Labor and the Globalization of Production Causes and Consequences of Industrial Upgrading
Edited by
William Milberg
...
50 downloads
1068 Views
1MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Labor and the Globalization of Production Causes and Consequences of Industrial Upgrading
Edited by
William Milberg
Labor and the Globalization of Production
This page intentionally left blank
Labor and the Globalization of Production Causes and Consequences of Industrial Upgrading Edited by
William Milberg Bernard Schwartz Center for Economic Policy Analysis New School University, New York
Editorial matter and selection and Chapter 1 © William Milberg 2004 Remaining chapters © Palgrave Macmillan Ltd 2004 All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No paragraph of this publication may be reproduced, copied or transmitted save with written permission or in accordance with the provisions of the Copyright, Designs and Patents Act 1988, or under the terms of any licence permitting limited copying issued by the Copyright Licensing Agency, 90 Tottenham Court Road, London W1T 4LP. Any person who does any unauthorized act in relation to this publication may be liable to criminal prosecution and civil claims for damages. The authors have asserted their rights to be identified as the authors of this work in accordance with the Copyright, Designs and Patents Act 1988. First published in 2004 by PALGRAVE MACMILLAN Houndmills, Basingstoke, Hampshire RG21 6XS and 175 Fifth Avenue, New York, N. Y. 10010 Companies and representatives throughout the world PALGRAVE MACMILLAN is the global academic imprint of the Palgrave Macmillan division of St. Martin’s Press, LLC and of Palgrave Macmillan Ltd. Macmillan® is a registered trademark in the United States, United Kingdom and other countries. Palgrave is a registered trademark in the European Union and other countries. ISBN 1–4039–3502–5 This book is printed on paper suitable for recycling and made from fully managed and sustained forest sources. A catalogue record for this book is available from the British Library. Library of Congress Cataloging-in-Publication Data Labor and the globalization of production: causes and consequences of industrial upgrading / edited by William Milberg p. cm. Includes bibliographical references and index. ISBN 1–4039–3502–5 1. Production (Economic theory) 2. Industries – Technological innovations. 3. Globalization – Economic aspects. 4. Labor market. 5. International trade. I. Milberg, William S., 1957– HB241.L32 2004 331.12—dc22 10 9 8 7 6 5 4 3 2 1 13 12 11 10 09 08 07 06 05 04 Printed and bound in Great Britain by Antony Rowe Ltd, Chippenham and Eastbourne.
2004042844
Contents List of Tables
vii
List of Figures
ix
List of Contributors
x
Acknowledgement
xi
1 Globalized Production: Structural Challenges for Developing Country Workers William Milberg
Part I
Reconceptualizing Globalized Production
2
Trade and Industrial Upgrading in Developing Countries Yilmaz Akyüz, Richard Kozul-Wright, and Jörg Mayer
3
Upgrading, Uneven Development, and Jobs in the North American Apparel Industry Jennifer Bair and Gary Gereffi
4
5
Technology versus Trade versus Social Institutions: Explaining Rising Wage Inequality in the Chilean Cosmetics Industry Janine Berg Why Do Firms Disintegrate? Towards an Understanding of the Firm-Level Decision to Subcontract and its Implications for Labor Asad Sayeed and Radhika Balakrishnan
Part II Skill, Gender, and Location Bias in Globalized Production 6
7
Skill Upgrading in Developing Countries: Has Inward Foreign Direct Investment Played a Role? Matthew J. Slaughter Does Trade Promote Gender Wage Equity? Evidence from East Asia Günseli Berik, Yana van der Meulen Rodgers, and Joseph E. Zveglich, Jr v
1
19 21
58
88
104
119 121
146
vi
8
Contents
Globalization in a Middle-income Economy: FDI, Production, and the Labor Market in South Africa Stephen Gelb and Anthony Black
Part III Challenges to the Regulation of International Production 9
Bargaining Power and Foreign Direct Investment in China: Can 1.3 Billion Consumers Tame the Multinationals? Elissa Braunstein and Gerald Epstein
179
207 209
10 Rethinking International Labor Standards Michael Piore
249
Index
266
List of Tables 2.1
Manufacturing employment as a share of total employment (%) by region, 1960–2000 2.2 Manufacturing output as a share of GDP (%), by region, 1960–2000 2.3 Selected trade and production indicators (%) for 26 developing economies, 1960–2000 2.4 Labor productivity in 26 developing economies and selected industrial sectors, 1980–2000 2.5 Sectoral shares (%) in manufacturing value added in selected developing economies, 1970–2000 2.6 Unit labor costs in 26 developing economies and selected sectors, 1980 and 2000 2.7 Indicators related to the international competitiveness of exporters of manufactures in 26 developing economies 2.8 Commodity structure of exports (% of total non-oil exports) from selected developing countries, 1980–2000 2.9 Indices of revealed comparative advantage for manufactured exports of selected economies, 1980–2000 2.10 Structural similarity indices for exports of manufactures and manufacturing value added, selected developing economies, 1980–81 and 1997–98 3.1 US apparel imports: total and 807/9802 trade, by Mexico and Caribbean Basin initiative (CBI) countries, 1994–2000 4.1 Cosmetics industry jobs: gender, skills, pay and labor demand 4.2 Existence and status of unions in participating firms 6.1 The skill mix of employment in foreign affiliates of US-headquartered multinationals 6.2 The share of US sales in ICT industries accounted for by the US parents of US-headquartered multinationals 6.3 The share of worldwide activity of US-headquartered multinationals accounted for by foreign affiliates 6.4 The relative importance for ICI services industries of US exports versus sales by foreign affiliates of US-headquartered multinationals 6.5 The relative importance of FDI in developing-country net capital inflows 6.6 List of industries and countries in estimation sample
vii
25 25 27 34 38 41 43 47 49
52 69 96 99 126 131 132
133 135 137
viii List of Tables
6.7
Estimation results for the effect of foreign-affiliate presence on skill upgrading 7.1 Employment, residual wage, and trade ratios by industry concentration, Taiwan: 1981–99 7.2 Employment, residual wage, and trade ratios by industry concentration, Korea: 1980–98 7.3 Pareto function estimates of domestic competitiveness for Taiwan and Korea 7.4 Determinants of change in the gender wage gap, Taiwan 7.5 Determinants of change in the gender wage gap, Korea 7A.1 Consistent manufacturing aggregates by Standard Industrial Classification Codes, Taiwan and Korea 7A.2 Key variable definitions and data sources 8.1 South Africa, working age population, 2003 8.2 Employment status within educational achievement categories 8.3 Employment indices by skill category and sector, 1985, 1995, and 2002 8.4 Sectoral output shares, 1995 prices 8.5 Foreign affiliates in South Africa, sectoral distribution and size in 2000 8.6 Geographic distribution of parent firms by sector 8.7 Market orientation by sector, South Africa 8.8 Domestic market share by mode of entry, South Africa 8.9 Availability of suitable skilled employees, sector means in 2000 8.10 Training expenditure 8.11 Black economic empowerment – all sectors 9.1 Accumulated FDI stock in China by source countries, 1983–99 9.2 FDI as a percent of gross investment by area 9.3 Measures of liberalization and openness 9.4 Wage equations for 29 provinces, using data for 1986–99 9.5 Employment equations for 29 provinces, using data for 1986–99 9.6 Investment equations for 29 provinces, using data for 1986–99 9.7 Government revenue equations for 29 provinces, using data for 1986–99
139 159 160 164 166 167 172 173 182 183 184 185 188 189 191 193 198 200 202 216 217 222 224 228 231 234
List of Figures 1.1 1.2 1.3
1.4 2.1
2.2
3.1 4.1 7.1 7.2 7.3 7.4 9.1
World GDP, FDI, and trade, 1970–2001 Broad and narrow measures of international outsourcing in US manufacturing, 1958–2001 Imported inputs as a share of total intermediate inputs, Canada, Japan, United Kingdom, and United States, selected years and sectors Intra-firm trade as a percentage of total trade: United States, 1977–98 Change in manufacturing value added and exports of manufactures in relation to changes in gross fixed capital formation: 1990–2000 compared to 1980–90 Changes in manufacturing value added in relation to changes in exports of manufactures: 1990–2000 compared to 1980–90 North American apparel employment, 1985–2000 Returns to schooling for university, high-school, and primary-school educated workers, Greater Santiago, 1958–98 Manufacturing-sector female/male wage ratios: cross-country evidence Gender wage ratios by manufacturing industry status, Taiwan and Korea Gender wage ratios and manufacturing trade ratios, Taiwan and Korea Gender wage ratios by industry concentration, Taiwan and Korea FDI flows into China, 1986–99
ix
4 4
5 7
30
31 66 90 153 154 156 157 215
List of Contributors Yilmaz Akyüz was formerly Director of UNCTAD’s Division on Globalization and Development Strategies and currently holds the Tun Ismail Ali Chair in Monetary and Financial Economics, University of Malaya. Jennifer Bair is Assistant Professor of Sociology at Yale University. Radhika Balakrishnan is Associate Professor of Eonomics at Marymount Manhattan College. Janine Berg is Labour Economist at the International Labor Office in Geneva. Günseli Berik is Associate Professor of Economics at the University of Utah. Anthony Black is Associate Professor & Head, School of Economics, University of Cape Town. Elissa Braunstein is Assistant Research Professor at the Political Economy Research Institute of the University of Massachusetts-Amherst. Gerald Epstein is Professor of Economics at University of MassachusettsAmherst and Co-Director of the Political Economy Research Institute. Stephen Gelb is Executive Director of The EDGE Institute, and Visiting Professor, Development Studies, University of the Witwatersrand, both in Johannesburg. Gary Gereffi is Professor of Sociology at Duke University. Richard Kozul-Wright is Senior Economic Affairs Officer at UNCTAD’s Division on Globalization and Development Strategies in Geneva. Jörg Mayer is Economic Affairs Officer at UNCTAD’s Division on Globalization and Development Strategies in Geneva. William Milberg is Associate Professor of Economics at New School University, and Program Coordinator at the New School’s Bernard Schwartz Center for Economic Policy Analysis. Michael Piore is Professor of Economics at MIT. Yana van der Meulen Rodgers is Associate Professor of Women’s and Gender Studies at Rutgers University. Asad Sayeed is Director, Collective for Social Science Research, Karachi, Pakistan. Matthew J. Slaughter is Associate Professor of Business Administration at the Tuck School of Business at Dartmouth. Joseph E. Zveglich, Jr is Economist at the Asian Development Bank, Manila, Philippines. x
Acknowledgement This project was generously supported by the Bernard Schwartz Center for Economic Policy Analysis at New School University.
xi
This page intentionally left blank
1 Globalized Production: Structural Challenges for Developing Country Workers William Milberg1
1
From industrialization to industrial upgrading
According to the standard theory of international trade, trade liberalization will result in the equalization of wages globally (the factor price equalization effect) and a reduction in wage inequality in countries that are abundant in low-skill labor (the Stolper–Samuelson effect). Decades of trade liberalization, and ten years after the formation of a World Trade Organization with liberalization of trade as its main goal, it would appear that neither of these predictions has been borne out. The theory emphasizes the gains from trade resulting from the efficiencyincreasing shifts in the structure of production brought on by trade liberalization. If instead of emphasizing the substitutions of capital and labor as factors of production, we focus on the asymmetry with which these factors enter the global economy – capital significantly mobile internationally and labor only marginally so – then the challenges globalization poses for workers become obvious: the greater international mobility of capital relative to labor puts workers from a given location at an immediate disadvantage, both in terms of bargaining power with owners of capital (whose threats to move gain greater credibility) and with respect to the state (as governments are more able to tax immobile than mobile productive factors). But this perspective – which arguably has led the anti-globalization movement to propose the adoption of controls on international capital movements, international codes of corporate conduct, international labor standards, and reduced barriers to international migration – also does not capture the full picture of the structural challenges facing workers in developing countries in the early twenty-first century. Historically, economic development has hinged crucially on industrialization, as the transition from agriculture to industry has involved capital 1
2
William Milberg
investment, technological progress and growth in labor productivity that have raised living standards. But global changes over the past 25 years – in communications, transportation, technology and, most importantly, corporate strategies and government policies – have greatly changed the terrain on which industrialization occurs. Even as import substitution industrialization strategies gave way to export promotion strategies for many countries in the 1980s, it was not possible at that time to envision just how globalization would alter the conditions necessary for successful industrialization. Export promotion appeared to be consistent with welfare-optimizing economic theory: free trade and comparative advantage would bring benefits from specialization and exchange that were not available under the protected cocoon of import substitution. The key to industrialization, then, would be to gradually shift production (and thus exports) into more capitalintensive sectors which would bring higher productivity and wages. And export promotion seemed to work, although perhaps in only a few countries and not for the reasons given by many economists. The East Asian “miracle,” that is the relatively rapid industrialization of South Korea, Taiwan, and Hong Kong, eventually brought significant wage gains, a more educated work force, higher labor standards, and political democratization. While there is debate over the conditions that encouraged this wave of industrialization, there is considerable evidence that it occurred with selective government protection from imports and inward foreign direct investment, subsidies for export promotion and, very significantly, regular checks and controls on businesses who often had to meet performance and investment standards to receive continued subsidy or protection.2 Openness played a role, but may have been as much the result rather than the cause of economic development. Can the East Asian miracle be replicated in other countries? There is certainly some evidence that it can – see especially the successes of a second tier of Asian countries, as well as some positive signs from Brazil, Chile, India, Turkey, and others.3 Yet it is also clear that the conditions for such success have changed in a number of ways. First, past success by some industrializing countries has left less room for success by others. Mayer et al. (2002) call it a problem of “fallacy of composition,” in that a few countries may successfully upgrade to higher value-added exports, but if all countries expand in the same sectors simultaneously, many will not succeed and the resulting capacity expansion may dampen prices such that even successful countries will see their revenues lowered. Blecker (2002) describes the problem similarly as “diminishing returns to export-led growth.” Second, the intensification of the globalization of production, especially since 1980, has changed the structure in which the international division of labor is determined, thus altering the channels through which higher value added can be captured. Globalized production means that industrialization
Globalized Production 3
today is different from the final goods export-led process of just 20 years ago. Now the issue facing firms and even governments is often not simply that of finding new, more capital-intensive goods to sell to consumers in foreign countries. Instead, it seems to require moving up through the chain of production for a particular commodity or set of commodities, so-called “industrial upgrading.” This involves fitting into existing corporate strategies and, given the increasingly liberal international trade and investment environment, establishing close ties to a potentially diverse group of leading firms. It is in this context that the battle takes place for more jobs, higher wages and improved work conditions in developing countries. In this introductory chapter, I first give a brief profile of globalized production. I then consider both technological and social (i.e. economic) explanations of this process, focusing in particular on the degree of competition among firms along global value chains and the concomitant distribution of value added. Finally, I give an overview of the rest of the book. The two central questions addressed in these essays are: How does industrial upgrading occur? And what are the consequences of industrial upgrading efforts for workers? The chapters to follow show that not only are the obstacles to successful industrial upgrading significant and varied, but that the consequences of upgrading efforts are often far from the unambiguous positive effect that the term itself connotes. Countries that promoted low-skill labor-intensive manufacturing (most prominently through the establishment of export processing zones) as a stepping stone to higher value-added activities have often found themselves unable to move up, caught in what Harrison (1994) termed a “low-level equilibrium trap.” Mousiolek (2001) reports that during the 1990s, the garment sector in Eastern Europe underwent industrial downgrading, whereby full package producers went bankrupt, replaced by low-skill intensive, sweatshop operations serving the major European designers and retailers. Even in successful cases of upgrading, profits may rise without wage increases or some workers may benefit while others lose. The picture is a complicated one, but an informed analysis of globalization can only come from an understanding of the variety in developing countries’ experiences.
2
Global production sharing
The most commonly cited indicator of the degree of globalization is the share of trade in output. The trade share has risen steadily for the world as a whole, especially since around 1980, although not as much as Foreign direct investment (Figure 1.1). But the globalization of production means more than just an increase in the share of the world’s output that is traded internationally. Globalized production has also involved a change in the structure of international trade, in particular the growth of trade in intermediate goods, that is, in international outsourcing. Components may be produced in one location,
4
1000 900 800 700 600 500 400 300 200 100
World trade (index)
00
98
20
96
19
94
19
92
World GDP (index)
19
90
19
88
19
86
19
84
19
82
19
80
19
78
19
76
19
74
19
72
19
19
19
70
0
World FDI (index)
Figure 1.1 World GDP, FDI, and trade, 1970–2001
% 25
20
15
10
5
19
5 19 8 6 19 0 6 19 2 6 19 4 6 19 6 6 19 8 70 19 7 19 2 74 19 7 19 6 78 19 8 19 0 82 19 8 19 4 8 19 6 8 19 8 9 19 0 9 19 2 9 19 4 9 19 6 98 20 00
0
Broad
Narrow
Figure 1.2 Broad and narrow measures of international outsourcing in US manufacturing, 1958–2001 Note: “Broad” is the percentage of intermediate inputs that are imported. “Narrow” is only the percentage of intermediate inputs imported from the same (2-digit SIC code) industry. Source: Bivens (2003).
Globalized Production 5
shipped to another for assembly and then shipped to a third location for sale as a final good. I will refer to this as “global production sharing.”4 Is there evidence that heightened global production sharing constitutes a distinctly new era of globalized production? Figure 1.2 shows that US international outsourcing took off in the 1980s, and has continued to increase since, reaching significant levels of imported input use of above 20 per cent. Figure 1.3 shows the upward trend in the import of inputs in a variety of manufacturing industries in the major industrialized countries. The classic examples of the global production sharing are Ford’s “world car,” with components produced in over 14 countries and assembly performed in another three or four locations, or Nike’s shoe production, in
Country
1974
Canada Japan United Kingdom United States
15.9 8.2 13.4 4.1
1984
1993
All manufacturing industries 14.4 7.3 19.0 6.2
20.2 4.1 21.6 8.2
Chemical and allied products Canada Japan United Kingdom United States
9.0 5.2 13.1 3.0
8.8 4.8 20.6 4.5
15.1 2.6 22.5 6.3
Canada Japan United Kingdom United States
17.7 2.1 16.1 4.1
Canada Japan United Kingdom United States
13.2 3.1 14.9 4.5
17.1 3.4 23.6 6.7
30.9 2.9 34.6 11.6
Canada Japan United Kingdom United States
29.1 1.8 14.3 6.4
Transportation equipment 37.0 2.4 25.0 10.7
49.7 2.8 32.2 15.7
Industrial machinery (non-electrical) 21.9 1.9 24.9 7.2
26.6 1.8 31.3 11.0
Electrical equipment and machinery
Figure 1.3 Imported inputs as a share of total intermediate inputs, Canada, Japan, United Kingdom, and United States, selected years and sectors Source: Campa and Goldberg (1997).
6
William Milberg
which the American parent company employs 50 times more workers in Asia than it does in the United States. But the phenomenon has now spread to many manufacturing and service sectors, including finance and information technology. A description of Barbie Doll production and outsourcing gives a sense of the spread of the phenomenon to even low-value-added operations: The raw materials for the doll (plastic and hair) are obtained from Taiwan and Japan. Assembly used to be done in those countries, as well as the Philippines, but it has now migrated to lower-cost locations in Indonesia, Malaysia, and China. The molds themselves come from the United States, as do additional paints used in decorating the dolls. Other than labor, China supplies only the cotton cloth used for dresses. Of the $2 export value for the dolls when they leave Hong Kong for the United States, about 35 cents covers Chinese labor, 65 cents covers the cost of materials, and the remainder covers transportation and overhead, including profits earned in Hong Kong. The dolls sell for about $10 in the United States, of which Mattel earns at least $1, and the rest covers transportation, marketing, wholesaling and retailing in the U.S. The majority of value-added is therefore from U.S. activity. The dolls sell worldwide at the rate of two dolls every second, and this product alone accounted for $1.4 billion in sales for Mattel in 1995.5
3 Technological and social (economic) explanations of global production sharing Gereffi (1994) emphasizes the distinction between buyer-driven and producer-driven global value chains, the distinction depending on the nature of the lead firm in the chain. A producer-driven chain is typical in industries characterized by scale economies, and is often driven by transnational corporations who may outsource production but who keep R&D and final good production within the firm. Automobiles, computers, and aircraft are examples of this. Buyer-driven commodity chains occur mainly in consumer goods such as apparel, footwear, and toys. In this case the global commodity chain is driven by large retailers (e.g. Wal-Mart, The Gap), that is, firms that do no manufacturing themselves, but concentrate on design and marketing, subcontracting the actual production of the good. Whether the driver is a producer or a buyer, the motivation for global production sharing is normally the search for reduced costs or increased flexibility. Cost reduction can come from lower labor compensation or reduced taxes. Flexibility may increase with the use of short-term supply contracts or by using a non-unionized or unregulated labor force. When the cost of international communication and transportation fall, the return on outsourcing rises and global production sharing will increase. Such a trend in costs would
Globalized Production 7
% 50 45 40 35 30 25 20 15 10 5
19 77 19 82 19 83 19 84 19 85 19 86 19 87 19 88 19 89 19 90 19 91 19 92 19 93 19 94 19 95 19 96 19 97 19 98
0
Exports
Imports
Figure 1.4 Intra-firm trade as a percentage of total trade: United States, 1977–98
provide a technological explanation for the rise in global production sharing, and there is ample evidence that transportation and communication costs have fallen significantly in the past 20 years.6 Because of the simultaneous expansion of the activities of transnational corporations (foreign direct investment in Figure 1.1) and growth in trade in intermediate inputs (Figures 1.2 and 1.3), many analysts assume that the growth in the world trade share is the result of a growth in intra-firm trade, that is international trade within a single transnational company. Surprisingly, the share of trade that is intra-firm has been relatively constant for the past 25 years. Figure 1.4 shows that US intra-firm trade as a share of total US exports and imports for the period 1977–98 has been remarkably flat. A similar pattern is found in the intra-firm trade from Japan and Sweden, the only two other countries for whom reliable intra-firm trade data exist.7 With outsourcing increasing and intra-firm trade constant, the rise in the share of trade in intermediates must be the result of arm’s-length transactions, that is, international subcontracting outside the confines of the transnational corporation. Why should arm’s-length outsourcing be of increasing importance in a world where transnational corporations play such a large role? Here we must turn the theory of foreign direct investment on its head. Hymer (1976), and others, argued that the transnational firm is a nonmarket institution in the
8
William Milberg
Coaseian sense that the internal operations of the firm are not market processes. Use of such nonmarket processes reflects their apparent superiority, perhaps because of the transactions cost savings they bring compared to market transactions. Such savings, or rents, could result from the firm’s intangible assets related to technology, management, or marketing. While the internalization of international operations through foreign investment in this sense constitutes a market failure, the protection of such knowledge assets is widely recognized as the prime reason for firms to invest abroad rather than serve foreign markets in other ways, such as exports.8 If intra-firm trade is the result of firm internalization strategies, then the observed rise in non-arm’s-length subcontracting should be explained by externalization. That is, the expected relative return of arm’s-length subcontracting (as compared to foreign direct investment) must be rising. Why should cost reductions be increasingly prevalent externally rather than within firms? Langlois (2003) attributes this decline of the multi-unit, vertically integrated firm – a reversal of a trend identified by Chandler (1977) as central to the rise of industrial capitalism in the nineteenth century – to the growth in both the breadth and the depth of markets. Langlois writes: Rather than seeing the continued dominance of multi-unit firms in which managerial control spans a large number of vertical stages, we are seeing a dramatic increase in vertical specialization – a thoroughgoing “de-verticalization” that is affecting the traditional Chandlerian industries as much as the high-tech firms of the late twentieth century. In this respect, the visible hand – understood as managerial coordination of multiple stages of production within a corporate framework – is fading in a ghostly translucence … Costs of coordinating through markets may be high simply because existing markets – or, more correctly, existing market-supporting institutions – are inadequate to the needs of new technology and of new profit opportunities. But when markets are given time and a larger extent, they tend to “catch up,” and it starts to pay to delegate more and more activities rather than to direct them administratively within a corporate structure. (Langlois 2003, pp. 352–3) Certainly as productive capacity and quality have increased in many developing countries, the gains from flexibility that arm’s-length relations create have risen. Lead firms can set relatively short-term subcontracts, allowing the ability to respond more rapidly to changes in final good demand conditions or changes on the supply side, on issues ranging from product design, to wages, exchange rates, or government tax or regulatory policies in the countries with suppliers or potential suppliers. External outsourcing can itself stimulate competition among suppliers, reducing costs beyond what could be accomplished within the realm of internal operations. This competitive pressure on suppliers can translate into pressure on labor costs or on labor standards.9
Globalized Production 9
The end result is an asymmetry of market structure along global value chains. Entry barriers (often the result of brand marketing, but perhaps also due to proprietary technology or scale) allow the persistence of oligopolistic market structure at the top of the chain. Outsourcing promotes a competitive structure at the bottom. The process may be endogenous to the strategy of lead firms.10 The endogenous asymmetry of market structure in global commodity chains begins to explain the two, seemingly incongruous, tendencies that can be discerned in the evolving structure of global industry. On one side, and despite the popular association of globalization with more competition, there is a tendency toward greater concentration of industry globally (Nolan, 2003). The global wave of merger and acquisition activity constituted a consolidation of the oligopoly position of lead firms who, in the process, focused their efforts on “core competence” and outsourced other activities. On the other side, there is evidence that more and more developing countries are entering manufacturing industries at the low end of the value chain, seeming to introduce more, not less, competition at the world level (Mayer et al., 2001). Certainly most of the world’s largest firms are based in developed countries. Just 5 percent of Fortune 500 Companies and 3 percent of Financial Times 500 companies were based in low-income countries. Of the 27 developing country firms on the Financial Times 500 list, 24 were from Asia and only 3 were from Latin America. Of the 100 largest nonfinancial multinational enterprises in the world in 2000 (ranked by foreign assets), just five are from developing countries and two of these are petroleum producers (Petroleos Venezuela and Petronas of Malaysia).11 The evidence of greater dispersion of production across a wide variety of generally low value added manufacturing sectors is consistent with a number of recent econometric studies of competition in developing countries. Roberts and Tybout (1996) present a series of country studies that focus on entry and exit conditions. Summarizing evidence on Chile, Columbia, Mexico, Morocco, and Turkey for the 1970s and 1980s, they write that “… [E]ntry and exit rates are substantial … Despite the popular perception that entry and the associated competitive pressures are relatively limited in developing countries, these entry figures exceed the comparable figures for industrial countries.”12 Another study focuses on profitability and its persistence in seven developing countries – Brazil, India, Jordan, Korea, Malaysia, Mexico, and Zimbabwe – and compare it to estimates for industrialized countries. The authors find that “Surprisingly, both short- and long-term persistence of profitability for developing countries are found to be lower than those for advanced countries.” (Glen et al., 2002, p. 1). Finally, a study from the labor market perspective also confirms the competitive picture in developing countries. Brainard and Riker (1997) estimate the wage elasticity of labor demand across affiliates of US transnational corporations. Low-wage affiliates have little effect on employment in the home operation, but a large and significant effect on employment in other low-wage affiliates of the same firm.13
10 William Milberg
4 Upgrading in global commodity chains and consequences for labor The asymmetry of product market structures in global value chains reintroduces the “ruinous competition” from which capitalism escaped in the late 1800s with the rise of oligopoly in many industries, and has immediate implications for value added along global value chains.14 The asymmetry (endogenous or not) may lie behind the current situation in which developing countries have greatly expanded their share of global manufacturers exports while seeing their share of global value added in manufacturing rise by proportionally much less. This is a key issue, since the value added from export oriented production is an important source of income for reinvestment and consumption demand in these countries, both crucial elements of any successful industrial upgrading process. Competitive product markets, in theory, generate no economic profit and lower wages. Competitive firms have no rents to share with employees, and can survive only if wages are kept at a minimum. The increased use of sweatshop labor today, which has come with the rise in arm’s-length outsourcing, can be seen as tied to global production sharing.15 While the labor market effects of trade liberalization have been the subject of much research and heated debate among economists for many years, the effects of globalized production and in particular outsourcing have only begun to be studied. The irony is that precisely at the moment when computerization has led to a revolution in the mechanization of production, the ability to outsource has reasserted the importance of the labor component of production costs. Instead of becoming inconsequential as the result of technological change, labor costs are now an important determinant in the production location decision as firms increasingly slice up the value chain. According to Paul Krugman (1995: 336–7): It is often said that labor costs are now such a low share of total costs that low wages cannot be a significant competitive advantage. But when business people say this, they … mean that because of the growing vertical disintegration of industry the value added by a given manufacturing facility is likely to be only a small fraction of costs, which are denominated by the cost of intermediate inputs. But this vertical disintegration, or slicing up of the value chain, creates a greater, not a smaller opportunity to relocate production to low-wage locations. Why have trade and investment liberalization been associated with rising wage inequality in developing countries, contrary to the predictions of the Stolper–Samuelson theorem? Trade economists in the mid-1990s dubbed the phenomenon “skill-enhancing trade,” according to which, increasingly, specialization in low-skill intensive sectors still constituted an increase in the
Globalized Production 11
demand for skills.16 But skill-enhancing trade was more of an ex-post rationale than a full-blown theory, and the essays in this book show that there is much more involved than changes in technology induced by trade and investment liberalization. Other factors are also at work, including a mix between domestic and foreign firms creating variation in the overall skill demand, a weakening of labor laws and the reinforcement of existing labor market segmentation, locational disadvantages, domestic political tensions that mediate global pressures, and even moral codes. The (now) standard econometric finding in the literature on the positive relation between trade liberalization and relative (skilled/unskilled) wages may be spurious given all the other factors at work. While I have focused on the degree of competition in global value chains, the chapters that follow address the question of industrial upgrading from the perspective of particular nations and regions. In Chapter 2, UNCTAD economists Yilmaz Akyuz, Richard Kozul-Wright, and Joerg Mayer give a broad overview of developing countries’ patterns of investment, export growth, and industrialization, revealing stark differences in the East Asian and Latin American experiences. First-tier East Asian countries – Korea, Taiwan, and Hong Kong – have reached levels of industrialization similar to those in the wealthiest countries, and now maintain their level of industrial activity amidst healthy economic growth. Second-tier East Asian economies – Indonesia, Malaysia, Philippines, and Thailand – continue to increase the share of manufacturing in output while manufacturing productivity growth rises. Latin America seems to have experienced deindustrialization “prematurely,” that is, the manufacturing share has stopped growing relative to the rest of the economy while per capita incomes are much lower than those in the advanced industrialized countries. Sub-saharan Africa has also begun to deindustrialize, but in the context of declining output growth. Akyuz, Kozul-Wright, and Mayer emphasize the positive feebacks among capital investment, exports, and growth. Again, it is Indonesia, Malaysia, and Thailand that have been most impressive in their ability to continue to industrialize and raise manufactured exports while investment also rises. Mexico, on the other hand, has greatly expanded manufactured exports, but had a decline in manufacturing value added in GDP – perhaps the starkest case of the “low-level equilibrium trap” referred to above. In both Mexico and Brazil, the opening to foregin direct investment led to restructuring, toward more capital-intensive production, but diminished investment in more dynamic sectors intensive in R&D and engineering. Thus in the 1980s and 1990s, sectors associated with industrial upgrading grew in Korea, Malaysia, Taiwan, and Turkey, and they stagnated in Brazil, Mexico, Chile, and Argentina. The authors also identify a profit squeeze for many developing country exporters in the 1990s, consistent with the competitive pressures associated with asymmetric market structures in global value chains. This occurred in part because much developing country export
12 William Milberg
growth in the 1990s took place while productivity was stagnant. Thus the authors identify a high- and low-road to manufacturing expansion, the former associated with rising productivity, the latter with stagnant or falling productivity, in which export growth was driven by wage suppression or currency depreciation. The high-road was associated with “upgrading” to more medium- and high-tech exports. The low-road was associated more with expansion of assembly operations in consumer electronics and even automobiles. The next two chapters fill out the picture painted in Chapter 2, with detailed case studies from Latin America. Mexico, as seen in Jennifer Bair and Gary Gereffi’s example of the North American apparel industry in Chapter 3, seems to have required extensive subcontracting relations with American designers and retailers before moving into more sophisticated areas of production, ultimately to so-called “full package” production rather than the subcontracted role of cloth cutter it had traditionally played for American firms. Bair and Gereffi focus on the regional dynamics resulting from the North American Free Trade Agreement (NAFTA), and contrast the Mexican experience with that of countries in the Caribbean Basin to show the impact of distinct trade policies on export-oriented development. They argue that NAFTA is creating upgrading opportunities for some Mexican firms to move from the low value-added, export-oriented assembly (or maquila) model to full-package production. But upgrading in the apparel sector has been associated with an uneven development pattern across North America. Bair and Gereffi explore the unevenness of upgrading dynamics through a comparison of two blue jeans manufacturing clusters in the United States and Mexico: El Paso and Torreon. The comparison shows the importance of local, national, and regional institutional contexts in shaping inter-firm networks and their development impact. In Chapter 4, Janine Berg brings a detailed analysis of the Chilean cosmetics industry to bear on the general debate over the causes of rising wage inequality in developing countries, and in the process on the question of how difficult industrial upgrading is in an open economic environment. The Chilean cosmetics industry was closed to trade under the period of importsubstitution industrialization. The sector began to be liberalized in 1974 and has since faced increasing competition, particularly during the 1990s. The increase in foreign competition arising from free trade has had two principal effects depending on whether the firm is a foreign-owned multinational or a domestic, Chilean-owned firm. Multinationals have changed their competitive strategy in Chile, leading to a loss of production jobs, and increased relative demand and wages for skilled workers employed in management and sales. Domestic firms, on the other hand, have responded to the more competitive environment by making investments that expand and upgrade their manufacturing facilities. The technological changes have been low-skilled biased, leading to an increase in the relative employment of
Globalized Production 13
low-skilled workers. Yet the employment increases have not led to concomitant increases in the relative wages of low-skilled workers. Domestic firms have used the weakened labor relations environment to hold down wage increases for low-skilled workers, allowing the firms to continue underpricing the multinationals and retain market share. The case study reveals a complexity in the adjustment process that is not captured in standard models of trade liberalization and wages. In particular, with high levels of unemployment of low-skill workers and a weakening of institutions protecting labor bargaining power, a higher demand for low skills is not inconsistent with reduced wages for low-skill workers. In Chapter 5, the concluding essay of Part I, Radhika Balakrishnan and Asad Sayeed give an overview of theoretical debates over the “make or buy” decision by lead firms in global value chains. They propose a “push” and “pull” explanation for subcontracting outside the firm, both linked to cost competition. Firms are pushed into subcontracting when cost reductions are possible through lower wages alone. The pull factor occurs when competition also requires productivity enhancement, perhaps through greater flexibility or with the introduction of a new production technique. The consequences of industrial upgrading, even when it is achieved, are also not easy to generalize. Part II contains three essays on the skewed effects of globalization across groups of people and across space. In Chapter 6, Matthew Slaughter finds that inward foreign direct investment from the United States (as measured by the TNC share of total employment) has been associated with a rise in the demand for more skilled labor. Such skillupgrading associated with US foreign direct investment is found for all countries, but the effect is greater for developing countries. This is a broad and important finding, consistent with the results of Feenstra and Hanson’s (1997) study of oursourcing and wages in Mexico’s maquiladoras. Slaughter emphasizes that multinational firms affect both the demand for and supply of skills in host-country labor markets. On the demand side, inward foreign direct investment can stimulate demand for more-skilled workers in host countries through several channels. To date, most empirical evidence indicates that these channels work mainly within multinationals themselves, rather than through knowledge spillovers to domestic firms. On the supply side, the question of how inward foreign direct investment influences the development of human capital is much less clear, with possible links at both the micro- and macro-levels. This chapter offers some new empirical evidence on the links between inward foreign direct investment and withinindustry skill upgrading for a country–industry–year panel spanning both developed and developing countries. The main empirical finding is a robustly positive correlation between skill upgrading and the presence of affiliates of US multinationals, with this correlation even stronger among the subsample of developing countries. This correlation is consistent with inward foreign direct investment stimulating skill upgrading in these
14 William Milberg
developing countries. Slaughter concludes by putting the results in context, showing that the bias in the multinational enterprise demand for skills has a double edge in that it seems an essential part of the industrial upgrading but also implies rising wage inequality between low- and high-skill workers. In Chapter 7, Gunseli Berik, Yana van der Meulen Rodgers, and Joseph Zveglich explore another type of wage inequality: between men and women workers. They find that trade expansion in Taiwan and Korea has sometimes brought a rise in the gender wage gap, even when they control for all measured labor quality differences. They thus provide another example of the double edge of greater insertion into the global economy. The authors begin by elaborating the main theories of trade and relative wages by gender. The dynamic implications of Gary Becker’s theory of discrimination lead one to expect that increased competition from international trade would reduce the incentive for employers to discriminate against women. This effect should be more pronounced in concentrated sectors of the economy, where employers can use excess profits to cover the costs of discrimination. In nonneoclassical theory, wage discrimination is expected to increase with growing trade in a context of employment segregation that limits women’s ability to achieve wage gains. To test these competing theories, Berik, van der Meulen, and Zveglich study the impact of competition from international trade on the gender wage gap in Taiwan and South Korea between 1980 and 1999. They include controls for differences in market structure across industries in order to isolate the effect of competition from international trade. The estimation results are not consistent with Becker’s theory: greater international competition in concentrated sectors is associated with larger wage gaps between men and women. In particular, the authors find that rising import competition in the case of Taiwan is strongly associated with rising discrimination against women workers in manufacturing, while a decline in export competition in Korea is weakly associated with an improvement in women’s relative wages. They conclude with a discussion of the policy options facing countries seeking more gender fairness in the globalization process. The options relate mainly to the development of new, and implementation of existing, anti-discrimination legislation, showing once again the importance of strong democratic domestic political processes for the fair management of globalization. In Chapter 8, Stephen Gelb and Anthony Black examine the limits to the globalization of production and obstacles to it, by focusing on foreign direct investment in South Africa. The paper first provides a brief historical background on South Africa’s historical industrialization pattern, emphasizing the emergence of “mid-tech” industries most inward-focused, low rates of labor absorption resulting in very high levels of unemployment and extreme inequality (reinforced by apartheid) and the contribution of market- and resource-seeking foreign direct investment to industrial development. The authors present results from a recent firm-level survey of foreign direct
Globalized Production 15
investors who entered South Africa for the first time after 1990. The survey results confirm macro-level data suggesting disappointingly low inflows of foreign direct investment, suggesting that South Africa is not deeply integrated into global production chains and networks, notwithstanding the relatively advanced level of industrial development and the presence of large pools of unemployed (and unskilled) labor. Gelb and Black then turn to look more closely at three industrial sectors – autos (a producer-led chain), clothing (a buyer-led chain), and financial services – to examine some of the reasons for the low degree of globalized production. In each of these sectors, production has been globalized to some degree, but has encountered obstacles to further advancement. All three sectoral cases confirm the survey results that globalization of production in South Africa has progressed to a limited degree only, driven in large measure by the contribution of foreign investors, while domestic firms still find it difficult to orient toward exports and to insert themselves into global production chains and networks. What is the role for policy in molding a globalization process that facilitates upgrading and especially improved employment and pay conditions for the developing world? We began this introduction with a laundry list of proposals from the anti-globalization movement. The final two chapters identify two very different challenges to the effective regulation of globalized production. Elissa Braunstein and Gerald Epstein (Chapter 9) consider the possibility that a large domestic market can lead to more effective use of inward capital flows. In particular, the authors consider the nature of China’s bargaining power in relation to transnational corporations and foreign direct investment. They explore the questions of how China has used that power, what leakages have occurred as the government has tried to exploit its bargaining power, and what have been the impacts of the bargaining process on the Chinese population. The Chinese central government closely managed the process of foreign investment so that it would focus on exports rather than the domestic Chinese market. At the same time, the decentralized nature of some aspects of the policy made it difficult to manage all components of the foreign direct investment process. Braunstein and Epstein then present new empirical results assessing the impact of foreign direct investment on employment growth and wage growth. The regression analysis shows that foreign direct investment’s impact has been positive but rather limited in size. Significantly, they find that foreign direct investment has crowded out domestic investment. The authors also find the impact of foreign direct investment on local tax revenue to be negative, suggesting that, at least at the provincial level, the social benefits of foreign direct investment have been dissipated. The findings of this chapter call into question the desirability of making significant concessions to attract foreign investment in China. This caution applies even more strongly to other countries with much less bargaining power than China, which is to say, virtually every other country in the developing world.
16 William Milberg
Michael Piore’s discussion of the production process in the relatively lowtech woodworking sector in Ciudad Hidalgo in northern Mexico (Chapter 10) puts the issue of welfare and industrial upgrading in a broader context. Industrial upgrading, that is a shift to a new, more mechanized production process is key to raising labor standards. Such standards develop endogenously as part of the transformation of the production process, its labor relations, and the social norms that result. When firms move to more mechanized production, labor standards not only tend to improve, they also become easier to enforce. Imposing higher labor standards in a production environment in which such standards are unrelated to local conditions is not fruitful. Piore gives the example of the child labor regulations. Children run regularly about the woodshop. “Much of what is going on, in fact,” he remarks, “was not child labor but child care.” Labor standards only make sense if the production system itself is transformed in a way that is compatible, both technically and socially, with those standards. Why has the surge in global trade and foreign direct investment not resulted in a more significant boost in employment, wages, working conditions, technological change, industrial diversification, and export revenue for most developing countries? The studies in this book provide a variety of answers, giving a rich sense of the reasons that industrial upgrading is difficult and not necessarily welfare-improving for labor, certainly not for all labor. Overall, the essays show that globalized production, far from being a panacea for developing countries, creates a new set of challenges to economic development – for entrepreneurs, workers, governments, and international organizations. These new challenges, I have argued, result not from the extent of globalization per se, but from structural changes in the sphere of production and policy. The challenges posed by globalization for the process of economic development vary greatly from place to place, and thus no single policy should be adopted in all countries or even in different regions of the same country. As the “architects” of international economic policy turn their attention away from finance and increasingly toward global production, this lesson should loom large in their deliberations.
Notes 1. I am grateful to Yana van der Meulen Rodgers and Gunseli Berik for helpful comments on the first draft of this chapter. 2. See Amsden (1989). 3. See Amsden (2001) and Akyuz et al. (1998). 4. Synonyms abound, including “the international disintegration of production” (Feenstra, 1998), “fragmentation,” (Arndt and Kierzkowski, 2001) “the slicing up of the value chain,” (Krugman, 1995) “global production networks” (Harrison, 1994) and the “global commodity or value chain” (Gereffi, 1994). 5. Feenstra (1998). 6. See Arndt and Kierzkowski (2001).
Globalized Production 17 7. See UNCTAD (2001). 8. See, for example, Dunning (1988). 9. Similarly, arm’s-length relations with suppliers reduces the buyer firm’s responsibility for social standards in the supplying firm. A company like Nike can provide account for labor practices in Nike-owned production facilities, but is much less likely to be held accountable if the supplier is not owned by Nike. 10. See Milberg (2003) for a development of the notion of endogenous asymmetry of market structure. 11. Fortune (2000), Financial Times (2000), and UNCTAD (2002). 12. Roberts and Tybout (1996), p. 191. 13. Brainard and Riker (1997). 14. Shapiro (2002). 15. In this view, sweatshops are best understood as a historical or developmental phenomenon, rather than an ethical or moral category. See Piore (2000) and Chapter 10 in this volume. 16. Feenstra and Hanson (1997).
References Akyuz, Y. et al. (eds) (1998) Replicating the East Asian Miracle. Amsden, A. (1989) Asia’s Next Giant, New York: Oxford University Press. Amsden, A. (2001) The Rise of the Rest: Challenges to the West from the Late-Industrializing Nations, New York: Oxford University Press. Arndt, S. and H. Kierzkowski (eds) (2001) Fragmentation: New Production Patterns in the World Economy, Oxford: Oxford University Press. Bivens, L. J. (2003) “International Outsourcing and American Labor,” Working Paper, Washington, DC: Economic Policy Institute. Blecker, R. (2002) “Diminishing Returns to Export-Led Growth,” in P. Davidson (ed.), A Post Keynesian Perspective on Twenty-First Century Problems, Cheltenham: Edward Elgar. Brainard, S. L. and D. Riker (1997) “Are U.S. Multinationals Exporting U.S. Jobs?” Working Paper 5958, March, NBER. Campa, J. and L. Goldberg (1997) “The Evolving External Orientation of Manufacturing Industries: Evidence from Four Countries,” Working Paper 5919, NBER. Chandler, A. (1977) The Visible Hand: the Managerial Revolution in American Business, Cambridge: The Belknap Press of Harvard University Press. Dunning, John (1988) “The Eclectic Paradigm of International Production: A Restatement and Some Possible Extensions,” Journal of International Business Studies, 1–31. Feenstra, R. (1998) “Integration of Trade and Disintegration of Production,” Journal of Economic Perspectives, Fall. Feenstra, R. and G. Hanson (1997), “Foreign Direct Investment and Relative Wages: Evidence from Mexico’s Maquiladoras,” Journal of International Economics, 42 (3/4), May, 371–39. Financial Times (2000) “FT500,” May 4. Fortune (2000) “The Fortune 500” annual issue. Gereffi, G. (1994) “The Organization of Buyer-Driven Global Commodity Chains: How U.S. Retailers Shape Overseas Production Networks,” in G. Gereffi and M. Korzeniewicz (eds), Commodity Chains and Global Capitalism, Westport, CT: Greenwood Press.
18 William Milberg Glen, J., K. Lee, and A. Singh (2002) “Corporate Profitability and the Dynamics of Competition in Emerging Markets: a Time Series Analysis,” ESRC Centre for Business Research, University of Cambridge, Working Paper No. 248, December. Harrison, B. (1994) Lean and Mean: the Changing Landscape of Corporate Power in the Age of Flexibility, New York: Basic Books. Hymer, S. (1976) The International Operations of National Firms: a Study of Direct Foreign Investment, Cambridge: MIT Press. Krugman, P. (1995) “Growth in World Trade: Causes and Consequences,” Brookings Papers in Economic Activity. Langlois, R. (2003) “The Vanishing Hand: the Changing Dynamics of Industrial Capitalism,” Industrial and Corporate Change 12 (2): 351–85. Mayer, J., A. Butkevicius, and A. Kadri (2002) “Dynamic Products in World Exports,” Discussion Papers, Geneva: UNCTAD. Milberg, W. (2003) “The Changing Structure of Trade Linked to Global Production Systems: What are the Policy Implications?” Background Paper for World Commission on the Social Dimension of Globalization, Geneva: International Labor Organization. Mousiolek, B. (2001) “Industrial Downgrading and the State of Work in East Europe’s Apparel Industry,” Memo prepared for conference on Labor and the Globalization of Production, New York: Center for Economic Policy Analysis, New School University. Nolan, Peter (2003) “Industrial Policy in the Early 21st Century: the Challenge of the global Business Revolution,” Chapter 14 in Chang, Ha-Joon (ed.), Rethinking Development Economics, London: Anthem Press. Piore, M. (2000) Trade and the Social Structure of Economic Activity, in S. Collins, (ed.), Imports and the American Worker, Washington: Institute for International Economics. Roberts, M. and J. Tybout, (eds) (1996) Industrial Evolution in Developing Countries: Micro Patterns of Turnover, Productivity, and Market Structure, New York: Oxford University Press for the World Bank. Shapiro, N. (2002) “Competition,” in J. King (ed.), Elgar Companion to Post Keynesian Economics, Aldershot: Edward Elgar. UNCTAD (2001, 2002) World Investment Report, Geneva: United Nations.
Part I Reconceptualizing Globalized Production
This page intentionally left blank
2 Trade and Industrial Upgrading in Developing Countries Yilmaz Akyüz, Richard Kozul-Wright, and Jörg Mayer1
1
Introduction
It is generally accepted that capital accumulation can help raise per capita income and living standards in an economy simply by allowing a fuller use of underutilized labor and natural resources without altering the efficiency with which resources are utilized. Long-term economic success, however, depends on sustained improvements in productivity; each worker producing more from any given level of effort provides the basis for rising incomes and living standards. In this sense, it is productivity gains, and not simply additional jobs, that characterize a virtuous process of accumulation and growth. Such a process is invariably associated with structural changes in output and employment as a result of both shifts in economic activities across agriculture, industry, and services and upgrading to higher valueadded activities within each sector through the introduction of new products and processes. The importance of structure to the development process is partly due to the fact that the overall level of income is closely linked to the allocation of resources among sectors, and the sectors can show, at any point in time, significant variations in productivity levels. But it also derives from differences in the potential of various sectors for technical progress and productivity growth. Such differences emerge not only in the broad division of sectors into agriculture, mining, manufacturing, and services, but also in intra-sectoral structures. The importance of establishing a domestic industrial base to respond to development challenges lies in its potential for strong productivity and income growth. That potential derives, on the supply side, from a predisposition to scale economies, specialization, and learning and, on the demand side, from favorable global market and price conditions. Successful development experiences have established a close relationship between the growth rate of industrialization and of productivity (Kaldor, 1967), as well as between an acceleration of growth and a shift of labor from the lowproductivity primary sector into higher-productivity industry (Kuznets, 21
22 Yilmaz Akyüz, Richard Kozul-Wright, and Jörg Mayer
1955). These observations have been confirmed most recently by the experience of the East Asian Newly Industrialized Economies (NIEs) (see for instance Ros, 2000, pp. 19–30). Although changes in the structure of economic activity reflect some common underlying forces, there is also considerable potential for diversity across countries in their timing and extent, depending on the nature and composition of investment (both in machinery and equipment and in human and physical infrastructure), resource endowments, size and location. Foreign trade also exerts an important influence on the evolution of economic structure, insofar as it can help overcome domestic supply- and demand-side constraints on industrialization and growth. However, as with investment, the extent to which trade feeds into a more or less dynamic and virtuous industrialization process owes a good deal to policy choices and interventions. Following a discussion of the industrialization process in economic development, this chapter assesses how the main factors associated with building and maintaining industrial capacity, productivity, and the pattern of trade have changed in developing countries over the past two decades. Particular attention is given to changes in international specialization within the industrial sector through upgrading. Throughout, the chapter compares and contrasts the performance of economies in East Asia and Latin America and, to a lesser extent, Africa, with respect to structural change, productivity growth, international competitiveness, and trade.
2
Structural change and economic development
Industrialization and growth The challenge of narrowing income gaps with richer countries depends crucially on the creation of leading industrial sectors, along with related technological and social capabilities, in the context of the process of structural change that accompanies economic development (Abramovitz, 1986). A common pattern is discernible in most of the successful experiences. An initial sharp drop in the share of agriculture in total employment is followed by its continuous decline, steadying at a very low level as the economy matures. A weak rise in demand for foodstuffs, combined with relatively strong productivity growth in agriculture, explains the declining weight of the primary sector in overall economic activity. This is associated with a sharp rise in the share of industry in terms of both employment and output. During the “industrialization stage” mechanization spreads to the primary sector, thereby sustaining the fall in agricultural employment. At the same time, strong complementarities with the service sector ensure a steady rise in employment and output in commercial services, transportation, and finance. As the economy grows, the differential growth in productivity and demand between industry and services brings about further structural changes in
Trade and Industrial Upgrading 23
employment and output. While the growth in demand for manufactures slows down as incomes rise, relatively rapid productivity growth is maintained. As a result, industrial output keeps pace with demand without additional employment, and the share of industry in total employment starts to fall. If aggregate demand becomes sluggish, the industrial sector may start shedding labor, and hence lose employment in absolute as well as relative terms. On the other hand, relatively slow productivity growth of the service sector, coupled with a steady growth in the demand for its products, implies that this sector begins to absorb the employment released by industry. This process is accompanied by a persistent rise in the share of services in total employment and output, reaching over two-thirds at higher levels of income. These trends describe the process of “deindustrialization” that has occurred in mature high-income economies (Rowthorn and Wells, 1987). Such structural changes rarely occur as a smooth or harmonious process. Indeed, they pose new and difficult economic challenges for policy makers. In particular, as labor is released from agriculture its absorption is not assured, with a consequent risk of disguised or open unemployment. In the earlier stages of industrialization, rapidly increasing demand for manufactures could lead to balance-of-payments difficulties and threaten sustained economic growth if the primary sector is unable to provide the necessary foreign exchange earnings. At a later stage of industrialization, as seen in many European countries over the past three decades, slow growth in aggregate demand could mean that labor released from industry cannot be productively absorbed in the services sector, resulting in persistent unemployment. This process can be called “negative deindustrialization” as opposed to “positive deindustrialization,” the latter taking place in the context of rapid growth and full employment (Rowthorn and Wells, 1987, p. 25; UNCTAD, 1995: part three, chapter III). There has been a good deal of diversity in the pace and scale of industrial development across countries. Such diversity is clearly influenced by factors susceptible to strong policy influences and choices, including the pace and nature of capital accumulation, trade, and international competitiveness (see, for instance, Gomory and Baumol, 2000; Amsden, 2001). Resource endowments, size, and geographical location also have a strong bearing on the timing and extent to which labor shifts into industrial activities. Countries rich in natural resources can delay industrialization even as they experience faster growth, resulting in a lower share of employment in manufacturing at any given level of income. However, they cannot always avoid pressure to establish dynamic industrial sectors, since it may not be possible to reach high-income levels without a strong industrial base. Moreover, the pressure to diversify into industrial activities is likely to intensify if efforts to expand incomes are hindered by adverse terms of trade and external payments difficulties that prevent them from meeting the demand for manufactures. Indeed, those economies that have relied more heavily on
24 Yilmaz Akyüz, Richard Kozul-Wright, and Jörg Mayer
primary commodity exports to achieve higher levels of income, such as Australia, Canada, and some of the Scandinavian countries, have all experienced periods of strong industrial development and diversification as essential components of their sustained economic growth. For late starters, the industrialization process tends to be more capitalintensive, offering greater opportunities for rapid productivity growth due to their access to the technology and capital equipment produced in the more advanced economies. As a result, successful industrialization in developing countries is expected to create fewer jobs in industry at any given level of income. But opportunities for some late industrializers to become “workshop economies,” producing large quantities of labor-intensive products for export, can go some way toward offsetting this tendency. Thus, they can expand manufacturing employment beyond the limits set by the domestic market. In the same vein, a mature economy, with a competitive edge in key industrial sectors and a surplus in manufacturing trade, can normally employ more labor in those activities and delay deindustrialization. Deindustrialization associated with strong productivity growth in manufacturing has been a visible trend in the advanced industrial economies over the past few decades.2 Pooled data regressions on the share of manufacturing in total employment for 18 developed countries during the period 1963–94 suggest that the level of per capita income at which deindustrialization typically becomes a visible trend is between $8000 and $9000 (measured at constant 1986 prices), a figure already reached in the 1960s in a number of advanced industrial economies (Rowthorn and Ramaswamy, 1999). To the extent that most developing countries are well below this level of income, they should be expected to experience a steadily rising trend in the share of manufacturing in total employment and output. Indeed, as Tables 2.1 and 2.2 show, this was generally the case in most developing regions until the early 1980s. In almost all regions in Table 2.1, manufacturing employment started rising from the 1960s onward while confirming the tendency, noted earlier, for late industrializers to exhibit a relatively lower share of employment in manufacturing than the early industrializers. Latin America already had a high share of manufacturing in total employment during the 1960s, and this was maintained in the two subsequent decades, except in the Southern Cone countries where it declined sharply during the 1970s from the higher levels of the 1960s. East Asia started from a lower level but caught up rapidly with Latin American countries during the 1960s and, in particular, in the 1970s, while sub-Saharan Africa made modest progress during those decades. Output trends broadly paralleled this shift in employment. Again, in the early 1960s, Latin America, particularly the Southern Cone countries, had a higher share of manufacturing in GDP than other developing regions, and these shares were broadly maintained throughout the subsequent two decades. The high shares of both employment and output as early as the
Trade and Industrial Upgrading 25 Table 2.1 Manufacturing employment as a share of total employment (%), by region, 1960–2000 Region
1960
1970
1980
1990
2000
Sub-Saharan Africa Latin America Southern Cone West Asia and North Africa South Asia East Asia (excl. China) First-tier NIEs China Developing countries Developed countries
4.4 15.4 17.4 7.9 8.7 8.0 10.5 10.9 10.0 26.5
4.8 16.3 20.8 10.7 9.2 10.4 12.9 11.5 10.8 26.8
6.2 16.5 16.2 12.9 10.7 15.8 18.5 10.3 11.5 24.1
5.9 16.8 16.6 15.1 13.0 16.6 21.0 13.5 13.6 20.1
5.5 14.2 11.8 15.3 13.9 14.9 16.1 11.5 12.5 17.3
Notes: Sub-Saharan Africa includes Benin, Botswana, Burkina Faso, Cameroon, Central African Republic, Chad, Congo, Côte d’Ivoire, Democratic Republic of the Congo, Gabon, Ghana, Kenya, Lesotho, Malawi, Mauritania, Mauritius, Niger, Nigeria, Rwanda, Senegal, South Africa, Togo, Zambia, and Zimbabwe; Latin America includes the Southern Cone countries and Colombia, Costa Rica, Dominican Republic, Ecuador, El Salvador, Guatemala, Honduras, Jamaica, Mexico, Nicaragua, Panama, Paraguay, and Peru; Southern Cone includes Argentina, Brazil, Chile, and Uruguay; West Asia and Middle East includes Algeria, Egypt, Morocco, Oman, Saudi Arabia, Tunisia, and Turkey; South Asia includes Bangladesh, India, Pakistan, and Sri Lanka; East Asia includes Hong Kong (China), Indonesia, Malaysia, the Philippines, the Republic of Korea, Singapore, Taiwan Province of China, and Thailand. Source: Calculations, based on data provided by the International Labour Organization.
Table 2.2 Manufacturing output as a share of GDP (%), by region, 1960–2000 Region
1960
1970
1980
1990
2000
Sub-Saharan Africa Latin America Southern Cone West Asia and Middle East South Asia East Asia (excl. China) First-tier NIEs China Developing countries Developed countries
15.3 28.1 32.2 10.9 13.8 14.6 16.3 23.7 21.5 28.9
17.8 26.8 29.8 12.2 14.5 20.6 24.2 30.1 22.3 28.3
17.4 28.2 31.7 10.1 17.4 25.4 29.6 40.6 24.7 24.5
14.9 25.0 27.7 15.6 18.0 26.8 28.4 33.0 24.4 22.1
14.9 17.8 17.3 14.2 15.7 27.0 26.2 34.5 22.7 18.9
Note: For definitions of country groupings see Table 2.1. Sources: Calculations, based on data on manufacturing output and GDP at current prices from World Bank (1984 and 2003); and Government Statistical System of the Republic of China, online.
