in East Asia: accumulation versus assimilation', Brookings Papers on Economic Activity. Cordero, J. (1999), 'El Crecimiento Econ242mico y la Inversion en Costa Rica', Working Paper, Instituto de Investigaciones en Ciencias Economicas, Universidad de Costa Rica, San Jose. Dr232ze, J. and A. Sen (1989), Hunger and Public Action, Oxford: Clarendon Press. Easterly, W. and Kraay, A. (1999) 'Small states, small problems?', Development Research Working Paper 2]39, Washington DC: World Bank. Gindling, T. H. and Terrell, K. (1995), 'The nature of minimum wages and their effectiveness as a wage floor in Costa Rica, 1976-91', World Development 23, 1439-58. Gonzalez-Vega, C. (1999) 'Costa Rica: mismanagement of the coffee boom', P. Collier, J. W. Gunning and Associates, Trade Shocks in Developing Countries: Vol. 2, Oxford: Oxford University Press: 42-81. Graham, E. and Floering, I. (1984), The Modern Plantation in the Third World, London: Croom Helm. Hirschman, A. O. (1977), 'A generalized linkage approach to development, with special reference to staples', in M. Nash (ed.), Essays on Economic Development and Cultural Change in Honor of Bert F. Hoselitz, Chicago 111: University of Chicago Press: 67—98. Jenkins, M., G. Esquivel and P. Larrain (1998), 'Export processing zones in Central America', Development Discussion Papers on Central America 4, Cambridge MA: HIID. Kunte, A., K. Hamilton, J. Dixon and M. Clemens (1998), Estimating National Wealth: Methodology and Results. Environment Department, Washington DC: World Bank. Morley, S. A. (1995), 'Structural adjustment and the determinants of poverty in Latin America', in N. Lustig (ed.), Coping with Austerity: Poverty and Inequality in Latin America, Washington DC: The Brookings Institution: 42-70. Nelson, J. D. (1990), Economic Crisis and Policy Choice, Princeton NJ: Princeton University Press: 169-213.

238

G. Barboza and J. Cordero

Rottenberg, S. (1993), The Political Economy of Poverty, Equity and Growth: Costa Rica and Uruguay, New York: Oxford University Press. Saborio, S. (1990), 'Central America', in J. Williamson (ed.), Latin American Adjustment: How Much has Happened? Washington DC: Institute for International Economics: 279-302. Shafer, D. M. (1994), Winners and Losers: How Sectors Shape the Development Prospects of States, Ithaca: Cornell University Press. Ulate, A. (1999), Crecimiento, Empleo y Equidad en America Latina. Costa Rica, Economic Commission for Latin America and the Caribbean, Santiago (mimeo). United Nations (1998), Human Development Report 1998, United Nations, New York. World Bank (1999), World Development Indicators 1999, World Bank, Washington DC. Young, A. (1995), 'The tyranny of numbers: confronting the statistical realities of the East Asian growth experience', The Quarterly Journal of Economics, 40, 641-80.

15 Growth, Capital Accumulation, and Economic Reform in South Africa CAROLYN JENKIN S

1 5 . 1 . INTRODUCTIO N South Africa is a mineral exporting country. It is the world's largest gold producer, and it has commercially viable reserves of all major minerals except oil and bauxite. The economy is essentially market-driven, although historically there was heavy government intervention, in order, amongst other things, to maintain a system of institutionalized racial discrimination. Of its population of about 44 million, over 80 per cent are indigenous blacks. Although its income per head is over US$3,000, its measured income inequality is very high,1 creating political pressures for redistribution of income and wealth (see Table 9.4). Compared with most of the rest of sub-Saharan Africa, the economy is large, diversified and developed. Nevertheless, throughout the twentieth century mining output has dominated exports, and from the 1970s fluctuations in the gold price, in particular, have had a significant influence on macroeconomic performance. The international isolation of the country in the apartheid era had wide-ranging effects, deepening the aspiration for self-sufficiency as an aim of economic policy and providing an additional disincentive to reform a protectionist trade regime designed to promote the welfare of the relatively skilled and educated minority. Even so, South Africa began a process of trade liberalization some two decades before joining the WTO in 1994. The process has been gradual and sustained, and is not yet complete. Since 1980, macroeconomic performance has been poor. Growth has decelerated to virtually nothing; unemployment is very high; and inflation was in the 10-20 per cent range from the early-1970s until the mid-1990s. This economic performance parallels that of other resource-rich middle-income countries that pursued an importsubstitution strategy over a long period—especially those which experienced social conflict. In fact, South Africa makes an interesting case study of a resource-rich country in transition. As Chapter 14 has shown, while policy matters, and matters a great deal, initial conditions do constrain the effectiveness of economic policy. This chapter shows 1

In 1993 the Gini coefficient was 0.62.

240

C. Jenkins

that, although South Africa embarked on reform over a quarter of a century ago, it still has not established the conditions required to achieve its growth potential. Part of the reason for this is that reform was undertaken during conditions of increasing isolation under sanctions and increasing domestic uncertainty. Physical investment actually declined. In addition, racial segregation of both education and the labour market, together with increasing political polarization, meant that skills were acquired slowly while social capital ran down. The new majority government is in a better position to tackle these problems but improvements have been slow in coming. Some of the reasons for this, like contagion effects of international financial crises, are outside of the government's control. Others arise from the conscious decision to sacrifice some growth in the pursuit of redistribution imperatives. Redistribution is clearly necessary. Even so, greater focus is needed on education and training, and on dealing with the root causes of the haemorrhage of skills and reluctance of domestic investors to commit to new projects. The next section sets out the characteristics of the typical development trajectory of a mineral-driven economy, which provides a context for assessing whether South Africa's experience fits the usual pattern. Section 15.3 compares South Africa's growth experience with that of other middle-income countries to give an indication as to what factors conditioned its growth trajectory. Section 15.4 considers in more detail the magnitude and composition of investment, and how these have changed over time, and Section 15.5 assesses the current policy reform process, focusing on the constraints imposed by the need to redistribute both income and wealth to the recently enfranchised black majority.

1 5 . 2 . ECONOMI C PERFORMANCE AN D DEVELOPMENT STRATEG Y 15.2.1. Economic performance Like that of many other developing countries dependent on exporting primary commodities other than oil, South Africa's economic growth decelerated as a result of a series of shocks in the 1970s. Growth performance was particularly weak after commodity prices collapsed in the early 1980s (Table 15.1). The turning point was probably 1974, when the long-run growth rate began to decline and the cyclical pattern of economic activity became more unstable.2 The 1974 crisis was the culmination of various structural problems, which, amongst other things, reduced the profitability of investment. The range of causes of slower growth is summarized below. Many of these appear to be as predicted by the model of resource-abundant growth outlined in Chapter 8, although South Africa's unique political problems exacerbated these factors. In the 1970s labour productivity was falling because of skills shortages, partly the result of historical under-investment in education. Political unrest, frequently expressed by strikes, was also driving up real unit labour costs. The growth of the 192,75 appears to be a turning point for many countries experiencing a collapse in economic growth, with sharper falls in those countries divided by social conflict (Rodrik 1998).

Economic Reform in South Africa

241

Table 15.1. Economic performance indicators (% p.a.)

Real annual average growth rates Population growth Average private investment ratio Average private savings ratio Inflation

1951-70

1971-80

1981-90

1991-98

5.1 n.a. 16.2* 17.4* 3.6*

3.9 n.a. 19.3 19.9 9.5

1.6 2.4 18.5 22.2 14.4

1.1 2.1 14.5 20.0 13.2

*1960-70. Source: Calculated from SARB data.

capital-labour ratio accelerated (in an attempt to maintain productivity). Consequently, the capital-output ratio rose sharply, reflecting the falling profitability of investment (Gelb 1991: 19-23). Insufficiently developed export capacity, together with a high propensity to import investment goods, placed a foreign-exchange constraint on growth. This was exacerbated by an overvalued exchange rate and, at least in prospect, by calls for international sanctions. Although a meteoric increase in the price of gold helped to partially insulate South Africa from the balance-of-payments crises that were a factor in driving many developing countries into debt, falling world demand contributed to falling production and rising unemployment. The government's almost complete disinterest in stimulating alternative employment for blacks (Thomas 1990: 255), and the deliberate dualism in both agricultural development and skills acquisition made it difficult to absorb new job seekers: job creation fell from 157,000 annually between 1960 and 1974 to 57,000 a year from 1974 to 1985. Financial sanctions imposed against South Africa in 1985—86 did precipitate a debt crisis, although, unlike other developing countries, debt-service difficulties arose from the structure of the foreign debt (predominantly short-term loans) rather than its magnitude. Increasing social conflict, political uncertainty and tight monetary conditions necessitated by financial sanctions caused private investment to fall. Costs of production rose sharply: of capital, because of a shift in policy towards maintaining positive real interest rates and a weaker currency; of labour, because union pressures saw wage increases outstripping productivity improvements and because of a huge increase in politically-motivated strikes; and of raw materials, because of both domestic inflation and a weaker currency. The economy stagnated. With the change of government in 1994 came the lifting of economic sanctions and a surge in confidence. Growth accelerated (modestly) for a period of about three years. However, the tight anti-inflationary stance of the central bank (successful in reducing inflation to 6.9 per cent by 1998) stifled growth. This, together with a tighter fiscal stance, significant structural changes occurring as a result of rapid trade liberalization, and nervousness about emerging markets in the wake of a series of financial crises, meant that South Africa's growth performance in the 1990s was disappointing. Although South Africa has not experienced a prolonged period of contraction, it does display a sustained deceleration of economic growth. It is possible to say that

242

C. Jenkins

South Africa demonstrates many of the characteristics of mineral-driven development, with the resulting misallocation of investment and slow growth, and that many of these characteristics were heightened by apartheid, which, amongst other things institutionalized skills acquisition and income differentials along racial lines and prevented investment in social capital. This becomes even more evident when changes in policy and capital accumulation are described.

15.2.2. The evolution of trade and industrial policy This section traces out South Africa's trade and industrial policy, which has been one of the key features of its development strategy. The events are summarized in appendix Table 15.1. South Africa was one of the first of the world's current middle-income countries to explicitly adopt, in 1925, a policy of import-substituting industrialization (ISI). ISI was expected to assist with developing greater economic independence from Britain and with creating employment for whites in manufacturing. There was an expansion of both parastatal and private investment, some of which was financed from taxes and profits generated in the mining sector, with substantial gains both in employment and in industrial expansion and diversification. Although protection weakened competitiveness, many industries became reliant on protection for survival, and its maintenance became a political issue, particularly after 1960 when the threat of economic isolation added a strategic need for continuing ISI. Manufacturing growth began to slow towards the end of the 1960s. Estimates of the contribution of ISI to growth show that import substitution was an important source of overall growth during the 'easy' stages of ISI; thereafter, given the size of the domestic market, further import substitution into capital goods did not drive growth (Scheepers 1969). Increasingly, further industrial expansion was dependent on the ability to import capital, and, therefore, on primary export earnings, as a marked anti-export bias existed in manufacturing. The failure to expand exports, together with the recognition that ISI was no longer driving growth, initiated a major policy shift in the early 1970s. There was concern that South Africa was too dependent on a single commodity, gold (at that stage subject to a fixed price). In any event, imports were rising too rapidly for the trade deficit to be financed by gold exports and capital inflows. The government introduced a policy of export promotion through a range of incentives, but left the protective tariff and import control structure largely unchanged. The strategy was reinforced by the real devaluation of the rand in the mid-1970s. In the years that followed the shift in policy, exports grew in volume terms more rapidly than imports, although these effects are better attributed to the depreciation of the currency than to any programme of export promotion (Holden 1990: 267). The increase, between 1974 and 1985, came mainly from expanding mining exports (from 57 per cent to 62 per cent of the value of total exports) and from processed minerals (from around 5 per cent of total exports (15 per cent of manufactured exports) to

Economic Reform in South Africa

243

17.5 per cent of the total (52 per cent of manufactured exports)). With few exceptions, most other categories of manufactures were sold mainly in the domestic market. Generally speaking, the 1980s saw a shift towards market-oriented instruments of economic policy, especially monetary and exchange-rate policy. In 1983 the government began to liberalize trade by replacing quantitative restrictions (QRs) with import tariffs.3 The liberalization was not unambiguous: import surcharges, introduced in 1985 in response to a tightening of economic sanctions, raised rates of effective protection substantially (IDC 1990), and, as in other countries, increases in tariffs were used to offset the effects of the removal of QRs on industry. It has been argued that the dismantling of QRs initiated what has been called 'South Africa's second liberalization episode' (Bell 1993), although, if anything, effective protection rose. The process was supplemented with more effective incentives to exporters, and supported by a more appropriate exchange-rate policy. The sanctionsprecipitated debt crisis of 1985 increased the urgency with which this was pursued, as it was necessary to turn the historical current-account deficit into a substantial surplus to enable the servicing of the foreign debt. The effect was a dramatic increase in nongold export volumes that increased by an average of 7.7 per cent annually between 1984 and 1990, in spite of trade sanctions. The proportion of gold in total exports fell from 48 per cent in 1983 to 27 per cent in 1992. There were fears that the new majority government elected in 1994 would be under pressure to reverse the reforms, opting for more interventionist policies. However, the new government recognized the need for South Africa to become internationally more competitive, and opted for a more neutral trade regime, with a simpler, transparent tariff structure, and tariffs reduced in line with GATT requirements. An agreement was signed with the WTO in 1994. After 1994 the new government accelerated the agreed programme of tariff reduction, and introduced supplementary measures to deepen the liberalization, including privatization and the abolition of most agricultural marketing boards.4 A phased relaxation of exchange controls occurred simultaneously with the freeing of trade, although neither process is yet complete. In summary, South Africa has experienced not so much a trade liberalization episode as a gradual shift of trade policy in a more liberal direction, accelerated with the transfer to majority rule and the removal of international sanctions. The complex structure of import protection meant that the system was open to lobbying by sectional interests, although this began to change even before the change of government in 1994. Liberalization coincided with economic sanctions, which both intensified the urgency with which it was pursued and offset some of the potential benefits. 3 The share of imports (by value) requiring permits fell from 77% in 1983 to 23% in 1985. By 1989 the proportion was 15%. The government was not, however, in favour of free trade. A 1985 white paper (RSA 1985: 14) states: 'The government wishes to make it very clear that it has never espoused any so-called "free-trade" policy. On the contrary, like all previous South African governments since 1924, and indeed earlier, it has decisively implemented a policy of protection for industry.' 4 The marketing boards had power over imports, exports and prices of most agricultural products. In part, their abolition is in line with WTO requirements, but objectives also include breaking the monopolistic stranglehold over agricultural markets and the opening up of the sector to those previously excluded.

244

C. Jenkins 1 5 . 3 . GROWT H EXPERIENCE I N C O M P A R A T I V E PERSPECTIVE

South Africa is fairly typical of other developing countries classed by the World Bank as 'upper middle income'. In this section its growth performance is compared with that of other countries in its group, and an attempt is made to isolate the possible reasons for its declining growth rate. Ideally, one would want to estimate a growth model for South Africa in order to ascertain what determines the country's growth path, but data are too poor to undertake theoretically-consistent econometric work. The methodology employed in this section is, therefore, to identify those variables that have been found to explain growth in other developing regions and then to compare South Africa's averages for these variables with the averages for other upper middle-income developing countries. The starting point is a paper by Sala-i-Martin (1997), written as a response to earlier econometric studies involving a range of methodologies and hundreds of variables to explain growth, which attempts to rank explanatory variables according to their robustness in regressions. Sala-i-Martin draws on the empirical growth literature to identify variables that 'explain' economic growth. Three fixed variables are almost always found to be important: initial per capita income, initial life expectancy, and the initial primary school enrolment rate. A model incorporating these is then estimated with 62 additional variables that are also argued to contribute to growth performance. The results are reported for those that are found to be strongly correlated with growth. Sala-i-Martin's 'robust' variables5 and their average values for upper middle income countries6 are recorded in Table 15.2. The countries whose names appear in bold type are those where mining is an important component of GDP and/or where primary exports are a significant proportion of total exports7 In order to avoid the problem of noncomparability of data, the data used have been taken from the six largest of the original cross-country studies. The period covered by growth regression analysis is typically 1960-90, as these are the years for which GDP adjusted for purchasing power parity are available.8 Half of the twelve countries reported in the table are dependent on primary commodities. In five, over 10 per cent of GDP is accounted for by mining (nearly one quarter in the case of Saudi Arabia), and in most of these the proportion of primary products (minerals and agriculture) in total exports is high. Diversification away from dependence on primary exports is a consistently significant regressor in cross-country growth models. Generally speaking, the per capita growth rate is higher, the lower the proportion of primary exports in total exports (or in GDP). This is not obvious from 5

Religious variables are omitted. World Bank Environmental Economics and Indicators Unit classification. ' Note that this definition of natural-resource abundance differs from that used elsewhere in this volume. Here we arc constrained by the information used in cross-country regressions. 8 Because of the importance of 'initial conditions', some of the variables in Table 15.2 are intended to apply to several decades ago. In the case of longer series of data, most are out of date, simply because more recent information is not available across all countries in the sample. 6

Table Table 15.2. 15.2. The The determinants determinants of of growth in in GDP GDP per capita capita in in comparative comparative perspective perspective South South Mauritius Mauritius Mexico Mexico Argentina Argentina Brazil Brazil Africa Africa Growth Growth Growth Growth of of real GDP GDP per capita capita (1985 international international prices) prices) (1985

0.014 0.014

0.023 0.023

0.022 0.022

0.002 0.002

Chile Chile

Uruguay Uruguay Trin. Trin. & & Korea Korea Tobago Tobago

0.027 0.027

0.011 0.011

0.004 0.004

0.014 0.014

0.063 0.063

1745 1745

2898

3847

5630 5630

883 883

0.95 0.95

11

11

0.88

0.94 0.94

57.3 57.3

67.3

Malaysia Malaysia Saudi Saudi Greece Greece Source Source? Arabia Arabia 0.042 0.042

n.a. n.a.

0.039 0.039

B&L B&L

n.a.

2066

B&L B&L

0.12

11

B&L B&L

68.8 68.8

B&L

Fixed variables variables Fixed

Level income, 1960 Level of ofincome, 1960 (1985 prices) prices) (1985 Primary Primary school school enrolment in 1960 Life Life expectancy in in 1960 1960

2107 0.89 0.89 49.2

2818 0.98 0.98 59.4

2798 0.80 0.80

3294 0.98 0.98

57.3 57.3

65.2 65.2

54.9 54.9

63.9

54.2 54.2

1381 1381 0.96 0.96 54.3 54.3

44.7

Additional variables

Non-equipment n.a. 0.16 investment/ GNP Equipment n.a. 0.00 Equipment investment/GNP Number of years open, 1965-90 25 1965-90 0 Fraction GDP in mining 0.11 0.00 Fraction of ofGDP Primary 0.17 0.29 Primary exports in 1970 1970 Rule o fflaw (score)"a Rule law (score) Revolutions 0 Revolutions and coups (fraction) 0.04 War dummy 11 0 Wardummy Political rights (score)b 4.78 2.17 Civil liberties (score)b 5.67 2.39 (score)"c Degree o f capitalism (score) 3 55 Absolute latitude 29° 20° SD 12 S D black market premium 12 Real exchange rate distortion 72 126 126

0.17 0.17

0.12 0.12

0.16

0.10

0.17

0.02

0.02

0.06 0.02 0.02 0 14 0 14 0.02 0.16 0.00 0.16 0.05 0.15 0.09 453 5 3 0.12 0.2 0 00 11 0 3.39 5.78 4.22 3.39 4.67 4.17 3 3 5 20° 34° 35° 16 331 28 16 331 28 97 100 92 100

0.04 0 0.16 0.08 4 0.05 0 1.56 1.56 1.89 1.89 3 10° 10° 7 96

50 0.03 0.02 4 22 0 00 11 3.67 3.67 3 55 17° 17° 15 15 71

0.02 0.05 0.96 3.94 3.56 37° 67 67 113 113

0.13

0.16

0.19

0.06 0.04 22 25 0.00 0.10 0.02 0.37 3 44 0.4 0.04 0 0 4.44 3.11 5.17 4.06 3 5 38° 3° 28 2 28 110 88 110

n.a. n.a. 0 0.23 0.60 n.a.. n>a 0.12 00 11 6.06 6.33 3 44 23° 24 n.a.

0.23

D&S D&S

0.07 D&S 25 S&W 0.02 H&J 0.04 S&W 3 &L EE&L 0.2 B&L B &B&L L 1.94 B&L 1.94 2.22 B&L H H&J & J 38° H&J 4 LL&R &R 116 L&R 116

Notes: Bold type indicates dependence on primary commodity production. Notes: "1980s; high scores scores are bad; huge gaps gaps in the data, treat with caution. "1980s; b !972-89; high scores are bad. 1972-89; c Score 4 or 5 indicates high degree degree of capitalism. Sources: 1. Barro & Lee (1994), 2. Levine & Renelt (1992), 3. Easterly & Levine (1997), 4. Hal! & Jones Jones (1998), 5. Sachs & Warner (1997), 6. DC Long & Summers (1993) Sources: (1993)-

246

C. Jenkins

Table 15.2, where Asian countries and the only European country have outperformed those in Africa and Latin America over the thirty-one years 1960-90, irrespective of whether or not they are dependent on primary exports. Between 1960 and 1990, South Africa's real output per head grew at a rate similar to that of Chile and of Trinidad and Tobago, both of which are mineral economies. These three countries outperformed Argentina and Uruguay, neither of which produce and export minerals in significant proportions—although agriculture is important in Uruguay. However, they, in turn, were outperformed by the other countries in Table 15.2, some of which are primary exporters and others of which are not. It appears that factors in addition to the resource endowment must be considered when assessing the comparative growth performance of the countries in Table 15.2.

15.3.1. Initial socioeconomic conditions Convergence processes suggest that the initial level of income is a factor in determining the speed at which economic growth occurs, with initially poorer countries (here Korea, Malaysia, and Brazil) growing faster if they are able to take advantage of technological developments in more advanced countries, thereby converging on their own long-run growth paths. South Africa and Greece had very similar levels of GDP per head in 1960 (just over $2,000 in 1985 international prices) which suggests that, in theory and other things being equal, they should have been experiencing similar growth rates. Other things were not equal, however. Indicators of human capital (education and health) were poorer in South Africa than they were in any of the countries now classified as upper middle income, with the exception of Mexico and Saudi Arabia. The level of education of the workforce is particularly important, as expenditure on schooling generally raises productivity (and employability) over time, and basic education is important for developing exports of manufactured goods (Wood 1994). Low investment in human capital may explain, in part, why South Africa has grown more slowly (see Chapter 4).

15.3.2. Investment Neo-classical growth models stress the importance of physical investment in driving more rapid expansion of output, and the correlation between the rate of investment and the rate of growth of output is strong in all studies that analyse the determinants of economic growth. This is particularly true of equipment investment (rather than buildings and infrastructure). The average ratio of total investment to GNP (over the period 1960—85) for all of the faster growing economies in Table 15.2 approximates or exceeds 20 per cent. Equipment investment to GNP is notably higher in the countries recording above-average growth (6—7 per cent in Brazil, Korea, and Greece). Nonequipment investment also tends to be higher in the faster growing countries. Unfortunately the data sets used do not include figures for South Africa. Calculations using data published by the South African Reserve Bank suggest that equipment investment averaged 8 per cent of GNP from 1960 to 1990 (and total investment

Economic Reform in South Africa

247

2.3 per cent of GNP), but there is no way of knowing whether these figures are comparable with those used for the calculations in Table 15.2. It is possible that South Africa has had comparatively high rates of physical investment, as relative prices were skewed to favour capital. If this is the case then other explanations must be found for lower than average rate of economic growth.

15.3.3. Economic policy Indicators of economic policy and stability which have been found to correlate strongly with economic growth include the extent of openness to international trade, the degree to which markets are allowed to operate, the extent of exchange-rate distortion, and the variability of the premium on the black-market exchange rate. The countries in the sample which have tended to grow faster are those which have been open for at least some of the period 1965-90. The growth literature has a lot to say about the relationship between openness to international trade and economic growth. There is an emerging consensus that outward orientation facilitates growth in both developed and developing countries.9 A lower degree of distortion of the exchange rate—and a lower degree of variability in the black-market rate (usually associated with lower levels of black-market trading)—are evident in the fastergrowing economies. These indicators are symptomatic of macroeconomic instability and policy uncertainty, and may also indicate overvaluation caused by Dutch disease. Although South Africa displayed some distortion of the exchange rate, and a small black market premium, these were moderate by developing country standards. This is entirely believable, as government spending did not ever race out of control to the same extent as in many other mineral economies, and, even during the period of political uncertainty, the authorities maintained macroeconomic stability sufficient to limit the extent of capital flight. Although frequently significant in cross-country regressions, it is not obvious from the averages presented in the table that fast-growing economies are associated with a high degree of market orientation. This implies at least that considerable government intervention in the economy is not inimical to growth, as Malaysia demonstrates: it is the nature of the intervention that matters (Chapter 9).

15.3.4. Civil liberties and human rights The type of state often emerges as important in determining economic growth (see Chapter 8), although, again, it is not obvious from the data presented in the table that a higher level of civil liberty and democracy accompanies economic growth. South Africa rates worse on civil and political liberties than other countries in the sample, with the exceptions of Chile, Korea, and Saudi Arabia. It is very likely that the See, 9erfor example, Collier and Gunning (1999), Dollar (1992), Edwards (1993), and Sachs and Warner (1997). These findings are not undisputed (Krishna et al. 1998; Rodriguez and Rodrik 1999). Trade policies are, to some extent endogenous, so the direction of causation may not be obvious. Moreover, protection of manufacturing often produces high growth initially but slower growth in the longer run.

248

C. Jenkins

interaction of social conflict and external shocks were responsible for the economy's stagnation in the 1980s and early 1990s. This section has given an indication, by placing South Africa in comparative perspective, of the possible determinants of its growth performance. In most of the variables which have been found worldwide to explain economic growth, South Africa has tended to underperforrn faster-growing countries. This is most observable in the areas of human and social capital, social and political stability, and trade policy, which is at least partially consistent with the experience of other resource-abundant countries. The evolution of the country's industrialization strategy was sketched earlier. The following sections consider the structure of the capital stock, and social and political factors.

15.4. THE M A G N I T U D E AND COMPOSITION OF W E A L T H 15.4.1. The stock of capital in 1994 It appears that South Africa has underperformed other middle-income developing countries in investment in human capital but not in physical capital formation. This is only partially consistent with the experience of other natural resource-abundant countries, where natural resource rents often distract governments from the need to develop human resources and divert private investors away from wealth creation into rent-seeking activities (Ranis 1991). The World Bank (1999) has estimated the stock (in 1994) of human, produced, and natural capital for 92 countries.10 Table 15.3 reports the estimated proportions of each form of capital in total wealth for the twelve upper middle income countries. The figures are converted to US dollars for comparison, but are not adjusted for purchasing power parity. In terms of the structure of wealth, South Africa appears to be representative of its group, showing no severe distortion. Less than 10 per cent of South Africa's estimated wealth in 1994 was in the form of natural capital (agricultural land as well as minerals). The proportion of wealth held in the form of produced capital is particularly high, and previous investment in human resources is about average for the group. In per capita terms, however, South Africa compares unfavourably with the other countries: its human capital stock per person ($37,000) is almost the worst of the group, and its stock of produced capital per person ($11,000) is third from the bottom. Moreover, these figures imply nothing about either the efficiency of the capital stock or its ownership distribution. In South Africa, much previous investment in both 10 As these data are in an early stage of compilation, there may still be inaccuracies (see Hamilton, this volume). The data for South Africa are particularly odd. It is distinctly possible that natural resources are undervalued. In this case, human resources will be overestimated, as they are a residual. 11 When adjusted for purchasing power parity, these percentages are not significantly different: 75 per cent, 20 per cent, and 5 per cent respectively. 12 The EEI Unit's figures for wealth embodied in minerals and metals is the estimated present value of reserves. The value of South Africa's reserves is probably underestimated.

Table 15.3 . Wealth estimates for upper middle-income developing countries, 1994

Trin. & Kore a Malaysi Tobago Afric US$bn Total weath

83

US$ per head Total wealth Human resources Produced assets Natural capital

64,400 33,212 19,078 12,109

% of total wealth Human resources Produced assets Natural capital

0.52 0.30 0.19

6,025

a Mauritiu s Sout h Saud i Mexic a Arabi a

1,280

57

2,168

135,193 65,639 109,979 41,942 22,278 11,881 2,937 11,817

51,874 40,287 10,351 1,236

52,135 36,969 10,962 4,205

0.81 0.16 0.02

0.64 0.18 0.18

Source: Calculated from World Bank EE1 Unit (1999).