1960s suggest that the import-substituting industrialization pursued in the region did have a significant impact on the process of structural change. In East and South Asia, the share of manufacturing progressed steadily from the relatively low levels of the early 1960s, and it showed a noticeably steeper
26 Yilmaz Akyüz, Richard Kozul-Wright, and Jörg Mayer
rise in the first-tier NIEs. China already had a high share of manufacturing in GDP in the early 1960s due to its strong industrialization drive, and this increased even further in the two subsequent decades. In sub-Saharan Africa, there was also an upward trend during the period 1960–80, but it was much weaker than in Asian countries. In West Asia and North Africa, the rise in the share of manufacturing in GDP during the 1970s was reversed in the subsequent decade, due to the increasing importance of crude oil production in the region. The period since 1980 has been marked by a significant degree of divergence. The East Asian economies continued to industrialize at a rapid pace, with the first-tier NIEs reaching industrial maturity. The second-tier NIEs, thanks to their large natural-resource base, started to industrialize later, their industrialization gaining momentum from the late 1970s. China’s pattern of early industrialization clearly shows the influence of central planning. From the 1980s, following its increasing shift towards a market economy and the expansion of foreign-funded enterprises, there were reversals in terms of both employment and output (UNCTAD, 2002: chapter V). In sub-Saharan Africa, the share of manufacturing in total employment stagnated during the 1980s, associated with a decline in the share of manufacturing output in GDP; however, both stabilized in the 1990s at relatively low levels. Latin America seems to have experienced deindustrialization prematurely. The region as a whole saw a sharp drop in the share of manufacturing in total output during the 1980s and 1990s in the context of a significant slowdown in overall economic growth, while the share of manufacturing in employment started to fall in the 1990s after remaining relatively stable in the 1980s. The reversal was particularly pronounced in the Southern Cone countries. The share of manufacturing in total output in Latin America is now similar to that of the major industrial countries, while its level of per capita income is much lower.3 Turning to country experiences, only 8 of the 26 economies listed in Table 2.3 succeeded in raising the share of manufacturing value added in GDP from the 1980s to the 1990s. Surprisingly, perhaps, three of these countries were in Africa (Côte d’Ivoire, Egypt, and Ghana), but their shares nevertheless remained at modest levels. The Republic of Korea and Taiwan Province of China appear to be set to enter a phase of positive deindustrialization in the context of rapid growth. On the other hand, the East Asian second-tier NIEs (except the Philippines), which are in the intermediate stages of industrialization, experienced continuous and strong growth in the share of manufacturing value added in GDP and employment. By contrast, almost all Latin American countries listed in the table saw significant declines in the share of manufacturing value added in GDP. This was most pronounced in Argentina and Chile following their introduction of economic reforms, during the 1970s and 1980s in Chile and during the 1980s and 1990s in Argentina. In Brazil and Mexico too, the sharp fall in the share of
Trade and Industrial Upgrading 27 Table 2.3 Selected trade and production indicators (%) for 26 developing economies, 1960–2000 Country/economy
Manufacturing value added as a share of GDP 1960–69 1970–79 1980–89
1990– 2000
Exports of manufactures as a share of exports of goods and services 1980–89
1990– 2000
Argentina Bolivia Brazil Chile China Colombia Côte d’Ivoire Ecuador Egypt
38.6 — 28.2 23.8 29.0 18.9 10.3 18.6 —
35.3 — 30.0 24.2 37.3 23.0 9.4 17.8 15.7
29.3 — 32.6 19.7 35.8 22.0 16.0 19.4 14.6
20.3 15.8 23.7 18.0 34.0 17.0 18.8 20.9 17.8
25.9 2.8 44.2 6.6 67.5 15.4 8.3 1.6 7.8
26.4 15.3 46.8 10.6 78.0 23.9 11.9 5.4 10.0
Ghana India Indonesia Kenya Malaysia Mexico Morocco Nigeria Pakistan
11.4 13.6 9.0 10.5 9.5 20.1 16.2 5.0 14.3
11.1 15.3 10.4 12.0 16.8 22.7 16.7 4.8 15.9
8.0 16.4 15.1 11.8 20.3 23.2 18.0 8.2 16.0
9.2 16.4 22.8 11.2 27.3 20.6 17.6 4.9 16.6
— 16.2 29.6 7.1 27.7 29.3 26.4 — 55.3
7.0 55.4 45.1 15.8 63.0 62.3 33.7 1.1 73.4
Peru Philippines Rep. of Korea Taiwan Prov. of China Thailand Turkey Uruguay Venezuela
16.9 20.4 16.5 16.7 14.2 12.7 — 15.4
21.4 25.7 25.0 28.4 19.0 13.4 23.8 16.1
26.8 25.0 29.8 34.4 23.5 18.7 26.5 19.5
15.3 23.2 29.5 28.9 28.8 18.3 21.0 17.4
11.9 18.1 81.6 81.8 30.6 45.2 32.7 5.4
13.2 47.7 77.5 81.9 56.7 44.9 28.9 11.0
Sources: Calculations based on UN/DESA, Commodity Trade Statistics database; World Bank, World Development Indicators (2002); and Thomson Financial Datastream.
manufacturing activity in the 1990s coincided with an intensification of market-based reforms. In these cases, the decline in the relative importance of manufacturing activities occurred at a level of per capita income that was much lower than in either the industrialized economies or the East Asian economies. This might conceivably be interpreted as a desirable return to their comparative advantage in resource-based sectors, following their shift from
28 Yilmaz Akyüz, Richard Kozul-Wright, and Jörg Mayer
import-substituting industrialization to an outward-oriented strategy. However, while this shift was associated with a significant acceleration of growth in Chile, this was not the case for Argentina, Brazil, or Mexico. Furthermore, a comparison with European economies that have a wellendowed natural-resource base, such as the Scandinavian economies, shows that resource-rich Latin American countries, including Chile, are lagging considerably in industrialization, even allowing for their rich resource endowments. Available evidence on employment shows that the share of manufacturing employment in Chile and Argentina in the late 1990s was between one-half and one-third the level reached by the Scandinavian economies in the early 1960s, when they were at comparable income levels. In the resource-rich Scandinavian economies, the share of manufacturing employment started to fall from a much higher level of income – after having reached a higher peak – than the levels attained in Latin America. In Chile, while it is possible that manufacturing activity may pick up, in both relative and absolute terms, once the opportunities in the primary sectors are exhausted, the current level of industrialization does not appear to contain the many dynamic elements needed for such a transformation. This point has been made in a recent assessment of Chile’s pattern of structural change between 1960 and 1990, using an input–output accounting framework to gauge the strength of its industrial sector. The legacy of a decade of reforms, beginning in the mid-1970s, appears to be weaker backward and forward linkages in this sector, particularly in the technologically sophisticated segment of manufacturing. The downgrading of “heavier industries” is reflected, in particular, in a sharp jump in imported inputs to nearly two-thirds of total inputs (compared to less than 40 percent in the East Asian economies at a comparable stage of development) as well as a significant weakening of competitiveness in technology-intensive sectors (see the section on Trends in international specialization). These raise concerns about the longer-term technological prospects, self-sustainability, and overall stability of this pattern of structural change (Albala-Bertrand, 1999). Taken together, the above evidence shows that, unlike the advanced industrial economies and the East Asian NIEs, the deindustrialization trend in many developing countries in Latin America and sub-Saharan Africa has not been a benign product of differential productivity growth in the context of a steady economic expansion. Rather, it has coincided with a widespread slowdown in output growth. Indeed, a recent study of developing countries in the decades before and after the debt crisis in the 1980s has shown that across much of South and East Asia productivity remained high in the postcrisis period, often accelerating, and employment growth remained strong (Pieper, 2000). This pattern was supported by persistently high rates of economic growth in the periods before and after the crisis. Latin American economies have exhibited a different pattern. In most cases, growth in employment remained largely unchanged between the two periods while
Trade and Industrial Upgrading 29
output growth was similar or lower. As a result, productivity growth remained slow or negative. The exceptions to this pattern are Brazil, where higher productivity growth in manufacturing was achieved at the cost of employment, which fell sharply, and Chile and Costa Rica, both of which enjoyed stronger output and productivity growth in the later period. Economies in sub-Saharan Africa generally showed very little change across the two periods, reflecting widespread and persistent stagnation. Capital accumulation, trade, and industrialization As noted above, the pace and pattern of industrialization are greatly influenced both by the pace and pattern of capital accumulation and the participation of countries in international trade. Indeed, successful industrialization in developing countries is often based on mutually reinforcing dynamic interactions between capital accumulation and exports. This dynamic export– investment nexus is well known and is described in some detail in UNCTAD (1996) in relation to the evolution of the East Asian NIEs. Exports broaden the size of the market and thus allow scale economies to be exploited; they also provide the foreign exchange needed for capital accumulation, in view of the dependence of most developing countries on imported capital goods. At the same time, investment improves export potential by adding to production capacity and improving competitiveness through productivity growth. Such a process is typically characterized by rising investment, exports, and manufacturing value added, both in absolute terms and as a share of GDP. Over time, both foreign exchange and savings gaps close as exports and domestic savings begin to grow faster than investment. Figures 2.1A and B relate changes in the investment ratio over the period 1980–2000 to changes in the shares of manufacturing value added and exports in GDP, while Figure 2.2 relates changes in the latter two to each other. They show that countries which have been successful in sustaining a virtuous process of accumulation at the initial and intermediate stages of industrialization are the ones that have been able to combine rising investment with expanding manufacturing value added and exports. This is particularly the case for the three dynamic second-tier NIEs, namely Indonesia, Malaysia, and Thailand. On the other hand, the Republic of Korea and Taiwan Province of China are in more mature stages of industrialization, combining still rising investment ratios with falling or stable shares of manufacturing value added and exports in GDP. In both cases, the share of manufacturing value added in GDP is still at much higher levels than in advanced industrial countries, and the share of manufacturing in total exports has been stable, at high levels that had already been attained in the 1970s and 1980s (Table 2.3; see also Amsden, 2001: table 6.9). Most Latin American countries combined a declining share of investment in GDP with a declining share of manufacturing value added. While in many of these countries the share of manufactured exports in GDP remained
30 (A) 8
Indonesia Malaysia
Change in manufacturing value added (percent of GDP)
6 Thailand 4 Egypt
Côte d'Ivoire 2
Ecuador
–8
Pakistan Morocco 0 –3 Kenya Venezuela Philippines–2 Mexico
India
Rep. of Korea Turkey 7
2 Bolivia China
Chile
Nigeria –4
–6
Colombia Uruguay Taiwan Prov. of China
–8
Brazil
Argentina Peru
–10
Change in gross fixed capital formation (percent of GDP) (B) 20
Change in exports of manufactures (percent of GDP)
Malaysia (42) Thailand
Philippines 15
Mexico 10
Indonesia China
5 Pakistan Morocco India Bolivia Venezuela Kenya Turkey Colombia Côte d'Ivoire Ecuador Chile Egypt Argentina 0 Peru –8 –3 2 7 Brazil Uruguay
Rep. of Korea
Taiwan Prov. of China -5 Change in gross fixed capital formation (percent of GDP)
Figure 2.1 Change in manufacturing value added and exports of manufactures in relation to changes in gross fixed capital formation: 1990–2000 compared to 1980–90 (percent of GDP). A. Change in the share of manufacturing value added in GDP in relation to the change in the share of GFCF in GDP. B. Change in the share of exports of manufactures in GDP in relation on the change in the share of GFCF in GDP Sources: UNCTAD secretariat calculations, based on World Bank, World Development Indicators (2002); and Thomson Financial Datastream.
Trade and Industrial Upgrading 31 20 Malaysia (42) Thailand
Change in exports of manufactures (percent of GDP)
Philippines 15
Mexico 10
Indonesia
China
5 Pakistan Morocco India Kenya Bolivia Turkey Colombia Venezuela Côte d⬘Ivoire Argentina Chile Ecuador Egypt 0 Brazil −10 −5 0 5 Peru Rep. of Korea Uruguay
10
Taiwan Prov. of China −5 Change in manufacturing value added (percent of GDP)
Figure 2.2 Changes in manufacturing value added in relation to changes in exports of manufactures: 1990–2000 compared to 1960–90 (percent of GDP) Source: See Figure 2.1.
unchanged or fell, in others the falling or stagnant share of manufacturing value added in GDP was associated with a rising share of manufactured exports. This latter group consists of two sets of countries. First, the middleincome, primary-commodity exporters, where the increase in manufactured exports was moderate and started from a very low base (e.g. Colombia, Ecuador, and Venezuela); and second, countries where the increase in manufactured exports was due to their participation in labor-intensive assembly activities. The most important example of the latter is Mexico, which combined a lower share of investment and manufacturing value added in GDP with a rapidly expanding share of manufactured exports.4 In East Asia, only the Philippines manifests the same characteristics as Mexico in combining rising manufactured exports as a share of GDP with a falling share of manufacturing value added in GDP. As in Mexico, the share of investment in GDP fell in the Philippines during 1980–2000. While Malaysia also participates in international production networks through labor-intensive assembly activities, it succeeded in increasing both its manufactured exports and value added, although the increase in the former was much stronger (Table 2.3). China saw a small decline in the share of manufacturing value added in GDP, but a large increase in manufactured
32 Yilmaz Akyüz, Richard Kozul-Wright, and Jörg Mayer
exports, a disparity due partly to the extremely high share of manufacturing in GDP in the 1980s associated with central planning, and partly to the country’s participation in low-value-added assembly activities within international production networks. Chile was the only country in Latin America to combine a strong investment performance with a lower share of manufacturing value added in GDP and a moderate increase in manufactured exports, from a very low base. The support and protection given during the import-substituting industrialization of the 1960s and 1970s undoubtedly allowed industry in Latin America, and to a lesser extent in Africa, to expand considerably faster than would have been possible under competitive conditions. Unlike in East Asia, however, which also made extensive use of industrial policies, these strategies in Latin America and Africa were not always able to promote viable industries (Hirschman, 1968). Consequently, with big-bang liberalization and the withdrawal of support and protection, industries in these regions, confronted with stiff competition, were forced to downsize, rationalize, or perish. In this sense, the deindustrialization process, associated with the shift in the development paradigm, can be seen as a corrective step in the context of a Schumpeterian process of “creative destruction.” However, after so many years of reform and adjustment, there is little sign of creative forces initiating a new virtuous process of accumulation, growth, and structural change.
3
Productivity growth: inter-industry patterns
The close correlation observed in East Asia between high rates of investment, rising shares of manufacturing in GDP and strong export performance is underpinned by a rapid growth in productivity. However, the link between investment on the one hand and productivity growth and trade performance on the other is not automatic. While a shift to industrial activities is essential for attaining rapid productivity growth and high-income levels, it is not always clear how the allocation of investment across sectors influences the speed with which the productivity gap with advanced industrial countries is narrowed. Indeed, there is no consensus as to whether productivity gains associated with investment can best be captured in sectors with large or small productivity gaps with advanced industrial economies. While some authors (Gerschenkron, 1962) have invoked the “advantage of backwardness” to support the view that sectors with the largest productivity gap tend to attract the most investment and narrow that gap the fastest, others (e.g. Krugman, 1990) have suggested that developing countries tend to narrow the productivity gap at equal rates across industrial sectors. Productivity growth also depends on how investment is combined with learning in the context of technological progress. Even where technology is embodied in imported capital equipment along with complementary
Trade and Industrial Upgrading 33
codified knowledge, certain aspects of any technology are tacit, and thus can be acquired only through learning-by-doing. Furthermore, using any imported technology efficiently would necessitate modification to suit specific local conditions. Thus, a country’s initial knowledge base, combined with step-wise learning, determines how well it copes with and applies new technologies. From this perspective, technological change is the joint outcome of investment in modern capital equipment and learning how to use it efficiently (Nelson and Winter, 1982; Abramovitz, 1986; Lucas, 1993; and Nelson and Pack, 1999). Targeted technology policies also have a direct bearing on the outcome. An important development that has influenced the sectoral pattern of technology transfer and absorption is the increasing vertical integration of production into distinct value-added stages located in different countries, and the greater participation of developing countries in such global production networks. The kinds of industrial activities most easily relocated from more to less developed countries are those that use easily traded intermediate products, and in which the share of wages in production costs is high, because such activities benefit from a variation in wage costs across locations. Significant imports of both machinery and equipment and intermediate goods appear to accompany increased participation in such networks (UNCTAD, 2003: chapter IV). Although this has been seen as a possible basis for technological leapfrogging and rapid acceleration of productivity growth, the technology transfer and learning processes in such networks are increasingly circumscribed by global strategies of transnational corporations (TNCs), rather than by national development strategies of the recipient countries. Thus the pace of productivity growth in developing countries and the speed with which the productivity gap with developed countries in different sectors can be reduced are affected by the nature of their participation in international production networks, as well as by technology and capital goods imports and the process of learning and adaptation. Table 2.4 shows the evolution of labor productivity in various developing countries and in the United States, the world’s technology leader. It covers the manufacturing sector as a whole as well as a selection of low-, medium-, and high-skill industries. The data are presented for 1980, 1985, and 2000 to enable an assessment of how the debt crisis and extensive policy reforms, which in most countries occurred between 1985 and 1990, might have affected sectoral productivity trends. Since the data reported sometimes refer to different periods, and large proportional changes often occur in countries with small industrial bases, the evidence needs to be interpreted with caution. In all Asian countries for which data is available labor productivity has improved significantly, and in most cases continuously across all sectors, over the past two decades, while no such trend is discernible in Latin America (with the exception of Chile). Moreover, in many countries in these regions, productivity levels fell during the 1990s (i.e. the period after widespread trade
34
Table 2.4 Labour productivity in 26 developing economies and selected industrial sectors, 1980–2000 (Index numbers, 1990 100) Country/ economy
Total manufacturing (ISIC 300) 1980
Argentina Bolivia Brazil Chile China Colombia Côte d’Ivoire Ecuador Egypt Ghana India Indonesia Kenya Malaysia Mexico
— 77.0 — 80.2 — 75.2 — 79.8 55.6 — 55.2 54.0 83.7 67.1 —
1985 122.1 58.3 79.9 115.3 — 87.5 — 77.2 89.4 — 72.5 68.9 88.8 93.9 102.5
Food products (ISIC 311)
2000
1980
85.1a 90.8b 114.0c 144.6 242.1d 101.3 — 117.3d 90.7b — 152.4 124.2d 89.4e 171.1 108.0
— 85.8 — 97.7 — 67.8 — 86.5 73.5 — 34.6 39.9 94.0 90.6 67.8
1985 134.7 171.6 89.4 117.0 — 97.1 — 101.7 94.6 — 78.3 54.6 95.6 96.4 101.1
2000
Textiles (ISIC 321) 1980
88.1a — 122.8 115.5 108.9c — 149.6 79.8 311.5d — 105.5 63.0 — — 97.8d 99.4 81.6e 80.1 — — 174.0 69.9 113.2 45.8 98.8e 104.2 162.7 60.2 101.3 111.7
1985 109.9 93.1 91.8 97.9 — 72.9 — 95.1 94.0 — 67.7 67.4 88.4 61.5 115.9
Clothing (ISIC 322) 2000
1980
55.7a — 98.0 149.3 76.9c — 121.7 98.8 181.7d — 51.3 91.8 — — 101.3d 157.2 86.4e 93.1 — — 107.4 43.3 158.1 39.0 74.0e 111.7 208.6 62.8 82.3 —
1985 118.4 103.6 108.6 123.9 — 89.7 — 106.4 134.0 — 52.1 73.8 112.9 73.9 118.0
Electrical machinery (ISIC 383) 2000
1980
94.8a — 109.7 150.6 78.3c — 184.8 49.5 224.4d — 105.6 74.9 — — 93.1d 119.9 181.3e 90.9 — — 107.5 64.2 147.6 56.3 105.8e 25.1 151.2 66.6 85.2 113.6
1985 101.2 79.1 83.2 83.5 — 86.7 — 107.6 232.7 — 69.3 76.7 32.9 98.6 96.8
2000
Transport equipment (ISIC 384) 1980
64.2a — 81.0 192.0 102.0c — 104.4 98.3 285.1d — 78.4 52.9 — — 61.5d 69.1 162.3e 76.0 — — 173.1 60.8 155.7 47.8 90.9e 105.6 219.3 40.9 107.4 111.6
1985 128.0 47.9 74.2 110.6 — 65.5 — 94.4 128.9 — 73.9 44.0 128.9 55.6 95.0
2000 103.9a 84.6 180.6c 174.6 — 62.2 — 109.8d 262.8e — 123.2 187.2 69.8e 116.8 158.1
Morocco 85.8 Nigeria — Pakistan 63.4 Peru 107.3 Philippines 74.1 Rep. of Korea 50.7 Taiwan Prov. 61.9 of China Thailand — Turkey 61.3 Uruguay 88.0 Venezuela 83.6 Memo item: United States
80.6
81.8 — 86.0 107.0 79.9 65.1 72.3
117.0d 110.8 — — 89.5 120.8a 82.0a 117.5 75.0 150.0b 231.8 57.2 57.3 127.1a
114.8 — 106.6 74.6 72.3 67.2 82.8
131.5d 79.0 — — 118.8a 41.2 57.5a 120.9 149.5 88.7 205.8 61.0 110.6a 51.3
99.9 — 61.0 116.7 49.7 78.5 66.7
99.2d 57.6 — — 106.1a 61.0 76.1a 119.6 140.2b 77.1 233.1 58.6 127.4a 70.1
103.4 — 67.0 101.3 50.6 62.4 78.3
116.4d 70.3 — — 133.8a 49.7 124.5a 101.3 145.3b 59.9 196.5 38.7 92.2a 56.4
88.4 — 61.9 85.9 46.6 64.0 67.9
85.9d 57.8 — — 218.9 64.0 66.2 173.0 96.4b 63.5 330.0 41.8 148.6a 54.1
81.1 — 65.5 105.5 28.8 65.4 61.8
85.3d — 200.2 68.1 152.5b 187.6 118.0a
— 71.5 125.9 91.3
— 121.3 127.5d 114.1e
— 65.4 70.6 92.7
— 71.8 121.7 103.1
— 134.4 111.7d 183.0e
— 75.3 76.3 99.0
— 77.7 114.6 124.7
— — 114.2 59.5 115.3d 98.2 45.9e 142.1
— 68.9 97.7 156.4
— — 148.5 62.8 66.6d 69.6 90.3b 105.9
— 76.5 118.4 110.2
— — 135.6 54.5 81.1d 66.3 98.2b 137.3
— 59.9 80.9 142.1
— 135.4 48.2d 260.8b
89.0
114.7c
79.5
92.1
113.2
84.1
89.4
98.6
149.9
118.0
82.7
93.0
144.1
78.5
88.3
220.4
80.3
Notes: Labour productivity calculated as real value added (in national currency) per worker. Nominal value added deflated by the GDP-deflator. Electrical machinery includes semiconductors and telecommunications equipment. a 1996; b 1997; c 1995; d 1999; e 1998. Sources: Calculations based on UNIDO, Industrial Statistics Database (2002); World Bank, World Development Indicators (2002); and Thomson Financial Datastream.
35
36 Yilmaz Akyüz, Richard Kozul-Wright, and Jörg Mayer
and financial liberalization), and in some cases they dropped below the levels observed in 1985 (i.e. in the middle of the debt crisis). A number of countries in Latin America, such as Argentina, Brazil, and Mexico, experienced a particularly sharp productivity decline in traditional labor-intensive sectors such as textiles and clothing. By contrast, productivity performance was better in transport equipment than in manufacturing as a whole, and productivity growth in that sector in Brazil and Mexico even exceeded that observed in the United States. Productivity growth in food products in countries such as Brazil, Colombia, and Mexico kept up comparatively well with that in the United States. In Asia, the Republic of Korea achieved higher rates of productivity growth than the United States, both in total manufacturing and in each of the sectors in the table. The evidence further suggests that other Asian economies, including India, Indonesia, Malaysia, Pakistan, the Philippines, and Taiwan Province of China, as well as Turkey and Chile, also successfully reduced the productivity gap with the United States for manufacturing as a whole, but the rate of productivity growth varied widely across industries. A number of these countries have imported a substantial amount of machinery from developed countries over the past three decades (UNCTAD, 2003: chapter IV). Significantly, Mexico, which also imported large amounts of machinery, did not share in this productivity trend, except in transport equipment and, to a lesser degree, in electrical machinery. Differences across countries in the development of labor productivity in various industries are also reflected in changes in the relative weight of individual sectors in total industrial value added. Table 2.5 shows the increasing importance of industries based on natural resources in the major Latin American economies, with the exception of Mexico, and the declining importance of the metalworking industries (including the information technology subsectors), with the exception of Brazil. This implies that opening up to international competition and FDI led to a shift in the production structure toward the relatively capital-intensive sectors involved in processing abundant natural resources, while those activities intensive in research and development (R&D) and in engineering lost weight in total industrial output, thereby reducing the potential for productivity growth and innovation (Cimoli and Katz, 2001). Regarding structural change, Mexico holds an ambiguous position which reflects the weight of the automobile sector and maquiladora industries – sectors with different labor intensities – in its recent pattern of industrialization. However, an examination of the structural-change index of the United Nations Industrial Development Organization (UNIDO) suggests that, while there was a restructuring in manufacturing after market-oriented reforms, structural change remained the same for the period 1984–94 under the new policy regime as it had been for the 1970–81 period under import substitution (Moreno-Brid, 1999, pp. 48–9; Máttar et al., 2002, pp. 27–8). Moreover,
Trade and Industrial Upgrading 37
the sectors with the largest increase in their share in GDP during 1987–94 (namely automobiles, motors, and accessories, and nonelectrical machinery and equipment) also had the largest increase in 1970–81. Thus more than a decade of economic reforms appears not to have radically changed the structure and dynamics of manufacturing activity. These changes in the share of different sectors in manufacturing value added in Latin American countries significantly differ from those observed in the Asian economies shown in Table 2.5. While the intensity of these changes differed considerably across the Asian economies, the metalworking and automotive industries gained in importance in all of them. The shift toward metalworking was accompanied by a sharp decline in the importance of natural-resource-based and labor-intensive industries. A recent study provides evidence that the share of those manufactures commonly associated with successful industrial upgrading (electrical and nonelectrical machinery and transportation equipment) grew particularly rapidly in the Republic of Korea, Malaysia, Thailand, and Turkey, but much less so in Brazil, Chile, Mexico, and, particularly, Argentina (Amsden, 2001). While structural change within the manufacturing sector was limited during the period 1980–94, it was greatest in economies such as Indonesia and Malaysia, which experienced substantial growth in manufacturing value added as a share of GDP, and lowest in semi-industrial economies, such as Argentina and Mexico, as well as in more mature industrial economies (the Republic of Korea and Taiwan Province of China), where that share contracted. Thus the weak performers included countries that experienced both “positive” and “negative” patterns of deindustrialization.5
4
Trade and the pattern of structural change
Close integration into the international trading system through rapid liberalization has been the cornerstone of economic reform in developing countries since the mid-1980s. This could be expected to influence the pattern of structural change, along with resource endowments and geography. However, the impact of trade integration largely depends on the circumstances under which it takes place, and on the policies pursued during the integration phase. Integration in Latin America and Africa marked a sharp shift in development strategy, occurred in a big-bang manner, and followed the debt crisis (i.e. in a period of weakness). This contrasts with the integration process in East Asia that occurred from a position of strength and was characterized by a continuous and purposeful strategy of gradual opening up. China represented in some ways an exception, as it combined a rapid pace of integration with accelerated growth, largely because its opening up was from a position of strength common to other Asian countries. It is likely that the pattern of industrialization and structural change across the developing world since the debt crisis has been related to these underlying patterns of trade integration.
38
Table 2.5 Sectoral shares (%) in manufacturing value added in selected developing economies, 1970–2000 Sectors
I II IIIIV V
Argentina 1970 1980
1990 1996
1970
1980
15.6 9.9 36.2 38.2
14.3 8.5 46.7 30.5
18.8 9.9 35.8 35.5
— — — —
Sectors
I II IIIIV V
Brazil
— — — —
13.7 7.7 48.6 30.0
Mexico
Chile
1990 1996 — — — —
22.8 8.7 42.4 26.1
Colombia
1970
1980
1990
2000
1970
1980
1990
2000
14.9 7.7 43.2 34.2
7.7 2.6 61.5 28.2
8.1 1.8 64.6 25.5
6.7 2.3 66.8 24.1
10.7 2.9 45.7 40.7
10.2 4.0 50.1 35.6
9.6 4.3 51.2 34.9
7.3 2.3 53.0 37.4
Malaysia
Republic of Korea
Taiwan Province of China
1970 1980
1990 2000
1970
1980
1990 2000
1970
1980
1990
2000
1973
1980
1990
1996
13.3 5.5 46.8 34.4
14.1 14.4 48.8 22.6
9.8 3.2 49.5 37.5
21.4 4.3 41.2 33.1
30.5 5.6 36.8 27.1
9.1 5.4 45.5 39.9
16.6 6.1 39.5 37.8
29.1 10.5 31.6 28.8
39.7 11.8 27.9 20.6
21.1 5.0 35.7 38.3
21.5 6.3 37.5 34.6
28.7 7.8 35.8 27.7
36.2 8.0 35.3 20.4
— — — —
11.8 18.9 43.8 25.4
48.9 4.1 29.0 18.0
Notes Sector I: Metal-working industry (ISIC 381, 382, 383, and 385). Sector II: Transport equipment (ISIC 384). Sector III IV: Food, beverages, and tobacco (ISIC 311, 313, and 314) and natural resource processing industries (ISIC 341, 351, 354, 355, 356, 371, and 372). Sector V: Labor-intensive traditional industries (ISIC 321, 322, 323, 324, 331, 332, 342, 352, 361, 362, 369, and 390). Sources: Cimoli and Katz (2001); UNIDO, Industrial Statistics Database (2002).
Trade and Industrial Upgrading 39
Industrialization and competitiveness It is generally agreed that a country’s pattern of participation in international trade is determined to a large extent by its resource endowments and the efficiency with which resources are utilized. Conventional wisdom suggests that greater international mobility of capital should increase the importance of relative endowments of high- and low-skilled labor in shaping the effect of trade on the pattern of industrialization. In particular, in most developing countries, industries using low-skilled labor should be expected to attract capital, making that labor more productive and those industries more competitive on international markets. According to this view, developing countries should specialize in low-skilled, labor-intensive manufactured goods and import high-skill-intensive goods from advanced countries. This would lead to a narrowing of the wage gap between unskilled and skilled workers in developing countries. Changes in industrial specialization and linkages between industry, trade, and development have been seen in recent years through the lens of international competitiveness. However, a degree of caution is needed in applying this concept to economic challenges facing developing countries. In the first place, strictly speaking, the concept may be useful to define the position of individual enterprises vis-à-vis each other, but not for comparisons among economies as a whole or even among industries comprising many firms with different characteristics (Krugman, 1994). Many countries that have highly competitive firms in certain industries find it necessary to protect others against foreign competition, and this is true at almost every level of industrialization and development. Furthermore, from a private perspective it may matter little whether the international competitiveness of an enterprise is improved through productivity growth, wage cuts, or a devaluation of the currency, but from a broader socioeconomic point of view, these have totally different implications for economic and social stability and welfare. Finally, competitiveness is a relative concept and there is an adding up problem; it is not possible for all countries to simultaneously improve the competitiveness of their firms in a given industry. On the other hand, the success of several developing countries in simultaneously raising productivity and wages can improve their overall economic welfare without altering their relative competitive positions in the sectors concerned. Studies on international specialization often assume that in an economy rates of productivity growth differ widely across industrial sectors, while wages develop more equally across industries.6 Such asymmetry can provide an important source of structural change. In a dynamic context, uneven productivity growth across industries, combined with even growth in wage rates, implies that workers in industries with relatively higher productivity are not fully compensated. The productivity gains are thus spread over the whole economy through general wage increases and changes in relative prices. Such wage and productivity dynamics – and hence the development of relative unit
40 Yilmaz Akyüz, Richard Kozul-Wright, and Jörg Mayer
labor costs – will have a major bearing on the comparative cost advantages of different countries in specific industries. If a country experiences relatively faster productivity growth in some industries than in others, while wages are growing at similar rates across industries, it will gain comparative advantage in the catching-up sector, provided that productivity and wage developments in other countries do not follow the same pattern.7 A comparison of unit labor costs in the sample of 26 developing countries relative to the United States for a number of manufacturing categories in 2000 does not reveal a consistent pattern, as ratios differ substantially for individual countries in different industrial categories, as well as for individual categories in different countries (Table 2.6). The fact that the table includes data for countries with small industrial sectors, which therefore often experience considerable fluctuations in their levels of wages and productivity over time, may partly explain this absence of a consistent pattern. For all industrial categories and all the selected countries taken together, there have been almost equal numbers of upward and downward changes in the ratios over the past two decades. But it is noteworthy that the Republic of Korea is the only country in the table where the ratio of unit labor costs to that in the United States fell in all five sectors. In a number of other countries (notably India), labor costs relative to the United States also fell in traditional labor-intensive industries such as clothing. By contrast, this ratio rose in other sectors for a large number of countries, particularly in electrical machinery. It is perhaps surprising that this is the case even for economies that have become, in varying degrees, successful exporters of telecommunications equipment and semiconductors within international production networks, such as Malaysia, Mexico, the Philippines, Taiwan Province of China, Thailand, and Turkey. Strong productivity growth appears to have been a key determinant of export success in these products in Malaysia and Taiwan Province of China, while relatively slow productivity growth in Mexico and the Philippines (see Table 2.4) suggests that other factors, including wages and exchange rates, played a more important role in these countries in retaining competitiveness. An industry or a business enterprise can be called internationally competitive if it can sell its products at the same price (or slightly below) and earn the same return as its competitors. While this definition of competitiveness is straightforward, measuring changes in the international competitiveness of a country’s tradables sector is more complicated, particularly for developing countries for which the required data are often unavailable. The real exchange rate is a widely used index of the competitiveness of a country’s tradables sector. A popular definition of the real exchange rate relies on the purchasing power parity approach, according to which the real exchange rate equals the nominal exchange rate multiplied by the ratio of the foreign price level to the domestic price level. An assessment of the international competitiveness of a country’s industrial sector would, ideally, be based on
Table 2.6 Unit labor costs in 26 developing economies and selected sectors, 1980 and 2000 (Ratios to the United States level) Country
Food products 1980
2000
Textiles 1980
Clothing
2000
1980
2000
Electrical machinery 1980
2000
Transport equipment 1980
2000
Argentina Bolivia Brazil Chile China Colombia Côte d’Ivoire Ecuador Egypt Ghana
0.87a 0.86 0.53c 0.63 0.68 0.60 0.92 1.36 1.45 1.00
1.95b 0.61 0.74b 0.80 — 0.62 1.50d 0.88e 1.45f 0.82b
0.48a 0.93 0.42c 0.65 0.26 0.47 0.85 0.91 1.27 0.80
1.28b 0.76 0.65b 0.89 — 0.66 1.06d 0.30e 1.21f 0.96b
0.48a 0.82 0.39c 0.55 0.08 0.58 0.73 0.82 0.99 0.45
0.64b 0.65 0.47b 0.51 — 0.47 1.02d 0.34e 0.38g 0.60b
0.70a 0.51 0.52c 0.88 0.59 0.48 0.78 0.96 1.00 1.08
2.11b 1.00 0.81b 0.90 — 1.01 1.34d 1.20e 1.10g 0.39b
0.79a 0.47 0.60c 0.46 0.42 0.53 0.36 0.86 1.51 0.84
1.78b 1.34 0.53b 0.74 — 0.97 1.69d 0.55e 0.71g 1.63b
India Indonesia Kenya Malaysia Mexico Morocco Nigeria Pakistan Peru
1.74 0.97 1.16 0.60 1.00 2.08 0.99 — 0.43
1.29 0.71 1.31e 1.08 0.90 1.61e 0.29b — 1.02b
1.25 0.61 1.00 0.75 0.85 1.19 0.85 — 0.43
1.57 0.42 2.20e 0.59 0.88 1.38e 0.80b — 0.62b
0.96 0.95 0.94 0.82 0.69h 1.25 0.52 — 0.66
0.47 0.45 0.96e 0.84 0.64 1.05e 0.11b — 0.46b
1.01 0.49 1.47 0.71 0.73 1.42 0.56 — 0.37
0.98 0.62 0.74e 1.01 1.06 1.49e 0.56b — 0.95b
1.24 0.40 1.10 0.67 0.49 1.34 0.09 — 0.25
1.43 0.26 3.34e 0.69 0.43 0.92e 0.04b — 0.50b
Philippines Rep. of Korea Taiwan Prov. of China Thailand Turkey Uruguay Venezuela
0.63 0.81 0.94 0.46i 1.12 1.65 1.34
0.65d 0.73 1.93b 0.92j 1.09 1.64e 0.93d
0.60 0.74 1.09 0.46i 0.70 0.84 1.14
0.67d 0.63 1.45b 0.87j 0.69 0.74e 0.72d
0.80 0.71 0.44 0.67i 0.62 0.76 1.03
0.59d 0.62 0.80b 1.07j 0.43 0.69e 0.49d
0.60 0.82 0.97 0.35k 0.72 1.03 0.98
0.80d 0.56 1.81b 0.65j 0.97 1.52e 0.68d
0.47 0.78 0.78 0.48k 0.98 0.72 0.86
0.40d 0.71 1.17b 0.41j 0.65 1.22e 0.17d
Source: UNCTAD secretariat calculations, based on UNIDO, Industrial Statistics Database (2002).
41
Notes: Unit labour costs calculated as wages (in current dollars) divided by value added (in current dollars). a d g j 1984. 1997. 1998. 1994. b e h k 1995. 1999. 1984. 1982. c f i 1985. 1996. 1979.
42 Yilmaz Akyüz, Richard Kozul-Wright, and Jörg Mayer
the relative price of foreign to domestic production baskets of internationally traded industrial goods. But as data on this are unavailable for most countries, an assessment is made here based on consumer-price indices. An alternative index of changes in a country’s degree of competitiveness refers to relative unit labor costs (i.e. the ratio of nominal wages expressed in dollars to labor productivity also expressed in dollars, relative to the same ratio in the United States). This definition of the real exchange rate is particularly useful as it allows the decomposing of changes in international competitiveness into the relative impact of changes in nominal wages, labor productivity, and the nominal exchange rate. A combination of a virtuous and sustainable improvement in social welfare and a high degree of international competitiveness is characterized by strong productivity growth associated with a rise in investment and increased or stable employment, a rate of growth of real wages that keeps pace with productivity, and a nominal exchange rate that maintains purchasing power parity. Table 2.7 summarizes the findings regarding the changes in international competitiveness and export performance of domestic manufacturers over the past two decades.8 The growth rate of manufactured exports is a key performance indicator, but since it may be misleading with respect to countries that start from a small base of manufactured exports, it is supplemented in the table by the share of manufactures in total non-oil merchandise exports in 2000. Of the 26 economies in the table, Côte d’Ivoire, Nigeria, Peru, and Uruguay have shown poor growth in manufactured exports over the past two decades. Despite a more rapid growth performance, the share of manufactures in total non-oil exports has remained very low in Chile, Ecuador, Ghana, and Kenya. This group includes three countries that experienced the strongest improvement in the competitiveness indicator: Ghana, Ecuador, and Nigeria. These countries raised the competitiveness of their manufactures through wage repression or a sizeable currency depreciation, rather than through strong productivity performance, which suggests that the improvement in their competitiveness represented a correction of the imbalances between low productivity and relatively high domestic wages and prices.9 By contrast, the Republic of Korea and Taiwan Province of China registered strong growth in manufactured exports based on a significant increase in labor productivity. As a result, manufacturers in these economies were able to maintain competitiveness, while at the same time achieving the fastest increase in wages among all economies listed in the table. China, Malaysia, and Mexico experienced particularly strong growth in their manufactured exports, which today account for about 90 percent of their total non-oil exports, but their performance in terms of labor productivity growth was much less impressive; a phenomenon consistent with the observation above that the increase in manufacturing value added in these countries has lagged behind that in manufactured exports, although to varying degrees.
Table 2.7 Indicators related to the international competitiveness of exporters of manufactures in 26 developing economies (Index numbers for 2000 with 1980 100, unless otherwise indicated) Country
Argentina Bolivia Brazil Chile China Colombia Côte d’Ivoire Ecuador Egypt Ghana India Indonesia Kenya Malaysia Mexico Morocco Nigeria Pakistan Peru Philippines Republic of Korea
Real dollar exchange rate based on consumer price indexa
(1984–96) (1980–97) (1985–95) (1980–99) (1980–97) (1980–99) (1980–97) (1980–95) (1980–99) (1980–99) (1984–2000) (1980–99) (1980–96) (1980–96) (1980–96) (1980–97)
Real dollar exchange rate based on unit labor costsb
Real effective exchange ratec
Nominal wage per workerd
Workerd productivitye
Unit labour costsf
Real wagesg
Memo items (%) Average annual growth of exports of manufactures
Share of manufactures in total non-oil merchandise exports in 2000
47.7 164.8 50.9 183.7 — 173.2 195.8 244.6 92.4
23.3 106.7 39.9 162.3 — 127.6 140.7 340.5 148.7
66.7 159.1 43.3 155.8 343.2 153.4 150.9 218.1 —
240.5 94.8 152.2h 168.3 — 191.1 107.2 44.6 146.1
50.5 73.9 114.8h 180.4 142.3 138.2 110.2 105.9 158.8
101.9 66.7 96.3 82.1 — 101.0 106.9 36.7 42.5
73.5 78.7 137.4 148.0 — 136.0 107.8 54.0 69.3
13.9 19.0 8.6 14.0 27.4j 11.7 3.4 16.0 11.8
40.1 33.2 60.2 16.2 91.3 57.0 18.4 17.4 63.7
698.3 215.8 331.3 153.0 187.5 78.2 173.0 119.7 188.7
178.0 300.1 285.5 175.9 160.2 67.0 202.0 864.3 —
651.3 215.6 332.2 — 151.8 73.9 131.8 232.4 180.7
58.6 141.3 114.7 97.9 241.1 213.4 96.8 28.8 —
77.9 279.9 228.2 120.1 255.2 113.0 136.3 183.3 177.1
81.0 52.8 81.7 61.8 84.9 90.2 60.8 25.3 95.2
221.5 145.9 188.0 74.1 216.5 100.7 82.9 18.1 181.4
12.5 12.0 24.8 10.0 22.1 23.8 10.6 3.9 12.8
16.0 81.1 76.5 22.6 89.7 92.5 66.5 57.5 86.1
35.3 120.6 129.1
52.1 105.3 130.4
— 118.9 129.0
227.3j 263.2 533.5
140.1j 202.6 459.5
47.4 80.5 72.1
36.2 163.0 329.8
4.9 17.5 12.1
21.9 92.9 96.0
43
Country
Taiwan Prov. of China Thailand Turkey Uruguay Venezuela
44
Table 2.7 Continued Real dollar exchange rate based on consumer price indexa
Real dollar exchange rate based on unit labor costsb
Real effective exchange ratec
Nominal wage per workerd
(1980–96)
86.7
49.7
91.4
550.7
(1982–94)
108.5 139.3 113.8 122.4
75.4 184.6 120.0 453.3
171.3 108.8 92.0 161.6
141.6 161.7 175.4 42.4
(1980–99) (1980–98)
Workerd productivitye
Unit labour costsf
Real wagesg
205.9
121.0
98.6 197.0 146.6 136.2
140.9 54.5 68.0 19.2
Memo items (%) Average annual growth of exports of manufactures
Share of manufactures in total non-oil merchandise exports in 2000
248.6
12.9
96.4
105.9 107.8 98.5 26.3
30.4 17.4 6.8 15.0
79.8 83.1 47.5 63.7
Notes a Index of bilateral exchange rate with the United States dollar multiplied by the ratio of index of United States consumer prices to the index of domestic consumer prices; an index number higher than 100 indicates a real depreciation of the local currency. b Ratio of domestic unit labor costs to United States unit labor costs. c Based on relative consumer prices. d Calculated on the basis of dollar values. e Real value added per worker calculated by deflating value added (in United States dollars) per worker by the GDP-deflator. f Ratio of nominal wages in manufacturing (deflated by the consumer price index) to value added in manufacturing (deflated by the GDP-deflator). An index number higher than 100 indicates an increase in the share of labor in the functional distribution of income. g Nominal wage per worker deflated by the consumer price index. h 1990–95. i 1985–99. j 1982–96. Sources: UNCTAD secretariat calculations, based on IMF, International Financial Statistics (2002); World Bank, World Development Indicators (2002). UNIDO, Industrial Statistics Database (2002) and Thomson Financial Datastream.
Trade and Industrial Upgrading 45
Most countries that sought to increase their international competitiveness, but achieved little or no improvement in labor productivity, appear to have had to resort to wage suppression or sharp depreciations. Thus the level of wages fell in most African and Latin American countries in the table. Evidence further suggests that since the mid-1980s rapid trade liberalization in these regions has also been associated with growing wage inequality between skilled and unskilled labor (UNCTAD, 2001; ILO, 2001). While various explanations have been offered for this trend, the extent to which countries have responded to competition from emerging, low-cost producers of labor-intensive manufactures by cutting wages or replacing less educated with better educated labor, rather than by new investment and upgrading, appears to be of particular importance. The competitiveness of Latin American manufacturers was further undermined by sharp appreciations. Indeed, the sharpest nominal appreciations among all the countries listed in the table occurred in Argentina, Brazil, and Peru, and this has been a major factor in the very strong deterioration in the international competitiveness of these countries’ manufacturers over the past two decades. The limited data on the sectoral distribution of investment available for a number of Latin American countries (ECLAC, 2001: table I-6) suggests that there is indeed a positive link between the development of industrial investment as a share of GDP and labor productivity in the manufacturing sector. In Chile, the sectoral investment coefficient of industry more than doubled between the early 1980s and the late 1990s and, as can be seen from Table 2.7, this rise was accompanied by a strong increase in labor productivity in Chilean manufacturing. By contrast, the sectoral investment coefficient of industry in Peru fell during the 1990s to about half its average level of the 1970s and 1980s, a drop that was accompanied by a sizeable decline in the country’s labor productivity in manufacturing between 1980 and 1996. In Bolivia and Colombia, the sectoral investment coefficients changed little between the mid-1980s and mid-1990s, with only a slight rise in manufacturing labor productivity. In Brazil, as noted above, there was a sizeable improvement in manufacturing productivity, attained through labor-shedding rather than investment. However, none of the countries that improved their competitiveness by wage suppression or massive devaluations achieved sustained improvements in export and value-added performance to a similar extent as countries that had succeeded in raising productivity and wages in a virtuous process of capital accumulation and employment expansion. Upgrading exports As already noted, some production and export patterns are more favorable to industrialization and growth than others. It is possible to establish a virtuous circle between investment, exports, and growth by investing in sectors with significant productivity and market potential, and using the
46 Yilmaz Akyüz, Richard Kozul-Wright, and Jörg Mayer
export proceeds to finance imports of capital goods and intermediate inputs required for further productivity increases. Exports of more technologyintensive manufactures are of key importance in this context because primary sectors often face adverse terms-of-trade movements as well as limits to raising productivity, and markets for labor-intensive manufactures exported by developing countries are rapidly becoming oversupplied (UNCTAD, 2002: chapter IV). In examining the links between industrial upgrading and exports, five broad groups are used here based on a distinction between primary commodities (Group I) and manufactures; the latter are further distinguished according to whether their production processes rely mainly on labor and natural resources (Group II), and whether they are characterized by lowtechnology intensity (Group III), medium-technology intensity (Group IV), or high-technology intensity (Group V).10 Table 2.8 shows that between 1980 and 2000, the share of primary commodities in total non-oil exports declined rapidly in all the economies, for some of them from an already low level, as in the Republic of Korea and Taiwan Province of China. Exceptions are Chile and Côte d’Ivoire, where the decline was much more modest. The fall in commodity prices relative to manufactures clearly played an important role in this trend. But only two countries in the table, Côte d’Ivoire and Colombia, experienced an absolute decline in export earnings from primary commodities due to sharp declines in wood and coffee exports respectively. While a number of countries also experienced sharp falls in export earnings in certain commodities (Argentina in cereals and sugar, Brazil in cocoa and coffee, and Turkey in cotton and live animals), their total earnings from commodity exports increased. Chile has been particularly successful in changing the composition of its primary commodity exports; it has raised the share of food and other agricultural products and reduced that of nonferrous metals, especially copper. The sharp fall in the share of primary commodities in non-oil exports contrasts with the steep rise in the share of medium- and/or high-technologyintensive products in three major Latin American economies (Argentina, Brazil, and Mexico) and the three East Asian economies shown in the table (Malaysia, the Republic of Korea, and Taiwan Province of China). India, Morocco, and Turkey experienced the largest increase in the share of laborand resource-intensive manufactures, while the Republic of Korea and Taiwan Province of China are the only economies in the table where this product group declined in importance along with the drop in the share of primary commodities. However, success in upgrading differs significantly among the Asian and Latin American economies that shifted to technology-intensive products. Industrial upgrading of exports from the Republic of Korea and Taiwan Province of China has been based on a comparatively wide range of medium- and high-tech products. This has led to an increase in the relative
Table 2.8 Commodity structure of exports (% of total non-oil exports) from selected developing countries, 1980–2000 Commodity group
Group I Group II Group III Group IV Group V Other manufactures
Republic of Korea
Taiwan Prov. of China
2000
1980
1990
2000
1980
1990
2000
1980
1990
2000
1980
1990
2000
9.9 42.5 19.1 8.2 16.8 3.5
5.5 33.3 14.7 13.3 27.9 5.3
4.0 14.8 11.5 21.5 46.3 1.9
10.7 40.6 8.6 12.3 18.6 2.9
6.7 28.3 10.3 18.7 27.5 5.6
3.6 14.1 10.5 19.5 48.2 1.2
74.9 6.7 0.7 3.0 14.3 0.4
33.7 12.3 3.2 8.5 39.0 3.2
10.3 9.3 1.9 10.9 66.1 1.5
— — — — — —
20.4 38.9 5.9 13.7 14.8 6.3
8.7 33.2 8.4 15.7 26.2 7.8
40.9 38.5 5.7 7.0 5.1 2.7
26.6 51.1 4.8 6.6 9.3 1.6
18.9 52.6 6.6 6.6 11.7 3.6
Brazil
Chile
Colombia
Mexico
1980
1990
2000
1980
1990
2000
1980
1990
2000
1980
1990
2000
1980
1990
2000
76.0 8.7 3.1 5.4 6.7 0.2
68.2 10.0 8.1 5.4 8.2 0.2
59.9 8.8 4.9 14.9 11.1 0.3
60.3 9.4 6.4 15.0 8.3 0.6
46.3 12.7 13.5 15.6 11.0 0.9
39.8 12.9 8.6 19.3 18.8 0.7
91.4 1.8 2.0 0.9 4.0 0.0
89.7 3.7 1.4 0.9 3.9 0.2
83.8 5.2 1.4 3.0 6.3 0.3
79.6 13.2 1.3 2.5 2.8 0.6
59.8 26.0 5.1 2.5 5.6 1.0
43.0 22.1 6.2 8.6 18.7 1.4
40.6 8.7 2.2 19.4 24.9 4.2
30.3 7.8 7.5 35.8 16.5 2.1
7.5 12.8 4.5 43.6 29.3 2.3
Côte d’Ivoire
Group I Group II Group III Group IV Group V Other manufactures
India
1990
Argentina
Group I Group II Group III Group IV Group V Other manufactures
China
Malaysia
1980
Morocco
Turkey
Egypt
Ghana
1980
1990
2000
1980
1990
2000
1980
1990
2000
1980
1990
2000
1980
1990
2000
93.3 3.2 0.5 1.6 1.4 0.1
87.2 9.9 0.7 0.6 1.5 0.1
81.6 8.8 1.4 2.6 5.2 0.2
75.3 12.8 0.2 0.6 10.8 0.3
45.8 27.2 0.9 2.6 23.0 0.5
33.5 39.8 1.0 4.7 20.4 0.7
70.6 22.2 1.4 3.0 2.4 0.4
30.6 42.0 13.6 4.9 8.0 0.8
16.9 44.2 9.9 16.4 10.3 2.4
69.6 26.5 1.9 0.1 1.7 0.2
39.8 44.8 6.7 1.2 7.0 0.5
36.3 36.9 5.3 5.5 13.1 2.8
98.4 0.6 0.2 0.2 0.5 0.0
94.9 2.4 1.7 0.3 0.3 0.3
84.0 9.0 1.1 4.3 1.3 0.3
Sources: UNCTAD secretariat calculations, based on United Nations Commodity Trade Statistics tapes and estimates by the United Nations Statistical Office.
47
Note: Total non-oil exports refer to SITC sections 0–8 less section 3. The commodity group “Other Manufactures” includes sanitary and plumbing products; toys and sporting goods; office and stationary supplies; works of art; jewellery; musical instruments; and other miscellaneous manufactures. For further explanations see text.
48 Yilmaz Akyüz, Richard Kozul-Wright, and Jörg Mayer
importance of electrical and nonelectrical machinery, road motor vehicles, industrial chemicals, and electronics, as well as ships and boats in the Republic of Korea and iron and steel in Taiwan Province of China. By contrast, industrial upgrading in Malaysia has been based on a much narrower range of products, concentrating on electrical and nonelectrical machinery, as well as electrical and electronic goods, in the context of the participation of the economy in international production networks. The increase in the share of manufactures in the three large Latin American economies has also been based on a relatively narrow range of products. In Mexico, the share of automobiles in total non-oil exports has grown strongly, along with that of the electronics and labor- and resource-intensive industries, such as clothing and wood products. Argentina and Brazil have also experienced a strong increase in the share of automobile exports, as well as chemicals, pharmaceuticals, and aircraft. There appears to be a close relationship between the evolution of the structure of exports and the inter-industry pattern of investment in the major Latin American countries over the past three decades. Evidence provided by ECLAC (2001: table A-7) suggests that there has been little change in the ranking of industrial sectors regarding the allocation of investment, and that the share of the five most important sectors taken together strongly increased in all countries except Chile. Generally, there has been no significant shift in investment toward technology-intensive industrial categories in any of the Latin American countries for which data are available. Indeed, in almost all cases where a substantial change occurred in the inter-sectoral pattern of investment, there was a shift toward resource-based or laborintensive products (i.e. metal products in Chile, paper and pulp in Chile and Colombia, and clothing in Peru). The main exception is transport equipment in Brazil and Mexico, where investment registered a strong increase. The automobile sector in these two countries has experienced substantial restructuring over the past few years, based on investment by TNCs. This, however, has increased the technology content of automobile exports without leading to a similar increase in their domestically generated contents (UNCTAD, 2002: chapter III). Trends in international specialization An assessment of the extent to which changes in the industrial composition of exports examined above have been associated with a consolidation of countries’ positions in international trade requires a comparative analysis of changes in international specialization. For this purpose, changes in international trade patterns are analysed by comparing the sector-specific indices of revealed comparative advantage (RCA) based on export data for 21 industrial sectors and the periods 1980–84 and 1996–2000 (Table 2.9).11 The evidence shows that in the Republic of Korea and Taiwan Province of China, the greatest increase in the RCA indices was in the medium- and
Table 2.9 Indices of revealed comparative advantage for manufactured exports of selected economies, 1980–2000 Industrial sector
Rep. of Korea
Malaysia
Taiwan (Province)
China
Argentina
Brazil
Chile
Mexico
Period Average Period Average Period Average Period Average Period Average Period Average Period Average Period Average annual annual annual annual annual annual annual annual I II change I II change I II change I II change I II change I II change I II change I II change Labour-intensive and resource-intensive Textiles 2.3 Clothing 9.2 Footwear, 4.9 leather and travel products Wood, cork 9.7 and furniture Paper, print 6.7 and publishing Non-metallic 5.2 mineral products
manufactures 4.6 9.0 8.3
4.1 3.7 5.2
8.4 3.5 5.5
4.7 6.2 1.6
4.5 4.5 0.9
14.4 14.6 11.5
2.9 3.6 1.9
6.0
1.6
5.0
17.5
2.2
4.6
1.1
6.0
4.1
-1.4
3.3
2.0
1.5
4.9
3.6
1.2
0.2
2.2
0.9
1.8
2.2
Manufactures with Iron and steel Fabricated metal products Simple transport equipment Shipbuilding
low technology content 1.8 2.3
7.7 4.4
7.2 1.7
2.2 0.0
1.3 1.4
2.3 1.0
2.6 0.3
5.6 3.8
Manufactures with Rubber and plastic products Non-electrical machinery Electrical machinery
2.9
5.0
8.9
10.9
1.6
1.3
14.5
7.7
2.7
1.2
4.7
5.3
5.5
6.0
7.3
3.3
medium technology content 1.6
0.5
-2.3
9.1
5.4
1.4
4.1
6.7
11.1
6.6
0.8
0.2
2.8
0.9
5.2
0.5
6.0
5.1
5.7
10.1
3.7
3.8
0.7
1.8
49
Industrial sector
50
Table 2.9 Continued Rep. of Korea
Malaysia
Taiwan (Province)
China
Argentina
Brazil
Chile
Mexico
Period Average Period Average Period Average Period Average Period Average Period Average Period Average Period Average annual annual annual annual annual annual annual annual I II change I II change I II change I II change I II change I II change I II change I II change Road motor vehicles
Manufactures with high Industrial chemicals Pharmaceuticals Computers and office equipment Communications equipment Aerospace Professional & scientific equipment Other manufacturing
5.9
3.1
23.3
12.0
0.4
2.4
8.3
technology content 8.8
8.7
3.4
4.0
0.3
0.3
2.7
5.0
18.0
2.8 10.1
6.9 12.7
10.1 32.9
10.0 20.7
1.6 17.3
1.4 10.5
7.4 0.4
4.4 0.8
0.7
4.2
7.3
0.7
9.8
2.0
2.3
0.4
0.3 2.2
10.3 2.5
0.5 3.0
13.4 0.3
14.7 0.7
7.3 0.2
1.8 0.1
3.0 1.0
2.2
3.5
2.5
0.8
5.6
1.6
8.8
2.5
Notes: RCA is measured as an economy’s share in total world exports in a given sector divided by the economy’s average export share in all manufacturing sectors. A plus sign indicates a RCA of 1 or above, a minus sign indicates a RCA of less than 1. Manufactured exports refers to SITC Rev. 2, 5–8 less 68. Other manufacturing includes SITC Rev. 2, 812 and 894–9. Period I refers to 1980–84, except for China where it refers to 1987–90; period II refers to 1996–2000. Source: UNCTAD secretariat calculations, based on COMTRADE and estimates by the United Nations Statistical Office.
Trade and Industrial Upgrading 51
high-technology manufacturing categories, and the sharpest declines were in the labor-intensive and resource-based manufacturing categories. The data for Chile show the opposite picture, with the largest increase being in labor- and natural-resource-intensive manufacturing sectors and the greatest decline in the high-tech sectors. While changes in the RCA indices of Argentina and Brazil are more varied, the pattern is similar to that of Chile. The main feature that distinguishes the three Latin American countries from each other is the strong increase of these indices in the aerospace sector in both Argentina and Brazil (moving from below to above unity in the case of Brazil) and their sharp increase in Argentina in communications equipment, and automobiles. This is particularly noteworthy, since automobiles and aerospace are the two industries in Argentina and Brazil that have been supported by industrial policy in recent years, despite extensive market-oriented reforms.12 Changes in the RCA indices of Mexico and Malaysia reflect the increasing involvement of these two economies in assembly-based activities within international production networks. Both countries acquired or increased their RCA in computer and communications equipment manufacturing. Similarly, Mexico experienced a strong increase in its RCA index for road motor vehicles. Given the greater import intensity of production and exports in developing countries, the examination of the pattern of exports does not provide adequate indications as to the evolution of inter-industrial patterns of production and value added in these countries. In this sense, improvements in the pattern of exports, particularly in terms of a shift toward high-tech products, does not necessarily indicate a concomitant improvement in the pattern of production and manufacturing value added. Analysis of sectorspecific evidence (Table 2.10), comparing bilateral structural similarity indices for exports of manufactures and manufacturing value added for selected country groups and economies, shows that for a developing country, a higher degree of similarity with respect to any of the leading developed countries in terms of the pattern of manufactured exports does not necessarily imply a corresponding similarity in its pattern of manufacturing value added.13 Indeed, while the structure of manufactured exports of developing countries as a whole became increasingly similar to that of developed countries as a whole between 1980 and 1998, this was much less so for the pattern of manufacturing value added. The composition of both manufactured exports and value added of most Asian economies shown in the table came to resemble more closely that of the major developed countries, but this was not generally true for the other countries. The evidence also suggests that there is little correlation between the growing similarity in the structures of manufactured exports and manufacturing value-added. Among the developing countries listed in the table, the Republic of Korea stands out for having reached a manufacturing valueadded structure that was by far the closest to that prevailing in the leading
Structural similarity with United States Exports 1980–81
Japan
Value added
1997–98 1980–81 1997–98
Exports
Germany Value added
1980–81 1997–98 1980–81 1997–98
Exports
Value added
1980–81 1997–98 1980–81
1997–98
Asia Hong Kong, China Rep. of Korea Singapore Taiwan Prov. of China
1.26 1.06 0.74 1.08
1.01 0.53 0.70 0.57
0.95 0.61 0.47 0.66
0.73 0.38 0.57 0.64
1.24 0.90 0.63 0.97
1.03 0.52 0.36 0.57
0.94 0.52 0.47 0.55
0.79 0.36 0.57 0.55
1.29 0.94 0.72 1.05
1.17 0.58 0.89 0.67
1.03 0.59 0.46 0.59
0.93 0.31 0.51 0.52
Malaysia Philippines
1.32 1.30
0.71 0.92
0.71 0.75
0.67 0.67
1.19 1.35
0.71 0.93
0.59 0.77
0.68 0.63
1.31 1.25
0.88 1.05
0.72 0.79
0.61 0.71
China India
1.14 1.26
0.89 1.27
0.68 0.69
0.62 0.68
1.31 1.34
0.90 1.34
0.61 0.58
0.57 0.63
1.08 1.24
0.99 1.19
0.60 0.61
0.60 0.66
Turkey
1.59
1.21
0.74
0.73
1.55
1.24
0.62
0.67
1.50
1.14
0.66
0.74
Latin America Chile Colombia Costa Rica Mexico Venezuela
1.33 1.17 1.22 0.90 0.95
1.15 1.10 0.86 0.47 0.93
0.74 0.69 0.78 0.91 0.59
0.82 0.76 0.76 0.80 0.78
1.50 1.35 1.29 0.93 1.06
1.33 1.27 0.94 0.45 1.19
0.69 0.67 0.75 0.82 0.51
0.76 0.74 0.79 0.74 0.73
1.30 1.16 1.16 0.91 0.98
1.08 0.97 0.97 0.50 0.97
0.84 0.73 0.82 0.85 0.63
0.88 0.85 0.88 0.73 0.79
Memo item: Developing countries
Structural similarity with developed countries’ average 0.87 0.57 0.46 0.37
Notes: The index values are the sum of the absolute differences between the home country and foreign country in the shares of different sectors of manufacturing industry in total export of manufactures or in total manufacturing value added. This measure varies between zero and two; a value of zero indicates identical sector compositions of the two economies, and a value of two indicates complete dissimilarity of sectoral structures. Sources: UNCTAD secretariat calculations, based on data from Nicita and Olarreaga (2001). The structural similarity indices have been calculated using a method suggested by Krugman (1991).
52
Table 2.10 Structural similarity indices for exports of manufactures and manufacturing value added, selected developing economies, 1980–81 and 1997–98
Trade and Industrial Upgrading 53
developed countries. The manufactured export structure of China, Malaysia, Mexico, the Philippines, and Singapore also began to resemble that of the major developed countries, but this was much less so for the structure of their manufacturing value added. For the majority of Latin American countries, the structure of their manufactured exports became moderately more similar to that of the major industrial countries, while the structure of their manufacturing value added became less similar. In sum, available evidence suggests a strong divergence in the evolution of international specialization between Asian and Latin American developing countries. The Republic of Korea and Taiwan Province of China have gained RCA in medium- and high-technology manufactures, and the production structures of these economies have become significantly similar to the economies of the major industrial countries, both in production and exports of manufactures. In Malaysia and Mexico, the pattern of specialization has moved toward the assembly of computers and office equipment, communications equipment, computers, and (particularly in Mexico) automobiles. The processing of natural resources has come to dominate production and export activities in Argentina, Brazil, and Chile, although the automobile and aerospace industries have gained in importance in Argentina and Brazil. Taken together, the evidence suggests that among the major developing countries, only the first-tier East Asian NIEs have succeeded in simultaneously upgrading their production and export structures. By contrast, in other countries the change in the pattern of specialization of production has not involved a shift toward high-value-added activities.
5
Conclusions
The analysis in this chapter suggests that regarding the process of accumulation, industrialization, trade, and structural change, it is possible to distinguish between five broad categories of economies: ●
●
A first group includes the first-tier NIEs, notably the Republic of Korea and Taiwan Province of China, which have already achieved a considerable degree of industrial maturity through a rapid accumulation of capital, and growth in industrial employment, productivity, and output, as well as in manufactured exports. In both economies the share of industrial output is well above the levels of advanced industrial countries, but the pace of expansion of production capacity and output in the industrial sector has slowed down compared to previous decades. A second group consists of countries that are progressing rapidly in industrialization. They are increasing the share of manufacturing in employment, output, and exports, and upgrading from resource-based and laborintensive products to medium- and high-tech products in both output and trade. These include the dynamic second-tier NIEs, notably Malaysia
54 Yilmaz Akyüz, Richard Kozul-Wright, and Jörg Mayer
●
●
●
and Thailand. China and, to a lesser extent, India should also be considered in this group of rapid industrializers, even though they are at earlier stages of industrialization compared to the second-tier NIEs. A third group comprises countries that have rapidly integrated into international production networks by focusing on simple assembly operations in labor-intensive manufactures. These countries have seen a sharp rise in industrial employment and manufactured exports, but their performance in terms of investment, manufacturing value added and productivity growth, as well as overall economic growth, has been poor. Two countries that stand out in this group are Mexico and the Philippines. A fourth group comprises countries that have reached a certain level of industrialization, but have been unable to sustain a dynamic process of industrial deepening in the context of rapid growth. These include Brazil and Argentina, where investment performance has been poor, industry has been losing its relative importance in total employment and value added, productivity growth has been cyclical (resulting from laborshedding rather than faster accumulation and technical progress), industrial upgrading has been limited, and exports have continued to be dominated by primary products and low-value-added manufactures. In these countries, progress achieved in certain industries such as aerospace and automobiles, has not gone deep enough to establish a dynamic momentum in industry. Many African countries are also in this group in terms of sluggish progress in their industrialization and structural change, even though they are at a much lower level of industrial development. A final category consists of countries that have achieved sustained and strong growth by intensifying exploitation of their rich natural resources through a rapid pace of capital accumulation. However, their industrial performance has been weak both in terms of manufacturing value added and exports, and prospects for further structural change and productivity growth appear to be limited. The most outstanding example is Chile.