0.78 0.20 0.02

0.71 0.21 0.08

2,608

o Argentin a Brazi l Chil e Urugua y Greec e

6,267

4,680

9,453

958

264

1,077

149,024 68,224 52,853 49,405 24,296 12,187 71,876 6,632

136,911 114,504 12,557 9,850

59,400 41,492 10,846 7,061

68,249 45,919 7,890 14,440

83,351 59,682 8,864 14,806

103,513 75,046 23,261 5,207

0.70 0.18 0.12

0.67 0.12 0.21

0.35 0.16 0.48

0.72 0.18 0.10

0.84 0.09 0.07

0.72 0.11 0.18

0.72 0.22 0.05

250

C. Jenkins

Figure 15.1 . Ratios of GDI and CDS to GNP (%), 1970-1994 Source: Derived from World Bank data.

physical and human capital is arguably inefficient,13 and its distribution is highly unequal, held by an economic elite defined, until recently, on racial grounds and protected by a legal and economic policy framework.

15.4.2. Trends in capital accumulation Until the early 1970s South Africa was successful in attracting large inflows of both direct and portfolio foreign investment, mainly into mining. Relatively high domestic savings financed local capital accumulation and meant that a domestic capitalist class developed. After the mineral boom of 1979-81 both GDI and GDS fell as a proportion of GNP (Fig. 15.1). The economy stagnated, consumption absorbed a larger proportion of output, and investment fell. As a proportion of GNP, public spending on education rose sharply during the 1970s, falling very slowly thereafter (Fig. 15.2). Prior to that, it was a matter of explicit policy not to provide more than a basic education for Africans, who were viewed as an unskilled workforce. This is entirely consistent with the considerable investment in physical capital, which required a small highly-skilled workforce and with the strategy of ISI, one aim of which was to provide employment for whites. The country's mineral wealth made the strategy appear sustainable, and reduced the need to develop labourintensive industry so as to absorb the black population into the modernizing economy, at least until the 'easy' stages of ISI were over and growth slowed in the 1970s. As political protests intensified, and as the government began to move pro-actively 13 It is not really possible to generalize about the efficiency of investment. Some huge capital-intensive projects, pursued as part of the self-sufficiency strategy, were a complete waste of money; others, including some parastatals, have become so efficient that estimates of South Africa's comparative advantage have concluded that the country has a comparative advantage in large capital-intensive production (see Holden and Holden 1978).

Economic Reform in South Africa

251

Figure 15.2 . Education/GNP (%), 1970-1994 Source: Derived from World Bank data.

Table 15.4. Extended genuine saving, South Africa and income group averages, 1993

Extended genuine Hig saving incom % of GNP US$per cap.

h Uppe r Lowe r Lo e middle-incom e middle-incom

14% 3433

10% 39

11% 415

w Sout e incom e Afric 11% 147

h a

5% 144

Source: World Bank EE1 Unit (1999).

towards promoting export diversification, investment in black education rose rapidly, although it is still inadequate. In other words, the aims of apartheid and industrialization policy have overlapped to a considerable extent for most of the century. Saving (out of which future investment is financed) is declining. In 1993 South Africa's gross domestic saving was 17.2 per cent of GNP, which is low for a country at its level of development. By 1999, the ratio had fallen to 14.5 per cent. Sustainable development requires even higher rates of gross savings (see Chapter 3). If the definition of 'saving' is extended to take into account both natural resource depletion (deduction) and the accumulation of human capital (addition), the ratio of genuine saving to GNP was only 6.7 per cent; deducting the costs of pollution damage causes the ratio to fall to around 5 per cent (Table 15.4). This is considerably lower than the averages for countries in all income groups, including the poorest. If current spending on education is excluded from the concept of genuine saving, then the rate of resource depletion has almost entirely offset any wealth accumulation in the 1980s and 1990s (Fig. 15.3). This is consistent with the experience of other countries that have experienced growth collapses: natural capital (along with other forms of capital) tends to be eroded. It is imperative that the rate of saving be raised,

252 ^3^


Figure 15.3 . Extended genuine saving II/GNP (%), 1970-1994 Source: Derived from World Bank data. Genuine Saving I = Gross Saving — Depreciation — Resource Rents (Depletion of Natural Resources). Genuine Saving II = Genuine Saving I — CO2 Damage (lower curve). Extended Genuine Saving II = Genuine Saving II + Education Spending (upper curve).

that the rate of environmental exploitation be reduced, and that the rate of natural resource depletion be at least offset by rates of accumulation of human and produced capital if the country is to pursue a sustainable growth path.

15.4.3. Trends in the accumulation of natural resource rents The ability to accumulate natural resource rents not only reduces investment in human and produced capital, but it supports ISI long after its contribution to development has ceased to be positive (Chapter 8 this volume; Ranis 1991: 76-7). It is certainly true that significant reform of South Africa's industrial strategy was postponed until the change of government in 1994, two decades after it became evident that ISI was no longer driving growth, and that the gold boom, which coincided with the oil-price shocks, made it appear that both a difficult world environment and economic sanctions could be survived. Was it also true that natural resource rents influenced the pattern of capital accumulation? Unfortunately, data on the rents accruing from natural resources are available only from 1970, which coincides with the beginning of the reform process in South Africa. These are plotted as a proportion of GNP in Fig. 15.4. A comparison of Figs. 15.1, 15.2 and 15.4 reveals that the surge in natural resource rents—caused by the 1970s surge in commodity prices—coincided with the period in which government spending on education and private investment increased. As the rents declined so, in fact, did current spending on education and gross investment as a proportion of GNP. This pattern is consistent with that following mineral booms elsewhere in Africa (Collier and Gunning 1994; Deaton 1992): spending of all kinds, especially investment, tends to rise. The problem is usually that the private sector

Economic Reform in South Africa

253

Figure 15.4. Resource rents/GNP (%), 1970-1994 Source: Derived from World Bank data.

treats the boom as temporary, while governments frequently do not, generating Dutch disease effects. Although South Africa's natural resources typically generate 'point' socioeconomic linkages rather than 'diffuse' ones, one of the most significant differences between South Africa and other 'point' resource-abundant countries, is that the mineral resources are largely privately owned, and the state has access to rents mainly via taxation. This may be an important factor in reducing the Dutch disease effects of commodity-price booms. Although rents allowed economic liberalization to be delayed, for reasons given earlier, the government began to spend more on black education in the 1970s. Higher tax revenues from the mining sector enabled this to be financed without significantly reducing other forms of spending. Moreover, because the bulk of the rents accrued to the private sector, it used the windfall to invest not only in mining but also in manufacturing.14 The fall in rents in the 1980s may have accelerated the shift in thinking towards liberalization, particularly as the economy was under strain from financial sanctions. However, there were other factors which played a role in encouraging the transition. South Africa has always placed considerable emphasis on international integration. ISI was not, as so often elsewhere, part of a strategy of withdrawal from world markets. After 1960, only South Africa withdrew from international institutions to preempt expulsion. As world economic ideology moved rightwards during the 1980s, government and business-sector thinking shifted in concert—at least part of the way. Moreover, by the end of the 1980s it was apparent that a change in the political regime was inevitable. Fears of African socialist-type policies helped to prompt moves towards M14ining conglomerates themselves diversified their activities, moving into manufacturing activities of all kinds.

254

C. Jenkins

economic liberalization and the restructuring of state assets, so that, by diluting the resource rents, they would not fall into the hands of the new state. It is difficult therefore to separate the effects of natural resource rents from other factors which were present in the shift towards less interventionist economic policy. 15.5. THE R E F O R M OF POLICY IN SOUTH A F R I C A Although South Africa embarked on reform over a quarter of a century ago, it still has not established the conditions required to achieve its growth potential, in spite of pursuing a largely orthodox macroeconomic policy strategy and avoiding the extreme distortions found in other resource-abundant developing countries. The policy reform process commenced in the 1970s, and proceeded in phases in the 1980s. Liberalization of all aspects of the economy—with the exception of the labour market—has been accelerated since the transfer to majority rule in 1994. The reforms that have taken place have not been part of a general and formal structural adjustment programme, but have occurred in response to perceived economic problems. This piecemeal introduction of policies means that the policy regime has consisted of both overlapping and sometimes contradictory policies. However, because liberalization measures are designed domestically, with the aim of addressing particular problems, and generally, with the support of the business community and the trade unions, the government is committed to each reform, and there is no significant credibility problem. Of course, this does not imply that the government has necessarily responded to problems sufficiently quickly. The existence of large natural resource rents made delays in reform possible in the past, and they are likely to do so again, especially with respect to labour-market liberalization.15 One reason why reform has not produced faster economic growth is because it was initially undertaken during conditions of increasing external isolation and domestic uncertainty. Physical investment, previously high, actually declined. In addition, racial segregation of both education and the labour market, together with increasing political polarization, meant that skills were acquired slowly while social capital ran down. The new majority government is in a better position to tackle these problems but improvements have been slow in coming. In part, there are problems outside of the control of government which have contributed to lower-than-potential growth. In particular, a precarious foreign reserves position has meant that it has been necessary to maintain very high real interest rates over a sustained period to protect the economy from the contagion effects of international financial crises. It helps that, with the exception of the organs of institutionalized racial discrimination, institutions and organizations have been maintained. This reduces the instability which accompanies such a major political change. 15 This is common amongst natural resource rich countries where labour movements tend to be strong (Ranis 1991). 16 The concept of political institutions includes not only formal rules and informal constraints, but also the form of the interactions of private entities with the state (e.g. patron-client networks). Organizations, on the other hand, are the players: political, social, economic, and educational bodies (North, 1995: 23).

Economic Reform in South Africa

255

However, there are areas in which policy is deficient. Education is probably the most serious failure of the new majority government. It is recognized that, in spite of now absorbing almost a quarter of non-interest government expenditure, the availability and standard of education for the previously disadvantaged majority is yet to show a marked improvement. There is a lack of clear direction in education policy, in the context where radical change is badly needed. Thus, despite the increasing share of education expenditure, which is high by international standards, its contribution to improving either quality or access is limited—evidence that simply spending more money does not necessarily solve problems. In addition, new labour legislation may undermine some of the objectives of current industrial policy, if it makes it more difficult to improve competitiveness.17 The government's primary objective is employment creating growth; its other priority is to redistribute wealth as well as income. While change is clearly necessary, a fundamental restructuring which shifts resources away from the elite creates a strong impetus to emigrate, particularly among the better-educated. This creates a delicate problem for the new government. Against the pressure for radical redistribution must be set the need to keep skills and savings within the domestic economy, and to allow sufficient wealth accumulation to attract foreign investors. This will inevitably slow down the rate at which redistribution can be achieved. In the longer run this may be better for the country's economic growth rate, as redistribution of wealth can retard economic growth (Alesina and Rodrik 1994). The problem is that the existing capital stock, which represents a sunk cost, was developed under a different development strategy, and may be inappropriate for the new outward-oriented, preferably labour-intensive growth which the government is seeking to encourage. If development is path-dependent, then the question remains as to how to shift on to a different growth path. Perhaps the following is worth noting. Compared with other upper middle-income economies, the structure of aggregate investment does not show significant distortion, with the exception of under-investment in human capital. The guideline for sustainable development is that the accumulation of human and produced capital must occur at the same rate as the depletion of natural capital. The government can work on improving human capital; indeed, this is a priority not only for growth but also for redistribution. Investment in physical capital is better handled by the private sector, and this requires the right policy environment. Macroeconomic stability and more liberal trade tend to emerge in regression analysis as linked with higher investment and faster growth. If this is the case, then the current economic strategy is probably the best available. This does not constitute a radical break with the past (as has occurred in some other middle-income countries with a history of ISI, like Mexico, Chile, Poland and former Czechoslovakia). However, destruction of the existing capital stock and institutions is costly, and current policy is nosing the economy towards a development strategy 17 In the recent World Economic Forum's Global Competitiveness Index, South Africa ranked 38 (out of 53).

256

C. Jenkins

which has generated faster growth in the resource-deficient developing countries. For South Africa this may, in fact, be the wisest approach. At present the government's energy is concentrated on overcoming the legacy of apartheid. Investing in human resources and facilitating private investment, while liberalizing the economy, may become increasingly attractive if its impact on the demand for unskilled labour is positive and if it generates faster growth. Slow reform may be more sustainable in the longer run.

1 5 . 6 . IMPLICATION S FOR OTHE R TRANSITION ECONOMIES It is difficult to draw generalizable conclusions from the experience of economies in transition. It is clear that the sustained pursuit of a largely orthodox macroeconomic policy strategy helped South Africa to avoid the extreme distortions which characterized other resource-abundant economies which suffered Dutch disease— although South Africa did not avoid the trap altogether. It is also clear that gradual liberalization has not generated faster economic growth as expected, although the conditions under which liberalization has occurred are to a large extent unique: economic isolation, the erosion of institutionalized racial discrimination, and widespread financial instability among emerging economies. What is not clear is whether growth would have been faster had adjustment occurred under more favourable external and domestic conditions. Nor is it clear whether the constraints of history are binding. Nor is it clear whether all that is needed is time. These are questions to which South Africa's experience cannot, as yet, provide answers.

Appendix 15.1 Year(s)

Policy change , importan t commissions, relevant event s

1925—72 1925 1948 1958

The period of import-substitution industrialization (ISI) Adoption of ISI with the Customs Tariff and Excise Duty Amendment Act Introduction of QRs Viljoen Commission recommends continued ISI, but using tariffs rather than QRs or subsidies SA government announces its intention to lift QRs under pressure from the GATT and the IMF, but does nothing

1969 1972—83 1972 1972 1972-76 1975-79 1978 1979-80 1983—91 1983 1983-85 1983 1983-85 1985

The first trade liberalization episode Reynders Commission recommends export promotion Export incentive measures are introduced Some relaxation of QRs The rand is devalued Further assistance to exporters introduced in line with the Van Huyssteen Committee's proposals Rand appreciates sharply

1990 1990 1991

The second trade liberalization episode Kleu Study Group recommends a move away from ISI The reduction of QRs is resumed The dual exchange-rate system is abolished The external value of the rand falls significantly Government white paper recommends a dual approach to industrial policy: ISI and export promotion Debt crisis; dual exchange-rate system re-introduced Substantial import surcharges introduced BTI begins to move proactively towards trade policy reform QR removal continues 'Structural adjustment' export incentives introduced for clothing, textiles, automobiles and automobile components General Export Incentive Scheme (GEIS) is introduced The phasing out of the import surcharge begins; not completed An accelerated depreciation tax scheme is introduced

Current 1994 1995 1995 1995 1995 1996 1997 1997 1999

The third trade liberalization episode The conversion of QRs to tariffs is completed Import surcharges are eliminated Tariff reduction in line with GATT requirements begins The financial rand is abolished Negotiations with the European Union over trade preferences commence The SADC free trade protocol is signed Further exchange control liberalization is announced GEIS removed and replaced with WTO-compatible export incentives Agreement reached with the EU on trade preferences

1985 1985 1987 1989 1989

258

C. Jenkins REFERENCES

Alesina, A. and D. Rodrik (1994), 'Distributive policies and economic growth', Quarterly Journal of Economics, 109/2, 465-90. Barro, R. and J. W. Lee (1994), 'Data set for a panel of 138 countries', available at http:// www.worldbank.org/html/prdmg/grthweb/. Bell, T. (1993), 'Should South Africa further liberalise its foreign trade?', in M. Lipton and C. Simkins (eds.), State and Market in Post-Apartheid South Africa, Johannesburg: Witwatersrand University Press. Collier, P. and J. W. Gunning (1994), 'Trade shocks: consequences and policy responses in developing countries', International Center for Economic Growth Occasional Paper 51. Collier, P. and J. W. Gunning (1999), 'Explaining African economic performance', Journal of Economic Literature, 37, 64-111. De Long, J. B. and L. H. Summers (1993), 'How strongly do developing countries benefit from equipment investment?', Journal of Monetary Economics, 32, 3; data available at http:// www.worldbank.org/html/prdmg/grthweb/. Deaton, A. (1992), 'Commodity prices, stabilization and growth in Africa', Center for International Studies Research Program in Development Studies Discussion Paper 166. Dollar, D. (1992), 'Outward-oriented developing economies really do grow more rapidly: Evidence from 95 LDCs, 1976-1985', Economic Development and Cultural Change, 40/2, 523-44. Easterly, W. and R. Levine (1997), 'Africa's growth tragedy: Policies and ethnic divisions', Quarterly Journal of Economics, 112; data at http://www.worldbank.org/html/prdmg/ grthweb/. Edwards, S. (1993), 'Openness, trade liberalisation and growth in developing countries', Journal of Economic Literature, 31/3. Gelb, S. (1991), 'South Africa's economic crisis: an overview', in S. Gelb (ed.), South Africa's Economic Crisis, Cape Town: David Philip. Hall, R. and C. I. Jones (1998), 'Why do some countries produce so much more output per worker than others?', Stanford University; data available at (http: //www-leland.stanford.edu/~chadj/), (mimeo). Holden, M. (1990), 'The choice of trade strategy: past reflections and future prospects' in N. Nattrass and E. Ardington (eds.), The Political Economy of South Africa, Cape Town: Oxford University Press. Holden, M. and P. Holden (1978), 'Effective tariff protection and resource allocation: a nonparametric approach', Review of Economics and Statistics, 60/2. Industrial Development Corporation (IDC) (1990), Ondersoek na die tarief heskermings-beleid: ontleding van die tariefstruktur, analytical document and five annexes, Sandton. Krishna, K., A. Ozyildirim and N. Swanson (1998), 'Trade, investment and growth: nexus, analysis and prognosis', NBER Working Paper 6861. Levine, R. and D. Renelt (1992), 'A sensitivity analysis of cross-country growth regressions', American Economic Review, 82, 4; data at http://www.worldbank.org/html/ prdmg/ grthweb/. North, D. (1995) 'The new institutional economics and Third World development', in J. Harriss, J. Hunter and C. Lewis (eds.) The New Institutional Economics and Third World Development, London: Routledge, 17—26. Ranis, G. (1991), 'Towards a model of development', in L. Krause and K. Kihwan (eds.), Liberalization in the Process of Economic Development, Berkeley: University of California Press.

Economic Reform in South Africa

259

Republic of South Africa (RSA) (1985), White Paper on Industrial Development Strategy, Pretoria: Government Printer. Rodriguez, F. and D. Rodrik (1999), 'Trade policy and economic growth: a skeptic's guide to cross-national evidence', NBER Working Paper 7081, Cambridge, MA. Rodrik, D. (1998), 'Trade policy and economic performance in sub-Saharan Africa', NBER Working Paper 6562, Cambridge MA. Sachs, J. and A. Warner (1997), 'Sources of slow growth in African economies', Journal of African Economies, 6/3: 335-76; data at http://www.worldbank.org/html/prdmg/grthweb/. Sala-i-Martin, X. (1997), 'I just ran two million regressions', American Economic Review Papers and Proceedings, (May) 87/2, 178-83. Scheepers, C. (1969), 'The effect of import substitution on the volume and structure of South Africa's imports, 1926/27—1963/64', Finance and Trade Review. Thomas, W. (1990), 'Unemployment and the job creation challenge', in R. Schrire (ed.), Critical choices for South Africa: an agenda for the 1990s, Cape Town: OUP. Wood, A. (1994), North-South Trade, Employment and Inequality, IDS Development Studies Series, Oxford: Clarendon Press. World Bank (1999), Environmental Economics and Indicators Unit data available at http://www-esd.worldbank.org/eei/.

16 Reforming Resource-Abundant Transition Economies: Kazakhstan and Uzbekistan R I C H A R D M . AUTY

1 6 . 1 . INTRODUCTIO N In the mid-1990s, comparative studies of the economies in transition strongly suggested that macro policy determined successful adjustment (de Melo et al. 1996; Fischer et al. 1996; Aslund et al. 1996). Table 16.1 shows why. It summarizes the performance of the five categories of reformers over the period 1989-94 and suggests that faster reform of macro policy is preferable to gradual reform because faster reformers experience less loss of output and government revenue than slower reformers, and that they resume economic growth sooner. This approach to transition reform was championed by the international financial institutions and it is termed here the rapid reform model. It rests upon three central premises. First, rapid stabilization and tight inflation control are necessary prerequisites for sustained economic growth. Second, renewed growth also requires prompt realignment of internal and external prices as well as low fiscal deficits. Moreover, price reform calls for privatization and improved financial regulation. Third, and finally, the pace of economic recovery is linked to the comprehensiveness of reform rather than to a recovery in the level of investment. However, more recent data cast doubt upon the ability of the rapid reform model to explain the differing transition outcomes. For example, the intermediate reformer Kazakhstan has not achieved the economic rebound predicted by the model whereas Uzbekistan, a very slow reformer, has not experienced the accelerating growth collapse that was expected. A key omission from the rapid reform model is the impact of the initial conditions on the transition, notably the natural resource endowment, social capital, and the central planning legacy of economic distortions, especially obsolete produced capital (de Melo et al. 1997; Havrylyshyn et al. 1998; Kenny 1999). A principal components analysis (de Melo et al. 1997) of the initial conditions reveals two key parameters for transition performance, namely, the degree of policy distortion and the extent of over-industrialization. These two parameters identify four clusters of transition economies. The two sets of successful reforming countries exhibit

Table 16.1 . Liberalization, grnwth, and inflation in transition economies 1989-1994

Reform stage/

Per capita

country

cropland (ha)

Liberalization index

(A)

(B)a

Average inflation 1993-94 (%)

Average GDP growth 1993-94 (%)

Ratio of

GDP trough

1994 GDP to 1989 GDP (%)

(as of % 1989)

Advanced" High-intermediate0 Low-intermediate11 Slow0 Otherf

0,33 0.52 0.95 0.46 0.23

3.90 2.55 1.67 0.90 2.11

0.85 0.72 0.55 0.27 0.50

23 124 933 1,968 3,102

1.7 0.03 -15.6 -13.2 -14.5

83 67 57 72 45

80 65 51 66 34

Kazakhstan Uzbekistan China'1 Vietnam11

2.08

1.31 1.11 1.60 1.80

0.37 0.37 n.a. n.a.

1,870 640 19 8

-18.5 -2.5 12.6 8.4

57 89 166 139

49 88

0.20g

0.08 0.09

—

Notes: "Accumulative reform index 1989-94; B = Average liberalization 1993-94. b Slovenia, Hungary, Poland, Czech Republic, and Slovak Republic. "Baltic States, Bulgaria, Albania, Romania, and Mongolia. ^Russian Federation, Kyrgyz Republic, Moldova, Kazakhstan. "Uzbekistan, Belarus, Ukraine, Turkmenistan. tountries affected by regional tensions (Croatia, Macedonia, Armenia, Georgia, Azerbaijan, Tajikistan). s Since 1970, rapid population growth halved this figure, which is also sensitive to increased irrigation capacity. Additional renewable and mineral resources place Uzbekistan within the resource-abundant category. ^Liberalization index based on World Bank (1996), 14. Source: de Meio et al. (1996), 405, except WRI (1998), 286-7 for per capita cropland.

262

R. M. Auty

less initial policy distortion than the two less successful groups. They are also relatively resource poor. They comprise the economies in transition of Northeast Europe and East Asia. The two less successful groups of reformers (Central Asia and then the Baltic States, Russia and part of the Caucasus) exhibit all three features that the rapid reform model neglects, namely: • natural resource abundance (cropland per hectare in excess of 0.55 compared with a mean of 0.30 for the other two clusters (WRI 1998)), • deficient institutional capital (4.2 on an index of legal safeguards compared with 7.7 for eastern Europe (Johnson et al. 1997)) and • a highly distorted economy so that many firms retreat into the unofficial economy (estimated at 32.4 per cent of official GDP in the former Soviet Union (FSU) compared with 19 per cent for eastern Europe (Johnson et al. 1997)). Among the developing market economies these three features are associated with a high propensity for a growth collapse. This chapter shows how the predictive power of the rapid reform model is improved by taking account of the impact of initial conditions on the transition, with particular reference to natural resource abundance. It draws upon a longer review of the literature (Auty 1999a). That literature suggests that resource abundance can have two adverse effects. First, the rents can diminish the urgency of reform and feed corruption (Frye and Shleifer 1997; Lane and Tornell 1996; Leite and Weidmann 1999). Second, resource abundance can amplify the contraction of the non-booming tradable sector due to Dutch disease effects (Rosenberg and Saavalainen 1998; Sachs and Warner 1995). This promotes a dual economy that heightens income inequality and exacerbates social tensions because the nonresource tradables sector is a particularly important source of non-government employment and skill provision. These adverse features of resource abundance for the transition are likely to be more severe where the resource endowment creates point socioeconomic linkages, as in mining, as opposed to the diffuse linkages associated with crop production by smallholders. This chapter examines the impact of natural resources on transition reform with reference to two mid-sized resource-abundant economies, Kazakhstan and Uzbekistan. The structure of the chapter is as follows. The next section briefly describes the pace of reform in the two case study countries and compares their transition trajectories with the predictions of the rapid reform model. Section 16.3 analyses whether the hesitant recovery of Kazakhstan can be explained by its large mineral endowment. Section 16.4 evaluates alternative theories for the shallow transition recession of Uzbekistan. Section 16.5 summarizes the conclusions and draws the policy implications. 1 6 . 2 . PREDICTE D AN D ACTUA L TRANSITIO N TRAJECTORIES: K A Z A K H S T A N A N D U Z B E K I S T A N The rapid reform model predicts that late reformers will experience a steep decline in GDP and government revenues before economic growth resumes. If the natural resource endowment is factored in, then in the case of Kazakhstan, foreign investment

Transition Economies: Kazakhstan and Uzbekistan

263

in mining will tend to accelerate the post-transition rebound of the real exchange rate. This, in turn, will intensify the demise of the nonresource tradable sector, raise income inequality, and strengthen pressure to maintain public sector employment. In addition, corruption may be substantial due to the concentrated rent stream and point socioeconomic linkages of mining, and this will further depress the rate of economic growth. In contrast, crop-driven Uzbekistan should experience a more broad-based recovery, but due to its even tardier espousal of reform, it will be expected to do so only after a deeper and more prolonged recession. Deviation from these predictions will imply that even after factoring in the natural resource endowment, the rapid reform model is still too narrowly calibrated.

16.2.1. Differences in the speed of reform Kazakhstan and Uzbekistan are both late reformers. But Kazakhstan, after initially lagging behind Uzbekistan, accelerated the pace of reform from 1994 whereas Uzbekistan continued at a slow speed and subsequently reversed some reforms. By 1995, the level of reform on the liberalization index averaged 0.63 for the Baltics, Russia, and other FSU countries (BRO) as a group compared with 0.60 for Kazakhstan, while Uzbekistan trailed with an index of 0.50 (Table 16.1). Price liberalization proceeded faster than privatization in Kazakhstan and was essentially completed in 1994. All export and import licensing schemes were removed in 1995 and financial reform began with a reduction in interest rates and confirmation of central bank independence. Compared with Kazakhstan, the reform policy of Uzbekistan has been even more cautious. Pomfret and Anderson (1997: 20) characterize Uzbek reform as 'inconsistent gradualism'. Price reform is partial: half or more of wheat and cotton production is purchased at official prices that are set well below world levels. In addition, both the interest rate and the exchange rate continue to be artificially set and the black market premium on the national currency exceeded 100 per cent from the mid-1990s (Pomfret and Anderson 1997). Finally, some of the reforms were reversed after the IMF agreement broke down. For example, average tariffs rose from 12 to 17 per cent in 1997. The government of Uzbekistan has attempted to follow the Chinese model of agricultural reform by allowing private land use rights while retaining formal state ownership of land and operating a dual price system. The state farms were collectivized by 1994 and the land allotted to private plots rose. Most cooperatives adopted the Chinese household responsibility system whereby individual farmers contracted to produce a given volume at a fixed price and were free to sell any surplus and retain the revenues. Outside of agriculture, small and medium-sized firms were privatized but they were subject to tight regulation and heavy taxation. Overall, barely one-fifth of the large and medium-sized SOE factories have been privatized and the remainder benefit from monopolistic markets, subsidized energy, and preferential access to foreign exchange and credit on favourable terms. The Uzbek development strategy transfers resources from agriculture and the nonfarm private sector to protected industry and social entitlements. The entitlements

264

R. M. Auty

are sustained at more generous levels than elsewhere in the FSU: pensions are maintained while schooling remains free to all and medical services have been minimally affected by the collapse of the FSU (Pomfret and Anderson 1997). Meanwhile, the economic disruption of the transition is also cushioned by the extended family and the retention of the historical system of local headmen, which help sustain a relatively equitable distribution of income (Akhad et al. 1995), if not of wealth. Although these institutions sustain social cohesion, they also impede the enhancement of social capital as far as the transition to a market economy is concerned. This is because the surviving institutions forestall the need to create new businesses and to develop commercial skills in the informal sector.

16.2.2. Differences

in the transition trajectories

Table 16.1 shows that the GDP growth rate of Kazakhstan conformed to the pattern associated with relatively slow reformers up to the mid-1990s, whereas Uzbekistan had a much better performance than would be expected of a slow reformer. Revised figures show that the decline in Kazakhstan's output between 1991 and 1998 bottomed out at a level one-third below its 1991 level. But the decline for Uzbekistan was only 16 per cent before the growth rate turned mildly positive, a fall that is smaller than for any other transition economy (Taube and Zettelmeyer 1998: 4). In addition, whereas government revenues collapsed to barely 16 per cent of GDP in the case of Kazakhstan, they stabilized at twice that level in Uzbekistan. One possible explanation for the more buoyant performance of Uzbekistan is inaccurate data because the statistics do show a systematic overstatement of its economic output. Furthermore, the contribution of this factor towards giving the appearance of a milder contraction may be heightened by the much smaller share of the informal sector in the GDP of Uzbekistan. Johnson et al. (1997: 182) estimate from adjusted electricity consumption that the level of activity in the unofficial sector halved in the case of Uzbekistan to around 6 per cent of total GDP over the years 1989—95. In contrast, the figure for Kazakhstan tripled to almost one-third of total GDP, very close to the average for the BRO countries (Taube and Zettelmeyer 1998). The low Uzbek figure results from the subsidization of large SOEs and consequent maintenance of employment, and also from more repressive state control. The use of data for electricity consumption as a proxy for output does narrow the gap in output decline between Uzbekistan and the comparator countries, compared to the GDP measure. Nevertheless, the fact remains that Uzbekistan still has the smallest decline in electricity consumption and the lowest cumulative deterioration in economic activity so that some of the difference in performance must be explained by factors other than statistical inaccuracy. There are two leading explanations for the anomalous behaviour of Uzbekistan. First, a favourable resource endowment reduced the urgency for reform (Taube and Zettelmeyer 1998). Second, Uzbek policy promotes economic diversification by protecting manufacturing from import competition (the explanation favoured by the government of Uzbekistan). The actual transition trajectories of Kazakhstan and Uzbekistan are now compared with the

Transition Economies: Kazakhstan and Uzbekistan

265

predictions of the rapid reform model, adapted to incorporate the natural resource endowment. 16.3. T R A N S I T I O N WITH POINT SOURCE R E S O U R C E S IN K A Z A K H S T A N

16.3.1. The severe transition recession and entrenchment of corruption Although Kazakhstan sustained very high levels of investment from the early-1970s this did not offset a long-term accelerating deterioration in capital efficiency and labour productivity so that growth collapsed. Nevertheless, Kazakhstan became the most prosperous republic in the otherwise lagging Central Asian region (de Melo et al. 1997). To that extent, it has more in common with the FSU countries to the north and west, with their greater legacies of obsolete industrial capital. The industrial sector was already considerably larger than the agricultural sector by the early-1990s (Table 16.2). Both the key tradable sectors, agriculture and industry, initially contracted faster than the economy as a whole so that between 1990 and 1995, the trough of the downswing, the share of manufacturing in GDP fell by more than half and that of agriculture by two-fifths (De Broeck and Kostial 1998: 19). These relative declines are much deeper than those experienced by Uzbekistan. Moreover, when subsidies were withdrawn in the mid-1990s, the output of agriculture declined slightly faster than that of industry. By 1997, agriculture generated only 11 per cent of GDP in Kazakhstan (but still absorbed two-fifths of the workforce) compared with 26 per cent for Uzbekistan (Table 16.2). The collapse in government revenues in Kazakhstan (to a meagre 16 per cent of GDP by 1997) was even sharper than the collapse in its output of tradables. Even then, the low figure of 16 per cent of GDP includes privatization receipts equivalent to 3.4 per cent of GDP. Meanwhile, despite the privatization windfall, the budget deficit in 1997 was 3.8 per cent of GDP, of which almost two-thirds was funded by foreign borrowing. Consequently, given the role played by privatization receipts, the underlying structural deficit was closer to 7 per cent of GDP, or one-third of public Table 16.2. Sectoral composition of GDP, Kazakhstan and Uzbekistan 1993-1997 (% GDP) Kazakhstan

Agriculture Industry Construction Transport and communications Trade and catering Other (mainly services)

Uzbekistan

1993

1997

1993

1997

16.4 28.7 8.3 10.0 10.4 34.6

10.9 20.4 4.1 10.5 17.0 54.1

27.9 22.4 9.0 6.7 10.4 36.6

25.89 16.5 8.1 6.3 17.0 54.1

Source: IMF (1998): 29; Taube and Zettelmeyer (1998): 19.

266

R. M. Auty

expenditure. This created acute problems in maintaining social entitlements. Income inequality rose sharply (Becker and Urzhumova 1998). However, a further rise in tax rates in order to maintain social transfers is precluded because compliance is already a problem due to the relatively high rates set by the reforms (EBRD 1996). In this context, many politicians are tempted to postpone hard political decisions and also consider how they and their associates may personally take advantage of the fact that Kazakhstan may have sufficient energy resources to become one of the ten largest global oil producers. Table 16.3 presents three scenarios for Kazakhstan oil rents, based on estimates of production costs by Planecon (1997) and oil prices by the World Bank (1998). It shows that under the most optimistic price/output combination, the oil revenues could reach US$4.75 billion, equivalent to 23.8 per cent of Kazakhstan GDP. Such a windfall is more than double that which, relative to its GDP, Venezuela received during the 1974-78 and 1979-81 oil booms, and slightly higher than the windfalls that Indonesia and Nigeria received at that time (Gelb and Associates 1988). However, the 'middle' scenario in Table 16.3 is more consistent with the long-term forecasts of the World Bank (1998). It yields a much more modest rent stream that is equivalent to only 3.2 per cent of GDP. The large difference between the two scenarios is due to the fact that the rent is reduced disproportionately by lower oil prices because it is a residual after meeting all costs, including the risk-related return on capital (Auty 1998). The Caspian region is a relatively high cost oil producer on account of high overhead costs, including transportation. Even so, the medium price scenario would still boost public revenues by one-fifth, if the level of public expenditure is assumed to stabilize at 16 per cent of GDP. Yet there is also the prospect of no rents through the medium-term if the third scenario proves more accurate. The low price scenario raises the unpalatable prospect that Kazakhstan will see its real exchange rate Table 16.3. Projected Kazakhstan oil rent stream, three scenarios

North Sea oil price ($/bl) Caspian oil price ($/bl)a Netback/rent ($/bl) Total rent ($ billion) 0.5 million bpd 1.5 million bpd 2.5 million bpd Total rent (% GDP)b 0.5 million bpd 1.5 million bpd 2.5 million bpd

Low

Medium

High

12.00 10.00 0.00

16.00 14.00 1.15

20.00 18.00 5.15

— — —

0.210 0.630 1.049

0.940 2.820 4.699

— — —

1.1 3.2 5.2

4.7 14.1 23.5

Notes: a Discounted by $2 per barrel from North Sea price. b GDP = $20b at 1997 prices (assumes 5%/yr growth to 2020). Source: Auty (1998).

Transition Economies: Kazakhstan and Uzbekistan

267

Table 16.4. Social capital indices, by natural resource endowment group

Group Governanc

index (

e Stat e intervention Corruptio n Brib e tax % frequency) indexa ( % revenue)

Central and East Europe 1.48 Mineral-rich FSU 1.45 Cropland-rich NFSU and 1.22 South-east Europe

23.2 39.7 30.8

2.3 7.4 6.1

3.3 5.3 4.8

Resource-poor SFSU Kazakhstan Uzbekistan

50.0 23.7 46.6

8.5 6.5 7.0

6.7 4.7 5.7

1.27 1.27 1.83

Source: EBRD (1999): 123 and 125, except a EBRD (1997): 39.

recover due to heavy investment in the oil sector which will depress non-oil tradable activity without any appreciable offsetting improvement in government finances. There is anecdotal evidence that ministers have held back on politically difficult decisions because they expect an oil boom to resolve fiscal problems. There is also speculation that large sums of public money (from contract signature bonuses, for example) have been siphoned into private accounts, consistent with the exploitation of point resources under inadequate public sector accountability. Further evidence of corruption and deficient institutional capital emerges from the scale of the collapse in government revenue and the sharp expansion of the unofficial economy to one-third of official GDP (Johnson et al. 1997). Finally, Kazakhstan languishes with Azerbaijan and Georgia at a low level on indices of tax fairness, a measure of corruption, and legal safeguards (Simon et al. 1997). However, Table 16.4 suggests that perceived corruption is even worse in the resource-poor FSU countries, although a direct comparison is misleading because civil strife in three out of the four resource-poor FSU countries retarded reform and severely weakened governance and social capital. Even if the postponement of the oil windfall should lower the opportunity for highlevel corruption, corruption may still adversely impact the economy. This is because a continued squeeze on state revenues will perpetuate low salaries that encourage government officials to impose illicit imposts on private firms, which imposts depress investment (Auty 1999b). In order to discourage public sector workers from abusing their positions, the civil service must be cut in size and its remuneration improved. But this also requires an expansion of labour-intensive employment in the private sector (Becker and Urzhumova 1998) in order to provide alternative opportunities for displaced government workers. This is proving especially hard to achieve in mineral-rich Kazakhstan, as discussed below.

16.3.2. The real exchange rate rebound and the faltering Kazakh recovery Investment in the Kazakh oil sector may be partially responsible for the faltering economic recovery. Capital inflows into Kazakhstan, as well as into oil-rich Azerbaijan

268

R. M. Auty

and gas-rich Turkmenistan, are large compared to other transition economies within the CIS. For Kazakhstan, the inflow in the mid-1990s was equivalent to 5.6 per cent of GDP annually (Rosenberg and Saavalainen 1998: 22). This foreign investment has been associated with a faster rebound in the real exchange rate of Kazakhstan than in the resource-deficient economies in transition. Kazakhstan's real exchange rate has recovered much closer to its equilibrium level than in the less resource-abundant CIS countries. By 1998 the dollar wage in Kazakhstan was $120 compared with $12-40 in resource-poor Tajikistan, Georgia, Armenia, and Kyrgyzstan, $50-55 in resource-rich Uzbekistan and Turkmenistan and $200 for Russia on the eve of the August 1998 financial crisis (Valdevieso 1998: 9, Gurgen et al. 1999). Significantly, economic growth faltered in 1998 after having turned positive in 1996 and reaching 2 per cent in 1997 (IMF 1998). Yet, Kornelia and Zettelmeyer (1997) argue that the real exchange rate does not appear to have overshot its post-transition equilibrium. For example, Kazakhstan exporters succeeded in expanding sales to the rest of the world in the face of the collapse of traditional export markets in the former Soviet Union, an outcome that does not suggest any lack of competitiveness. The economic slow-down in Kazakhstan in 1998 appears to have more to do with unexpected external shocks (falling oil prices and collapsing CIS markets) than with any systematic deterioration in external competitiveness. However, this conclusion ignores the need to diversify into more labour-intensive exports for social reasons and in order to diversify and strengthen the economy, and thereby avoid falling into a staple trap. In fact, Kazakhstan in common with other mineral-rich FSU economies experienced a redistribution of employment 1990-97 that is consistent with Dutch disease effects. Table 16.5 compares actual employment shifts with those expected for a country of similar size and level of development. Resource-poor Kyrgyzstan experienced a surge of labour into agriculture that boosted the employment share of the tradable sector above the level expected by 1997. However, Kazakhstan had a minimal shift of employment into agriculture and its tradable sector, overall in 1997 employed one-fifth less of the workforce than expected (compared with one-quarter less than expected for mineral-rich Russia). Interestingly, crop-rich Uzbekistan slightly exceeded its expected tradable employment share.

16.3.3. Social costs of uneven structural change in the Kazakh economy The steep decline in tradable output during the transition created a two-sector economy in Kazakhstan. A well-funded, largely foreign-owned export sector contrasts with a capital-starved domestic sector that withholds wages, pensions, and utility payments in order to sustain its cash flow (Financial Times 1998). The fragile economic recovery of 1996-98 reflects the rebound of HCI as a result of successful privatizations that include the 6 million tonne Ispat-owned integrated steel plant and the 250,000 tonne Samsung-owned copper refinery. The fuel sector and metallurgy sector each generate about 25 per cent of industrial output. Oil and gas also provide

Transition Economies: Kazakhstan and Uzbekistan

269

Table 16.5, Dutch disease effects 1990-1997 (actual c.f. expected employment share) Index

1990

1997

Departure from norm in 1997

Agriculture Industry Distortion

0.33 0.28 +0.10

0.48 0.14 +0.06

+0.14 -0.08

Agriculture Industry Distortion Agriculture Industry Distortion

0.22 0.31 +0.05 0.13 0.42 +0.03

0.25 0.18 -0.08 0.14 0.32 -0.09

-0.03 -0.05

Agriculture Industry Distortion

0.39 0.24 +0.08

0.40 0.19 +0.02

Resource poor

Kyrgyzstan

Mineral rich Kazakhstan

Russia

-0.17 +0.08

Crop rich

Uzbekistan

+0.05 -0.03

Source: EBRD (1999).

one-third of exports and nonfuel minerals supply two-fifths. Sharp recoveries are also expected in alumina and gold. Trade reform has initially realigned the economy towards primary production, especially minerals and mineral processing. However, both mining and HCI are highlycapital-intensive so that other sectors need to expand direct and indirect employment. The experience of oil-rich countries like Saudi Arabia suggests that such employment should occur in the private sector and not the public sector (see Chapter 12). The additional employment that is required seems initially likely to come from services rather than from either non-mineral manufacturing or agriculture. Although diversification into manufacturing should eventually occur, there is likely to be a lag for the reasons discussed in Chapter 14. For example, a vigorous expansion of food processing in Kazakhstan (IMF 1998: 30) was offset by continued sharp contractions in other subsectors. The latter include chemicals, fabrics, farm machinery, white goods, and electrical goods. Their output declined from FSU levels from two-thirds (machinery and metalworking) to seven-eighths (light industry), compared with one-third for minerals and metallurgy. Unfortunately, the medium-term employment prospects are poor in agriculture as well as in mining and manufacturing. Although Kazakhstan has one of the highest endowments of cropland per capita among the developing countries (Table 16.1), only 6 per cent of the cropland is irrigated. Moreover, cropland was over-extended under central planning, especially in the grain-growing northern region so that it required sizable subsidies, estimated at 10-12 per cent of GDP (De Broeck and Kostial

270

R. M. Auty

1998: 40). The withdrawal of those subsidies due to the reforms made some contraction in output inevitable, albeit not as great as the 40 per cent fall from 1990 levels that occurred. The sharp decline in production is due to the failure of private rural credit to emerge and an associated reduction in the supply of farm inputs. For example, inputs per hectare of grain fell by almost two-thirds and yields dropped to 0.7 tonnes per hectare. Yet, although grain output should recover to 20 million tonnes, agriculture will not employ more than a fraction of those now dependent on it. Agriculture has functioned as a labour reservoir during the transition recession because it has absorbed workers, albeit less than in the non-mineral economies. But although subsistence farming may have alleviated rural poverty, the 40 per cent of the population presently dependent on the rural economy require rapid growth in nonfarm jobs. Given the surplus labour in the nonfarm public sector, private services will initially have to provide most employment in Kazakhstan. This will have two adverse consequences. First, it will maintain the heavy specialization of the tradable sector in mining (and enhance the vulnerability of the economy to external shocks and corruption). Second, it will heighten income inequality due to the capital—intensive nature, and associated high wages, of mining and HCI. The mining sector therefore acts as a double-edged sword in the transition of Kazakhstan: it has driven the relatively rapid appreciation of the real exchange and when this is combined with diminished expectations of mineral rents, the prospect for economic recovery through the medium-term is disappointing. These conclusions do not invalidate the original rapid reform model, but they do suggest that its predictions of economic recovery can be improved by explicitly considering the natural resource endowment.

16.4. RESOURC E ABUNDANCE , MACR O POLICY , AN D THE UZBEKISTA N TRANSITIO N ANOMAL Y 16.4.1. The misguided development strategy of Uzbekistan The Uzbek government attributes the greater resilience of the Uzbekistan economy compared with that of Kazakhstan to subsidizing manufacturing within a policy of gradual reform. De Melo et al. (1997) acknowledge that low-income economies like those of East Asia may be able to sustain a gradual reform process with rapid economic growth. Uzbekistan may at first seem to qualify because it was one of the poorest states in the FSU with a per capita income barely half that of Kazakhstan on the eve of the transition. Cotton and gold together provided 30 per cent of GDP and 60 per cent of exports in the early-1990s (Taube and Zettelmeyer 1998). However, de Melo et al. (1997) attribute the East Asian success to relatively low levels of economic distortion and a cultural edge, both of which Uzbekistan lacks. This is because the very low East Asian income levels mean that agriculture remains overwhelmingly dominant and it has proved sufficiently flexible to rapidly expand output and release a large stream of labour for competitive manufacturing. Transfers to the relatively small industrial

Transition Economies: Kazakhstan and Uzbekistan

271

sector were not on a scale to depress incentives within this dynamic and competitive tradable sector. Zettelmeyer (1998) uses a regression model in order to assess the relative importance of economic policy and resource endowment, respectively, in explaining the mild transition recession in Uzbekistan. He finds that agricultural production (and cotton exports in particular) is strongly and positively associated with the modest nature of the GDP decline in Uzbekistan (Table 16.2). The country is the world's third largest exporter of cotton. Buoyant revenues from cotton (along with those from gold) helped to sustain government: finances after independence in the face of a loss of transfers from the Soviet Union, estimated at 12 per cent of GDP. Agriculture also helped to cushion the adjustment to the transition by increasing its share of employment as a result of a net migration to rural areas. The share of the rural population rose slightly to 61 per cent, even as rural wages dropped from 15 per cent above the national average to one-third below the national average (Pomfret and Anderson 1997). The policy of subsidizing industry may sustain output over the short- and mediumterm, but the staple trap model shows that it has two adverse consequences for longterm economic growth. First, it misallocates resources by transferring revenue in excess of the resource rents from the potentially efficient primary sector to less efficient activities. Second, the transfer of a sizable fraction of the resource rents to the government voids the potential advantage of the diffuse socioeconomic linkages that crop-led developing countries possess compared with mineral-led ones. That advantage arises from a wider dispersal of rents through the economy and consequent reduced vulnerability to predatory government behaviour.

16.4.2. Repressing the buoyant primary sector Uzbekistan attempted to emulate the Chinese model of agricultural reform, but in contrast to the dynamism that such reform unleashed in China, agricultural production contracted in Uzbekistan. This is because the local elite retained more control in Uzbekistan than it did in China and the central government also came to depend heavily on farm revenue. The continued manipulation of major commodity prices and regulation of the exchange rate by the central government severely distort primary commodity prices for domestic producers (Pomfret 1998). In the late-1990s the government secured almost one-third of its revenue from excise duties (possibly onehalf of this from cotton). One result of the government policy is to depress rural incomes by more than the overall decline in economic output, thereby severely reducing incentives to producers. A second consequence of government policy is that the share of trade in Uzbek GDP halved to 26 per cent between 1994 and 1997, whereas in Kazakhstan it remained stable at 31 per cent of GDP (EBRD 1998). The transfers from primary product exports were not just directed into protected industry, however, they were also targeted at boosting food self-sufficiency. Yet it is far from clear that this is in line with the comparative advantage of the Uzbek economy. Some 360,000 hectares of land were transferred from cotton to grain (predominantly wheat), implying a substantial loss of cotton exports (World Bank 1998).

272

R. M. Auly

Grain production was also expanded onto marginal land, including 490,000 hectares (30 per cent of the land under grain) that was not irrigated and is subject to drought. Yet, despite the extra resources, grain output in 1996 was barely three-fifths the target and met only half the country's requirement. The expansion of grain farming is a rnisallocation of resources. Self-sufficiency is not undesirable per se: Uzbekistan did benefit from selfsufficiency in energy production. The expansion of oil and gas production is at least in a commodity in which the country might expect to hold some comparative advantage. Moreover, by maintaining energy self-sufficiency, Uzbekistan minimized the disruption that other countries experienced that remained dependent on energy imports from the FSU. Increased energy production freed Uzbekistan from its dependence on Russian imports and in 1995 energy accounted for 15 per cent of exports, just under one-quarter the size of cotton exports. Zettelmeyer (1998) confirms that energy selfsufficiency had a significant positive effect on economic growth, although he finds the export of energy is negatively correlated with economic growth because of barter deals and nonpayment by importers. He concludes that the combination of buoyant cotton revenues and energy self-sufficiency relieved the import compression that other transition economies experienced and that these two factors are key causes of the mild transition recession in Uzbekistan.

16.4.3. Cosseting the manufacturing sector risks a growth collapse Zettelmeyer (1998) finds that investment in industry is particularly weak as an explanatory variable in his growth model, whereas those modest macro reforms that were achieved had a beneficial impact on the recovery. In other words, the recovery of the Uzbek economy occurred despite the industrial policy that the government espoused. Although manufacturing attracted some foreign direct investment, notably from Daewoo and BAT, most investment came from the state and was directed at large SOEs. The scale of government promotion of industry is substantial, amounting to three-fifths of national investment in 1996. Most large SOEs were captured by their managers who secured access to low-interest loans and also enjoyed soft budgetary constraints and little competition. This system also helps to account for the relatively high perceived corruption index (Table 16.4) despite the fact that the Uzbek government experienced less revenue loss and retained greater policy autonomy than many FSU economies. Yet despite the huge transfer of resources to the manufacturing subsector (and the expansion in energy production), Table 16.2 shows that a significant decline in the share of industry in the GDP of Uzbekistan still occurred. The industrial policy had some success in expanding the production of autos, television sets, and video recorders with foreign investment (Taube and Zettelmeyer 1998: 20). Elsewhere, the production of bulky HCI products such as fertilizer, ferrous metals, and cement was sustained by subsidies. Most other sectors recorded declines, however, despite state assistance. The products made by the SOEs were often of low quality and proved difficult to sell (EIU 1996). Meanwhile, the SOE factories tend to be capital-intensive and to provide little

Transition Economies: Kazakhstan and Uzbekistan

273

employment. Yet a rise in employment is required in order to absorb a workforce that is set to grow rapidly, given that the rate of population growth is 2.6 per cent. The working age population is only 48 per cent of the total and the worker/dependency ratio remained high (Table 2.4). The staple trap model suggests that resources will continue to be misallocated in the absence of hard budget constraints and price signals that reflect scarcity value. Yet Uzbekistan manufacturing should not require subsidies. It has the potential to serve regional markets for manufactured goods. This is because it offers investors the region's largest national market and the rebound in its real exchange rate will lag those of its mineral-rich neighbours like Kazakhstan. Estimates of the US dollar wage level for Uzbekistan suggest that it remains at barely one-third of its equilibrium level, whereas that of Kazakhstan is already close to parity. However, by postponing reform and diverting resources from sectors with competitive advantage to protected sectors that have little incentive to become efficient, current policy is likely to result in a growth collapse and a further decline in real wages. It will require a more vigorous expansion of the private sector if the economic potential of Uzbekistan is to be realized and rapid employment-intensive growth is to be sustained.

1 6 . 5 . CONCLUSION S AND POLIC Y IMPLICATION S The rapid reform model stresses macro policy and neglects differences among the transition economies in their endowments of natural resources, social capital, and obsolete produced capital. The incorporation of the natural resource endowment into the rapid reform model improves its explanatory power. Consistent with the extended model, both Kazakhstan and Uzbekistan delayed reform; compared with the more resource-constrained countries of Eastern Europe and East Asia the two governments exhibit higher levels of rent-seeking behaviour and corruption (Table 16.5). In the case of Uzbekistan, a natural resource endowment that yielded buoyant crop revenues to ease the foreign exchange constraint helps to explain why the growth collapse and the loss of government revenue are less than the unadjusted rapid reform model predicts for such a slow reformer. To that extent, the transition to date has indeed proved easier in a lower-income crop-driven resource-abundant country than in mineral-driven Kazakhstan or Azerbaijan (Auty 1999b). This explanation for the Uzbek anomaly is still too simple, however, because Uzbekistan also benefits from more robust social capital and a smaller legacy of obsolete industry. These distinctive attributes combine with favourable export prices to retard the decline in government revenue and reduce the apparent urgency of reform. The experience of Uzbekistan therefore also confirms that a favourable natural resource endowment may be used to postpone difficult reforms, when the stock of obsolete capital is low. Yet unlike the resource-poor gradual reformers of East Asia, Uzbekistan is not likely to escape major reform. The resource-constrained rapid reform model suggests that the Uzbek policy of gradual reform, unlike gradual reform in East Asia, intensifies economic distortions and represses the export sector. This locks the economy into a staple trap that the developing market economies suggest will

274

R. M. Auty

lead to a growth collapse. Ironically, manufacturers in Uzbekistan should not require state assistance if they take advantage of a lower exchange rate and large home market to export manufactures and farm products to their mineral-rich neighbours. Investment in minerals in Kazakhstan has strengthened the exchange rate and retarded economic diversification so that economic recovery is slower than the original rapid reform model predicts. However, although corruption is high, it is no worse than in other FSU economies. More specifically, heavy foreign investment in mining and HCI pushed the economy of Kazakhstan much closer to its equilibrium wage than in most other transition economies. The diversification of the economy into more employment-intensive activity is constrained and Kazakhstan has endured a sharp contraction in output. The result is a dual economy in which the potentially employment-intensive sectors lack investment. Moreover, the front-loading of the capital recovery stage of the HCI investment means that the government of Kazakhstan may yet experience the unpalatable combination of a relatively high real exchange rate, weak economic growth, and rising income inequality without compensation from higher mineral revenues. Yet the appreciation of the real exchange rate need not retard growth through the medium-term and long-term because any unwelcome appreciation can be curbed by sterilizing the inflows and increasing the rate of domestic saving, thereby lowering the demand for foreign capital inflows. This policy has the added benefit of making funds more readily available to a wider range of domestic investors than is the case with foreign savings. Such a pattern of investment will not only boost production, but it will also slow the rise of income inequality because it will be directed towards more labourintensive activity. Higher saving is required because even with the inflow of foreign capital for mining, total investment slumped to a level of only 11 per cent of GDP by 1996-97 (IMF 1998). This is barely half the expected level for a country of the size and per capita income of Kazakhstan, according to the Syrquin and Chenery (1989) norms. The genuine saving rate of Kazakhstan is negative, in part because of the very low rate of domestic saving and in part because of the scale of the oil depreciation (Chapter 3). It is therefore necessary to remove the disincentives to domestic saving and investment that arise out of inadequate financial markets and predatory public sector behaviour. Households can be encouraged to save, by a move from income tax to value added tax, by incentives to accumulate individual pension funds and by improved reliability of the banking system. Business saving can be raised through institutional improvements and lower profit taxes. However, these deficiencies also serve to highlight the need to go further and incorporate social capital more explicitly into the resource-constrained rapid reform model. In addition, for the policy prescriptions to be most effective, the political state needs to evolve from its present predatory form (with or without a streak of paternalism) towards either a developmental autonomous benevolent state or a consensual democracy. It seems most likely that the greater propensity of North-eastern European countries to successfully implement rapid reform stems from their superior social capital. This, in turn, reflects the advanced development of these economies before the introduction of central planning, and also their relatively short exposure to central

Transition Economies: Kazakhstan and Uzbekistan

275

planning. In this sense, Table 16.1 picks up differences in GDP recovery that reflect variations in social capital rather than merely differences in the speed of reform as is assumed by the unadjusted rapid reform model. Although social capital is also important in conditioning the response of the remaining transition economies, their lower income brings a greater reliance on the primary sector and consequently a stronger role for natural resources in governing the nature of reform.