Countries in any one of these groups may also manifest characteristics of those belonging to the other groups. For instance, China and Malaysia have also expanded their manufactured exports much faster than value added by participating in international production networks, but unlike Mexico, their investment and growth performance has been impressive. This explains why manufacturing productivity has been growing much faster, and the share of manufacturing value added in GDP has been stable (China) or rising (Malaysia). There are also borderline cases between the second group of rapid industrializers and the fourth group of “laggards.” For instance, Turkey is closer to the former, while Colombia is closer to the latter group. The evidence examined here shows that Latin American countries have generally been unable to remove structural and institutional impediments to rapid and sustained accumulation, growth, and structural change, despite
Trade and Industrial Upgrading 55
drastic changes to their development strategy introduced in response to the debt crisis of the 1980s. This persistent and rapid deterioration in the position of the region relative to both the industrialized countries and the successful developing economies of East Asia raises serious doubts about the underlying logic of the policy approach based on the Washington Consensus, that an import-substitution growth strategy could effectively be replaced by an outward-oriented strategy simply by eliminating inflation and opening up markets to foreign trade and investment flows.
Notes 1. The work for this paper was originally carried out for chapter V of UNCTAD’s Trade and Development Report 2003. The authors are grateful to H. Flassbeck, D. Kotte, G. Palma, and M. Shafaeddin for various comments and suggestions and to E. Borneck, J. Pizarro, and A. Rapin for statistical assistance. 2. The exact timing and pace of this shift in different developed countries is influenced by policy choice as well as supply shocks, such as the sudden discovery of raw materials, which can accelerate the shift away from manufacturing employment. This happened in the Netherlands, following the discovery of natural gas (hence the term “Dutch disease” sometimes used to describe this process), and in some favorably located smaller European economies, thanks to a sharp rise in earnings from tourism and financial services. By contrast, trade has helped sustain industrial employment in Japan, despite relatively stagnant domestic markets. 3. For a comparison with industrial countries, see Arrighi, Silver, and Brewer (2003). 4. When measured at constant prices, the share of gross fixed capital formation (GFCF) in GDP rose, on average, by less than one percentage point during the 1990s compared to the 1980s while the share of manufactured value added rose by some 1.5 percentage points. Applying the same price index to manufactured exports as to value added, the share of such exports in GDP measured at constant prices would show an increase of 12.7 percentage points during the same period, that is, 8.5 times the increase in the share of manufacturing value added in GDP. 5. These results are based on calculating the growth rate of real manufacturing value added for every percentage point of structural change, using a structural change index from Moreno-Brid (1999). 6. See, for example, Dornbusch et al. (1977) and Gomory and Baumol (2000). 7. A switchover in the structure of comparative advantage can also occur in a Heckscher–Ohlin model when a country changes its endowment structure faster than others. However, this appears to be empirically less relevant, given that the relative position of country groups with respect to their endowments in human capital, natural resources, and labor has changed little over the past 40 years, as discussed in UNCTAD (1998, p. 186). 8. Figures in Table 2.7 do not indicate relative competitive positions of the manufacturing sectors of the countries concerned (as the positions also depend on the situation in the base year of the index), but simply the direction of change in each country. 9. These results, based on labor productivity measured in dollars, differ from the index numbers given in Table 2.4 based on national currencies. The difference is small for most countries except Bolivia, China, and Ecuador.
56 Yilmaz Akyüz, Richard Kozul-Wright, and Jörg Mayer 10. For further discussion of these categories, see UNCTAD (2002). For a similar analysis, focusing on Latin America, see Katz and Stumpo (2000). 11. This analysis of export patterns is based on a revised version of the Balassa index of RCA; the approach follows Proudman and Redding (2000); and Redding (2002). The measure of RCA used by these authors evaluates an economy’s export share in a given sector relative to its average export share in all manufacturing sectors, rather than to the weighted sum of export shares in all manufacturing sectors. For the advantages of this modification, see Proudman and Redding (2000, p. 394). 12. Industrial policy in the automobile sector has been closely linked to regional policies in the context of the Southern Common Market (MERCOSUR). Thus part of the strong export performance of the automobile sector in both countries is likely to reflect intra-industry trade between Argentina and Brazil (UNCTAD, 2002), while the export success of Brazil’s aerospace industry owes a great deal to Embraer’s move in the mid-1990s into what was then a niche market for civilian regional jets (Goldstein, 2002). 13. It is clear that a specific index of similarity to a leading developed country does not have the same significance for all developing economies. Indices with respect to Germany, which can be taken as a proxy for indices with the European Union, are likely to be more important for Turkey, while indices with the United States are of overriding importance to Latin America.
References Abramovitz, M. (1986) “Catching-up, forging ahead, and falling behind,” Journal of Economic History, 46(2). Albala-Bertrand, J. M. (1999) “Industrial interdependence change in Chile: 1960–90 a comparison with Taiwan and South Korea,” International Review of Applied Economics, 13 (2). Amsden, A. (2001) “The Rise of ‘the Rest’: Challenges to the West from Late-Industrializing Economies” Oxford: Oxford University Press. Arrighi, G., B. J. Silver, and B. D. Brewer (2003) “Industrial convergence, globalization, and the persistence of the North–South divide,” Studies in Comparative International Development, 38(1). Cimoli, M. and J. Katz (2001) “Structural reforms, technological gaps and economic development,” Mimeo. http://www.druid.dk/conferences/nw/paper1/cimoli-katz.pdf. Dornbusch, R., S. Fischer, and P. Samuelson (1977) “Comparative advantage, trade, and payments in a Ricardian model with a continuum of goods,” American Economic Review, 86(5). ECLAC (2001) Investment and Economic Reform in Latin America. Santiago, Chile: Economic Commission for Latin America and the Caribbean. Gerschenkron, A. (1962) Economic Backwardness in Historical Perspective, Cambridge, MA: Belknap Press of Harvard University Press. Goldstein, A. (2002) “The Political Economy of High-Tech Industries in Developing Countries: Aerospace in Brazil, Indonesia and South Africa,” Cambridge Journal of Economics, 26(4). Gomory, R. E. and W. J. Baumol (2000) Global Trade and Conflicting National Interests, Cambridge, MA and London: MIT Press. Hirschman, A. (1968) “The political economy of import-substituting industrialization in Latin America,” Quarterly Journal of Economics, February. International Labour Organization (ILO) (2001) “Trade liberalization and employment,” (GB.282/WP/SGD/2). Geneva, November.
Trade and Industrial Upgrading 57 Kaldor, N. (1967) Strategic Factors in Economic Development, Ithaca, NY: Cornell University, W. F. Humphrey Press. Katz, J. and G. Stumpo (2001) “Sectoral regimes, productivity and international competitiveness,” CEPAL Review, 75, Santiago, Chile: Economic Commission for Latin America and the Caribbean. Krugman, P. (1990) “A ‘technology gap’ model of international trade,” in P. Krugman (ed.), Rethinking International Trade. Cambridge, MA and London: MIT Press. Krugman, P. (1991) Geography and Trade, Cambridge, MA: MIT Press. Krugman, P. (1994) “Competitiveness: a dangerous obsession,” Foreign Affairs, 73, March–April. Kuznets, S. (1955) “Economic growth and income inequality,” American Economic Review, 45. Lucas, R. (1993) “Making a miracle,” Econometrica, 61. Máttar, J., J. Moreno-Brid and W. Peres (2002) “Foreign investment in Mexico after economic reform,” Serie Estudios y Perspectivas, 10, México: DF, Economic Commision for Latin America and the Caribbean. Moreno-Brid, J. (1999) “Reformas macroeconómicas e inversión manufacturera en México,” Serie Reformas Económicas, 47. Santiago, Chile: Economic Commission for Latin America and the Caribbean. Nelson R. R. and H. Pack (1999) “The Asian miracle and modern growth theory,” Economic Journal, 109. Nelson R. R. and S. G. Winter (1982) An Evolutionary Theory of Economic Change, Cambridge, MA: Harvard University Press. Nicita A. and M. Olarreaga (2001) “Trade and production, 1976–99,” Working Paper, no. 2701, Washington, DC: World Bank. Pieper U. (2000) “Deindustrialisation and the Social and Economic Sustainability Nexus in Developing Countries: Cross-Country Evidence on Productivity and Employment,” Journal of Development Studies, 36. Proudman J. and S. Redding (2000) “Evolving Patterns of International Trade,” Review of International Economics, 8(3). Redding S. (2002) “Specialisation Dynamics,” Journal of International Economics, 58. Ros J. (2000) Development Theory and the Economics of Growth, Ann Arbor, MI: University of Michigan. Rowthorn R. and R. Ramaswamy (1999) “Growth, Trade and Deindustrialization,” IMF Staff Papers, 46, Washington, DC: International Monetary Fund. Rowthorn R. and J. R. Wells (1987) De-Industrialization and Foreign Trade, Cambridge: Cambridge University Press. UNCTAD (2001) “Globalization and the Labour Market” (UNCTAD/GDS/MDPB/ Misc.14). Paper prepared by the Secretariat for the meeting of the ILO Working Party on the Social Dimension of Globalization, 12 November 2001. Geneva, October. UNCTAD (various issues) Trade and Development Report. New York and Geneva: United Nations publication.
3 Upgrading, Uneven Development, and Jobs in the North American Apparel Industry* Jennifer Bair and Gary Gereffi
Globalization is changing the nature of work and business in the contemporary world economy. Debate today revolves around the implications of globalization’s transformative influence for firms and workers, particularly in the developing world. While capital is increasingly mobile, workers remain relatively place bound, and this tension between the global and the local demands new tools for policymakers studying labor issues, as well as new strategies for labor activists. The International Labor Office has focused on the relationship between globalization and employment in numerous studies, which explore a range of issues from working conditions in maquiladoras to the impact of information and communication technologies on the quantity, quality, and location of jobs (see ILO, 2001). Many of these studies have focused on the cross-border production and trade networks that are at the heart of economic globalization, asking about the impact of these networks in the communities where they touch down. The consensus that emerges from this literature is that cross-border networks can have positive as well as negative developmental consequences: “Globalization in a regional framework can boost development opportunities, but it may also undermine established local networks of backward and forward linkages” (ILO, 1996: 120). This chapter offers a theoretical and empirical assessment of the developmental consequences of globalization. We contend that the global commodity chains framework, which provides a network perspective on the production and distribution of goods and services in the global economy, is a useful tool for analyzing globalization’s implications at multiple levels of analysis, including for local communities and workers. Our discussion focuses on the apparel industry, one of the oldest and most global export industries in the world (Dickerson, 1999). Garment making is the typical “starter” industry for countries engaged in export-oriented industrialization, and as has been amply documented, it played the leading role in East Asia’s early export growth (Bonacich et al., 1994; Gereffi, 1998, 1999). The East Asian experience is notable not only for the importance of 58
North American Apparel Industry 59
clothes in the export profile of newly industrializing economies such as Hong Kong and Korea, but also because the region is home to countries that have managed to parlay their export activities into upgrading trajectories. This upgrading dynamic reflects a shift between two different production systems that characterize export-oriented production in the contemporary world economy: the assembly model of industrial subcontracting and the full-package model of commercial subcontracting. In this article, the recent emergence of full-package production in the Mexican apparel industry is seen as part of a wider transformation of the North American apparel commodity chain. This process produces winners as well as losers and it creates uneven upgrading opportunities benefiting some firms and workers more than others. Uneven outcomes reflect the ways in which particular communities become linked to international production networks, and the impact of the various institutional environments (supranational, national, and local) in which they are embedded. Section 1 introduces the concept of the apparel commodity chain and offers a brief typology of its lead firms or chain drivers. In the Section 2, we discuss the concept of industrial upgrading and provide a description of the main export roles characterizing apparel-producing countries in the global apparel industry. Section 3 focuses on the regional dynamics resulting from the North American Free Trade Agreement (NAFTA). We contrast the Mexican experience with that of the region’s other dynamic apparel exporters, the countries of the Caribbean Basin, to show the impact of trade regimes on export-oriented development trajectories. In Section 4, we explore the unevenness of the upgrading process within North America through a discussion of two blue jeans manufacturing clusters in the United States and Mexico: El Paso and Torreon. In the final section we summarize our findings regarding upgrading through networks in the North American apparel industry and we conclude by underscoring the importance of linking the commodity chains perspective with territorially based approaches that emphasize social and institutional contexts in order to advance the debate about globalization and its consequences.
1
The apparel commodity chain
The rise of globalization as a central concern of contemporary social science has generated a vast literature regarding its implications for development. Among the key issues that have been posed about the relationship between globalization and development is that of scale: At what level of analysis should we focus our efforts to understand contemporary processes of development, and at what level should policy interventions be directed? While some scholars focus on the rise of supranational institutions and processes that challenge traditional constructions of sovereignty and transform the role of states (Sassen, 1996), contributors to the so-called new regionalism
60 Jennifer Bair and Gary Gereffi
underscore the renewed importance of the sub-national locality as a space of development (Storper, 1997; Amin, 1999). While disagreements in this literature abound, one can identify an emerging consensus that globalization has led to what Peter Dicken and his colleagues call a “relativization of scale.” They argue that “a distinctive feature of contemporary capitalism is the ability to operate on multiple scales, but none of these scales should, in themselves, be considered a privileged level of analysis” (Dicken et al., 2001: 95). These authors conclude that the study of globalization requires a grounded, network-based approach capable of ranging across multiple spatial scales while elucidating the dynamic relationships between them. We believe that the global commodity chains (GCC) framework contributes to the development of such an approach. It emphasizes the organizational dynamics of contemporary capitalism – that is, the role of firms in constructing transnational networks for the production and distribution of goods and services. While commodity chains are often global in the sense that they traverse national borders and incorporate firms and workers in several different countries, they are also local because particular links of the chain are rooted in distinct communities. The chain approach illuminates the flows of capital, goods, services, and labor between and across space, the implications of these flows for the various places that are incorporated into global chains, and the ways in which local contexts shape and mediate these networks, even as they are, in turn, transformed by them. Commodity chains have four characteristics: (1) an input–output structure, which describes the process of transforming raw materials into final products; (2) a territoriality, or spatial dispersion of the activities involved in this transformation; (3) a governance structure, which describes the power relations that are exercised along and through the chain; and (4) an institutional context that shapes the inter-firm networks that connect the various links in the chain and mediate the outcomes associated with the operation of the chain in different environments (Gereffi, 1995). The governance dimension is particularly critical for understanding the potential of the GCC approach to shed light on the distributional consequences of globalization by highlighting the ways in which firms and workers in developing economies become incorporated into the global economy. In his work, Gereffi (1994, 1999) identifies two ideal types of governance structures in global commodity chains, which correspond to different organizational forms of international economic coordination. Producer-driven commodity chains are established by industrial capital in sectors such as automobiles, aircraft, and heavy machinery, while buyer-driven commodity chains are controlled by commercial capital in industries such as apparel, footwear, and toys. Producer-driven chains are typical of capital- and technology-intensive industries in which large, usually transnational, manufacturers play the central roles in coordinating production networks. Buyer-driven chains, in contrast, refer to those industries in which
North American Apparel Industry 61
retailers, marketers, and branded manufacturers play the pivotal roles in setting up decentralized production networks.1 The distinction between producer- and buyer-driven chains recognizes that there are various forms of coordination and control in global industries, depending on the type of firm that is dominant. The task for commodity chain analysis lies in identifying the lead firms in a particular industry, whose strategic position in the chain allows them to create and appropriate higher returns relative to other chain participants (Gereffi, 2001). The power of lead firms and their ability to control actors at other segments of the chain are derived from their ability to construct and reproduce various barriers to entry at the segments of the chain they occupy (Kaplinsky, 1998). The forms of governance in particular chains affect the distribution of profits and costs among firms and workers that participate in these economic networks: “What it [the concept of governance] does is throw light on those factors determining the nature of the insertion of different producers into the global division of labor. … [I]t is not just a matter of whether producers participate in the global economy which determines their returns to production, but how and on what terms they do so” (Kaplinsky, 2000: 12). Through empirical analysis of global industries, the GCC framework makes possible an understanding of how local developmental outcomes are affected by the networks through which particular firms (and thus the workers they employ, and the communities in which they are located) become incorporated into cross-border chains. The buyer-driven designation aptly captures the dynamics of the global apparel industry, which has been transformed in recent decades by the dramatic increase in offshore production by domestic firms, the emergence of regional production blocs, the phase-out of the Multi-Fiber Arrangement and its replacement with the Agreement on Textiles and Clothing, the prominence of branding as a key strategy, and a dramatic consolidation at the retail end of the chain (Taplin, 1994; Jones, 1998; Klein, 1999). The last two trends capture the growing importance of the organizational buyers or lead firms in this industry. One of the main characteristics of many lead firms that fit the buyer-driven model, including retailers like Wal-Mart, Sears Roebuck, and JC Penney, athletic footwear companies like Nike and Reebok, and fashion-oriented apparel companies like Liz Claiborne, Gap, and The Limited, is that these companies design and/or market – but do not make – the branded products they order. They are part of a new breed of “manufacturers without factories” that separate the physical production of goods from the design and marketing stages of the production process. Profits in buyerdriven chains derive not from scale, volume, and technological advances, as in producer-driven chains, but rather from unique combinations of highvalue research, design, sales, marketing, and financial services that allow the retailers, designers, and marketers to act as strategic brokers in linking overseas factories and traders with evolving product niches in their main consumer markets (Gereffi, 1994).
62 Jennifer Bair and Gary Gereffi
Diversity of lead firms There are three types of lead firms in the apparel commodity chain: retailers, marketers, and branded manufacturers (Gereffi, 2001). As apparel production has become globally dispersed and the competition between these firms intensifies, each type has developed extensive global sourcing capabilities. While moving out of production, they are fortifying their activities in the high value-added design and marketing segments of the apparel chain, leading to a blurring of the boundaries between these firms and a realignment of interests within the chain. Retailers. In the past, retailers generally purchased the merchandise produced and marketed by apparel manufacturers, but the relationship between retailers and apparel manufacturers has been transformed by the advent of “lean retailing” and the rise of “private label” or store brands, which are sold alongside competing national brands (Abernathy et al., 1999). For example, retailer JC Penney now offers its customers a choice between traditional brands of jeans, such as Levis or Wrangler, and its own highly successful private label, called Arizona Jeans. Almost a third of the women’s apparel sold in the United States today consists of private label products. Marketers. A notable feature of buyer-driven chains has been the creation since the mid-1970s of prominent marketers with well-known brand names, but that carry out no production in their own factories. They include companies like Liz Claiborne, Nike, and Reebok, which were “born global” since their sourcing has always been overseas. As pioneers in global sourcing, branded marketers were instrumental in providing overseas suppliers with knowledge that subsequently allowed them to upgrade their position in the apparel chain. Branded manufacturers. From a GCC perspective, the main significance of branded manufacturers is that they generally coordinate industrial subcontracting networks, while retailers and marketers coordinate commercial subcontracting networks. This set of firms includes companies such as Sara Lee (which owns the Hanes and L’Eggs brands), Levi Strauss and Company, and VF Corporation (which markets Wrangler and Lee jeans). Branded manufacturers have tended to rely heavily on offshore production, often in owned-and-operated plants. However, many branded manufacturers are abandoning production altogether, suggesting that their strategies will become more similar to those employed by marketers and retailers. Identifying the various lead firms that drive the apparel commodity chain is important because of their role in establishing and coordinating the international production and sourcing networks that are the infrastructure of global apparel trade. The commodity chains perspective directs our attention to the relationship between the type of lead firm, the kind of networks the
North American Apparel Industry 63
firm establishes with suppliers and contractors, and the prospects for upgrading associated with these networks. In the next section, we discuss these upgrading prospects in terms of the main export roles that characterize the global apparel industry.
2
Industrial upgrading as export role shifts
Industrial upgrading, as we define it here, entails moving to higher-value activities in global supply chains. For the apparel industry, industrial upgrading can be conceptualized as a series of role shifts involved in moving from export-oriented assembly to more integrated forms of manufacturing and marketing associated with the original equipment manufacturing (OEM) and original brand name manufacturing (OBM) export roles, respectively (Gereffi, 1999). Assembly manufacturing teaches apparel exporters about the price, quality, and delivery standards used in global markets.2 The most typical upgrading move following assembly is OEM, which involves the devolution of a greater range of activities from the lead firm to the contract manufacturer that receives the order. The OEM role has emerged across a variety of industries, including electronics, automobiles, pharmaceuticals, and apparel. While the terminology may be slightly different across industries, in each case the contract manufacturers “provide complete bundles of manufacturing-related services … including investment in production facilities (in both domestic and international settings), component and material sourcing, the manufacturing process itself, quality assurance, in-bound and out-bound logistics, etc” (Sturgeon, 2001: 15). In the apparel industry, OEM is referred to as the full-package model, and the upgrading consequences of the shift from the assembly to the full-package role can be noted at both the level of the individual firms and at the level of the exporting region. At the level of the firm, full-package can be considered a form of upgrading because it expands a producer’s customer base beyond the branded manufacturers that have traditionally favored industrial subcontracting (or assembly) networks in order to include the retailers and marketers that prefer commercial subcontracting (or full-package) networks. Compared with the mere assembly of imported inputs, full-package production changes the relationship between buyer and supplier in a direction that, at least potentially, gives more autonomy and opportunities for learning to the supplier. Full-package production is needed because the retailers and marketers that order garments typically have little interest or experience in manufacturing apparel. Thus, the suppliers must learn how to do everything, and this process of learning often takes place in the context of a relatively long-term relationship with the buyers. The more stable and open the relationship between the client and the supplier, the more favorable is the environment for observing and learning from the buyer.3 At the level of the exporting region or cluster, full-package production can be considered a form of upgrading when it stimulates linkages between
64 Jennifer Bair and Gary Gereffi
different segments of the apparel commodity chain. Unlike assembly networks, which usually require imported inputs, full-package networks provide opportunities for exporting firms to find local suppliers for materials such as fabric, buttons, and thread. One of the main criticisms leveled at exportprocessing zones is that they are islands of assembly production that are essentially unconnected to the domestic economies of the countries that host them. While export-processing zones generate foreign exchange through exports and they also create jobs, the absence of linkages to local firms means that they generate minimal growth beyond the export enclave. Because fullpackage production is more likely to stimulate local linkages to component suppliers, it may be considered a form of intrasectoral upgrading. The term “full-package production” refers to any production arrangement between a client (i.e. the buyer) and a contractor (i.e. the manufacturer), whereby the contractor receiving the order is responsible for purchasing the raw materials (e.g. thread, fabric) and coordinating all the different parts of the production process. The full-package manufacturer is usually a textile or apparel producer, though theoretically, a broker with no production capacity can also be considered a full-package “manufacturer” if he buys the raw materials and coordinates the various production tasks that are outsourced to subcontractors. Full-package production therefore does not necessarily imply a more integrated form of manufacturing with backward linkages to local suppliers (Schrank, 2002), though we argue that full-package production is far more likely to generate such linkages than the export-oriented assembly model, which virtually always entails the importation of foreign-made components. There are five defining elements in this historically and organizationally grounded GCC approach to industrial upgrading. First, sequences of export roles are contingent, not invariant, features of industrial upgrading. While the progression from assembly to OEM to OBM export roles is quite typical, success in one role does not guarantee success in subsequent ones. Backsliding is possible and the sequences may vary, especially for more advanced forms of upgrading. Nor are these export roles mutually exclusive. In fact, most nations and many firms are tied to the world economy in multiple ways. Second, industrial upgrading involves organizational learning in global supply chains to improve the position of firms or regions in international trade and production networks. Participation in GCCs is important for industrial upgrading because it puts firms and economies on potentially dynamic learning curves. However, there are many obstacles to moving up these chains, and the barriers to entry for each export role typically are more demanding as one moves along the upgrading trajectory. Third, industrial upgrading requires not only physical and human capital, but also social capital – that is, relevant and effective networks. Economic theories of upgrading indicate that as capital (both physical and human) becomes
North American Apparel Industry 65
more abundant relative to labor and the endowments of other countries, nations develop comparative advantages in capital- and skill-intensive industries. Industrial upgrading within the apparel commodity chain involves building and coordinating networks with different kinds of lead firms that have access to distinct pools of design, production, and marketing resources needed to create new forms of competitive advantage. Fourth, sustaining the upgrading process within a particular commodity chain involves both forward and backward linkages from production, and the kind of learning that occurs across these segments. There are various ways that industrial upgrading can proceed once the capabilities for integrated manufacturing required by the OEM role have been mastered, whether by individual companies or by networks of firms. One upgrading option is to move forward along the supply chain from production to marketing. For example, Hong Kong apparel companies have gone from OEM to OBM by establishing new retail chains featuring their brands. US apparel giants like Levi Strauss and Company and Sara Lee have chosen to lessen their commitment to manufacturing in order to put more resources into building global brands, which are the most profitable part of the softgoods value chain, while textile manufacturers have integrated into apparel supply precisely to enhance their manufacturing capabilities and enlarge their potential customer base. Fifth, the upgrading processes of firms in terms of shifts along or between commodity chains is an important, but not a sufficient, condition for ensuring positive development outcomes. The upgrading process is often uneven across regions within one country and between firms within the same cluster. In order to highlight upgrading dynamics and their consequences, particularly for workers, one must move from a discussion of upgrading at the industry level to an examination of how particular places become incorporated into such chains, and how this process is mediated by the institutional and politico-economic characteristics of the environments in which firms and workers are embedded. In the following section, we focus on the role of regional trade regimes in shaping the upgrading prospects of textile and apparel exporters in two different parts of North America: Mexico and the Caribbean Basin.
3
Upgrading in the North American apparel industry
No industry better captures the development dilemmas that shape North America’s regional integration efforts in the past decade than apparel. Apparel shipments from Mexico and the Caribbean Basin countries to the United States are pacing the boom in manufactured exports from these economies, generating both jobs and foreign exchange. Yet complaints abound about the quality of these jobs, the stability of the export earnings, and the declining standard of living confronted by workers. A central question in the North American integration debate is whether NAFTA should be considered a good deal or not, and for whom. Critics of
66 Jennifer Bair and Gary Gereffi
NAFTA in the United States claim that it has escalated the destruction of the US manufacturing base, and they bolster this view with evidence from the apparel industry, which accounts for 30 percent of all NAFTA-related job losses registered by the US Department of Labor between 1994 and 1999 (Spener and Capps, 2001). Between 1985 and 2000, employment in the US apparel industry fell from 1.12 million to 633 000 workers. The number of US apparel jobs lost between NAFTA’s implementation in 1994 and 2000 (340 800) was more than twice as great as the number lost between 1985 and 1994 (146 000). In fact, the decline in US apparel employment since NAFTA is very close in magnitude to the number of apparel jobs created in Mexico (325 700) over the same period (see Figure 3.1). These aggregate statistics make it look as though NAFTA has created a zero-sum game between the United States and Mexico.4 Actually, the story of NAFTA’s winners and losers in textiles and apparel is far more complex than these numbers reveal. A disproportionate share of job losses in the US apparel industry has been borne by southeastern states, such as North and South Carolina, and Georgia – places that specialize in the mass production of standardized garments like blue jeans and underwear that Mexico has been exporting in large volumes since NAFTA went into effect.5 However, these same states have big textile industries, and textile firms
1 500 1 357 6.6%
Total employment (000s)
1 200
10.8%
1 392 4.6% 15.9%
1 288 6.5% 18.0%
1 366 6.4%
33.3% 43.4%
900
600
1 284 7.3%
82.6%
79.5% 75.6% 60.3% 49.3%
300
0
1985
1991 Canada
1994 Mexico
1997
2000
United States
Figure 3.1 North American apparel employment, 1985–2000 Sources: Statistics Canada, Employment, Earning and Hours; US Bureau of Labor Statistics; Censos Económicos.
North American Apparel Industry 67
supported NAFTA as a defensive maneuver to protect the North American market against a flood of cheap Asian exports. Carlos Moore, president of the American Textile Manufacturers Institute, concludes that NAFTA has indeed benefited the US textile industry and its workers: “Simply put, apparel imports from Mexico help our industry and our workers; apparel imports from the Far East hurt us … [I]f we didn’t have NAFTA, job losses in the textile industry would have been far more drastic because United States garment-making would have continued to move to the Far East and we would not have nearby markets for our textiles” (Moore, 1999). This account suggests that the winners from NAFTA are US textile companies and the Mexican apparel manufacturers that are sewing US fabrics into garments. Indeed, since NAFTA went into effect on January 1, 1994, Mexico has emerged to challenge China as the top US apparel supplier of apparel.6 However, free trade has proven a lot tougher for North Carolina textile companies than Moore anticipated. Strong supporters of NAFTA, such as Burlington Industries, Galey & Lord, Guilford Mills, and Malden Mills, have all filed for Chapter 11 bankruptcy protection in 2001 and 2002, blaming their misfortunes on factors such as the Asian currency devaluations, the strong US dollar, a glut of textile products on the world market, and China’s transshipments of textile goods through Mexico to evade US quotas7 (ATMI, 2001; Neff et al., 2002). Overwhelming debt has pushed several high profile apparel firms into Chapter 11 bankruptcy as well, including Warnaco Group, the manufacturer and distributor of Calvin Klein jeans and Speedo swimsuits, and the underwear giant, Fruit of the Loom. In response to the textile crisis, a new lobbying organization spearheaded by Roger Millikin, head of one of the Carolinas’ largest textile companies, Millikin & Co., and UNITE (Union of Needletrades, Industrial and Textile Employees) President Bruce Raynor, was launched in March 2002 to take a more aggressive protectionist stance on textile trade issues. The new group, known as the American Textile Trade Action Coalition (ATTAC), is protesting the loss of about 180 000 textile jobs, nearly one-third of the industry’s total workforce, and the shuttering of at least 220 textile plants in the past five years (Mecia, 2002). The move is a throwback to the mid-and late 1980s, when a coalition of labor unions, textile mills, apparel firms, cotton growers, and chemical companies pushed three bills through the US Congress that would have capped the growth of textile imports to 1 percent per year. These efforts by the textile lobby fell short, with Presidents Reagan and Bush vetoing the protectionist trade bills in 1985, 1988, and 1990. Of course, NAFTA and subsequent trade deals alone cannot be blamed for US job losses in labor-intensive industries such as apparel. Companies in the United States have been investing overseas for decades in order to cut costs. But free-trade agreements do accelerate the trend toward global sourcing, and the pain is distributed unevenly. Only one-fifth of the 50 000 North Carolinians displaced by trade-related layoffs since 1998, when the state
68 Jennifer Bair and Gary Gereffi
began keeping count, enrolled in worker-retraining programs that they are entitled to through NAFTA. Less than one quarter of the 10 000-plus workers in North Carolina who completed retaining programs landed the kind of jobs they trained for, and those experiencing trade-related layoffs who did find a job earned 12 percent less than before the layoff (Rives and Neff, 2002). Mexico and the Caribbean Basin: the maquila model Within Latin America, the labor-intensive assembly of manufactured goods from imported components is disproportionately concentrated in Mexico (whose in-bond factories are called maquiladoras) and the Caribbean Basin (where assembly production is usually carried out in Free Trade Zones) because of these countries’ low wages and proximity to the US market. The US market dominates Mexico and the Caribbean Basin’s export profile and is the destination for over 90 percent of their apparel exports. Mexico’s maquiladora industry, which was established in 1965 by the Border Industrialization Program, is made up of assembly plants that use imported components to make goods for export to the US market. Historically, Mexico’s maquiladora plants typified low value-added assembly, with virtually no backward linkages. In the 1980s, a second generation of maquiladora plants began to push beyond this enclave model to a more advanced type of production, making components for complex products like automobiles and computers. By the mid-1990s, a third generation of maquiladoras was said to be emerging (although this claim was based largely on a case study of Delphi, a General Motors plant), signaling a move beyond manufacturing to involve activities such as research and development and product design (Carrillo and Hualde, 1998). The maquila sector in Mexico has expanded dramatically since NAFTA. Maquiladora employment more than doubled between 1990 and 2000 from 439 000 to 1.3 million workers. In 1993, the year prior to NAFTA, there were 400 apparel maquiladoras employing 66 000 workers; by 2000, 1120 apparel maquiladoras provided jobs to nearly 290 000 Mexicans (Bair and Gereffi, 2002: 33). This expansion mirrors the extraordinary export dynamism of Mexico’s apparel industry in the post-NAFTA period, which is shown in Table 3.1. Mexico’s total apparel exports increased from $1.9 billion in 1994 to $8.1 billion in 2001 and the vast majority of this growth occurred after NAFTA’s implementation in 1994. As the last column of Table 3.1 reveals, 807/9802 trade (linked to the US production-sharing program) has historically dominated the US import profile in garments, accounting in 1994 for 78 percent of imports from Mexico and 80 percent of those from Central America and the Caribbean. Because NAFTA has reduced the incentive of Mexican companies to register their exports under the maquiladora program, we estimate that the percentage of Mexico’s exports that were fullpackage (i.e. non-807/9802 goods) in 2001 to be about one-third, instead of the one-half indicated in Table 3.1.8
North American Apparel Industry 69 Table 3.1 US apparel imports: total and 807/9802 trade, by Mexico and Caribbean Basin Initiative (CBI) countries, 1994–2000 Year
Total apparel imports (US$ millions)
807/9802 Trade (US$ millions)
807/9802 Trade as a share of total imports (percent)
World 1994 1995 1996 1997 1998 1999 2000 2001
36 878 39 438 41 679 48 287 53 874 56 376 64 181 63 789
5 707 7 631 8 719 11 322 12 791 13 474 12 953 12 273
15 19 21 23 24 24 20 19
Mexico 1994 1995 1996 1997 1998 1999 2000 2001
1 889 2 876 3 850 5 349 6 812 7 845 8 730 8 128
1 470 2 282 2 967 4 096 5 102 5 417 5 071 4 030
78 79 77 77 75 69 58 50
CBI countries 1994 1995 1996 1997 1998 1999 2000 2001a
4 538 5 487 6 076 7 664 8 349 8 889 9 702 9 602
3 617 4 497 4 999 6 411 6 929 7 301 7 181 7 602
80 82 82 84 83 82 74 79
Notes: a 2001 807/9802 imports from the CBI include apparel imported under the Caribbean Trade and Development Act (CBTDA). CBTDA import data is compiled from statistics of the Office of Textiles and Apparel (available at http://www.otexa.ita.doc.gov). Sources: Compiled from official statistcs of the US Department of Commerce, International Trade Administration, Office of Textiles and Apparel; US imports for consumption, customs value.
The development of full-package capabilities among Mexican firms has been driven by NAFTA, and specifically the Agreement’s rules of origin that govern trade in textile products. The 807/9802 regime in the United States restricts preferential trade access to the US market to goods assembled abroad from US components. NAFTA has changed the rules of the game by extending preferential access to the US (and Canadian) market to all Mexican goods that meet the rules of origin established in the Agreement. Goods are
70 Jennifer Bair and Gary Gereffi
considered as complying with NAFTA rules of origin if they contain a minimum percentage of “North American” (Canadian, US, or Mexican) content. For textile and apparel products, NAFTA enshrines a triple transformation rule (also known as “yarn forward”), which means that apparel and textile products qualify as North American as long as the yarns that they contain were made in one of the three NAFTA countries. Following NAFTA’s passage, countries in the Caribbean Basin voiced concern that their apparel exports, which were still subject to the valueadded tariff, would no longer be competitive with Mexico’s.9 When NAFTA took effect, “Mexico enjoyed the equivalent of a six-point tariff-rate advantage in the US market, was no longer subject to import quotas on many apparel items, and, most notably, could count Mexican inputs as part of the requisite NAFTA content. That gave Mexico a huge advantage compared to the Caribbean Basin countries” (Mortimore, 2002: 300). Eventually, political pressure led the US Congress to grant assemblers in the Caribbean Basin a version of “NAFTA parity” in the form of the United States–Caribbean Trade Partnership Act, which was approved in May 2000, and enacted as Title II of the Trade and Development Act of 2000 (Rodriguez-Archila, 2000). This act provides both duty- and quota-free treatment for garments exported by certain countries of Central America and the Caribbean that use fabric made in the United States. In reality, the Caribbean Basin countries have not received NAFTA parity. Although the Trade and Development Act levels the playing field between the Caribbean Basin countries and Mexico, it does not replace the old 807 rules of origin with new North American ones, and therefore, unlike NAFTA, it does not encourage the development of a local supply base for textiles and other inputs in the region. One scholar, who asked if “the Dominican Republic’s export processing zones can survive NAFTA,” concludes that the Caribbean Trade and Partnership Act’s restrictions on local value-added limit the possibilities for vertical integration of the garment industry beyond the stage of export-oriented assembly (Mathews, 2002: 316). Mexico’s ability to export clothing that is assembled from fabrics manufactured in Mexico gives it an edge over other apparel exporters, and Mexico thus figures prominently in the strategies of the lead firms that drive the North American apparel commodity chain. While the institutional context created by NAFTA sets the stage for the development of full-package production in Mexico, inter-firm networks linking the apparel industry’s lead firms to local manufacturers are the mechanism enabling the shift from the assembly to the export role. Reconfiguring the apparel supply chain and Mexico’s transition to full-package production The key to Mexico’s transition from assembly to more integrated export production are networks organized by US firms that want to increase their
North American Apparel Industry 71
security and enhance profits by coordinating the activities associated with full-package supply in North America. Large firms in different segments of the apparel chain, mainly from the United States, are vying to become coordinating agents in new North American networks that could strengthen Mexico’s capabilities to carry out full-package supply (Bair and Gereffi, 2002): ●
●
●
●
●
Synthetic fiber companies in the United States and Mexico have been lobbying down market with US apparel manufacturers and retailers, trying to get the apparel firms to develop products using their fibers and trying to get retailers to bring these orders to Mexico. Textile mills have been forging alliances with apparel suppliers that could allow for more integrated textile and apparel production in different regions of Mexico; in addition, textile firms are exploring the possibility of creating their own product development teams for select apparel categories. US branded apparel manufacturers are rationalizing their supply chains in Mexico, looking for smaller numbers of more capable suppliers, or “deverticalizing” their domestic and offshore production operations by divesting themselves of manufacturing assets in favor of building up the marketing side of their business, with an emphasis on global brands. A handful of Mexican integrated apparel manufacturers, who own modern plants that go from spinning and weaving through apparel production and finishing, have developed strong reputations with US retailers and marketers that are looking to place full-package orders in Mexico. US and Latin American retailers are beginning to set up sourcing networks in Mexico, aided by government-supported vendor certification programs.
Among the most prominent players in the post-NAFTA restructuring of the North American apparel industry are US textile companies. A prime example of this trend is North Carolina-based Burlington Industries, which grew through acquisitions to become the world’s largest textile maker by 1980. Burlington had only three plants in Mexico in 1994, two of which were for the production of cotton and synthetic fabrics. In 1998, Burlington announced that it would invest $80 million over the next three years in five apparel plants as part of its effort to develop garment services. The Casual Wear Division of Burlington launched a strategy of integrated denim apparel production in Mexico, which included an owned-and-operated sewing factory and an industrial laundry for pre-washing the jeans it produced in Mexico. One Burlington executive described this move as “one-stop shopping” for the company’s clients: “The strategy is to offer fabric in garment form as a service to branded customers, many of whom have to outsource production anyway” (Hill, 2000: 29). Burlington Industries filed for Chapter 11 bankruptcy protection in November 2001, and both the apparel plant and Burlington’s share of the laundry joint venture in the Casual Wear Division were sold as part of the
72 Jennifer Bair and Gary Gereffi
company’s restructuring effort. However, the company continues to produce fabric in Mexico and uses a number of Mexican contractors to fill its fullpackage apparel orders for brand-name clients (Bair, 2001). Guilford Mills, under the leadership of its late CEO Chuck Hayes, has also been active in Mexico. Unlike some of his counterparts in the US textile industry, Hayes believed that NAFTA provides an opportunity to strengthen the fiber–textile–apparel chain in North America, thereby repatriating textile and apparel production that had gone to Asia. Hayes was the driving force behind the creation of “NuStart,” an industrial park dedicated to garment production located outside of Cuernavaca in central Mexico. In addition to Guilford, Alpek (the petrochemical division of the Mexican conglomerate, Grupo Alfa) provided financial support for the project, and eventually DuPont and Burlington Industries also signed on. NuStart was inaugurated in July 1997, and as of June 2000, seven companies from Canada, the United States, and Mexico were operating in the park. Despite a host of problems that have plagued NuStart, Guilford Mills is heading an initiative to create a second industrial park in Tamaulipas. A Mexican textile company called Kaltex is also attempting to secure its position in the post-NAFTA apparel commodity chain. Founded in 1925, Kaltex is one of the largest companies in the Mexican textile industry. Kaltex expanded beyond its textile roots in yarn and fabric production to include garment making, including full-package apparel programs in 1994. Kaltex inaugurated its first denim mill in 1996, and it is considered one of the largest denim manufacturers in Mexico, as well as a major player in the fullpackage jeans market. Kaltex’s denim branch, Denimex, exports virtually all its denim to the United States in the form of fabric or apparel. Denimex has a close relationship with the Lee jeans company, and much of its apparel production is sold to Lee’s parent company, VF Corporation. US retailers are a growing presence in Mexico. JC Penney established a buying office in Mexico City in 1994 with the goal of sourcing apparel from Mexican manufacturers for its private label lines. In 1994, it procured $7 million of apparel in Mexico, an amount that increased to about $100 million in 1999. JC Penney sources from 22 Mexican companies, and tee-shirts, underwear, and jeans are the principal products. Some of the apparel for JC Penney’s most successful private label line, Arizona Jeans Wear, is manufactured in the northern Mexican town of Gómez Palacio by the Original Mexican Jean Company (OMJC), which is a joint venture between a US manufacturer, Aalfs, and a Mexican partner. The networks linking JC Penneys, Aalfs, OMJC, and local subcontractors in Torreon exemplify the role of interfirm networks in restructuring the North American apparel commodity chain in the context of the institutional environment created by NAFTA (Bair and Gereffi, 2001). In the next section we underscore the impact of this restructuring in particular communities, returning to the debate about who benefits and who loses as a result of NAFTA.
North American Apparel Industry 73
4 NAFTA and uneven development: the rise and fall of the blue jeans industry in El Paso and Torreon One of the most intriguing chapters in the North American apparel story is the comparison of Torreon, a city located in the cotton-growing region of La Laguna in northern Mexico, and El Paso, Texas, whose former title of the “Blue Jeans Capital of the World” has been claimed in recent years by Torreon.10 In 2000, there were about 350 apparel factories operating in Torreon, virtually all oriented to the export market. According to industry estimates, these factories made about 6 million pairs of jeans a week. The year before NAFTA’s implementation, the cluster’s total production volume was only 500 000 garments a week – a volume that is only slightly greater than the production capacity of the single largest apparel manufacturer in Torreon today. Approximately 75 000 of Torreon’s population of a half million people were employed in the apparel and textile industries in 2000, up from just 12 000 workers in 1993 (Bair and Gereffi, 2001: table 1, p. 1889). Dramatic growth in garment employment in Torreon can be contrasted with a striking decline in apparel employment in El Paso, Texas,11 where the number of apparel jobs fell by half between 1993 and 2001 (from 23 581 to 11 851). Of course, one cannot conclude that El Paso’s losses have translated directly into Torreon’s gains. In the post-NAFTA period, garment manufacturing has migrated to many low-cost locales, not just Torreon, and this process has meant a decline in apparel employment in cities across the United States, not just El Paso. However, no place has been as hard hit as El Paso, which has suffered more job losses attributable to NAFTA (as certified by the US Department of Labor under the Transitional Adjustment Assistance program) than any other American city. Though El Paso’s experience of deindustrialization is typical among America’s urban manufacturing centers, aspects of the city’s profile make it a particularly compelling illustration of the dynamics of NAFTA’s uneven development. While Mexican Americans and Mexican immigrants comprise the majority of El Paso’s overall labor force (65 percent in 1990), their dominance in the apparel industry is nearly total. Virtually all of El Paso’s garment workers are of Mexican origin, with a full two-thirds of these workers having been born in Mexico. Thus, as El Paso’s apparel factories close and a portion of these jobs are relocated to Mexico, large numbers of Mexican-born workers who migrated north in search of better employment opportunities find themselves losing jobs to compatriots south of the border. Despite the recent attention NAFTA has received as the widely perceived and most proximate reason for local job losses (Medaille and Wheat, 1997; Moreno, 1997), the fate of El Paso’s apparel industry has been linked to developments on the southern side of the US–Mexico border for more than a decade prior to NAFTA. Its trajectory underscores the importance of trade regimes in shaping commodity chains, as well as the precariousness of the
74 Jennifer Bair and Gary Gereffi
industrialization and employment patterns that result and the vulnerability of local communities to a change in the regulatory environment. El Paso’s apparel industry peaked in the early 1970s, when it employed 60 percent of all manufacturing workers and 20 percent of all private-sector workers. By 1990, these percentages had declined to 33 and 8 percent, respectively (Spener, 2002: 142). Among other factors, including labor strife following the historic strike at local manufacturer Farah in the mid-1970s, pre-NAFTA job losses were attributable to the rise of maquila production in Mexico. Several companies were drawn to Mexico following the sharp devaluations of the Mexican peso in the early 1980s, which made Mexican labor only one-tenth to one-eighth as expensive as minimum-wage labor in El Paso. However, while the expansion of maquiladora exports from Mexico contributed to the loss of local sewing jobs, the regulations governing this crossborder production sharing actually helped to stabilize El Paso’s overall apparel industry employment, which increased between 1990 and 1993. The maquiladora trade regime extended preferential access to the US market to apparel assembled in Mexico from fabrics cut and formed in the United States. In addition to cutting, post-assembly processes such as laundering had to be done in the United States. The result of this trade regime was the consolidation of El Paso as a cut-and-finish center, whose firms complemented the sewing activities carried out in Mexican maquiladoras. In addition to this 807-related employment, the early 1990s witnessed an expansion in the production operations of several large local employers. In the immediate pre-NAFTA period, “El Paso was in the midst of a boom in garment employment, with the local industry dominated by three large employers. Levi Strauss and Company, Lee, and Wrangler had, respectively, 4,600, 2,000, and 2,600 El Paso employees” (Spener, 2002: 146). When NAFTA was implemented in 1994, the complementarity between El Paso’s cutting rooms and laundries and Mexican maquiladoras that had helped sustain the local industry was destroyed. The result was a relocation of cutting and laundering operations from the southern United States to export-oriented apparel clusters in Mexico, such as Aguascalientes, Puebla, and especially Torreon. In the remainder of this section, we discuss the results of research we conducted in Torreon in the summers of 1998 and 2000. Our study of the region’s post-NAFTA export boom reveals the role of inter-firm networks in the restructuring of the North American apparel commodity chain and the uneven developmental consequences that this process entails for communities on both sides of the US–Mexico border.12 Interviews with firms and industry associations in both the United States and Mexico revealed that the networks between retail, textile, and apparel companies in Torreon and the United States are creating a dynamic apparel cluster in the area and creating upgrading opportunities for local firms. In particular, Torreon has emerged as Mexico’s leading center for exportoriented blue jeans production. Its development as the world’s new blue
North American Apparel Industry 75
jeans capital has been driven, in part, by the same firms that were once major employers in El Paso, such as Levi Strauss and Company, VF Corporation (which makes both Lee and Wrangler jeans), and Sun Apparel. Levi Strauss and Company’s presence in the region has been indirect, as a major client for a number of Torreon’s export-oriented manufacturers. Several of the 39 contractors that Levi Strauss trained and certified in Latin America between 1994 and 1998 are located in Torreon (Spener, 2002). Outsourcing to garment manufacturers (many of whom are located in Mexico, as well as the Caribbean Basin and China) has allowed Levi’s to reduce its US employment by more than one-half since NAFTA – from 36 500 in 1994 to 16 700 in 2001. The El Paso region has been reeling from Levi’s plant closings since 1997, at which point the company still employed over 17 000 workers in 31 factories in the United States. When Levi Strauss and Company announced in February 1999 that it would be closing 11 US plants and laying off 5900 workers, the move was thought to be the culmination of a series of layoffs that had left the company with only 11 plants in the United States (Emert, 1999). However, by 2002 Levi’s was operating only eight factories in the United States, and in April of that year it announced that it would close six of them as part of its effort to cut worldwide staff by an additional 20 percent. When completed, this round of lay-offs will leave Levi’s with fewer than 1 000 US employees (Bloomberg News, 2002). VF Corporation and Sun Apparel are major employers in the Torreon cluster. VF Corporation has built a factory in a new industrial park in Torreon for its Wrangler line of denimwear, where fabrics are cut and finished jeans are laundered. This facility is the hub in a “hub and spoke” system that includes sewing facilities in three nearby towns. Sun Apparel was one of the first US apparel manufacturers to develop a presence in the region, arriving in Torreon in 1988. The company owns a local maquiladora called Maquilas Pami. Since NAFTA, Sun Apparel has expanded its enormous base of subcontractors in the Torreon region and throughout Mexico. In addition, Sun Apparel, which was purchased by Jones of New York in 1998, has added cutting and laundering capabilities in Torreon. Investments like the ones made by VF Corporation and Sun Apparel in cutting and laundering facilities are upgrading Torreon’s apparel manufacturing base. In anticipation of the elimination of tariffs on the cutting of denim cloth in Mexico on January 1, 1999, many jeans manufacturers moved their cutting operations to Mexico in 1997 and 1998. Post-production links of the chain have also emerged in Torreon, most notably the laundering of jeans and other pants. Companies could begin to wash their production in local laundries in 1994, as this was one of the production processes liberalized immediately under NAFTA’s phase-in schedule. NAFTA reduced the duty on stone-washing (a common laundering process for blue jeans) from 17 percent to 0 percent, encouraging several manufacturers to establish industrial laundries in the region.
76 Jennifer Bair and Gary Gereffi
The expansion of the apparel commodity chain beyond the garment assembly link, which is the raison d’etre of the maquiladora industry, is demonstrated by the fact that four of Torreon’s ten largest firms are full-package manufacturers. In other words, all of these companies receive orders from their clients, finance the purchase of the fabrics and other materials needed, and deliver the finished product (Bair and Gereffi, 2001: table 3, p. 1895). Aside from the small number of the cluster’s largest companies that are able to finance full-package production, there is a second tier of firms in Torreon that are known as “half-package” manufacturers – that is, they carry out all the production activities (cut, sew, launder, and finish), but do not buy the fabric, usually because they have limited amounts of working capital. This process of “lengthening” the commodity chain to include segments of the apparel chain beyond assembly constitutes upgrading at the level of the local cluster. In addition, the particular firms that are making the fullpackage and half-package transitions are upgrading via the acquisition of new capabilities, which increases their competitiveness by adding greater value to the manufacturing process and providing more services to the client. There has also been growth in Torreon’s textile industry, spurred by US companies that have established manufacturing operations in the area. Among the firms making new textile investments in the Torreon region is Cone Mills, a denim manufacturer that has established a joint venture with the Mexican textile firm, Compañía Industrial de Parras. Together these two companies built Parras-Cone, a denim mill located in the town of Parras, about two hours from Torreon. Grupo Lajat, a large Mexican corporation with interests in agriculture and oil, also helped expand Torreon’s textile base. Although the company had no previous experience in making textiles, its owners believed that NAFTA would stimulate increased demand for Mexicanmade fabrics. The textile mill that they built, Textiles Lajat, began manufacturing denim in 1995. By 1998, it was producing around three million yards of denim a month, most of which was used in garments destined for the United States. Although Textiles Lajat was sold to Parras in December 1998, Grupo Lajat stays involved in the apparel business through another of its subsidiaries, Kentucky Lajat. This company began in 1995 as a joint venture with the US-based apparel firm, Kentucky Apparel. Together the two companies built a cutting room, two industrial laundries, and two finishing plants (where jeans are pressed, inspected, and packed) in the Torreon cluster, as well as six sewing factories located in small towns around Torreon. In a decided reversal of most joint ventures between US and Mexican companies, Grupo Lajat bought out its smaller US partner in July 1999. Like several of Torreon’s larger full-package manufacturers, Kentucky-Lajat counts among its present or former clients a number of the US apparel industry’s most prestigious firms, such as Gap and Tommy Hilfiger. The client profiles of local exporters such as Kentucky-Lajat indicate a shift in the customer base of Torreon’s apparel manufacturers, which is a critical feature
North American Apparel Industry 77
of the post-NAFTA upgrading process in the cluster. Prior to 1994, most of the region’s export-oriented apparel production was organized and managed by a few large US apparel companies, such as Farah and Sun Apparel, that owned and operated local maquiladoras. The two branded manufacturers that were once the largest employers in El Paso’s apparel industry – VF Corporation (which owns the Lee and Wrangler labels) and Levi Strauss and Company – have also played an important role in Torreon’s recent rise as North America’s leading blue jeans production center. Although brand-name marketers and retailers have been the driving force behind the development of international production and sourcing networks in the global apparel industry, prior to NAFTA these types of lead firms placed no orders directly in Torreon. By 2000, there were about 20 brand-name clients sourcing jeans from Torreon. These included retailers, such as K-Mart, Gap, and JC Penney, as well as marketers, such as Tommy Hilfiger and Calvin Klein (Bair and Gereffi, 2001: table 2, p. 1892). This diversification of the customer base reflects Torreon’s shift from the maquila to the full-package export role. The maquila model of export production was linked to the large US apparel manufacturers that coordinated the region’s pre-NAFTA garment production and provided the inputs for assembly. After NAFTA, retailers and marketers also arrived in the region. These new organizational buyers were looking for full-package manufacturers because they dedicate themselves to design, distribution, and marketing as opposed to production. The networks between brand-name foreign buyers and the local manufacturers that shifted from maquila to full-package production to service them are the organizational infrastructure of Torreon’s transformation from a maquila outpost to the blue jeans capital of the world. Four of Torreon’s major full-package manufacturers are Mexican-owned firms (Bair and Gereffi, 2001: 1895–6). One such company is Pafer-Huichita. This full-package manufacturer used to work as a maquiladora for foreign firms, at which time it sent its fabrics to be cut in El Paso so that the apparel it assembled in Mexico could be imported under the 807 regime. Since NAFTA’s implementation, Pafer-Huichita has opened its own cutting room in Torreon. The emergence of domestic full-package companies like PaferHuichita competing alongside US-owned contractors shows that postNAFTA, export-oriented production has produced significant opportunities for the cluster’s largest and most capable local firms. Having gained experience through maquila production for US clients and earned the trust of foreign buyers, Mexican firms are now developing direct links to export markets. These full-package firms upgrade by eliminating middlemen like a broker or trading company, which allows them to enjoy the often higher profits and relatively greater security full-package production offers as compared to maquila orders. In addition to the upgrading dynamics we were able to identify at the cluster and firm levels, the post-NAFTA arrival of new organizational buyers
78 Jennifer Bair and Gary Gereffi
may also be having an upgrading effect at the plant level. Most of the brandname buyers placing orders in Torreon inspect the local factories producing their apparel to insure that they are meeting a host of standards that apply not only to the quality of their products, but also to the quality of their workplaces. The standards range from environmental regulations to safety measures designed to reduce the risk of workplace injury. This trend reflects the anxiety generated by recent publicity surrounding several cases of labor abuses in sweatshops ranging from Los Angeles to Guatemala to Southeast Asia, which has made many US companies more cautious about producing and sourcing apparel offshore. This is particularly true for branded manufacturers such as Levi Strauss and Company, marketers like Tommy Hilfiger, and retailers like JC Penney. Because the consumer associates these companies’ products so closely with their brand names, negative publicity surrounding labor abuses in their production networks can seriously tarnish the enormous investment that carefully cultivated brand names represent. The importance branded firms attach to the preservation of their public image has become an Achilles heel of major US apparel companies, which make ideal targets for anti-sweatshop campaigns sponsored by human rights groups, student activists, and labor organizations (Gereffi et al., 2001). As a result of their heightened sensitivity to negative publicity, companies producing or sourcing apparel in places like Torreon appear to be raising the standards of sewing factories in terms of maintaining clean, well-ventilated and well-lit facilities. Equipment in the cluster’s larger, full-package factories is similar to or better than what is typically found in US plants producing for the same buyers. While this is not to suggest that labor abuses do not occur in Torreon, or that all of the companies producing in the area have facilities identical to the largest firms, we believe that Torreon’s transition from an assembly site to a dynamic manufacturing cluster is having two positive impacts on employment in the region. First, the development of new segments of the chain, such as textiles, laundering, and cutting, is bringing new types of jobs to the region that complement the growing number of sewing jobs. These new jobs include not only basic production activities such as those involved in cutting fabric, but also the supervisory and technical positions needed to maintain highly automated, capital-intensive operations like Parras-Cone. Second, there is a process of upgrading at the plant level, in terms of working conditions in the cluster’s large- and medium-sized factories, which can be traced to the arrival of prominent marketers and retailers with corporate codes of conduct that make their business with local firms at least nominally contingent on the manufacturer’s ability to meet the standards and requirements enumerated in these codes. The limitations of upgrading at the local level We have presented evidence suggesting that the emergence of full-package apparel production in Torreon is facilitating upgrading at the level of the
North American Apparel Industry 79
cluster by extending the region’s apparel commodity chain beyond apparel assembly and encouraging the development of related activities such as textile manufacturing. At the firm level, some of the cluster’s largest companies are also able to upgrade via networks connecting them to powerful foreign clients placing full-package orders. However, there are also a number of pitfalls created by the type of export-led industrialization that the Mexican apparel industry exemplifies. First, Mexico’s adoption of an export-oriented growth strategy entails a significantly heightened degree of dependence on the US market. The decline that Mexico’s export sector has experienced as a result of the economic downturn in the United States is a case in point. During the peak of the post-NAFTA boom in the late 1990s, export dynamism and employment creation in Mexico were attributed to tighter integration of the Mexican and US economies. However, the recent crisis in the country’s maquiladora industry is the opposite side of this NAFTA coin. Between October 2000 and March 2002, over 280 000 maquila jobs were lost (INEGI, 2002). Many of these jobs were in maquiladoras making garments for the US market, as suggested by the fact that Mexico’s apparel exports to the United States, which had grown by over 10 percent per annum since NAFTA, fell almost 7 percent between 2000 and 2001. The hardship caused by the decline in US demand has occasioned some Mexican apparel manufacturers to consider the possibility of exporting to Europe and South America in order to reduce their dependence on the American market (Romney, 2001). Second, even in the midst of the post-NAFTA boom, a relatively small number of firms were reaping most of the benefits generated by Torreon’s export dynamism. While the presence of both domestic and foreign manufacturers in the Torreon cluster might be interpreted as a positive development outcome, the emergence of full-package networks in Torreon is primarily benefiting a wealthy local elite whose control over the industry is strengthened by its access to the US companies placing orders in the region. Torreon’s full-package club is an exclusive group, and its membership roster to date is limited to a small set of very large and well-connected companies. The Torreon apparel cluster is criss-crossed not only by production networks and subcontracting relationships, but also by kinship networks and social relationships. The extent to which the leading firms in the local industry are related by blood or marriage is one of the most intriguing findings to emerge from our research. Of the cluster’s ten largest firms, six are owned by members of two families. Thus, Torreon’s recent export dynamism is serving not only to promote economic upgrading, but also to reinforce the stratified nature of the local industry structure. When viewed in this light, the consequences of the cluster’s export boom are not unambiguously positive. Third, the vertical structure of the cluster’s export-oriented production networks has negative consequences for the region’s smaller firms and their workers. While a few large manufacturers in Torreon are receiving fullpackage orders (either directly or indirectly) from big US buyers, these orders
80 Jennifer Bair and Gary Gereffi
are actually being filled by several tiers of contractors and subcontractors organized into hierarchical networks that are controlled by the cluster’s dominant firms. Because the buyers in the United States are benchmarking Mexican full-package manufacturers against other exporters, firms in Torreon are under pressure to reduce their production costs to a minimum so that they can offer a competitive price. They, in turn, exert pressure on their subcontractors by trying to procure assembly services for the lowest possible price per piece, and this search for ever lower costs has negative implications that cascade down the network pyramid, from manufacturer to contractor to subcontractor. The end result of this competitive dynamic is significant downward pressure on the manufacturers’ profit margins, and consequently on workers’ wages. Small firms that are incorporated at the bottom rungs of these export networks experience the highest risks, the lowest wages, and the poorest working conditions. Fourth, because Mexico is characterized by substantial inter-regional variations in terms of labor costs, rising wage rates in Torreon may well result in the relocation of apparel assembly to less expensive areas of the country. However, while the “footloose” nature of the apparel industry means production is unlikely to be permanently rooted in Torreon or any other area, the nature of the full-package production networks in Torreon makes the area less vulnerable to a total collapse if apparel assembly moves elsewhere. Unlike the basic sewing factory needed for apparel assembly, the other segments of the commodity chain that are moving to Torreon – textile mills, modern laundries, and computerized cutting rooms – represent investments that companies are less likely to walk away from when wages rise, largely because the more capital-intensive nature of these operations make labor costs less critical than other factors, such as the availability of water and the cost of electricity. Apparel is unlikely to remain a top industry in Torreon over the long term, however. All countries or subnational regions that have been successful apparel exporters have experienced a similar trajectory of growth and eventual decline when the relatively narrow window of opportunity in the industry is fully exploited. Torreon already hosts other manufacturing industries (such as auto parts and machinery). Although the apparel industry is likely to remain important for the local economy in the near future, it should not be viewed as a permanent pole of Torreon’s industrial development. A more likely trajectory for Torreon is one of intrasectoral upgrading within apparel, with the area emerging as a coordination center for full-package networks throughout Mexico and perhaps eventually even the Caribbean, or intersectoral upgrading with other manufacturing sectors emerging as the area’s leading industries. Apparel production has begun to move from the urban areas of Torreon and Gómez Palacio to the rural hinterland, partly a response to the high turnover rates that characterized the cluster’s tight labor market in the late
North American Apparel Industry 81
1990s.13 Employers in the Torreon area that we interviewed identified excessive turnover as their most serious problem. Local manufacturers met to try to develop a unified front against the practice of “pirating” workers from competing firms by bidding up wages, but they have not yet developed a solution. A few have tried to mitigate the turnover by relocating production to outlying rural areas. Many of the collective farms that anchored Mexico’s agricultural program for decades are located around Torreon in the Laguna region, and they were privatized in the early 1990s under the administration of Mexican President Carlos Salinas. The privatization of these cooperatives, known as ejidos, created a supply of landless rural workers with few employment opportunities. The incorporation of former ejido lands into the Torreon apparel cluster reflects the importance of the national economy as an institutional arena affecting development outcomes. Mexico’s neoliberal turn has entailed a wholesale disavowal of the import-substituting industrialization strategy that was the region’s development orthodoxy for several decades and the enthusiastic adoption of an export-oriented growth strategy in its place (Dussel Peters, 2000). This process of economic and political transformation also entails the destruction of the Mexican Revolution’s institutional legacies. It is in this context that the privatization and sale of the ejidos – the centerpiece of a land reform process that was one of the Revolution’s most far-reaching products – must be understood. In other words, changes in the fabric of Mexico’s politico-economic institutions are reflected in the transformation of the rural areas around Torreon and their development as low-wage production sites on Torreon’s periphery. As the example of Torreon’s ejidos illustrates, the nation-state remains a critical player in shaping globalization and mediating its outcomes. In fact, the national economy is the context most directly affecting wages and working conditions in Mexico’s textile mills, garment factories, cutting rooms, and industrial laundries. Despite the efforts of anti-sweatshop campaigns and the rise of corporate codes of conduct among leading US apparel firms doing business in Mexico, the declining standard of living that many Mexican workers have experienced since the 1994 devaluation cannot be understood solely by analyzing the ways in which these workers have become incorporated into the apparel commodity chain. Instead, the extent to which Mexicans benefit from the processes of industrial restructuring, political-economic reform, and trade liberalization depends in large part on the nested institutional environments within which they live and work (Bair, 2002).