REFERENCES Akhad, A., A. Alisher and S. Kahramon (1995), 'Economic reform and investment priorities in the Republic of Uzbekistan', Comparative Economic Studies, 37/3, 27. Aslund, A., P. Boone and S. Johnson (1996), 'How to stabilise: lessons from post-Communist countries', Brookings Papers on Economic Activity, 1, 217—314. Auty, R. M. (1998), 'Mineral wealth and the economic transition: Kazakhstan', Resources Policy, 24/4, 241-9. Auty, R. M. (1999a), 'The IMF model and resource-abundance transition economies: Kazakhstan and Uzbekistan', UNU/WIDER Working Papers 169, Helsinki: UNU/WIDER. Auty, R. M. (1999b), 'Why is the diversification of manufacturing not occurring in Azerbaijan?' paper prepared for the ECSPE, World Bank, Washington DC. Becker, C. M. and D. S. Urzhumova (1998), 'Pension burdens and labour force participation in Kazakhstan', World Development, 26, 2087-103. De Broeck, M. and K. Kostial (1998), 'Output decline in transition: the case of Kazakhstan', IMF Working Paper 98/45, Washington DC: International Monetary Fund, de Melo, M., C. Denzier and A. Gelb (1996), 'Patterns of transition from plan to market', World Bank Economic Review, 10, 397—424. de Melo, M., C. Denzier, A. Gelb and S. Tenev (1997), 'Circumstance and choice: the role of initial conditions and policies in transition economies', Washington DC: World Bank (mimeo). EBRD (1996), Transition Report 1996, London: EBRD. EBRD (1997), Transition Report 1997, London: EBRD. EBRD (1998), Transition Report 1998, London: EBRD. EBRD (1999), Transition Report 1996, London: EBRD. EIU (1996), Country Report: Kyrgyz Republic, Tajikistan, Turkmenistan and Uzbekistan, Fourth Quarter, London: Economist Intelligence Unit. Financial Times (1998), 'Kazakhstan: a survey', Financial Times (17 June). Fischer, S., R. Sahay and C. A. Vegh (1996), 'Stabilization and growth in transition economies: the early experience', Journal of Economic Perspectives, 10, 45-66. Frye, T. and A. Shleifer (1997), 'The invisible hand and the grabbing hand', American Economic Association: Papers and Proceedings, 87/2, 354—8. Gelb, A. H. and Associates (1988), Oil Windfalls: Blessing or Curse? New York: Oxford University Press. Gurgen, E., H. Snoek, J. Craig, J. McHigh, I. Izvorski and R. van Rooden (1999), Economic reforms in Kazakhstan, Kyrgyz Republic, Tajikistan, Turkmenistan and Uzbekistan, IMF Occasional Paper 183, Washington DC: IMF.

276

R. M. Auty

Havrylyshyn, O., I. Izvorski and R. van Rooden (1998), 'Recovery and growth in transition economies: a stylized regression analysis', IMF Working Paper 98/141, Washington DC: International Monetary Fund. IMF (1998), 'Republic of Kazakhstan: recent economic developments', IMF Staff Country Report 98/84, Washington DC: IMF. Johnson, S., D. Kaufman and A. Shleifer (1997), 'The unofficial economy in transition', Brookings Papers on Economic Activity, 2, 159-239. Kenny, C. (1999), 'Why aren't countries rich? Weak states and bad neighbourhoods', Journal of Development Studies, 35, 26-47. Kornelia, K. and J. Zettelmeyer (1997), 'Competitiveness in transition economies: what scope for real appreciation?' IMF Working Paper 97/149, Washington DC: IMF. Lane, P. R. and A. Tornell (1996), 'Power, growth and the voracity effect', Journal of Economic Growth, 1, 213-41. Leite C. and J. Weidmann (1999), Does Mother Nature corrupt - natural resources, corruption and economic growth', IMF Working Paper 99/85, Washington DC: IMF. Planecon (1997), 'A sophisticated way to model earnings', Meridian International Conference Centre, Washington DC (mimeo). Pomfret, R. (1998), 'Agrarian reform in Uzbekistan: why has the Chinese model failed to deliver?' Working Paper, University of Adelaide. Pomfret, R. and K.H. Anderson (1997), 'Uzbekistan: welfare impact of slow transition', UNU/ WIDER Working Papers 135, Helsinki: UNU/WIDER. Rosenberg, C. B. and T. O. Saavalainen (1998), 'How to deal with Azerbaijan's oil boom? Policy strategies in a resource-rich transition economy', IMF Working Paper 98/6, Washington DC: IMF. Sachs, J. D. and A. Warner (1995), 'Natural resource abundance and economic growth', NBER Working Paper 5.198, Cambridge MA: National Bureau of Economic Research. Simon, J., D. Kaufmann and A. Shleifer (1997), 'The unofficial economy', Brookings Papers in Economic Activity, 2, 159-221. Syrquin, M. and H. B. Chenery (1989), 'Patterns of development, 1950 to 1983', World Bank Discussion Paper 41, Washington DC: World Bank. Taube, G. and J. Zettelmeyer (1998), 'Output decline and recovery in Uzbekistan: past performance and future prospects', IMF Working Paper 98/132, Washington DC: International Monetary Fund. Valdevieso, L. M. (1998), 'Macroeconomic developments in the Baltics, Russia and other countries of the former Soviet Union, 1992-97', IMF Occasional Paper 175, Washington DC: International Monetary Fund. World Bank (1996), World Development Report 1996: from Plan to Market, Washington DC: World Bank. World Bank (1998), Commodity Markets and the Developing Countries 5/2, 36—7, Washington DC: World Bank. WRI (1998), World Resources 1998-99, New York: Oxford University Press. Zettelmeyer, J. (1998), 'The Uzbek growth puzzle', IMF Working Paper 98/133, Washington DC: International Monetarv Fund.

17

Reforming a Large Resource-Abundant Transition Economy: Russia A N I L M A R K A N D Y A A N D ALIN A A V E R C H E N K O V A

1 7 . 1 . INTRODUCTIO N The process of economic reform has been a turbulent one in all the former planned economies of Central and Eastern Europe (CEE) and the Newly Independent States (NIS, former members of the Soviet Union). With the exception of those countries that have been at war, however, in few has the transition been more difficult and subject to such severe dislocation, as in the case of Russia. Of all the 26 countries of CEE and the NIS, Russia is probably the best endowed with natural resources. The total value of Russia's natural resources (agricultural land, forests, and mineral wealth) is, per capita, probably bigger than any of them, although that is difficult to establish given the difficulties in the collection of reliable data. This chapter seeks to understand why the transition process has been difficult in Russia, and in particular to what extent these difficulties arise from its rich natural resource endowment. It also offers some advice on how the transition process might be better managed in order to minimize the difficulties arising from the resource abundance and indeed to use that abundance to its advantage. The paper is structured as follows. Section 17.2 provides a brief summary of what has been happening in Russia since the start of the economic reforms (dating from 1989 onwards). These changes fall under various categories: economic flow variables such as GDP, investment, government budgets, and inflation; economic stock variables such as physical capital and human capital; social indicators such as poverty and inequality; and transition indicators that measure the degree to which the economy has made the reforms towards a market economy. Special attention is given to the changing role played by natural resources in the economy. Section 17.3 discusses the relevance to Russia of the main explanations offered for the relatively We would like to thank Richard Auty, Gustavo Barboza, Colin Lawson, and the participants of the workshop on Resource Abundance and Economic Growth held in Helsinki in July 1999 for useful comments that have been included in the revised version. Final responsibility for the contents remains, of course, ours. 1

Since there are no data for Bosnia and Herzegovina, comparisons are restricted to at most 25 countries.

278

A. Markandya and A. Averchenkova

poor performance of resource rich countries in making the transition to a market economy. In particular, it focuses on the changing way in which rents from the natural resource sector have been used. Section 17.4 concludes the chapter, with a discussion of the prospects for achieving the transition to an economic and social system that is sustainable, that is, one that can sustain living standards at levels comparable to those of the OECD countries. Clearly further reforms will be needed if this target is to be met.

17.2. THE B A C K G R O U N D TO THE R U S S I A N REFORMS

17.2.1. Socioeconomic flow indicators In the 15 years before its collapse, the Soviet Union had a sluggish economy, but one that was maintaining slow, albeit patchy, growth in 'Net Material Product' (the measure closest to the GDP of the Western National Accounts System).2 Between 1976 and 1980, average annual growth was 4.3 per cent; between 1981 and 1985 it slowed down to 3.6 per cent. With a slightly falling population this translated into per capita output rising at a little more than these rates. Measures for the next five years are difficult to obtain, and there are discrepancies between sources, but the broad conclusion seems to be that the USSR achieved virtually no growth in Net Material Product over that period. The data for the Russian Federation start in 1991 and show a rapid decline in GDP over the following years. In 1991 the fall was 13 per cent; in 1992, 19 per cent; in 1993 12 per cent, and in 1994 15 per cent (EBRD 1999).3 By 1995, the alarming rate of fall in GDP was slowed but not stopped. In summary, Russia has paid a high price for the transition process. Between 1989 and 1998 GDP fell by about 45 per cent. This can be compared with the 25 countries that form the CEE/NIS transition group and for which data are available. The average fall in GDP for all of them was about 26 per cent (Table 17.1). So Russia did worse than the average but was not 'bottom of the class'. Countries that did worse were Armenia, Tajikistan, Turkmenistan, Ukraine, and Moldova. Countries with a roughly similar performance include the Former Yugoslav Republic, Kyrgyzstan, Latvia, and Lithuania. Some of these countries, however, faced serious disruption through wars and civil unrest. Of course, GDP is only one guide to economic welfare, and a very imperfect one at that. It is important to look at other indicators, such as inflation, life expectancy, unemployment, and poverty. The inflation experience of the Russian Federation is well established. There were massive price increases in the early 1990s—from 875 per cent in 1993 to 197 per cent in 1995. Over the next three years these slowed down so

2 Some recent research has estimated ex post GNP for the Soviet Union from 1971-98 (White et al. 1997). The figures for GNP cited there for the 1970s are much lower than the official Net Material Product data. One implication of the new data is that the decline in GNP is overestimated. This remains a research question. " There are discrepancies in these figures too. The World Economic and Social Survey of the UN gives growth in 1993 as —8.6 per cent and in 1994 as —12.8 per cent.

Table 17.1 . Comparison of transitional indicators and GNP per capita 1994-1998 Country

Poland Uzbekistan Kazakhstan Russian

Change in private sector share 10 25 35 20

Federation

Change Change i n in large small scal e scale privatization privatization

Change i n Change i n government price and enterpris e liberalization restructuring

0.3 0.3 1 0.3

0.3 0 2 1

0 1 1 0

0 0.3

Change Change in in trade competition and policy foreign exchange system

0.3

7.1 -5

-0.8

1.3 0.8 0.3 -1

-11.4

18 -12 -37 -45

0

2

-0.8

-0.8

-11.4

-65

0.1

0.5

0.3

0.3

5.1

-26

t

-0.3

-0.3

1

2

25

1

Average

17.7

0.6

1.3 0.5

-0.21

-0.07

-0.52

-0.2

-0.26

-0.36

0.26

-0.2

-0.32

-0.39

-0.46

-0.3

-0.54

0.24

Source: EBRD (1999).

27.7

1.3 0.8 0.3 -1

0.3

Moldova

Correlation with % GNP 1994-98 Correlation with % GNP 1989-98

GNP/head GNP/head Change i n banking % change % change reform and 1994-98 1989-98 interest rat e liberalization

280

A. Markandya and A. Averchenkova

that by August 1998, when the financial crises hit Russia, inflation was running at an annual rate of about 7 per cent. Since then, however, it has jumped to over 15 per cent a month. According to World Bank data, the average inflation rate for the 26 transition countries of the region for the period 1990 to 1995 was 888 per cent per annum. The average for Russia was 710 per cent placing it 16th among the group. Apart from Croatia, all the worst performers were NIS states. Another important indicator of the deteriorating situation in Russia is adult mortality and life expectancy. For males, life expectancy at birth has fallen by 3.8 years between 1990 and 1996 (from 63.8 to 60 years), a fall that is unprecedented in peacetime. For females, the fall is less but still notable—from 74.3 years in 1990 to 73 years in 1996. The fall in male life expectancy is the greatest among the transition countries. Among the economies in transition, Russia has one of the lower recorded rates of unemployment (1.25 per cent of the labour force). However, this is a gross underestimate of the true level of unemployment. From surveys that are regularly carried out, the total number out of work in the RF is estimated at about 8.8 million or 11.4 per cent of the economically active population. These surveys have been conducted since 1992, and there has been very little change in the numbers (Bezgreblnaya 1995). The above estimates of unemployment also exclude some aspects of hidden unemployment, in the form of workers who are on enforced 'holidays', or are only provided part time employment even though they are counted as fully employed. This unemployment experience is not dissimilar to other transition countries. Comparisons are difficult to make but a detailed study undertaken by one of the authors showed much the same picture for Bulgaria, with real unemployment being, if anything, even higher (Markandya 1996). Indices of inequality are much more difficult to construct and to compare over time. Auty (1998) reports values for the Gini coefficient for transition economies as between 0.2 and 0.3 in the late 1980s. Since then, all transition countries have seen some increase in inequality but in Russia, it has been particularly marked. The most recent figure is around 0.45, which is well above the OECD average of 0.35. The World Bank 1996 World Development Report analysed Gini coefficients for 5 CEE and 3 NIS countries. The data in the NIS countries results in greater uncertainty about the change in inequality, but it looks most likely that the increase in the value of the coefficient was highest in the three NIS countries, with Russia and Kyrgyzstan at the top of the NIS list. More telling than the inequality measures, such as the Gini coefficient, are the poverty indicators. According to the World Bank estimates, 30.9 per cent of the population was below the national poverty line in 1994. The number of poor in this year was about six times higher than in 1991. It is likely that the figure was even less 4 In 1995 life expectancy for men was even lower, at just 59.2 years. The main reason was the increase in mortality of middle-aged and older groups caused by heart and circulatory diseases and external causes (suicide, accidents, etc.). This trend is connected with the change in lifestyle, loss of stability (particularly unemployment) and confidence in the future. In connection with unemployment see Markandya (1999).

Transition Economy: Russia

281

than 5 per cent before the reforms started, when there was virtually universal provision of health care, housing, and employment. UNICEF (1994) on the basis of a definition of relative poverty argues that in 1992, 61.3 per cent of the Russian population was poor, and 23.2 per cent extremely poor (that is defined as 60 per cent of the poverty line). Data for 1993-94 shows that 35 per cent of families with one child, 50 per cent of families with two children, 72 per cent of families with three children and 88 per cent with four or more children had per capita income less than the subsistence minimum. Again, international comparisons are hard to make but the situation in Russia is unlikely to be better than in other transition countries. Certainly the comparison between 1991 and 1994 is telling. In the face of these problems, the government has been impotent. Although Russian public expenditures were up to 40 per cent of GDP, they remain highly regressive. According to Aslund (1999) 70 per cent of all social transfers went to the 30 per cent of most wealthy households in 1998. Similarly, a World Bank study in 1994 showed that the poorest 20 per cent of the Russian population received only 6 per cent of all social assistance (Milanovic 1998, cited by Aslund 1999). At the same time 4 per cent of GDP is being spent on enterprise subsidies that breed corruption and organized crime, while impeding enterprise restructuring (Aslund 1999). All in all, human welfare in Russia has declined a great deal since the transition process began. It appears that the people of Russia have paid a higher price for this process than those of many other transition countries.

17.2.2. Transition indicators The EBRD has developed a set of indicators to measure the performance of the transition countries in moving to a market based economy. The set was started in 1994 and has been added to over time. Initially the following were measured: share of activity in the private sector, large-scale privatization, small scale privatization, government and enterprise restructuring, price liberalization, trade and foreign exchange reform, and banking and interest rate liberalization. In 1998 the list included all the above plus security market and non-bank financial institution reform, legal transition indicators, and infrastructure management reforms. Since most of these indicators are qualitative, it is necessary to give a judgmental ranking for the performance of each country. This was calibrated from 0-3, giving values in units of 0.5. The conversion of a qualitative ranking to a quantitative one is controversial but the data are interesting enough to warrant further analysis.5 In 1994, Russia was above the average value for all seven indicators. Hence in the period from 1989 to 1994 it was in a leading group of reforming countries, at least as measured by these indicators. Table 17.1 gives the change in the 1994 set of indicators between 1994 and 1998. The average performance for all 24 countries and that for Russia are also given. For the change during these four years, Russia was above average 5 For further description of the indicators and their construction, see the EBRD Annual Reports 1994—98. A negative value indicates that reforms have been reversed.

282

A. Markandya and A. Averchenkova

for two indicators (share of output in the private sector and small-scale privatization), but below in the other five. The combined effect of the changes pre- and post-1994 are shown in Table 17.2, where 8 more indicators are listed. For the original seven, Russia was above the average for three (share of private sector, large-scale privatization, smallscale privatization), very close to the average for two (government and enterprise restructuring and price liberalization), and below for another two (trade and foreign exchange reforms and banking reforms). For the new set of indicators it was above average in 5 of the 8, close to the average in one and below in two (effectiveness of legal transition and security market and non-bank financial institution reform). Overall, the measure of reform in Russia by 1998 is well within the range of performance among the transition countries and above the average for the majority of indicators. Moreover, studies by the EBRD (1999) and other researchers (De Melo et al. 1996) have shown that the higher the level of reform, the better the economic performance as measured by the growth in GDP. This positive correlation between cumulative change in the index and growth is also observed in Table 17.2, where the simple correlation coefficients between the reform indicator and GDP growth are given in the last row. In all cases these are positive. Hence Russia is an 'outlier'; its growth performance is poorer than the reform indicators would suggest. Some reasons for this are discussed in Section 17.3.7

17.2.3. Natural resource endowments Russia is without doubt one of the best-endowed countries in terms of natural resources. With a population that is less than 2.5 per cent of the world total, it has over 12 per cent of the land area. It is the largest producer of natural gas in the world, the third largest crude oil producer, and one of the principal sources of many minerals, such as coal, gold, diamonds, and ferrous metals. The World Bank initiated studies that estimate the wealth of different countries. The first set of results (World Bank 1997) do not include the countries in transition, as the initial efforts revealed too many data problems and the numbers were not considered reliable. This makes it impossible to carry out an analysis on how the wealth of Russia is made up, and how it has changed in the period since transition. Table A3.1 provides estimates of the genuine savings in 1997 taking into account the depletion of natural capital (covering mostly minerals and excluding many renewable resources). The negative value of this indicator for Russia suggests that the current level of reinvestment of the resource rents into human and produced capital is insufficient and leads to an unsustainable development path. However, estimates of national trends and 6

Indeed this is exactly what the regressions in De Melo show. ' An additional factor that could explain differences in growth performance is the rate of change in the indicators. Reform that is too fast could be more costly in terms of overall growth than reform that is slower. In fact, the last row of Table 18.1 shows exactly that. Most correlations between the change in the indicators and growth for the period 1994-98 are negative. Hence if the level is positively correlated with growth but the change is negatively correlated, this suggests that there is an optimum rate of transition. This issue is not developed further here, but is the subject of another paper (Lawson et al. 1999).

Table 17.2. Transition indicators for selected countries, 1998 Country

Poland Slovenia Uzbekistan Kazakhstan Kyrgyzstan Russian Bosnia and Average Correlation Federation Herzegovina with growth in GDP since 1989

Year 1998 Population 38.7 Private sector share 65.00 of GDP Large scale 3.25 privatization Small scale 4.25 privatization Government and 3.00 enterprise restructuring Price liberalization 3.25 Trade andforeign 4.25 exchange system Competition policy 3.00 Banking reform and 3.25 interest rate liberalization Securities markets 3.25 and non-bank F.I.s

1998 2 55.00

1998 23.6 45.00

1998 15.7 55.00

1998 4.6 60.00

1998 147.2 70.00

1998 4.1 35.00

10.77

0.4

3.25

2.25

3.00

3.00

3.25

2.00

0.57

0.28

4.25

3.00

4.00

4.00

4.00

2.00

0.75

0.45

2.75

2.00

2.00

2.00

2.00

1.75

0.45

0.56

3.00 4.25

2.00 1.75

3.00 4.00

3.00 4.00

2.75 2.25

3.00 2.00

0.55 0.70

0.12 0.27

2.00 3.00

2.00 1.75

2.00 2.25

2.00 2.75

2.25 2.00

1.00 2.00

0.42 0.50

0.53 0.48

3.00

2.00

2.00

2.00

1.75

1.00

0.47

0.73

Table 17. 2 (Continued) Country Polan

Legal Transition Indicators Extensiveness Effectiveness Total Infrastructure Telecommunications Railways Electric power GNP Growth (1989-98) Source: EBRD (1999).

d Sloveni a Uzbekista n Kazakhsta n Kyrgyzsta n Russia n Bosni a an d Averag e Correlatio n Federation Herzegovin a wit h growt h in GDP since 1989

4.00 4.00 4.00

3.00 3.00 3.00

2.25 2.00 2.00

2.25 2.00 2.00

3.00 2.00 2.00

3.75 2.00 3.00

2.00 1.00 1.00

0.56 0.50 0.50

0.14 0.22 0.29

3.25 3.25 3.00 18.00

2.25 3.25 2.25 3.00

2.00 2.00 1.00 -12.00

2.00 2.00 3.25 -37.00

2.00 1.25 2.25 -40.00

3.00 2.25 2.00 -45.00

1.25 2.00 2.00

0.44 0.45 0.45 -2.62

0.35 0.4 0.27

285

Transition Economy: Russia

international comparisons are difficult to make due to the high levels of uncertainty and incomplete data. Some more information on the importance of the natural resource based sectors can be gleaned from the output data. Table 17.3 traces the evolving structure of production for the period 1992-97. It is constructed from a number of sources which vary in their reliability, so that the numbers must be considered as only indicative of the trends. During these six years, the natural resource based sector has declined by 60 per cent in value terms and its share of GDP has fallen from 42 per cent to 23 per cent The percentage fall has been greatest for minerals, followed by fuels and agriculture The natural resource sectors have taken a bigger cut than the other sectors of the economy. Services actually increased by about 23 per cent in real terms, construction has fallen by 46 per cent and industry by 27 per cent. The result is that the economy is much less dependent on natural resources now than it was in 1992. Although the structure of output has changed significantly since 1992, the change in the structure of external trade has remained more consistent. Table 17.4 gives the key external trade statistics for Russia from 1994 to 1997. Net exports of fuels and other Table 17.3. Structure of the economy oj the Russian Federation: 1992—1997

1992 Trillion Rubles 1997 prices Fuels Minerals Agriculture and forestry

1993

1994

1995

1996

1997

585.6 530.7 347.7

364.6 355.4 285.8

242.8 241.0 194.0

262.3 250.8 212.6

257.6 201.3 176.0

237.0 185.8 167.8

Total for natural resource based activity

1463.9

1005.8

677.8

725.7

634.9

590.6

Construction Services Industry and other

278.1 1147.3 587.4

253.9 1460.2 454.4

249.4 1496.4 347.6

186.0 1461.6 284.2

138.5 1174.3 616.8

150.4 1412.2 431.8

Total constant 1997 prices

3476.7

3174.2

2771.1

2657.5

2564.5

2585.0

Fuels Minerals Agriculture andforestry Total for natural resources based activity Construction Services Industry and other

16.8% 15.3% for10.0 42.1%

11.5% 11.2% 9.0% 31.7%

8.8% 8.7% 7.0% 24.5%

9.9% 9.4% 8.0% 27.3%

10.0% 7.9% 6.9% 24.8%

9.2% 7.2% 6.5% 22.8%

8.0% 33.0% 16.9%

8.0% 46.0% 14.3%

9.0% 54.0% 12.5%

7.0% 55.0% 10.7%

5.4% 45.8% 24.1%

5.8% 54.6% 16.7%

Total

100%

100%

100%

100%

100%

100%

Shares

Note: Shares for agriculture, construction and services up to 1995 were taken from UNECE (1996). Estimates for the same sectors for 1996 and 1997 are based on Russian Economic Trends (1999), Data for fuel and mineral sectors were constructed from data provided by Goskomstat. The estimates for 'Industry and Other' are obtained as a residual.

286

A. Markandya and A. Averchenkova Table 17.4. External trade for Russia: 1994-1997

1994

1995

Net Exports ($ bn) Fuels Other Minerals Wood Agricultural products

52.1 27.2 2.1 —7.9

Total natural resource net trade Manufactured products Total all exports Shares (% total) Fuels Other minerals Wood Agricultural products Manufactured products

1996

1997

63.7 30 3.3 -10.5

69.4 37.7 2.1 —8

65.8 37.3 1.7 —11.6

73.5 -8.1

86.5 -6.4

100.9 —6.8

93.2 -10.7

67.1

80

86.9

83.9

79.9% 43.4% 2.4% -9.6% -7.8%

78.4% 44.5% 2.0% -13.8% -12.8%

77.6% 40.5% 3.2% -11.8% -12.1%.

79.6% 37.5% 4.1% -13.1% -8.0%

Note: Manufactured products include machinery, chemicals and textiles. Source: Goskomstat (1998).

minerals increased substantially during this period (fuels by 26 per cent and minerals by 37 per cent, in dollar terms). Exports of wood and wood products, agriculture, and manufactured products on the other hand have continued to decline. In terms of shares, fuels and other minerals dominate; they account for more than 100 per cent of net exports (agriculture having a negative net export has a negative share as well). The picture that emerges therefore is one where there is a difference between the fuels and minerals on the one hand and agriculture and forestry on the other. Although both subsectors are declining in terms of total domestic production, the net exports of fuels and minerals are increasing as a percentage of net trade, whereas agricultural products are decreasing in terms of net trade as well. The most obvious reasons for this pattern are the decline in domestic demand for fuels and other minerals, the continued fall in agricultural and forest sector productivity and competitiveness, and the effects of appreciation of the exchange rate on the net trade in these products. These issues are discussed further in Section 17.3, but at this point it is worth noting that the net exports of other 'tradables', such as machinery, chemicals, and textiles declined much less than agriculture. In value terms they fell by about 32 per cent but their share of net exports has remained constant.

1 7 . 3 . TH E R U S S I A N REFOR M P R O G R A M M E AN D THE NATURA L RESOURC E SECTOR This section looks at a number of explanations for the pattern of development in Russia during the transition phase that have been highlighted above, focusing on the role of resource abundance. A number of hypotheses have been advanced for the relatively poor performance that such countries have exhibited, particularly among

Transition Economy: Russia

287

Table 17.5. Transition indicator sum for CEE countries to 1998 Country

Indicator sum

Turkmenistan Belarus Bosnia and Herzegovina Tajikistan Uzbekistan Azerbaijan Ukraine Albania Kyrgyzstan Kazakhstan Russian Federation Armenia Georgia

14.75 21.25 24.00

Ranking

Country

Indicator sum

Ranking;

1 2 3

Moldova FYR Macedonia Slovak Republic

38.00 38.25 39.50

14 15 16

25.00 28 30.00 31.5 32.00 35.25 35.75 36.25

4 5 6 7 8 9 10 11

Latvia Croatia Bulgaria Lithuania Slovenia Romania Estonia Czech Republic

40.53 40.75 42.00 42.00 42.25 42.50 47.25 47.75

17 18 19 20 21 22 23 24

37.50 37.75

12 13

Poland Hungary

49.00 52.50

25 26

Average

36.6

transition economies (see Chapter 16). This section evaluates the three principal ones, namely: deferral of reform; an appreciation of the real exchange rate so that economy suffers from the 'Dutch disease'; and difficulty both in creating new institutional capital and in curbing the corrosion of social capital. Each of these hypotheses is discussed further below.

17.3.1. Slower reforms in resource rich countries As noted in the previous section, there is little evidence that Russia proceeded with reform at a slower rate than the other transition countries. Table 17.5 summarizes the indicators of transition as estimated by the EBRD. The figure given is a simple arithmetic sum of 15 indicators discussed in the previous section. It may be argued that this is a crude view of the achievements of transition, but it is nevertheless a useful one The values given for these indicators have been arrived at mostly on a judgemental basis, from the experience of those who are most familiar with what has been happening in these countries. The figures show that the average value of the sum for the 26 countries was 36.6. Russia was almost exactly at the average, with a value of 36.3 The other large resource exporters were a little below: Azerbaijan at 30.0 and Kazakhstan at 35.8.8 8

Chapter 16 looks at the relative performance of Kazakhstan and Uzbekistan. It notes that Uzbekistan had a better economic performance than Kazakhstan, although by the indicators in Table 17.5 it was slower in making the transition. It explains this difference mainly in relative performance by (a) the greater negative impact of mineral resource abundance on the transition (Kazakhstan has much more mineral wealth than Uzbekistan whose endowment is mainly agricultural) and (b) the speed of the transition, which it argues did not help Kazakhstan and allowed Uzbekistan to experience only a modest collapse in growth and to maintain public revenues. We would agree with both these observations; our own study, looking across all transition countries particularly supports observation (b).

288

A. Markandya and A. Averchenkova

Countries below the average were mostly the least developed, and those with the fewest resources to implement the transition. On the basis of this evidence it would be hard to sustain the view that Russia was particularly poor in implementing the reforms required for transition. However, it is possible that, were it not for the presence of the subsidies and support available from its natural resources, the transition would have been even faster in Russia. This question is considered further under the discussion of rents.