5
Conclusion
The commodity chains framework is a valuable tool for studying the implications of globalization because it is able to move across multiple scales: the global, regional, national, and local. It focuses on interorganizational
82 Jennifer Bair and Gary Gereffi
networks in global industries to illuminate the dynamics of contemporary capitalism. This chapter has shown that the apparel commodity chain in North America is driven by lead firms that coordinate international production and sourcing networks, and their decisions have compelling implications for workers in many communities. Numerous studies of exporting clusters in developing countries, whether they utilize the commodity chains framework (Kessler, 1999; van Dooren, 2002) or other approaches such as the industrial districts or cluster paradigms (Rabellotti, 1999; Vangstrup, 1999), have confirmed the importance of external linkages in shaping the experiences of local firms in export-oriented apparel industries.14 These findings suggest, as we have argued elsewhere (Bair and Gereffi, 2001), that commodity chain analyses can usefully complement studies of exporting clusters in industrializing countries by highlighting how local producers become incorporated into global markets. However, we agree with those who point out that institutional environments are a critical factor shaping international networks and mediating their consequences in the particular communities where the links of commodity chains touch down. Upgrading prospects, and developmental outcomes more generally, are determined not just by the organizational dynamics of commodity chains, but also by several layers of institutional environments: global (e.g. the Multi-Fiber Arrangement or the World Trade Organization’s Agreement on Textiles and Clothing); macro-regional (e.g. regulatory regimes such as NAFTA and the Caribbean Basin Trade and Development Act); national (e.g. welfare states and industrial relations systems); and local (e.g. industry associations and labor markets). In this respect, we concur with Palpacuer and Parisotto (2003: 99), who argue that the GCC framework “should be combined with other, ‘territorially based’ approaches … in order to better grasp the social and institutional context in which global networks operate at the local level, and how this context influences network dynamics.” Our analysis of the North American apparel industry endeavors to meet this challenge. We have analyzed the structure of the global apparel industry as a sequence of export roles, whereby lead firms construct different types of networks that connect local exporters to global markets. However, we have also focused on North America as a regional production and trade bloc within the global apparel industry and highlighted the way in which NAFTA-era restructuring is affecting firms and workers in the United States, Mexico, and the Caribbean Basin. The emergence of full-package supply networks linking Mexican firms and the US market has created upgrading opportunities for Mexico vis-à-vis the Caribbean, while simultaneously underscoring the difficulties regional integration presents for workers in US textile and garment centers, such as North Carolina and El Paso, Texas, respectively. Even within Mexico’s most dynamic apparel cluster, Torreon, the benefits generated by post-NAFTA export dynamism are contingent and
North American Apparel Industry 83
transitory. Globalization entails uneven development for firms and workers both within and across regions and nations, and viewing the process through the lens of the commodity chains framework contributes to our understanding of who wins and who loses, and why.
Notes * This article was originally published in Global Networks, Vol. 3, No. 2 (April 2003), pp. 143–169. Permission to reprint is gratefully acknowledged. 1. See also Palpacuer and Parisotto (2003) for a discussion of the producerdriven/buyer-driven typology. 2. Assembly manufacturing is the first rung on our ladder of export roles. However, for countries that are undergoing trade liberalization, the assembly model of exportoriented production can represent a “downgrading” for local firms that have designed and manufactured their own brands for sale in often highly protected national markets. Companies unable to compete with imports at home, and unable to export their own brands to foreign markets, may turn to assembly subcontracting for foreign firms as a survival strategy (Bair, 2001, 2002; Carrillo et al., 2002). 3. However, see Dussel Peters et al. (2002) for a discussion of the “limits of foreign partners as teachers.” 4. Canada’s apparel employment increased 13 percent since the passage of NAFTA, from 82 800 in 1994 to 93 700 in 2000. 5. In contrast, the United States’ two principal garment-manufacturing poles of Los Angles and New York have witnessed far less dramatic declines in apparel employment since NAFTA (Kessler, 2002; Palpacuer, 2002), although there is some concern that a resurgence of sweatshops in the United States is contributing to the competitiveness and flexibility of these urban centers, whose firms specialize in more fashion-sensitive apparel segments such as women’s outwear and sportswear (Ross, 2002). 6. In 2001, China exported almost $8.9 billion worth of apparel to the United States, while Mexico was in second place among US suppliers with $8.1 billion. Hong Kong was a distant third, at $4.3 billion in apparel exports (US Department of Commerce, US imports for consumption). 7. George Henderson, chairman and chief executive officer of Burlington Industries, pointedly criticized “the U.S. government’s history of using the textile industry as a bargaining chip in international relations” (Rives, 2001). This statement followed an announcement by the current Bush administration that it is considering cutting tariffs on textiles coming into the United States from Pakistan, a key ally in the Afghan war. 8. See Bair (2002: 192–6) for a more detailed discussion of the difficulties in measuring full-package exports. 9. According to one account, trade diversion from the Caribbean Basin to Mexico after NAFTA was responsible for 150 plant closures and 123 000 lost jobs between 1995 and 1996 (Rohter, 1997). 10. Torreon’s apparel industry actually encompasses the adjoining cities of Torreon and Gómez Palacio, which are located in the neighboring states of Coahuila and Durango, respectively. 11. This account of El Paso’s apparel industry is taken from Spener (2002), unless otherwise indicated.
84 Jennifer Bair and Gary Gereffi 12. See Bair and Gereffi (2001); Gereffi et al. (2002); and Bair (2002) for a more detailed analysis of the Torreon case. 13. Turnover rates in Torreon’s apparel industry vary between plants, with some companies reporting turnover rates of 4 percent a month while others face monthly rates of 10 percent or more. 14. See the introduction to a special issue of World Development edited by Schmitz and Nadvi (1999) for a concise overview of this literature.
References Abernathy, Frederick, John Dunlop, Janice Hammond, and David Weil (1999) Stitch in Time: Lean Retailing and the Transformation of Manufacturing – Lessons from the Apparel and Textile industries, New York: Oxford University Press. Amin, Ash (1999) “An institutionalist perspective on regional economic development,” International Journal of Urban and Regional Research 23: 365–78. ATMI (American Textile Manufacturers Institute) (2001) “Crisis in U.S. textiles,” August. Washington, DC: American Textile Manufacturers Institute. Bair, Jennifer (2001) “Casos exitosos de pequeñas y medianas empresas en México: la industria del vestido en Aguascalientes,” in Enrique Dussel Peters (ed.), Claroscuros: integración exitosa de las pequeñas y medianas empresas en México, Santiago de Chile: United Nations Economic Commission for Latin America and the Caribbean, pp. 63–105. Bair, Jennifer (2002) “Is Mexico sewing up development? NAFTA and the North American apparel industry,” PhD Dissertation, Department of Sociology, Duke University. Bair, Jennifer and Gary Gereffi (2001) “Local clusters in global chains: the causes and consequences of export dynamism in Torreon’s blue jeans industry,” World Development 29 (11) (November): 1885–1903. Bair, Jennifer and Gary Gereffi (2002) “NAFTA and the apparel commodity chain: corporate strategies, inter-firm networks, and local development,” in Gary Gereffi, David Spener, and Jennifer Bair (eds), Free Trade and Uneven Development: The North American Apparel Industry after NAFTA, Philadelphia, PA: Temple University Press, pp. 23–50. Bloomberg News (2002) “Plant closings by Levi’s will include 3 in Texas,” Houston Chronicle, April 9. Bonacich, Edna, Lucie Cheng, Norma Chinchilla, Nora Hamilton, and Paul Ong (eds) (1994) Global Production: The Apparel Industry in the Pacific Rim, Philadelphia, PA: Temple University Press. Carrillo, Jorge and Gepedo Hualde (1998) “Third generation maquiladoras? The Delphi-General Motors case,” Journal of Borderlands Studies 13 (1) (Spring): 79–97. Carrillo, Jorge, Alfredo Hualde, and Araceli Almaraz (2002) “Commodity chains and industrial organization in the apparel industry in Monterrey and Ciudad Juarez,” in Gary Gereffi, David Spener, and Jennifer Bair (eds), Free Trade and Uneven Development: The North American Apparel Industry after NAFTA, Philadelphia, PA: Temple University Press, pp. 181–99. Dicken, Peter, Philip F. Kelly, Kris Olds, and Henry Wai-Chung Yeung (2001) “Chains and networks, territories and scales: towards a relational framework for analysing the global economy,” Global Networks 1 (2), 89–112. Dickerson, Kitty G. (1999) Textiles and Apparel in the Global Economy, 3rd edition. Englewood Cliffs, NJ: Prentice Hall.
North American Apparel Industry 85 Dussel Peters, Enrique (2000) Polarizing Mexico: The Impact of Liberalization Strategy, Boulder, CO: Lynne Rienner. Dussel Peters, Enrique, Clemente Ruiz Durán, and Michael Piore (2002) “Learning and the limits of foreign partners as teachers,” in Gary Gereffi, David Spener, and Jennifer Bair (eds), Free Trade and Uneven Development: The North American Apparel Industry after NAFTA, Philadelphia, PA: Temple University Press, pp. 234–45. Emert, Carol (1999) “Levi’s to slash U.S. plants: competitors’ foreign-made jeans blamed,” San Francisco Chronicle, February 23. Gereffi, Gary (1994) “The organization of buyer-driven global commodity chains: how U.S. retailers shape overseas production networks,” in Gary Gereffi and Miguel Korzeniewicz (eds), Commodity Chains and Global Capitalism, Westport, CT: Praeger, pp. 95–122. Gereffi, Gary (1995) “Global production systems and third world development,” in Barbara Stallings (ed.), Global Change, Regional Response: The New International Context of Development, New York: Cambridge University Press, pp. 100–42. Gereffi, Gary (1998) “Commodity chains and regional divisions of labor in East Asia,” in Eun Mee Kim (ed.), The Four Asian Tigers: Economic Development and the Global Political Economy, San Diego, CA: Academic Press, pp. 93–124. Gereffi, Gary (1999) “International trade and industrial upgrading in the apparel commodity chain,” Journal of International Economics 48 (1) ( June): 37–70. Gereffi, Gary (2001) “Shifting governance structures in global commodity chains, with special reference to the Internet,” American Behavioral Scientist 44 (10) ( June): 1616–37. Gereffi, Gary, Ronie Garcia-Johnson, and Erika Sasser (2001) “The NGO-industrial complex,” Foreign Policy 125 ( July–August): 56–65. Gereffi, Gary, Martha Martínez, and Jennifer Bair (2002) “Torreon: the new blue jeans capital of the world,” in Gary Gereffi, David Spener, and Jennifer Bair (eds), Free Trade and Uneven Development: The North American Apparel Industry after NAFTA, Philadelphia, PA: Temple University Press, pp. 203–23. Hill, Suzette (2000) “Burlington down, but coming up swinging,” Apparel Industry Magazine 61 (6): 66. ILO (International Labor Office) (1996) Globalization of the Footwear, Textile and Clothing Industries, Geneva: ILO. ILO (International Labor Office) (2001) World Employment Report 2001: Life at Work in the Information Economy, Geneva: ILO. INEGI (Instituto Nacional de Estadistica, Geografia, e Informatica) (2002), Banco de información Economica database, Mexican government statistics. Available at http://dgcnesyp.inegi.gob.mx; website consulted December 10. Jones, Jackie (1998) “Apparel sourcing strategies for competing in the U.S. market,” Industry, Trade, and Technology Review (December): 31–40. Kaplinsky, Raphael (1998) “Globalisation, industrialization, and sustainable growth: The pursuit of the nth rent,” Discussion Paper 365, Institute for Development Studies, University of Sussex. Kaplinsky, Raphael (2000) “Spreading the gains from globalization: what can be learned from value chain analysis?” Working Paper 110, Institute for Development Studies, University of Sussex. Kessler, Judi A. (1999) “The North American Free Trade Agreement, emerging apparel production networks and industrial upgrading: the Southern California/Mexico connection,” Review of International Political Economy 6 (4): 565–608. Kessler, Judi A. (2002) “The impact of North American economic integration on the Los Angeles garment industry,” in Gary Gereffi, David Spener, and Jennifer Bair
86 Jennifer Bair and Gary Gereffi (eds), Free Trade and Uneven Development: The North American Apparel Industry after NAFTA, Philadelphia, PA: Temple University Press, pp. 74–99. Klein, Naomi (1999) No Logo, New York: Picador. Mathews, Dale (2002) “Can the Dominican Republic’s export processing zones survive NAFTA?,” in Gary Gereffi, David Spener, and Jennifer Bair (eds), Free Trade and Uneven Development: The North American Apparel Industry after NAFTA, Philadelphia, PA: Temple University Press, pp. 308–23. Mecia, Tony (2002) “Textile industry’s clout in Washington unravels,” The Charlotte Observer, Charlotte, NC, April 29. Medaille, Bill and Andrew Wheat (1997) “Faded denim NAFTA blues,” Multinational Monitor 18 (12): 23–27. Moore, Carlos (1999) “NAFTA: A good deal or not?” Greensboro News & Record, Greensboro, NC, February 21. Moreno, Jenalia (1997) “El Paso’s torn fabric; Border city struggles to mend loss of apparel industry,” The Houston Chronicle, December 14. Mortimore, Michael (2002) “When does apparel become a peril? On the nature of industrialization in the Caribbean Basin,” in Gary Gereffi, David Spener, and Jennifer Bair (eds), Free Trade and Uneven Development: The North American Apparel Industry After NAFTA, Philadelphia, PA: Temple University Press, pp. 287–307. Neff, Joseph, John Wagner, and Karin Rives (2002) “Leaders scramble as job losses mount,” News & Observer, Raleigh, NC, August 20. Palpacuer, Florence (2002) “Subcontracting networks in the New York garment industry: changing characteristics in an era of globalization,” in Gary Gereffi, David Spener, and Jennifer Bair (eds), Free Trade and Uneven Development: The North American Apparel Industry after NAFTA, Philadelphia, PA: Temple University Press, pp. 53–73. Palpacuer, Florence and Aurelio Parisotto (2003) “Global production and local jobs: can global enterprise networks be used as levers for local development?” Global Networks 3 (3) (April), pp. 97–120. Rabellotti, Roberta (1999) “Recovery of a Mexican cluster: devaluation bonanza or collective efficiency?” World Development 27 (9): 1571–85. Rives, Karin Schill (2001) “Textile maker Burlington Industries files for bankruptcy protection,” News & Observer, Raleigh, NC, November 16. Rives, Karin, and Joseph Neff (2002) “Remedy misses the mark,” News & Observer, Raleigh, NC, August 19. Rodriguez-Archila, Laura (2000) “Apparel market: new U.S. legislation places CBERA countries on a more equal competitive basis with Mexico,” Industry, Trade and Technology Review (July): 21–32. Rohter, Larry (1997) “Impact of NAFTA pounds economies of the Caribbean, jobs flowing to Mexico,” New York Times, January 30, p. 1. Romney, Lee (2001) “As U.S. demand shrinks, Mexico’s garment makers alter strategies,” Los Angeles Times, April 22. Ross, Robert J. S. (2002) “The new sweatshops in the United States: how new, how real, how many, and why?,” in Gary Gereffi, David Spener, and Jennifer Bair (eds), Free Trade and Uneven Development: The North American Apparel Industry after NAFTA, Philadelphia, PA: Temple University Press, pp. 100–22. Sassen, Saskia (1996) Losing Control: Sovereignty in an Age of Globalization, New York: Columbia University Press. Schmitz, Hubert and Khalid Nadvi (1999) “Clustering and industrialization: introduction,” World Development 27 (9): 1503–14.
North American Apparel Industry 87 Schrank, Andrew (2002) “Downgrading by diffusion: technology transfer, learning-bywatching, and the export-led trap in the Americas,” Paper presented at the 2002 Annual Meeting of the American Sociological Association, Chicago, IL, August 16–19. Spener, David (2002) “The unraveling seam: NAFTA and the decline of the apparel industry in El Paso, Texas,” in Gary Gereffi, David Spener, and Jennifer Bair (eds), Free Trade and Uneven Development: The North American Apparel Industry after NAFTA, Philadelphia, PA: Temple University Press, pp. 139–60. Spener, David and Randy Capps (2001) “North American free trade and changes in the nativity of the garment industry workforce in the United States,” International Journal of Urban and Regional Research 25 (2): 301–26. Storper, Michael (1997) The Regional World: Territorial Development in a Global Economy, New York: Guilford. Sturgeon, Timothy J. (2001) “How do we define value chains and production networks?” IDS Bulletin 32 (3): 9–18. Taplin, Ian (1994) “Recent manufacturing changes in the U.S. apparel industry: the case of North Carolina.” in Edna Bonacich et al. (eds), Global Production: The Apparel Industry in the Pacific Rim, Philadelphia, PA: Temple University Press, pp. 328–44. van Dooren, Robine (2002) “TexMex: linkages in a binational garment district? The garment industries in El Paso and Ciudad Juarez,” in Gary Gereffi, David Spener, and Jennifer Bair (eds), Free Trade and Uneven Development: The North American Apparel Industry after NAFTA, Philadelphia, PA: Temple University Press, pp. 161–80. Vangstrup, Urik (1999) “Collective efficiency and regional industrial clusters in Mexico – assessment of a theory of local industrial development,” PhD Dissertation, Department of Geography and International Development Studies, Roskilde University.
4 Technology versus Trade versus Social Institutions: Explaining Rising Wage Inequality in the Chilean Cosmetics Industry1 Janine Berg
1
Introduction
One of the arguably more disappointing outcomes of globalization has been the rise in wage inequality in developing countries. With few exceptions, mainly in East Asia, wage inequality has increased under trade and financial liberalization. In Latin America, including Chile, this outcome has been particularly acute. The association between openness and rising wage inequality has been well established in econometric studies on the Chilean economy (Robbins, 1994; Beyer et al., 1999). The studies assume that free trade causes an increase in demand for skilled workers, a fall in demand for low-skilled workers, and thus, a rise in the relative wage gap. Similar studies on other developing countries (Aitken et al., 1996; Feenstra and Hanson, 1997), have associated increased multinational activity in developing countries with rising wage inequality. Multinationals are believed to cause technological spillovers, which increase the demand for skilled workers, increasing relative wage dispersion.2 What remains unclear, however, is how these effects occur. What are the dynamic processes that cause openness to lead to increased wage dispersion? Are there other forces that are not considered, and if so, how do they play out? This chapter gives evidence of how restructuring under free trade affects workers, through a case study of the Chilean cosmetics industry. The Chilean cosmetics industry was closed to trade under the period of import-substitution industrialization. The sector began to be liberalized in 1974 and has since faced increasing competition, particularly during the 1990s. The increase in foreign competition arising from free trade has had two principal effects depending on whether the firm is a foreign-owned multinational or a domestic, Chilean-owned firm. Multinationals have 88
Labour and the Globalization of Production 89
changed their competitive strategy in Chile, leading to a loss of production jobs and increased relative demand and wages for skilled workers employed in management and sales. Domestic firms, on the other hand, have responded to the more competitive environment by making investments that expand and upgrade their manufacturing facilities. The technological changes have been low-skilled biased, leading to an increase in the relative employment of low-skilled workers. Yet the employment increases have not led to concomitant increases in the relative wages of low-skill workers. Domestic firms have used the weakened labor relations environment to hold down wage increases for low-skilled workers, allowing the firms to continue underpricing the multinationals and retain market share.
2
Free trade and wages
Since 1973, when the military government unleashed its program of economic liberalization, including trade reform, wage differentials in Chile between skilled and low-skilled workers have steadily increased. By the 1990s, the returns to schooling of university-educated workers had approached a 30 percent wage premium, after controlling for other worker characteristics.3 At the same time, the returns to schooling of high school educated workers had steadily fallen, from a difference of about five percentage points compared with university-educated workers in the 1960s under import-substitution industrialization, to more than a fifteen percentage-point difference by the end of the 1990s (see Figure 4.1).4 The increase in the returns to schooling for university-educated workers is surprising given the tremendous increase in educational attainment among Chilean workers over the four decades. In the 1957–65 period, only 6.3 percent of the Greater Santiago labor force had more than 12 years of schooling, yet by 1991–96, 22.3 percent of the labor force had more than 12 years of schooling and 14.6 percent were university educated (17 years or more). At the same time, the number of workers with less than 12 years of schooling was cut in half from 83.7 percent to 42.9 percent, as the government pursued its policy of universal secondary education (Bravo and Marinovic, 1997). The increase in wage inequality is also surprising as the shift to an open economy is expected to compress wages in developing countries where labor, particularly low-skilled labor, is the more abundant factor of production. In Chile, it was thought that import-substitution industrialization had created a bias towards greater incorporation of capital goods into production that would be reversed under free trade, since without the preferential terms of trade, industries would shift production towards more readily and cheaply available low-skilled labor (Corbo and Meller, 1982). As predicted by Hecksher-Ohlin theory, low-skilled labor, the relatively more abundant factor of production, would reap the returns of the policy shift. Moreover, inequality was expected to reduce, as the increased demand for low-skilled
90 Janine Berg 0.35
0.30
0.25
Return
0.20
0.15
0.10
0.05
19
5 19 8 6 19 0 6 19 2 6 19 4 6 19 6 6 19 8 7 19 0 7 19 2 7 19 4 7 19 6 7 19 8 8 19 0 8 19 2 8 19 4 8 19 6 8 19 8 9 19 0 9 19 2 9 19 4 9 19 6 98
0.00 Year Junior education High school education College and graduate education Figure 4.1 Returns to schooling for university, high-school, and primary-school educated workers, Greater Santiago, 1958–98 Source: Berg and Contreras (2004), based on University of Chile Employment Survey.
labor under free trade would, under the Stolper-Samuelson theory, increase the remuneration of low-skilled labor while lowering the remuneration of skilled labor. For supply and demand shifts alone to explain the substantial rise in wage inequality in Chile since liberalization, it would imply that demand shifts in favor of skilled workers were great enough to mitigate the increase in supply of skilled workers. It would also mean that production did not shift towards the relatively more abundant factor of production, as predicted by Heckscher-Ohlin. This belief supports the view of some economists who have argued that trade liberalization in Chile has been associated with skillbiased technological change. Robbins (1994) and Gindling and Robbins (2001) have advocated this viewpoint, arguing that machinery imported since liberalization has necessitated employment of skilled workers, while reducing the need for low-skilled workers. Similarly, Beyer et al. (1999) regress an index for openness on Chilean wage differentials over time and
Labour and the Globalization of Production 91
find a positive and significant coefficient. They suggest that this can be explained from the technological change that was biased against low-skilled labor in the 1980s and 1990s. These arguments, however, are based on econometric work that measures correlations and thus cannot account for variables that are not easily proxied for, such as labor market deregulation.5 Labor market deregulation is an important issue since along with the trade reforms, the government adopted policies to make the labor market more “flexible,” as well as to de-politicize labor relations (Meller, 1992). When the government assumed power in 1973, it threw out the existing labor code, replacing it in 1979 with a wholly revised code that restricted unions and collective bargaining. Hence, the possibility remains that the econometric results given in the literature to explain rising wage inequality in Chile since liberalization are spurious, with trade reforms, machinery imports, and wage inequality occurring simultaneously, but not necessarily as the direct result of one another.
3
Methodology
To analyze the causes of increased wage inequality in Chile simply by identifying an outcome and then hypothesizing about its causes provides little understanding of the mechanisms by which trade liberalization alters demand for workers in developing countries. This chapter aims to understand the dynamic processes that cause openness to lead to increased wage dispersion, through the use of a case study. The goal of case study research is to “expand and generalize theories,” as opposed to statistical work which enumerates frequencies. Thus findings are to be used to generalize to theoretical propositions as opposed to assessing incidence (Yin, 1994, p. 10). Industry decompositions for Chile indicate that around 80 percent of relative employment and wage shifts since liberalization have been within industry rather than between industry (Meller and Tokman, 1998), suggesting that a within-industry study is an appropriate analytical approach. With this case study, I am able to consider the effects of changing strategies of multinational firms, increased competition, technological change, and a changed labor-relations environment, on the employment and pay of skilled and low-skilled workers. The analysis provides several insights into the applicability of the traditional view of trade and wages (HOS), versus alternative explanations, such as skill-biased technological change (SBTC) and labor market rigidity. Doing so highlights the need to consider other changes in the economy besides just the price and use of certain labor. A variety of factors were taken into consideration when choosing the case, including feasibility, but also whether the case offered an interesting study. I deliberately avoided dying industries, such as apparel and footwear, which have suffered tremendous and continual job loss since the 1970s. Employment in the cosmetics industry, on the other hand, has expanded
92 Janine Berg
over the past several decades despite significant restructuring. In 1979, cosmetics firms with factories employed 4359 workers growing to 6239 by 1997.6 Another important factor in my decision to study the cosmetics industry was its technological accessibility. Cosmetics products have been successfully produced in Chile for a century, thus their production did not pose a scientific hurdle for the country. Also, simple perusal of supermarket and pharmacy shelves indicated that there was substantial domestic and foreign presence in the sector, offering the possibility for an analysis of foreign direct investment. The fall in trade barriers, transportation costs and more aggressive market seeking by multinational cosmetics firms offered enough challenges to make its story worth telling. The field research is based on interviews with management and union leaders at nine cosmetics companies, four of which are multinational subsidiaries, the other five are Chilean owned. Other sources of information included data from the Chilean annual manufacturing census, available at the four-digit SIC level; factory visits; interviews with outside experts; and information from annual reports and the business press. The use of different sources of evidence is encouraged in case study research because it allows the researcher to triangulate, meaning that researcher enlists multiple sources of evidence to show convergence upon a line of inquiry. Doing so results in more accurate and convincing study conclusions. Thirteen firms were contacted for interviews, of which nine agreed to participate in the study. Firms were selected to achieve a balanced representation of multinational and domestic firms and different-sized firms; within these categories the selection was random. Based on sales and market share information, the sample includes roughly 30 percent of the Chilean cosmetics market.
4
Features of the Chilean cosmetics industry
The cosmetics industry is a fast-moving consumer good industry that includes skin care, hair care, fragrance, make-up and personal hygiene products. For 2000, the world cosmetics industry was valued at US$122.2 billion (Koser, 2001). The Chilean cosmetics industry is composed of approximately 50 companies in production, distribution or both. Roughly 60 percent of the market is foreign-owned, either through multinational subsidiaries, distributorships or licensing agreements. Products are targeted at three consumer segments: selective (4 percent), semi-selective (20 percent) and mass-market (52 percent); the remaining 24 percent of the population does not have the resources to purchase cosmetics products (U.S. Department of Commerce, 2000). According to the Chilean Cosmetics Industry Chamber of Commerce, retail sales in 1999 were US$680 million, which accounts for roughly 1 percent of GDP and annual per capita spending of US$45. With an increase in real incomes beginning in the late 1980s and continuing in the 1990s, more Chileans are regularly buying cosmetics products and many have upgraded their market segment.
Labour and the Globalization of Production 93
5
Restructuring under Free Trade
Although initially alarmed by the government’s 1974 announcement that tariffs would be reduced to 10 percent by 1979, Chilean cosmetics firms faced little competition from imports during the 1970s and 1980s. The economic crisis of the early 1980s, coupled with the temporary increase in tariffs following the crisis, resulted in a real average annual fall of final good cosmetics imports of 4.2 percent for the decade. During the 1970s and 1980s, cosmetics and personal hygiene imports were limited to selective niche products, such as name-brand perfumes sold in department stores. In the 1990s, imports diversified in terms of product range, as well as consumer market group. A strong economy along with multinational fervor to capture new markets led to a real annual increase of final good cosmetics imports of 66.5 percent between 1990 and 2000. By 2000, imports totaled US$100 million (at wholesale prices), a significant portion of the US$680 million domestic retail sales market. Importing economically priced mass-market products in the 1990s threatened the survival of companies that had been previously untouched by import competition. At the same time, cosmetics companies faced cost pressures on the retailing end with the growth of “hypermarkets” (defined as having 10 000 m2 or more of floor space) as well as supermarket and pharmacy chains. The US supermarket model was not introduced into Chile until the 1960s, yet by century’s end, 63 percent of the country’s population shopped regularly at supermarkets. The industry is extremely consolidated, with the market leader, D y S, holding almost 30 percent market share and 21 hypermarkets accounting for 28 percent of industry sales (ASACH, 2001). The consolidated buying power of the retailers has allowed them to exert pressure on cosmetics firms to keep prices down. Further contributing to the decrease in margins is the practice by large chains and hypermarkets of renting their shelf space, which has increased cosmetics firms’ selling costs. Cosmetics multinationals in Chile: resurgence and regionalization The first wave of cosmetics multinationals entered Chile during the period of import-substitution industrialization by hopping the tariff and constructing greenfield factories. Between 1970 and the debt crisis of 1982, there was little multinational investment as firms felt hampered by the political and economic turmoil in Chile as well as their own internal problems as a result of the oil crisis. Following the debt crisis, multinational expansion re-surged, attracted by the ample brownfield investment opportunities as a result of the high bankruptcy rate following the 1982 crash. In the 1990s, cosmetics multinationals in Chile altered their policy of setting up, either as greenfield or brownfield investments, local production centers in the country. This policy was not as a response to internal conditions in Chile, but rather was dependent on the economic policies of other
94 Janine Berg
countries in the region, as well as on the internal policies of the countries. Many factories that had opened in Latin American countries to “tariff-hop” during import-substitution industrialization closed in the 1990s after the region embraced free trade, making it possible to establish regional production centers to serve the different country markets. Overall, Chile has fared negatively under the strategy of regionalizing production since the multinationals prefer to locate production in Brazil or Mexico where the domestic markets are larger, rather than in small countries like Chile. Moreover, Chile’s surrounding Andes Mountains make transportation to other Latin American countries difficult and costly. The effect of the regionalization policy was particularly perverse in some instances when multinationals acquired successful Chilean cosmetics firms, only to then close the factory and import the products from their regional production centers. Complementing the strategy of regionalizing production was the decision by multinationals to narrow their product lines, removing country specific products so that the same good would be sold in all countries of the region. In doing so, production facilities could increase economies of scale. Although cosmetics production is still produced in batches – meaning that production is organized discontinuously as a succession of workshops, rather than organized as a continuous line – firms save money by having fewer, but larger, batches. Also product standardization can allow automation of some laborintensive tasks, such as putting on a bottle top. More importantly, firms can save money on raw material purchases since the larger quantities needed enable them to negotiate better prices with their suppliers. For examples, firms can order chemical supplies by the tank wagon, as opposed to the more costly drums. Factories can also make better use of high-cost “magic” ingredients, such as botanical extracts used in shampoos. Concentration levels of the magic ingredients are minimal (less than 1 percent), yet chemical producers will sell a minimum of ten kilos, nine of which may not be needed by a small production facility. What has this meant for the Chilean subsidiaries? Previously the subsidiaries operated as quasi-independent firms with product lines oftentimes specific to the country, with production done locally by in-house factories, and with marketing, sales and distribution undertaken by the country office. Now multinational subsidiaries in Chile, for the most part, are limited to marketing, sales and distribution. The benefit of the regionalization strategy is that the reduced production costs have allowed the multinationals to lower their prices and compete for the mass market, a segment that had previously been relegated to the domestic firms. The strategy has been largely successful: multinational firms have captured market share from consumers previously unable to afford their products.7 Domestic cosmetics firms: fighting to retain market share Three, somewhat intertwined, factors have threatened the competitive position of the Chilean firms: the increase in imports, the drop in multinational
Labour and the Globalization of Production 95
product prices, and the consolidation of the retail sector and subsequent decline of traditional mom and pop outlets. The response of the domestic firms to the competitive threats include upgrading technology, diversifying the product line to offer a wider range of cosmetics products, and expanding into third-party production. Prices have been lowered as well. Among Chilean cosmetics firms, exports comprise a minor share of production (roughly 8 percent of sales) and the industry remains constrained by the small size of its domestic market. Production is thus done in small batches making it preferable to produce with semi-automatic rather than automatic equipment. Nevertheless, the domestic firms have taken steps to update their stock of machinery in an effort to improve product quality. The new technologies, however, have not affected the organization of work. Most of these investments have been through the purchase of second-hand equipment from closed multinational factories. Additionally, firms have invested in increasing the range of products they offer, thereby diversifying their risk and staking out market share in new areas. Product diversification also allows the firms to make the best use of the small-batch production model. It also enables the firms to offer more products on the shelf space that must now be rented at the supermarkets. Another way to increase sales and take advantage of installed capacity has been to accept subcontracting work from other cosmetics companies, from supermarkets and pharmacies that have generic lines, or from hotels. The domestic firms have been successful in retaining the low- and middleincome segments of the market, where sales are based more on price than on marketing and advertising. Some domestic firms have also been able to successfully compete with the multinationals in the higher-end market segment, exploiting their advantage of in-country product development and production. The at-home advantage allows them a quicker turn-around in developing new products that respond to the demands of Chilean consumers.
6 How have the shifts in competitive strategies affected the industry’s workforce? The competitive policies of multinational corporations and the defensive responses of domestic firms have had direct ramifications on who is or isn’t hired by these two competitor groups, as summarized in Table 4.1. For workers employed at multinationals, the strategy to regionalize production resulted in a closing of local factories that forced the dismissal of the mostly low-skilled workforce, employed to place tops on bottles and other simple, yet labor-intensive, tasks. Other more skilled production workers, such as the semi-skilled mixers and machinery adjusters, as well as the few personnel hired to ensure quality control, have also lost their jobs. Moreover, in the 1990s, multinational subsidiaries had adopted their headquarters’ policy of outsourcing non-core personnel such as the
96 Table 4.1 Cosmetics industry jobs: gender, skills, pay and labor demand Job title/ duties
Sex
Skill/ schooling
Pay (domestic firms) US$
Labor demand Multinationals
Domestic
Management Both
High-skilled. University degree.
$7 300/month (technical dir.) $5 500/month (HR manager)
Unchanged
Unchanged
Marketing and sales
Both
Marketing staff has university degree. Sales staff has high school or more.
Marketing: $1 400– 2 600/month Sales: $1 200– $1 600/month
Increase
Increase dependent on consumer segment: more exclusive increased demand
Quality control
Mostly Female
University or technical school studies.
Head of QC $1 800/month; Assistants $550/month
Decrease – job loss with closing
Increase from expansion and emphasis on product quality.
Compounder Male
Semi-skilled. OJT.
$500/month
Decrease – job loss with closing
Increase with greater output
Mechanic
Semi-skilled. OJT. May have technical school studies.
$500/month
Decrease – job loss with closing
Increase with greater output
“Envasadora”/ Female filling and assembly
Unskilled, some high school
$275/month
Decrease – job loss with closing
Increase with greater output
Storage and distribution
Male
Unskilled, some $275/month high school
Decrease from outsourcing and computerization.
Some increase, though controlled w/computerization
Non-core services (cleaning, security, grounds, cafeteria)
Cleaning/ Unskilled cooking – female; security/ grounds – male
Decrease from outsourcing
Unchanged
Male
“Promotora”/ Female product promoter
$250/month
Unskilled, $275/month appearance most important
Increase; needed Increase; for selling at emphasis department stores on low-cost promotional efforts
Note: Wages are US$ per month and were calculated at C$550 US$1. Data are averages for the category and are based on interviews with general managers and union wage information. Source: Author’s compilation based on interviews, factory visits and union contracts.
Labour and the Globalization of Production 97
maintenance, cleaning and security staff. Although these workers may then be rehired through a subcontracting firm, their wages are lower than they were when hired directly by the multinational. The study confirmed the well-known finding that multinationals pay more than domestic firms – in this industry, almost double the domestic firms’ low-skilled wage rate of roughly US$275 per month. Unfortunately, managers could not convincingly explain why they pursued this wage policy. In the end, however, few low-skilled workers benefited from the higher wages paid by multinationals as many of these jobs were eliminated, either through regionalization or outsourcing.8 The workers that kept their jobs at the multinational subsidiaries are employed in management, marketing and sales; all well-paid, skilled positions with workers, who if not university educated, have at least completed high school and have work experience. Yet the size and function of the subsidiaries’ staff is now greatly reduced. The net result of labor demand changes at multinationals has been a large upward shift in the relative demand for skilled compared with low-skilled workers. Domestic firms, on the other hand, have shown an increase in the employment of production workers as companies have focused more on production-related activities, accepting subcontracting jobs from multinational firms and retailers. The competitive strategies of product diversification and third-party production coupled with the characteristics of small-batch production have meant that technological investments have not caused a drop in employment. Instead, more low-skilled workers have been needed for the labor-intensive tasks of monitoring runs, fitting bottle tops and packaging goods. Even when machinery is more productive, such as upgrading from manual to semi-automatic equipment, increases in output have been accompanied by increases in employment. The following example illustrates the employment effects of the machinery investments undertaken by domestic producers. In 1999, a domestic firm purchased a new reactor from Argentina for the purpose of producing facial cream. The new reactor produces two and a half times the volume of the reactor they owned, which dated from the 1970s. The digital temperature display on the new reactor is more accurate and more readable than the mechanical temperature display of the old one; the reactor itself is also easier to clean. The domestic firm still uses the old reactor, which is located in the workroom next to the new reactor. Compared with the old reactor, the new one improves quality by having improved temperature control, by making larger batches, and by mixing the batches more quickly. In sum, it is more productive. Yet at the packaging stage of production, more workers (envasadoras) are needed to fill and put the caps on the jars, and then package the boxes to prepare them for delivery. In the aggregate, the labor saved from having a better reactor is cancelled out by the labor intensity of the packaging operations, leading to an increase in employment. Yet overall
98 Janine Berg
productivity is greater and the final product is of higher quality. As this example shows, it is incorrect to interpret technological upgrading in this industry as a capital–labor trade-off. Also contributing to the increase in employment has been the low-cost approach that domestic firms have taken towards marketing and sales. Because domestic firms prefer to use the savings from advertising to offer lower prices, spending on product promotion has involved hiring lowskilled product promoters ( promotoras) to introduce the products, as opposed to other more costly forms of advertising. Thus, this competitive strategy has also increased low-skilled employment. The result has been an overall increase in the relative demand for lowskilled labor, as confirmed by manufacturing census data. According to the census, employment at cosmetics factories increased by 83 percent between 1980 and 1995 for low-skilled workers, while skilled workers’ employment increased by 49 percent during the same period.9 Yet the employment gains that favored low-skilled production workers did not lead to an improvement of their financial position in the firm. Between 1980 and 1995, productivity increased annually by 3.8 percent, skilled workers’ wages increased by 4.2 percent annually, but low-skilled workers’ wages increased by just 1.5 percent annually. Thus the strong relative employment gains for lowskilled workers were not matched by wage gains.
7
What explains the wage divergence?
To understand the relative lack of wage gains for low-skilled workers in domestic firms we must consider labor relations. Labor unions in Chile are divided along class lines, with low-skilled and semi-skilled production workers belonging to a union, and skilled production workers, administrative staff and managerial personnel, generally not unionized. Thus, collective bargaining negotiations affect just one class of workers. The decline in union power, as a result of repression and the introduction of neo-liberal labor laws, affected low-skilled workers as opposed to skilled workers. Prior to the military coup, the cosmetics industry was a highly unionized sector, with its own industry federation that dates back to the early 1940s. Today, the few unions that exist are in firms that were in business prior to the military coup; firms that started after 1973 do not have unions. Moreover, there are fewer unions today as domestic cosmetics firms acquired by multinationals in the 1980s and 1990s were closed under the regionalization strategy. Among the firms that have unions, the affiliation rate is low. When workers have effectively organized, their efforts have often been thwarted by employers who promote union leaders to managerial positions, creating a leadership vacuum, or give higher pay increases to non-union workers, causing the union to eventually dissolve (see Table 4.2).
Labour and the Globalization of Production 99 Table 4.2 Existence and status of unions in participating firms Firm
Union?
What happened? Condition of union?
Anti-union practices?
M1
No
Dissolved with factory closing.
M2
No
Dissolved 2 years prior to factory closing.
M3
No
Dissolved with factory closing.
M4
No
Subcontracts manufacturing.
C1
Yes
Only 20% of production workers are members.
C2
No
Workers sign collective agreement.
C3
Yes
Most of the 22 production workers belong. Negotiations concern wage increases.
C4
Yes
Union exists for both white-collar and blue-collar workers. Negotiations concern wage increases
Allegations of anti-union practices by promoting union leaders to non-union positions.
C5
No
Workers sign collective agreement
Management admitted paying non-union workers more in a deliberate and successful attempt to destroy the union.
Allegations of anti-union practices by offering early retirement package to union leader and bonuses to workers to leave union.
Allegations of anti-union practices by promoting union leader to non-union position.
Note: M1–M4 are multinational firms; C1–C5 are domestic firms. By negotiating a collective agreement as opposed to a collective bargaining contract, the workers forfeit their right to strike. Source: Interviews with management and union leaders; collective bargaining contracts.
The low affiliation rate and lack of dedication to collective efforts on the part of the workers has made it easier for domestic cosmetics firms to constrain wage growth. Yet even well functioning unions with capable leadership often have difficulty securing wage increases for their workers due to the highly competitive external environment that the firms face. For example, in 1998 wage negotiations at one domestic firm, management placed
100 Janine Berg
greater emphasis on the heightened state of competitiveness within the cosmetics industry, and in particular, on the market power of multinational firms, as opposed to the slowdown in the Chilean economy. In a response to union demands for a wage increase, management wrote: Nowadays competition is noticeably greater as a consequence of market liberalization; we not only face internal competition, but also very strong competition from the most powerful transnationals of the sector … . (Contract Negotiation Document, C4, 1998, translated from Spanish) Management mentioned how a multinational had acquired the licenses to distribute products that it had been distributing, causing a fall in company sales and lower market share. In the end, the workers received wage increases of one percent above the previous year’s rate of inflation. The wage policy of the firms differs noticeably, however, towards skilled marketing and quality control staff and certain semi-skilled workers. When discussing the wage policy towards more skilled workers, managers stated how they paid these workers greater wages to reduce turnover. For the semiskilled workers responsible for mixing large batches of expensive raw materials or who work with chemical explosives, the efficiency wage policy is appropriate. It also seems appropriate for the skilled workers responsible for formulating new lines of products and conducting quality control tests. Yet the same reasoning is less understandable towards skilled marketing and sales workers, particular since managers answered that it was easy to find these skilled workers. The divergence in wage growth stems from the increasingly competitive environment, which has decreased the ability of firms to share rents with workers. Within this setting, firms have favored sharing rents with more skilled workers. By controlling wage growth of low-skilled workers’ wages, more funds were made available to compensate the more skilled workers. In this respect, skilled workers benefited from the drop in low-skilled workers’ wages, as this increased the amount of revenue available for their wages.10 Unions have difficulty countering management’s wage policy as their inability to retain members lessens their representation in the firm, hampering their negotiating power. The result is little real wage growth for the industry’s easily replaceable, low-skilled workers.
8
Conclusion
Increased competition as a result of free trade and changing multinational strategies coupled with a weakened union movement have resulted in little real wage growth for low-skilled workers in the Chilean cosmetics industry. The story is invariably more complex when told by an industry case study
Labour and the Globalization of Production 101
as opposed to econometric output. The findings from this case study give some insight into the causes of widening wage dispersion in Chile under free trade and, on a more theoretical level, insights into the relevance of competing theories on trade and labor. To begin with, it is not simply that Chile liberalized its trade in 1974 which prompted a shift in multinational strategy, but rather that the region reduced trade barriers in the 1990s prompting multinational firms to redesign the competitive strategies formed under ISI. Regionalization has enabled the multinationals to be more aggressive about controlling costs and thus more able to enter the mass-market consumer segment. It has also directly forced competition with domestic firms, leading to a drop in prices and a lowering of rents. Although price drops have been good for the Chilean consumer, to the extent that wages are associated with firm profits, it has not been good for some of the industry’s workforce. The shift in multinational strategy also has implications regarding traditional views about the benefits of multinational investment for a host country. Although the purchase of second-hand multinational equipment by domestic firms could be considered a “spillover,” for Chile, it was a one-shot benefit. Furthermore, the association between multinationals and greater relative employment of skilled workers is the result of the elimination of low-skilled positions, a shifting away from local production to a country presence that is limited to marketing, distributing, and selling foreign products in the host country. Thus, the skill bias is not the outcome of production being more advanced in these firms. Moreover, multinational firms, unlike their domestic counterparts, outsource non-core service workers, a practice that has narrowed internal labor markets as well as reducing wages for the outsourced staff. It is evident from the analysis of the cosmetics industry that skill-biased technological change has not been a factor in the growth of relative wage differentials in the industry. In multinationals, the greater employment of skilled workers is not technologically driven, but rather, the result of deindustrialization in the sector. In domestic firms, machinery investments have improved quality and increased output, but the strategy of product diversification has resulted in an expansion of low-skilled employment, as more workers are needed for the manual tasks of assembly and packaging. Moreover, domestic firms have preferred low-cost competitive strategies that have favored employment of low-skilled workers, along Heckscher–Ohlin lines. With greater competition domestic firms have chosen to control costs by limiting the growth of low-skilled workers’ wages. The weakness of the union movement, as evident in low affiliation rates and waning commitment by leaders and the rank and file, has enabled managers to force cost reduction on low-skilled workers. Although one can only speculate what the outcome of the sector would have been with a strong union movement, it
102 Janine Berg
is likely that it would have resulted in a reduction of skilled workers’ wage growth and a decrease in wage differentials. At a minimum, the study highlighted the importance of institutional change in determining the distribution of wages and showed how decisions at the firm level can be used to constrain wage growth, beyond the constraints imposed by the aggregate economy. The goal of this case study was to offer insights into employment and wage change under free trade in Chile, the mechanisms that have triggered these changes, and the relevance of economic theory. Still, appeasing critics concerning the applicability of these findings to the larger economy, requires that similar studies be undertaken of other Chilean industries. Nevertheless, this study gives us a starting point for questioning prevailing wisdom about the relationship between trade, employment, and wages in open developing economies.
Notes 1. Financial support of field research in Chile during 2000–01 was provided by the Fulbright-Hays Doctoral Dissertation Research Abroad Fellowship program. I am grateful to William Milberg for many helpful comments and suggestions on this research. 2. See the chapter by Slaughter (this volume) for a more in-depth discussion of the possible effects of multinational activity on developing country workers. 3. Based on data from the University of Chile employment survey which covers the Greater Santiago area. 4. Other measures of inequality reflect similar patterns. For example, the Gini coefficient for Greater Santiago increased from 0.506 in 1964–69 to 0.586 in 1987–90. Alternatively, estimates based on the national household survey, CASEN, give a Gini coefficient of 0.560 in 1987, falling slightly to 0.553 in 1996 (Larrañaga, 1999). 5. Household and firm-level data sets in Chile do not contain information on union membership or other similar indicators. 6. Based on data from annual manufacturing survey (ENIA) for SIC 3523, toiletry products. I was unable to locate comprehensive data on employment in cosmetics firms without factories. 7. For example, one multinational introduced in 1996 its line of mass-market makeup products, after purchasing the leading Chilean make-up company, dissolving its product line and importing the products. By 2000, the multinational was second to a domestic firm in number of units sold and first in value of sales (AC Nielson market study, January–February 2000). 8. For this reason, “envasadora” pay in Table 4.1 reflects the current industry average, after the closing of many multinational factories. 9. The census defines low-skilled and skilled according to the job performed and not whether the worker is involved in production. Nevertheless, there remains a strong correlation between the low-skilled/production versus skilled/production classification. 10. Bacha and Taylor (1978) propose a similar explanation for rising wage inequality in Brazil in the 1960s. In their view, hierarchy in firm, rather than marginal
Labour and the Globalization of Production 103 productivity, determines a worker’s earnings. As managers share in the residual income after workers are paid, they attempt to squeeze laborers’ pay.
References Aitken, Brian, Ann Harrison and Robert Lipsey (1996) “Wages and Foreign Ownership: A Comparative Study of Mexico, Venezuela and the United States,” Journal of International Economics 40 (3–4 May): 345–71. Asociación Gremial de Supermercados de Chile (ASACH) “Participación de Ventas Supermercados de Chile,” 1996–2000. Bacha, Edward and Lance Taylor (1978) “Brazilian Income Distribution in the 1960s: ‘Facts,’ Model Results and the Controversy,” Journal of Development Studies 14 (3): 271–97. Berg, Janine and Dante Contreras (2004) “Political-Economic Regime and the Wage Curve: Evidence from Chile, 1957–1996,” International Review of Applied Economics 18 (2): 151–65. Beyer, Harald et al. (1999) “Trade Liberalization and Wage Inequality,” Journal of Development Economics 59: 103–23. Bravo, David and Alejandra Marinovic (1997) “Wage Inequality in Chile: 40 Years of Evidence,” mimeo, Universidad de Chile. Corbo, Vittorio and Patricio Meller (1982) “Alternative Trade Strategies and Employment Implications: Chile,” in A. Krueger (ed.), Trade and Employment in Developing Countries, Cambridge: NBER. Feenstra, Robert and Gordon Hanson (1997) “Foreign Direct Investment and Relative Wages: Evidence from Mexico’s Maquiladoras,” Journal of International Economics 42: 371–93. Gindling, Thomas and Donald Robbins (2001) “Patterns and Sources of Changing Wage Inequality in Chile and Costa Rica during Structural Adjustment,” World Development 29 (4): 725–45. Koser, Glen (2001) “State of the Industry 2001,” Global Cosmetics Industry 168 (6): 20–30. Larrañaga, Osvaldo (1999) “Distribución de ingresos y crecimiento económico en Chile,” Serie Reformas Económicas 35, Santiago: CEPAL. Meller, Patricio (1992) “Labor Reforms” in O. Muñoz, ed., Economic Reforms in Chile, Ocassional Paper, Washington, D.C.: Inter-American Development Bank. Meller, Patricio and Andrea Tokman (1998) “Chile: Apertura Comercial, Empleo y Salarios,” Santiago: OIT/ETM. Robbins, Donald (1994) “Relative Wage Structure in Chile, 1957–1992: Changes in the Structure of Demand for Schooling,” Estudios de Economía, vol. 21, Santiago: University of Chile. Slaughter, Mathew (2002) “Skill Upgrading in Developing Countries: Has Inward Foreign Direct Investment Played a Role,” in William Milberg (ed.), Labor and the Globalization of Production, this volume. Yin, Robert (1994) Case Study Research: Design and Methods, New York: Sage Publishers.
5 Why Do Firms Disintegrate? Towards an Understanding of the Firm-level Decision to Subcontract and its Implications for Labor Asad Sayeed and Radhika Balakrishnan1
Introduction The post cold-war global economy is characterized by the twin phenomena of disintegration of the production process and the integration of the world economy through trade (Feenstra, 1998). Although the phenomenon of disintegration of the production process has been observed over the last three decades, it has accelerated with trade and financial sector liberalization in the post cold-war world order. In this context, flexible production systems with vertical and at times horizontal disintegration are seen as increasingly profitable. This is especially true with regard to out-sourcing of production to developing countries. Korzeniewicz and Martin (1994) conceptualize this phenomenon as characterized by a new global division of labor. Subcontracting is a relationship where a firm externalizes part of the production or processing of their product to another separate entity but according to the specifications of the firm that is subcontracting its work (Dickens, 1998, p. 230). By subcontracting production outside its legal precincts, firms reduce their labor cost and the transaction cost of monitoring and metering labor. More important is the fact that subcontracting enables the firm to externalize part of its capital and operational costs emanating from the regulatory environment (Appay, 1998, pp. 161–84).2 The phenomenon of subcontracting raises important questions with regard to its impact on employment, wages, working conditions, gender dynamics and productivity. Ong (1997, p. 61) argues that since the 1970s the process of accumulation is typified by flexible labor regimes, which in turn, are increasingly based on female and minority workers in the third world 104
Decision to Subcontract and Implications for Labor 105
and in poor regions of metropolitan countries.3 Therefore, the spatial dispersion of work has largely been accompanied by a decrease in wages, deterioration of working conditions and increasing feminization of the work force.4 Since much of this work has been un-semi skilled, it has drawn in economically and socially oppressed workers, that is, women, children and illegal migrant workers, who are prepared to work for long hours at low wages (Standing, 1989 and 1999). “This trend is most pronounced in the manufacturing industries in low and middle income countries at the early stages of industrialization.5 Feminization is pronounced where gender wage gaps are wide, and there is evidence that as those gaps close, feminization reverses” (Seguino and Grown, 2002, p. 12). This decline in formal employment has also been accompanied with an increase in more insecure subcontracted small unit-based or home-based employment of women workers (Ghosh, 2001). Dispersion of work has also worked as a labor-disciplining device by capital. Segregation of the production process enables firms to evade existing labor legislation and makes it increasingly difficult for workers to organize. Others have argued, on the other hand, that the spatial dispersion of work has an employment and/or productivity enhancing aspect. This body of work suggests that work that is generally contracted out internationally is low skilled. This therefore implies that there is an increase in demand for high-skilled workers in the industrialized countries made possible by labor saving technological change. There has been some evidence that there has been an increase in productivity in the industrialized countries. The wage gap in the industrialized countries has increased and they suggest that there is possibly an increase in relative wages for high-skilled workers in lowincome countries (Feenstra, 2000).6 Firm level decision to go for labor flexibility, it is argued, is driven by productivity enhancement and therefore, in principle, there should be no negative impact on wages.7 The increase in demand for low skilled labor in the third world has not necessarily resulted in an increase in wages of workers in these countries. There has, however, been some evidence in a few countries that women workers in subcontracted industries, particularly those working in small sweat shops engaged in home-based work are new entrants to the labor force and therefore their ability to earn wages can be seen as an improvement in their relative status (Balakrishnan, 2002). In this chapter we seek to specify different conditions that compel previously integrated firms to disintegrate production and the impact that those specific conditions have on wages and working conditions, particularly women workers. In Section 1 of the chapter we revisit the literature on the “make or buy” decision of the capitalist firm, primarily because this literature provides important conceptual tools that are needed to specify different conditions in which firms decide to sub-contract. Revisiting this literature will also
106 Asad Sayeed and Radhika Balakrishnan
provide some clues to changes in economic organization that has resulted in increasing disintegration of the production process. In Section 2 we present a conceptual framework that explores the relationship between long-run profit maximization and subcontracting. We start with the proposition that the firm maximizes its long-run profits essentially through minimizing unit labor costs. A modified unit labor cost equation is introduced which incorporates different transaction costs that are highlighted by the literature on “make or buy” decisions. In Section 3, we specify the ex-post decision of the firm in terms of a push and pull towards subcontracting. Based on the theoretical framework developed earlier we specify the productivity enhancing character of subcontracting (the pull mechanism) and the cost reducing motivation (the push mechanism) along with their attendant impact on wages and working conditions. Section 4 concludes.
1
Make or buy? A brief literature review
The decision of a firm to make or buy raises fundamental questions about economic organization in market economies. Coase (1937) posed the question as to why firms exist in a market economy where prices are expected to provide signals for efficient allocation of resources to all factors of production. Coase’s answer to this question was that the hierarchical structure of the capitalist firm enhances the efficiency of resource allocation. He characterized this phenomenon as an ‘island of conscious power’ in the sea of price-coordinated market exchange. Much of the subsequent literature that seeks to answer the question regarding the existence of the capitalist firm does so through two seemingly opposing explanations. Neoclassical economists have followed Coase in explaining the existence of the capitalist firm in terms of enhancing efficiency of the production process.8 The (neo)Marxists, on the other hand, have argued that the essential purpose of the capitalist firm is to control the labour process.9 See Putterman (1986) for a useful summary on the existence of the capitalist firm that is couched in terms of its efficiency or control characteristics. The efficiency argument has been developed by introducing the notion of transaction costs. Transaction costs are most succinctly defined as the cost of exchange that firms have to undertake to buy inputs from the market and to conduct the production process efficiently (Williamson, 1985). The cost of writing, negotiating, policing and enforcing contracts are some transaction costs in this realm. Transaction costs are, however, also incurred within the firm. Supervision and monitoring of the labour process is an essential transaction cost (Alchian and Demsetz, 1972). Firms, therefore, integrate the production process if the cost of managing market interactions is less than the monitoring cost. Because of the hierarchical nature of the firm the internal transaction costs are generally deemed to be less than those encountered externally. Going by the operational use of transaction costs in explaining
Decision to Subcontract and Implications for Labor 107
the integration of economic activity, Eggertson (1990, pp. 15–16) has shown that these costs can be categorized exclusively as arising out of either high information costs or the existence of asymmetric information.10 The Marxist conceptualization of the control function of the firm has its origins in Marx’s distinction between the spheres of circulation and production where surplus value is created in the production sphere, that is, the firm. Without necessarily invoking the surplus value argument, the neo-Marxists demonstrate the process of labor control through de-skilling as the principal basis of the existence of the firm. The work of Braverman (1974) on the Fordist mode of integrated, assembly line production is taken as the benchmark in much of these analyses. That the hierarchy within the firm and the freedom of market exchange coexist in capitalist economic organization is recognized by exponents of both neo-classical and Marxist positions (see Putterman, 1986, pp. 25–9). Both will also agree that regardless of motivations to integrate production, the bottom line in corporate governance remains the maximization of long-run profits. We, however seek to understand the opposite phenomenon of Why Firms Disintegrate? The purpose of the brief review of the literature on firm integration is helpful in borrowing important conceptual tools used in that literature. Whether firms decide to make or buy is determined by the firm objective of profit maximization. Firms achieve this objective in a milieu where technology and institutions facilitate to reduce transaction costs through disintegrating the production process and/or increasing capitalist control over labor power.
2
The conceptual framework
The long-run objective of the capitalist firm axiomatically is to maximize profits. The decision to make or buy will thus hinge on this bottom line. Profitability can be expressed simply in equation (1) as: P R (Kc UL(Q))
(1)
where P profitability, R total revenue and is the product of price per unit and output, Kc capital consumption, measured through the rate of depreciation on the given capital stock and UL is the modified unit labor cost equation which takes the form. UL W L/Q O
(2)
where W total labor related costs, L total labor employed,11 Q total quantity of goods produced and O per unit overhead economic costs such as utilities, taxes, levels of inventory and non-labor regulatory costs (environmental, health, etc.)
108 Asad Sayeed and Radhika Balakrishnan
The modification in equation (2) from the standard unit labor cost conceptualization is first, the inclusion of overheads (O) and second, instead of taking W as nominal wages, we take the entire cost on labor as W w Nw S
(3)
where w nominal wages, Nw non-wage firm level benefits to workers12 and S cost of supervision and monitoring, including the pay package of those supervisors and managers whose job is to monitor and regulate labor. Moreover, since efficiency is essentially gauged by improvements in productivity (Q/L), equation 2 can be re-written as A R/L UL W/A O
(4)
Going back to equation (1), we assume that the cost of capital is given13, the firm is a price taker in both input and output markets and a range of techniques exist which are more or less labor (capital) intensive. Firms then seek to minimize their unit labor costs. The decision to subcontract is, therefore, made at the point where long run profits from subcontracting are greater than those of producing in-house. Institutional characteristics The decision to make or buy will vary with variation in the institutional environment in which firms operate. One such important institutional characteristic is the ‘formal–informal’ sector divide. When firms disintegrate production within a country, they typically move production out of the ‘formal’ sector to the ‘informal’ sector. The formal sector is characterized by adherence to the legal structure prevalent in a country. For our purposes, this will translate into adherence of labor laws and the payment of taxes. Labor laws usually refer to the existence of a minimum wage and non-wage benefits. Important non-wage benefits are over-time compensation, a holiday schedule, health benefits,14 including maternity leave, housing allowances and bonuses. The right to unionize is also typically granted in most countries for registered firms.15 The “cost” of being formal, therefore is that the total wage (W) in the formal sector is relatively high.16 Compared to that such constraints do not exist in the informal sector. In countries where such distinct sectors exist, it is legitimate to ask as to why all firms and economic activity does not shift to the informal sector? While there are no theoretical answers to this question, we conjecture that beyond a certain capital and turnover threshold it is difficult to evade legality. More substantively, access to formal sector credit (which tends to be less costly than informal sector credit) is only possible for a legally registered firm. The formal–informal divide has to be seen not only in terms of firms and products but also labor markets. Flexibility of work in terms of increasingly
Decision to Subcontract and Implications for Labor 109
insecure employment, limited job mobility and few or no benefits has been observed for all workers. But the incidence of these features for women’s employment greatly exceeds that for men (Seguino and Grown, 2002). This is one important form of labor market segmentation among others.17 In terms of skill levels, wages, working conditions and collective action, workers in the informal sector are ‘poor cousins’ of their formal sector comrades. Moreover, the threat of informalization has also decreased the benefits in the formal sector. When workers in the formal sector know that any demand for increase in wages or improvements in working conditions can possibly make the management react by pushing production outside the firm, job security becomes the paramount consideration.18 The same intra-country logic can be applied to outsourcing of work across countries. Apart form the fact that labor is cheaper in developing countries compared to developed countries, labor rights are institutionalized in greater measure in OECD countries compared to developing countries. Most OECD countries have a legislated minimum wage and provide substantive non-wage benefits to workers across the spectrum.19 Similarly environmental and health related regulations are much tighter in OECD countries compared to developing countries. This difference in labor legislation has shifted work from the north to the south and has decreased the female share of employment in the north. Kucera and Milberg (2000), find evidence of declines in females, share of manufacturing employment in a number of industrialized economies in response to north–south trade. That is, women employed in the formal manufacturing sector in the north have been displaced in response to increased trade with southern countries that are more intensively using women in labor-intensive industries. This important institutional characteristic lies at the cornerstone of subcontracting and outsourcing work when profits are threatened by increasing unit labor costs.
3
The push and pull characteristic of subcontracting
To understand the impact of subcontracting on labor it is important to specify the conditions in which work is subcontracted. Building on the argument above that in a rapidly integrating global financial and trade environment, minimizing unit labor costs is critical for long-run profit maximization. We now move on to analyze different forces which prompt firms to subcontract work. We distinguish situations in which firms are pulled or pushed into subcontracting. Firms are pulled to subcontract when labor costs are reduced through improvements in productivity (or A in equation 4). In contrast, firms can be pushed into subcontracting when unit labor costs are minimized solely through cost minimization (W and O) without any attendant productivity improvements. A pull into subcontracting creates conditions for improvements in returns to labor over time.20 A push into subcontracting,
110 Asad Sayeed and Radhika Balakrishnan
on the other hand, is purely exploitative as one cannot envisage any improvements in wages and working conditions for workers involved in this form of subcontracting.21 Pull into subcontracting The simplest pull mechanism towards subcontracting is the principle of expanded reproduction that states that the division of labor is determined by the extent of the market. As demand for a particular industry using inputs increases, the minimum efficient scale of those products used as inputs increases and leads to that product being manufactured independently (Stigler, 1968). This form of subcontracting is generally associated with capital-intensive, continuous-flow methods22 of production and is amenable to production technologies where economies of scale are central. If economies of scale determine a minimum efficient scale of production which is greater than the need of the parent firm, then it will decide to buy rather than produce. The cost of that particular input reduces, which in turn reduces O and has a positive impact on UL. The above form of subcontracting, however, is different from subcontracting in sectors where technology is generally based on batch production – where production is broken down into distinct and fully contained tasks. Skillintensive sectors dominate this profile. In these sectors, innovations in the division of labor occur primarily due to technological change where general purpose and divisible machinery is brought together with skilled and trained workers. The flexible specialization paradigm á la Piore and Sabel (1984) illustrates this phenomenon. Divisibility of capital enables the large firm to outsource production and reap the gains from economies of specialization. Software development, computer-aided designing, manufacture of automobile parts and electronics are some examples of such activity. Specifically, in the IT industry in India, women are better represented than in the formal sector as a whole though class and caste segmentation are evident. Most of the workers in the industry are educated, English speaking and urban. Though educated women are benefiting from this type of employment, roughly 61 percent of females above the age of seven cannot read and write in India (Ghosh, 2001). High skill levels of workers and a change in their contractual arrangement from an input-based time wage to an output-based piece rate induce this form of subcontracting. In such a situation, skill intensity has a positive impact on labor productivity A, through skill embodied in L. Moreover, the total wage bill (W) reduces because supervision costs (S) are reduced if the work is subcontracted out on piece-rates.23 The nominal wage (w) will depend on the human capital endowment of the work force and the relative ease/difficulty in acquiring skills. If it is easy to impart or acquire requisite skills then (w) may remain depressed or may even decline and vice versa. In the case where (w) increases the only condition is that the increase in (w) should be less than
Decision to Subcontract and Implications for Labor 111
or equal to the reduction in supervision costs (S). By subcontracting work out overhead costs (O) also reduce. A combination of reduction in total wages (W) and overhead costs(O), along with an improvement in labor productivity (A) will lead to a decline in the unit labor cost (UL). Such forms of subcontracting usually takes place within the formal sector. In labor intensive sectors based on batch production and where output can be broken into distinct and fully contained tasks, the supervision cost is reduced through piece rating. But in order to induce workers to perform better, a higher nominal wage (w) is paid.24 The higher nominal wage (w) can have an efficiency wage impact and thus improve work effort, thereby improving labor productivity (A). The impulse to subcontract work is that it reduces overheads (O) and supervision costs (S) while non-wage benefits (Nw) are eliminated. Thus even after an increase in the nominal wage (W), total wages (W) as a whole are reduced. This form of subcontracting is most prevalent in lower value-added exports such as garments, footwear and carpet manufacture. Usually formal sector export firms subcontract such work out to the informal sector. The “efficiency wage” effects (i.e. when higher than market wages are paid to induce and retain workers in these sectors) are also less likely to take place because firms can relocate to lower wage sites as wages increase, even before productivity gains can be captured (Seguino and Grown, 2002). The same results will occur if the supply of woman workers in the informal sector occurs at a lower reservation wage than the rest of the economy.25 In industries where “nimble” fingers are assumed to be needed and/or women can be coerced into working long hours at subsistence wages, labor productivity (A) will improve and at the same time non-wage benefits (Nw), supervision costs (S) as well as overhead costs (O) will reduce. However, such improvements in productivity will be short-lived, simply because reproducing labor power at subsistence wages for women who have to bear the double burden of the care economy has its obvious physical limits.26 This form of pull subcontracting will thus soon degenerate into a push subcontracting (see below) or will move into an efficiency wage like situation described above. The latter two forms of subcontracting arrangements can be observed at the low-end of the export market, such as products sold at Wal-Mart or open market sales in developed countries. This trend can also be seen in buyer-driven commodity chains, which refer to “those industries in which large retailers, brand named merchandisers, and trading companies play the pivotal role in setting up decentralized production networks in exporting countries, typically located in the Third World” (Gereffi, 1994, p. 97). These industries are not manufacturers as such, but merchandisers who farm out all of the production to different agents around the world (Gereffi, 1994). The pull into subcontracting is thus dependent on the ability of the firm to forego the supervision and metering cost. The particular mix of skill intensity, demand for the product and increased work effort as a result of the
112 Asad Sayeed and Radhika Balakrishnan
efficiency wage can create a situation where subcontracting work results in a positive-sum outcome. On the one hand, the firm is able to minimize its unit labor cost and thereby increase profits and on the other hand it creates conditions for the realization of better wages and/or working conditions for skilled and informal sector workers. Of course the realization of better wages and working conditions is relative. If production has moved out of developed countries or the formal sector in developing countries, then the comparison is invalid because of different labor market and institutional characteristics. For developing country informal sector workers, productivity enhancement, ipso facto creates labor market dynamics for improvements in wages and working conditions. The same can be said about skill intensity in outsourcing. As Feenstra (1998) states “… outsourced activities are un-skilled labor intensive relative to those in the developed economy but skilled labor intensive relative to those in the less developed economy”. Thus skills, as well as skill intensity, should be seen on a spectrum.27 Push into subcontracting Push subcontracting is characterized always by a move from the formal (internationally or domestically) to the informal, unregulated product and labor markets. In contrast to the pull towards subcontracting, firms can be pushed into outsourcing in three cases: (i) increase in economic costs (O), (ii) increasing product market competition and (iii) the possibility to circumvent the regulatory environment (both labor and others). Production of consumer nondurables at the lower end of the market segment and much of home-based work are examples of the push into subcontracting. The outsourced work process is usually the least skill-intensive, involving minimal capital outlays, and the labor process is generally repetitive and monotonous. Rather than improving productivity and product quality, reducing costs to survive in the market is the dominant criterion. Much of this push into subcontracting can be explained by technical change and the onset of neo-liberal economic policies instituted in the developing as well as the developed world. In developing countries these policies have resulted in familiar stabilization and structural adjustment programs. In the developed world, a more gradual process of deregulation and dismantling of the welfare state – and thereby social protection – has been observed. Deregulation coupled with technical change has made firms more risk averse in a milieu of increasing competition. As Standing observes: Stimulated by high unemployment, by new technology, by more aggressive international competition (notably from Japan and the newly industrialized countries), by deregulation and the erosion of union strength, and by the desire to overcome the uncertainty induced by the international economic instability, enterprises every where are devising means of reducing the fixed costs of labor. There is a global trend to reduce the reliance on full – time wage and salary workers earning fixed wages and
Decision to Subcontract and Implications for Labor 113
various fringe benefits. Companies and the public sector enterprises in both the developed and developing economies are increasingly resorting to casual or temporary workers, to part-timers, to subcontracting and to contract workers (Standing, 1999, p. 1078). The first case we examine is that of firms being pushed into subcontracting as a response to increasing economic costs of production (O). The price of raw materials increase due to a fast deteriorating exchange rate or high interest rates due to a contractionary monetary policy and/or an increase in the cost of utilities increases because of structural adjustment/stabilization policy packages. Since the overall increase in production costs is beyond the control of individual entrepreneurs, they resort to reducing labor and overhead costs by outsourcing those aspects of the production process wherever it is technologically and administratively feasible to do so. Such outsourcing will not necessarily lead to any productivity enhancement, and the expected impact on wages and working conditions is negative. In terms of our formulation, an increase in overhead and economic costs (O) will lead to an increase in the unit labor cost (UL). Thus the incentive to reduce labor cost.28 Such a move also reduces the overheads of the firm in terms of renting space and electricity but the increase in other elements of (O) will be high. It is clear in this situation that (A) is not affected and the (UL) reducing mechanism operates essentially through a reduction in (W). The second case of push subcontracting is where firms are subjected to excessive levels of price-based competition. At the lower end of the market spectrum,29 where price rather than quality is the determining criterion for capturing market share, cost reduction takes place through subcontracting work, typically to home-based workers. Removal of domestic entry barriers and deregulation of investment has resulted in increasing the level of domestic competition in many developing countries. As a result push subcontracting is a frequent phenomenon in a number of such industries. Reduction in tariffs due to trade liberalization policies can also exacerbate price-based competition. Only a small segment of the third world consumer market corresponds to the quality-conscious western consumer, where niche markets for customized products are increasingly ruling the roost. In such markets, healthy “competition” exists, where quality and design rather than price is the determining factor. Much of the consumerism in the developing world is based on either cheap (read affordable) imitations or “modern” necessities. As tariff barriers come down, intra-third world competition for imitation Gucci handbags or Levi jeans or simply garments or toothbrushes, soaps and slippers intensifies. And since price is the only criterion through which market share is to be captured among these labor-intensive industries, competition ensues between poor countries over whose labor is cheapest. If the Pakistani slipper maker is to survive on the back streets of Karachi selling her product, she has to pay less to her worker than her Chinese or Vietnamese counterpart. The bottom line is thus clearly defined.
114 Asad Sayeed and Radhika Balakrishnan
Niches in this case shift from products to workers. Employers/producers then prowl for women and children of the poorest households, who are usually migrants either from rural areas or from war-ravaged or calamity-hit neighboring countries. Once such segmentation in the labor market is intensified, then the wage rate in the labor market as a whole also drops. If the producers are successful in reducing the overall wage level, then in popular parlance they have attained competitiveness.30 In terms of unit labor costs, the impact of excessive competition is similar to that of increasing overhead costs. The pressure to decrease the price of the product is paramount with quality (measured through supervision costs (S) and labor productivity (A) in equations 3 and 4 above) having little relevance. In this case, work is subcontracted out to eliminate none-wage benefits (Nw) and reduce the nominal wage (W) to the lowest possible level. This form of subcontracting might be done at the level of small sweatshops or home-based work, depending on the level of direct monitoring required. Garments, plastic products and other consumer non-durables are good examples of this. The trend towards small shop and particularly home-based work has been increasing internationally. As Ghosh (2001) points out “in the garment industry alone, the percentage of home workers to total workers has been estimated at 38 percent in Thailand, between 25 and 29 percent in the Philippines, 30 percent in one region in Mexico, between 30 and 60 percent in Chile and 45 percent in Venezuela.” The shift to home-based work has a significant impact on the defeminization of formal sector employment in these sectors. The predominance of women in home-based work has possibly contributed to a shift in workers from factories to lower paid subcontracted work. In Thailand there was evidence that many of the garment factories that closed as a result of the 1997 financial crisis, transferred their production to small shops and homes (Balakrishnan, 2002). The third condition for push subcontracting to emerge is in response to legislation which protects labor rights. Such legislation, prevalent in South Asia and Latin America has resulted in increasing (w) through a minimum wage legislation and by increasing Nw.31 Push subcontracting eliminates Nw and minimizes w. The impact on labor productivity of this push for subcontracting is at best neutral and may even decline. But given the institutional setting, firms are able to lower their unit labor costs. In countries where pro-worker, particularly pro-women workers, legislation has taken place there has been a decrease in the employment of women and there is evidence that the casualization of work through subcontracting has increased (Ghosh, 2002). A similar outcome will emerge when firms seek to evade environmental and/or health related regulations or union power prevalent in OECD countries. They will then farm out production to developing countries where such regulations either do not exist or are weakly enforced.32
Decision to Subcontract and Implications for Labor 115
4
Conclusion
In this chapter we have sought to conceptually specify different conditions in which firms disperse their production activities. Our primary concern has been the impact that this increasingly prevalent mode of industrial organization has on labor in general and gender dynamics of the labor market in particular. By focusing on the long-run profitability concern of firms, we seek to understand the bottom line benefits of decentralized production at the firm level. This framework also offers a window into examining subcontracting at the national level inside developing countries. Most of the theoretical literature in the field has focused on multinational firms subcontracting and integrating developing country firms into the production process. The process of subcontracting is also taking place within national industries in the third world for the national market. By focusing on profitability we intend to capture both the effect of international trade regimes as well as national and international macro economic policy.
Notes 1. Comments on an earlier draft by Elissa Braunstein, Shahrukh Rafi Khan, Farhan Sami, David Gillcrist and Rajeev Patel are gratefully acknowledged. Research help from Hourig Messerlian is also greatly appreciated. We are particularly grateful to Will Milberg for his encouragement and very helpful comments on two earlier drafts. All errors of omission and commission remain ours. 2. Regulatory costs will include compliance with financial, social, and environmental regulation including health and safety related regulations as well as fringe benefits to workers. 3. Ong argues that the world recession in the 1970s compelled capitalists in metropolitan countries to restructure production in the face of rising labor costs and increasing competition and to employ low-paid workers in informal settings. This same strategy is used in developing country contexts as well. 4. The trend toward feminization of employment, that is increased share of women among paid employees, has been hastened as firms compete more intensely to reduce costs, with women’s wages universally lower than that of men. However, there is some evidence to show that women’s share of employment in the formal sector is decreasing. In some regions the move toward informalization is dominated by women. 5. However, the example of East and South East Asia shows that the share of women in employment has fallen in the latter part of the 1990s even before the financial crash of 1997 but particularly after (Ghosh, 2001). 6. See essays in Feenstra, Robert (ed.) 2000. The impact of International Trade on Wages, Chicago: University of Chicago Press. 7. Presumably the argument will be that the loss in terms of foregone non-wage benefits will be compensated by employment generation. 8. Principal exponents of this position are Williamson (1975 and 1985), Alchian and Demsetz (1972), and Jensen and Meckling (1976). 9. Marglin (1974), Edwards (1979), and Bowles (1985) have theorized the phenomenon along these lines.
116 Asad Sayeed and Radhika Balakrishnan 10. Because there is no conception of power in neo-classical economics, non-economic or transaction costs appear to occur only because of information failures. Otherwise agents would be able to negotiate perfect contracts based on their respective utility functions. 11. L will embody human capital as well as asset specificity. 12. Apart from regulation, the existence of unions and their level of effectiveness will influence non-wage benefits. 13. Financial globalization has resulted in increasing convergence of interest rates globally. Coupled with flexible exchange rate regimes and removal of barriers on the flow of financial capital have meant that the actual cost as well as transaction costs have made national boundaries increasingly irrelevant so far as capital mobility is concerned. 14. Maternity benefits are particularly important in this context for women workers. 15. In developing countries, those production facilities that are not registered with state authorities are generally termed as part of the informal sector. By not registering, they remain outside the remit of the legal structure governing production activities. 16. In the case where there is no minimum wage and legislated worker benefits are not substantial but the right to unionize exists, union action can increase W by negotiating higher wage demands and non-wage benefits for workers. 17. In developing countries, labor market segmentation along ethnic or caste lines is commonly observed. 18. A large mutlinational corporation in Pakistan used this tactic to keep its unionized workers from demanding wage and non-wage benefits. 19. There may be small islands of third world like “informal sector” wage and employment conditions, but that is the exception rather than the rule. 20. This of course will depend on a number of factors, such as the degree of segmentation in the labor market, the regulatory environment and the future course of technical change. 21. Given the abundant supply of labor in developing countries an increase in N does not mean an increase in W. 22. Continuous flow technologies are those where the production process cannot be separated and stopping the process in midstream is costly in terms of time. Steel and cement production are some examples of such technologies. 23. S is reduced, as instead of continuous monitoring of work effort input, only discrete monitoring of the output has to be undertaken. 24. Note that this higher nominal wage is paid to workers already in the informal sector. Because they have never become accustomed to Nw, this is an increase in their total real wage. 25. See Standing (1999) where empirically this is seen as a frequent feature in developing countries. 26. This problem can be ameliorated through constant turn over in the labor force as is evidenced in many export-processing zones where workers are generally kept only for a few years and then substituted with other similarly placed workers. Internationally, firms can move from one country with very low wages to another for this purpose. 27. The only defining criterion must be specific training acquired to undertake a specific task. 28. More specifically to eliminate Nw and reduce W. 29. These are goods aimed primarily at the lower income end of the domestic market.
Decision to Subcontract and Implications for Labor 117 30. See Dickens for increase in intra third world trade from mid 1980s to mid 1990s. For example in 1985, less than one-fifth of the Asian Newly industrializing economies were sold within the region whereas by 1994 nearly two-fifths were sold in the region. The intra third world trade has increased. 31. See ILO labor standards to compare the ratification of minimum wage regulations in several countries in these two regions. 32. See Rodrik for discussion of the effect of globalization on labor legislation. There is some evidence that environmental regulations effect firm location, there is political pressure against governments that environmental regulations are too costly. The actual threat of location of industry is highly debated. For an interesting discussion see Eban Goodstein’s “Malthus Redux? Globalization and the Environment” in Globalization and Progressive Economic Policy edited by Dean Baker, Gerald Epstein and Robert Pollin. Cambridge University Press. See also Tony Cleaver 2001. Understanding the World Economy Routledge Press.
References Alchian, A. and Demsetz, H. (1972) “Production, Information Costs and Economic Organization,” American Economic Review 62: 777–95. Appay, Beatrice (1998) “Economic Concentration of the Externalization of labor,” in Economic and Industrial Democracy 19: 161–84. Balakrishan, R. (2002) (ed.), The Hidden Assembly Line: Gender Dynamics of Subcontracted Work in a Global Economy, Bloomfield, Conn. Kumarian Press. Bowles, S. (1985) “The Production Process in a Competitive Economy: Walrasian, NeoHobbesian and Marxian Models,” American Economic Review 75: 16–36. Braverman, H. (1974) Labor and Monopoly Capital, NY: Monthly Review Press. Coase, R. (1937) “The Nature of the Firm,” Economica 4 (6). Dickens, Peter (1998) Global Shift: Transforming the World Economy Third Edition NY: Guilford Press. Edwards, R. (1979) Contested Terrain: The Transformation of the Work Place in the Twentieth Century, NY: Basic Books. Eggertson, T. (1990) Economic Behaviour and Institutions, Cambridge: Cambridge University Press. Feenstra, R. (1998) “Integration of Trade and Disintegration of Production in the Global Economy.” Journal of Economic Perspectives 12 (4). Gereffi, Gary (1994) “The Organization of Buyer-Driven Global Commodity Chains: How U.S. Retailers Shape Overseas Production Networks,” in G. Gereffi and M. Korzeniewicz (eds), Commodity Chains and Global Capitalism. Westport: Preager. Ghosh, Jayati (2001) “Globalisation, Export-Oriented Employment for Women and Social Policy: A Case Study of India,” Working Paper. Jensen, M. and Meckling, W. (1976) “Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure,” Journal of Financial Economics 3: 305–60. Korzeniewicz, Roberto and William Martin (1994) “The Global Distribution of Commodity Chains,” in G. Gereffi and M. Korzeniewicz (eds), Commodity Chains and Global Capitalism. Westport: Preager. Kucera, D. and Milberg, W. (2000) “Gender Segregation and Gender Bias is Manufacturing Trade Expansion: Revisiting the ‘Wood Asymmetry,’” World Development 28 (7) 2000. Marglin, Stephen (1974) “What Do Bosses Do? The Origins and Functions of Hierarchy in Capitalist Production,” The Review of Radical Political Economy 6 (1).
118 Asad Sayeed and Radhika Balakrishnan Ong Aihwa (1997) “The Gender Labor Politics of Post Modernity,” in Lowe and Lloyd (eds), The Politics of Culture in the Shadow of Capital. Durham: Duke University Press. Piore, M. and Sabel, C. (1984) The Second Industrial Divide. NY: Basic Books. Putterman, Luis (ed.) (1986) The Economic Nature of the Firm, Cambridge: Cambridge University Press. Rodrik, Dani (1997) Has Globalization Gone too Far, Institute for International Economics. Seguino, Stephanie (2000) “Accounting for Asian Economic Growth: Adding Gender to the Equation,” Feminist Economics (November). Seguino, Stephanie and Caren Grown (May 2002) “Gender Equity, Growth, and Trade Policy: Are There Win-Win Options?” Working Paper. Standing, Guy (1989) “Global Feminization through Flexible Labor,” World Development 17 (7). Standing, Guy (1999) “Global Feminization Through Flexible Labor: A Theme Revisited,” World Development 27 (3). Stigler, G. (1968) “The Division of Labour is Limited by the Extent of the Market.” Reprinted in G. Stigler, The Organisation of Industry. Homewood, Ill.: R. D. Irwin. Williamson, O. (1975) Markets and Hierarchies: Analysis and Anti-Trust Implications, NY: Free Press. Williamson, O. (1985) The Economic Institutions of Capitalism, NY: Free Press.
Part II Skill, Gender, and Location Bias in Globalized Production
This page intentionally left blank
6 Skill Upgrading in Developing Countries: Has Inward Foreign Direct Investment Played a Role?1 Matthew J. Slaughter
1
Introduction
An important part of globalization in recent years has been the ongoing rise in foreign direct investment (FDI). UNCTAD (2000) reports that from 1979 to 1999, the ratio of world FDI stock to world gross domestic product rose from 5 percent to 16 percent and the ratio of world FDI inflows to global gross domestic capital formation rose from 2 percent to 14 percent. One consequence is that an increasing share of countries’ output is accounted for by foreign affiliates of multinational enterprises (MNEs). The foreign-affiliate share of world production is now 15 percent in manufacturing and other tradables (Lipsey et al., 1998). How do these multinational affiliates influence host labor markets in developing countries? In this chapter, I offer some insights on this question are offered by examining the issue of “skill upgrading,” which will be defined both in terms of labor demand and supply. Each side of the labor market will be addressed in turn. On the demand side, the academic literature on multinationals suggests several channels by which inward FDI stimulates demand for more-skilled workers in host countries. These include technology transfer to host-country affiliates; technology flows – both market-mediated and via spillovers – to hostcountry firms; and investments in physical capital related to new technologies. Both the theoretical concepts and empirical evidence on these various channels will be discussed in this chapter. There is compelling evidence on the importance of within-firm technology transfer and capital investment as modes of boosting host-country demand for more-skilled workers. The evidence is much more mixed on technology flows to domestic firms, particularly 121
122 Matthew J. Slaughter
via spillovers. But contrary to what is commonly assumed, it is argued that lack of spillovers is not necessarily a bad thing in light of the stronger evidence on the roles for within-firm technology transfer and capital accumulation. On the supply side, the question of how inward FDI influences the development of human capital is much less clear. Two different modes by which MNEs can facilitate investments in human capital are distinguished. One is the short-term, firm-level activities by which individual firms interact with host-country labor markets through on-the-job training, support for local educational institutions, and the like. The other is long-term, country-level activities by which MNEs collectively contribute to the overall macro environment in which fiscal policy drives education policy. To the extent that MNEs contribute to a good macro environment in host countries – through raising worker productivity, providing a relatively stable source of foreign capital, paying host-country taxes – they contribute to the ability of host countries to fund education. If successful generalizations of firm-level educational initiatives may be hard to come by, as these efforts continue sight should not be lost of the country-level contributions. After laying out these supply-and-demand issues, some new empirical evidence on the links between inward FDI and within-industry skill upgrading for a country–industry–year panel that spans manufacturing disaggregated into seven industries for 16 countries – developed and developing – from 1982 through 1990 have been presented in this chapter. To construct this panel I combined information on host-country inputs and outputs with information on the presence in these countries of affiliates of US-headquartered multinationals has been combined. The main empirical finding is a robustly positive correlation between skill upgrading and the presence of US affiliates, with this correlation even stronger among the sub-sample of developing countries. This correlation is consistent with inward FDI stimulating skill upgrading in these developing countries. The rest of this chapter is organized as follows. Section 2 addresses the theory and empirical evidence on how MNEs affect the demand for skills in host countries. Section 3 turns to the supply-side issues. Section 4 presents the new empirical evidence, and Section 5 concludes.
2
Multinationals and the demand for skilled labor
Theoretical links In what ways does the nationality of ownership influence the demand for labor of firms in developing countries? Note that if there were no such ways, then MNEs would merit no particular attention when thinking about skill upgrading. There is widespread agreement among researchers in many fields that a distinguishing feature of these firms is their possession of knowledge assets – patents, proprietary technology, trademarks, and so on – that can be deployed in plants outside the parent country. This knowledge intensity is important for understanding the nature of MNE labor demand in host countries.
Skill Upgrading in Developing Countries 123
From an industrial-organization perspective, Dunning (1981) formalized a framework in which MNEs are firms possessing three particular sets of advantages, known together as “OLI.” First is the ownership advantage, that is, the ownership of a firm-specific asset. Second is the location advantage, that is, it must be cost efficient for the firm to exploit that asset abroad rather than in just the home country. And third is the internalization advantage, that is, the firm must be better off using its asset itself rather than contracting with another independent firm. In international trade, over the last two decades there has been substantial progress in modeling multinational firms in general equilibrium. This theoretical literature contains mostly uni-dimensional theories of multinationals, which focus on either horizontal or vertical FDI. The vertical-FDI view is that multinationals arise to take advantage of international factor-price differences.2 Firms engage in two activities: headquarter services to develop and maintain the firm’s knowledge assets, and production of output. Headquarter services are intensive in physical and human capital, while production is intensive in manual labor. When factor prices differ across countries, firms become multinational by locating production in countries where manual-labor costs are low and headquarters in countries where skilled-labor costs are low. Even though these production activities may be low-skill intensive relative to headquarter services, for host countries they likely will be skill-intensive relative to their initial activity mix. The horizontal-FDI view is that multinationals arise because trade barriers make exporting costly.3 The formal setup is one in which firms have a highfixed-cost headquarters and one or more production plants. When trade costs are low, a firm produces all output in domestic plants and serve foreign consumers through exports. When trade costs are high, a firm becomes multinational by building production plants at both home and abroad, each serving just that country’s consumers. This type of FDI is called horizontal because the multinational does the same activities (here, production) in all countries. One tell-tale sign that MNEs are knowledge-intensive firms is their intensity of research and development (R&D). In the aggregate, evidence consistent with this is the overlap between countries that perform lots of R&D and countries that headquarter lots of MNEs. It is commonly calculated that approximately 90 percent of the world’s R&D is performed in just five countries: the United States, the United Kingdom, France, Germany, and Japan (Keller, 2001). These five countries are also among the largest source countries for world FDI flows. At the firm level, Slaughter (1998) reports that over the past 20 years the US parents of US-headquartered MNEs – only 2727 firms in 1994 – have consistently performed over half of all US R&D. Having established that MNEs tend to be very knowledge-intensive firms, we can now elaborate on how this knowledge intensity can help raise hostcountry demand for skills. There are at least three important channels to identify.
124 Matthew J. Slaughter
One is technology transfer. The simple idea here is that MNE use of knowledge assets often entails the transfer of technology from parents to affiliates. Inward FDI, then, can mean new production technologies for the host country, which in turn can boost demand for more-skilled labor within host-country affiliates to the extent that the innovations are skill-biased. This can occur whether the inward FDI is horizontal or vertical in nature. In either case, FDI expansion is likely to entail skill-biased technological change (SBTC).4 A second demand channel is that these new technologies may also reach domestically owned firms in host countries. This may happen through market-mediated arrangements such as patent licensing, in which domestic firms pay MNEs for the right to use their technologies. New technologies may also reach domestic firms thanks to “productivity spillovers” via nonmarket channels. Either way, the presence of inward FDI may stimulate domestically owned firms to demand more and more-skilled labor. Spillovers of knowledge from affiliates to domestic firms are an oftenclaimed benefit to inward FDI, so it is worth outlining possible spillover channels. The general idea that interaction among firms can generate spillovers dates back to at least Marshall (1920). Caves (1974, 1996) has had an early and ongoing interest in analyzing this possibility for multinationals interacting with host-country firms. Mansfield and Romeo (1980) present some early survey evidence in which US multinationals reported the frequency and pace at which their technology deployed in foreign affiliates reached host-country competitors, all consistent with multinational spillovers. Theoretical work on the mechanics of spillovers ranges from general discussions, often leavened with anecdotes, to formal general-equilibrium models. Broadly speaking, spillovers are commonly hypothesized to fall along industry or regional lines. An example of multinational spillovers along industry lines is Rodriguez-Clare (1996), in which affiliates increase a host country’s access to specialized varieties of intermediate inputs, the improved knowledge of which raises the total factor productivity (TFP) of domestic producers. Less formally, it is often hypothesized that domestic firms learn from affiliates in the same industry via a range of informal contacts: for example, trade shows; supplier/distributor discussions; exposure to affiliate products, marketing, and patents; technical support from affiliates; reverse engineering. Depending on how narrowly or broadly industries are defined, if the key contacts are between suppliers and/or distributors, then spillovers may be classified as intra-industry or inter-industry.5 Other spillover mechanisms may operate along regional lines. One commonly proposed avenue (since at least Marshall, 1920) is via labor turnover. If at least some of the knowledge particular to foreign affiliates is embodied in their labor force, then as affiliate employees leave to work for domestic firms this knowledge may move as well. For example, Song et al. (2001) use
Skill Upgrading in Developing Countries 125
US patent records to trace the movement of scientists between domestic and foreign firms (also see Motta et al., 1999; Moen, 2000). This knowledge need not be firm-specific (e.g. inventory-control or management techniques). If inter-regional labor mobility within a country is low, then these spillovers are likely to be concentrated within regions where the affiliates operate rather than dispersed country-wide. More generally, regional labor-market spillovers can be thought of as one important kind of agglomeration economy that can induce firms to locate near each other in space. Krugman (1991) offers some formal models of agglomeration issues.6 A third channel for boosting host-country skilled-labor demand, for both foreign and domestically owned firms alike, is capital investments. Implementing new technologies often entails making new capital investments (e.g. computers and office products). To the extent that capital and skills are complements in firms’ factor demands, skill upgrading may arise not just directly from new technologies but also indirectly from the capital investments induced by these new technologies. Empirical evidence on these links For the empirical evidence on how MNEs influence the mix of host-country labor demand, consider in turn each of the three channels – within-MNE technology transfer, affiliate to local technology transfer, and capital investment. An important implication of within-MNE technology transfer from parents to affiliates is that, relative to host-country domestic firms, this transfer and/or its resultant boost in demand for skilled workers should lead affiliates to pay higher wages. This implication enjoys a lot of empirical support. Many studies – of both developed and developing – have found that establishments owned by multinational firms pay higher wages than domestically owned establishments, even controlling for a wide range of observable worker and/or plant characteristics such as industry, region, and overall size.7 To the extent that production technology is largely unobservable in these studies, the regularity of this “multinational wage premium” may stem from the superior technology and thus labor-demand mix of these firms.8 More direct evidence on the transfer of technology and resultant labordemand mix for MNEs can be obtained from data on US-headquartered MNEs. Since the late 1970s, the Bureau of Economic Analysis (BEA) within the US Department of Commerce has collected data on both the US and foreign operations of these companies. One piece of evidence consistent with rising within-firm technology transfer is the rising share of MNE-wide R&D performed by foreign affiliates. In 1982 affiliates performed 6.4 percent of worldwide R&D for these firms; by 1994 that share had nearly doubled, to 11.5 percent. If one role of R&D is to facilitate technology transfer, then this rising R&D share suggests rising technology transfer. As for the labor-demand mix, in its census years the BEA requires foreign affiliates in manufacturing to distinguish non-production from production
126 Matthew J. Slaughter Table 6.1 The skill mix of employment in foreign affiliates of US-headquartered multinationals Country
1977 P Emp
1977 NP Emp
1977 NP Share
1994 P Emp
1994 NP Emp
1994 NP Share
World Brazil India Malaysia Mexico Philippines Singapore Taiwan
2 370.0 179.5 19.3 19.0 103.1 44.7 27.4 45.3
1 400.0 116.7 14.9 6.7 68.1 16.4 5.1 13.0
0.37 0.39 0.44 0.26 0.40 0.27 0.15 0.22
2 093.4 133.9 8.2 67.3 284.2 29.1 44.5 22.3
1 422.6 91.4 7.6 42.9 131.0 23.1 24.6 15.7
0.41 0.41 0.48 0.39 0.32 0.44 0.36 0.41
Notes: “P Emp” indicates production–worker employment, in thousands. “NP Emp” indicates nonproduction-worker employment, in thousands. “NP Share” indicates the share of total employment accounted for by nonproduction workers. All data come from the Bureau of Economic Analysis.
employment. Following a number of studies in the trade-and-wages literature, one can define the former to be more-skilled and the latter to be less-skilled.9 Table 6.1 reports the skill mix of affiliate employment in 1977 and again in 1994 (the most recent year for which these data are available) for the overall world and for a set of prominent developing countries. The key message of Table 6.1 is a widespread shift in the skill mix of affiliate employment. In 1977, manufacturing affiliates of US MNEs employed 2.37 million production workers worldwide. By 1994 this number had fallen to just 2.09 million. Over that same period non-production employment actually increased slightly, from 1.40 million to 1.42 million. This means that the skilled-labor share of total manufacturing employment in affiliates has been rising, from 37.2 percent to 40.5 percent. This rise is matched in all individual countries in Table 6.1 except Mexico, regardless of whether the overall levels of affiliate employment was rising or falling. To put these share changes in context, during this same period in the United States the skilled-labor share of employment across all manufacturing plants rose from 26.1 percent to 30.0 percent. This was a period of widespread SBTC in US industries (e.g. Haskel and Slaughter, 2002), yet the share increase in the United States is only slightly larger than that in US affiliates. And this share increase in affiliates does not simply reflect shifting relative size among industries of different but constant skill intensities. Slaughter (2000) reports that for the standard 32 BEA industry groups within overall manufacturing, 24 had rising skilled-employment shares over the period covered in Table 6.1.10 The shift in relative employment in Table 6.1 is suggestive of technology transfer that stimulates demand for more-skilled workers. More generally, this rising within-affiliate relative employment of more-skilled workers has been
Skill Upgrading in Developing Countries 127
widely documented in many countries – both developed and developing – in recent decades (e.g. Berman et al., 1998; Berman and Machin, 2001; Haskel and Slaughter, 2002) such employment shifts in the face of flat or rising relative wages for more-skilled workers is commonly cited as evidence consistent with SBTC.11 Taken together, this evidence on affiliate wages, R&D, and skill-mix is all consistent with the idea that affiliates stimulate demand for more-skilled workers thanks to technology transfer from the parent firms. Hold for now the question of whether this technology somehow reaches other firms in host countries, and turn first to the issue of capital accumulation. Many studies have documented how new technologies are often embodied in new capital goods (as opposed to simply changing production techniques for unchanging capital goods). The most prominent recent example of this is the recent surge in innovation in information and communication technology (ICT) products: desktop and laptop computers, fax machines, pagers, cell phones. In turn, there is also a large body of evidence (see the survey in Hamermesh, 1993) that capital investment stimulates firms’ demand for more-skilled workers. Data on US MNEs are again consistent with this. From 1977 to 1994, the affiliate share of worldwide capital within these firms rose from about 19 percent to 23 percent. Within manufacturing, the affiliate capital share rose in 22 of the 32 primary BEA industries. The coincidence of affiliate capital deepening and shifting relative employment is consistent with capital–skill complementarity related to technology transfer. More generally, to the extent that FDI involves, by definition, host-country capital accumulation by foreign-owned firms, it is not surprising that affiliate expansion should be a force raising demand for more-skilled workers. What about the transfer of technology from affiliates to domestic firms in host countries? Existing evidence on whether there are productivity spillovers is of three types. The first are case studies. Cases can offer rich description about episodes and exemplify general issues, but they do not always offer quantitative information and do not easily generalize. Second, there are industry-level studies (e.g. Caves, 1974; Blomstrom, 1986). Many have documented a positive correlation between FDI inflows and productivity. But the causal meaning of this industry-level correlation is unclear. It may be that inward FDI raises host-country productivity via spillovers. It may also be that inward FDI raises host-country productivity by forcing the exit of low-productivity domestic plants, or simply by raising the market share of foreign firms who are, on average, more productive. Or it may be that multinationals tend to concentrate in high-productivity industries. This latter interpretation is consistent with the knowledge–capital models of multinational firms outlined above. The third set of studies are micro-level analyses. These studies examine whether the productivity of domestic plants (or firms) is correlated with FDI
128 Matthew J. Slaughter
presence in the industry and/or region of the domestic plants. These microlevel studies are the best suited for identifying productivity patterns consistent with spillovers that industry-level studies cannot. For developing countries, however, there is very little micro evidence supporting knowledge spillovers. Haddad and Harrison (1993) find increased industry-level FDI is correlated with lower, not higher, domestic-plant productivity in Moroccan manufacturing plants. Aitken and Harrison (1999) find the same negative result for Venezuelan manufacturing. For developed countries the results are more mixed. Chung et al. (1998) report no evidence that Japanese automobile firms operating in the United States boosted the productivity of their American component-supplier firms via technology spillovers. Haskel et al. (2002) report some of the strongest micro-level evidence of FDI spillovers. For a panel of plants covering the entire UK manufacturing sector from 1973 through 1992, they estimate a significantly positive correlation between a domestic plant’s TFP and the foreign-affiliate share of activity in that plant’s industry. Why is the evidence on FDI spillovers so mixed? One possible explanation is the pro-competitive effects of affiliate operations. It may be that foreign entrants take domestic firms’ market shares as they stimulate product– market competition, and thereby force domestic incumbents up their averagecost curves. This argument is consistent with Baily and Solow (2001), who survey a wide range of micro evidence that international competition of many forms – including both FDI and trade – tend to spur competitive responses in exposed firms. Another explanation may be that only certain kinds of inward FDI – for example, greenfield construction of new facilities rather than mergers and acquisitions – matter for generating spillovers. Still another explanation may simply be that domestic firms in developing countries do not have sufficient absorptive capacity to realize knowledge transfers from affiliates. But I know of only two micro-level spillover studies for developing countries, so it is unclear how generalizable these two data points are. This will be an important point of consideration when evaluating the empirical evidence in Section 4. Summary of empirical evidence and policy implications Consistent with standard models of MNEs, there is compelling evidence that affiliate demand for skilled labor is stimulated by their receipts of parent technology and their investments in physical capital. Purely through a compositional shift, then, more inward FDI can raise host-country demand for skilled workers. The evidence is mixed, however, as to whether affiliates also stimulate this host-country demand via technology transfer to domestic competitors. The ambiguous evidence, at best, on knowledge spillovers from foreign to domestic firms may strike some as unfortunate. Such spillovers are an oftentouted benefit of inward FDI. It is important to emphasize, however, that
Skill Upgrading in Developing Countries 129
externalities of this kind are, by definition, market failures. In theory, if profit-maximizing MNEs are aware of their ability to generate spillovers, then their operational decisions may be endogenous to this possibility and thus may try to minimize spillovers’ benefits to competitors. The survey evidence on MNE expansion strategies in Mansfield and Romeo (1980) supports this minimization argument. For example, US MNEs report they transfer older technologies to affiliates that are joint ventures – thus where knowledge spillovers are more likely – than they do technologies to wholly owned affiliates. The analysis of Shaver and Flyer (2000) is also consistent with this story. They argue that when firms vary in their inherent technology abilities and other measures of firm performance, then these firms differ in the net benefits they realize from agglomerating near each other. “Best practice” firms have the least to gain and the most to lose from clustering: few other firms can offer them new ideas, yet their good ideas can benefit a large number of other firms. Thus, with heterogeneous firms agglomeration may be characterized by adverse selection, where the firms with the most to offer by clustering have the least incentive to do so. Their analysis of location choices of greenfield investments into the United States supports this idea: betterpractice foreign plants, proxied by measures such as size, were less likely to locate near domestic firms. An important policy implication of this endogeneity of knowledge spillovers is that host-country policies that aim to encourage knowledge transfer can have the paradoxical effect of aggravating rather than solving the underlying market failure – and thereby of reducing, not enhancing, the host-country benefits of foreign presence. The recent work by Moran (2001) makes precisely this point. He carefully examines two industries with extensive global activity in FDI, automobiles and computers/electronics. For each industry he distinguishes two types of host countries. One is those that allow parents to maintain tight control over affiliate operations and thereby allow affiliates to be integrated into MNEwide production networks as the firms see best. The other is countries that impose relatively stringent and/or widespread performance standards on affiliates – for example, ownership caps, domestic-content requirements, and various technology-sharing mandates. Moran’s (p. 32) description of the latter group presents a striking set of differences between the two types. The implications for the development prospects of the host are not favorable. Resources are wasted. Not only are host country consumers penalized, but so too are host country producers that rely on the use of the resulting goods and services to establish their own competitive positions in the marketplace … the plants utilize older technology, and suffer lags in the introduction of newer processes and products in comparison to wholly owned subsidiaries without such requirements. At considerable
130 Matthew J. Slaughter
variance with the dynamic infant industry perspective, the plants are locked systematically into a position well behind the cutting edge of the industry. Put differently, there is compelling evidence that inward FDI brings new technology and capital investment to host countries within the boundaries of affiliate operations. The evidence that this technology spills over to domestic firms is much more mixed. But one should not automatically assume that more of the latter would be better, because in general equilibrium it may come at the cost of less of the former. Policy makers need to keep this in mind. If policy makers care only about raising aggregate productivity, then they should be indifferent about the nationality of ownership of their more-productive firms. Let me offer two examples of this point. First is a country, Ireland. Ireland enjoyed a booming economy throughout most of the 1990s, which has been widely touted as driven in large part by a surge in inward FDI – and thus in technologies and capital investment – concentrated in high-technology sectors like computers and pharmaceuticals. Today, foreign affiliates of US firms account for about 16 percent of total Irish GDP, and within Irish manufacturing foreign affiliates of all firms account for about 75 percent of employment. Does it matter for Ireland whether its surge in output related to strong technology and investment gains has been largely or even entirely within the boundaries of foreign affiliates operating there? By extension, does it matter for any other country either? My second example is an industry, ICT products. In recent years ICT industries have been central to the aggregate economic performance of the United States and other countries. For the United States, about two thirds of the acceleration since 1995 in labor-productivity growth is accounted for by the combination of the production and the use of ICT products (Slaughter, 2002). Prominent ICT-producing firms such as Microsoft and Intel are widely regarded as worldwide best-practice firms, and policy makers worldwide profess their intention to host a rising share of worldwide ICT activity. It is instructive to note the large and rising role for MNEs in these ICT sectors. Table 6.2 offers some evidence on this. For two key ICT industries, industrial machinery and electronic goods, it reports the share of total US sales accounted for by the sales of goods of US parents of MNEs whose main line of business is in those industries. Shares are reported for 1982, 1989, and 1996; similar shares for overall manufacturing outside these ICT industries are also reported.12 Over the 1980s and into the 1990s, US parents of MNEs account for over 60 percent of total US sales in these two prominent ICT industries. Moreover, the importance of these industries has generally been rising over time. In machinery this sales share rose from 54.8 percent in 1982 to 62.2 percent in 1996. In electronics this sales share actually declined over the 1980s, but
Skill Upgrading in Developing Countries 131 Table 6.2 The share of US sales in ICT industries accounted for by the US parents of US-headquartered multinationals Industry
1982
1989
1996
Machinery Electronics Other manufacturing
54.8 73.2 45.0
58.6 66.6 49.1
62.2 77.6 49.3
Notes: Cell entries report the share (in percentage terms) of each industry’s US sales accounted for by the sale of goods of US parents of US-headquartered multinationals whose main line of business is that relevant industry. Other manufacturing is all manufacturing less machinery and electronics.
surged in the 1990s from 66.6 percent to 77.6 percent. This prominent presence for US parents in these industries is far larger than their presence in the rest of manufacturing. The parent sales share for other manufacturing rose from 45 percent in 1982 to 49.3 percent in 1996. And during the 1990s this share was virtually unchanged in the rest of manufacturing, while it was rising substantially in the two ICT industries. All this suggests that MNEs account for a sizeable share of total US ICT activity, a share which has been both rising over time – particularly over the 1990s – and which appears larger than in most other industries. It is also of interest to know how prominently foreign affiliates appear in the worldwide activity of these firms. Do MNEs in ICT industries look more global than those in other industries in terms of affiliates being a higher share of firmwide activity? Data answering this question are in Table 6.3. This reports the share of worldwide firm value added and employment accounted for by majorityowned foreign affiliates. These shares are reported for 1982, 1989, and 1997 for machinery, electronic goods, and all industries together. Table 6.3 shows that in 1997, foreign affiliates in these central ICT industries accounted for between 26 and 40 percent of worldwide firm value added and employment. These shares were generally rising by several percentage points over the 1980s and 1990s. They were also uniformly higher by 1997 than for the broad economy, where the increases were generally smaller. The global presence of these ICT industries involves not just high-income but many middle- and low-income countries as well. The McGraw Hill Companies et al. (2000) reports that in several central ICT industries, several low-income countries such as Mexico, Malaysia, Philippines, and South Korea are large exporters running net trade surpluses. This is consistent with global production networks in which these countries tend to import ICT intermediates, add value to these intermediates, and then export them on to additional countries. There are also compelling studies of international
132 Matthew J. Slaughter Table 6.3 The share of worldwide activity of US-headquartered multinationals accounted for by foreign affiliates Industry
Activity measure
1982
1989
1997
Machinery Electronics All industries
Value added
28 14 22
39 18 23
38 26 25
Machinery Electronics All industries
Employment
26 23 21
34 26 21
40 33 25
Notes: Cell entries report the share (in percentage terms) of worldwide activity of US-headquartered multinationals accounted for by foreign affiliates (where data are available for majority-owned affiliates only).
production networks for very specific activities within ICT sectors – for example, McKendrick’s et al. (2000) coverage of producing hard-disk drives. A generation ago these disk drives were physically produced in the United States, but since then production has migrated to lower-cost regions in various lower-income countries primarily in Southeast Asia. Consistent with all this, Hanson et al. (2001) report that from 1982 through 1998, for affiliates of US multinationals the fastest growing industry–region combination was computers/office products in Southeast Asia. It is also important to emphasize that for many producers of ICT products, foreign customers may be served much more effectively through foreign affiliates rather than exports. This may be particularly true for ICT services, many of which require firms to interact on-site with customers. Affiliates of MNEs, then, can also figure prominently in terms of serving foreign markets. Table 6.4 demonstrates this predominance of foreign-affiliate sales over U.S. exports for the key ICT industries of computer services, data-processing and network services, and electronic-information services. For these industries, this table reports both total foreign sales by majority-owned affiliates and total US exports in three years over the 1990s–1992, 1994, and 1998.13 Affiliate sales were about eight times larger than exports in 1992, and by 1998 this gap had grown to nearly 20 times. This shows that for many ICT services, foreign affiliates have become an increasingly important channel for serving foreign markets. So not only do the US parents of MNEs account for a high and rising share of US activity in central ICT industries (Table 6.2), but within these firms in these industries a high and rising share of total activity (Table 6.3) and total foreign-market sales (Table 6.4) is accounted for by their foreign affiliates. Together, all this suggests that MNEs mediate an important share of total world ICT activity. Again, from the standpoint of policy makers interested in attracting ICT firms, does this matter? If these firms choose to minimize knowledge spillovers – a plausible assumption for such information-intensive
Skill Upgrading in Developing Countries 133 Table 6.4 The relative importance for ICT services industries of US exports versus sales by foreign affiliates of US-headquartered multinationals Year US exports Affiliate sales
1992
1994
1998
1 417 11 545
2 332 25 859
3 412 65 056
Notes: Cell entries report the value of either total US exports or total affiliate sales, in billions of current dollars, for the ICT industries of computer services, data-processing and network services, and electronic-information services (where data cover majority-owned affiliates only).
sectors – does that mean that host countries should not try to attract them? Ideally, with micro panels covering many industries like ICT in many countries, more evidence could be brought to bear on these questions; see Section 4.
3
Multinationals and the supply of skilled labor
On the supply side, the question of how inward FDI influences the development of human capital is much less clearly answered. This link is, correctly, at the center of this book, as not a lot is known about it. I will distinguish two different modes by which MNEs can facilitate investments in human capital. Multinationals and the short-term, firm-level supply of labor MNEs can facilitate investments in skilled labor through the short-term, firm-level activities in which individual firms interact with host-country labor markets through on-the-job training, support for local educational institutions, and the like. MNEs might directly affect labor supplies, as their transferred knowledge might boost the skills of their employees (and, with spillovers, the skills of domestic employees as well). They might also indirectly affect labor supplies, for example, by influencing the educational infrastructure of host countries in terms of curriculum choices and vocational training. For example, Hanson (2000) reports that Intel recently chose to establish a large assembly and testing facility in Costa Rica, in part thanks to Costa Rica’s agreement to expand high-school training in electronics and English. There is recurring discussion of the “skills gaps” multinationals encounter in host-country labor markets. Knowing how individual firms try to overcome these gaps may hold lessons for the educational initiatives of hostcountry governments.
134 Matthew J. Slaughter
I offer two related points. First, in the training literature it is well documented that educational initiatives by firms tend to be for firm-specific skills, not general skills (see Lynch, 1992). This focus on firm-specific skills is understandable in light of the inability of firms to capture the returns on investment in general skills. Second, the knowledge of MNEs is often times of competitive value. Government initiatives to have this information flow beyond affiliates may have unintended consequences, as outlined in Section 2. Taken together, these two points do not convey that individual MNEs cannot engage the institutions of host-country labor markets to help build skills. But they do indicate that the methods of MNE human-capital development are likely to often be firm-specific rather than aimed at general human-capital issues of numeracy, literacy, and problem-solving. Multinationals and the long-term, national supply of labor The other way in which MNEs can facilitate human-capital development is a set of long-term, country-level activities by which MNEs collectively contribute to the overall macro environment in which fiscal policy can support education policy. To the extent that MNEs contribute to a good macro environment in host countries, they contribute to the ability of host countries to fund education. First, MNEs foster skills acquisition economy-wide to the extent that their affiliate activities of technology transfer and capital investment boost demand and thus wages for skilled workers. These labor-demand drivers were discussed in Section 2. Economy-wide, if MNEs contribute to rising demand and wages for skilled workers, then over the long-run they contribute to the general-equilibrium incentive of individuals in host countries to acquire skills through education and/or training. If individuals in host countries have access to these methods of skills acquisition, then they should respond to the price signals coming from the labor market. Second, the rise in economic activity from MNE affiliates means a rise in host-country tax revenue (whether taxes are levied on labor, capital, or both). This broadening of host-country tax bases can allow greater government investment in education and training. Of course, FDI output and taxes therefrom do not automatically imply greater investment in human capital. But FDI output and taxes therefrom do free up budget constraints and thereby make possible these greater investments. This broadly accords with the recent findings of Dollar and Kray (2000), who document for a large set of developing countries that overall economic growth tends to coexist with growth in incomes for countries’ poorest groups. Third, FDI inflows can improve not just the level of host-country economic activity but also its volatility. Many developing countries rely on foreign capital to help fund domestic investment opportunities. Table 6.5 (from World Bank, 2000) reports the composition of net capital flows into developing countries over the 1990s. One prominent fact is the declining
Skill Upgrading in Developing Countries 135 Table 6.5 The relative importance of FDI in developingcountry net capital inflows Year
1990
1995
1997
1999
Total Official Private FDI Equity Debt
98.5 55.9 42.6 24.1 2.8 15.7
257.2 53.9 203.3 105.0 36.1 62.2
343.7 39.9 303.9 170.3 30.2 103.4
290.7 52.0 238.7 192.0 27.6 19.1
Notes: Cell entries report the value of various kinds of developing-country net capital inflows, in billions of US dollars. These data come from World Bank (2000).
relevance of official aid flows, whose share of the total fell from nearly 60 percent in 1990 to under 20 percent in 1999. A second prominent fact is that within private flows, FDI has grown in both absolute and relative importance. By 1999 FDI accounted for about two thirds of total capital inflows and nearly 80 percent of private inflows. A notable feature of FDI relative to other forms of capital flows is its low volatility. For most of the world’s developing countries over the 1990s, yearon-year variation in FDI flows has been much lower than in equity and debt flows. Table 6.5 shows this to be the case during the second-half of the 1990s with the run up and subsequent crash down of debt financing and, to a lesser extent, equity flows. In contrast, FDI flows grew steadily over the decade. This pattern in Table 6.5 has been documented in many studies. For example, World Bank (1999) reports that for a sample of 21 developing countries from 1978 through 1997, FDI inflows were less volatile (in terms of sample coefficient of variation, as a share of GDP) than were non-FDI capital inflows. Similar evidence can be found in Reisen and Soto (2001). All this means that over time, for many countries a rising share of their total international capital inflows have been of relatively stable FDI. From the standpoint of macroeconomic policy, stable capital inflows are much easier to manage. Accordingly, these FDI inflows help foster macroeconomic stability in which educational investments can better flourish. Again, macro stability – like tax-revenue growth – may not be a sufficient condition for FDI to stimulate human-capital development. But it may, again, be a necessary condition. A fourth issue is that FDI inflows may inhibit “brain drain.” In many developing countries, an ongoing policy concern is the loss of highly educated natives to employment opportunities abroad (either as these people get education locally and then emigrate, or as they get education abroad and then do not return home). To the extent that FDI inflows bring to host countries those attractive employment opportunities, they may inhibit the brain drain.14
136 Matthew J. Slaughter
Again, consider the example of Ireland. The 1990s boom, due in large part to the inward FDI surge, is widely perceived as having boosted demand for skilled Irish workers – with a resulting surge in labor supply driven largely by reverse migration of young Irish back into the country from locations like England and the United States. Over the 1990s the Irish labor force rose by about 60 percent, with a commensurate rise in the population from 2.8 million in 1961 to 3.8 million today (Brumley, 2001). For several decades before the 1990s, annual net emigration out of Ireland was about 35 000 per year. During the 1990s this reversed to net immigration of about 45 000 per year, of which the majority were Irish returnees. Of course, the Irish experience may be somewhat unique, but is exemplifies well the general idea of the interaction between inward FDI reverse migration. Summary of multinationals and labor supply To the extent that successful generalizations of firm-level educational initiatives may be hard to come by as these efforts continue, sight should not be lost of the country-level contributions that MNEs can make to humancapital development. This is particularly true for policy aimed at longer time horizons. High MNE wages as an important labor-market signal, higher and more-stable macroeconomic growth and tax revenue, and reduced emigration incentives: through all these channels MNE affiliates can stimulate acquisition of long-term skills in host countries.
4 New empirical evidence on inward FDI and developing-country skill upgrading Sections 2 and 3 discussed the possible mechanisms by which inward FDI can stimulate skill upgrading in host countries. This section offers some new empirical evidence on this FDI-to-skill upgrading link for a country–industry–year panel that spans manufacturing disaggregated into seven industries for 16 countries – developed and developing – from 1982 through 1990. To construct this panel combined information on host-country inputs and outputs with information on the presence in these countries of affiliates of US-headquartered MNEs. The data on host-country inputs and outputs by industry–country–year come from the United Nations (UN) General Industrial Data Base. In these UN data industries are defined at the threedigit ISIC level, of which there are 28 within manufacturing. The data on US MNEs come from the BEA within the US Department of Commerce, as described in Section 2. I choose to measure FDI presence in host countries by their total employment; results are very similar using alternative activity measures such as affiliate sales. Matching these two data sets required aggregating the ISIC industries because the publicly available BEA data by industry–country–year tend to suppress a substantial amount of observations (to prevent disclosure of individual companies) as the BEA industries are sliced more finely, especially
Skill Upgrading in Developing Countries 137
for smaller countries. I settled on aggregating the 28 up to seven. For the country dimension of the data, I included any middle-income or lowincome country for which both the UN and BEA data were available for most potential observations; I also chose a set of high-income countries as a comparison group. This left me with 16 countries. Finally, the time dimension of the data run 1982 through 1990. BEA data are not available annually for earlier years, and after 1990 the UN data stop distinguishing workers by occupation, which is essential for a measure of skill upgrading. The overall country–industry–year panel contains 951 observations. It is an unbalanced panel due mainly to UN data missing for some country-years. Table 6.6 lists the industries and countries ordered by their 1985 per worker gross domestic product (GDP). In the empirical analysis I will distinguish two groups of developing countries. One group is defined simply as the eight below-median income countries in Table 6.6. The other group is defined more narrowly as the bottom five countries in Table 6.6, with this cut-off Table 6.6 List of industries and countries in estimation sample Industriesa Food and Kindred Products Chemicals Metals and Metal Products Machinery Electronics Transportation Products Miscellaneous Manufacturing Countries Canada Sweden Denmark Finland United Kingdom Spain Ireland Japan Venezuela Mexico Greece South Korea Panama Chile Columbia India
Per worker GDP (1985)b 31 147 26 504 23 861 23 700 22 987 21 169 19 197 18 820 18 362 17 036 16 270 10 361 10 039 9 768 9 276 2 719
Notes a Industries are as classified by the US BEA. See text for details. b Per worker GDP are in US dollars as reported in the Penn World Tables.
138 Matthew J. Slaughter
chosen based on the several-thousand-dollar jump in income above this point in the table. I measure the skill intensity of each industry by its share of total wage bill accounted for by non-production workers. The results are virtually identical using the non-production share of total employment, but I choose the wagebill shares for greater comparability with many studies of skill upgrading in the SBTC literature that assume the underlying production technology to be translog, which is more flexible than Cobb-Douglas or CES (e.g. Berman et al., 1994; Haskel and Slaughter, 2002). The sample mean for this variable is 0.476 (s.d. of 0.195). I measure the presence of MNEs in an industry by the share of its employment accounted for by US affiliates. Ideally I would use data on affiliates from all source countries and not just from the United States, but such data simply do not exist. The sample mean for this variable is 0.052 (s.d. of 0.113). Thus, the typical observation in the data is a country–industry–year in which non-production workers account for just under half of the total wagebill and in which US affiliates account for about one in 20 jobs. I estimate regressions of wage-bill shares on MNE presence and various other controls. In related skill-upgrading studies, these controls commonly include capital and/or output per worker. My results are robust to including or excluding these controls; as capital-stock data contain more missing observations and are probably of highly-variable quality across countries, I report results without these controls. Instead, I exploit the panel nature of the data to control for determinants of skill intensity across time, countries, and industries. In all specifications I include a time trend, as many SBTC studies have found the 1980s to be a decade of widespread skill upgrading in many countries. In all specifications I also include a full set of country fixed effects, to control for unobserved country-level forces – for example, overall level of development, technology, and market transparency – that may influence skill intensity. I experiment with including and excluding a full set of industry fixed effects. Given the use of time and country controls, including industry controls as well does not leave much independent variation in the MNE regressor of interest. Regressions are reported for OLS in levels. Results are robust to weighted least squares using, for example, industry employment as weights. Results are also robust to first-differencing the data. Differenced specifications are common in the skill-upgrading literature, and control for all industry and country fixed effects. But time-differencing aggravates measurement error – an estimation issue that almost surely is important given the range of countries and data sources drawn together in this panel. Consistent with this, relative to the reported results differenced results have similar coefficient estimates but larger standard errors. Table 6.7 reports estimation results, where each column is for a different specification and each row is for a particular regressor. Table cells report coefficient estimates and the corresponding t-statistics that are robust to
Table 6.7 Estimation results for the effect of foreign-affiliate presence on skill upgrading Sample of countries
All countries (1)
Multinational Emp. share Time
0.340 (2.99)*** 0.005 (2.98)*** Country 821 0.63
0.638 (3.99)*** 0.007 (2.99)*** Country 391 0.70
Sample of countries
All countries (6)
Developing countries, broad definition (7)
Multinational Emp. share Time
0.036 (0.61) 0.004 (2.56)** Country, Industry 821 0.72
0.183 (2.13)** 0.006 (5.18)*** Country, Industry 391 0.86
Controls Observations R-squared
Controls Observations R-squared
Developing countries, Developed countries, broad definition broad definition (2) (3) 0.271 (2.05)** 0.002 (1.03) Country 430 0.61
Developing countries, narrow definition (4)
Developed countries, narrow definition (5)
0.530 (3.93)*** 0.002 (0.98) Country 212 0.63
0.298 (2.31)** 0.005 (2.97)*** Country 609 0.63
Developed countries, Developing countries, broad definition narrow definition (8) (9) 0.025 (0.28) 0.001 (0.54) Country, Industry 430 0.69
0.118 (1.57)* 0.001 (1.10) Country, Industry 212 0.86
Developed countries, narrow definition (10) 0.006 (0.09) 0.004 (2.29)** Country, Industry 609 0.71
139
Notes: Robust t-statistics in parentheses. * significant at 10 percent level; ** significant at 5 percent level; *** significant at 1 percent level. In each specification the dependent variable is the non-production share of total wagebill. The broad definition of developing countries consists of those in the bottom half of Table 6.6. The narrow definition of developing countries includes only the bottom five countries in Table 6.6. See text for details.