17.3.2. The appreciation of the real exchange rate The exchange rate for the Russian Federation has certainly appreciated in real terms in the period since 1992. Table 17.6 provides estimates of the appreciation of the real exchange rate for Russia and for four other countries in transition for varying periods ranging from 1991 to 1998. It is interesting to note that all these countries have seen some appreciation in the real exchange rate. This has been greatest in Kazakhstan, followed by Russia and Ukraine. The appreciation for Ukraine to 1997 was very similar to that for Russia, although the latter is better endowed with natural resources for which there is an international demand. The two non-NIS countries included (Hungary and Poland) showed a small real appreciation but there too, it was not insignificant. The depreciation of the rouble in 1998 has corrected the overvaluation considerably, reducing it from 72 per cent to 45 per cent. This was forced on Russia as a result of several factors, including a collapse in the price of oil and a loss of confidence

Table 17.6 . Real exchange rate appreciation for selected transition countries

Country Perio

Hungaryb Kazakhstanc Polandd Russiae Russiae Ukrainef

d Actua l rate Rat e that would exist Percentag e at end of wit h no appreciation relativ e to appreciatio n period beginnin g o f perioda 1991-97 1992-97 1991-97 1992-97 1992-98 1992-97

203.5 75.6 3.52 5,960 20,841 1.89

230.0 3,234.0 5.20 20,986 37,985 7.5

11.5 97.6 32.3 71.6 45.1 74.8

Notes: Rates are calculated by the authors on the assumption that for no appreciation the local rate should change by the difference between the domestic rate of inflation and the US rate of inflation. Both are measured using national consumer price indices. These are not ideal, but consistent export price data were difficult to obtain. b Currency rate is Forints/Dollar. c Currency rate is Tenge/Dollar. d Currency rate is Zlotys/Dollar. c Currency rate is Old Roubles/Dollar. f Rate is Hryvnias/Dollar. a

Transition Economy: Russia

289

in the government's ability to meet its international obligations. As a result, Russia has experienced a resurgence in inflation, but also a major boost to the domestic manufacturing sector. This suggests that the overvaluation of the exchange rate during the period 1992-97 did impede domestic growth. In summary, the table shows that there has been a real exchange effect in all countries but it has been more noticeable in the resource rich countries like Russia and Kazakhstan. The next question is the extent to which this appreciation has affected the economic performance of Russia. As Table 17.4 shows, net trade in manufactures has gone into a slightly bigger deficit, but by no means an enormous one. Exports have performed quite well. This may reflect the nature of the products, which include military and other hardware, for which trade is dictated by considerations other than simple economics. Agricultural products have moved massively into deficit, but this is only partly the result of the exchange rate change. It reflects the fact that the domestic sector is in serious disarray, with institutional reform being very slow (Esty 1997). The same applies, to a lesser extent, to the forestry sector. On the other side, the evidence from the devaluation of 1998 does suggest that domestic manufacturing does respond to relative exchange rates, as output has increased in the import competing and export sectors. Taking all sectors, it must be concluded that the real exchange rate appreciation in the transition countries has been widespread, but with the higher rates in the resource abundant countries. Moreover, Russia has experienced a significant appreciation, but one that is not dissimilar to Kazakhstan and Ukraine. The latter is less resource rich than Russia. The appreciation has had some effect on exports and import substitute sectors and thereby on domestic growth but the size of this effect may not have been very big. Other factors relating to institutional reform are likely to be more important.

17.3.3. Implementing institutional reforms and maintaining social cohesion The problems of implementing institutional reforms in resource abundant countries arise in part because resource rents are available. If they are used to support subsidies to consumers then the government is not under so much pressure to change the market structure and open up the affected industries to competition. At the same time, the process of privatization and private sector involvement open up possibilities to acquire some of the rents that hitherto went to the state. Where state assets have been divested in a way that allows a few individuals to get their hands on valuable assets at knockdown prices, there is a transfer of wealth that is resented and that creates social conflicts. Each of these is considered in the context of the Russian experience. The rents from key natural resources can be estimated as follows. For each resource the output has been multiplied by the world price for Russian exports, f.o.b. From this is subtracted the cost of production, including transport to the point of export. The remaining amount is the potential rent. The actual rent is less than this to the extent that different groups of consumers receive the commodity at below the international price.

290

A. Markandya and A. Averchenkova

After deducting the subsidies to different groups, the rent is left that is available for sharing between the state and the enterprise that owns the resource. For Russia it is possible to calculate the actual rent and the potential rent for oil, gas, coal, and iron ore (detailed data on other minerals is not accessible). The breakdown of this rent between the state and the enterprises is more difficult because full tax details for the enterprises are not available. Hence this is only possible at some aggregate level. Table 17.7 gives the figures for the years 1994 to 1997. The figures show that the country continued to transfer a large part of the rent from the main natural resources to consumers, through subsidies on the prices of energy. These amount to about half of the potential rent, although the share of subsidies has been falling. In 1994 consumers captured 62 per cent of the potential rent; in 1997 this Table 17.7. Potential and actual rents from natural resources in Russia ($billion) Resource

1994

1995

1996

1997

Oil Potential rent Actual rent Subsidy to CIS Subsidy to consumers Gas

31.98 15.32 1.38 13.69

33.19 18.08 0.88 8.8

40.17 23.96 0.83 12.22

36.28 21.89 0.08 10.61

Potential rent Actual rent Subsidy to CIS Subsidy to consumers

40.04 11.40 1.67 29.95

36.55 11.44 2.01 30.73

44.68 16.30 0.88 31.26

43.25 15.20 1.11 31.71

8.13 2.67 0.04 5.99

7.30 3.90 0.05 4.39

8.80 5.94 0.03 3.28

7.45 4.32 0.04 3.67

1.14 1.14 0.00 0.00

1.41 1.14 0.00 0.00

1.83 1.83 0.00 0.00

1.64 1.64 0.00 0.00

80.15 29.39 3.09 49.63

77.07 33.42 2.94 43.88

93.66 46.20 1.68 46.76

86.98 41.41 1.18 45.99

Coal

Potential rent Actual rent Subsidy to CIS Subsidy to consumers Iron Ore

Potential rent Actual rent Subsidy to CIS Subsidy to consumers Total

Potential rent Actual rent Subsidy to CIS Subsidy to consumers

Note: Actual rents plus subsidies should add up to potential rents. If they do not, it is because of rounding errors. Actual rent has been calculated at the assumption that the difference between the extracted and exported amount of the resource is sold at the price reported for the domestic market. However, taking into account the scale of non-cash modes of payments in Russia, these figures are likely to be overvalued. Gaddy and Ickes (1998) estimate the share of barter payments among industrial enterprises to be above 50%. Source: Authors' calculations.

Transition Economy: Russia

291

had fallen to 53 per cent. Nevertheless it remains very high and provides evidence that pressure for reforms in some sectors of the economy was reduced as a result of the availability of these rents. One mechanism for this rent to be transferred to industrial consumers is through the barter system, which is prevalent in much of Russian commerce (50 per cent in 1997). Gaddy and Ickes note that utilities such as Gazprom provide energy to enterprises, which pay them in goods. These are then used to pay tax in kind to the government, but at overvalued rates. By this means barter and other forms of non-cash payments facilitate tax avoidance and evasion. In 1997, 40 per cent of taxes paid to the federal government were in non-monetary form. The overvaluation of manufactured goods through such non-cash systems of payment has helped to keep inefficient enterprises unrestructured and created a value-extracting 'virtual economy' (see Gaddy and Ickes 1998). As far as state revenues are concerned, the level coming from natural resources shows little trend but the share has been falling. Table 17.8 provides some key data. Natural resource revenues for the consolidated budget have remained around 100 trillion roubles since 1992. The percentage share, however, has declined from around 28 in 1992-94, to around 22 in the years 1995-97. As a percentage of GDP, the share also appears to be dropping, although trends are difficult to discern. For example, the ratio of the share of resource based sectors in taxes, to their share in GDP (the definition of the relative tax burden, as given by the IMF), increased from less than one in 1992-93, to over one in 1995-97. This indicates that the government is relying more on this sector for its revenues, when allowance is made for its declining share in GDP. Additional evidence not shown in Table 17.8 indicates that the burden of the oil sector has increased even more than for natural resources as a whole. That too would suggest that increasing reliance on natural resource rents has reduced the pressure for tax reform. The other aspect of institutional reform in Russia relates to the transfer of natural resource and other assets through privatization. There is no doubt that the process has resulted in large fortunes being accumulated by some former managers of the large enterprises, although the process by which this has happened is not straightforward. Aslund (1999) argues that the privatization process itself did not create benefits for the managers. Rather it provided opportunities for rent-seeking and extraction of export rents, subsidies, and other sources of transfers from the state. The fact that great wealth was being accumulated and at the same time many people were facing increasing desperate poverty has been a major source of social conflict. Table 17.8. Taxation of natural resources in Russia (consolidated federal and regional)

Total revenue from natural resources (trillion 95 roubles) As % of total revenues Relative tax burden Source: See text.

1992

1993

1994

1995

1996

1997

103

76

115

94

102

116

28.3 0.88

20.8 0.92

29.0 1.66

21.5 1.11

22.3 1.25

21.8 1.33

292

A. Markandya and A. Averchenkova

A number of researchers have drawn attention to the importance of this phenomenon in explaining the poor growth performance of some countries, including Russia (Aslund et al. 1996; Rodrik 1998). As Rodrik notes, When social divisions run deep and the institutions of conflict management are weak, the economic costs of exogenous shocks ... are magnified by the distributional conflict triggered. Such conflicts diminish the productivity with which a society's resources are utilized in a number of ways: by delaying needed adjustments in fiscal policies and key relative prices, by generating uncertainty in the economic environment, and by diverting activities from the productive sphere to the redistributive one.

This has been a key problem in Russia in the transition phase. Another way of looking at the problem is in terms of social capital. As Knack and Keefer (1997) note, where social capital is high, there is less resort to litigation and/or criminal racketeering to enforce contracts. The latter is particularly a problem in Russia, where the legal situation is uncertain in many areas. In this context, the role of state intervention remains very important. All economic activities are subject to licensing that is provided by over 60 state agencies. Required enterprise inspections create opportunities to extort bribes (Aslund 1999). A natural consequence is that corruption is pervasive. In its Corruption Perceptions Index for 1998, Transparency International puts Russia among the most corrupt, as 76th among 85 countries (Russian Economic Barometer 1999). The key question, and one that is difficult to answer clearly, is to what extent has this arisen because of resource abundance? As noted above, the process of transferring assets is widely seen as inequitable. Much of the wealth that has been transferred originated in the natural resource sector. Arguably, without the creation of wealth inequalities in this form the process of transition would have been easier and would have generated less conflict than it did. This view is supported by Stiglitz's analysis of the Russian performance (Stiglitz 1999). He emphasizes the failure of the privatization programme in creating a post-communist society where corporate governance was effective. One of the implications has been that the government is unable to collect taxes from a public deeply cynical about the actions to control the public finances.

17.4. CONCLUSION S The transition experience in Russia has produced benefits for some people, but it has not generally been a happy one. Although Russia was not the worst performer on many of the indicators (except the drop in life expectancy and possibly inequality), its ranking was relatively poor. Given its many initial advantages, especially its immense natural resource wealth, a better outcome might have been expected. This chapter has addressed the question: was the high cost of transition inevitable for Russia, and to what extent did it arise as a consequence of the resource abundance? It analyses three hypotheses. First is the view that resource abundant countries are slow to make reforms because they have a cushion of rents to rely on. The transition indicators that have been

Transition Economy: Russia

293

developed by the EBRD give qualified support to the view that the resource abundant countries have made less progress towards a market-based economy. Russia's overall performance was about average in terms of the change in all IS indicators, but it was ranked below the median. The poorest performances were not from the resource abundant countries, but from some of the countries with few resources, several of which were, however, also facing wars and civil unrest. Nevertheless, it is certainly true that Russia had access to significant rents from oil, gas, and minerals. Availability of resource rents allowed Russia to maintain barter and other non-cash systems of payments creating a value-extracting economy with inefficient and wasteful manufacturing sector. The rents have also helped the government meet its revenue requirements, making tax reform not as pressing an issue as it would otherwise have been. Therefore, reforms in public finances and enterprise restructuring would have been faster had it not been for the availability of natural resources. The second explanatory factor is the impact of resource abundance on the real exchange rate. Again it is the case that the real rate appreciated greatly in Russia. From the recent data and history it can be concluded that resource availability allowed the government to pursue a strong rouble policy. But it must be understood that this was a choice of the government. It could have followed polices that allowed the rouble to devalue in line with inflation, but it took the view that this would have reinforced the inflationary pressures (as imports play such a large part in consumer prices), when control of inflation was seen as a primary goal. This is a point made throughout this volume: resource abundance does not predetermine the fate of nations; it is within their capacity to adjust policies so that the resources work in favour of overall economic development rather than against it. The impact of the real exchange rate appreciation is seen in Russia in an increased deficit on agricultural trade and in trade on manufactured goods. The latter, however, was small compared to the size of the appreciation. Much of this trade remains managed, and the government has been fairly successful in maintaining its exports. The third factor is a social and institutional one. Resource abundance creates social conflict, as the benefits are unequally distributed. The process of reform in Russia undoubtedly resulted in an increase in inequality, much of that was the consequence of a privatization programme that allowed the erstwhile managers to acquire valuable state assets at very low prices. This inequality has been greatly resented, especially as it has been accompanied by other changes that have reduced the living standards of millions of people. Prominent among these is inflation, which has wiped out the savings of many, increased uncertainty of employment and a decline in the state health and education services. The associated increase in social conflict has resulted in a devaluation of the social capital of the country. Relations of trust are lost and corruption and criminal activity increased. The risk of investment for the long term has become much greater and the real economy has suffered. Although all three factors have played a significant part in explaining the relatively poor performance of Russia during the transition period, the third explanation, along with the presence of rents slowing down institutional reforms in key areas, is the most potent. For the future, the government must pay the highest attention to reducing the

294

A. Markandya and A. Averchenkova

social conflict. This requires a crackdown on corruption and criminal activity. Furthermore, when reforms are introduced, the government must ensure that these are not used as a vehicle for a few people to extract more rents from the system. A more careful accounting of the rents from natural resources and for allocating them to the highest value uses in a transparent way would be beneficial. If these actions are taken, and complemented by stable macroeconomic policies, there is some hope that the transition will yet be successful in this vast country. At the same time as reforming the government, it is also desirable to make the government sector smaller. Others, including Aslund (1999) have also noted the need for this. Total state revenues should be reduced, leading to a decline in the value of tax exemptions. This in turn would both reduce rent-seeking and provide incentives to abolish barter. A reduction in the share of barter transaction should lead to increased incentives for investment in an efficient industrial sector and for the restructuring of value-subtracting sectors (Gaddy and Ickes 1998).

REFERENCES Aslund, A. (1999), 'Why has Russia's economic transformation been so arduous?', paper prepared for the Annual World Bank Conference on Development Economics, 28—30 April 1999. Aslund, A., P. Boone and S. Johnson (1996), 'How to stabilise: lessons from post-communist countries', Brookings Papers on Economic Activity 1: 217-314. Auty, R. M. (1998), 'Resource abundance and economic development: improving the performance of resource-rich countries'. Research for Action 44, Helsinki: UNU/WIDER. Bezgreblnaya, I. (1995), Labour Markets in Russia, paper prepared for the Harvard Institute for International Development, Moscow Office. De Melo, M., C. Denzier and A. Gelb (1996), 'Patterns of transition from plan to market', World Bank Economic Review, 10, 397—91. EBRD (1999), Transition Report 1999: Ten Years of Transition, European Bank for Reconstruction and Development. Esty, D. C. (1997), 'Environmental protection during the transition to a market economy', in W. T. Woo, S. Parker and J. D. Sachs (eds.) Economies in Transition: Comparing Asia and Europe, Cambridge MA: MIT Press: 357-85. Gaddy, C. and B. Ickes (1998), A simple four-sector model of Russia's 'virtual' economy (May). Goskomstat (1998), Russia in Figures: Concise Statistical Handbook, Goskomstat of Russia, Moscow. Knack, S. and P. Keefer (1997), 'Does social capital have an economic pay-off? A cross-country investigation', Quarterly Journal of Economics, 112, 1250—88. Lawson, C. W., A. Markandya, and M. Pemberton (1999), 'Explaining the effects of the level and rate of transition on economic growth', University of Bath Discussion Paper (mimeo). Markandya, A. (1996), 'ODS phase out in Bulgaria: evaluating the effectiveness of alternative strategies when enterprises are in transition' Report to the World Bank. Markandya, A. (1999), 'Employment and environmental protection: the trade-offs in an economy in transition', Environmental and Resource Economics, 13, 1-27. Rodrik, D. (1998), 'Where did all the growth go? external shocks, social conflict, and growth collapses', J. F. Kennedy School of Government, Harvard University, MA. Russian Economic Barometer (1999), (http://www.cspp.strath.ac.uk/).

Transition Economy: Russia

295

Russian Economic Trends (1999), Stockholm Institute of Transitional Economics and East European Economies, (http://www.hhs.se/site/ret/ret.htm). Stiglitz, J. (1999), Whither Reform?, Annual World Bank Conference on Development Economics, World Bank. UNECE (1996), Economic Survey of Europe in 1995-96, New York and Geneva. UNICEF (1994), Crisis in Mortality, Health and Nutrition, New York: UNICEF. White, S., A. Pravda, and Z. Gitelman (1997), Developments in Russian Politics 4, Macmillan Press. World Bank (1997), Expanding the Measure of Wealth, Washington DC: World Bank.

18 A Nordic Perspective on Natural Resource Abundance THORVALDUR GYLFASON

1 8 . 1 . TH E NORDI C COUNTRIE S SINC E 197 0 This chapter analyses the Nordic countries as relatively late industrializers and asks what evidence there is that resource abundance continues to affect their economic performance. Since about 1970, the Nordic economies generally have grown less rapidly than those of many other industrial countries have. Their living standards have also diverged from one another: they were approximately the same around 1990, but that is no longer the case. In 1997 for example, Norway's ppp-adjusted per capita GNP, the highest in the group, was 28 per cent higher than that of Sweden, the lowest, compared with a difference of less than 1 per cent in Sweden's favour, in 1990 (World Bank 1999). If we gauge living standards by ppp-adjusted GDP per hour worked, a better measure because it mirrors labour productivity, then we find a 25 per cent difference between Norway and Sweden, a 35 per cent difference between Norway and Finland, and a difference of about 50 per cent between Norway and Iceland (Table 18.1). It is not enough, however, to look at current income flows and the hours of work necessary to sustain them in order to assess the wealth of nations and their living standards. It is also necessary to view the underlying trends, including the status and movement of key macroeconomic stock variables like natural-resource endowments, including the environment, and other national assets and liabilities. However, many of these assets and liabilities—natural-resource endowments and social capital among them—are notoriously hard to measure. Table 18.2 shows the World Bank's estimates of the level and composition of total ppp-adjusted national wealth per person in 1994 in the Nordic countries (all but Iceland). According to these estimates, the national wealth of each of the Nordic countries is above the West-European average, but below that for North America. Human capital is by far the most important component of wealth everywhere. The share of natural capital is accordingly small everywhere, ranging from 4 per cent in Denmark to 10 per cent in Norway. Natural resources Richard M. Auty, Ronald Findlay, Matti Pohjola, Moshe Syrquin, and an anonymous referee provided useful comments and suggestions. For a longer version of the paper see Gylfason (1999a).

Nordic Perspective on Natural Resources

297

Table 18.1 . Nordic living standards (1)

(2)

GDP per hour worked 199 7 GNP pe r capit a 199 7 (ppp-adjusted, US$) (ppp-adjuste d US$) Denmark Finland Iceland Norway Sweden

23,450 19,660 22,500 24,260 19,010

28 26 23 35 28

Source: World Bank (1997) and OECD (1997).

Table 18.2 . Nordic wealth: human, physical, and natural, 1994 (1) (2 ) (3) (4 ) Total pe r capita Huma n capita l Physica l capital Natura l capita l national wealth ( % of total) ( % of total) ( % of total) (US$ thousands , ppp-adjusted)

Denmark Finland Iceland Norway Sweden North America Western Europe

295 241 — 302 260 326 237

72 56 — 57 68 76 74

24 37 — 33 27 19 23

4 7 — 10 6 5 2

Source: World Bank (1997).

however, account for a considerably larger share of national wealth in the Nordic countries than in Western Europe and North America on average. The natural resource abundance of the Nordic countries is fairly concentrated. Norway's oil and natural gas account for two-thirds of its natural capital compared with just 11 per cent in Denmark (whose agricultural land accounts for two-thirds of its relatively small natural wealth). In Finland and Sweden, forests constitute about two-thirds of natural capital. Hence, the natural capital of Norway, Finland, and Sweden is quite heavily concentrated in a single industry—oil and gas in Norway and forests in Finland and Sweden. Unlike natural capital that cannot be accumulated, but only managed and maintained, physical and human capital accumulation requires investment in machinery, equipment, and education. Let us now take a look at each of these in turn and relate them to economic growth performance and prospects. Table 18.3 reviews some indicators of investment, education, external trade, and economic growth in the five Nordic countries. High-quality investment clearly is good for growth at least in the 1

Social capital could not be included because no numerical estimates of it exist.

Table 18.3. Nordic investment, education, trade, and growth: An overview

(1)

(2)

(3)

(4)

(5)

Investment Investment Expenditure on Expenditure on Share of

Denmark Finland Iceland Norway Sweden World/OECD

(6) Exports

(7) Exports

(8) Economic

(average (1996, % of universities 1960-96, GDP) (1994, % of % of GDP) GDP)

education (1995, % of GNP)

17-34 year olds (average (1996, % of growth in tertiary 1960-96, GDP) per capita education % of GDP) (ppp, % year (1995, %) 1980-96)

21 25 24 27 21 22

8.2 7.6 5.0 8.1 8.1 4.8

10.8 14.0 8.5 12.8 9.2 10.8*

17 16 18 21 16 24

2.1 1.9 1.0 2.1 2.2 1.5*

32 27 35 38 28 32

34 38 36 41 40 36

1.6 1.2 0.9 2.2 0.8 1.4

Note: The bottom line shows unweighted averages. An asterisk denotes an unweighted average for OECD countries. The Norwegian figure in column 3 is an estimate. Source: World Bank (1999) and OECD (1997).

Nordic Perspective on Natural Resources

299

medium term, and perhaps even in the long run. The same argument applies to education and external trade and, in fact, to any significant contribution to increased efficiency. Both education and trade lift the level of output that can be produced from given inputs through increased efficiency. Table 18.3 shows the ratio of gross domestic investment to GDP in 1960—96 as well as in 1996 alone. Investment in 1996 was well below its historical average and also below the world average in all five countries. The declining trend of investment need not be a matter of grave concern however, because Nordic investment generally seems to have been of fairly high quality, despite the bad banking that helped trigger the acute financial crises of the 1990s. Rapid growth despite relatively little investment would be a clear sign of efficiency, but sluggish growth with slow investment is not. More saving and more and better investment would clearly be good for growth. The Nordic countries' commitment to tertiary education as well as to education in general is, with one exception, well above the world average (columns 3 and 4). There are signs, however, at least in Iceland and Sweden, that excessive wage compression in centralized labour markets and blunted incentives due to various tax wedges and welfare policies have reduced the demand for higher education (column 5). Insufficient education tends to impede economic growth. Table 18.3 also shows the ratio of exports to GDP in 1960-96 as well as in 1996. Exports in 1996 were above their historical average in all five countries. In 1996, the export ratio was at or above the world average in four of the five countries, but that comparison is flawed because it does not take the small size of the Nordic economies individually into account. This makes a difference because small countries are more dependent than larger ones on external trade to extend their home markets beyond their national borders. Therefore, a comparison of export shares should pit the Nordic countries with less than five million inhabitants per country on average, against other small countries—say, all countries with a population often million or less. There are 65 such countries that report to the World Bank, and their average export ratio in 1996 was 42 per cent. None of the Nordic countries matches that, even if Norway and Sweden come close. Small countries that neglect to make up for the small size of their home markets through judicious specialization and vigorous trade in world markets, may expect to have to pay for this neglect through slower economic growth than would otherwise be available to them in the long run. In Norway and Iceland, in particular, foreign trade has been stagnant, or worse, for decades. This means that the rapid expansion of oil exports from Norway since the mid-1970s has crowded out non-oil exports krone for krone. In Iceland, the export ratio has hovered around one-third at least since 1945, an extremely low ratio in a country with only 275,000 inhabitants.3 No other industrial countries have experienced declining or stagnant export ratios in the post-war period. For comparison, the unweighted world average export ratio rose from 25 per cent in 1960 to 36 per cent in 2 A similar trend appears in the industrial countries as a whole, but the pattern is more pronounced in the Nordic countries. 3 For comparison, the average export share of thirty countries with fewer than two million people, all such countries reporting export ratios to the World Bank, was 50 per cent in 199S.

300

T. Gylfason

Figure 18.1 . Nordic exports, 1960-1997 Source: World Bank (1999).

1996, reflecting partly the increase in the number of small countries since 1960, but partly also increased openness to international trade. Figure 18.1 shows the ratio of the export of goods and services to GDP in the Nordic countries year by year from 1960 to 1997. The Norwegian and Danish export ratios have remained virtually unchanged over this period, while the Icelandic export ratio has actually declined. Only in Finland and Sweden has the export share increased significantly. Why have the Nordic countries' exports in general been so sluggish?—at least when compared with some other small open economies like Ireland. And why have their exports developed so differently? One can think of at least two possible reasons. The first has to do with inflation. The Nordic countries have a history of somewhat higher inflation than, say, the member countries of the European Union. This means that the real exchange rates of the Nordic currencies have been somewhat higher than they otherwise would have been, and this has hurt exports. Inflation can have real effects at least as long as nominal exchange rates do not adjust fully and instantaneously to changes in domestic or foreign prices. By driving real exchange rates too high above their long-run equilibrium levels, periodically or permanently, and possibly also through other channels, inflation seems to discourage exports across countries (Gylfason 1999b). This helps explain the perceived need for repeated devaluation in Finland, Norway, and Sweden in the 1970s and 1980s.4 This is indeed one of the main symptoms of the Dutch disease (Corden 1984; Wijnbergen 1984). A second possible explanation for sluggish export performance and for the different trends observed in different countries has to do with primary exports. Figure 18.2 shows the evolution of primary exports relative to merchandise exports in the Nordic Ts.his phenomenon was also much in evidence in Iceland from 1960 at least until the mid-1990s.

Nordic Perspective on Natural Resources

301

Figure 18.2 . Nordic primary exports, 1963-1997 Source: World Bank (1999).

countries since 1963. Finland and Sweden, whose total exports have increased most rapidly, are the two countries in the Nordic group whose dependence on primary exports has been the least since the early 1960s, having gradually declined below 20 per cent of the total as manufacturing exports expanded. Norway, by contrast, due to its spectacular oil discoveries since the mid-1970s, has seen its primary export share increase from about 50 per cent in 1963 to almost 80 per cent in 1997. Iceland's primary export share has fallen, yes, but from almost 100 per cent in 1963 to only a little less than 90 per cent in 1997.5 Even so, exports offish account for a bit more than a half of total exports of goods and services and about one sixth of Iceland's GDP. Why all this concern about exports? Of the three pillars of economic growth stressed above, that is, investment, education, and trade, the last is perhaps the least obvious. Exports and related variables have figured hardly at all among the many significant determinants of economic growth suggested by recent econometric research (Barro and Sala-i-Martin 1995: Chapter 12). But this is hardly surprising because, at least within an endogenous-growth framework, trade and growth are jointly determined. This means that some of the variables that have been found to affect economic growth across countries and over time, may actually do so in part through exports. One of the reasons why inflation, for example, seems to impede economic growth is that inflation hurts exports and thereby also imports, not only of goods, services, and capital, but also of ideas, information, innovation, and know-how (Gylfason and Herbertsson 1996). Foreign trade, like virtually all other sources of increased efficiency, is a likely source of economic growth, directly and indirectly (Sachs and Warner 1995a; Edwards 1998). But this does not mean that all trade is equally good for growth. High-tech trade 5 The almost 90 per cent share of primary exports in merchandise exports from Iceland includes aluminium and ferro-silicon exports, which account for about 10 per cent of the total.

302

T. Gylfason

seems more likely to encourage economic growth through technological spillovers than low-tech, labour-intensive trade. Taken together, the figures reviewed in Table 18.3 do not seem to suggest a particularly growth-friendly environment in the Nordic region. Only Norway and Denmark grew more rapidly than the world economy at large in 1980-96 (column 8). Economic growth in Iceland and Sweden was especially weak in this period, partly because of severe economic downturns in the early 1990s, but partly also for other reasons. The rest of this chapter discusses some channels through which economic growth in the Nordic countries may have been affected by their abundant natural resources. In earlier work, the share of natural resources in national wealth and the share of primary production in the labour force have been shown to be inversely related to investment, education, and exports among other things, and thereby also to economic growth. Here the aim is to explore the linkages between primary exports, total exports, investment, education, and growth with respect to the Nordic countries, and Norway and Finland in particular.