140 Matthew J. Slaughter
arbitrary forms of heteroskedasticity. The main empirical finding of Table 6.7 is a robustly positive correlation between skill upgrading and the presence of US affiliates, with this correlation even stronger among the sub-sample of developing countries. This result holds for the first five specifications, which include a full set of country effects, and for the second five as well, which control for both country and industry effects. It also holds for either the broad or narrow definition of developing countries. Note that in all specifications in Table 6.7 the coefficient on time is positive, significantly so in many cases. This is consistent with the finding in related studies of a secular trend in skill upgrading. In every specification, (unreported) F-tests indicate the industry effects and/or country effects that are jointly significant. The evidence in Table 6.7 suggests that inward FDI stimulates skill upgrading in developing countries. Again, Section 2 laid out several mechanisms by which this skill upgrading might arise. Without plant- and/or firm-level data these mechanisms cannot be sorted out. Nevertheless, the evidence in Table 6.7 suggests that attempts to obtain and analyze these sorts of micro panels could yield very interesting results.15
5
Conclusions
This chapter has discussed how multinational firms affect both the demand for and supply of skills in host-country labor markets. On the demand side, multinational affiliates raise demand for more-skilled workers as they utilize firm-specific knowledge assets and as they invest in physical capital. All this may also occur in domestic firms in host countries if these knowledge assets are somehow transferred, but evidence on this – particularly for externality spillovers – is rather mixed. On the supply side, multinationals can raise the supply of more-skilled workers both at the micro-level of individual affiliates training workers in-house and via interactions with host-country education and training institutions. They can also do this at the macro-level through channels such as helping raise and stabilize output and affecting migration decisions. This chapter also presented new empirical evidence that inward FDI is correlated with skill upgrading in developing countries. Data permitting, these results may merit closer research along several lines. On the demand side, the issue of how these firms control the within-firm and cross-firm flows of information may matter for how broadly these knowledge assets spread within host countries. At this point, there simply is not a large, systematic body of evidence on knowledge flows into FDI-host countries. This is in part because data requirements to distinguish alternative stories are high: microlevel data on plants, firms, and individuals are really needed. With greater data efforts, much may be learned. In closing, it is important to point out that multinational firms and FDI are not the only channel by which countries can gain access to the
Skill Upgrading in Developing Countries 141
technology and capital required for economic growth. It has been widely documented that in recent years most governments worldwide have made their policies much more friendly to foreign firms (UNCTAD, 2000). That said, history offers many examples of governments (e.g. Japan and South Korea) pursuing development strategies instead of or in tandem with an FDI strategy: joint ventures, licensing, and exporting have been common. The focus of this chapter on the role of FDI in facilitating host-country skill upgrading is best seen in the broad context of all development strategies.
Notes 1.
2.
3.
4.
5.
6.
7.
8.
9.
For helpful comments I thank David Howell, Will Milberg, and other participants at the New School University’s CEPA Conference, “Labor and the Globalization of Production,” March 8, 2002. See Helpman (1984) and Helpman and Krugman (1985). This view is also related to models of foreign outsourcing, in which the vertical separation of production occurs without multinationals. See Markusen (1984), Horstmann and Markusen (1987, 1992), and Markusen and Venables (1998, 2000). Trade models of this variety are similar to older theories of tariff-jumping FDI. See Caves (1996) for a discussion. There have been some attempts to integrate models of horizontal and vertical FDI into a single framework. See, for example, Markusen (2001). In general-equilibrium trade models with multiple sectors, the “sector bias” of technological change – that is, what industries these innovations are occurring in – can matter for economy-wide labor-demand changes above and beyond any factor bias to these innovations. See Haskel and Slaughter (2002). For example, suppose that knowledge spills over from automobile producers, which fall within US Standard Industry Code (SIC) 3711, to automobile-parts suppliers, which fall within US SIC 3712. Then at the four-digit SIC level spillovers would be inter-industry, but at the two- or three-digit level they would be intra-industry. Spillovers via labor turnover might hinge on the mobility of administrative workers – for example, managers with knowledge of organizational techniques. But it might instead hinge on the mobility of production workers as well – for example, assemblers with knowledge of production-line efficiencies. For example, Howenstine and Zeile (1994) and Doms and Jensen (1998) document these wage differentials among US manufacturing plants. Globerman et al. (1994) present similar evidence for Canada; Aitken et al. (1996) for Mexico and Venezuela; and Te Velde and Morrissey (2001) for five African countries. The multinational wage premium may be caused by forces other than superior technology. For example, Budd et al. (2001) argue that if MNEs are, on average, more profitable than domestic firms, then international rent sharing within MNEs could explain this wage premium. For a panel of MNEs in Europe over the 1990s, they estimate a robust correlation between affiliate wages and parent profitability, consistent with this profit-sharing story. Other explanations might be compensating differentials due to perceived disamenities related to working for multinationals, or (especially for operations in lower-income countries) concerns about individual attitudes toward these firms as somehow “exploiting” host-country workers. Berman et al. (1994) document for the United States that employment trends for this job-classification measure track quite closely employment trends measured
142 Matthew J. Slaughter
10.
11.
12.
13. 14.
15.
by the white-collar/blue-collar job classification – which in turn closely reflects the college/high-school classification. That said, this broad occupation classification is an admittedly imperfect measure of labor–market skills. In particular, the skill mix can change within each occupation group – and thus there may be skill upgrading within the separate pools of non-production and production workers. Also notable is the fact that many developing countries had non-production employment shares below that of the overall world. This broadly suggests that MNE employment demands respond to cross-country differences in factor prices. Many researchers (e.g. Berman et al., 1994, 1998) have argued that SBTC has been a primary force behind rising income inequality across skills within countries. But this view is by no means unanimous. For research arguing the links among skill upgrading, SBTC, and wages is far from clear, see, for example, Howell and Wieler (1998) and Card and DiNardo (2002). Within the widely used Standard Industrial Classification (SIC), many studies term ICT sectors part or all of electrical and non-electrical machinery (SIC 36 and 35, respectively). These industries contain much of the ICT hardware such as computers and office products (SIC 357) and semiconductors (SIC 3674). Other ICT sectors include telecommunication services (SIC 48) and information services (SIC 737). Sales data for these industries in the overall United States come from the National Bureau of Economic Research (2001). Sales data for the US parents of American companies with global operations come from the BEA. What is reported for these parents is their sales of goods only, not of goods and services. This is to maximize comparability with the US industry-wide sales data. That said, one potential limitation of these parent data is they classify all of a parent’s sales of goods into the single industry in which that parent is classified. To the extent that some parents span multiple lines of business, and thus sell goods across multiple industries, these data may be noisy. That said, for a smaller number of years sales data are also classified by industry of sales, rather than by industry of parent. Sales data across these two methods are very close to each other. In fact, for ICT industries parent sales by industry of sales are slightly larger than parent sales of goods by industry of parent, so this alternative sales measure would make US parents look even more prominent than they already do in Chart A. The ICT service industries in this chart together constitute SIC industry 737. I know of no empirical evidence on this issue. One concern may be that if multinationals in host countries attract highly talented locals, if those locals move on to employment within these firms in other countries then brain drain may be exacerbated. I did experiment with lagging the measure of FDI presence by one or two years, with the idea that if host-country skill-upgrading impacts take time to materialize, then lagged FDI presence might show even stronger correlations than those in Table 6.7. Such stronger correlations did not appear, however; the results were qualitatively unchanged.
References Aitken, Brian J. and Ann E. Harrison (1999) “Do Domestic Firms Benefit from Foreign Direct Investment? Evidence from Venezuela,” American Economic Review ( June): 605–18. Aitken, Brian, Ann Harrison, and Robert E. Lipsey (1996) “Wages and Foreign Ownership: A Comparative Study of Mexico, Venezuela, and the United States,” Journal of International Economics 40 (3–4) May: 345–71.
Skill Upgrading in Developing Countries 143 Baily, Martin Neil and Robert M. Solow (2001) “International Productivity Comparisons Built from the Firm Level,” Journal of Economic Perspectives 15 (3), Summer: 151–72. Berman, Eli, John Bound, and Zvi Griliches (1994) “Changes in the Demand for Skilled Labor within U.S. Manufacturing: Evidence from the Annual Survey of Manufactures,” Quarterly Journal of Economics, 367–97. Berman, Eli, John Bound, and Stephen Machin (1998) “Implications of Skill-Biased Technological Change: International Evidence,” Quarterly Journal of Economics 1245–80. Berman, Eli and Stephen Machin (2000) “Skill-Biased Technology Transfer: Evidence of Factor-Biased Technological Change in Developing Countries,” Mimeograph, January. Berman, Eli and Stephen Machin (2001) “Globalization, Skill-Biased Technological Change and Labour Demand,” in E. Lee and M. Vivarelli (eds), Globalization, Employment and Poverty Reduction, International Labour Office, Geneva. Blomstrom, Magnus (1986) “Foreign Investment and Productive Efficiency: The Case of Mexico,” Journal of Industrial Economics 35 (1): 97–110. Brumley, Bryan (2001) “IT Reverses Brain Drain,” Australian IT, October 23. Budd, John W., Jozef Konings, and Matthew J. Slaughter (2001) “International Profit Sharing in Multinational Firms,” Mimeograph, November. Card, David and John DiNardo (2002) “Skill Biased Technological Change and Rising Wage Inequality: Some Problems and Puzzles,” National Bureau of Economic Research Working Paper #8769. Caves, Richard E. (1974) “Multinational Firms, Competition, and Productivity in Host-Country Industries,” Economica 41: 176–93. Caves, Richard E. (1996) Multinational Enterprise and Economic Analysis, 2nd edition, Cambridge: Cambridge University Press. Chung, Wilbur, Will Mitchell, and Bernard Yeung (1998) “Foreign Direct Investment and Host Country Productivity: The American Automotive Component Industry in the 1980s,” Mimeo. Dollar, David and Aart Kraay (2000) “Growth is Good for the Poor,” Development Research Group. Washington, D.C.: World Bank. Doms, Mark E. and J. Bradford Jensen (1998) “Comparing Wages, Skills, and Productivity Between Domestically and Foreign-Owned Manufacturing Establishments in the United States,” in Robert Baldwin, Robert Lipsey, and J. David Richardson (eds), Geography and Ownership as Bases for Economic Accounting, Chicago: University of Chicago Press, 235–55. Dunning, J. H. (1981) International Production and the Multinational Enterprise, London: George, Allen, and Unwin. Eaton, Jonathan, and Samuel Kortum (1999) “International Technology Diffusion: Theory and Measurement,” International Economic Review 40 (3) August: 537–70. Globerman, Steven, John C. Ries, and Ilan Vertinsky (1994) “The Economic Performance of Foreign Affiliates in Canada,” Canadian Journal of Economics 27 (1): 143–56. Haddad, M. and Ann E. Harrison (1993) “Are there Positive Spillovers from Direct Foreign Investment?” Journal of Development Economics 42: 51–74. Hamermesh, Daniel (1993) Labor Demand, Princeton: Princeton University Press. Hanson, Gordon H. (2000) “Should Countries Promote Foreign Direct Investment?” G24 paper. Hanson, Gordon H., Raymond Mataloni, and Matthew J. Slaughter (2001) “Expansion Strategies of U.S. Multinational Firms,” in Dani Rodrik and Susan Collins (eds), Brookings Trade Forum 2001, 245–94.
144 Matthew J. Slaughter Haskel, Jonathan E., Sonia C. Pereira, and Matthew J. Slaughter (2002) “Does Inward Foreign Direct Investment Boost the Productivity of Domestic Firms?” National Bureau of Economic Research Working Paper #8724. Haskel, Jonathan E. and Matthew J. Slaughter (2002) “Does the Sector Bias of SkillBiased Technological Change Explain Changing Skill Premia?” European Economic Review 46 (10): 1757–83. Helpman, E. (1984). “A Simple Theory of Trade with Multinational Corporations,” Journal of Political Economy 92: 451–71. Helpman, E. and P. Krugman (1985) Market Structure and Foreign Trade, Cambridge, MA: MIT Press. Horstmann, I. J. and J. R. Markusen (1987) “Strategic Investments and the Development of Multinationals,” International Economic Review 28: 109–21. Horstmann, I. J. and J. R. Markusen (1992) “Endogenous Market Structures in International Trade,” Journal of International Economics 32: 109–29. Howell, David R. and Susan S. Wieler (1998) “Skill-Biased Demand Shifts and the Wage Collapse in the United States: A Critical Perspective,” Eastern Economic Journal 24 (3): 343–66. Howenstine, Ned G. and William J. Zeile (1994) “Characteristics of Foreign-owned U.S. Manufacturing Establishments,” Survey of Current Business, 74 (1) January: 34–59. Keller, Wolfgang (2001) “The Geography and Channels of Diffusion at the World’s Technology Frontier,” National Bureau of Economic Research Working Paper No. 8150, March. Krugman, Paul R. (1991) Geography and Trade, Cambridge: MIT Press. Lipsey, Robert E., Magnus Blomstrom, and Eric Ramstetter (1998) “Internationalized Production in World Output,” in Robert Baldwin, Robert Lipsey, and J. David Richardson (eds), Geography and Ownership as Bases for Economic Accounting, Chicago: The University of Chicago Press, 83–135. Lynch, Lisa M. (1992) “Private-Sector Training and the Earnings of Young Workers,” American Economic Review 82 (1): 299–312. Mansfield, Edwin and Anthony Romeo (1980) “Technology Transfer to Overseas Subsidiaries by U.S.-Based Firms,” Quarterly Journal of Economics 95 (4): 737–50. Markusen, J. R. (1984) “Multinationals, Multi-Plant Economies, and the Gains from Trade,” Journal of International Economics 16: 205–26. Markusen, J. R. (2001) Multinational Firms and the Theory of International Trade, Cambridge, MA: MIT Press. Markusen, J. and A. Venables (1998) “Multinational Firms and the New Trade Theory,” Journal of International Economics, 46: 183–203. Markusen, J. and A. Venables (2000) “The Theory of Endowment, Intra-Industry and Multinational Trade,” Journal of International Economics, 52: 209–34. Marshall, Alfred (1920) Principles of Economics, 8th Edition, London: Macmillan. McKendrick, David G., Richard F. Doner, and Stephan Haggard (2000) From Silicon Valley to Singapore: Location and Competitive Advantage in the Hard Disk Drive Industry, Stanford, CA: Stanford University Press. Moen, Jarle (2000) “Is Mobility of Technical Personnel a Source of R&D Spillovers?” National Bureau of Economic Research Working Paper #7834, August. Moran, T. H. (2001) Parental Supervision: The New Paradigm for Foreign Direct Investment and Development, Washington, D.C.: Institute for International Economics, August. Motta, M., A. Fofur, and T. Ronde (1999) “Foreign Direct Investment and Spillovers through Workers’ Mobility,” CEPR Discussion Paper #2194.
Skill Upgrading in Developing Countries 145 Reisen, Helmut and Marcelo Soto (2001) “Which Type of Capital Flows Foster Developing-Country Growth?” International Finance 4 (1): 1–14. Rodriguez-Clare, Andres (1996) “Multinationals, Linkages, and Economic Development,” American Economic Review 86 (4) September: 852–73. Shaver, J. Myles and Frederick Flyer (2000) “Agglomeration Economies, Firm Heterogeneity, and FDI in the United States,” Strategic Management Journal 21 (12): 1175–93. Slaughter, Matthew J. (2002). U.S. Trade and Investment Policy and the Growth of Information Technology, Washington, D.C.: Emergency Committee for American Trade. Slaughter, Matthew J. (2000) “Production Transfer Within Multinational Enterprises and American Wages,” Journal of International Economics 50 April: 449–72. Slaughter, Matthew J. (1998) American Investments, Global Returns, Washington, D.C.: Emergency Committee for American Trade. Song, J., P., Almeida and G. Wu (2001) “Learning-by-Hiring: When is Mobility Useful?” Paper presented at “Technological Innovation and Evolution,” Conference, March. Te Velde, D.W. and O. Morrissey (2001) “Foreign Ownership and Wages: Evidence from Five African Countries,” CREDIT Discussion Paper. The McGraw-Hill Companies and United States Department of Commerce/International Trade Administration (2000) US Industry & Trade Outlook (2000), New York: The McGraw-Hill Companies. United Nations Conference on Trade and Development (2000) World Investment Report: Cross-Border Mergers and Acquisitions and Development, New York: United Nations. World Bank (1999, 2000) Global Development Finance, Washington, D.C.: The World Bank.
7 Does Trade Promote Gender Wage Equity? Evidence from East Asia Günseli Berik, Yana van der Meulen Rodgers, and Joseph E. Zveglich, Jr
1
Introduction
During the last several decades, industry and export mixes in Taiwan and South Korea have shifted toward higher-skill, technology-intensive products, while lower-skill, labor-intensive industries have been moving abroad. At the same time, women’s relative educational attainment and skill levels improved considerably. Yet some trends have differed across these two economies. Taiwan became increasingly open to trade, with a steady rise in the total trade to output ratio from a low of 48 percent in the early 1980s to a high of almost 90 percent by the late 1990s. Against this backdrop, Taiwan’s average female–male wage ratio in manufacturing dropped fairly steadily from 66 percent in 1981 to 60 percent in 1993. Only in the mid1990s did the wage ratio begin a strong climb upward, reaching 67 percent by 1999. In contrast, Korea’s slow and steady improvement in women’s relative wages – from 47 percent in 1980 to 58 percent by 1998 – was accompanied by a slight decline in trade openness.1 Industrial structure and policy also differ between the two economies, with Taiwan having a higher proportion of small firms and relatively less emphasis on selective government intervention. These divergent features in two of the most successful practitioners of the export-led growth model make Taiwan and Korea well-suited to examine the impact of international trade competition on the gender wage gap. Increased participation in the global economy is expected to generate growing pressures on firms to engage in cost-cutting practices that may affect female employees differently than male employees, particularly if sex discrimination plays a role in hiring and pay decisions of firms. The dynamic implications of Becker’s (1959) theory of discrimination lead us to expect that if discrimination is costly, then increased competition from international 146
Does Trade Promote Gender Wage Equity? 147
trade will reduce the incentive for employers to discriminate against women. This effect should be more pronounced in concentrated sectors of the economy, where employers can use excess profits to cover the costs of discrimination. Alternatively, wage discrimination may increase with growing trade in a context of employment segregation that limits women’s ability to achieve wage gains. For example, trade competition can entail technological restructuring that contributes to women’s absorption into lower paying jobs and a relative loss of power in wage bargaining situations. Following the approaches in Borjas and Ramey (1995) and Black and Brainerd (2000), the empirical strategy controls for differences in market structure across industries in order to isolate the effect of competition from international trade. The methodology is applied separately to Taiwan and Korea, and the results are compared. To control for the effects of changes in worker productivity characteristics, estimations are performed using average residual wage gaps as well as average unadjusted wage gaps by industry. The residual wage gap – the portion of the wage gap that cannot be explained by observed productivity differences between men and women – is commonly attributed to wage discrimination against women. This study contributes to scholarship on globalization by yielding new evidence on the impact of competition from international trade on the gender wage gap in two of the world’s most trade-oriented economies. By integrating micro data on labor markets with macro data on industrial structure and trade, the study moves beyond the unsatisfactory treatment of discrimination in the labor literature as an unexplained residual. At the same time it can include controls for worker productivity characteristics that are lacking in studies that pursue an industry-level analysis of gender wage differentials. Previous research on the effect of increasing competition on wage discrimination is still inconclusive and limited in geographical coverage, leaving the impact of globalization on the gender gap an empirical puzzle. The econometric analysis does not support the dynamic implications of Becker’s theory. Results indicate that competition from international trade has a positive relationship with residual wage gaps between men and women in both Taiwan and Korea. While this result is associated more strongly with import growth in Taiwan and with export growth in Korea, we conclude that in both cases the situation of persistent gender wage inequality can be best remedied with passage and enforcement of anti-discrimination legislation.
2
Competition and the gender wage gap: existing research
The Neoclassical Approach. According to neoclassical theory based on Gary Becker’s The Economics of Discrimination (1959), long-run competition is not compatible with persistent discrimination. The existence of an equally productive but underpaid group of workers will attract profit maximizing
148 Günseli Berik et al.
firms that don’t have discriminatory tastes against this group. Hence the expansion of demand for the group that is discriminated against will gradually eliminate the wage differential. Discrimination will persist only in the least competitive sectors. There employers and managers, who choose to indulge their taste for discrimination by avoiding to hire the lower paid workers and employing higher cost labor instead, can continue to engage in costly discrimination as long as they are making excess profits. Under increasing competition in these industries – through, for example, deregulation of the industry, entry of new firms, or competition from cheaper imports – the use of the underpaid labor is expected to grow and gradually the increase in demand will bid up the wages of this group of workers, reducing the wage gap.2 Consistent with this theoretical notion of discrimination, empirical studies operationalize discrimination as the wage differential that remains after accounting for all measured productivity differences between two groups. This differential is labeled as the “unexplained residual” wage difference.3 A fairly small number of empirical studies have examined the hypothesis that increased competition leads to less gender discrimination and an improvement in women’s relative wages. Among studies of competition in domestic product markets, Black and Strahan (2001) find that by reducing rents that previously accrued to men, banking-sector deregulation has led to improvements in women’s relative wages beyond what can be explained by improvements in women’s skill characteristics. Also for the United States, Hellerstein, Neumark, and Troske (1997) find evidence of wage discrimination against female employees among plants with higher levels of product market power, whereas this finding is not evident in plants with lower levels of market power. However, because discriminatory plants with market power do not grow more slowly than non-discriminatory plants, the dynamic implies that market forces over time have little impact on reducing discrimination. In the labor market, less domestic regulation can contribute to the weakening of anti-discrimination laws and their enforcement, thus generating the result that greater competition enlarges the gender wage gap. For example, Meng (1996) reports that the unexplained portion of the gender wage gap is considerably larger for the rural industrial sector in China than for the urban state sector. The author concludes that wage discrimination by gender is more problematic in the rural industrial sector because the relative lack of government involvement has strengthened traditional attitudes toward the unequal roles of men and women in the labor market. But other researchers argue that more competitive labor markets should work to women’s advantage by penalizing firms that discriminate against women, whereas less competitive labor markets increase the potential for firms to engage in gender discrimination. For example, Winter-Ebmer (1995) finds that in Austria, monopsony power among firms contributes to larger gender wage gaps.4
Does Trade Promote Gender Wage Equity? 149
Among studies of competition from international trade, Black and Brainerd (2000) find that in United States manufacturing industries that are both concentrated and increasingly exposed to competing imports, the residual gender wage gap narrows to a greater extent than in non-concentrated, trade impacted industries. The authors interpret this result as evidence supporting Becker’s hypothesis that discrimination cannot persist in an increasingly competitive environment. The authors also find that increased trade openness in general, controlling for domestic and international competitiveness, increases the US gender wage gap. Curiously, the authors attribute this latter result to the adverse effect of trade on the wages of unskilled workers (presumed to be predominantly female), even though regressions with residual wages already control for worker skill characteristics. A drawback of the study is its application to data from the United States, a relatively closed economy. Artecona and Cunningham (2001) apply Black and Brainerd’s methodology to examine the impact of trade liberalization in Mexico and obtain similar but weaker results. Similar to Black and Brainerd, the authors also find that trade liberalization alone widened the residual gender wage gap and attribute this result to the increasing premium for men’s greater experience. In a cross-country analysis using data for 3 middleincome and 13 high-income countries, Behrman and King (1999) find some evidence that increased competitive pressure due to trade openness is associated with higher gender wage ratios. Although this result controls for the impact of competition in domestic markets, as measured by the number of firms relative to production by sector, it is not robust across sample specifications. By contrast, Zveglich, Rodgers, and Rodgers (1997) and Rodgers (1998) strike a cautionary note against optimism about trade openness. These studies determine how changes over time in measured skill characteristics and their market returns contribute to the evolution of the wage gap in Taiwan and Korea. They conclude that women experienced growing losses in unmeasured gender-related factors over time – losses that worked against women’s relative improvements in education and experience. These unexplained losses more than offset women’s measured productivity gains in Taiwan during the 1980s and early 1990s, while they partially offset women’s productivity gains in Korea in the same period. While these studies do not explicitly relate trade competition to trends in residual gender wage gaps, they do suggest a possible rise in wage discrimination during a period of outward-oriented growth. In sum, the empirical studies that examine the impact of increased exposure to competitive forces on the gender wage gap, often controlling for contemporaneous changes in detailed worker characteristics, have yielded results that are largely inconclusive. In addition to this group of studies, there are a smaller number of studies that have adopted the non-neoclassical perspective on the labor market and have sought to examine the relationship
150 Günseli Berik et al.
between trade and the gender wage gap in the context of East Asian countries. The non-neoclassical approach. In the non-neoclassical approach, wages are determined by not only labor quality, but also industry characteristics, which are the product of competition among firms within an industry and across industries.5 Competition is the tendency for uniform profit rates to emerge across industries under conditions of capital mobility. In contrast to the neoclassical concept, competition is not antithetical to, but encompasses, monopoly and tendencies toward industrial concentration, since concentration may facilitate mobility of capital and equalization of the profit rate. Moreover, competition involves not only rivalry and conflict among different firms within and across industries but also competition among groups of workers. According to this approach, discrimination is not an aberration that arises when competitive conditions fail to hold. On the contrary, it is a product of competition among firms and groups of workers, each behaving in economically rational ways. In this approach, the existence of a hierarchy of jobs and the specter of joblessness are the preconditions for discrimination. In response to these employment conditions, established groups in the labor market become primary agents of discrimination. These privileged groups act collectively to secure favorable employment conditions in a selfinterested manner and engage in exclusionary behavior. The groups’ respective bargaining power, shaped by their recent social class history, will determine how successful each group will be in securing favorable employment conditions. In this approach, the agency of the established group of workers, by virtue of its greater collective power, will shape the pattern of employment segregation and the position where each group is able to locate itself in the job hierarchy. In this tradition, feminist economists have long argued that organized action by male workers has shaped the gender-segregated structure of jobs in industrial economies and the associated lower pay for women workers (Hartmann, 1979; Milkman, 1990). In the context of contemporary East Asia, given the weakness of organized labor for much of this period, male workers may not have been as instrumental in shaping the gendersegregated patterns of employment. A number of studies present evidence to show that, instead, employment segregation by gender in Korea and Taiwan is produced by employer and state hiring, training, employment, and labor control policies in a context of patriarchal gender norms (Cheng and Hsiung, 1994; Nam, 1994; Seguino 1997). Women are excluded from higher paying, skilled occupations because they are denied access to training, or steered away by gender specific job advertisements. Their employment experience and potential for wage increases are cut short by the employer practice of the “marriage bar.” Furthermore, women’s wage growth is dampened by the
Does Trade Promote Gender Wage Equity? 151
state’s promotion of home-based work and relaxation of immigration rules to expand the supply of labor. In turn, the restriction of women’s employment opportunities via these policies limits women’s bargaining power vis-á-vis employers, and therefore their ability to secure wage gains. The state’s emphasis on suppressing collective bargaining in export industries, where women workers are concentrated, has translated into relatively more suppression of women’s bargaining efforts. The state may have implemented some of these policies with the objective of supporting continued export competitiveness, rather than perpetuating gender inequalities; nonetheless, their effect is likely to have perpetuated gender wage inequalities. Feminist economists also argue that Taiwan and Korean firms may have sought to preserve labor peace by favoring men over women in their employment policies. Furthermore, wage discrimination against women workers in export industries may have been used as a vehicle by Korea’s chaebol, which own firms in a variety of industries, to maintain profitable operations in their capital-intensive firms.6 The last two decades in Taiwan and Korea represent a confluence of a number of developments that complicate the analysis of how competition from international trade affects the gender wage gap. Trade competition, outward foreign direct investment, and technological upgrading in several industries are all linked in the effort to maintain or extend competitiveness at the firm and industry levels. Empirical studies using the non-neoclassical framework have examined the impact of trade and foreign direct investment on changes in the gender wage gap. In the Taiwan context, Berik (2000) uses industry-level panel data to provide an explicit test of the effect of export orientation in Taiwan on (unadjusted) gender wage ratios. She finds that, after controlling for employment segregation by gender and other industry characteristics, greater export orientation is associated with smaller wage differentials between men and women. However, this result is due to the fact that export orientation has a larger adverse impact on men’s wages than women’s wages, suggesting the importance of examining absolute wage levels as well as relative wages. Seguino (2000) finds support for her thesis that divergent trends in the (unadjusted) gender wage ratio in Taiwan and Korea over the 1981–92 period are related to the differences in the nature of foreign direct investment flows in the two countries. Accordingly, greater capital mobility in Taiwan’s labor-intensive, female-intensive sectors leaves women workers more vulnerable to losses of bargaining power in wage negotiations. In Korea, however, the government exercised greater control over capital outflows in labor-intensive, export-oriented sectors. Relative to Taiwan, Korean capital was more likely to flow out of capital-intensive, male-dominated sectors. Subsidized credit to export-oriented firms for productivity-enhancing investments also helped to facilitate women’s relative wage increases in Korea. In an environment where capital is more mobile, the reality or threat
152 Günseli Berik et al.
of moving to lower wage sites limits the ability of workers to secure higher wages. An environment of lesser capital mobility, on the other hand, encourages firms to maintain competitiveness by other strategies, such as technological upgrading and improvements in product quality. Unexamined so far are the effects of technological upgrading that are likely to be differentiated by gender. In a period of technological restructuring, men may seek to preserve the new, stable, and higher paying jobs for themselves in the face of competition from lower-waged women workers (Darity, 1989). Women, in turn, may experience greater job losses as unskilled jobs disappear, with three possible outcomes: first, women may not find new employment because new jobs require more skills and women may not qualify without training, but from which they are excluded; second, women may get absorbed into reclassified (lower paying) jobs; and third, women may be absorbed into new jobs that are not held by men. Each of these scenarios indicates perpetuation of and perhaps increase in wage discrimination against women workers. While adherents of the non-neoclassical approach do not predict an increase in wage discrimination in technologically dynamic sectors, one could plausibly provide such an account: industries that are undergoing technological upgrading are likely to shed labor associated with the old technology, and hire new, more skilled labor. Such a process is likely to intensify job competition among groups of workers. This process may widen the gender wage gap as the emerging employment structure makes it even less likely that women will find employment in higher paying positions, and their longer job queues for such positions will keep their wages in check. In sum, non-neoclassical studies have either found or predicted adverse effects of greater international openness on the gender wage gap. These studies have explained employment and pay discrimination by sex in the context of employer incentives, the power of established groups of employees, state objectives to maintain export competitiveness, bargaining power, and gender norms. In such a context, major structural shifts due to capital mobility, trade, and technological restructuring have affected not only changes in labor demand across sectors, but also persistent pay inequities between men and women within sectors. Two recent studies have attempted to disentangle changes in employment and pay in explaining overall changes in Taiwan and Korea’s gender wage gaps. Seguino (2000) argues that the redistribution of employment across industries appears to matter less than gender pay discrepancies within industries. A similar conclusion is reached in an analysis by occupation groups rather than industries: pay discrepancies within detailed occupation groups rather than shifts in employment structures across occupations explain much of Taiwan’s wage inequality during the 1980s and 1990s (Zveglich and Rodgers, 2002). These results imply that women in Taiwan and Korea are not receiving equal pay for work of equal value within industry and occupation groups, results that
Does Trade Promote Gender Wage Equity? 153
could be explained by the presence of wage discrimination in the Taiwan and Korean labor markets. The following descriptive and regression analyses provide more detail on sectoral patterns in employment and pay differentials and how these differentials are related to competition from international trade.
3
Descriptive analysis
Wage structures. Korea is notorious for having one of the lowest gender wage ratios in East Asia. Figure 7.1 provides wage ratio estimates for selected Pacific Rim economies. In 1990 only Malaysia had a lower wage ratio, and by the late 1990s both Korea and Japan ranked at the bottom with ratios of 58 percent. Taiwan’s gender wage ratio during the 1990s compared favorably with other East Asian economies but stood below that of the industrialized countries, including Australia, New Zealand, and the United States. According to Figure 7.1, during the 1990s the wage ratio generally rose across countries.
90
Female/male wages (%)
80
70
60
50
40
Ze al an Si ng d ap or So e ut h Ko re a Sr iL an ka Ta iw an Th ai la nd Un ite d St at es
ia
Ne w
al ay s M
Ko ng Ja pa n
Ho ng
Au st
ra l
ia
30
1990
1999
Figure 7.1 Manufacturing-sector female/male wage ratios: cross-country evidence Note: Wage data are for 1990 and 1999, or for the closest available years. Sources: International Labour Office (2000); authors’ calculations for South Korea and Taiwan.
154 Günseli Berik et al.
Turning next to more detailed trends in the Taiwan and Korean gender wage ratios, Figure 7.2 shows that without adjusting for any gender differences in skills, the female-male wage ratio in manufacturing fell considerably for much of the period, from 66 percent in 1981 to 60 percent in 1993, before Panel A: Taiwan
85 Female/male wages (%)
80 75 70 65 60 55
99
98
19
97
19
96
19
95
19
94
19
93
19
92
19
91
19
90
19
89
19
88
19
87
Residual ratio, manufacturing
Residual ratio, non-manufacturing
Unadjusted ratio, manufacturing
Unadjusted ratio, non-manufacturing
Panel B: Korea
90
Female/male wages (%)
19
86
19
85
19
84
19
83
19
82
19
19
19
81
50
80 70 60 50 40
98 19
96 19
94 19
92 19
89 19
86 19
83 19
19
80
30
Residual ratio, manufacturing
Residual ratio, non-manufacturing
Unadjusted ratio, manufacturing
Unadjusted ratio, non-manufacturing
Figure 7.2 Gender wage ratios by manufacturing industry status, Taiwan and Korea
Does Trade Promote Gender Wage Equity? 155
rising again to 67 percent by 1999. In contrast, women’s relative wages in non-manufacturing industries remained fairly stable at roughly 71 percent until the mid-1990s, when they started to climb by about a percentage point a year. Once gender differences in education and experience are taken into account, women’s relative earnings fall sharply until the mid-1990s, particularly in manufacturing, with a recovery in both sectors during the late 1990s.7 Patterns in Korean gender wage ratios are reported in the second panel. Like Taiwan, Korean women’s relative wages are lower in manufacturing industries than in non-manufacturing industries, but unlike Taiwan, Korea’s female to male wage ratio rose fairly steadily throughout the period in both parts of the economy. Once gender skill differentials are taken into account, residual wage ratios are again lower in manufacturing than elsewhere in the economy, with some upward trend in the early to mid-1980s and not much movement thereafter. Notably, unadjusted wage ratios are quite low in Korea relative to Taiwan while residual wage ratios are considerably higher in Korea. This result occurs because observed worker characteristics explain a larger portion of the gender wage differential in Korea than they do in Taiwan, so once worker characteristics are taken into account, the Korean wage differential narrows more for Korea than for Taiwan.8 To illustrate how competitive forces from international trade might be related to women’s relative wages, Figure 7.3 plots the manufacturing-sector residual wage ratios against aggregate manufacturing trade ratios in both Taiwan and Korea.9 In Taiwan, the period of falling residual wage ratios coincides with a fairly steady increase in trade ratios, while the recovery in women’s relative residual wages toward the end of the period coincides with a flattening in both trade ratios. These patterns, which appear to be slightly stronger for the import series, suggest that increasing openness over time is negatively related to patterns in women’s relative earnings. In Korea, the slow but steady increase in the residual wage ratio contrasts noticeably with a decline in the aggregate export ratio for most of the period, while there is no clear relationship with imports.10 Competitive forces from domestic market structures may also have played a role in determining women’s relative wages in the manufacturing sector. Figure 7.4 reports both wage ratios and wage levels by industry concentration for Taiwan and Korea. For Taiwan, Panel A shows that both the unadjusted and residual wage ratios are higher in concentrated industries than in non-concentrated industries. Further analysis (Panel B) shows that wage levels for both men and women are higher in concentrated industries than non-concentrated industries – though men in non-concentrated industries earn substantially more than women in concentrated industries.11 The most likely explanation is the ability of firms in concentrated industries to pay higher wages to all workers, reflecting the technological dynamism of concentrated industries such as electronics, chemical products, and chemical materials (Chen and Chen, 1995; Amsden and Chu, 2002). Chen and Chen (1995)
156 Günseli Berik et al.
Panel A: Taiwan
70
70
60
60
50
50 40
40
30
30
99
98
19
97
19
96
19
95
Imports/gross output
19
94
19
Residual wage ratio
Panel B: Korea
70
90
60
80
50 70 40 60 30 50 20
Imports/gross output
98 19
96 19
94 19
92 19
19
19
19
19
Exports/gross output
89
30 86
0 83
40
80
10
Female/male residual wages
93
19
92
19
91
19
90
19
89
19
88
19
19
19
19
19
19
19
19
19
87
0 86
0 85
10
84
10 83
20
82
20
Female/male residual wages
80
80
Exports/gross output
Exports/output, imports/output (%)
90
90
81
Exports/output, imports/output (%)
100
Residual wage ratio
Figure 7.3 Gender wage ratios and manufacturing trade ratios, Taiwan and Korea
infer technological capability from sectoral patterns in foreign direct investment. Sectors in which outward foreign direct investment to high wage economies exceed that directed to low wage economies, are sectors where firms are engaged in efforts to expand markets by differentiating products from home production. These industries are also likely to have firms engaged in internal restructuring in Taiwan, introducing new product lines
Panel B: Male and female wage levels, Taiwan
Panel A: Gender wage ratios, Taiwan 180 Real hourly wages (NT$)
75 70 65 60 55
160 140 120 100 80 60 40
50
Residual ratio, concentrated Unadjusted ratio, concentrated
19 8 19 1 8 19 2 8 19 3 8 19 4 8 19 5 8 19 6 8 19 7 8 19 8 8 19 9 9 19 0 9 19 1 9 19 2 9 19 3 9 19 4 9 19 5 9 19 6 9 19 7 9 19 8 99
19 8 19 1 8 19 2 8 19 3 8 19 4 8 19 5 8 19 6 8 19 7 8 19 8 8 19 9 9 19 0 9 19 1 9 19 2 9 19 3 9 19 4 9 19 5 9 19 6 9 19 7 9 19 8 99
20
Residual ratio, non-concentrated Unadjusted ratio, non-concentrated
Male wages, concentrated Female wages, concentrated
Panel C: Gender wage ratios, Korea
6000 Real hourly wages (Won)
80 70 60 50 40
5000 4000 3000 2000 1000
Male wages, concentrated Female wages, concentrated
Figure 7.4 Gender wage ratios by industry concentration, Taiwan and Korea
98 19
96 19
94 19
92 19
89
86
19
19
19
80
98 19
96 19
94 19
92
Residual ratio, non-concentrated Unadjusted ratio, non-concentrated
Male wages, non-concentrated Female wages, non-concentrated
157
Residual ratio, concentrated Unadjusted ratio, concentrated
19
89 19
86 19
83
0 19
80
30 19
Male wages, non-concentrated Female wages, non-concentrated
Panel D: Male and female wage levels, Korea
83
Female/male wages (%)
90
19
Female/male wages (%)
80
158 Günseli Berik et al.
to replace the transplanted ones, and reemploying skilled employees in newly created departments. Industries that have limited capacity to upgrade products or pay higher wages, on the other hand, include several of the nonconcentrated industries in this study. These industries experience absolute declines in output as firms replicate home production in low-wage economies in Southeast Asia. Trends over time are similar in Taiwan’s concentrated and non-concentrated industries. Both experienced a steep decline in wage ratios in the 1980s, followed by relatively flat wage ratios in the early 1990s, and a slight rise in the late 1990s. This commonality across industries suggests that wage trends are shaped by forces other than changes in industrial policy or industrial structure, which were both relatively stable during the period. In the latter half of 1990s, the residual wage ratio in non-concentrated industries began to converge to the residual wage ratio in concentrated industries. Panel B shows that this convergence reflects the fact that men in non-concentrated industries experienced virtually no wage growth after 1995, while wages of women in non-concentrated industries continued to rise. In contrast, the Korean residual wage ratio in non-concentrated industries has been relatively flat while the ratio in concentrated industries has been rising (Panel C). Panel D shows that, similar to Taiwan, men and women in Korea’s concentrated industries earn more than their counterparts in non-concentrated industries.12 Employment distributions. Changes in labor demand may provide further insight into the changes in the gender wage gaps. According to Becker’s theory, trade expansion is likely to reduce discrimination by increasing not only women’s relative wages, but also the relative demand for women workers. To shed some light on quantity effects during the period of analysis, Tables 7.1 (Taiwan) and 7.2 (Korea) present more detailed information on employment shares by gender, wage ratios, and trade shares across sectors for the 1980–99 period. Within manufacturing Taiwan and Korea share some similar trends, in that both female and male employees recorded dramatic shifts out of lowskilled jobs – especially textiles and garments – into higher-skilled jobs – especially electrical and electronic equipment. In Taiwan in 1981, over half of women’s employment in manufacturing (62 percent) was in textiles and apparel, electronics, and plastics. In 1999 women’s employment continued to be concentrated in the three major employers of women – now, electronics, textiles and apparel, and metals (basic and fabricated), which accounted for 61 percent of women’s employment. Men’s employment was more evenly distributed across sectors than women’s, but became more concentrated in a few sectors over time. The top three employers of men accounted for 38 percent of male employment in 1981 and 50 percent in 1999. Employment patterns in Korea are similar. The largest employers of women were textiles, apparel, and electronics in both 1980 and 1998. In
Table 7.1 Employment, residual wage, and trade ratios by industry concentration, Taiwan: 1981–99 (in percentage points) Female employment share Base 1981 Non-concentrated industries Basic Metal; Fabricated metal Wood products Machinery and equipment Paper; printing Misc. industrial products Plastic products Concentrated industries Food, beverage, and tobacco Textile and apparel Non-metallic mineral products Chemical products Leather, fur, related products Transport equipment Rubber products Electrical, electronic Precision instruments Chemical materials, petroleum
Male employment share
Percent female
Residual wage ratio 1999–81
Import ratio 1999–81
Export ratio 1999–81
Change 1999–81
Base 1981
Change 1999–81
1981
1999
3.3 4.1 1.4 2.3 6.7 8.6
4.9 1.4 1.5 1.4 2.6 2.9
17.8 8.2 8.3 5.0 3.9 6.6
3.7 4.2 0.7 0.2 1.5 1.1
12.6 28.1 11.6 26.7 57.5 50.2
21.1 32.1 18.4 33.3 54.5 41.9
5.9 3.7 3.3 3.1 1.0 13.9
6.7 24.2 39.8 12.8 8.6 6.6
9.5 3.1 243.4 3.5 6.1 14.2
5.1 36.1 4.5 1.4 4.4 1.7 0.9 17.3 1.4 0.8
0.2 17.5 1.7 0.8 2.7 1.6 1.7 16.5 0.4 0.8
7.2 9.7 5.9 2.4 2.5 6.0 1.2 10.0 1.6 3.6
1.3 3.2 1.5 0.2 1.5 0.3 0.4 9.0 0.7 0.8
35.7 74.4 37.3 31.5 57.8 17.9 36.8 57.5 41.0 14.8
38.7 66.6 30.5 41.5 53.5 26.5 52.6 55.5 44.1 20.3
2.5 11.5 8.5 3.9 14.5 2.6 2.4 2.8 6.4 6.6
12.4 8.2 10.3 7.7 13.5 15.0 2.1 0.5 282.9 11.5
6.4 25.1 4.8 8.3 67.4 9.5 8.1 8.6 62.0 18.6 159
Female employment share
Male employment share
Percent female
Residual wage ratio 1998–80
Import ratio 1998–80
Export ratio 1998–80
Base 1980
Change 1998–80
Base 1980
Change 1998–80
1980
1998
Non-concentrated industries Furniture and fixtures Misc. industrial products Wood products and cork Printing and publishing Wearing apparel Fabricated metal products Plastic products Machinery and equipment Other non-metallic mineral Precision instruments Leather, fur, related products Textile industries Pottery, china, and glass
0.3 4.6 1.4 1.0 18.4 2.2 0.8 0.6 0.9 1.4 2.8 30.5 1.5
1.0 1.7 0.2 1.3 8.3 0.5 0.8 3.9 1.4 0.4 0.3 15.7 1.2
0.9 2.4 2.8 2.4 4.2 8.0 1.5 6.6 4.4 1.3 2.8 11.3 2.2
0.1 0.5 1.2 0.7 2.2 3.6 1.3 4.4 1.8 0.1 1.2 5.1 1.2
25.0 63.4 30.3 27.7 80.0 19.8 33.3 7.8 16.0 48.8 47.5 71.0 37.4
31.0 35.3 19.6 21.1 64.4 17.9 17.2 12.8 23.8 33.5 41.5 45.9 8.9
3.6 6.3 1.0 7.0 0.5 13.4 0.6 10.9 8.8 3.6 14.7 5.4 21.0
2.3 0.6 31.2 1.6 8.2 12.9 2.4 83.8 0.1 31.1 16.5 7.6 11.1
2.7 9.0 34.2 4.8 40.4 14.6 18.4 21.8 10.9 18.5 14.4 41.8 2.0
Concentrated industries Food manufactures Beverage and tobacco Paper and paper products Transport equipment Electrical, electronic Rubber products Chemical products Iron and steel Chemical materials Petroleum and coal
5.1 0.5 1.1 0.7 16.1 6.3 3.0 0.4 0.4 0.2
4.8 0.5 0.0 3.4 11.0 5.5 1.7 0.6 0.3 0.8
5.6 1.5 2.5 7.8 10.8 4.4 4.3 7.5 2.4 2.4
0.6 0.9 0.5 5.9 6.9 2.8 0.5 2.7 1.1 0.8
45.0 22.3 29.0 7.1 57.4 56.6 38.5 4.3 12.3 5.8
40.9 12.5 11.5 9.5 35.2 15.3 25.5 6.9 6.3 9.5
9.2 21.8 8.1 11.2 13.0 5.1 10.8 4.4 2.5 21.1
2.0 3.1 6.6 31.7 17.3 0.2 3.3 0.7 8.6 10.7
1.8 0.6 14.6 0.1 36.2 107.6 6.0 4.7 25.7 28.7
160
Table 7.2 Employment, residual wage, and trade ratios by industry concentration, Korea: 1980–98 (in percentage points)
Does Trade Promote Gender Wage Equity? 161
contrast to Taiwan, the concentration of women in these three industries declined over time, whereas the top three employers of men accounted for a larger share of male employment in 1998 (42 percent) than in 1980 (30 percent). In both economies, the industries where women’s employment is concentrated are also among the most female intensive sectors. Table 7.1 does not indicate support for Becker’s theory. Increasing trade competition in Taiwan is not associated with an increase in both the wage ratio and the relative demand for women, particularly in concentrated industries where this effect may be expected. Among concentrated industries, only the chemical products and transport equipment industries experienced rising import (or export) ratios that accompanied a rise in residual wage ratios and an increase in women’s share of industry employment. In non-concentrated industries, the relative demand for women rose along with higher trade shares in all but two industries, but this change was accompanied by a decline in wage ratios, suggesting absorption of women into lower paying positions. The non-concentrated sectors have been especially subject to relocation in Southeast Asia and mainland China and are now the source of rising import competition in Taiwan. Except in the plastics industry, these sectors experienced growth in their export shares along with rising import shares, which suggests that they continue to be competitive in export markets. Declining residual wage ratios in all of the non-concentrated industries, not only in industries undergoing capital flight, may indicate the threat of capital mobility on women’s relative wages.13 Changes in employment, pay, and trade for Korea differ considerably. Table 7.2 shows that in Korea’s concentrated industries, increases in either import or export shares tend to accompany rising wage ratios. However, in most cases these gains are associated with a decline in the female share of industry employment. The relative demand for women also fell in most nonconcentrated industries, and several of these declines were associated with declining export orientation and rising import competition.14 Focusing on the largest employers, Tables 7.1 and 7.2 further show that in both countries import competition in textiles and apparel is associated with greater decline in job opportunities for women than for men. At the same time, this industry has become more export-oriented in Taiwan, while in Korea only textiles experienced an increase in its export share, in both cases most likely on the basis of higher quality products.15 Consistent with the decline in relative demand for female workers, the residual wage ratio in Taiwan’s textile and apparel industries declined over time, and fell in textiles but was relatively constant in Korea’s apparel industry. The lower wage ratio may reflect not only that the jobs held by women are rapidly disappearing but also that women are being absorbed into new lower-paying positions in textiles and apparel as the industry undergoes restructuring. The electrical and electronic goods industries in both Taiwan and Korea also saw a decline in the percent of the workforce that is female. In Taiwan, this change was
162 Günseli Berik et al.
associated with a decline in trade ratios and in the wage ratio, which suggests that women are being absorbed into lower paid jobs and are unable to secure wage gains relative to men as this concentrated sector undergoes technological restructuring. In Korea, however, the decline in women’s representation among electronics workers was accompanied by a higher wage ratio and by greater trade shares.16 These patterns among the largest employers signal the contrasting experiences of female workers in Taiwan and Korea since 1980 as labor markets responded to the forces of trade and technological change.
4
Methodology
The regression analysis determines the extent to which the gender wage gap responds to increased competition from international trade. The basic strategy underlying the empirical approach is to separate the effects of greater openness in international trade from other contemporaneous shocks that are unrelated to international competitiveness. The effect of greater openness in international trade is isolated by controlling for differences in market structure across industries. Increased openness in international trade is hypothesized to place pressure on firms to reduce discrimination against women. Firms in relatively more concentrated industries are considered to face less domestic pressure to engage in competitive behaviors such as reducing costly discrimination against women, compared to firms in relatively less concentrated industries. Any reduction in the gender wage gap in relatively concentrated industries should then be attributed to the competitive forces from international trade. Intuitively, the wage effects from trade openness are identified by comparing the wage impact of trade openness in more concentrated industries to less concentrated industries. The estimation model, as employed in Black and Brainerd (2000), tests the idea that changes in competitiveness over time affect changes in wage gaps over time. Let Wist be the average log wage of workers with sex s in industry i and year t, and let sWist denote the difference between men and women in average log wages in industry i and year t. Further, let t denote the difference in any variable between two years within any industry. The estimation equation is: t sWist 0 1Ci 2(tMit) 3(Ci tMit) error where the dependent variable is the change over time in the average gender log wage gap by industry. The notation Ci is the industry-specific measure of domestic concentration, t Mit is the change over time in the industry-specific import ratio (and alternatively the export ratio), and Ci t Mit represents the interaction between industry concentration and the change in the trade share. The coefficient 1 represents the effect from domestic concentration,
Does Trade Promote Gender Wage Equity? 163
and 2 represents the effect from increased exposure to international trade. The final coefficient 3 represents the wage gap effect in a sector that is more open and concentrated; it is the estimate of changes in international competitiveness on changes in the wage gap over time. Each variable is disaggregated by detailed industry codes. Regressions are weighted using probability weights based on average-period employment levels by industry. All data used to estimate the model are described in the Data appendix. The dependent variable is constructed from time series information on average wages by sex, industry, and year. The above equation is estimated using both average unadjusted wage gaps by industry and average residual wage gaps by industry. The residual wage gap – the portion of the wage gap which cannot be explained by observed productivity differences between men and women – is commonly attributed to wage discrimination against women. To calculate these figures, a human capital earnings function is estimated for male employees in each year. Here, the dependent variable is log monthly earnings. Because the independent variables include log monthly hours worked, the analysis effectively estimates an hourly wage equation.17 The remaining observable characteristics include a standard set of human capital indicators, such as education, experience, regional locations, and worker status.18 Occupation dummy variables are excluded since occupational attainment may involve some degree of labor market discrimination. The coefficients from the male regression are used to calculate predicted log wages for male and female workers.19 Residual wages are the difference between actual log wages and predicted log wages, and the residual gender wage gap is the difference between male and female residual wages. The industry-specific measure of domestic concentration is constructed from Pareto function estimates of concentration across revenue categories and industries. Following Curry and George (1983), the number of firms of size x or larger in a given industry can be written as a Pareto distribution of the following form: P(x) (x0/x) for x greater than a minimum size x0 where denotes a constant. Taking the logarithm of both sides yields the following estimation equation: logP(x) log x0 log x Intuitively, on a log-log scale, the distribution of firms along increasing size categories is described as a downward sloping line with slope .20 The steeper the slope coefficient, the less concentrated – or more competitive – the industry. This method can be applied to data that is readily available for many developing countries, including both Taiwan and Korea, whereas other popular indices of concentration have more stringent data requirements.
Taiwan
1981
Korea
1983
1. Basic metal industries; fabricated metal products 2. Wood Products, bamboo products, and furniture 3. Machinery and equipment 4. Paper and paper products; printing 5. Miscellaneous industrial products 6. Plastic products 7. Food, beverage, and tobacco manufactures 8. Textile industries; wearing apparel & accessories 9. Non-metallic mineral products 10. Chemical products 11. Leather, fur, and related products 12. Transport equipment 13. Rubber products 14. Electrical, electronic machinery & equipment 15. Precision instruments 16. Chemical materials, petroleum, and coal products
0.77 0.76 0.76 0.71 0.71 0.67 0.62 0.61 0.61 0.58 0.56 0.53 0.52 0.51 0.51 0.26
1. Furniture and fixtures 2. Miscellaneous industrial products 3. Wood products and cork products 4. Printing and publishing 5. Wearing apparel 6. Fabricated metal products 7. Plastic products 8. Machinery and equipment 9. Other non-metallic mineral products 10. Precision instruments 11. Leather, fur, related products, and footwear 12. Textile industries 13. Pottery, China, and glass products 14. Food manufactures 15. Beverage and tobacco manufactures 16. Paper and paper products 17. Transport equipment 18. Electrical, electronic machinery & equipment 19. Rubber products 20. Chemical products 21. Iron and steel, non-ferrous metal industries 22. Chemical materials 23. Petroleum and coal products
0.75 0.73 0.73 0.71 0.66 0.65 0.61 0.58 0.57 0.57 0.54 0.54 0.53 0.47 0.45 0.44 0.40 0.37 0.35 0.24 0.23 0.22 0.04
Note: Figures report the coefficient estimates of Pareto functions using data on number of enterprises by size categories of operating receipts (Taiwan) or shipments (Korea).
164
Table 7.3 Pareto function estimates of domestic competitiveness for Taiwan and Korea (industry ranking in descending order of competitiveness)
Does Trade Promote Gender Wage Equity? 165
The Pareto function estimates are used to specify domestic concentration as a continuous variable and, alternatively, as a dummy variable. For the latter, industries are ranked in descending order of competitiveness, and industries above a critical breakpoint are assigned a value of zero for non-concentrated, while industries below a critical breakpoint are assigned a value of one for concentrated. To check for robustness in our determination of concentration across industries, the Pareto function is estimated using alternative proxies for size categories and alternative years of data. In particular, revenue size categories are measured by operating receipts for Taiwan and shipments for Korea. Alternatively, asset size categories are measured by assets for Taiwan and paid-in-capital for Korea. For each series the function is estimated with data from a beginning year and an end year. Table 7.3 presents results for both countries’ industry-specific domestic competitiveness using the Pareto function approach applied to revenue size categories using data from a beginning year. Rank correlation coefficients indicate that the rankings do not change much over time and are robust to alternative measures of industry concentration.21 Finally, the import ratio variable is constructed from time series information on the ratio of imports to domestic production (and alternatively, the ratio of exports to domestic production). Trade shares may be a less precise indicator of trade openness compared to measures based on tariffs, non-tariff barriers, and export subsidies, particularly for economies such as Taiwan and South Korea whose governments used an extensive array of trade policies to protect their export- and import-substitute sectors. However, even though the tradeables sectors in these countries benefited from government protection, the protection was linked with carefully structured incentives for competition, thus making trade shares a reasonable indicator of international competitiveness.22 In the regression analysis, trade ratios are specified as natural logs, so the coefficients on the trade ratio variables are interpreted as elasticities.
5
Regression results
Estimation results are found in Table 7.4 (Taiwan) and Table 7.5 (Korea). The results are reported as changes in the male–female wage gap measured in log points. The male–female gap may be converted to a female to male ratio by exponentiating its negative. Both tables report results from estimations where domestic concentration is specified as a time-invariant dummy variable (columns 1 and 3) and as a time-invariant continuous variable (columns 2 and 4). Both tables report results for the full period of analysis (columns 1 and 2), and to avoid possible distortions from the East Asian financial crisis, for the period ending in 1996 (columns 3 and 4).
166 Table 7.4 Determinants of change in the gender wage gap, Taiwan (in log points 100; standard errors in parentheses)
Panel A: Unadjusted wage gap C t Mt C t Mt Constant C t Xt C t Xt Constant Panel B: Residual wage gap C t Mt C t Mt Constant C t Xt C t Xt Constant
(1)
(2)
(3)
(4)
15.23*** (2.94) 4.73 (3.43) 10.66* (5.08) 7.35*** (1.89)
54.13*** (8.88) 28.12 (22.05) 43.37 (32.76) 34.89*** (5.68)
18.35*** (5.36) 3.76* (2.07) 17.83*** (4.71) 12.02*** (2.88)
69.28*** (18.58) 61.72*** (18.68) 85.72*** (25.85) 42.34*** (13.00)
9.51** (3.61) 1.79 (1.61) 1.04 (5.36) 5.25 (2.98)
44.66** (16.53) 2.20 (10.37) 6.46 (15.11) 28.40** (9.71)
4.18 (4.62) 4.26 (5.24) 4.22 (11.33) 7.80** (2.74)
6.86** (3.00) 1.17 (1.86) 11.07** (3.72) 8.39*** (1.72)
20.28** (8.08) 37.28** (13.10) 49.85** (17.86) 8.31 (5.91)
11.03 (6.96) 1.46 (4.16) 17.74*** (5.57) 12.62** (5.47)
2.01 (3.71) 5.73 (3.73) 8.64 (6.04) 9.25*** (2.53)
20.44 (15.11) 14.73 (9.38) 25.90 (15.01) 4.85 (9.26)
0.66 (6.51) 4.29 (10.29) 9.11 (14.47) 12.93** (5.58)
19.23 (20.97) 16.55 (16.49) 29.17 (24.87) 6.91 (13.66) 39.93 (24.27) 72.00*** (17.71) 96.21*** (25.20) 19.87 (16.10) 4.99 (31.11) 8.23 (22.27) 14.81 (35.87) 15.22 (18.55)
Notes: Columns differ as follows: (1) C dummy variable, estimation period is 1981–99; (2) C continuous variable, estimation period is 1981–99; (3) C dummy variable, estimation period is 1981–96; and (4) C continuous variable, estimation period is 1981–96. The notation *** indicates statistically significant at the .01 level; ** at the .05 level; * at the .10 level. Each regression has 16 observations and is weighted using probability weights based on average-period employment levels by industry. The R2 statistics range from 33 percent to 52 percent for regressions with import ratios, and 1 percent to 32 percent for regressions with export ratios.
167 Table 7.5 Determinants of change in the gender wage gap, Korea (in log points 100; standard errors in parentheses) (1) Panel A: Unadjusted wage gap C t Mt C t Mt Constant C t Xt C t Xt Constant Panel B: Residual wage gap C t Mt C t Mt Constant C t Xt C t Xt Constant
(2)
(3)
(4)
10.83* (5.82) 1.09 (1.94) 4.81 (4.81) 10.76*** (3.24)
3.65 (27.06) 5.14 (15.91) 6.66 (26.62) 14.04 (15.11)
4.60 (5.81) 0.72 (1.10) 8.68 (10.82) 13.95*** (2.75)
5.48 (19.53) 29.75 (18.40) 46.71 (28.73) 13.93 (10.60)
16.21** (7.67) 5.33** (2.17) 7.17 (9.61) 11.66*** (2.42)
35.26 (30.76) 20.80** (8.60) 29.47* (14.32) 36.96* (17.81)
8.06 (4.74) 4.92*** (1.32) 7.68 (9.21) 12.01*** (1.45)
4.70 (32.21) 17.26* (9.66) 23.65* (13.56) 19.40 (16.72)
10.66* (5.38) 1.86 (2.80) 1.36 (3.80) 1.40 (4.57)
10.99 (18.38) 5.82 (9.60) 6.36 (15.94) 11.71 (9.28)
6.44 (4.77) 0.61 (2.41) 8.20 (5.87) 2.43 (4.16)
11.13 (14.71) 25.92** (9.86) 38.88** (16.28) 11.28* (6.30)
15.96*** (4.93) 1.21 (5.25) 7.05 (6.34) 0.78 (3.88)
37.39 (22.41) 14.06*** (3.60) 21.46* (10.45) 26.87** (11.45)
9.84** (4.49) 3.60 (3.48) 3.57 (4.78) 0.33 (3.58)
21.63 (22.00) 11.25** (4.20) 13.27 (8.07) 16.06 (10.16)
Notes: Columns differ as follows: (1) C dummy variable, estimation period is 1980–98; (2) C continuous variable, estimation period is 1980–98; (3) C dummy variable, estimation period is 1980–96; and (4) C continuous variable, estimation period is 1980–96. The notation *** indicates statistically significant at the .01 level; ** at the .05 level; * at the .10 level. Each regression has 23 observations and is weighted using probability weights based on average-period employment levels by industry. The R2 statistics range from 1 percent to 27 percent for regressions with import ratios, and 18 percent to 34 percent for regressions with export ratios.
168 Günseli Berik et al.
A strong result emerges for Taiwan: an increase in international competitiveness over time in concentrated industries is associated with a widening in the gender wage gap. This result is robust to alternative end years in estimation, alternative weighting methods, and most alternative specifications of industry concentration. However, the result holds only when trade openness is measured using the ratio of imports to output by sector. Table 7.4 shows that after controlling for worker productivity characteristics, an increase in international competitiveness from imports in concentrated industries (C t Mt) between 1981 and 1999 is associated with a statistically significant increase in the residual wage gap. The result could imply that wage discrimination against women rises in the face of increased competition from imports. While an increase in competitiveness from exports in concentrated industries (C t Xt) also appears to be associated with an increase in the residual wage gap, the result is not significant. For unadjusted wage gaps in Panel A, the main result holds, but the coefficient estimate on C t Mt for the 1981–99 period loses its precision once concentration is measured using a continuous variable. When 1996 is used as the endpoint, increased competitiveness from imports in concentrated industries is similarly associated with wider wage gaps across specifications for both the residual and unadjusted wage gaps. This result suggests that the main conclusion is not sensitive to any spillover effects from the financial crisis. In additional tests for robustness (not reported), the positive coefficient on the interaction between concentration and the import ratio for both the unadjusted and residual wage gap regressions maintains its precision when concentration is measured using the assets-based Pareto estimates, regardless of the end year of the estimation period. A similar robustness across alternative measures of concentration is in evidence when 1997 is used as the end year. When concentration is measured in terms of the number of enterprises/output, the estimates for C t Mt are more sensitive to the end year chosen and only become significant in most specifications that use 1996 and 1997 as the end year. The conclusion that trade competition in concentrated sectors is positively associated with the gender wage gap still holds for Korea, but the results are not as strong, and competition from exports appears to matter more than competition from imports. The first panel of Table 7.5 shows that increased competitiveness from exports is associated with a widening in the unadjusted wage gap in concentrated industries when domestic concentration is specified as a continuous variable but not as a dummy variable. This conclusion also holds for the residual wage gap for the 1980–98 period only. In the shorter period that controls for the effects of the East Asian crisis, it appears that competitiveness from imports, rather than exports, is positively associated with widening residual wage gaps. Similar conclusions are reached in robustness tests when concentration is measured using the number of establishments/output by sector.