18.2. SELECTE D SYMPTOMS: D I A G N O S I N G TH E DUTCH DISEAS E Most authors have emphasized two closely related symptoms of the Dutch disease, among others: an overvalued currency that impedes non-primary exports and perhaps total exports as well, and a heavy dependence on natural resources which, in times of resource booms, is viewed as the root cause of the real overvaluation of the currency. However, as argued in earlier chapters, there is some evidence that natural resource abundance tends to be associated not only with sluggish non-primary exports, but also with slow investment and deficient education among other things, and thereby also with slow economic growth. Let us now review each of these in turn with one eye on the Nordic countries.

18.2.1. Trade There are two main ways in which the Dutch disease can manifest itself through exports, namely by making the composition of exports less favourable to economic growth and by reducing total exports and economic growth. The real appreciation of the currency may hurt just the kind of high-tech capital-intensive or high-skill labourintensive manufacturing and service exports that are particularly conducive to rapid growth. The more intriguing case, however, is the one where a boom in primary exports reduces non-primary exports krone for krone as in Norway, or worse. A priori, one might expect a positive correlation between the ratio of exports of goods and services relative to GDP and the ratio of primary exports to merchandise 6

See Sachs and Warner (1995b, 1999), Sachs (1999), Gylfason (1999b), and Gylfason el al. (1999). See Corden and Neary (1982), Wijnbergen (1984), Neary and Wijnbergen (1986), Gelb (1988), Matsuyama (1992), and Auty (1995). 7

Nordic Perspective on Natural Resources

303

exports. This is because resource-rich countries experience primary export booms at regular intervals through new discoveries and so on, and thus have higher export ratios than resource-deficient countries as long as non-primary exports are less than fully crowded out, ceteris paribus. The absence of a correlation might be viewed as a sign of full crowding out. Data for 158 countries over the years 1960—97, however, display a negative correlation between primary exports and total exports across countries (see Gylfason 1999a: Fig. 6), but the relationship is weak (t= 1.6, r= -0.12).8 A tenpoint increase in the primary export share from one country to another is accompanied by a decrease in the export ratio by one percentage point. But although all the Nordic countries are quite close to the regression line, the data do not provide any general indication that their export propensities are directly affected by their primary exports. Weak though it is, the relationship described above may stem from the tendency of natural-resource-related booms to lift currency rates and real wages, thus reducing exports, other things being equal, but exports depend on other factors as well. At any rate, there is no evidence of a positive relationship between primary exports and total exports. A simple correlation does not entail causation, however. It is conceivable that increased openness reduces the need for primary exports rather than, or as well as, the other way round. It is also possible that primary exports and total exports respond to third factors in ways that generate the pattern observed. No conclusions are being drawn here as to cause and effect. However, the weak relationship described above accords reasonably well with the results of a multivariate regression analysis, where an attempt is made to distinguish cause from effect (Gylfason 1999b). Moreover, the data cover a large and diverse group of countries at different stages of development as is customary in empirical growth research. This custom stems from the fact that the data for the high-income countries, if they were scrutinized separately, generally do not give rise to significant conclusions about the determinants of economic growth across countries, partly because there are simply not enough of them.

18.2.2. Investment Primary industries, not least agriculture and fisheries in developing countries, tend to be relatively low-tech and low-skill labour-intensive, so that their own investment needs as well as their encouragement of investment in other industries may be limited. Figure 18.3 shows the relationship between the ratio of gross domestic investment to GDP in 1960-97 and primary exports in 156 countries. Again, the correlation is significantly negative (t = 2.3, r= —0.19). A 22-point increase in the primary export share from one country to another goes along with a decrease in the domestic investment ratio by one percentage point. In its turn, a decrease in domestic The 8acorrelation equals the square root of R2. The significance of the correlation is established by a t-test of the significance of the slope of the regression line through the scatterplot.

304

T. Gylfason

Figure 18.3 . Primary exports and domestic investment

investment by 1 per cent of GDP typically shaves 0.1-0.2 percentage points off the annual rate of growth of output per head, ceteris paribus. Over the years, Finland and Denmark have invested just about the amount predicted by the regression shown in Fig. 18.3, Sweden less, Norway and Iceland, more. However, average investment rates mask a declining trend in all five countries since the 1960s. In 1996, the investment rates of all five Nordic countries were below the regression line in Fig. 18.3.

18.2.3. Education A third possible source of an inverse relationship between natural resource abundance and economic growth has to do with education. A strong emphasis on primary exports, not least agriculture in developing countries, by not calling for much highly trained manpower, tends to generate not only less investment in physical capital than otherwise, but also in human capital (Gylfason et al. 1999). The relationship between enrolment in secondary schools in 1980-96 and primary exports in 104 countries shows linkage that is significant (see Gylfason 1999a: Fig. 13), both economically and statistically (t = 8.4, r= —0.64). The regression suggests that a 1189-point increase in the primary export share from one country to another is associated with a one-point drop in the secondary-school enrolment rate. All but two of the almost 40 countries with a secondary-school enrolment rate of 30 per cent or less, have a primary export share of about 60 per cent or more. Here there is actually a discernible pattern even in the Nordic part of the picture: among the five countries, average secondary-school enrolment, which ranges from 86 per cent in Iceland to 94 per cent in Finland, is approximately inversely related to the primary export share. If natural resource abundance deters education, then this linkage may produce an inverse relationship also between primary exports and economic growth—through education. Even so, the causation may run the other way: primary production may be

Nordic Perspective on Natural Resources

305

Figure 18.4. Primary exports and economic growth, 1960-1997

decreasing the level of education. Most probably, though, primary production inhibits education, and vice versa.

18.2.4. Economic growth Let us now wrap up the argument by viewing the cross-sectional relationship between primary exports and economic growth since 1960. If natural resource abundance tends to hamper trade, investment, and education, it should also impede economic growth across countries, other things being the same. Figure 18.4 shows the relationship between the average annual rate of growth of per capita GNP from 1960 to 1997 and primary exports in 147 countries. The slope of the regression line is significantly negative (t = 7.2,r= —0.30). A 40-point increase in the primary export share from one country to another is associated with a reduction in per capita growth by one percentage point. This is not a small effect (if it is an effect, that is, as opposed to a mere correlation), because per capita growth rates move in a narrow range compared with primary export shares. Even if the analysis is limited to the 57 high- and uppermiddle-income countries in the sample, the pattern remains roughly the same as in Fig. 18.4. The inverse relationship between primary exports and economic growth is not confined to poor countries. The pattern shown in Fig. 18.4 comes fairly close to the quantitative results obtained from several recent multivariate regression analyses of economic growth across countries. A representative result from cross-sectional and panel studies is that an increase in the primary export share by 25—30 percentage points (e.g., from 50 per cent of merchandise exports to 75-80 per cent) reduces the rate of per capita growth from one country or period to another by one percentage point, ceteris paribus. See9. Gylfason (1999a), Gylfason and Herbertsson (1996), and Gylfason et al. (1999).

306

T. Gylfason

Similar results are obtained when natural resource abundance is measured by the share of primary production in the labour force.9 How do the Nordic countries fare in Fig. 18.4? Denmark and Finland are quite close to the regression line. Sweden lies a bit below the line. Even so, it seems unlikely that the relatively low average shares of primary exports in merchandise exports in those three countries, ranging from 22 per cent in Sweden to 45 per cent in Denmark compared with 70 per cent on average for the sample as a whole, have had much to do with their growth performance over the years. The share of natural capital in their national wealth ranges from 4 to 7 per cent (Table 18.2), compared with an average of 2 per cent in Western Europe, 5 per cent in North America, and 12.5 per cent for the world as a whole. The natural-resource-based industries in the three countries are thus relatively small compared with the world at large, even if they are large compared with those of other high-income countries. But what about Norway? Figure 18.4 shows that Norway has grown more rapidly than predicted by its primary export share alone, but this is not surprising. For one thing, economic growth obviously depends on a host of factors other than the primary export share. For another, the period covered by the figure, 1960-97, starts a full decade and a half before Norway became a significant oil exporter. The next section compares Norway and Finland in more depth as examples of countries with some dependence on point and diffuse resources, respectively.

1 8 . 3 . CONTRASTIN G NATURA L RESOURC E DEPENDENCE I N NORWA Y AN D FINLAN D 18.3.1. Norway Primary export dependence is one of the chief hindrances, real or imagined, to Norwegian accession to the European Union (EU) and the Economic and Monetary Union (EMU).10 There is, therefore, a need for Norway to deal with its naturalresource-based obstacles to EU membership if it is to be able to weigh the benefits and costs of membership on an equal footing with other prospective and present EU and EMU members. This requires the implementation of a market-friendly, fair, and property-rights-oriented solution to the problem of how best to regulate access to, and allocate the rents from, limited common-property natural resources. In fact, Norway has charted a long-run-oriented, tax-based, and reasonably market-friendly approach to the management of its vast oil resources.11 By law, the title to petroleum deposits on the Norwegian continental shelf is vested in the State. In principle, therefore, all the rent from oil and gas should accrue to the Norwegian people through their government. Exploration and production licences are awarded for a small fee to domestic and foreign oil companies alike. Why small? Because the Norwegian government has decided to expropriate the oil and gas rent through taxes and fees as well as direct involvement in the development of the resources rather than through sales or 10 11

This argument also applies to Iceland, see Gylfason (1999a). Estimates of Norway's oil wealth range from 50 to 250 per cent of GNP (Th216gersen 1994).

Nordic Perspective on Natural Resources

307

auctioning of exploration and production rights (OECD 1999: Chapter 3). The State has a direct interest in most offshore oil and gas fields and, like other licensees, receives a corresponding proportion of production and other revenues, roughly 40 per cent of the total. Through its direct partnership with other licensees as well as through various taxes and fees, the Norwegian State has managed to absorb about 80 per cent of the resource rent since 1980.12 Thus, in 1997, revenues from petroleum activities accounted for more than a fifth of total government revenues and were equivalent to 9-10 per cent of Norway's mainland GNP, or 8-9 per cent of total GNP, including oil The oil revenue is deposited in the Norwegian Petroleum Fund, which is being built up and invested mostly in foreign securities. Oil exports account for about a third of total exports of goods and services from Norway. The oil industry contributes about one-sixth of Norway's GDP (in 1997). At the same time however, a variable proportion of each year's net oil-tax revenue is transferred from the Government Petroleum Fund to the fiscal budget, essentially to cover the non-oil budget deficit. The proportion of net tax revenues from petroleum thus transferred to the government budget was about one-fourth in 1997 and almost 40 per cent in 1998, but is envisaged to drop to less than 10 per cent in 1999. Even so, the Norwegians have not been tempted to expand their central government beyond reasonable limits as a result of the oil boom. Even 20 years after discovering their oil, the Norwegians continue to content themselves with a smaller central government than Denmark, Finland, and especially Sweden. The upshot of all this is, first, that the Norwegians are already preparing themselves with care for a (fairly distant) future without oil and, second, that 'sharing the oil with foreigners' and related concerns does not arise in connection with Norway's oil wealth in the discussion of the pros and cons of potential Norwegian EU membership. Thanks to the market-oriented approach to oil-resource management as well as to the legal status of Norway's oil reserves as a taxable common-property resource, oil does not stand in the way of Norway's entry into the EU, if this is where the Norwegian people want to go at the end of the day. The Norwegians' management of their fish resources is rather different from their handling of their oil wealth. Norway's fishing industry is actually tiny, employing, like the oil sector, less than 1 per cent of the country's labour force. Agriculture, forestry, and fishing together account for about 2 per cent of GDP, and their share is declining. Of this small share, the fisheries account for less than a half. Government subsidies to the fishing industry increased successively from the 1950s onwards until they peaked at about 70 per cent of the incomes of fishermen and boat-owners in 1981 (Hannessori 1996: 23-24). Since then, however, the subsidies have been reduced in stages, down to almost nothing. Even so, the government carries the cost of managing the fisheries and of enforcing fishery regulations; this cost is considered equivalent to about 10—15 per cent of the gross value of the catch (ibid: 30). Moreover, virtually all the resource rent from the fisheries, roughly estimated at 20-25 per cent of the gross value of the catch 12 The main revenue items are corporate tax and a special resource surtax, but also royalty, area fee, and carbon-dioxide tax.

308

T. Gylfason

(ibid: 29), has been allowed to dissipate through excess capacity and overmanning. This matters because the fishing industry's vociferous protests were seemingly the single most important factor contributing to the Norwegians' rejection of EU membership in the referendum of 1994 as well as in 1972. Part of the problem is that the Norwegian fishing industry is perceived to be much larger than it actually is. This is partly because the fishing industry is quite important to individual coastal communities, even if it is unimportant in a macroeconomic sense to the Norwegian economy as a whole. A vocal fishing lobby also does its best to insure that this false perception does not fade from the public consciousness. Anyhow, it is inefficient to tie regional support to particular industries, such as fisheries or agriculture. It would be more efficient to aim subsidies or other regional-policy instruments at the regions concerned with no strings attached rather than at specific industries, and thus to allow the recipients themselves to decide whether they want to continue to fetch fish from the sea or do something else. However, even if the expansion of oil exports from Norway since the mid-1970s seems to have left total exports essentially unchanged relative to GDP, the Norwegians have nonetheless been able to manage their oil resources in a way that has removed any oil-related hindrances from the road that could lead them into the EU, if this is where they want to go. In view of their market-friendly management of their oil wealth, there is no economic reason why the Norwegians could not in the same manner improve the management of their fish resources through fishing fees or through the sale of transferable catch quotas so as to remove the chief remaining hindrance on their way to full membership of the EU.

18.3.2. Finland There have been concerns in Finland that the country's considerable and longstanding dependence on its forest resources, which account for two-thirds of its natural capital and about a third of its exports, might render Finland exposed to the Dutch disease. As before, let us organize the argument around trade, investment, and education. First, Finnish exports have expanded quite rapidly since 1960: they have almost doubled, from 22 per cent of GDP in 1960 to 38 per cent in 1997 (Fig. 18.1). Export growth and the willingness to participate in and contribute to European integration, thus do not seem to have been held back by an overwhelming emphasis on forestry-related exports at the expense of other exports. Even so, Finland, like Sweden, Denmark, and even Norway, is only an 'average' exporter in the sense that its exports are about the same relative to GDP, as they are in other countries of the same size. Second, the Finnish forestry sector is quite high-tech-intensive and seems to have stimulated other sectors through various technological spillovers. For example, Finland's share in forestry-related machinery and equipment in world markets is actually larger than its share in wood, paper, and pulp. Finland has a strong manufacturing sector whose share in total exports increased from 56 per cent in 1963 to 83 per cent in 1997 (Fig. 18.2). Nokia grew out of an industrial firm that was partly, albeit only to a small extent, in the forestry business before.

Nordic Perspective on Natural Resources

309

Third, the capital intensity of forest resource products has understandably called for large investments in Finnish forestry, thereby attracting capital from other sectors. This seems to have changed not only the sectoral composition of capital accumulation, but may also have helped change its total amount, because investment in Finland has fallen quite dramatically since 1960 from just under 30 per cent of GDP to just over 16 per cent. This type of more than full crowding out of non-primary investment is, however, unlikely to be the sole explanation for Finland's low aggregate investment because, for one thing, the slump in investment in Finland coincided with the economic crisis in the 1990s and, therefore, seems likely to have other causes as well, including various structural flaws and policy problems that contributed to the crisis in addition to the collapse of the Soviet Union.13 Some observers have argued that the rapid growth of the Finnish economy from 1960 to 1990 resulted in large measure from forestry-related, partly state-directed, investment projects which kept total investment at 25-30 per cent of GDP throughout this period (Tainio et al. 1999). When financial markets in Finland were liberalized in the late 1980s however, and real interest rates rose, the argument goes, some of these investments proved inefficient, and total investment plunged. Fourth, there is no discernible sign of a lack of commitment to education in Finland. In 1995, the Finnish government spent 7.6 per cent of GNP on education compared with 4.8 per cent on average in the world as a whole.14 Expenditure on education in Finland has increased in recent years from just under 5 per cent of GDP to around 7.5 per cent. Even so, Finnish outlays on education lag a bit behind those of Denmark, Norway, and Sweden (Table 18.3), but the differences are small. Fifth and last, forest ownership in Finland is fairly evenly distributed among the Finnish population, and has been that way for a long time. Forestry-related booms in the Finnish economy have not created a super-rich but small class of politically powerful rent seekers. Rather, the benefits that have emanated from the country's natural resources have been spread widely across the whole population without creating political or social tensions. There is, therefore, no discernible evidence of the recurrent economic policy failures that tend to characterize rent-seeking societies over long periods—including tendencies towards protectionism against trade and other forms of economic integration, and towards thinking that natural resources are more important than human resources and so on. True, like the fishing industry in Norway and Iceland (and, though to a much lesser extent, the oil industry in Norway), the forest industry in Finland has considerable political clout. This explains, in part, the repeated devaluation of the markka in the 1970s and 1980s, which was aimed mainly at preserving the profitability of the forest industry. Government support of the forest industry has not, however, taken the form of direct subsidies or trade restrictions. In sum then, Finland passes the test for the Dutch disease on at least four scores out of five: (a) its exports have grown rapidly relative to GDP since 1960, even if they are 13

See Honkapohja et al. (1996) and Jonung et al. (1996). Thisce, put Finland in 14th place among 170 countries in 1995, while Denmark landed 8th place, Norway 9th, Sweden 12th, and Iceland 76th.

310

T. Gylfason

only 'average' by world standards; (b) the share of primary exports in total exports has fallen sharply since 1963 as the natural-resource-based industry seems to have encouraged other industries through technological spillovers; (c) investment has fallen sharply relative to GDP in the 1990s, true, but it does not seem reasonable to ascribe but a part of the fall directly to Finland's dependence on forest resources; (d) education compares favourably with other OECD countries; and (e) the ownership of and access to the main natural resource are rather widely shared so that macroeconomically counter-productive rent seeking is not a serious problem. It would seem that Sweden and Denmark also pass this five-pronged test (Gylfason 1999a). Norway, however, passes the test only in part. Norwegian exports have been stagnant since 1960, albeit at a respectable level, primary exports still account for three-quarters of total exports, and investment has decreased over time. On the other hand, Norway shows no signs of having faltered in its commitment to education, nor has the natural resource rent been allowed to fall into too few hands.

18.4. CONCLUSIO N It has been argued that natural resource abundance may retard economic growth in three important ways: by reducing total exports and thereby, ultimately, also imports of goods, services, and capital relative to national income; by reducing investment in physical capital; and by reducing investment in human capital (education). Consistent with earlier findings in this book, however, this is not an inevitable outcome if appropriate institutional safeguards are in place, whether the natural resource has diffuse or point socioeconomic linkages. Even so, an echo effect of resource dependence may be detected. Norway shows that the trade channel does not merely involve exports and imports, it may also operate through a political conflict between natural resource dependence and the propensity to participate in, and benefit from, international economic integration. This mayhelp explain why the authorities in Norway still show no signs of wanting to join Denmark, Finland, and Sweden in the European Union, even as the Central and Eastern European countries are queuing up outside the gates. To those Norwegians (and Icelanders) who want their countries to be full participants in European integration, this factor alone is, perhaps, ample reason to fear that heavy natural resource dependence may be, at best, a mixed blessing in the long run.

REFERENCES Auty, Richard M. (1995), Patterns of Development: Resources, Policy and Economic Growth, Edward Arnold, London. Barro, Robert J. and Xavier Sala-i-Martin (1995), Economic Growth, New York: McGraw-Hill. Corden, W. Max (1984), 'Booming sector and dutch susease ecomicscs: survey and conolidation', Oxford Economic Papers, 36, 359—80.

Nordic Perspective on Natural Resources

311

Corden, W. Max and J. Peter Neary (1982), 'Booming sector and de-industrialization in a small open economy', Economic Journal, 92, 825-48. Edwards, Sebastian (1998), 'Openness, productivity and growth: what do we really know?', Economic Journal, 108 (March), 383-98. Gelb, Alan (1988), Windfall Gains: Blessing or Curse?, New York: Oxford University Press. Gylfason, Thorvaldur (1999a), 'Natural resources and economic growth: a Nordic perspective on the Dutch disease', UNU/WIDER Working Papers No. 167, Helsinki: UNU/WIDER. Gylfason, Thorvaldur (1999b), 'Exports, inflation, and growth', World Development, 27 (June): 1031-57. Gylfason, Thorvaldur and Tryggvi Thor Herbertsson (1996), 'Does inflation matter for growth?', CEPR Discussion Paper 1503. Gylfason, Thorvaldur, Tryggvi Thor Herbertsson and Gylfi Zoega (1999), 'A mixed blessing: natural resources and economic growth', Macroeconomic Dynamics, 3 (June): 204—25. Hannesson, Rognvaldur (1996), Fisheries Mismanagement: The Case of the North Atlantic Cod, Oxford: Fishing News Books. Honkapohja, Seppo, Erkki Koskela and Jouko Paunio (1996), 'The depression of the 1990s in Finland: an analytic view', Finnish Economic Papers, 9, 37-54. Jonung, Lars, Joakim Stymne and Hans Tson Soderstrom (1996), 'Depression in the north. Boom and bust in Sweden and Finland 1985-1993', Finnish Economic Papers, 9, 55-71. Matsuyama, Kiminori (1992), 'Agricultural productivity, comparative advantage, and economic growth', Journal of Economic Theory, 58, 317—34. Neary, J. Peter and Sweder van Wijnbergen (1986) (eds.), Natural Resources and the Macroeconomy, Oxford: Basil Blackwell. OECD (1997), Education at a Glance, OECD Indicators, Paris. OECD (1999), Economic Surveys: Norway, 1999, Paris. Sachs, Jeffrey D. (1999), 'Why economies grow', in Birgitta Swedenborg and Hans Tson S246derstr246m (eds.), Creating an Environment for Growth, Stockholm: SNS F246rlag: 17-31. Sachs, Jeffrey D. and Andrew M. Warner (1995a), 'Economic reform and the process of global integration', Brookings Papers on Economic Activity, 1, 1-118. Sachs, Jeffrey D. and Andrew M. Warner (1995b), 'Natural resource abundance and economic growth', NBER Working Paper 5398. Sachs, Jeffrey D. and Andrew M. Warner (1999), 'The big push, natural resource booms and growth', Journal of Development Economics, 59/1, 43-76. Tainio, Risto, Matti Pohjola and Kari Lilja (1999), 'Economic performance of Finland after the Second World War: from success to failure', in S. Quack, G. Morgan and R. Whitley (eds.), National Capitalism, Global Competition and Economic Performance, De Gruyter. Th216gersen,216ysten(1994), 'Economic policy, macroeconomic performance and the Norwegian petroleum wealth: a survey', Stiftelsen for Samfunns- og N230ringslivsforskning, Norwegian School of Business Administration, Bergen. Wijnbergen, Sweder van (1984), 'The Dutch disease: a disease after all?', Economic Journal, 94, 41-55. World Bank (1997), 'Expanding the measure of wealth: indicators of environmentally sustainable development', Environmentally Sustainable Development Studies and Monographs Series No. 17, World Bank, Washington, DC. World Bank (1999), World Development Indicators 1999, Washington, DC: World Bank.

This page intentionally left blank

PART VI

CONCLUSIONS

This page intentionally left blank

19 Conclusions: Resource Abundance, Growth Collapses, and Policy RICHARD M . AUTY

19.1. AN OVERVIE W O F RESOURC E ABUNDANC E AN D GROWTH COLLAPSE S 19.1.1. Temporal clustering of growth collapses in resource-abundant countries In the final decades of the twentieth century, a counter-intuitive relationship emerged between natural resource abundance and economic development. Since the mid1970s, the resource-poor countries have tended to grow significantly faster than the resource-abundant countries (Fig. 19.1). Although the larger resource-poor countries grew somewhat faster than the smaller resource-poor countries, the latter still easily outperformed all four categories of resource-abundant country (Table 2.3). Significantly, among the resource-poor countries, those with the most constrained resource-endowment and smallest rent stream, the resource-poor countries of East Asia, sustained exceptionally rapid economic growth that was also relatively egalitarian and propelled them to the status of high middle-income countries in barely two generations. At the other extreme, the resource-abundant countries with the highest share of natural resource rents in GDP, the small oil-exporting countries, have exhibited least economic resilience. Nevertheless, most countries within each of the three other resource-abundant subcategories experienced a growth collapse after falling into a staple trap of growing dependence on commodities whose competitiveness was eroding. Even the larger resource-abundant countries did so, although they had the most favourable conditions for evading a staple trap. Subsequently, the pace of recovery from the growth collapses has been disappointing, even when the resource-abundant countries strongly embraced rapid economic reforms. This is especially so for the smaller and most rent-rich countries. The underperformance of the resource-abundant countries is not inevitable however, because policy counts. Growth collapses occur if policies are adopted that retard competitive diversification so that the economy becomes locked into a staple trap. The economic development of a handful of small resource-abundant countries in differing cultural regions with differing permutations of natural resources (Malaysia, Thailand, Oman, Botswana, and Chile) supports this conclusion. So too does the historical

316

R. M. Auty

Figure 19.1 . Median annual growth in GDP per capita (const 1995 US$) resource-rich and resource-poor Source: World Bank (1999).

evidence presented in Chapter 6, which suggests that resource-abundance need not be a curse. During the First Golden Age of Economic Growth from 1870-1913, resourceabundant countries in both the temperate lands of new settlement and the tropics alike, experienced economic development if they availed themselves of the expanding trade opportunities afforded by sharply falling transport costs. During those years, the resource-abundant countries of 'new' settlement in North America and in the temperate zones of the Southern Hemisphere began to diversify into competitive manufacturing. Within the tropics, many primary product exporters of Asia and sub-Saharan Africa experienced economic growth, although they lagged the Latin American countries in industrial diversification. Their failure to industrialize may be the result of insufficient demand in their small domestic markets due to low wages caused by low productivity in tropical staple food production (Lewis 1978). One important consequence of this staggered start upon industrialization is that the lagging tropical countries still relied heavily on staples after the Second World War, when renewed trade liberalization and further falls in transport costs caused the world economy to boom again. In fact, the resource-abundant countries made a propitious start on economic growth that lasted until the early-1970s, but their growth then faltered while the resource-poor countries closed the income gap, and then forged ahead (Fig. 19.1). In addition to the historical evidence, further support for the non-deterministic nature of the underperformance of the resource-abundant countries in recent decades emerges from the short-run three-sector macroeconomic model of resource-abundant and resource-poor development presented in Chapter 7. The model captures important aspects of the dynamics that differentiate resource-abundant and resource-poor countries. In the resource-abundant case the model does show that a resource boom can lead to the rise of the non-traded sector at the expense of traded goods, and that a devaluation might be contractionary. An expansion in the money supply induced by

Conclusions

317

capital inflows amplifies these effects by expanding the economy. Conversely, when the model is calibrated for the resource-poor case, resource booms might even expand the non-resource traded sector, and devaluation may also be expansionary. However, the short-run macro model also demonstrates how resource-abundant countries can avoid a growth collapse with appropriate policies. More specifically, a resource-boom (or increase in transfers from abroad) should be accompanied by policies to limit exchange rate appreciation, or a policy-induced devaluation. In addition, policies that tax nontraded goods in order to foster traded goods production may also be effective under some circumstances. The model suggests that the key requirement is to pursue policies that sustain the competitive diversification of the economy. In fact, the policies adopted by many resource-abundant countries after the Second World War were based on scepticism concerning the role of markets. Impatience with commodity-driven growth due to the perceived long-run fall in the real prices of the commodity exports of the developing countries, and the instability of those prices, prompted the adoption of policies that have back-fired on the resource-abundant countries in two ways. First, at a national level, many developing countries adopted closed trade policies in order to reduce their dependence on commodity exports by forcing the pace of industrialization through protecting infant industry. This led to the policy being captured by those interest groups that benefited from it. The availability of natural resource rents prolonged the period of policy error and investment misallocation and so intensified the distortion of the economy. Commodity booms were frequently used to reverse or postpone necessary reform rather than to facilitate its implementation. In this way, the natural resource rents sustained policy error and retarded competitive structural change so that the resilience of the economy was impaired. The second policy error occurred at the global level and compounded the first policy error. Efforts to establish a New International Economic Order that would maintain or enhance the real prices of commodity exports triggered severe fluctuations in the rent streams during the 1970s. The resulting external shocks, whether negative or positive, strong or modest, impacted weakened economies and caused growth collapses in most resource-abundant developing countries (Fig. 19.1 and Table 2.3). The adverse relationship between natural resource abundance and economic development is therefore rooted in post-war policy prescriptions that overestimated the capacity of most developing countries to pursue a coherent economic policy. To the extent that policy error caused the growth collapses and policies are options, the inverse relationship between natural resource abundance and economic development is not a deterministic one.

19.1.2. Diverging structural change and development trajectories Nevertheless, the pattern of structural change with rising per capita income examined in Chapter 2 does favour the resource-poor countries whose early start on competitive industrialization triggers favourable self-reinforcing economic and social circles (see Fig. 8.1). In contrast, structural change causes resource-abundant countries to rely longer on commodity exports. This retards competitive industrialization, and

318

R. M. Auty

therefore urbanization, and slows passage through the demographic cycle. The postponed expansion of competitive industry severely tests the competence of the governments of resource-abundant countries in generating sufficient employment, raising skills, and building the social capital with which to reduce transaction costs and manage social tensions and economic shocks. Nevertheless, the tests posed by resource-abundant development were, in fact, met where local circumstances spawned a developmental state. The developmental state has both the capacity to pursue coherent economic policies and the aim of maximizing social welfare over the long-term. The consensual democracy in Malaysia provides one example, while the unlikely evolution of Chile after the collapse of the Allende populist boom into first an autonomous benevolent state and then a consensual democracy, affords another. Unfortunately, most resource-abundant countries had predatory or factional political states with limited capacity or inclination, or both, to pursue economic policies that maximize social welfare. In predatory and factional states, the natural resource rents are captured and redistributed by non-transparent means to favoured constituencies within protected industry, the public sector, the military or regional or ethnic groups.

19.1.3. Natural resources, political state, and policy The link between the natural resource endowment and type of political state is clear at either extreme of the natural resource endowment spectrum, but further research is required at more intermediate levels where generalization is more difficult. The benevolent autonomous developmental state is strongly associated with the resourcepoor countries of Northeast Asia where the pressure of population on scarce natural resources was extreme by the mid-twentieth century. These conditions have proved conducive to the realignment of the political state with the interests of the majority poor at a relatively low per capita income. There are four important reasons for this. First, extreme population pressure on scarce land resources appears to reduce the tolerance of the majority poor within the population for inequitable land distribution and predatory rent extraction. Second, it becomes clear at a relatively low income level in resource-poor countries that neither the primary sector nor competitive manufactured exports can support either a slow-maturing manufacturing sector or an overexpanded bureaucracy. Third, a spartan natural resource endowment places a premium upon the efficient use of scarce investment resources. This works against the relaxation of market discipline and fosters early investment in the potentially most abundant resources, human and social capital. A final reason is that resource-poor countries experience only muted Dutch disease effects so they are therefore less prone to close their trade policy and distort market signals. Other resource-poor countries subsequently evolved towards developmental states (usually the consensual democracy variant) and embarked on competitive industrialization as their per capita cropland fell to similarly low levels. Examples include Mauritius after the 1970s and, albeit less vigorously, Bangladesh since the 1980s. At the other extreme, generous resource-abundance is associated with the capture of resource rents by the government and their reallocation, often through measures with

Conclusions

319

low transparency, to advance the interests of cohesive groups. Even where resourceabundant countries did spawn developmental political states, as in many of the capitalsurplus oil-exporting countries, their governments frequently resorted to paternalism to manage social tensions and, in the process, they relaxed market discipline and distorted the economy. This occurred even without trade policy closure, so that Dutch disease effects do not automatically lead to trade policy closure as Sachs (1999) appears to suggest. This study also demonstrates that trade policy may be closed for reasons other than the Dutch disease, the most important of which is the espousal of policies to force industrialization by protecting infant industry. It is not yet established whether forced industrialization triggers factional and predatory political states or merely provides a means for them to consolidate their power. On the one hand, Latin America provides evidence that factional oligarchic states can evolve from structures set up by earlier landed elites, but on the other hand they may also spring from tribal rivalries after the withdrawal of colonial powers, as in sub-Saharan Africa. Ghana shows how quickly forced industrialization can entrench powerful urban interests that, in turn, capture government policy and retard reform in the face of faltering economic growth. Factional oligarchic states may degenerate into predatory states or they may shift towards democracies that are, however, likely to be polarized and to redistribute natural resource rents at the expense of long-term economic growth. Yet severe political or economic crises, like a growth collapse, may transform factional states into developmental consensual democracies. Even so, Costa Rica suggests that such a state may still experience a growth collapse, albeit a relatively mild one, if the social groups with a stake in competitive investment are relatively weak and a redistributive policy of 'support-led security' is adopted, rather than the 'growth-based security' of Malaysia. The fact that a consensual democracy grew out of a specific configuration of ethnic interests in Malaysia, and from a quite different combination in Botswana (under circumstances more commonly associated with a paternalistic autonomous benevolent state) reflects a complexity that, as yet, precludes stronger generalization. Summarizing, the natural resource endowment strongly impacts the development trajectory by differentiating the pattern of structural change and the character of the political state. This conditions the development strategy espoused, and so determines the efficiency with which inputs are deployed and the rate of competitive diversification of the economy. Moreover, policy error is likely to persist and the economy become more distorted, the larger the rent stream in relation to GDP and the less diffuse the socioeconomic linkages of the rent stream.

19.2. TH E COMPETITIV E I N D U S T R I A L I Z A T I O N MODE L

19.2.1. Competitive industrialization in the resource-poor countries Before exploring the policy implications, it is useful to summarize the model of competitive industrialization and contrast it with the staple trap model. Under a

320

R. M. Auty

developmental state, the resource-poor countries embark on a trajectory of competitive industrialization at a relatively low per capita income. This establishes virtuous interlocking economic and social circles to sustain rapid and equitable economic growth (see Fig. 8.1). The virtuous economic circle exhibits three key characteristics. First, the expansion of labour-intensive manufacturing, much of it for export, quickly absorbs surplus rural labour and pushes the labour market to its turning point. This creates pressure for wage increases that require the economy to diversify into competitive heavy and chemical industry that, in turn, reduces the vulnerability of the economy to external shocks. Second, early and rapid industrialization accelerates the rate of urbanization and, therefore, speeds up the demographic cycle. Consequently, the dependent/worker ratio falls and this stimulates saving and investment. The rate of investment quickly rises to 25 per cent of GDP or higher, with only a minor decline in investment efficiency (linked to the increased capital intensity of development), so that per capita GDP growth tends to double within a decade. Third, the early reform builds a constituency in support of efficient investment while the rapidly increasing complexity of the economy fosters a 'linear liberalisation' that constrains tendencies towards rent-seeking behaviour. The virtuous social circle also has three main features. First, the tightening of the labour market helps to maintain income equality by removing the drag of surplus rural labour on the wages of the poor. Meanwhile, the rapid upgrading of workforce skills compresses the income range by reducing the wage premium on higher skills, setting in train the equality with human capital accumulation circuit described in Chapter 4. Second, the relatively equitable distribution of income helps reduce transaction costs by boosting social capital accumulation, as described in Chapter 5. Third, these social trends combine to push the political economy towards a democracy that tends to be consensual rather than polarized, so that the coherence of economic policy is sustained while power is contested in a manner that reduces the risk of policy capture.

19.2.2. Competitive industrialization in the resource-abundant countries The longer initial dependence of resource-abundant countries on primary product exports, slows the rate of economic growth and postpones the labour-intensive stage of the model. This has three key consequences. First, diversification must initially occur into other primary products and this may prove especially difficult for the smaller countries with skewed natural resource endowments. Second, slower industrialization retards urbanization and delays the favourable middle stage of the worker/dependency cycle so that the accumulation of capital (including human capital) is slower in the absence of incentives, to encourage household saving or effective government saving on behalf of poorer groups. Third, the labour market turning point is postponed, so that surplus rural labour persists and skill acquisition is retarded, while income inequality, and its associated social tension, is amplified. The combination of slower economic growth and higher social tension places greater strain on the political

Conclusions

321

economy of the resource-abundant country and raises the risk that resource rents will be diverted to appease social groups at the expense of sustained economic growth. Yet, a handful of resource-abundant countries (noted earlier) show that a developmental political state can emerge in a resource-abundant country and achieve competitive industrialization (albeit much more slowly where the rent stream is unusually large and sustained, as in Botswana). Malaysia achieved competitive industrialization despite the risk of severe racial tension that has often been blamed for policy failure elsewhere. In fact, the riots of 1969 were the catalyst for reform since they led directly to the New Economic Policy. At the heart of this policy was a social contract that espoused a growth-based-security strategy, which redistributed assets in favour of the lagging rural Malay majority without choking off the commercial opportunities of the large Chinese minority. This strategy favoured prudent macroeconomic policy (prompt attention to fiscal and trade deficits) and retained market incentives (including a competitive exchange rate and an open trade policy). However, the Malaysian government went beyond the role of the 'enabling state' and accumulated assets on behalf of the majority poor, encouraged saving from a low per capita income level, and provided incentives for competitive manufacturing in export processing zones (EPZs).1 Its industrial policy was dualistic and more labour-intensive than protected infant industry while also avoiding burdening the primary sector (from which the government drew much of its political support) with onerous imposts. Subsequently, Mauritius, China, and most recently Namibia and Costa Rica (see Chapter 14) have benefited from export processing zones. Other Malaysian intervention was less successful: Edwards (1990) estimates that 4 per cent of GDP annually may have been transferred to infant industry during the mid-1970s while the HCI Big Push of the early-1980s was also ill-conceived. The scale of misallocation of resources was, however, small enough to be tolerated.

19.2.3. The staple trap model Yet, the developmental state proved the exception for resource-abundant countries during the Second Golden Age of Economic Growth. Dissatisfaction with the pattern of growth associated with a longer dependence on primary exports led to policies to force industrialization through infant industry protection and/or expansion of public sector employment. This study shows that in an intellectual climate that encourages an ambitious strategy of development, like forced industrialization, natural resource abundance is likely to adversely impact economic development (see Fig. 8.2). It does so by increasing the probability that: « the political state will be a factional or predatory state that consolidates its power by distributing natural resource rents to favoured social groups in the urban/industrial sector at the cost of a coherent economic policy for increasing social welfare,

Non1o-Moslem and non-Catholic countries might also wish to encourage birth control in order to accelerate passage through the demographic cycle.

322

R. M. Auty

• policies are adopted that nurture a non-tradable sector that includes slow-maturing capital-intensive industry and an unproductive public sector that demand transfers and foreign exchange from the primary sector. The transfers outstrip the rents and depress incentives in that sector in a negative economic circle, • slower rates of urbanization and demographic transition retard capital accumulation so that investment fails to match the levels under competitive industrialization, while more importantly, cumulating economic distortions depress capital efficiency, • the economy falls into a staple trap and specializes in exporting commodities whose viability is eroded so that the competitive diversification of the economy is not achieved and the vulnerability of the economy to shocks is enhanced, • the labour-intensive stage of industrialization is leapfrogged so that surplus labour depresses the wages of the rural poor and income inequality persists that slows social capital accumulation and intensifies sociopolitical tensions, weakening the political capacity to cope with shocks, • the rate of per capita GDP growth will slow and become erratic as governments seek to postpone reform in the face of strong political opposition from the non-tradable sector, • the economy will experience a growth collapse that damages all forms of capital, including natural and social capital, so that recovery is protracted. One important consequence of the staple trap is to cause high population growth to coexist alongside slow or negative economic growth and regressive structural change. This means that resource-abundant countries can move towards a condition of resource paucity within one or two generations. This has been the case in several regions of sub-Saharan Africa, notably the Sahel, as well as the Caucasus and Central America. Ghana, Armenia, and Costa Rica provide specific examples. The overall result is that in the absence of timely reforms, resource-poor countries can become locked into the staple trap so that an extra category of development trajectory might be added to Table 8.6. It could also be argued that the root cause of disappointing economic growth is a policy that maintains dependence on slow-growth staples, rather than simply the mismanagement of natural resource abundance. However, such a view loses the insight provided by an appreciation of how the mismanagement of natural resource rents brings this about. 1 9 . 3 . POLIC Y I M P L I C A T I O N S FO R DEVELOPIN G MARKE T AND TRANSITIO N ECONOMIE S

19.3.1. The importance of sound macroeconomic policies The international financial institutions (IFIs) have played a central role through the 1980s and 1990s in efforts to revive both the developing market economies that suffered a growth collapse and also the collapsed centrally planned economies. In exchange for financial assistance from the IFIs, the recipient countries have been required to adopt policies that stressed stabilization, market liberalization and tighter integration into the global economy. The findings of this study support the stress on

Conclusions

323

sound macro management. However, the study also demonstrates the extent to which most countries diverge from the developmental political state and lack the social capital required to sustain resource-driven development. Although the broad policies required for sound macro management are well-established, the difficulty lies in their implementation. Enhanced policy success requires institutional reinforcement that encourages the growth of a developmental state. Price stabilization is a prerequisite for renewed economic growth. It requires fiscal discipline so that budget deficits remain small enough to be financed without recourse to the inflation tax, which implies an operational deficit not exceeding 2 per cent of GDP (Williamson 1996). Since public expenditure is likely to be reduced, it needs to become sensitive to social returns, rather than to political returns. It must therefore be redirected away from vested interests towards hitherto neglected high-return areas that also improve income distribution like primary education, health care, and infrastructure. On the revenue side, overdependence on the primary sector requires tax reform to broaden the tax base and reduce marginal rates so that incentives are improved without excess regression. Yet, such a shift in the tax burden requires cutbacks in public investment (which are politically easier to make than cutbacks in current expenditure in times of fiscal austerity), and it may reduce private savings. Consequently, economic growth prospects receive a double blow because the slower growth in the economy during reform is accompanied by an overall fall in domestic savings and investment. Price liberalization calls, among other things, for exchange rate unification at a level to sustain long-term competitiveness. This will facilitate trade liberalization by the initial removal of quotas and progression towards a single low import tariff around 10 per cent and no higher than 20 per cent. Macro theorists (e.g., Davis 1995) wrongly assume that shifts in the real exchange rate will be clearly signalled so that the relative size of the non-tradable and non-mining tradable sectors will automatically adjust. During an upswing, the former will expand at the expense of the latter, only to reverse as the mineral boom tapers off. In fact, the adjustment is rarely as smooth, and capacity in the tradables sector that is lost during an upswing is not quickly replaced during a downswing. Moreover, sharp fluctuations in export revenue may place pressure on the real exchange rate and destabilize government expenditures. A real exchange rate appreciation can be avoided by sterilizing export revenue accruing directly to the government or from the proceeds of taxes on commodity rents. Central banks can also operate in exchange markets to prevent real exchange rate appreciation. EARA provides a rationale for allocating a portion of such revenues to a capital development fund, especially in the case of rents from nonrenewable resources such as minerals. Such rents might be held in offshore accounts to enhance sterilization and provide a source of future domestic investment in projects that promise a high return. This investment could subsequently help to counter the commodity boom-bust cycle. In addition, a commodity revenue stabilization fund (CRSF) can help to manage both the wwpredictability of the revenue stream over the long-run and the short-run, and also to deal with the predictable inflationary effects of an over-rapid absorption of

324

R. M. Auty

commodity rents. During commodity booms a CRSF automatically accumulates government revenues above a predetermined limit to be saved by the Treasury or Central Bank. This smoothes the deployment of the revenue streams over the cycle and limits the extent to which governments can use windfalls to buy political support. Such caution is preferable because the economic damage arising from a permanent favourable price shift that is assumed to be temporary is less than that arising if a temporary price improvement is assumed to be permanent. This is because of the dislocation, which arises when the revenue expansion is overestimated so that resources must then be withdrawn from the economy to restore economic balance. Sharp reductions in government expenditures that impact low-income groups as well as requiring suspension of essential capital projects, should be avoided. Financial liberalization towards market-determined rates is also desirable but in the face of uncertainty this may push interest rates to levels that stifle growth by deterring investment, bankrupting efficient firms, and squeezing public revenues through high debt service (Brewster 1994). Some compromise may therefore be required that allows intervention to sustain moderately positive real interest rates. Finally, property rights should be secured to encourage foreign investment and domestic private investors, including the informal sector (as an important source of employment creation and skill acquisition). This also requires a reduction in regulation to remove barriers to entry, reduce scope for rent-seeking and corruption, and concentrate effort on essential requirements like health, safety, and the environment.

19.3.2. The timing and adaptation of economic reform A controversy arose over the speed of the reform, with the IFIs tending to favour rapid 'big bang' reform over gradual reform. Fears of 'policy fatigue' may have driven the IFIs to focus upon the speed and sequencing of reform, rather than the strength of supporting institutions and other initial conditions. It may also be reasonable to assume that reform can exert sufficient pressure to create the requisite institutional changes and that, where change proves inadequate, a different approach to reform would not produce a better outcome. However, the recovery of the reforming economies has been slower than policymakers expected and this has raised doubts concerning the merits of the rapid reform policies prescribed by the IFIs. The compression of reform may bring several years of economic stagnation before a robust recovery is under way. The stagnation might be avoided by a combination of more gradual fiscal and monetary tightening and slower trade liberalization plus government spending and investment incentives that would have their principal impact on the lower income classes. Not surprisingly, economic reforms are often seen as harming those with low incomes, thus compounding the political difficulty of implementation. They pose special difficulties for factional political states because their governments are likely to respond to the short- and medium-term hardships of adjustment in a way that diminishes the long-term benefits (Mahon 1992). Opposition to economic reform may result in the election of a new government that abandons the reform. Therefore, the reform policies may require the government to boost demand

Conclusions

325

in order to mitigate unemployment and lower growth. This risks over-stimulating the economy, however, as in the case of the populist booms reported here in Mexico and Argentina during the mid-1970s. A more insightful discussion of the merits of different speeds of reform should emerge if the stock of social capital can be measured and incorporated into the growth equation. This would help to rationalize the political trade-offs in designing a reform programme and highlight the need to build constituencies that benefit from efficient investment in order to sustain the reform effort. Due consideration could also be given to alternative modes of compensating the losers from reform. However sound it may be in economic terms, reform is likely to elicit a strongly lagged response in the absence of a developmental state and the social capital that it deploys. Greater attention to initial conditions should help to explain reform outcomes and thereby contribute to improved policy prescription.

19.3.3. The impact of initial conditions Although the IFIs have been sensitive to the political trade-offs that politicians face in managing the social pressures arising out of reform, they have still underestimated the impact of variations in the initial conditions of reforming economies. These conditions include not only the natural resource endowment but also the level of public social capital and the legacy of obsolete produced capital. This study shows that natural resource abundance can retard reform and also prove particularly difficult for the smaller countries, especially the mid-income mineral economies. The foreign capital inflows that a reviving primary sector attracts, especially if it is capital-intensive sector, can cause the real exchange rate to rise if sterilization mechanisms are absent, and check economic diversification so that employment in the non-booming tradable sector, which is often more labour-intensive, is restrained. The largest countries tend to have a wider range of commodity diversification options while a larger fraction of their erstwhile-protected manufacturing sector can be re-oriented towards competitive markets (Londero and Teitel 1998). Small countries find economic diversification especially difficult, although microeconomics may achieve more flexibility by sustaining growth by supplying niche markets in dynamic large economies. They are slow to re-orient their manufacturing sectors yet they often lack scope to diversify their primary product exports. Post-reform employment creation is especially weak in small mid-income countries, in contrast to low-income countries like Indonesia, because wages tend to be high in relation to productivity. These problems are most acute for the smaller mid-income mineral economies and yield additional policy lessons. First, the establishment of EPZs at a relatively low per capita income level may provide useful catalysts for the competitive diversification of their economies. Second, small mineral economies have a greater need to minimize the Dutch disease effects during mining booms than most resource-abundant economies. Third, a sizable real depreciation of the exchange rate may be essential in order to

326

R. M. Auty

create a comparative advantage in the non-mining tradables and stimulate vigorous non-mineral-led growth. Since many governments lack the capacity to handle the challenges of managing resource rents, one way to avoid relying upon public sector spending would be to disperse the rents to private citizens, thereby decentralizing the decision-making process. Such a distribution system has the political appeal of allowing the citizens to make the decisions about the appropriate balance between consumption and investment. This study has provided both theoretical and empirical evidence of the advantages of diffuse socioeconomic linkages, where the rents emanate from booming crop prices and farmers retain a larger share of the windfall. Varangis et al. (1995) report sensible adjustments to the coffee boom of the late-1970s by Kenyan farmers who saved an estimated 60 per cent of the windfall. Elsewhere, Thai rice farmers adjust to higher prices in ways other than by the counterproductive strategy of over-planting. Unfortunately, in the case of minerals, the separation of the vast majority of rent recipients from the direct source of those rents may blunt the bias towards caution exhibited by farmers in receipt of crop rents. The importance of social capital has been neglected in implementing reforms. Policies should therefore be designed to build social capital wherever this is possible without depressing capital efficiency or establishing unsustainable entitlements such as the 'support-led security' of Costa Rica.

19.3.4. Policy lessons from transition and developed economies The superior social capital endowment of the transition economies of Northeast Europe compared with those in the FSU appears to explain much of their faster reform. If this is accepted, then the case for rapid reform in the lower-income transition economies may be weakened: their lacklustre performance arises from difficult initial conditions rather than from slow reform per se. Moreover, deficient social capital renders their higher dependence on natural resources more difficult to manage. Natural resource abundance can therefore retard the transition to a market economy for countries at a relatively low level of development (in which the primary sector assumes greater importance) and with low social capital and highly distorted economies. Although resource abundance does appear to have fed corruption in Russia, especially during privatization, the resource-poor and resource-rich countries of the southern FSU suggest the magnitude of corruption is inversely correlated with social capital. Corruption may be less overt under the more authoritarian government in Uzbekistan, but it remains pervasive. There is stronger evidence that an investment boom can trigger Dutch disease effects and stifle competitive diversification so that income inequality is heightened. In Kazakhstan (and also in oil-rich Azerbaijan), investment in the refurbishment of the mining sector triggered an early post -transition rebound of the real exchange rate that stunts private investment in the employment-intensive non-mining tradables. Yet the front-loading of capital recovery on such projects delays the expansion of tax revenues. The public sector is thereby fossilized in a vicious circle that depresses investment in

Conclusions

327

employment-intensive sectors, feeds corruption and prevents the transformation of the state into an agent for enabling economic development. There is firmer evidence still that natural resource rents retard reform. Rents from gold and cotton have been siphoned away by the government of Uzbekistan to feed the protected manufacturing sector. Yet, this strategy ignores the lessons of the marketdriven resource-abundant countries: in the absence of more market-sensitive reform, economic growth is likely to collapse. In Russia, natural resource rents have impeded reform because the government offset its declining revenues by using the rents to subsidize consumers. The rents also delayed the competitive diversification of the economy by permitting removal of hard budget constraints on enterprises. In addition, the robust response of the nonresource tradables sector to the negative oil shock of 1998, which may have cut GDP by 5 per cent and brought a sharp depreciation of the real exchange rate, is consistent with the Dutch disease thesis. Lastly, the privatization of natural resources amplified income inequality during the transition and corroded social capital. Fundamentally, the Russian reform programme failed to develop institutions for the effective capture of the oil rents and their deployment for the public good. Finally, the late-industrializing resource-abundant Nordic countries have managed to wean themselves from natural resource-dependency at different rates, albeit with different consequences for the long-term resilience of their economies. Denmark, the Nordic country least well-endowed with natural resources has experienced most success in diversifying its economy. Finland (like Sweden) has greatly reduced its former dependence on forestry and exhibits little evidence of the legacy of Dutch disease effects. Norway (like Iceland) still exhibits some Dutch disease effects. In addition, an echo effect persists from earlier resource dependence in the form of a fishing lobby whose political clout is disproportionately influential, given the shrunken size of the sector, and possibly detrimentally high. However, with regard to the oil sector, Norway has evolved a rational policy for handling the oil rents that clearly demonstrates the basic thesis of this book: even a large and volatile rent stream with point socioeconomic linkages can be well managed with appropriate social capital.

REFERENCES Brewster, H. (1994), 'Dutch disease in the age of adjustment', paper presented to the UNCTAD Conference on Development, Environment and Mining, Washington DC: World Bank. Davis, G. (1995), 'Learning to love the Dutch disease: evidence from the mineral economies', World Development, 23 (10), 1765-79. Edwards, C. B. (1990), Protection and Policy in the Malaysian Manufacturing Sector, Vienna: UNIDO. Lewis, W. A. (1978), Growth and Fluctuations 1870-1913, London: George Allen and Unwin. Londero, E. and S. Teitel (1998), Resources, Industrialization and Exports in Latin America, London: Macmillan.

328

R. M. Auty

Maddison, A. (1995), Monitoring the World Economy 1820-92, Paris: OECD. Mahon, J. E. (1992), Was Latin America too rich to prosper? Structural and political obstacles to export-led growth, Journal of Development Studies, 28, 241-63. Sachs, J. D. (1999), 'Resource endowments and the real exchange rate: a comparison of Latin America and East Asia', in Ito, T. and Krueger, A. O. (eds.), Changes in Exchange Rates in Rapidly Developing Countries: Theory, Practice and Policy Issues, Chicago IL: Chicago University Press. Varangis, P., T. Akiyama and D. Mitchell (1995), Managing Commodity Booms and Busts, World Bank, Washington DC. Williamson, J. (1996), 'Lowest common denominator or neoliberal manifesto? The polemics of the Washington consensus', in Auty, R. M. and Toye, J. (eds.), Challenging the Orthodoxies, Basingstoke: Macmillan: 13—22. World Bank (1999), World Development Indicators 1999, Washington DC: World Bank.