Does Trade Promote Gender Wage Equity? 169
6
Conclusion
This study has examined the impact of international competitiveness on industry-level gender wage gaps in Taiwan and Korea’s manufacturing sectors between 1980 and 1999. During much of this period the gender wage gap in Taiwan widened, only to begin a recovery in the mid-1990s. By contrast, Korea’s gender wage gap narrowed slowly and steadily through most of the period. While both countries are highly open to trade, trends in trade ratios also diverged over this period, with Taiwan becoming relatively more exposed to international competitive pressures than Korea. Taiwan experienced a near continuous rise in both export and import shares of its manufactured output, while Korea experienced a decline in its export share for much of the period along with a flat trend in its import share. Taiwan’s greater trade openness has been driven by not only increasing imports of labor-intensive goods from Southeast Asia – which has become the center of Taiwan’s investments – but also by efforts to technologically upgrade the export sector. Both countries had a relatively stable industrial structure and policy during the period, Taiwan being the more competitive and marketoriented of the two. The descriptive analysis, therefore, has shown that in Taiwan rising trade openness and a continued highly competitive industrial structure are associated with rising gender wage inequality. By contrast, Korea’s declining total trade openness, coupled with its less competitive industrial structure, is associated with a gradual narrowing of the manufacturingsector gender wage gap. The study specifically aimed to test the dynamic implications of Becker’s theory that greater competition erodes wage discrimination against women. The empirical analysis centered on the effects of trade on the gender residual wage gap in relatively concentrated industries. Residual wage gaps, which account for gender differences in observed productivity characteristics, are often attributed to wage discrimination against women. In concentrated industries, employers face less pressure from domestic competition to cut costs, so any reduction in the residual wage gap could then be attributed to the competitive forces from international trade rather than to domestic pressures. Consistent with the descriptive analysis, industry-level wage-gap regressions do not support Becker’s theory. Quite the contrary, results indicate that increasing competitiveness from trade is positively associated with the residual wage gap in concentrated industries in both countries. The more robust regression results emerge for Taiwan. They show that from 1981 to 1999, rising import shares are associated with rising wage discrimination against women workers in concentrated industries, which include the largest employer and major exporter sectors, such as the textiles/apparel and the electronics/electrical industries. For Korea, a slight decline in export orientation is weakly associated with a reduction in wage discrimination against female workers in concentrated industries between 1980 and 1998.
170 Günseli Berik et al.
The results for Taiwan may be explained in light of the restructuring in Taiwan’s labor-intensive, export-oriented industries, which experienced layoffs following the expansion of foreign direct investment in the same industries from Taiwan toward mainland China and Southeast Asia. Concentrated industries that are more successful in technological upgrading and capturing new export markets are also the ones that disproportionately employ higher-paid, male workers. Moreover, rising import competition in previously export-oriented sectors appears to be associated with more layoffs for women than men and greater downward pressure on women’s wages in an environment that affords little protection for workers. Through both channels, female workers appear to disproportionately bear the costs of greater exposure to international trade. In the Korean case, improvements in women’s relative pay occurred not in the presence of greater competition from international trade, as implied by Becker’s theory, but in the face of slightly weaker international trade pressures. The result is consistent with arguments in Seguino (2000) that the relative lack of capital mobility out of export-oriented sectors compared to Taiwan and the government’s policies to support upgrading efforts among high-volume firms in these sectors could have inadvertently benefited women. The comparative analysis of two highly successful export economies thus shows that increasing international openness does not ensure an improvement in women’s relative economic status. The major policy initiative to ensure that gains from globalization are shared more equally between men and women would be to pass and enforce anti-discrimination legislation. This initiative would not only open up access to jobs for women in a wider range of occupations and industries than the ones they are concentrated in but also ensure equal pay for equal work. The political economic context, however, indicates the enormous difficulty of passing, let alone enforcing, such legislation. Taiwan has had an equal pay for equal work clause on the books since enacting the 1984 Labor Standards Law, but evidence in Chiu (1993) shows that firms were less likely to comply with the equal pay clause than they were with other measures in the Labor Law. Taiwan still has no equal opportunity in employment legislation on the books. Chen’s (2001) discussion of the “Gender Equal Employment” Bill (GEEB), first introduced in the parliament in 1989, shows the resistance to the bill that came from business groups and economic ministries of the government entrusted with ensuring continued export competitiveness. The bill is a comprehensive law aimed to protect women’s equal rights in the labor market. It prohibits employment discrimination in various forms and introduces provisions such as parental leave to reduce the conflict between employment and family responsibilities. Chen states that representatives of business groups threatened to move firms out of Taiwan or not hire women, should the bill be passed, and in the recessionary environment of the early 1990s these threats were effective in preventing the passage of this bill into
Does Trade Promote Gender Wage Equity? 171
law. The threats suggest that wage discrimination against women workers continues to be an important means for continued export competitiveness, in spite of the manufacturing industry’s shift from the low-wage route to maintaining export competitiveness to a route where competition is more on the basis of export product quality. In the Korean case, the government has had equal opportunity in employment legislation on the books since 1987. Additional labor reforms in 1987 relaxed the government’s control on unions and liberalized collective bargaining procedures. These changes resulted in greater pressure on firms to comply with existing labor standards they had previously ignored, including an equal pay for equal work clause. The slight decline in export orientation and support from the government’s industrial policy in concentrated industries may have facilitated better enforcement of this anti-discrimination legislation during the period of analysis, helping to explain the decline in wage discrimination in export-oriented, concentrated industries.
Data appendix Taiwan’s standard industry classification has undergone six revisions since its inception in 1967. The first year of data used in this study is based on the second revision. Revisions 3 (1983), 4 (1987), and 6 (1996) contain changes primarily at the three-digit level, which are readily reconcilable with earlier versions for the two-digit classifications. Only the fifth revision, published in 1991, contains major changes at the oneand two-digit levels, but this change only affects the labor data from 1993 onwards. Korea’s standard industry classification has undergone seven revisions since its establishment in 1963. The first year of data used in this study is based on the fourth revision. Revisions 5 (1984) and 7 (1998) contain changes primarily at the four-digit and five-digit levels, which are readily reconcilable with earlier versions for three-digit classifications (Kang, 2001). Only the sixth revision, published in 1991, contains major changes at all levels of aggregation, but this change only affects the labor data from 1993 onward. Both the Taiwan and Korea industry classification systems are based on international standards. We categorize manufacturing industries by two-digit codes for Taiwan and three-digit codes for Korea since the labor data record industry of employment at that level. As shown in Appendix Table 7.A1, some additional aggregation was necessary to provide consistency over time and to avoid small cell sizes in the calculation of average wages by gender and industry.23 Appendix Table 7.A2 summarizes the key variables and their sources. The labor data for Taiwan are from the Manpower Utilization Survey for the period 1981–99. This household-level survey contains detailed information on earnings, employment, and worker characteristics. The Korean labor data come from the Occupational Wage Survey (OWS), but due to constraints on data availability, the sample is limited to the years 1980, 1983, 1986, 1989, 1992, 1994, 1996, and 1998. This establishment-level survey similarly contains comprehensive labor-market information for individual workers. Surveyed establishments are selected using a stratified random sampling method from establishments with at least ten workers. In addition to small firms, the OWS also excludes the self-employed, family workers, temporary workers, and public-sector workers. Wages for both countries are deflated using the consumer price index.
Table 7A.1 Consistent manufacturing aggregates by Standard Industrial Classification Codes, Taiwan and Korea Rev. 2–4 (1981–92)
1. Food, beverage, and tobacco manufactures 20, 21 2. Textile industries; manufacture of wearing apparel and accessories 22, 23 24 3. Manufacture of leather, fur, and related products 4. Manufacture of wood products, bamboo 25 products, and furniture 26 5. Manufacture of paper and paper products; printing 6. Manufacture of chemical materials; manufacture of petroleum and coal products 27, 29 28 7. Manufacture of chemical products 30 8. Manufacture of rubber products 31 9. Manufacture of plastic products 32 10. Manufacture of non-metallic mineral products 33, 34 11. Basic metal industries; manufacture of fabricated metal products 35 12. Manufacture and repair of machinery and equipment 36 13. Manufacture and repair of electrical and electronic machinery and equipment 14. Manufacture and repair of transport equipment 37 38 15. Manufacture of precision instruments 39 16. Manufacture of miscellaneous industrial products
Rev. 5–6 (1993–99) 11, 12 13, 14 15
16, 17 18, 19
21, 23 22 24 25 26 27, 28 29 31 32 33 39
Korean manufacturing industry aggregates
1. 2. 3. 4.
Food manufactures Beverage and tobacco manufactures Textile industries Manufacture of wearing apparel
Rev. 6–7 (1993–98)
311 313, 314 321 322
151–154 155, 160 171–173 181
5. Manufacture of leather, fur, related products, 323, 324 and footwear (except rubber or plastic footwear) 331 6. Manufacture of wood products and cork products 332 7. Manufacture of furniture and fixtures 341 8. Manufacture of paper and paper products 342 9. Printing and publishing 351 10. Manufacture of chemical materials 352 11. Manufacture of chemical products 12. Manufacture of petroleum and coal products 353, 354 13. Manufacture of rubber products 14. Manufacture of plastic products 15. Manufacture of pottery, china, and glass products 16. Manufacture of other non-metallic mineral products 17. Iron and steel basic industries; non-ferrous metal basic industries 18. Manufacture of fabricated metal products 19. Manufacture of machinery and equipment 20. Manufacture of electrical and electronic Machinery and equipment 21. Manufacture of transport equipment 22. Manufacture of precision instruments 23. Manufacture of miscellaneous industrial products
Source: DGBAS (1987, 1991) and NSO (2001).
Rev. 4–5 (1980–92)
182, 191, 192 201, 202 361 210 221, 222 241 242, 243 231–233
355 356 361, 362
251 252 261
369
269
371, 372
271–273
381 382 383
281, 289 291–293 300, 311–315, 319, 321–323 341–343, 351–353, 359 331–333 369, 371, 372
384 385 390
172
Taiwan manufacturing industry aggregates
Table 7A.2 Key variable definitions and data sources Variable
Description
Data source: Taiwan
Data source: Korea
Wage deflator
Average wages and average residual wages by sex and industry Consumer price index
Manpower Utilization Survey, DGBAS 1981–99 Monthly Bulletin of Statistics, DGBAS (various) 1981–99
Occupational Wage Survey, Ministry of Labor 1980, 83, 86, 89, 92, 94, 96, 98 NSO Online Statistical Database, NSO 1980–98
Trade shares Imports
CIF value by industry
Exports
FOB value by industry
Domestic production
Gross receipts (Taiwan), gross output (Korea) by industry
Exchange rate
Average exchange rate
International Economic Data Bank, ANU (2001) 1980–98 International Economic Data Bank, ANU (2001) 1980–98 Industrial Statistics Database, UNIDO (2000) 1980–98 Not applicable
Industry concentration
Pareto function estimates by industry
International Economic Data Bank, ANU (2001) 1981–99 International Economic Data Bank, ANU (2001) 1981–99 National Income Accounts, DGBAS 1981–99 Financial Statistics (IMF), Central Bank of China (various) 1981–99 Industrial and Commercial Census, DGBAS 1981, 1996
Real wage Nominal wage
Korean Industrial Census, NSO 1983, 1998 173
174 Günseli Berik et al. The US-dollar value of exports and imports by four-digit International Standard Industrial Category (ISIC) are from the International Economic Data Bank (IEDB) for the period 1980–99 for both Taiwan and Korea. IEDB converts UN trade data from Standard International Trade Categories to ISIC using concordance tables based on the structure of trade and production in the two economies. The trade ratios are relative to a measure of gross production by industry. For Taiwan, the data are gross receipts by two-digit level Taiwan industry codes (revision 6) in new Taiwan dollars. The production data are converted to US dollars using the annual average exchange rate. The trade data are regrouped according to two-digit Taiwan industry codes (revision 6) using a linking scheme based on concordance tables published by the United Nations and the government of Taiwan. For Korea, the production data are gross output in US dollars by three-digit ISIC codes from the UNIDO Industrial Statistics Database. For Taiwan, the data for estimating domestic competitiveness come from the Industrial and Commercial Census. This census, conducted once every five years, contains data by disaggregated industry on the number of enterprises in various operatingreceipts categories of increasing size. The census also has data on the number of enterprises by asset size categories and by industry. We use both series in alternative estimates of the Pareto function for a base year (1981) and an end year (1996). For Korea, the data for estimating domestic competitiveness come from the Korean Industrial Census. This census, conducted once every five years, contains data by disaggregated industry on the number of establishments in various shipment-value categories of increasing size. The census also has data by industry on the number of establishments by asset size categories, as indicated by paid-in-capital. Both series are used in alternative estimates of the Pareto function for a base year (1983) and an end year (1998).
Notes 1. Korea’s total trade ratio fell continuously from a high of 56 percent in 1985 to a low of 45 percent in 1993, thereafter rising again, owing to the collapse of output, to 51 percent in 1997 and 60 percent in 1998. 2. Becker’s original notion of competitiveness is based on characteristics of domestic industrial structure. More recent empirical studies on discrimination include international trade as a potential source of increased competitiveness. Behrman and King (1999) report that the idea that international trade increases competition or imposes market discipline is a longstanding one in the trade literature, and one that has recently found empirical support in Levinsohn (1993). 3. A common technique, based on Oaxaca (1973), uses cross-sectional data on workers and decomposes the wage gap in individual years into a portion explained by measured skill differences between men and women, and a residual portion commonly attributed to wage discrimination. 4. The totality of findings for China is quite mixed. For example, contrary to his previous work, Meng (2000) reports that men earn significantly more than women in domestic market-oriented, state sector employment whereas no gender differential exists in foreign-owned firms. The World Bank (2001) further discusses the mixed evidence on increased labor market competition and the gender wage gap. 5. The summary of theoretical frameworks is based on Darity (1989), Williams and Kenison (1996), Mason and Williams (1997), and Mason (1999). 6. See Seguino (1997), Cheng and Hsiung (1994), and Nam (1994) for further discussion of these arguments.
Does Trade Promote Gender Wage Equity? 175 7. For each country, the residual wage ratio in manufacturing and non-manufacturing industries is calculated using a wage regression based on the full set of non-farm employees. 8. For Taiwan, observed productivity characteristics explain about 40 percent to 60 percent of the gender wage differential during the 1980s and early 1990s (Zveglich, Rodgers, and Rodgers, 1997), while for Korea, observed characteristics explain about 70 percent to 80 percent of the gender wage differential during the same period (Rodgers, 1998). 9. The aggregate manufactured export share is constructed as total manufactured exports divided by total manufactured output, and similarly for imports. 10. The sharp rise in the aggregate export ratio between 1996 and 1998 reflects the dramatic drop in output due to the Asian crisis. 11. This inverse relationship between industry competitiveness and wage levels holds at the inter-industry level in a given year as well, especially in the early 1980s, and is robust to the measure of industry concentration used. 12. Similar to Taiwan, in a given year, the more concentrated the industry the higher the wages of both men and women, and this relationship is also robust to the measure of industry concentration used. 13. Further analysis shows that trends based on sectoral average changes over the entire period are consistent with the changes reported in Table 7.2. 14. Trends in Korea’s non-concentrated industries are sensitive to the choice of 1998 as the end year. Changes over time in sectoral averages show that the trend in nonconcentrated industries is one of increasing trade shares that are associated with a decline in wage ratios and in the female share of industry employment. 15. Chen and Chen (1995) indicate that in Taiwan, outward foreign direct investment in textiles and apparel is largely motivated by attempts to maintain firm competitiveness by relocation to lower wage sites. The relocation of apparel production to China and Southeast Asia stimulated expansion of exports of selected products in textiles to these countries, fostering cross-border linkage effects based on prior vertical integration of these industries in Taiwan. However, this linkage has not prevented the decline of Taiwan’s textile industry given the loss of European and US export markets and the simultaneous undermining of cross border linkages by some textile firms that invested abroad along with apparel factories. This intra-industry restructuring is consistent with a gendered restructuring of employment opportunities in Taiwan, whereby men’s employment expands in the growing, higher-paying sub-sectors of textiles, while women’s job opportunities decline or are reclassified as the lower tier jobs in textiles and apparel. 16. Closer inspection of Korean time series data for electronics indicates the outlier nature of the 1998 trade shares and the need to isolate the effects of the Asian crisis. Over the 1980–96 period, the rising wage ratio was associated with a decline in both export and import shares, but trade shares sharply rose after 1996 owing to the collapse of industry output following the crisis. 17. This approach is followed because the Taiwan survey reports monthly earnings and weekly hours worked. 18. In particular, the regressions for both Taiwan and Korea include a dummy variable for part-time worker (less than 40 hours per week for Taiwan or 160 hours per month for Korea), dummy variables for education level attained, years of potential experience (age, minus education, minus six) and its square, years of establishment-specific tenure and its square, and dummy variables for regional
176 Günseli Berik et al.
19.
20.
21.
22.
23.
locations. In addition, regressions for Taiwan include dummy variables for major area of study and a dummy variable for urban status. This approach, also followed in Blau and Kahn (1996), implicitly assumes that male coefficients more accurately reflect competitive returns to observable characteristics than female coefficients. An alternative approach combines coefficient estimates for both men and women following Oaxaca (1973). These two approaches yield the same measures for the total, predicted, and residual earnings gaps. The data source presents size categories as ranges. In the estimations, size levels are chosen to be the mid-point for each range. Because the final range has no upper bound, the size level is top coded as 1.2 times the lower bound. In further robustness checks, results are compared with rankings based on available data for number of enterprise (or establishment) units divided by output per industry, the K-firm concentration ratio, the Herfindahl index, and the profit rate by industry. See Amsden (2001) and Stiglitz and Yusuf (2001) for discussions of the competitive environment within the export- and import-competing sectors. Also, data constraints for measures related to trade policy in both Taiwan and Korea limit our options for measuring international competitiveness. The minimum cell size for a gender-industry group in the sample period was 13. Larger cell sizes would entail a tradeoff with degrees of freedom in the model estimations.
References Amsden, Alice H. and Wan-Wen Chu (2002) “Upscaling: Recasting Old Theories to Suit Late Industrializers,” in Peter C. Y. Chow (eds), Taiwan in the Global Economy: From an Agrarian Economy to an Exporter of High-Tech Products, Westport, CT: Praeger. Amsden, Alice (2001) The Rise of “The Rest”: Challenges to the West From LateIndustrializing Economies Oxford: Oxford University Press. Artecona, Raquel and Wendy Cunningham (2001) “Effects of Trade Liberalization on the Gender Wage Gap in Mexico.” World Bank, Background Paper to the Policy Research Report on Gender and Development. Australia National University (ANU) (2001) International Economic Data Bank: ISIC International Trade Data, Australia: ANU. Becker, Gary (1959) The Economics of Discrimination, Chicago, IL: University of Chicago Press. Behrman, Jere and Elizabeth King (1999) “Competition and Gender Gaps in Wages: Evidence from 16 Countries.” World Bank, Background Paper to the Policy Research Report on Gender and Development. Berik, Günseli (2000) “Mature Export-led Growth and Gender Wage Inequality in Taiwan,” Feminist Economics 6 (3): 1–26. Black, Sandra and Elizabeth Brainerd (2004) “Importing Equality? The Impact of Globalization on Gender Discrimination,” Industrial and Labor Relations Review 57 (4): 540–59. Black, Sandra and Philip Strahan (2001) “The Division of Spoils: Rent-Sharing and Discrimination in a Regulated Industry.” American Economic Review 91 (4): 814–31. Blau, Francine and Lawrence Kahn (1996) “Wage Structure and Gender Earnings Differentials: An International Comparison,” Economica 63 (250 supplement): S29–S62.
Does Trade Promote Gender Wage Equity? 177 Borjas, George and Valerie Ramey (1995) “Foreign Competition, Market Power, and Wage Inequality,” Quarterly Journal of Economics 110 (4): 1075–110. Central Bank of China (CBC) (Various years) Financial Statistics (IMF). Taipei. Chen, Fen-ling (2001) Working Women and State Policies in Taiwan, New York, NY: Palgrave. Chen, Tain-Jy and Yi-Ping Chen (1995) “Foreign Direct Investment and Deindustrialization: The Case of Taiwan,” Journal of Industry Studies 2 (1): 57–68. Cheng, Lucie and Ping-Chun Hsiung (1994) “Women, Export-Oriented Growth, and the State: The Case of Taiwan,” in Joel Aberbach, David Dollar, and Kenneth Sokoloff (eds), The Role of the State in Taiwan’s Development, Armonk, NY: M.E. Sharpe. Chiu, Su-fen (1993) Politics of Protective Labor Policy-Making: A Case Study of the Labor Standards Law in Taiwan. Ph.D. dissertation. Madison, WI: University of Wisconsin-Madison, 1993. Curry, B. and K. D. George (1983) “Industrial Concentration: A Survey,” Journal of Industrial Economics 31 (3): 203–55. Darity, William (1989) “What’s Left of the Economic Theory of Discrimination?” in Steve Shulman and William Darity (eds), The Question of Discrimination, Middletown: Wesleyan University Press. Directorate-General of Budget, Accounting, and Statistics, Executive Yuan (DGBAS) (1987) Standard Industrial Classification of the Republic of China, Revision 4, Taipei: DGBAS. —— (1991) Standard Industrial Classification of the Republic of China, Revision 5, In Chinese, Taipei: DGBAS. —— (Various years) Monthly Bulletin of Statistics, Taipei: DGBAS. Hartmann, Heidi (1976) “Capitalism, Patriarchy, and Job Segregation by Sex,” Signs: Journal of Women in Culture and Society 1 (2): 137–69. Hellerstein, Judith, David Neumark, and Kenneth Troske (1997) “Market Forces and Sex Discrimination,” National Bureau of Economic Research, Working Paper No. 6321. International Labour Office (ILO) (2000) Yearbook of Labour Statistics, Geneva: ILO. Kang, Myung Soo (2001) “A Study on Changes of Wage Distribution in Korea, 1976– 1998, From the Perspective of Skill-Biased Technological Changes,” Dissertation. Columbia, Missouri: University of Missouri, Columbia. Levinsohn, James (1993) “Testing the Imports-As-Market-Discipline Hypothesis,” Journal of International Economics 35: 1–22. Mason, Patrick (1999) “Male Interracial Wage Differentials: Competing Explanations,” Cambridge Journal of Economics 23: 261–99. Mason, Patrick and Rhonda Williams (1997) “The Janus Face of Race: Reflections on Economic Theory,” in Patrick Mason and Rhonda Williams (eds), Race, Markets, and Social Outcomes, Boston/Dordrecht/London: Kluwer Academic Publishers. Meng, Xin (1996) “The Economic Position of Women in Asia.” Asian-Pacific Economic Literature 10 (1): 23–41. —— (2000) Labor Market Reform in China, Cambridge: Cambridge University Press. Milkman, Ruth (1990) “Gender and Trade Unionism in Historical Perspective,” in Louise Tilly and Patricia Gurin (eds), Women, Politics, and Change, New York: Russell Sage Foundation. Nam, Jeong-Lim (1994) “Women’s Role in Export Dependence and State Control of Labor Unions in South Korea,” Women’s Studies International Forum 17 (1): 57–67. National Statistical Office, Republic of Korea (NSO) (2001) Korean Standard Industrial Classification Revision 7, Seoul: NSO. Oaxaca, Ronald (1973) “Male-Female Wage Differentials in Urban Labor Markets,” International Economic Review 14 (3): 693–709.
178 Günseli Berik et al. Rodgers, Yana (1998) “A Reversal of Fortune for Korean Women: Explaining the 1983 Upward Turn in Relative Earnings,” Economic Development and Cultural Change 46 (4): 727–48. Seguino, Stephanie (1997) “Gender Wage Inequality and Export-Led Growth in South Korea,” Journal of Development Studies 34 (2): 102–32. —— (2000) “The Effects of Structural Change and Economic Liberalisation on Gender Wage Differentials in South Korea and Taiwan,” Cambridge Journal of Economics 24: 437–59. Stiglitz, Joseph and Shahid Yusuf (eds) (2001) Rethinking the East Asian Miracle, Oxford: Oxford University Press. United Nations Industrial Development Organization (UNIDO) (2000) Industrial Statistics (ISIC 3-digit level) Database. CD-ROM. Vienna, Austria: UNIDO. Williams, Rhonda M. and Robert E. Kenison (1996) “The Way We Were?: Discrimination, Competition, and Inter-Industry Wage Differentials in 1970,” Review of Radical Political Economics 28 (2): 1–32. Winter-Ebmer, Rudolf (1995) “Sex Discrimination and Competition in Product and Labour Markets” Applied Economics 27 (9): 849–57. World Bank (2001) Engendering Development: Through Gender Equality in Rights, Resources, and Voice, Oxford: Oxford University Press. Zveglich, Joseph, Yana Rodgers, and William Rodgers (1997) “The Persistence of Gender Earnings Inequality in Taiwan, 1978–1992,” Industrial and Labor Relations Review 50 (4): 594–609. Zveglich, Joseph and Yana Rodgers (2004) “Occupational Segregation and the Gender Wage Gap in a Dynamic East Asian Economy,” Southern Economic Journal 70 (4): 850–75.
8 Globalization in a Middle-income Economy: FDI, Production, and the Labor Market in South Africa1 Stephen Gelb and Anthony Black
This chapter examines the impact of foreign direct investment during the 1990s on South Africa, a middle-income semi-industrialized economy with an established industrial base, a segmented labor market with very high unemployment, and real exchange rate depreciation during the 1990s. The paper argues on the basis of survey data and sectoral analysis that FDI has had some impact on the globalizing of production, in the sense of output being exported into global production networks. But most new FDI in South Africa has not been part of the latter processes, focusing rather on domestic and regional markets. Integration of domestic production processes into global networks remains limited. Furthermore, overall FDI levels have been low, so that policy objectives of increased output and employment from FDI have not been met. In particular, the contribution of FDI to savings, investment and thus employment-creating growth has been low, a matter of concern given unemployment rates in excess of 40 percent. On the other hand, FDI has contributed to skills upgrading, particularly among high-skilled blacks, which is important given the importance of ‘black economic empowerment’ in the current South African context. But in doing so, it may have also worsened the skills bias in the labor market. The chapter starts in Section 1 by providing an overview of the South African economy in the 1990s, looking particularly at the policy environment and at the evolution of labor demand. Section 2 provides an overview of FDI flows and their impact on output and employment growth, mainly using survey data. Section 3 examines the extent of globalized production linked to FDI, using both the survey data and brief sectoral “case studies” of the auto, clothing and financial services industries. Section 4 discusses selected labor market issues including skilled labor constraints, human capital accumulation and “black economic empowerment.” Section 5 concludes.
179
180 Stephen Gelb and Anthony Black
1
South Africa in the 1990s
Economic policy and performance Starting before the installation of the first democratically elected government in 1994, but continuing after this event, South Africa’s economic policy regime has become far more liberal and outward-oriented, a major reason being to attract new foreign investment.2 The single most important economic policy statement since 1994, the Growth, Employment, and Redistribution (GEAR) policy announced in June 1996, made increased levels of FDI one of its central objectives. Domestic policymakers have focused narrowly on a macroeconomic argument for FDI – low domestic savings are identified as the binding constraint on growth, to be alleviated by net capital inflows. Since GEAR was formulated partly in response to a capital account shock, FDI was seen as far preferable to volatile portfolio flows as a route to address savings shortages, which were in turn identified as the underlying reason for poor fixed investment rates (Government of South Africa, 1996). Industrial policy echoes the same concern – a recent major statement mentions FDI only once, arguing that “The promotion of domestic and foreign direct investment is critical given the low savings and investment rates in the economy” (Government of South Africa, 2002, p. 32).3 The fiscal deficit was 3.2 percent of GDP in 1990 but ballooned to 10.1 percent in 1993. It was brought below the 3 percent target by 1998, where it has remained. CPI inflation had averaged around 15 percent in the 1970s and 1980s, but was lowered to single digits by contractionary policies in the early 1990s. An inflation-targeting regime was put in place in 1999, with a 3–6 percent target band. South Africa became a GATT signatory in 1994, and average tariff levels were reduced from 27.5 percent in 1990 to 7 percent by 1997, while nearly 60 percent of imports faced a zero tariff in 2000. Both imports and exports grew rapidly, and the current account deficit averaged just over 1 percent of GDP during the 5 years to 2000.4 South Africa also signed the GATS, TRIPS, and TRIMS agreements on joining the WTO in 1995. Commitment to services liberalization under GATS was limited up to 2000, but has since been extended in communications, transport, and energy. Intellectual property legislation exceeded the minimum TRIPS standards before South Africa signed in 1995, though enforcement has been less effective. Just under half the capital stock was in public ownership in 1994, and there has been an explicit commitment to privatization. But implementation has been slow in the face of domestic political opposition and, since 1999, global market conditions. However, the single largest foreign investment since 1994 has been the 1997 sale to a strategic partner of 30 percent of the equity of the state telephone company. The IPO of a further 25 percent was delayed until 2003. The capital account and financial system have also been liberalized. The two-tier currency in place for most of the period since 1961 was abolished
FDI, Production, and the Labor Market in SA 181
in March 1995, and three-quarters of the foreign exchange control regulations in 1994 eliminated by 1998. In 1995, branches of foreign banks were allowed to operate and the Johannesburg Stock Exchange had a “little bang,” admitting foreign brokers and scrapping fixed commissions. By 2000, there were 12 foreign bank branches and 61 representative offices in South Africa. There has been considerable success in attracting portfolio inflows, as a result of developed financial markets and macroeconomic restraint. By 2000, gross nonresident transactions (purchases plus sales) represented 52 percent of turnover on the equity market, and 23 percent on the bond market (SA Reserve Bank, 2003). Between 1995 and 2002, South Africa received two-thirds of gross market-based capital flows to Sub-Saharan Africa, and 101 percent of net portfolio equity flows. South Africa’s share of all developing countries flows were 3.3 percent and 22 percent respectively (World Bank, 2003, tables A30 and A37). But these inflows have been very volatile, the downside of financial liberalization, and there have been three exchange rate crises since 1994. Against the US dollar, the Rand has depreciated from around R2.60 in 1990 to ZAR7.60 in 2000 to ZAR11.00 at end-2001, followed by an appreciation back to around ZAR7.50 by mid-2003. The substantial currency depreciation and associated capital outflows have offset the solid domestic macroeconomic performance reflected in the fiscal and inflation outcomes. Partly as a result of capital account volatility, but also due to constrained demand and political uncertainty, fixed investment has remained low, even though private sector profitability and productivity improved significantly during the 1990s. Private investment averaged only 12.13 percent of GDP between 1994 and 2002, compared with 10.5 percent in 1993. Even in the midst of a foreign debt standstill and forex-constrained growth in 1988, the investment rate was 12.75 percent. In consequence, GDP growth has been disappointingly low and also unstable, averaging only 1.82 percent per annum between 1990 and 2003, and 2.77 percent per annum between 1994 and 2003. Since population growth was 2.0 percent per annum from 1991 to 2003, per capita income was positive but very low in the latter period.5 The labor market The South African labor market is highly segmented. Table 8.1 distinguishes four categories within the working age population. Nearly a third of the working age population is not economically active, and less than 40 percent is in employment. Of those employed in 2003, only 63.6 percent were in the formal nonagricultural sector, with 7.5 percent in formal (commercial) agriculture, and 28.3 percent in the informal sector (unregistered businesses) and domestic service. In 2003, the “narrow” unemployment rate in South Africa was 31.2 percent, defined as a proportion of economically active people who had actively sought work during the previous four weeks. On the “broad” definition of the labor force – those who want to work but have
182 Stephen Gelb and Anthony Black Table 8.1 South Africa, working age population, 2003
Total population age 15–65 years (millions) Employed in formal sector Employed in informal sector & domestic service Unemployed (broad definition) Total economically active Not economically active
Number in ’000
% of popn
% of EA popn
29 555 8 223 3 270 8 421 19 914 9 569
100 27.8 11.1 28.5 67.6 32.4
n.a. 41.3 16.4 42.3 100 n.a.
Source: Statistics SA, 2003.
become discouraged from actively looking – unemployment was 42.1 percent. Unemployment rates are highly unequal across race and gender: 49.1 percent of Africans are unemployed but only 9.5 percent of whites, and 48.4 percent of women but only 35.9 percent of men (all broad rates). Only 28.6 percent of working age African women are in employment compared with 40.2 percent of working age African men, with 14.1 percent and 29.2 percent respectively are in the formal sector. Extremely high unemployment has resulted from the historical pattern of industrial development, which was based on diamond and gold mining from the late nineteenth century. The forced labor regime established by the mines provided the foundation for racial discrimination in the labor market. Mineral surpluses were directed to domestic manufacturing development from the 1920s, when iron and steel and energy works were established by the state, and tariff barriers and currency depreciation in 1933 enabled growth of labor-intensive consumer goods production for whites. After 1945, capital-intensive production expanded in both consumer durables (autos, electrodomestics) and heavy intermediate goods. Domestic market growth was linked to building political support among the urban white middle class and skilled industrial workers, and the racial income gap widened. In East Asia by contrast, laborintensive export-promotion followed the “easy” phase of labor-intensive import-substitution, contributing to employment and distributional equality. The fixed gold price in the Bretton Woods international monetary system underpinned export revenue stability, in contrast to other commodity exporters. Resource exports financed imports of capital goods, enabling an investment rate of over 25 percent in the 1960s. But by the start of the 1970s, the manufacturing sector was uncompetitive: labor productivity was low because of the apartheid labor and education systems, while low effective protection on machinery and assembled intermediates limited backward integration. As a result, import dependence in manufacturing was high, and the cost structure inflexible. Long-run manufacturing profitability began to falter from the late 1960s, reflecting these supply-side problems, and GDP growth declined from 5.5 percent per annum in the 1960s to 3.3 percent in the 1970s
FDI, Production, and the Labor Market in SA 183
and 1.2 percent in the 1980s (Gelb, 1991). Absorption of new black labor force entrants was low and open unemployment emerged. Unemployment was estimated at 11.8 percent in 1970 and 20.8 percent in 1980 (Simkins, cited in Standing et al., 1996, p. 107). Labor absorption was only 74 percent in the late 1960s and dropped to 12.5 percent by the late 1980s. With less than two-fifths of the population in employment, poverty and inequality are extremely high. Three million people (7 percent of the population) were living on less than $1 a day in 2000, and ten million people (23 percent of the population) on less than $2 a day (Woolard, 2002). The Gini coefficient was estimated at 0.57 in 2000, though it had declined significantly from an estimated 0.67 in 1991. In 2000, 64.9 percent of total income went to the top quintile in the population, as compared with 6.1 percent received by the lowest two quintiles. Poverty and inequality are highly correlated with race, gender, and location (urban versus rural), a legacy of apartheid’s migrant labor system. The poverty rate among female-headed households in 1995 was 60 percent, double that for male-headed households. In 2002, an estimated 11.4 percent of the population was HIV-positive, the second-highest prevalence rate in the world (HSRC, 2003). Prevalence among Africans was 12.9 percent and whites 6.2 percent. With a prevalence rate of 12.8 percent, women and girls are more vulnerable, compared with 9.5 percent for males. Though HIV/Aids is sometimes labeled a “poor peoples’ disease,” all socioeconomic groups are at risk, though the financial impact on poor people is clearly far greater. The unemployment and poverty rates underline the need for laborabsorbing growth, but there are huge disparities of skill across the different segments in the labor market, shown in Table 8.2. Nearly three-quarters of those with some postsecondary education are employed in the formal sector, Table 8.2 Employment status within educational achievement categories Up to Grade 4
Grades 5–10
Schoolleaving
PostAll secondary educational levels
Percent within educational category: Employed 38.6 31.5 40.7 Of which: Formal employment 17.8 19.2 33.6 Informal & domestic 20.5 12.1 6.8 Unemployed (broad) 25.0 27.8 35.7 Not eco active (broad) 36.4 40.7 23.6
77.3
39.0 73.2 3.3
27.8 11.0
13.2 9.6
28.6 32.4
Broad unemployment rate (share of eco active population)
39.3
46.9
46.8
14.6
42.3
Share of education level in total working age population
14.1
47.9
29.7
8.3
100
Source: Calculated from Statistics SA, 2003.
184 Stephen Gelb and Anthony Black
though their unemployment rate of 14.6 percent is high in international terms. But the employment rate is only 40.7 percent of those with schoolleaving certification, and only 33.6 percent are in the formal sector. The remaining 62 percent of the working-age population have at most some high school: only 19 percent of this group (columns 1 and 2) have formal sector jobs, and another 14 percent are in the informal/domestic sector. Among those employed, there are also disparities related to skills. Table 8.3 illustrates the “skills twist”: in every sector except construction, employment of highly-skilled workers has grown since 1985 while employment of unskilled workers has declined. Skilled employment has either grown slowly or declined.6 Though employment grew by more than 10 percent between 1995 and 1999 (while the labor force increased by 13.2 percent), there were significant compositional shifts, with job growth highest in financial services and in transport and communications. Employment in the primary and government sectors declined, and in manufacturing it rose by only 6.7 percent. This reflects changes in output structure during the 1990s, shown in Table 8.4. Mining, manufacturing, and other secondary industry (construction and utilities) all declined, while services increased their share, particularly transport and communications and financial services. Within manufacturing, there were also structural shifts: labor-intensive sectors – food and beverages, textiles, clothing, and footwear – grew slowly (about 0.2 percent per annum), and declined as a share of manufacturing valueadded from 23 percent in 1990 to 20 percent in 2000. Capital-intensive sectors – basic metals, wood products and chemicals – were the fastest growing (the former two at over 4 percent per annum), and increased their shares of MVA (Kaplan, 2003).
Table 8.3 Employment indices by skill category and sector, 1985, 1995, and 2002 (1985 100) Sectors
Agriculture Mining Manufacture Utilities Construction Trade Transport/comms Financial services Govt and community
Semi-/unskilled
Skilled
1985 1995 2002
1985 1995 2002
100 100 100 100 100 100 100 100 100
91 78 90 59 86 82 62 94 87
77 52 76 47 51 80 43 105 99
100 100 100 100 100 100 100 100 100
121 96 104 79 83 103 66 123 126
Source: TIPS, 2003. Online statistical and economic time series.
Highly skilled
130 77 97 84 50 111 51 131 125
1985 1995 100 100 100 100 100 100 100 100 100
197 145 145 173 98 117 117 165 128
2002 284 127 145 227 68 140 120 196 125
FDI, Production, and the Labor Market in SA 185 Table 8.4 Sectoral output shares, 1995 prices Share of gross value added, percent
Agriculture Mining Manufacturing Other secondary industry Transport and communication Financial services Govt and community services Trade and other services Total
Growth rate, 1990–2002
1990
1994
2002
5.0 7.3 22.0 6.9 7.9
5.0 7.4 20.5 6.7 8.3
4.3 5.4 20.2 6.6 11.7
0.8 0.5 1.3 1.7 5.5
15.6 16.0
16.3 16.5
19.6 13.5
4.0 0.6
19.4
19.3
18.8
1.8
100.00
100.00
100.00
2.0
Source: SA Reserve Bank, Online statistical and economic time series.
Rising capital-intensity was linked in part to trade. Between 1990 and 2002, there was a significant shift in exports from minerals to basic processed goods and machinery and equipment, minerals dropping from 61 percent to 36 percent of total exports while the two capital-intensive manufacturing sectors rose from 27.8 percent to 49 percent (TIPS, 2003). The share of manufacturing exports from capital-intensive sectors rose from 56.1 percent in 1993 to 60.8 percent in 1997, while lower trade protection meant import penetration in labor-intensive sectors rose from 55.5 percent to 67.5 percent (Edwards, 2001). Lewis (2001, pp. 46–7) sees “some evidence to suggest that trade liberalisation and increased trade … have induced a structural change in production towards capital-intensive sectors … South Africa has a low and declining share of exports that use unskilled labor, and a high share using more skilled labor,” a problematic outcome given the abundant supply of unemployed, unskilled labor. Looking at distribution, since 1990 the profits share of national income has grown from 42.8 to 48.6 percent. Notwithstanding this significant shift and the unemployment rate, real wage rates have risen rapidly: in the public sector by more than 2 percent per annum since 1985, and in the private sector by over 3 percent per annum since 1995.
2
Foreign direct investment in South Africa
Foreign direct investment has a long and complex history in South Africa. Foreign corporations have been present since Britain established a colony early in the 19th century and the financial system was established by London-based banks. Effective exploitation of the mineral resources from
186 Stephen Gelb and Anthony Black
the late nineteenth century required large capital-intensive operations, which depended on direct and portfolio investment flows from Europe, particularly London, and contributed to the early development of a domestic stock exchange in Johannesburg. Direct investment from the UK, the US, and Europe was important in the establishment and growth of new industrial sectors between the 1920s and 1970s.7 Although FDI continued to flow into mining, during the 1950s and 1960s it went mainly to manufacturing and services, entering behind tariff and regulatory protective barriers looking for market growth as white living standards rose. By the early 1970s, 40 percent of the FDI stock was in manufacturing and 25 percent in financial and business services, with only 15 percent in mining. From the early 1970s on, new FDI flows into South Africa slowed appreciably. Internationally, capital flows shifted in emphasis from direct to portfolio investment, and foreign investors in South Africa were increasingly subject to political pressures in their home countries from the international campaign against apartheid. During the 1980s, this campaign intensified as political instability in South Africa increased, while economic conditions in South Africa also weakened. Foreign direct investors began to exit, with about 225 US corporations and about 20 percent of UK firms departing between 1984 and 1988. In 1985, portfolio inflows also ceased, as foreign bank creditors imposed a debt repayment schedule on the South African government and public sector borrowers. The resulting capital outflows and further economic contraction were significant in shifting white political sentiment against apartheid, especially within business, since it became clear that access to international capital markets, and thus growth, were contingent on democratization. The unbanning of political movements in 1990 ended the disinvestment and trade sanctions phase. At this time, there were more than 450 foreign firms with direct investments in South Africa, and total foreign direct investment liabilities amounted to just below US$8 billion (at the ruling exchange rate), of which 85 percent was from Europe and 13 percent from North America. The reopening of the economy to international capital coincided roughly with the onset of the “emerging markets” investment boom, so that both domestic and international factors shaped the liberal, outwardoriented economic policy regime described above, which aimed to attract foreign investment. In addition to the broad policy stance, direct measures to attract foreign investment have also been put in place since 1994. Investment promotion and facilitation agencies have been established both at national level and in most of the nine provinces. These administer a large suite of more than 35 investment incentive schemes to national and foreign investors on a nondiscriminatory basis. Since 1994, South Africa has signed the TRIMS agreement requiring equal treatment between national and foreign investors with respect to importing, exporting, and access to foreign exchange, and free
FDI, Production, and the Labor Market in SA 187
repatriation of capital and dividends, as well as over 30 bilateral investment treaties, including with most OECD countries. New FDI since 1990 The overriding concern with the financial and macroeconomic aspects of FDI have resulted in a “beauty contest” approach to FDI policy in South Africa in recent years, the key questions being whether the country is receiving “enough” investment relative to “competitor” economies and what further measures are necessary to get more foreign companies “through the door.” Although industrial and trade policy have been heavily concerned with export promotion, there has been little explicit focus on the potential contribution of FDI in that regard, nor on the wider development impacts of FDI emphasized by international organizations such as the World Bank and UNCTAD, involving direct employment creation, technology transfer or productivity spillovers vertically or horizontally (UNCTAD, various years). Notwithstanding the emphasis in policy, FDI inflows have been disappointing in quantitative terms. Low growth of aggregate demand, currency volatility, and uncertainty related to the transition – would the new government’s policy stance be stable and investor-friendly – have all contributed to this outcome. Between 1990 and 1993, total inflows averaged $46 million per annum. Gross inflows rose rapidly after the transition in 1994, averaging $1861 million per annum to 2002.8 In 2002, South Africa received US$754 million, and all developing countries US$162 billion in total, so that South Africa’s share was just less than 0.5 percent (UNCTAD, various years). South Africa’s average inflow is equivalent to just over $41 per capita (based on the 2001 population), or 1.4 percent of per capita income. By comparison, average FDI inflow per capita for all developing countries was $40.42, equivalent to 3.5 percent of per capita income. Between 1995 and 2002, net inward FDI to South Africa was 1.5 percent of the total for all developing countries, though the country received 12 percent of net inward FDI flows to sub-Saharan Africa (World Bank, 2003, table A29).9 These shares are significantly lower than the corresponding proportions for portfolio flows cited above, reflecting much greater success in attracting the latter form of foreign investment. The outward movement of South African capital is worth noting here. Several major resource-based and financial services corporations have abandoned their apartheid-era conglomerate structure, relocated their head offices to the UK and the US, and become significant international investors. In addition, many other South African corporations have invested in subSaharan Africa, newly opened to South African firms by the end of apartheid. The stock of South African direct investment assets in Africa grew 18 percent per annum between 1995 and 2001 (SA Reserve Bank, 2001, 2003), with investments concentrated in mineral resource extraction and in marketseeking services, notably finance, retail, and infrastructure (EDGE Institute, African Investment Database, 2003).
188 Stephen Gelb and Anthony Black
The London Business School/EDGE Institute survey of foreign investors10 confirms the macro picture of limited success in FDI inflows. This was the first comprehensive survey since 1994 to examine FDI across all home countries and economic sectors. A sample of 162 firms was surveyed from a sample frame of 516 firms which had entered South Africa for the first time since 1990.11 The survey therefore excluded new investments by firms which entered South Africa during the apartheid era. The sectoral distribution12 and size by sector of affiliates is presented in Table 8.5. The affiliates are small: the sample medians are 90 workers and $1.94 million fixed capital respectively. Only 10 percent of the firms have a labor force above 1000 workers. This suggests FDI is unlikely to be a vehicle for either large capital inflows or for significant direct employment creation. Only the primary and the trade and hospitality sectors – each with a few firms only – have large medians. In the sample as a whole, 55 firms, more than one-third, had between 10 and 50 employees. There was an increase of 67 percent in the median size of firms’ labor force from their entry to 2000, while 44 percent of firms at least doubled the size of their workforce. However, excluding greenfield entries where the initial workforce is often a fraction of the intended complement at full capacity, the median growth in employment between entry and 2000 was only 14 percent, and only one-third of firms doubled in size. Nonetheless, although foreign firms are relatively small, their employment creation record has been relatively good.13 Nearly one-third of the sample is in skill- or knowledge-intensive sectors – financial and business services, IT and pharmaceuticals – where size is not correlated with turnover. Nearly 40 percent of firms with fewer than 50 workers are in the top three-fifths of firms by value of turnover, while a
Table 8.5 Foreign affiliates in South Africa, sectoral distribution and size in 2000
Primary Basic consumer goods Intermediate goods Machinery and equipment Infrastructure and construction Trade and hospitality Financial and business services Information technology Pharmaceuticals All sectors Source: LBS/EDGE survey.
No of firms
Percent of sample
Median size of workforce
Median capital stock($m)
5 21 27 31 19 8 33 13 5 162
3 13 17 19 12 5 20 8 3 100
1500 78 85 100 147 220 70 55 23 90
24.50 0.65 2.45 2.00 0.22 13.69 0.86 1.01 0.08 1.94
FDI, Production, and the Labor Market in SA 189
quarter of the firms in the top quintile for turnover have fewer than 250 workers. A significant minority of affiliates outsource much of their operation, especially production, themselves focusing on strategic management, marketing, and technical services. In some cases, this was different than other affiliates of the parent or a shift from their intention on entry to South Africa. This behavior is a response to risk, including political and social risk (a new and unknown government, HIV/Aids, crime), currency risk (secular depreciation of the ZA Rand), and market risk (slow economic growth in both South African and most of the region). Table 8.6 shows the distribution of home countries. Nearly 60 percent of the investors are from Europe, with 33 percent from the UK, Germany, and France who have approximately equal numbers of firms in South Africa. The European firms are distributed roughly equally across the three manufacturing sectors, infrastructure and, financial and business services. The North American firms are concentrated in materials processing, financial and business services and IT, while East Asian firms are concentrated in manufacturing, particularly materials processing and machinery and equipment. The parent firms cover the spectrum from small multinationals with operations in three or four countries to global giants, though many of the latter were in South Africa before 1990. The median number of affiliates is 20 and global labor force size 10 250 workers. Mean spending of parents on firmspecific assets – advertising and R&D – is between 1 to 2 percent of global turnover. Only 15 percent of the investors had no emerging market experience at all prior to South African entry, while more than half were already in three or more emerging market regions, and over one-third already had investments in Africa. Though nearly half the South African affiliates are tiny within the parent’s global turnover (share of less than
Table 8.6 Geographic distribution of parent firms by sector (number of firms)14 North America
Europe
East Asia
MENA
Other
Total
Primary Basic consumer goods Intermediate goods Machinery and equipment Infrastructure and construction Trade and hospitality Financial and business services Information technology Pharmaceuticals
3 1 8 5 1 2 10 6 0
1 17 10 16 15 6 17 6 4
1 2 9 9 2 0 2 0 0
0 0 0 0 0 0 0 0 0
0 1 0 1 1 0 4 1 1
5 21 27 31 19 8 33 13 5
Total
36
92
25
0
9
162
Source: LBS/EDGE survey.
190 Stephen Gelb and Anthony Black
0.5 percent), another quarter of affiliates contribute more than 5 percent of global sales. Looking at sectors, the basic consumer goods sector is dominated by European mid-size firms which have only recently started to expand into developing economies internationally – they have a small number of affiliates, mainly in Central and Eastern Europe and Asia, and would see South (or Southern) Africa as a potentially promising market. Materials processing firms from all three regions are somewhat smaller than the median, with the local affiliate’s capital stock and share of turnover relatively significant. The machinery and equipment parent firms are large and the South African affiliates’ share insignificant, perhaps because their products are better suited to economies of scope. They have a larger number of affiliates and a more diversified emerging markets presence than in materials processing. Most of the infrastructure firms have extensive emerging market experience, possibly related to recent expansion of the private sector’s role in developing countries. In financial and business services, firms have many widely dispersed affiliates with small workforces and low spending on both R&D and advertising, suggesting that they are providing high-end services relying on trust and personal networks. The South African affiliates contribute a very small share to global revenues. Finally, IT firms have a relatively large number of affiliates and substantial emerging market experience and spend a large share of turnover on R&D. Mode of entry is important for the economic impact of FDI. Mergers and acquisitions may imply smaller additions to capital stock and employment than new operations established via greenfield or joint-venture, but might nonetheless enable more rapid technology and skill transfer from the foreign investor to domestic partners to improve their international competitiveness and exports. Within the sample, full acquisitions accounted for 31 percent and partial acquisitions another 14 percent. South Africa was the only one of the four countries in the survey (Egypt, India and Vietnam were the others) with a significant share of acquisitions, reflecting the maturity of South Africa’s market for corporate control and suggesting that foreign investors find South Africa’s corporate environment familiar. Acquisitions were particularly prominent in materials processing and IT, and low in infrastructure and financial and business services, probably due to regulatory restrictions. New enterprises comprised only just over half the foreign investments of newly entering firms, of which greenfields (common among smaller parent firms) contributed 32 percent and joint ventures the remaining 23 percent. However, the direct contribution to output and employment growth was limited. Most greenfields are very small: 71 percent had fewer than 100 workers in 2000, and 50 percent had a capital stock value at start-up of less than $1 million. On the other hand, the largest affiliates are dominated by partial acquisitions, while full acquisition is the most frequent mode of entry among medium-size affiliates.
FDI, Production, and the Labor Market in SA 191
3
FDI and globalized production
Table 8.7 presents the means of affiliates’ sales into four distinct markets: domestic, regional,15 global, and non-arm’s length exports to other affiliates of the parent, indicating integration into international production chains or networks. The distinction between regional and global exports is important, since the former are often a “vent for surplus” which fluctuate inversely with domestic demand, whereas the latter involve “learning-by-exporting” and productivity improvement for domestic firms (Rankin, 2001), while for foreign affiliates, they reflect market-seeking versus efficiency-seeking motives for investment. Except in the primary and infrastructure sectors, Table 8.7 (which weights each firm equally irrespective of turnover) suggests that most foreign firms entered South Africa during the 1990s seeking markets: the mean share of domestic plus regional sales in the other sectors are all close to 80 percent of turnover. Outside of mining, market-seeking has been the main motive historically, and it appears that this strategy has now been extended into regional markets. Many firms indicated that their intention at entry was to establish themselves initially in the domestic market before moving to supply regional markets also, and the data reflect this strategy, as eight sectors have larger regional shares in 2000 than at entry, with a roughly constant total for domestic and regional sales in five. Several firms reported
Table 8.7 Market orientation by sector, South Africa (means of % of affiliates’ sales)16 Sector
Primary Basic consumer goods Intermediate goods Machinery and equipment Infrastructure and construction Trade and hospitality Financial and business services Information technology Pharmaceuticals All firms Source: LBS/EDGE survey.
Domestic market
Regional market
Global market
2000 Entry
2000
Other affiliates
Entry
2000
Entry
Entry 2000
17 94
13 85
10 5
0 7
73 1
87 8
0 0
0 0
84 86
73 77
1 2
6 10
15 6
19 8
0 7
1 5
67
69
3
7
24
22
6
3
80 84
82 78
1 6
2 10
19 6
16 8
0 5
0 3
85
53
1
26
8
14
7
7
96 81
87 73
4 3
11 9
0 12
0 15
0 4
1 3
192 Stephen Gelb and Anthony Black
that their Zimbabwean sales were much lower than anticipated as that economy (by far South Africa’s largest regional trading partner at the start of the 1990s) went into decline. Without the political crisis there, the regional sales share would probably have been significantly higher. The number of efficiency-seeking firms is small: the proportion of sales into global markets or to related firms is low, except for resource-seeking investors in the primary sector. Firms in materials processing, trade and hospitality (including tourism), and infrastructure may have entered intending to supply global markets, but in the latter two sectors, global markets’ share declined after entry. Global exports rose in consumer goods and IT possibly due to currency depreciation. Sales-weighted data confirm this picture. The sample is very unevenly distributed in this respect. Just over half the affiliates, 77 firms, are small, with sales below $10 million in 2000, and together account for only 4 percent of the sample’s aggregate turnover in all markets. This group of firms reoriented sales after entry most significantly from the domestic to the regional market, though the former still took 75 percent in 2000 and the latter only 12 percent. By contrast, 77 percent of aggregate sample turnover is contributed by 25 large firms with sales in 2000 above $75 million. Although 22 percent of this group’s aggregate turnover is sold on global markets or to other affiliates, these exports come from only nine large exporters, who sell on average 55 percent of their output outside Africa. This amounts to 80 percent of the sample’s exports outside Africa. Five of the nine are mining companies, and only two affiliates (in automotive components and industrial chemicals) have significant exports (by value) to related affiliates of the parent. The other two are in transport services and furniture and wood products respectively. Other research confirms the dominance of market-seeking, rather than efficiency-seeking, foreign investment. The 1999 National Enterprise Survey sampled 1425 firms, of which 68 fit the four criteria of the LBS/EDGE survey. Within this subgroup, the mean share of domestic sales was 68.5 percent of turnover, and exports to Africa were 15 percent (Gelb, 2001).17 As shown above, two-thirds of entries involve linkage with a South African firm in the form of full or partial acquisition or Joint Venture ( JV). The first column in Table 8.8 shows that market-seeking investors entering via these modes seek targets which already have significant market shares. Greenfields’ average share at entry was only 12 percent, but the increases in market share for both greenfields and partial acquisitions strongly suggest the competitive advantage of foreign firms. The limited degree of integration into global production chains and networks is something of a puzzle, given the relatively advanced level of industrial development and the presence of large pools of unemployed, unskilled labor. Brief examination of three sectors – autos, clothing, and financial services – provides further insights. All three sectoral cases confirm
FDI, Production, and the Labor Market in SA 193 Table 8.8 Domestic market share by mode of entry, South Africa (%) Market share
Greenfield Acquisition Joint venture Partial acquisition All firms
Entry
2000
12 34 31 25 26
19 35 33 37 30
Source: LBS/EDGE survey.
the survey results that globalization of production in South Africa has progressed to a limited degree only, but to the extent it does exist, it is driven in large measure by foreign investors that have established new, or adapted existing, operations for exporting. But as the LBS/EDGE survey shows to be the case for FDI generally, the numbers of firms, their employment and output, are small on average. South African firms on the other hand have largely been unable or unwilling to reorient to exports without the direct involvement of foreign firms, in part because of the existing industrial structure of production and consumption restricts their feasible choice set as they define it. Autos The auto sector illustrates the importance of the global producers at the apex of the producer-led commodity chain, who have integrated both their local assembly operations and parts of the local component industry into their global commodity chains.18 Historically, a rapidly growing domestic market and high tariffs had attracted foreign car manufacturers into South Africa. In the 1960s, white car ownership per capita lagged only the US, Canada, and Australia, and ownership doubled each decade from the 1940s to the 1980s. Local content requirements from the 1960s were weight- rather than value-based, and market growth slowed from the 1970s. By the 1980s, the industry was high-cost, low-volume, dependent on technology-intensive component imports and uncompetitive internationally. Combined with political pressures, foreign firms disinvested, and by the late 1980s only three of the seven vehicle manufacturers were majority foreign-owned, while most large component groups were locally owned. Over the past decade, a sectoral industrial policy, the Motor Industry Development Programme, was introduced involving tariff reductions and import–export complementation arrangements, to encourage a phased transition from completely knocked down (CKD) assembly to full manufacturing, to enable higher volumes and increasing localization of components. The policy forces carmakers to generate exports to offset duties on imported components
194 Stephen Gelb and Anthony Black
and vehicles, while increasing competitive pressure by reducing import duties. The policy has encouraged substantial export-linked FDI, led by the German firms. The major vehicle manufacturers have (re-)purchased equity stakes in their licensed operations and linked them into global supply processes, and have encouraged their foreign suppliers to enter, arranging large export contracts and facilitating access to their global networks (Gelb and Black, 2004b). Vehicle exports increased from 11 400 units in 1992 to about 110 000 units in 2002, with an even greater rate of growth in a small range of component exports, such as leather carseats and catalytic converters, the latter increasing from R197 million in 1994 to R4683 million in 2001. [change ‘or’ to ‘and] Other important component exports include tyres, exhausts, wheels, engine parts and wiring harnesses – the bulk of component exports reflect materials processing and resource beneficiation rather than complex manufacturing operations. The component subsector is divided between mainly foreign-owned export-oriented firms, and low volume, low local content firms supplying the domestic market. Many of the latter have been unable to adjust and have contracted, while those that have become export-oriented operate either in the aftermarket or as second-tier original equipment suppliers. Local operations acquired by foreign firms have done best: according to local management of the company acquired by the German airconditioner and radiator producer Behr, “the MIDP was starting to bite … [without a foreign investor, the company] could have stagnated into the aftermarket or even died” (cited in Gelb and Black, 2004b). Garments The South African garment industry developed behind high protective barriers for both garments and textile inputs. Its early growth was in the major cities, but apartheid legislation affected location by restricting African workers moving to metropolitan areas and subsidizing decentralized development in “bantustans” and “border areas” in “white South Africa.” The sector was closed, imports and exports each accounted for less than 10 percent of domestic output. The sector was dominated by South African retailers at the top of “buyer-led chains” producing for the domestic market, where high prices made wage premia possible. Foreign investment was minimal until Taiwanese firms began to take advantage of decentralization subsidies in the 1980s. Local firms also exploited the subsidies, but used them to compensate for very low productivity rather than adapting their urban “manufacturing systems” (Gibbon, 2002). As in autos, the 1990s have seen the lowering of trade barriers. In clothing, this has led to import competition and the decline of traditional production centres as firms sought locations with low wages and poor enforcement of regulations, moving to peri-urban sites or neighboring countries.20 On the other hand, export potential has increased significantly due to currency depreciation, lower input prices due to trade liberalization in textiles, and the
FDI, Production, and the Labor Market in SA 195
African Growth and Opportunity Act which has eased access to the US market since late 2001 and driven a new round of FDI in the industry. Between 1995 and 2001, exports grew rapidly at 9.4 percent per annum, but the ratio of exports to domestic output is still only 19.1 percent (having risen from 6.8 percent) while imports also grew rapidly and the sector had a small but persistent trade deficit (Moodley and Velia, 2002). But the export market is increasingly segmented. On one hand, mainly South African-owned firms in the Cape Town area are exporting mass-market woven products with domestic cloth to the EU market, as an addition to their production of higher-end products for the domestic market, where they are tied into local retailer-led chains. On the other hand, mainly foreign-owned (Asian) firms are producing low-end knitted garments using imported fabric for US retailers.21 These firms have little presence as domestic suppliers and operate in the peri-urban low-wage areas (Gibbon, 2002, p. 56). The latter group has become considerably more prominent during the 1990s as the destination of South African clothing exports changed significantly: the EU dropped from a share of around 60 percent to 20 to 25 percent, and the US increased from less than 10 percent to around 60 percent. Thus, as in autos, foreign firms are participating in raising exports from South Africa with local firms lagging behind. Gibbon (2002) attributes this outcome to different organizational cultures among local and foreign firms, due to their differing histories. On the one hand, the local firms were shaped by the domestic market in which high income inequality contributed to the emergence of powerful retail chains in different market segments. Local manufacturers have partly been unable to adapt to low wage, low skill environments allowing them to export, because links with domestic retailers required them to develop costly overhead structures (including management and technical capabilities) to facilitate frequent style and fabric changes and which were ill-suited to export markets. And they have partly been unwilling to establish themselves in export markets because domestic buyer-led chains provided a “comfort zone” of stable demand, with exports seen merely as a “vent for surplus”: Gibbon reports that SA firms have not moved into the United States less because of “difficulties meeting US market disciplines, but more a reluctance or difficulty in taking on major increases in volume” (2002, p. 60). Foreign producers, on the other hand, especially from Asia, brought to South Africa a “portable business model” learnt in the context of the MultiFibre Agreement, which allows them to adapt to low skill levels, and produce high volumes despite low specialization. In addition, they have preexisting networks which have given them access to foreign, especially US, buyers and allowed them to exploit opportunities such as AGOA. But it is important not to exaggerate the scale or potential: not only are exports still a small share of sector output, but the number of foreign firms is also small. Only six of a sample of 25 “exporting” firms (exporting more than 10 percent of turnover) were foreign-owned (Gibbon, 2002, p. 44). It is
196 Stephen Gelb and Anthony Black
unlikely that FDI alone will significantly expand exports and employment in the clothing sector, so that if local firms do not adapt their activities and expand, it will not happen. Financial services As services are produced and consumed simultaneously, services sector internationalization has involved “market-seeking” FDI to establish a commercial presence. But during the 1990s, much of FDI in services was aimed at constructing “global service networks,” as competitive pressures pushed firms to offer corporate clients a global “package deal” involving both product diversification and a wider geographical spread of operations (Dicken, 1998). This investment can be interpreted as “market-sustaining” rather than market-seeking in a narrow sense. In financial services, foreign entry facilitated integration into global financial markets, enhancing access to capital, so that emerging market governments and international banks had a mutual interest in deregulating and removing entry barriers. Foreign banks were allowed to open branches in South Africa from 1995,22 but were subject to two important restrictions: local liquid asset and minimum reserve requirements had to be maintained, and depositors had to maintain a minimum balance of $125 000, effectively excluding foreign banks from the retail market. These conditions significantly protected the “big 5” domestic commercial banks, which were not large by global standards,23 but controlled around 80 percent of the domestic market, where they were threatened in the early 1990s by non-banks in retail services. Foreign banks and securities traders were also allowed to operate on the Johannesburg Stock Exchange ( JSE) from 1995, with domestic broking houses given little protection. Even with the banking restrictions, the South African financial sector was attractive. In 1996, the JSE ranked fourteenth globally by market capitalization, and in the global bull market of the mid-90s, global institutions needed information and analysis on South African private and public securities. There were 638 listed companies on the JSE, many of whom were looking to expand trade and investment outside South Africa, creating a significant demand for offshore financial services. There were also around 1500 foreign companies in South Africa with direct investments or agencies and representative offices, to whom global service networks offered advantages. Finally, the expected wave of privatization and black economic empowerment deals promised lucrative investment and merchant banking returns. Between 35 and 40 foreign financial institutions established a presence in banking and securities trading in South Africa between 1996 and 2000, and others with an existing agency upgraded their operations. Acquisitions were not possible in banking due to the regulatory restrictions, but several of the large local securities traders were taken over by major international institutions. This increased competition in both the product market and also the labor market, where financial sector expertise was at a premium. The foreign entrants
FDI, Production, and the Labor Market in SA 197
have helped to integrate South Africa into global financial networks, enabling South African corporations to access global finance for their offshore operations, and enabling foreigners to purchase South African financial assets, as well as expanding the range of financial services available to foreign corporations in South Africa from the agency operations. Unlike the South African firms in the two manufacturing sectors discussed above, local banks have been able to move into “export” markets, establishing an international presence to offer services to buyers of South African assets in the European and US markets, and also moving into Africa where, though markets are small, they have significant competitive advantages over many international banks. But the global financial market downturn from mid-2000 followed by “9/11,” together with disappointing growth and competition in the overtraded domestic market, cut potential revenues in South Africa, while operating costs were high. Unlike autos and clothing, foreign financial firms do not have full control over their cost structures. The regulatory requirement to maintain a local balance sheet imposes costs, while externally supplied capital carries a country risk premium due to South Africa’s “emerging market” rating. In one major foreign bank, South African corporate loans faced hurdle rates at US blue chip levels, and 50 basis points higher than Central European corporates. Financial sector depth in South Africa – part of the attraction for foreign banks – also raised operating expenses, including the salary structure. As a result, many second-tier international banks have cut back or withdrawn entirely. The top-tier banks – Citibank, JP Morgan Chase, Deutsche and HSBC – have maintained high-cost international operations including in South Africa, aiming to raise their global market share, though even they have come under pressure in South Africa. The major domestic banks were able to protect their position in the face of foreign entry, though all the second-tier domestic banks disappeared.