Index

accumulation 47 Acheampong 168, 176 actual rent 289 Adas, M. 105 Adelman, I. 25, 77, 84 adult mortality 280 Africa 78 aggregate GDP 3 aggregate production function 46 agricultural frontier 97, 99 agriculture 20, 31, 65, 69, 72, 97, 107, 114, 167, 174, 195, 198, 213, 218, 229, 232, 244, 265, 269, 286, 303 Akhad, A. 264 Akira, S. 105 Al-Ebraheem, Y. 199 Alesina, A. 255 Algeria 82, 126 Ali, R. M. 150 Altiplano 187 America 96 Anand, S. 25 Andean valleys 187 Andersen, L. E. 190 Anderson, K. H. 263 Andrain, C. F. 127 Angola 82 apartheid 242 Argentina 13, 100, 208-20, 246 closed trade policies in 212—16 growth process in 101, 108 Armed Forces Revolutionary Council (AFRC) 169 AronJ. 79 Aslund, A. 281, 291 asset depreciation 39 asset values 39 Australia 82, 100-1, 212 autarky 19, 29, 30, 208, 212-14, 216 autarky price 97

auto assembly 216, 218 autonomous state(s) 128-9, 193 Autv, R. M. 26, 29, 82, 115, 174, 280 Azerbaijan 267

backward linkage 8, 31 Baker, C. 105 balance of payments 118, 120-1, 241 Balassa,B. 28,211 Baldwin 'South' model 8 Baldwin, R. E. 8, 30-1, 82, 228 bananas 103, 229, 235 Barro, R. J. 301 barter system 291, 294 BAT 272 Becker, C. M. 266 beef rearing 235 'beneficios' 228 benevolent state 128, 166, 193 Berge, K. 3, 20 Bergsman, J. 133 BerlinskiJ. 214 Bezgreblnaya, I. 280 bimodal agricultural strategy 70 biodiversity 235 Birdsall, N. 12 Bisat, A. 204 black-market rate 247 Bloom, D. E. 23, 26 Bolivia 13, 101, 130, 179-91, 217 'bonding' 79 Booth, A. 161 Boserupian thesis 26 Bosworth, G. 234 Botswana 30, 127, 130, 147, 150 Bowman, K. S. 25 Brazil 62, 70-2, 103, 246 import-substitution strategy of 71 inward-looking growth strategy of 72

330

Index

Brewster, A. 324 bridging 79 Brown, R. 20 Brundtland Commission report 36 Brunei 128, 193 Bruton, H. J. 104, 130 Buffie, E. F. 215 Buhner-Thomas, V. 103, 228 Bumiputeras 147, 1S9, 162 bureaucratic quality 85 Burma 105

CACM 233 Canada 99 capital 12, 96-8, 101, 138 accumulation 7, 119, 250-1 depletion 42 efficiency 265 importation of 68 inflow 150 stock 248, 255 surplus economy 193 surplus oil exporter(s) 193-4, 199 capital development fund (CDF) 180 capital-labour ratio 97, 240 capitalists 182 Cardoso, C. F. S. 103 Caribbean workforce 9 Cariola, C. 102 Caspian region 266 Cato, L. 103 Caves, R. E. 100 CDF 190 Central American Common Market (CACM) 229 Central American Integration Treaty 231 Central and Eastern Europe (CEE) 277 Ceylon 103 CGE model 134 charcoal 176 Chenery, H. 3, 10, 20, 25 child health 61 children 65 Chile 32, 102, 127, 147, 220, 246 Chinese household responsibility system 263 Chinese in Malaysia 148, 159 Chinese model of agricultural reform 263, 271 Chisari 214 Christopher Columbus model 96-8, 106 civic social capital 11,78-9 civil liberties 85, 247 civil war 76, 100, 126 coal 100, 282, 290 cocoa 106, 165, 167, 170-1, 177 coconut products 103

CODESA 230 Coes, D. 103 coffee 103, 228-9, 235 boom 230, 236 coffee/cocoa economies 84 Collier, P. 11, 76, 78, 103, 166, 252 Colombia 103 Comibol 182, 184, 187 commerce 233 commodity exporters 102 commodity revenue stabilization fund (CRSF) 323 competitive diversification 153-6 competitive industrialization 158-60, 165, 213, 229, 317 competitive industrialization model 135-8, 319-22 composition effect 62 compression effect 62 consensual democracies 137, 142, 148, 166, 177, 179, 189, 228, 318 consumer price index 117 consumption 59, 60, 117 goods 117 tax 122 contractionary devaluation 121 contractionary effect(s) 120—1 convergence processes 246 Corden, M. 7, 300 Corea, G. 103 corruption 14, 78, 263, 265-7, 292, 326 Cort233s Conde, R. 101 Costa Rica 14, 103, 131, 225-37 cropland/capita ratio 225 economic crisis in 231 growth process in 103, 108, 231-5 per capita income 108, 226 political economy of 227-9 cotton 100, 270 country size 20, 208-10 Craig, J. E. 104 crisis 231 crop-driven growth 3, 30 crops 141 crude oil 40 crude petroleum 147 Cuccorese, H. J. 101 cultivator 99

Daewoo 272 Dasgupta, P. 36 data analysis 83-9 data set 83 Davis, G. 32

Index Dayaks 147 De Broeck, M. 265 de Melo, M. 14, 260 Deaton, A. 252 debt 201, 218-19, 241 deficient demand 134 degradation 37 Deininger, K. 25 demand 117, 134 linkage 8, 31, 179 democracy 81, 137, 165, 247 democratic state 129 demographic cycle 26 demographic transition 26 demographic trends 174-5 Deng 220 Denmark 296-310, 327 Denoon, D. 101 dependants 23 dependency ratio 23, 26 dependency theory models 31 depletion 37, 42, 47 devaluation 117, 121, 123, 168, 289, 300, 309 developing countries 3, 83 development economists 6 development expenditure 197 development strategy 67, 236, 240-3, 270 development trajectory 167 DFI 236 diamonds 102, 282 Diaz-Alejandro, C. F. 101, 212 diffuse linkages 30 diffuse resource economies 84, 119 'diffuse' source resources 5 direct foreign investment (DFI) 230 discrimination 170 diversification 138, 182, 197, 203, 230, 233, 236, 244, 320 dollar 40, 202, 215, 273 Domenech, R. 212 domestic demand 117,172 domestic prices 7 domestic production 47 Drake, P. 182 Dr232ze,J. 236 drilling 182 dual economy 274 model 98-9, 107 dual exchange rate 212 Duncan, R. C. 7 Dutch disease 7,28,70, 115 and the economy (Bolivia) 186-8 effects 32, 57, 67, 72, 101, 129, 175, 180, 190, 232, 253, 300, 302, 318, 326-7

331

extreme form of 119, 123 models 116 symptoms of 302-6 dynamic general equilibrium model 20

East Asia 113 Easterly, W. 78, 85, 130 EBRD 14 Echevarria, C. 20 Economic and Monetary Union (EMU) 306 economic development 77-82 economic growth 20, 23, 28, 61, 64, 141, 169-71, 240, 244, 247, 299, 301, 305-6 economic infrastructure 153 economic instability 64 economic performance 240—2 economic policy 27, 167—9, 247 Economic Recovery Programme (ERP) 166, 171 economic reform(s) 235-6, 277, 324-5 economic stability 64 economic variables 277 economy 3 Argentine 210 Asian 220 Australian 101 Bolivian 102, 189-90 Brazilian 72, 190 Finnish 309 Ghana 173 Latin American/East Asian 119 Malaysian 158 Mexican 210, 218 mineral/non-minera! 3, 32 Nordic 296, 299 Saudi Arabian 194, 205 Siamese 105 South African 14 sustainable 36-7 Taiwan/Korea 210 Eder, G. 182 education 59-67, 246, 252 in Argentina 214 in Ghana 175 and income inequality 62 in Malaysia 160 in Nordic countries 299, 304 and output 60-5 in Saudi Arabia 204 in South Africa 57, 59, 62, 250, 255 education ratio 88 Edwards, C. B. 321 Edwards, S. 26, 301 Effective Rate of Protection (ERP) 157 Eken, S. 199

332

Index

El Salvador 82, 228, 235 El Serafy, S. 40 electricity consumption 264 emigration 174 Employee Provident Fund 150 employment 159, 175, 199, 229, 233, 242, 255, 269 endogenous growth theory 10, 61 Engerman, S. L. 130 enrolment (rate) 244, 304 environmental accounting 12 environmental and natural resource accounting (KARA) 180, 191, 205, 323 equilibrium 117, 120 equipment investment 246 equitable growth 59-67 erosion 235 ethnic diversity 130 ethnic fractionalization 78 Europe 96 European Union (EU) 203, 300, 306 Evans, P. 80 excess demand 134 exchange rate(s) 68, 71, 118, 171, 198, 215-16, 247, 263 exogenous shocks 76 expanded accounting 39 expanded national accounting 37 expenditure 197, 204 export(s) 118, 158, 169-70, 212, 214, 232, 242, 244, 285, 289, 300, 302 instability 159, 162 ratio(s) 299 shares 299 structures 77, 83 export base theory 8, 82, 228 export processing zone (EPZ) 138, 233, 236, 321, 325 extended family 264 external shocks 184-5, 210-12, 218, 248 external trade 169-71 extraction costs 41

factional state 127, 129 farmers 30, 228, 235, 326 Ferrer, A. 101 fertility 61 Fields,G. 25 financial liberalization 324 Findlay, R. 117, 127 Finland 296-310, 327 firewood 176 firms 80 fiscal adjustments 200—1 fiscal deficits 9, 141, 158, 212, 214

fiscal linkage 4,31 fisheries 301, 303, 307-8 food 96 foreign direct investment (FDI) 157, 173, 232, 272 foreign exchange 77 foreign finance 188 foreign investment 250 foreign loans 216 foreign trade 301 forestry 286 forests' 176, 225, 235, 297, 308-10 forward linkage 8, 31 Frankel, S. H. 102 free trade 243 free trade zones 157 Freedom House 85 Frimpong-Ansah, J. H. 167—8 FSU countries 265 fuels 285 Furnivall, J. S. 105

Gaddy, C. 291 GallupJ. L. 179, 189 gas 209, 268, 290, 306 GATT 243 GDP 5, 25, 44, 60, 65, 71, 117, 131, 147, 150, 171, 204, 210, 213, 278, 296, 299, 303 Gelb, A. H. 32, 193, 266 Gelb, S. 241 genuine saving(s) 38-19, 137, 152, 282 rate(s) 12, 38, 42-1, 47, 160, 175, 274, 251 Georgia 267 Ghana 13, 165-77 political economy analysis of 167-9 political state in 165-6 growth process in 108 Ghana Living Standards Survey (GLSS) 171 Gindling, T. H. 230 Gini coefficient(s) 26, 62, 85, 239, 280 GNP 250, 278, 296 gold 100, 102, 172, 239, 242, 270, 282 Gold Coast 106 gold rush 101 golden age 12, 95, 109, 208, 316, 321 Gonzalez-Vega C. 230 goods 96 government(s) 65-7, 135, 253, 324 Argentina 212, 214, 218 Costa Rican 227-8 Finnish 309 Ghana 175 Malaysia 158, 321 Mexican 211, 216-18

Index Norwegian 306-7 of mineral economies 32 of predatory and factional oligarchic states 142 of resource-poor countries 142 role of 66 Saudi 196-205 Taiwan 211 Uzbekistan 263, 270, 272 Graham, E. 228 grazing 235 Great Depression 108 Greece 246 'green' national accounting 37 gross investment 150, 252 gross national product (GNP) 38 growth 6, 8, 61, 66, 135-41, 173, 239, 241, 244-8, 255 growth accounting framework 10-11 growth collapse(s) 5, 9, 84, 315-19 in Argentina 208-10, 218 in Bolivia 179-91 in Costa Rica 227-31 in Ghana 165-77 in Mexico 208-10, 216-19 in resource-abundant countries 315-17 in Saudi Arabia 193-206 growth models 11,246 growth rate data 83 'growth spells' 25 growth theory 45 growth trajectory(ies) 5, 28, 76, 171, 212-16 growth-based security 236—7 Guevara, W. 182 Gulf War 200 Gunnarsson, C. 104 Gunning,]. W. 11, 166,252 Gylfason, T. 3, 26

Haggard, S. 127 Hall, R. 78 Hannesson, R. 307 Hansen, B. 99 Harbison, R. W. 103 hard coal 41 hard minerals 82 Harris-Todaro migration model 134 Hartwick rule 37 Hartwick, J. M. 37, 48 Havrylyshyn, O. 203 HCI Big Push 156-7,321 HCI 214, 216, 218, 268, 272 headmen 264 Heal, G. 36 health 246

heavy and chemical industry (HCI) 136, 153,212 heavy industry 20, 189 Herbertsson, T. T. 301 Herrick, B. 25 Hicks, J. R. 39 Hicksian income 48 Hill, C. B. 32 Hill, P. 106 Hirschman, A. O. 114, 228 Hlaing, U. A. 105 Hobart Houghton, D. 102 Holden, M. 242 Holmes, A. B. 106 Holton, R. H. 100 Hotelling rents 151 Hotelling rule 40 human capital 11, 27, 46, 60-5, 71, 82, 118, 123, 150, 161, 174-6, 181-4, 225, 246, 296 accumulation 20, 115, 123, 175, 214,251 investment 39, 57 theory 61 human rights 247 hydrocarbons 188, 219 hyperinflation 188, 191 hysteresis 114, 134, 230

Ibans 147 Iberian settlers 227 Iceland 296-310, 327 Ickes, B. 291 ICOR 218 identity 47 IFls 324-5 immigrants 101, 103-4 immigration 97, 101, 107 import substitution 242 import-substituting industrialization (!SI) 242 imports 118, 169,242,301 income 39, 255 income gini 85 income growth 26 income inequality 25, 61, 138, 141, 239, 266 and education 62 and growth 62 and natural resource abundance 25 and per capita income 25 income share 85 incremental capital output ratio (ICOR) 153,214 index 3, 12 consumer price 231

333

334 index (cont.): Corruption Perceptions 292 for domestic manufacturing 3 of inequality 280 of land endowment and agricultural potential 3 liberalization 263 of Saudi intra-industry trade 203 of sustainability 12 trade 196 of trade policy closure 212 of United Nations Human Development 225 India 220 Indians in Malaysia 148 indicators): of economic policy 247 of human capital 246 of inequality 85, 280 of social/political institutions 84-5, 88 of socioeconomic flow 278—81 of sustainability 38—42 of transition 277, 281-2, 287 Indonesia 32, 72, 127, 158, 193-4, 211, 266 industrial policy 242-3 industrialization 26, 67, 69, 103, 133, 137, 140, 142, 156-60, 167, 214, 229, 236, 320 industry 232-3, 265 inequality 67, 172, 293 inflation 172, 188, 239, 278, 289, 293, 300 Ingram, J. C. 105 Innis, H. A. 100 institutional capital 10 institutional credibility 78 institutional reforms 289-92 institutions 80, 90 interest rate(s) 97, 119,218,263 internal colonialism 181 international financial institutions (IFIs) 14, 322 intra/inter-community ties 79 investment 23-7, 59, 69, 119, 150-2, 173, 198, 218, 246, 299, 303 efficiency 130, 153, 214-15, 218 rate 150, 214-15 investors 203 iron ore 290 Ishak, S. 160 ISI 250, 252, 255 Ismail, M. S. 153, 161

Jamaica 32 Jamison, Dean T. 61 Jefferson, M. 101 Jenkins, M. 233 job creation 134, 235, 241

Index Johnston, Bruce F. 70 Jones, C. 78 Julius, D. 20

Kaletsky, A. 10 Kanbur, S. M. R. 25 Karl, T. 76 Kazakhstan 14, 260-75 recovery of 265—70 reform in 2 6 - 4 Keefer, P. 11,78,292 Kenya 66 Kharas, H. 150 Kilby, Peter 70 Killick, T. 165, 172-3 Kindleberger, C. P. 25 Kirsch, H. W. 102 Klein, H. S. 102 Knack, S. 11,78,292 Knight, John B. 61 knowledge 115 Korea 62, 70-2, 246 Kornelia, K. 268 Kostial, K. 265 Krause, A. S. 215 krone 299 Krucger, A. O. 133, 167 Krugman, P. 114, 121 Krugman, P. R. 7 Kunzel, P. 203 Kurer, O. 130 Kusi, N. K. 167 Kuwait 197 Kuznets curve 25 Kuznets, S. 25

La Porta, R. 78 labour 22, 27, 59, 61, 67, 69, 96-8, 138, 174, 320 movements 71 productivity 265 sink 69 Lal, D. 3, 9, 27, 126-7, 213 land 59, 69, 96-8, 248, 282, 297 colonization 32 frontier 97, 106, 235 gini 85 inequality 85 land-labour ratio 97, 99 Latin America 25-6, 70, 113 Lau, Lawrence J. 61 Leftwich, A. 127 leisure 59, 60

Index Levine, R. 78, 130 Lewis, W. A. 98, 108, 114 liberalization 137, 156, 243, 254, 256 life expectancy 244, 280 light industry 20, 187 lignite 41 Lim, D. 156, 159 linguistic fractionalization 78 linkage analysis 8 linkages 30," 32, 65, 100 liquefied natural gas (LNG) 147 litigation 292 living resources 38 loans 156, 201 Londono, J. L. 27 long term official capital (LTOC) 230 Looney, R. E. 203 Lougheed, A. L. 100

macro policy 270-3 Maddock, R. 101 mainstream economists 6, 27 Malaya 104, 147, 156 Malaysia 13, 65, 129, 147-63, 217, 246 dependency ratios of 26 ethnic tensions in 148, 161—2, 321 gini coefficients of 27 growth process in 147, 153-6 HCI Big Push 157, 162 human resource development in 160-1 political state in 137, 148 recession in 158 Maluccio, John 61 Mamalakis, M. J. 102 Mankiw, G. 10 manufactured exports 69, 71 manufactures 96, 114 manufacturing 7, 20, 29, 84, 97, 107, 114, 129, 195, 213, 253, 272 manufacturing sector 12, 26, 100, 199, 203, 219, 289 manufacturing-led growth 3 marketing boards 243 markka 309 Matsuyama, K. 7, 20 Mauritius 130-1 Mauro, P. 78 McLean, [. VV. 100 McMillan, M. 133 Meier, G. M. 100 Mellor, J. W. 3, 22 merchandise exports 300 Mercorsur 190, 203 metals 282

335

Mexico 13, 194, 208-20, 246 closed trade policies in 212 Mexico-Argentina, closed trade policies in 212-16 Meyanathan, S. D. 161 migration 31, 105-6, 173, 189 mineral accounting 39 mineral exporter(s) 4, 67, 101-2, 239 mineral-driven growth 32 minerals 40-1, 141, 151, 209, 244, 248, 274, 286 mines 182 mining (sector) 32, 179, 181, 270 mining 100-1, 141, 172, 218, 244, 253, 269 MNCs 159, 197 model of staples production 97 modernization 173 monetary sector 118 Morley, S. A. 233 Morris, C. T. 25, 77 Morris, T. 84 Muda River 65 multiplier effects 65 multivariate regression analysis 303, 305 Mundlak, Y. 212 Myint, H. 3,9,27 Myint, H.M. 114

Nankani, G. T. 193 national accounts method 204 National Front Coalition 148 national income 48, 117 National Liberation Council (NLC) 168 National Redemption Council (NRC) 168 National Revolution 181 nationalization 182, 196 Nattrass,]. 102 natural capital 10, 36, 174-6, 225, 234, 248, 251, 296 natural gas 40, 100, 282, 297 natural resource endowment(s) 42, 81-2, 126, 128, 131, 141, 147, 180, 227, 260, 282 natural resource rents 252 natural resources 46, 95, 108, 128-31, 150 in Argentina and Mexico 208 and autonomous states 128 in Costa Rica 234 and departure from welfare maximization 129 and economic development 95, 317 in Ghana 165 in Norway/Finland 306-10 in Nordic countries 297, 327 and political economy/state 126, 318-19

336

Index

natural resources (cont.): in Russia 282, 326 and the Russian reform programme 286-92 natural-resource endowments 296 Neary, P. J. 7, 114, 116 neo-classical theory 10 NEP 162 net income 46-9 net investment 150-1 Net Material Product 278 net national product (NNP) 39, 48 net national saving (NNS) 38 New Economic Policy (NEP) 159, 160, 189, 321 New International Economic Order 317 Newly Independent States (NIS) 277 Nigeria 67, 82, 266 nitrates 102 Nkrumah, K. 167-8, 173, 176 Nokia 308 Nominal Rate of Protection (NRP) 157 non-equipment investment 246 non-tradables 7, 101 non-traded sector 115-16, 119, 123 Nordic countries 14, 327 North America 99 North, D. C. 100 northern core 95 Norway 14, 296-310, 327 Norwegian EU membership 307 Norwegian Petroleum Fund 307

obsolete produced capital 260, 273 oil 82, 194, 209, 216-19, 267, 282, 290, 297, 301, 306-8 boom(s) 72, 152-3, 193, 196-200 crisis 37 exploration 182, 218 exporters 4 rents 200-5, 218, 266 shock(s) 5, 10, 137, 162, 165, 184, 195, 216, oil-exporters 10, 140 oil-importers 10 oligarchic regimes 82 oligarchic state 129, 212 Olson effects 78, 227, 236 Oman 147 open economy 197 organizations 254 output 116-17 over-industrialization 260 overall price index 117 overgrazing 235 overspending 134—5

overvaluation 171, 173, 247, 288 owners 82 ozone layer 36

palm oil 147, 153 Pampas 101, 208, 213, 219 'pan-African' ideology 167 Pastor, M. 181 pasture 235 paternalistic governments 205 Pearce, D. 36 peasant crops 103-6 'peasant hinterland' 98 peasants 99, 130, 187, 228 Penn World Tables 83 per capita cropland 3 per capita growth rate 244 per capita income(s) 3, 19-25, 226, 244 Pern era 108 performance 281 performance divergence 6, 9 exogenous explanations of 6 internal explanations of 9—10 Perkins, D. H. 208 Perloff, H. S. 100 Peron 213, 215 Peru 3, 32, 102 petro-dollars 10, 219, 230, 236 petrochemicals 203 Pezzey, J. 36 Phongpaichit, P. 105 physical capital 250, 255 physical investment 246, 254 Planccon 266 plantation crops 82, 103-4 point linkages 30 point rents 5 point resources 5, 141, 267 point source (natural) resources 77, 265-70 point source economies 84 polarized democracy 189, 220 policy distortion 260 policy failure 9 policy implications 14, 72, 90, 122-4, 322-7 policy issues 45 policy lessons 14 political economy 167-9, 228 political institutions 254 political liberties 85, 247 political polarization 240, 254 political scientists 126 political state(s) 126, 225 political unrest 240 pollutants 41, 46

Index pollution 39, 41 Pomfret, R. 263 Pope, D. H. 101 Popov, V. 14 population 159, 228 population growth 26 potential rent 289 poverty 291 poverty indicators 280 Prebisch, R. 6, 114 predatory state(s) 128-9, 167 present value approach 40 price(s) 41,97, 117, 169,214 price liberalization 263, 323 price stabilization 323 primary exports 28, 300, 303 primary sector 271 principal components analysis 260 Pritchett's mountains 79 private investment 204, 252 private sector 230, 252, 255, 267, 269, 289 privatization 243, 263, 268, 289, 291 produced capital 10, 150, 156, 248 production 46, 304 function(s) 30, 82, 117, 179 structure 172, 232 productive linkage 179 profits 98 Progress Party (PP) 168 prosperity 81-2 protection 133, 157, 187, 216, 218, 242, 247 Provisional National Defence Council (PNDC) 169 public accountability 81 public expenditure 197, 200, 323 public institutions 81 public investment 153-6, 183, 203, 230, 323 public sector 230, 232, 269 public social capital 11, 78, 79 public spending 217, 250 public support 236 pulp/paper 100 Putnam, R. 78

quality education 66, 72 quantitative restrictions (QRs) 243 quotas 71

racial discrimination 239 racial riots 162 racial segregation 254 racketeering 292 Ragayah, M. Z. 160

337

Ramsey growth model 116 ranching 235 rand 242 Ranis, G. 3 rapid reform model 260, 262, 273 rate of investment 23 Rawlings, J. J. 168 RBI 205 real estate 203 real exchange rate(s) 202-5, 217, 267-8, 288, 293, 300, 323 real wages 107, 141, 303 recession 234 redistribution 240 reform speed 263-4 reform(s) 141, 176, 187, 219, 227, 231-5, 240, 252, 254, 260, 282 regions of recent settlement 95, 99-102 rent deployment 196-200 rent reallocation 133—4 rentier autocrats 82 rentier capitalists 82 rentier economy 193 rents 132, 140, 176, 185, 230, 252, 289, 292, 327 Repetto, R. 40 resource abundance 57, 67-72, 158-60, 208-10, 262, 270-3, 315-19 resource boom(s) 114, 122, 185, 316 resource consumption allowances 151 resource depletion 251 resource prices 48 resource rent deployment 131-5 resource rents 40, 57, 67, 126, 131, 150, 156, 172, 289 resource sector 7, 219 resource-abundant countries 25, 67, 72, 147, 208, 239, 289 resource-based industry (RBI) 197 resource-poor countries 25 Reynolds, C. W. 215 Reynolds, L. G. 104 rice 105 Rimmer, D. 172 rival 202 Rodriguez 117 Rodriguez, F. 116 Rodrik, D. 78, 141, 255, 292 Romer, P. iM. 10 Rosenberg, C. B. 14 Rostow, W. W. 100 Rottenberg, S. 228 rouble 288, 293 rubber 103, 147, 153 rule of law 85 Russia 14, 277-94

338 Saavalainen, J. O. 14 Sabah 156 Saborio, S. 231 Sabot, Richard H. 61 Sachs,J. D. 3, 7, 10, 20, 28, 113, 115-16, 119, 193, 301 Sala-i-Martin, X. 244, 301 Santa Claus method 69, 199 Sarawak 156 Sarris, A. H. 171 Saudi Arabia 13, 28, 30, 193-206, 217, 246 open trade policies in 28, 193 paternalistic monarchy of 128, 194, 205 Saudi deployment strategy 196 saving(s) 23-5, 46, 59, 150-2, 173, 198, 251, 255 sawlogs/sawn timber 147 Scheepers, C. 242 schooling 66 self-sufficiency 19, 239, 272 sec also autarkySen, A. 236 service(s) (sector) 20, 173, 199, 232, 269 Shafer, D. M. 228 Shank, N. 204 Sheik Ahmed Yamani 76 Shell, K. 115 shocks 82, 161-3, 225, 240 Siam 105 Siamwalla, A. 32 silver 101 Simon, J. 267 Singer, H. 114 sink employment 173 Sirageldin, I. 199 skilled labour 27, 67 skills 61, 67, 160, 240, 255 SMC I 168 SMC II 168 smuggling 179 social capital 11, 81, 130, 135, 141, 166, 181-4, 225, 227, 236, 260, 292, 296, 326 index 235 social cohesion 14, 229, 264, 289-92 social conflict 291, 293 social costs 268—70 social development 84 social indicators 277 social infrastructure 153 social mines 182 social overhead capital (SOC) 153 social structure 81 social theory 77—81 social variables 77-81

Index societies 80 socioeconomic conditions 246 socioeconomic flow indicators 278-81 Sokoloff, K. L. 130 Solow growth model 78 Solow, R. M. 37 South Africa 14, 239-59 growth process in 102, 109, 244-8 reform in 239, 254-6 South African Reserve Bank 246 South Korea 26, 127, 211 abandoning policies in 129 competitive industrialization model 136 dependency ratios of 26, 136 gini coefficients of 26 HCI Big Push 137 per capita cropland of 26 Spanish colonization 227 spending 135, 252 Squire, L. 25 Sri Lanka 129 Stable Development 215 'stagflation' 135 stagnation 324 standard of living 236, 293, 296 staple trap 8, 30-2, 115, 124, 147, 214, 315 model 138-41, 165, 216, 229, 271, 273, 321-2 effects 57 staple(s) 8, 99, 101, 103 staples model 99, 106 state-owned enterprises (SOEs) 172, 175, 197, 229, 231, 272 states 80 steel 216, 218 StiglitzJ. 292 Stovel.J. A. 100 strikes 240 structural change 19-23, 25-7, 227, 268 structuralists 28 structuralist approach 7 structuralist economists 6 Sturzenegger, A. C. 213 sub-Saharan Africa 130 subsistence cultivator 99 subsoil resources 38 Sunkel, O. 102 support-led security 236 Supreme Electoral Tribune 229 sustainability 36, 46, 137, 175 indicator(s) 38-42 sustainable development 12, 36, 234, 251 Sweden 296-310, 327 Syrquin and Chenery norm(s) 195, 212, 232 Syrquin, M. 3, 20, 25, 208 Szereszewski, R. 106

Index Taiwan 127, 136,211 tariffs) 71, 156, 158, 212, 229, 243, 263 Taube, G. 264 tax(es) 121, 156, 182, 200, 307 taxation 32, 134, 141, 171, 185, 196, 198 Taylor, A. M. 101 Taylor, L. 121 tea 103 technical sophistication 203 Temple, T. 78 TendlerJ. 80 Terrell, K. 230 Thailand 31, 127, 147 Thomas, W. 241 three-factor model 7, 12, 96-8, 116, 316 three-sector model 116-19 timber 151 Timmer, C. P. 21 tin 102, 104, 147, 153 Tomich, T. P. 31 total domestic supply 47 total exports 303 total factor productivity (TFP) 204, 233 total rent approach 40 tradable sector(s) 213, 262, 265 tradables 7, 286 trade 29, 97, 169, 196, 242-3, 285, 289, 299, 301-3 balance 118, 120 deficit(s) 186, 214-5 liberalization 239, 243, 323 openness 166, 193 reform 269, 231 trade policy 28, 30, 247 closure 140, 212-16 export-oriented industrialization 156—8 and natural resource dependence 28 and primary product exports 30 traded sector115-16, 119, 123 trajectory development 9 transition anomaly 270-3 transition economies 14, 260-75, 277 transition indicators 277, 281-2, 292 transition recession 265-7 transition trajectories 262-4 Transparency International 292 transport 179, 189-90 Trinidad and Tobago 32, 246 tropical countries 95, 102-6 trust 78, 293 Tuncer, B. 133 Turkmenistan 268

Ulate, A. 231 Umikalsum, H. M. N. 156

339

UNCTAD 83 underdevelopment 124 unemployment 134, 239, 280 UNICEF 281 unimodal agricultural strategy 70, 72 unit rents 40 United Ghana Farmers' Council (UGFC) 170 United States (US) 82, 99, 212 urbanization 25-6, 138, 320 Uruguay 246 Urzhumova, D. S. 266 user cost 39 Utilitarian maximand 36 Uzbekistan 14, 260-75 reform in 262^1 transition anomaly theories 270-3

'vampire' state 167-8, 176 van Wijnbergen, S. 7, 114, 116, 300 Varangis, P. 326 variables 244 Veganzones, M.-A. 213 Venezuela 116, 126, 211, 266 Vincent, J. 48 VincentJ. R. 150

wage rate 97 War of the Pacific 102 Warner, A. M. 3, 10, 20, 28, 115, 193, 301 Watkins, M. H. 100 wealth accounting 40 wealth 38-9, 82, 148, 159, 248-54, 255, 282, 291, 296 Weissman, R. S. 171 welfare 25-7, 36, 38, 46, 128-9 Wetzel, D. 176 wheat 100, 198 Williamson's dependency cycle 131, 181 Williamson, J. 323 Williamson, J. G. 23, 25-6 windfall absorption 196-200 winners curse 113, 115 Winograd, C. 213 Withers, G. 101 wood 41 Wo3,20,,198, 246 98, 246 wool 100 worker dependency ratio 23-5, 174, 206 workers 23, 71, 199, 267 World Bank 10, 27, 184, 218, 244, 248, 266, 280, 296

340 World Development Indicators 83 world prices 40-1 World Values Survey 78 World War I 100 too203,243 243

Index Young, A. 10, 234 YPFB 183-4, 188 Zak, P. 78 Zettelmeyer,J. 264, 268,271