4
The labor market impact of FDI
It has been shown above that FDI inflows from new investors in South Africa have not contributed significantly to gross employment creation – the affiliates are small and relatively few in number in all sectors bar primary and trade and hospitality. The LBS/EDGE survey also examined a series of other issues of relevance to labor market development. Taken together, the results reported here suggest that new foreign firms entering South Africa are contributing to the skills bias and growing inequality in the labor market. Skills shortages The “skills bias” discussed in Section 2 has led to a commonly expressed view that growth in South Africa faces a binding skills constraint in the labor market (Lewis, 2001). This would be expected to discourage FDI but Table 8.9, reporting affiliates’ responses to a question about the availability of suitable
198
Table 8.9 Availability of suitable skilled employees, sector means in 2000 (means at entry in brackets)
Executive managers Professionals Operations managers Skilled workers All skilled employees
Primary
Basic consumer goods
3.60 (4.00) 4.40 (4.40) 4.20 (4.20) 4.20 (4.20) 4.10 (4.20)
3.62 (3.74) 3.90 (4.26) 3.90 (4.17) 4.14 (4.26) 3.89 (4.09)
Intermediate Machinery goods and equipment 3.56 (3.54) 3.96 (4.04) 3.74 (3.62) 4.15 (4.00) 3.85 (3.80)
Infrastructure
3.58 (3.16) 4.13 (4.03) 3.58 (3.58) 4.10 (4.06) 3.85 (3.71)
Source: LBS/EDGE survey. 1 “never available,” 5 “always available.”
3.47 (3.35) 4.16 (4.11) 3.33 (3.39) 3.74 (3.63) 3.71 (3.64)
Trade and hospitality
Financial and business services
3.75 (3.25) 4.50 (4.38) 4.00 (3.63) 4.13 (3.63) 4.09 (3.72)
3.42 (3.64) 4.13 (4.28) 3.97 (4.03) 4.06 (3.94) 3.90 (3.97)
Information Pharmaceuticals All technology
3.38 (3.23) 4.15 (4.00) 3.58 (3.50) 3.50 (3.58) 3.67 (3.56)
4.40 (4.00) 4.40 (4.40) 4.20 (4.00) 4.40 (4.40) 4.35 (4.20)
3.56 (3.48) 4.11 (4.16) 3.76 (3.76) 4.03 (3.96) 3.87 (3.84)
N firms
160 (157) 161 (158) 159 (156) 161 (158) 162 (159)
FDI, Production, and the Labor Market in SA 199
(in terms of quality and price) skilled employees,24 suggests that they are unconcerned about the issue. The overall mean score for the full sample is 3.87 in 2000 (compared with 3.84 at entry), which is very close to 4.0, meaning “mostly available.” All the means are greater than 3.0 (“sometimes available”), though some sectors had fairly significant shifts from entry to 2000. Suitable executive managers are the most difficult employees to find, though in some sectors – machinery and equipment, trade and hospitality and pharmaceuticals – there was significant improvement after entry. By contrast, professionals were easiest to find both at entry and in 2000, though scores declined, particularly in basic consumer goods. Financial and business services also declined after entry, while trade and hospitality, machinery and equipment, IT, and infrastructure improved substantially. Greenfield entrants were generally most pessimistic about skilled labor at entry (though obviously not discouraged from entering), but their perceptions improved substantially on exposure to the local labor market. Surprisingly, partial acquisitions who presumably had local knowledge prior to entry, moved in the opposite direction, from most optimistic at entry to significantly more negative views. There may be a difference in the expectations of the two groups of firms, with greenfields expecting less based on experience in other developing or middle-income economies and partial acquisitions forming expectations based on developed country norms. In responding to questions about the administrative and official institutional environment, which affects transactions costs for key assets, firms expressed strong concerns about immigration barriers to foreign workers, which raise the difficulty of bringing in employees from operations elsewhere. This is a much-discussed issue in South Africa and the LBS/EDGE survey confirmed more anecdotal evidence: this indicator was the only one of ten with a mean score below 3.0 (“somewhat conducive to profitable operations”), was ranked worst in seven of nine sectors, and declined between entry and 2000 in eight of the nine sectors. Human capital accumulation Foreign firms are seen to be a potentially important source of labor skills transfer to the domestic economy, and training spending provides an indicator of skills transfer via investment in human capital, even though it tells us nothing about the quality of the training, or its firm-specific characteristics. Table 8.10 shows that just under one-third of firms spend below 0.5 percent of turnover on training, and another third between 0.5 percent and 2 percent. Though we have just seen that foreign firms do not perceive there to be a serious constraint on high-skilled labor, these levels of expenditure seem low, even though foreign affiliates probably spend slightly more than domestic firms on training.25 There is little correlation between firms’ training expenditure and perceptions of skilled labor availability. There is also little variation by firm size, and the largest firms (over 1000 workers)
200
Table 8.10 Training expenditure (% of affiliates in sector) Training as % of sales
Primary
0 – 0.5 0.5 – 2 2–4 4–8 8 – 15 Over 15
20 40 20 0 0 20
Basic consumer goods
Source: LBS/EDGE survey.
22 28 6 22 22 0
Intermediate goods
Machinery and equipment
Infrastructure
Trade and hospitality
33 33 19 7 0 7
35 45 10 6 3 0
21 32 5 11 21 11
38 38 13 0 0 13
Financial Information Pharmaceuticals All sectors and business technology services
33 24 12 18 12 0
15 38 8 31 0 8
0 20 60 0 0 20
28 33 13 12 8 5
FDI, Production, and the Labor Market in SA 201
spend less on training than smaller firms, a disturbing result in terms of skills upgrading. It is surprising that affiliates in consumer goods appear to be spending more on training than those in machinery and equipment, which is likely to require higher skilled workers. In services, many infrastructure and IT firms also spend over 4 percent of turnover, but only one trade and hospitality firm is investing significantly in its labor force. There is a strong positive correlation between parent firms’ global R&D expenditure levels and affiliates’ training expenditure, presumably because the returns on R&D depend in part on employees’ ability to use technology. There is also a correlation between human resource investment and firms’ perception of their own performance, relative to their expectations at entry. Better performing firms spend more on training than poor performance firms, and it seems reasonable to hypothesize that causality runs from investment in human resources to performance, rather than the reverse. Black economic empowerment (BEE) In 1994, equity ownership and management of South African firms were overwhelmingly white, with an obvious need for “black economic empowerment” (BEE), or asset transfers to black people. The initial approach was “market-driven,” with white-owned corporations selling equity financed by loans secured by future earnings flows. This model inevitably suffered after the 1998 collapse of the Johannesburg Stock Exchange: at that time blackcontrolled companies accounted for about 7 percent of market capitalization, but by 2002 this had dropped to 2.2 percent. In 2001, a politically powerful Black Economic Empowerment Commission recommended a more interventionist strategy focused on ten-year sectoral targets for ownership, management and high-level occupations, and government procurement. These have now been formalized in sectoral “transformation charters.” The mining charter, for example, requires the current level of 14 percent black mining ownership to increase to 26 percent by 2014. The LBS/EDGE survey offered an opportunity to investigate BEE26 among foreign investors. The survey shows that foreign entry has not been a significant vehicle for expanding BEE ownership, but affiliates have been fairly effective in promoting black participation in high skill job categories. Given BEE’s recent origins, very few black-owned companies have been available for acquisition, and the first row in Table 8.11 shows that for the sample as a whole, the mean share of BEE ownership amongst all foreign investors was just 2 percent at the point of entry into South Africa. Seven percent of foreign firms had more than 10 percent BEE ownership,27 and in these firms, black owners had a relatively high stake averaging 41 percent. At the time of the survey (“now”), BEE ownership in the sample as a whole had risen to 3 percent, and the proportion of firms with BEE owners to 12 percent. Within this group, the share owned by BEE owners had dropped to 33 percent.
202 Stephen Gelb and Anthony Black Table 8.11 Black economic empowerment – all sectors
Ownership then Ownership now Executive management then Executive management now Professionals then Professionals now Operations management then Operations management now Skilled non-managerial then Skilled non-managerial now
% BEE all firms
% firms with BEE
% BEE per BEE firm
2 3 5 11 6 17 14 28 31 46
7 12 17 46 26 52 46 81 69 91
41 33 29 25 24 33 30 34 45 50
Source: LBS/EDGE survey.
Another survey, the National Enterprise Survey carried out in 1999–2000, included 157 foreign firms, whose BEE ownership share was very similar: mean level of BEE ownership was just below 3 percent, while 6 percent of foreign firms had more than 10 percent BEE ownership, and the average BEE share in that group was 41 percent (Gelb, 2001).28 Table 8.11 also presents data on the four skilled labor categories used above. BEE presence in the top end of the labor force is much more substantial than in ownership. BEE executive management in foreign affiliates has risen from 5 percent to 11 percent, with nearly half the affiliates (46 percent) having black executive managers at the time of the survey, compared with only about one in six (17 percent) at the time of entry. Among professionals, operations managers and other skilled categories, the proportion of companies with BEE employees, and the proportion of BEE employees in the sample as a whole, had both risen substantially since affiliates entered. There is significant sectoral variation, with trade and hospitality firms performing best, perhaps due to regulatory and public sector procurement requirements that firms have BEE ownership and management. Finance and business services and IT did surprisingly poorly, though the latter have increased rapidly from a low base. Investors’ home country matters: firms from English-speaking countries have transferred ownership and executive management positions, but not other high-skilled jobs, whereas East Asian firms have transferred no equity, but have numerous blacks in all high-skill jobs. Small firms (fewer than 100 employees) have done far better than large in all the skilled job categories, including executive managers. A BEE equity stake has value in certain regulated activities, so that BEE ownership and management in JVs was high at entry. Finally, although firms were largely
FDI, Production, and the Labor Market in SA 203
unconcerned about skilled labor availability, those firms which did experience skilled labor shortages more sharply were more willing to hire blacks for skilled occupations than firms less concerned about skill constraints.
5
Conclusions
The majority of new foreign investors entering South Africa during the 1990s established small or medium size affiliates with limited impact on employment creation and capital inflows. Nearly half the entrants acquired existing operations, rather than setting up new enterprises via greenfield or JV entry. Many investors mitigated risk by limiting the irreversibility of their investment, by outsourcing production or focusing on service provision rather than more capital-demanding manufacturing operations. Most firms entered for market-seeking purposes, though the relevant market encompasses both the domestic economy and neighboring countries. The entering firms are by and large well-established but mid-size multinationals with significant experience in developing economies. Some may have delayed entry to South Africa until the 1990s because of political factors. There appears to be more continuity than change in FDI flows into South Africa during the 1990s, compared with earlier inflows between the 1950s and the 1970s. Notwithstanding the shift in level and composition of global FDI flows during the 1990s, and the political and economic shifts in South Africa, the dominance of market-seeking FDI has not changed substantially. This will have disappointed not only government policymakers, whose primary concern has been the quantity of inflows and their macroeconomic impact, but also international institutions and policy analysts who have advocated FDI in recent years as a primary vehicle for export-oriented industrial development. The capital structure and domestic market orientation of South African industry at the beginning of the 1990s have begun to change during the 1990s, but this process has only just started, and only a small proportion of foreign firms have contributed to it thus far, notwithstanding the relatively high proportion of entrants via acquisition. Nonetheless, benefits have accrued to some local firms whether via acquisition or increased competition with foreign firms, as well as to black highskill labor, and the benefits of FDI may only emerge on a greater scale and scope only over the longer-run.
Notes 1. This chapter draws heavily on Gelb and Black, 2004a and 2004b, and on Gelb, 2003a, which were written as part of the London Business School project on FDI in Emerging Markets. We would like to thank the Centre for New and Emerging Markets at LBS for support, particularly Saul Estrin. Section 1 also draws on Gelb, 2003b.
204 Stephen Gelb and Anthony Black 2. Gelb 2003b contains a full discussion of the political economy of liberalization in South Africa. 3. There is also a single mention of “the important role of multinational corporations in integrated value matrices.” 4. From 1985–93, the current account was maintained in surplus to finance a capital outflow to repay debt. 5. South Africa’s population in 2001 was 44.5 million people, of whom 79 percent were african, 9.5 percent white, 9 percent “coloured” (mixed race) and 2.5 percent “indian,” according to the old apartheid racial categories. Real GDP per capita (1995 prices) was US$4013 (ZAR14554) in 2002, and in PPP terms, US$9401 (2000 data). 6. In some sectors, particularly construction, the data may reflect outsourcing and labor contracting, re-classifying unskilled workers into different sectors, but this is unlikely to change the overall picture. 7. Particularly France and Germany after 1945. The Japanese government barred direct investment in South Africa for political reasons, but many South African companies licensed technology from Japanese corporations. 8. This average is inflated by the 2001 figure of US$6789, which includes a very large capital flow resulting from the unbundling of Anglo American and de Beers, two major South African mining corporations with substantial cross-holdings since the 1920s. The former relocated to London in the late 1990s. The unbundling is formally defined as FDI but its economic effects were those of a portfolio flow. 9. Given substantial outward investment by SA corporations, net inflow data mask significantly larger gross inflows, the relevant variable for evaluation of inward investors. 10. The survey was part of a four-country study, including Egypt, India, and Vietnam, reported in full in Estrin and Meyer (2004). 11. The sample frame was compiled by The EDGE Institute using data from market research companies, media reports, foreign trade missions and international chambers of commerce. All firms in the population met four criteria: at least 10 percent foreign ownership; at least ten employees in South Africa; some valueaddition in South Africa; and initial entry to South Africa after 1990. Another 452 companies met the first three criteria but had entered before 1990. 12. The ISIC was re-structured into nine sectors to take account of specific factors of interest in the larger project. The sectoral and home region distribution of the sample (Tables 8.7 and 8.9) closely match the sample frame. 13. Of course, it is net job creation that is ultimately of concern: foreign firms may have increased their labor force by enhancing market share at the expense of domestic competitors who shed workers as a result. 14. “Other” includes Australia, India, Nigeria, Mauritius, and Turkey. 15. This may refer to Southern Africa or sub-Saharan Africa, depending on the respondent’s interpretation. 16. Since firms’ entry dates differ, the change from “entry” to “2000” cannot be interpreted as a growth rate. Data in Table 8.7 are not weighted by firm turnover. 17. The NES excluded the primary sector. 89 firms with more than 10 percent foreign ownership had entered South Africa before 1990, and these had mean sales of 85.7 percent to the domestic market and 8.9 percent to the region. Average turnover was much larger for pre-1990 entrants than post-1990. 18. Since most of the assemblers had been present in South Africa prior to 1990 as direct investors and/or licensors, they were not regarded as new entrants for the purposes of the LBS/EDGE survey.
FDI, Production, and the Labor Market in SA 205 19. This does not take account of currency depreciation over the period, but South Africa now provides 12 percent of global supply. 20. Wages in decentralized areas of South Africa are generally 25 to 50 percent of urban levels and frequently lower than neighboring states such as Lesotho (Gibbon, 2002: 30). 21. Several of the exporters in the clothing industry are pre-1990 entrants. The LBS/EDGE sample unfortunately does not include any Taiwanese firms due to logistical and linguistic difficulties in organizing interviews. Eight Taiwanese firms are in the sample frame. 22. Previously, only representative offices were allowed. Exchange control relaxation was also necessary. 23. They were ranked between 146 and 331 globally in 2001, by Tier 1 capital. 24. The excess supply of semi-skilled and unskilled labor made questions about their availability redundant. 25. This is based on a rough comparison with the National Enterprise Survey (Gelb, 2001). 26. In the survey, the term “Black” in BEE was defined as people from “historically disadvantaged communities,” that is, those classified as “African,” “coloured,” and “Asian” under the apartheid categories. In the case of ownership, the questionnaire referred simply to “BEE companies.” This is a term whose definition is highly variegated and contested in South Africa, and it seemed appropriate in the survey to allow the respondent to use his/her preferred interpretation of the term, rather than impose a single interpretation. 27. Another 3 percent of firms were in the midst of BEE deals at the time of the survey. 28. The NES results on BEE reported here do not distinguish between pre- and post1990 entrants among foreign firms, and the NES asked only about BEE ownership, not about BEE in skilled labor categories.
References Black, A. (2001) “Globalization and Restructuring in the South African Automotive Industry.” Journal of International Development 13 (6). Dicken, P. (1998) Global Shift. Transforming the World Economy. 3rd edition, London: Paul Chapman Publishing. Edwards, L. (1999) “Trade Liberalisation, Structural Change and Occupational Employment in South Africa.” TIPS Forum (September), www.tips.org.za Estrin, S. and Meyer, K. (eds) (2004) Investment Strategies in Emerging Markets. Cheltenham UK & Northampton MA, USA: Edward Elgar. Gelb, S. (1991) “South Africa’s Economic Crisis: An Introduction and Overview,” in Gelb, S. (ed.), South Africa’s Economic Crisis, Cape Town and Zed Press, London and New jersey: David Philip Publisher. Gelb, S. (2001) “Fixed investment in South Africa.” Report for Office of the President, Government of South Africa. Gelb, S. (2003a) “Foreign Companies in South Africa: Entry, Performance & Impact.” The EDGE Institute, Johannesburg, www.the-edge.org.za Gelb, S. (2003b) “Inequality in South Africa: Nature, Causes and Responses.” The EDGE Institute, Johannesburg, 2003b. www.the-edge.org.za Gelb, S. and Black, A. (2004a) “Foreign Direct Investment in South Africa,” in S. Estrin and K. Meyer (eds), Investment Strategies in Emerging Markets, Cheltenham, UK & Northampton MA USA: Edward Elgar.
206 Stephen Gelb and Anthony Black Gelb, S. and Black, A. (2004b) “South African Case Studies,” in S. Estrin and K. Meyer (eds), Investment Strategies in Emerging Markets, Cheltenham UK & Northampton MA USA: Edward Elgar. Gibbon, P. (2002). “South Africa and the Global Commodity Chain for Clothing: Export Performance and Constraints.” Centre for Development Research, Copenhagen. Government of South Africa, Department of Finance (1996) “Growth, Employment and Redistribution.” Pretoria. Government of South Africa, Department of Trade & Industry (2002) “Accelerating Growth and Development: The Contribution of the Integrated Manufacturing Strategy.” Pretoria. Human Science Research Council (HSRC). (2003) “Nelson Mandela/HSRC Study of HIV/AIDS: South African National HIV Prevalence, Behavioural Risks and Mass Media.” Pretoria. Kaplan, D. (2003) “Manufacturing Performance and Policy in South Africa – A Review.” TIPS Forum (September), www.tips.org.za Lewis, J. (2001) “Policies to Promote Growth and Employment in South Africa.” Informal Discussion Papers on Aspects of the Economy of SA, 16, World Bank. Moodley, S. and Velia, M. (2002) “An Analysis of the Recent Exporting Trajectory of the SA Clothing Value Chain: Upgrading or Downgrading?” University of Natal, Durban. Rankin, N. (2001) “Specialisation, Efficiency and Exports: Some Preliminary Results from South African Manufacturing Firms.” Centre for the Study of African Economies, University of Oxford. South African Reserve Bank (2001) South Africa’s Balance of Payments, 1946–2001. Supplement to Quarterly Bulletin (June). South African Reserve Bank (2003) Quarterly Bulletin (September). South African Reserve Bank. Online Statistical and economic time series. www. reservebank.co.za Standing, G. Sender, J. and Weeks, J. (1996) Restructuring the Labor Market: The South African Challenge. Geneva: International Labor Office. Statistics South Africa. (1999) October Household Survey, Pretoria. Statistics South Africa. (2003) Labour Force Survey. Statistical release P0210 (March). TIPS (Trade & Industry Policy Secretariat). (2003) South Africa HS Export Value Database. www.tips.org.za UNCTAD. World Investment Report. Geneva (various years). World Bank (2003) Global Development Finance 2003: Striving for Stability in Development Finance. Washington DC. I Woolard. An Overview of Poverty and Inequality in South Africa. Working Paper prepared for DFID South Africa, July 2002.
Part III Challenges to the Regulation of International Production
This page intentionally left blank
9 Bargaining Power and Foreign Direct Investment in China: Can 1.3 Billion Consumers Tame the Multinationals? Elissa Braunstein and Gerald Epstein
1
Introduction
Foreign direct investment (FDI) is commonly seen by economists and policy makers as a premier agent, not only of globalization, but also of economic growth and development. In fact, in light of the Asian and South American financial crises, in which portfolio flows proved to be flighty and unreliable, FDI is now treated more than ever as the capital flow of choice. FDI has thus become one of the most sought-after commodities in the global economy, the object of enormous investments of time and resources by policy makers who want to attract it, and the subject of an enormous amount of research and debate concerning its nature and impact. It is not surprising, then, that the role of foreign direct investment in the People’s Republic of China (PRC)1 provides fertile ground for studying the dynamics of FDI and globalization. The PRC has attracted a large amount of FDI over the last decade, the most of any developing country, and in recent years, it was among the top four recipients of FDI in the world. FDI flows to the PRC have occurred in a context of intense global competition for foreign investment by many countries, including developing countries in Asia. And the entry of China into the World Trade Organization (WTO) may dramatically affect the types and quantity of FDI flowing to China, and the government’s ability to manage and direct it. Indeed, many competitors in Southeast Asia and elsewhere worry that the PRC’s entry will lead to an acceleration of investment flows to the PRC and a corresponding reduction in flows to themselves. Hence, China provides more than just a case study of foreign investment; understanding the dynamics of FDI in China is essential to understanding the dynamics of FDI in the world economy as a whole.
209
210 Elissa Braunstein and Gerald Epstein
Globalization and FDI There are two, broad competing visions of globalization in general and the role of FDI in particular (Crotty, Epstein and Kelly, 1998). One is the “neo-liberal vision,” which treats FDI as an agent for spreading capital, technology and management skills across the globe and, therefore, as a crucial agent for economic growth and development. According to this view, any government that wants to partake of these benefits of FDI must implement a set of policies and institutional innovations, including openness to investment, modest regulation, government transparency, modest tax rates, and investor guarantees. The second vision is not as sanguine about the impact of globalization and FDI. Rather than envisioning a “race to the top” or a “convergence,” as forecast by neo-liberal adherents, this group fears a “race to the bottom.” According to this view, global competition for FDI places an enormous amount of bargaining power in the hands of multinational corporations (MNCs) and their allies in the International Monetary Fund (IMF), US Treasury and World Bank. This competition, in turn, forces countries to lower regulations, taxes, environmental protections, wages and working conditions in order to attract and retain capital. The result is a leveling down of wages, social protections, and government control across the globe (Barnet and Cavanaugh, 1994; Greider, 1997). These two views – neo-liberal versus race to the bottom – have become increasingly polarized, fought out in academic circles, policy fora and in the streets during annual meetings of the IMF, WTO, and other global institutions. As for policy, race to the bottom adherents argue for constraints on FDI in order to protect the living standards of workers in the developed economies, and to restrict the alleged exploitation of workers and communities in the developing world. By contrast, proponents of the neo-liberal view, and many academics and policy makers from developing countries, insist that proposed constraints on foreign investment are self-serving: they only raise the living standards of workers in the developed world at the expense of workers in developing countries. Crucial to this policy debate is an empirical question: Does FDI in fact raise the living standards of communities in the developing world? If FDI does make a substantial contribution to growth and development, as the neo-liberal advocates assume, then there may be a trade-off between protecting workers in the North and those in the South. However, if the race to the bottom view is correct and FDI in the current context does not help the majority of citizens in either developing or developed countries, then regulating FDI – for example by imposing codes of conduct or living wage legislation – might be in the common interests of workers in both North and South. As we have argued elsewhere, the impact of FDI will depend greatly on the institutional context in which it takes place (Crotty et al., 1998; Braunstein and Epstein, 1999). As in the race to the bottom view, we see bargaining
Bargaining Power and FDI in China 211
power between MNCs and governments, communities and workers as an essential determinant of the ability of countries and communities to benefit from FDI. In principle, under the right institutional circumstances, countries or regions could exert bargaining power over MNCs, and regulate them in such a way as to generate significant benefits for their communities. The questions are: what are the right circumstances? Do the countries try to do it? and, with what effect? FDI in China If any developing country would seem to have bargaining power relative to MNCs, it would be the PRC. China’s 1.3 billion consumers provide an enormous lure for companies wanting to get a foothold in the Chinese market. And the PRC’s almost limitless supply of cheap labor makes it an enormously attractive location for labor-intensive, export-oriented production. Moreover, the Chinese government has a strong regulatory apparatus in place to manage and control MNCs. Under these circumstances, one would expect that MNCs would be willing to make significant concessions to Chinese communities, workers, and governments. One possible measure of the PRC’s bargaining strength is the oft lamented “fact” that many foreign companies fail to make profits in China, nonetheless remain there in order to be well positioned in the long-run to sell to the Chinese consumer. An important question is whether the PRC government has been able to exploit its presumptive bargaining power to capture a greater share of the benefits from FDI than have countries with less bargaining power. If so, this would strengthen the argument that the free movement of FDI is in the interest of Chinese workers, even if it might not be in the interest of some workers in the United States. However, if China has not been able to use its potential bargaining power to benefit substantially from FDI, that would call into question the desirability of making significant concessions to attract FDI in China. And this caution would apply even more strongly to other countries with much less bargaining power than China, which is to say, virtually every other country in the developing world. There are many important issues associated with this question. Has the Chinese government tried to capture a significant share of the benefits and, if so, how has it tried to do so? For example, has the Chinese government attempted to achieve a high degree of technology transfer from MNCs? A substantial focus on export markets? High tax revenue? An improvement in wages and working conditions for workers? If so, has it been successful in this regard? Has it been more successful than other developing countries? A number of factors have potentially undermined the Chinese government’s ability to exploit its presumptive bargaining power relative to MNCs. Among the most important of these has been the extraordinary de-centralization of China’s investment management institutions. As part of the reforms begun in 1979, China’s central government ceded significant powers to provincial
212 Elissa Braunstein and Gerald Epstein
and local governments to attract and manage foreign investment. At the same time, the central government created significant political and economic incentives for provincial and local officials to try to attract investment, including signaling government officials that if they attracted more investment they would have better career prospects. Other potential factors reducing the benefits of FDI to the PRC’s general population include corruption, which allegedly pervades government– business relations, and the relative absence of independent labor organizations. Taken together, these three factors imply that benefits which could have accrued to the Chinese populace might have been siphoned off by government officials and shared with the MNCs themselves. However much the PRC has been able to exploit its bargaining power in the recent past, its entry into the WTO could dramatically reduce its bargaining power and the distribution of benefits from FDI. Among other effects, WTO regulations may limit China’s ability to engage in industrial policy, potentially compromising the PRC’s capacity to manage FDI in the interests of its economy. In this chapter we will elaborate on these points by investigating issues of the race to the bottom versus neo-liberal convergence in the case of FDI and China. In particular, we will consider the nature of China’s bargaining power in relation to MNCs and FDI, how it has tried to manifest and exploit that power, what leakages have occurred as the government has tried to exploit that bargaining power, and with what impacts on the Chinese population. We will also study how China’s likely entry into the WTO will alter China’s bargaining power with MNCs. Finally, we will discuss the policy implications of our findings. In particular, we will ask whether there is an inevitable conflict of interest between workers in China on the one hand and those in the developed world on the other; or whether, as the race to the bottom approach implies, workers in both the PRC and the North have a common interest in regulating MNCs and FDI. The next section will review the recent history of Chinese policy towards FDI. We emphasize that the central government closely managed the process of foreign investment so that it would focus on exports rather than the domestic Chinese market. At the same time, the decentralized nature of some aspects of the policy made it difficult to manage all components of the FDI process. In this section, we also present some stylized facts about the size and distribution of FDI in China in recent years. In Section 3 we present new empirical results assessing the impact of FDI on employment growth and wage growth. We show that its impact has been positive but rather limited in size. We then consider whether FDI has crowded in or crowded out investment, and we find that, in fact, it has crowded it out. In section 4 we consider the decentralized nature of FDI bidding and consider the impact of FDI on local tax revenue. We find that it is negative, suggesting that, at least at the provincial level, the social benefits of FDI have been dissipated. In the
Bargaining Power and FDI in China 213
penultimate section we look at the impact of China’s entry into the WTO on China’s ability to manage and benefit from FDI. In the final section we consider some larger implications of our findings.
2 Stylized facts of FDI in China: a brief review of Chinese policy towards FDI Introduction The history of China’s policy toward FDI is one of careful experimentation and management in an attempt to use FDI to simultaneously develop an export-led and import substitution strategy. The decentralized nature of Chinese administration has meant that some of these centrifugal tendencies have undermined certain aspects of the regulatory process. Over time, however, as the low-wage export-led strategy has run its limit, and as the Chinese government has broadened its liberalization strategy, China has attempted to attract a broader array of FDI, including joint ventures to serve the Chinese market. At the same time, the sources of investment have evolved from the Chinese diaspora to a broader set of countries, including those in the United States, Europe, and Japan. History At the second session of the Fifth National People’s Congress in July 1979, a joint ventures law was passed, granting foreign investment a legal status in China (Chen, 1996, p. 33). In this initial period FDI was restricted to joint ventures in China’s four special economic zones (SEZs) at the time (three in Guangdong province across the sea from Hong Kong (Shenzhen, Zhuhai – contiguous with Macao), and Shantou, and the fourth, Xiamen in Fujian Province, on the other side of the Straits of Taiwan (World Bank, 1994, p. 221). SEZs offered significant freedoms and advantages for foreign investors, including concessionary tax policies, exemption from export duties and import duties for equipment, instruments, and apparatus for producing export products, and an easing of entry and exit formalities (Chen, 1997b: 8). Pressure from other localities led the State Council in 1984 to extend economic freedoms similar to those of the SEZs to 14 additional “open” coastal cities, and in 1985 to the Yangtze and Pearl River Deltas as well as to a larger proportion of Fujian (World Bank, 1992). Specific encouragement of FDI really began in 1986, with passage of the Wholly Foreign-Owned Enterprise Law, which, in addition to permitting wholly foreign-owned enterprises, also reduced fees for labor and land use, established a limited foreign currency exchange market for joint ventures, and extended the maximum duration of a joint-venture agreement beyond 50 years (Chen, 1996; Huang, 1998). These policy initiatives coincided with a broadening of the reach of China’s Open Door Policy to include the entire coastal zone in 1988, a shift that became known as China’s coastal development
214 Elissa Braunstein and Gerald Epstein
strategy. Open policies for FDI now extended to the entire coastal region, stressing two main goals: (1) to develop labor-intensive industry in the coastal area; (2) to base the production of these industries in labor-intensive export processing of imported raw materials (Chen, 1997d: 12). The next watershed came in 1992, when Deng Xiaoping gave his now famous “Spring Wind” speech endorsing continued market reforms and rapid growth in the context of a post-Tiananmen conservative backlash (Shirk, 1994, p. 39), and the size of FDI flows into China soon accelerated, especially from industrialized countries. It was also at this time that the Chinese domestic market became more open to foreign firms (Cheng and Kwan, 2000, p. 213), certainly a strong incentive for developed source countries trying to get around China’s strict import controls. There was somewhat of a rollback on FDI liberalization in 1994, primarily to cool an overheating economy and discourage FDI in real estate (Cheng and Kwan, 2000, p. 226), but when the economy cooled down liberalization continued. In recent years, FDI policy has also focused on encouraging technologically intensive investment, as authorities have begun treating FDI as a means for acquiring foreign technology versus importing complete sets of advanced equipment (UNCTAD, 2000, p. 26). Since the mid-1990s, China’s policy towards FDI can be at least partly evaluated in terms of its desire to join the World Trade Organization. This desire can help to explain the Chinese Government’s attempt to rollback some of the special privileges for foreign investors. The idea is that as authorities reduce tariffs, they will also reduce preferential treatment for foreigninvested enterprises (FIEs), but preferential income tax treatment is expected to continue (Chen, 1996, p. 43). In that spirit the Chinese government announced the removal of duty-free status on capital goods imports by FIEs to begin in April 1996, a measure that was heavily qualified by grandfather clauses. Partly as a result, FDI fell off in 1996–97, and successful lobbying by FIEs as well as provincial officials eager for FDI is having some effect: it has since been announced that previous exemptions, such as exemption from import duties and value-added taxes on imports of equipment, have been restored (Henley et al., 1999, pp. 240–41; UNCTAD, 1999, p. 57). In sum, China’s policy towards FDI was clearly designed to encourage export-oriented FDI, looking externally to draw on both inputs and markets, and granting well-defined freedoms and incentives to the FIE sector. Policy makers, by developing the coastal development strategy that afforded SEZ-like privileges to the entire coast of China, created a kind of gigantic export processing zone, where free markets were defined not so much by geography but more by ownership (Naughton, 1996, p. 302). It is also important to note that these liberalizing policies were in line with the government’s own import substitution strategies, where FIEs and the limited free market in which they operated were largely separate from the centrally planned and inward-oriented sector (Kueh, 1992). It is only recently as the size of the
Bargaining Power and FDI in China 215
foreign-invested sector has continued to grow and sell to the domestic market that it has begun to exert important influences throughout the wider economy. These will be discussed more thoroughly below. The numbers Figure 9.1 traces the trajectory of nominal FDI inflows into China since 1986. The importance of policy shifts signaled by Deng’s Spring Wind speech in 1992 are clearly visible. Since that time, China has been the largest recipient of FDI in the developing world, with the exception of 2000, when Hong Kong outstripped China by about US$20 billion. This surge for Hong Kong is partly attributable to China, though, as many investors claimed they were “parking” investments in Hong Kong, making ready to enter China once the course of its accession to the WTO got clarified (UNCTAD, 2000, p. 25). In terms of source countries, Table 9.1 lays out accumulated FDI stock by source region between 1983 and 1999. Hong Kong has long been the main source of FDI in China, partly because of its role as an entrepot. It has also been the key thoroughfare for round-tripping, where domestic investment gets funneled through Hong Kong and back into China to take advantage of the privileges afforded to FDI. Estimates of round-tripped investment range between 15 and 25 percent of Hong Kong investment, a substantial amount (Harrold and Lall, 1993, p. 24; Huang, 1998, p. 63; Wei, 2000, p. 245). In recent years Hong Kong investment has lost some of its dominance due to increasing investments by industrialized countries, especially from Western Europe. There has also been a significant upswing in FDI from Latin America; this increase comes from developed country tax havens in the Virgin and Cayman Islands, which together account for over 90 percent of FDI coming from Latin America. Overall, FDI from developing countries
50 45 40 35 30 25 20 15 10 5 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 Figure 9.1
FDI flows into China, 1986–99
Source: China Statistical Yearbook, various years.
216 Elissa Braunstein and Gerald Epstein Table 9.1 Accumulated FDI stock in China by source countries, 1983–99 (1980 US$ million) 1983–91 Source countries
US$ million
%
NIES 9 920 61.75 Hong Kong 9 319 58.01 Taiwan 422 2.62 Singapore 179 1.12 South Korea 0 0 ASEAN 79 0.49 Japan 2 166 13.48 USA 1 817 11.31 Western Europe 1 047 6.51 Other DCs 193 1.20 Other Asia 124 0.77 East Europe 21 0.13 Latin America 17 0.11 Africa 2 0.01 Others 676 4.21 All LDCs 10 840 67.48 All DCs 5 223 32.52 Total 16 063 100
1992–95 US$ million
1996–99 %
45 372 74.12 36 105 58.98 6 003 9.81 2 013 3.29 1 251 2.04 1 175 1.92 4 062 6.64 4 529 7.4 2 686 4.39 708 1.16 1 170 1.91 85 0.14 321 0.52 39 0.06 1 065 1.74 49 227 80.42 11 986 19.58 61 213 100
US$ million
%
1983–99 US$ million
%
57 671 61.71 11 2963 66.17 41 466 44.37 86 890 50.89 6 682 7.15 13 107 7.68 5 934 6.35 8 126 4.76 3 588 3.84 4 839 2.83 1 725 1.85 2 979 1.74 7 833 8.38 14 061 8.24 8 048 8.61 14 394 8.43 8 897 9.52 12 630 7.40 1 349 1.44 2 250 1.32 1 065 1.14 2 359 1.38 117 0.13 223 0.13 5 692 6.09 6 030 3.53 245 0.26 286 0.17 811 0.87 2 552 1.49 67 325 72.04 1 27 292 74.62 26 126 27.96 43 335 25.38 93 452 100 1 70 728 100
Notes: ASEAN includes Thailand, Malaysia, the Philippines and Indonesia. “DCs” are developed countries; “LDCs” are less developed countries. Source: Years 1983–95 from Chen, 1997: Table 2; 1996–99 from authors’ calculations based on China Foreign Economic Statistical Yearbook 2000.
constitutes about 75 percent of FDI stock in China, assuming that Hong Kong can be classified as a developing country. This type of FDI tends to be concentrated in small, labor-intensive export-oriented manufacturing firms in China. Since the early 1990s, when flows from industrialized countries began to increase, so did the share of larger, more capital intensive joint ventures, which are also more likely to target China’s domestic market (World Bank, 1997; Kaiser et al., 1996; Zhu and Lu, 1998; Henley et al., 1999). Indeed, there is some evidence of a spectrum of investment intentions, categorized by nationality of investor: investors from the Chinese diaspora of Hong Kong and Taiwan, and others are at one end, investing primarily in labor-intensive, low wage-exporting firms; firms from Japan investing in intermediate goods products and higher quality products for the Japanese market; and US and European firms primarily investing in firms they hope will sell to the Chinese market. Of course, US firms are well known to hire low wage FIE firms on a contract basis to produce goods for export.
Bargaining Power and FDI in China 217
It is clear from the policy section above that most FDI is concentrated in the coastal provinces, partly as a result of the coastal development strategy and partly because of its proximity and historical linkages with the extended community of overseas Chinese in the NIEs. Between 1986 and 1999, about 90 percent of FDI inflows went to China’s coastal region. But considering these flows as a percent of gross domestic investment qualifies this picture somewhat. Table 9.2 indicates that as a percent of gross domestic investment, FDI is becoming more important as a source of investment for inland regions. Overall, FDI is strongly associated with international trade in China. In 1995, the trade share of FIEs was 39.1 percent of China’s total trade (Chen, 1997a, pp. 3–4). In terms of exports alone, the share of total exports produced by FIEs went from less than one percent of China’s total exports in 1984 to 27.5 percent in 1993 (Lardy, 1994, p. 72). In the past few years, trade by foreign affiliates accounted for as much as half of China’s total trade (UNCTAD, 2000, p. 54). Much of this increasing share is largely a result of the transfer of production of labor-intensive manufactured exports from Hong Kong, Taiwan, and to a lesser extent from other NIEs in East and Southeast Asia, to obtain lower production costs (Naughton, 1996, p. 314), as the timing of China’s FDI reform policies coincided with the upgrading of technology and economic restructuring of China’s higher income neighbors (Chen, 1997a). Timing was certainly not everything; vast networks of overseas Chinese in countries like Hong Kong, Taiwan, Singapore, and Malaysia were poised to take advantage of the confluence in policy and economic change. Despite increases in industrialized country investments into China, which tend to be larger and more capital intensive than investment from developing source countries, FIEs overall tend to be in labor-intensive industries (Chen, 1996, 1997b; Sun, 1998). In an empirical study of inter-industry variation in FDI using data from 1995, Chunlai Chen (1997b) finds a negative and significant relationship between an industry’s capital–labor ratio and FDI (as measured by FIE assets). He also finds that the share of FIEs in an industry tends to be higher the more labor intensive the industry, and that the share of FIEs is also higher in fast-growing export-oriented industries. Using data on China’s 3000 largest FIEs in 1994, which is biased towards larger and more capital intensive enterprises, Chen (1997c) splits industries Table 9.2 FDI as a percent of gross investment by area Year
Coast (%)
Inland (%)
1986 1992 1999
2.93 11.07 16.04
1.06 3.53 6.84
Source: Authors’ calculations based on China Statistical Yearbook, various years.
218 Elissa Braunstein and Gerald Epstein
into labor, capital, or technologically intensive categories, and counts the number of enterprises in each category.2 Even with data biased towards capital-intensive industries, of his sample he finds that 52 percent of FIEs are in labor-intensive industries, 25 percent in capital-intensive and 23 percent in technologically intensive sectors. Among FIEs, Hong Kong is by far dominant across almost all manufacturing industries. In addition to the investment patterns in manufacturing described above, the real estate sector has also been a significant recipient of FDI in China. Between 1984 and 1987, FDI into real estate increased rapidly, attracting more than a third of inward investment flows and peaking in 1986 at 49 percent of all inflows, mostly at the expense of inflows into industry. This proportion declined in 1988 as a result of tighter macroeconomic policies, but was on the rise again beginning in 1992, the beginning of another economic boom (Chen, 1996). In recent years real estate has held steady at around 12 percent of FDI inflows, partly as a result of central efforts to constrain speculative FDI.
3 The impact of FDI in China on growth, productivity, wages, employment, and investment The Chinese government ranks high among the world’s boosters for foreign investment. Along with many Chinese economists, the official line is that FDI has played an enormously important role in the development of the new China. Indeed, many observers believe that as FDI began to falter in the late 1990s, the Chinese government accelerated its efforts to join the WTO primarily to attract more FDI, presumably because of their belief in its importance for Chinese economic development. Yet, despite all the scholarly work devoted to the impact of FDI on China by Chinese economists and others, there is still very little hard evidence that FDI has had a large salutary impact on the Chinese economy, and if so, what exactly it has been. To help fill this gap, we have studied empirically the impact on a number of key macroeconomic variables. Growth and productivity Two of the areas where there has been a fair amount of empirical work has been on the impact of FDI on growth and productivity. Because China has been so tremendously successful in attracting FDI, this literature is an important part of assessing the promise of foreign investment in a developing country context. Beginning with growth, applications of the most standard type with some combination of capital and labor explaining GDP growth, find that FDI makes significant contributions to growth (Chen et al., 1995; Sun, 1998). These types of studies should be treated with caution, however, because FDI could be capturing the contributions of public policy to growth and
Bargaining Power and FDI in China 219
therefore overstating its effects. We deal with this issue by constructing the policy variable “liberalization,” discussed below. Also, these studies do not address the direction of causality: it is just as likely that GDP growth induces FDI as the other way around. In another study, Shan et al., (1999) test whether industrial growth in China is Granger-caused by FDI or vice versa. They find a two-way causality between industrial growth and FDI, that both FDI-led growth and growth drawing FDI are supported by the empirical evidence, indicating that the relationship between FDI and growth is more complex than simple studies would suggest. More complex approaches draw on the insights of endogenous growth theory, emphasizing the indirect effects of FDI on growth through productivity spillovers and forward and backward linkages. These analyses are based on the premise that much of the value of FDI comes in the form of ideas and technological spillovers, as multinationals bring with them managerial know-how, international connections, and technologically advanced production processes, all of which combine to enhance a locale’s overall productivity beyond what FDI directly contributes to domestic investment. Shang-Jin Wei (1996), in an analysis of Chinese city-level data covering 1988–90, finds that holding the growth of inputs constant, a 1 percent increase in the share of foreign-invested firms in city output in 1988 is associated with a 0.32 percent higher growth rate in output, although he does not control for policy. In a different approach that focuses on total factor productivity, Dees (1998) adds a policy variable by controlling for openness, and finds that FDI contributed to Chinese growth only in the 1990s. In a similar analysis, Wei (1995) finds that FDI becomes significant for growth beginning in the late 1980s. Like the standard growth regression studies, these studies also fail to test for the direction of causality, so it could be that multinationals are attracted to high productivity localities. An exception is the work of Gangti Zhu and Ding Lu (1998), who, applying a Granger causality test to a panel data set of 50 Chinese cities during 1985–95, find a causal linkage between the presence of FDI and productivity growth (as measured by value-added per employee), though they do not discuss the opposite causation, from productivity growth to FDI. They also find that spillover efficiencies from FDI are stronger at promoting labor productivity than in boosting total factor productivity, suggesting that the benefits of FDI in China have come from improvements in human resource allocation efficiency versus overall technological progress. Gordon Hanson (2001), in an UNCTAD study of the benefits of FDI for developing countries, is not very confident about the findings of these types of studies. He argues that although the early empirical literature was optimistic about the impact of MNCs on host-country productivity, its findings are open to the sorts of criticisms described above about the direction of causality as well as omitted variables, such as policy. He describes more recent and promising work done on the micro-level, where time series data
220 Elissa Braunstein and Gerald Epstein
of manufacturing plants provides solutions to these empirical problems by showing how the productivity of domestic plants changes over time in response to the presence of MNCs. Haddad and Harrison (1993), using data for Moroccan manufacturing plants in 1985–89, find a weak negative correlation between plant total factor productivity growth and the presence of MNCs in that sector. Aitken and Harrison (1999), using data on Venezuelan manufacturing plants for 1976–89, find productivity growth in domestic plants is negatively correlated with foreign presence in that sector. Hanson concludes that micro-level data undermines empirical support for productivity spillovers from FDI, perhaps indicating that MNCs confine competing domestic firms to less profitable segments of industry. Clearly, for the case of China, more work needs to be done on the micro-level to assess the productivity spillover effects of MNCs. Lacking data at the micro level, we choose not to revisit the already large literature on FDI, growth and productivity. Instead, using the insights gained from this work on growth and productivity, we decided to study more directly the impact that FDI has on workers and communities in China. We have refined our inquiry both in terms of questions asked and in accounting for policy in a more direct way. In terms of the former, in the next sections we take up the issue of FDI’s effects on wages, employment, and domestic investment. In all the empirical work we undertake, we use panel data from the 29 Chinese provinces over the period 1986–99.3 Before discussing the regression analysis, though, we develop a new policy variable that directly measures economic liberalization. Liberalization One of the biggest flaws in the previous literature has been its confounding of two processes occurring in China: one is liberalization and the other is openness to trade and foreign investment. While these have often gone hand in hand, they have not always done so. Liberalization – the freeing up of prices and the liberalization of private ownership of business and control over profits, has also been an important part of the Chinese economic story. It is important then to distinguish these two processes and their impact on China’s economic development. Liberalization need not be equivalent to openness. As any scholar of East Asian development will attest, high levels of trade integration can coexist with strong industrial policies that actively guide development or protect particular economic sectors from the effects of international integration. Empirical studies of China that do control for policy tend to use some measure of openness as a proxy for liberalization and reform. These measures most often reflect to what extent a geographical area has been incorporated into China’s Open Door Policy, ranging from SEZ status to whether a locality is situated in the coastal development zone (Wei, 1996; Dees, 1998; Cheng and Kwan, 1999, 2000). But China has exercised a clear industrial policy
Bargaining Power and FDI in China 221
where trade and FDI are concerned. Its increasing openness to trade and investment has coexisted with strong administrative controls, and so a distinct measure of liberalization is necessary to control for the effects of policy separate from openness. For the PRC, the transition from a centrally planned to a market economy – what we are calling its course of liberalization – has been a gradualist and ongoing one. In the early years of industrial reform, the state created a “dual track system” by maintaining clear delineations between production for the central plan and production in the new and growing sphere of market-based industry. By actually freezing the scope of the traditional mandatory balance plan, the state’s strategy meant that the marketized sector would overwhelm the planned sector over time in terms of economic significance, a process that Barry Naughton has termed “growing out of the plan” (Naughton, 1995). Part of this process entailed breaking up the government monopoly in industry and easing entry by nonstate firms. Although some new firms established at this time were state-owned in the sense of being started up by local governments, most of these types of additions came in the 1980s. The vast majority of new firms were “collectively” owned (an ownership category that in practice exhibits a wide variety of attributes and is thus treated as separate from the traditional state sector), but have become increasingly private or joint-venture in ownership structure as reforms have proceeded. In light of this history, a good way to capture liberalization is to use the ratio of state sector output to all industrial output.4 The data we use for this measure and the regressions below are annual provincial-level data between the years 1986 and 1999; 1986 was chosen as a starting point both because that is when policy towards FDI undergoes a marked shift, and because of data availability. Where possible, the data we use comes from the China Statistical Yearbook published by the PRC’s State Statistical Bureau, as it is regarded as more reliable than the alternative provincial statistical yearbooks that we used as a backup. There are a total of 29 provinces, listed in Table 9.3 below.5 Table 9.3 shows the liberalization measure (L2, or state output as a percent of all industrial output) for the beginning and ending of the time period in question, 1986 and 1999, and the percentage change between those two periods. Provinces are rank-ordered, from smallest absolute percentage change in L2 to largest. In order to get a sense of whether liberalization captures policy variables that are distinct from those that prescribe openness, we have also included average FDI as a percent of total investment (average FDI/I), as well as average trade as a percent of GDP (average T/GDP) over the same time period. If openness is equivalent to liberalization for the Chinese case, one would expect the liberalization variable to simply mirror FDI and trade. But this is clearly not the case. The correlation coefficients between L2 and average FDI/I and T/GDP are, respectively, 0.396 and 0.299; both are significant at the 95 percent level. While there is a significant relationship
222 Elissa Braunstein and Gerald Epstein Table 9.3 Measures of liberalization and openness Province
L2, 1986 L2, 1999 (%) (%)
Xinjiang Qinghai Ningxia Beijing Jilin Shaanxi Yunnan Guizhou Heilongjiang Hainan Gansu Shanghai Jiangxi Jiangsu Inner Mongolia Sichuan Guangxi Shanxi Shandong Liaoning Henan Tianjin Hebei Anhui Hubei Guangdong Hunan Zhejiang Fujian Mean Standard deviation
82.9 81.2 78.5 77.8 73.0 72.8 77.7 76.0 80.4 81.7 86.5 77.1 70.6 37.0 79.6 68.8 75.7 64.3 52.9 66.1 61.3 67.0 56.0 63.7 67.4 53.5 67.4 35.0 53.5 68.5 12.7
76.2 72.8 67.0 65.7 57.9 55.4 55.9 53.3 53.9 52.0 50.3 44.3 40.1 20.1 42.4 34.0 36.9 31.2 25.5 31.5 28.4 29.7 23.7 25.3 25.7 19.4 22.3 11.0 15.1 40.2 17.8
Change in L2 Average FDI/I 1986–99 (%) 1986–99 (%) 8.9 11.5 17.3 18.4 26.2 31.5 39.2 42.7 49.2 57.1 72.1 74.3 75.9 83.7 87.8 102.0 105.3 106.5 107.7 110.2 115.8 125.4 136.1 151.7 162.0 176.1 202.3 218.1 255.0 95.5 63.3
0.7 0.16 0.68 10.3 4.6 5.2 1.2 1.6 3.8 27.7 1.3 11.2 4.8 12.1 1.21 2.8 8.8 2.1 8.4 9.5 2.5 17.0 3.4 3.3 4.3 26.7 4.6 4.3 28.6 7.3 8.0
Average T/GDP 1986–99 (%) 10.0 6.0 8.6 19.9 16.0 9.9 9.3 5.1 11.5 43.8 5.5 53.4 8.3 23.1 8.9 6.5 12.9 7.7 16.1 26.6 5.2 42.1 9.8 7.3 8.8 88.1 8.4 18.5 39.5 18.5 18.3
Note: L2 is state sector output divided by all industrial output. Source: Authors’ calculations.
between these variables, there is also considerable independence. As such, we elected to use L2 as a policy variable that measures the impact of liberalization, independent of the impact of openness, in the regressions below. The impact of FDI on wages Two ways in which FDI can directly help workers is by raising wages and employment. In this section we analyze the impact on wages and in the next section we consider employment.
Bargaining Power and FDI in China 223
While wages are often used as an independent variable to explain FDI, it is rare to find the causality running the other way in the empirical literature on FDI in developing countries.6 But there is a clear causal link. First, FDI may affect labor demand (depending on whether it is greenfield investment or mergers and acquisitions, and on what competitive impact it has on domestic investment), thereby affecting wages. Second, spillover effects from potentially higher productivity (and paying) foreign enterprises could raise wages throughout the country. And last, because capital is internationally mobile and labor is not, FDI may enhance capital’s bargaining power relative to labor, thereby lowering wages (Paus and Robinson, 1998). Turning to the Chinese case, the core model we used is based on the notion that short-term (annual) changes in wages depend on labor demand and supply. Data are panel data for China’s 29 provinces between 1986 and 1999 (please see the data appendix for a fuller explanation of the data used).7 In regression equation (1) below the average provincial wage, measured as the average annual wage for a particular province, is a function of: total investment (I); foreign direct investment (FDI); total foreign trade (imports exports T); the total available labor force (LF, defined as the population fifteen and over); productivity, where GDP/employment q; and finally the liberalization variable, L2, the ratio of state sector output to all industrial output. Provincial fixed effects are ’s, a time trend has been added to control for uniform shocks, and is a serially uncorrelated random error. We use two measures of investment: gross investment (I) and “adjusted investment” (adjI), which subtracts FDI from gross investment to get a clearer sense of the effects of FDI on wages. The regression is run in logs to get elasticities, and first differences were used to address non-stationary in the variables.8 The results are detailed in Table 9.4. In wit i 1 lnIit 2 lnFDIit 3 lnTit 4 lnLFit 5 lnqit 6 lnLit2 timetrend εit Table 9.4 above presents a series of regressions that build up to the full model in regressions V and VA. Regressions I–V use the gross investment measure; regressions IA–VA use the adjusted investment measure. The potential for simultaneity – FDI, investment or trade could just as reasonably be argued to be the result of wage changes as the source of them – induced us to do some Granger causality testing. The results, which indicated mutual causation between these variables and wages, were not really satisfactory owing to the short time series and limited degrees of freedom. But the exercise did prompt us to use some lagged values on the right-hand side to partially address this problem. The regression results indicate that FDI has a statistically significant and positive impact on wages that is not significantly affected by the model’s other variables. Looking at the full model with gross investment (regression V),
224 Table 9.4 Wage equations for 29 provinces, using data for 1986–99 (estimation with first differences and fixed effects; dependent variable: in average wage) Variable InIt1
I
IA
0.255* (12.27)
II 0.225* (10.24)
0.224* (10.18)
InadjIt1
IIA
0.019*
InFDIt1
0.320** (2.10)
0.410** (2.12)
0.331** (2.22)
IIIA
0.212* (8.96) 0.198* (8.97) .025*
0.356** (2.32)
InTt1 InLFt1
III
IV
IVA
0.198* (7.73)
0.020*
0.019*
0.168* (6.90) 0.024*
0.021*** (1.37) 0.334** (2.25)
0.380* (2.47) 0.361* (2.38)
0.020 (1.17) 0.318** (2.06)
0.035** (2.10) 0.339** (2.16)
.091* (2.87) 0.371 306
Inqt1
R2
0.338
.263
0.379
0.339
0.386
0.356
0.073** (2.31) 0.394
N
346
346
335
335
334
334
306
T-statistics in parentheses. * Significant at the 99 percent level. ** Significant at the 95 percent level. *** Significant at the 90 percent level.
VA
0.193* (7.10)
0.181* (7.94) 0.025*
InL2
V
0.019*
0.160* (6.29) 0.024*
0.021 0.036** (1.22) (2.18) 0.327** 0.350** (2.11) (2.22) 0.047 0.075 (0.79) (1.24) 0.070** 0.086* (2.20) (2.70) 0.400 0.376 304
304
Bargaining Power and FDI in China 225
a 1 percent increase in FDI increases wages by about 0.02 percent. Put another way, if FDI increases by one standard deviation from its mean, wages will increase by 1.32 percent. In regression VA, the coefficient increases slightly so that a 1 percent increase in FDI increases wages by about .025 percent. The actual effect probably lies somewhere between these two figures, as some small proportion of FDI is comprised of financial flows that are not counted as a part of gross investment. Gross investment packs a lot more punch: in regression V a 1 percent increase in investment raises wages by 0.19 percent – an investment increase of one standard deviation from the mean raises wages by about 22 percent. Looking at the adjusted investment measure in VA lowers this estimate: a 1 percent increase in domestic investment raises wages by 0.16 percent. But it is still a much larger impact than FDI. Looking at the rest of the variables, we will confine the discussion to the regressions that use adjusted investment, keeping in mind that the true results are somewhere in between the regressions that use the gross and adjusted investment measures. Trade has increased the overall demand for labor in China, thereby raising wages: in regression VA, a 1 percent increase in trade raises wages by .036 percent. Put in terms of standard deviations, an increase of one standard deviation raises wages by 2.9 percent. The labor force variable, reflecting a reserve army effect where greater supplies of labor drive down wages, has the expected effect, and it is notably sizeable, suggesting that creating ample employment for China’s growing population is a validly key concern for policy makers. Productivity has the correct sign but is not statistically significant, probably due to noise in the measure. The results on liberalization are interesting for a couple of reasons. First, including liberalization does not affect the other variables that one might posit are picking up liberalization effects: neither FDI nor trade, sometimes treated as proxies for liberalization, are altered significantly by the inclusion of the liberalization variable. At least for the Chinese case, then, these openness measures are not picking up the effects of liberalization on the Chinese economy. Second, there is a consistently positive and statistically significant relationship between liberalization and wages. If the proportion of state sector output to all industrial output decreases by 1 percent, wages will increase by 0.086 percent; if L2 declines by one standard deviation from the mean, wages will increase by 3.06 percent. This result suggests that the freedom from wage controls, and the overall increase in incomes (especially rural) that came with liberalization, were important positive benefits of the economic reform program begun in 1979. But it is important to recall that we are only measuring money wages. The substantial non-cash benefits that can come along with a state sector job, such as housing allowances, still render these jobs among the most prized, at least for less educated or lower-skilled workers who would otherwise have trouble accessing workplace benefits. Some interesting work on the impact
226 Elissa Braunstein and Gerald Epstein
of FDI on relative wages in China illustrates this point. In a study that compares state and FIE sector wages for unskilled workers, it was found that after including non-wage benefits like pensions, housing, and medical care, state sector jobs were much better for unskilled workers who had been assigned or expected to receive public housing (Zhao and Xu, 2000).9 The same estimate was not done for the college-educated, but it was hypothesized that wages between the state and foreign sectors for skilled workers would be much smaller since FIEs offer good benefit packages to the college-educated. The work of Yaohui Zhao (2000) bears on this issue. He argues that because of the segmented nature of labor markets in China, FDI does indeed raise the relative wages of skilled workers, but not because FIEs demand more skilled workers. Using urban household survey data in 1996, Zhao finds that education is used to access the privileged state sector. Unskilled workers are thus more abundant in the unprivileged or informal sectors, giving FIEs easy access to unskilled workers. Conversely, FIEs must compete with the state sector for skilled workers, bidding up the relative wages of skilled workers in the foreign sector. He also finds that the proportion of skilled to unskilled workers is similar in the state and foreign sectors, suggesting that FIEs do not employ more skilled labor than state firms, casting doubt on the notion that relative wage increases are a result of FIEs enjoying higher productivity. While FIEs do pay higher wages (including bonuses and cash subsidies) than state firms, these higher wages are unevenly distributed among workers of different educational levels. Including noncash benefits such as housing, less educated workers earn less in FIEs than in state enterprises, but more educated workers earn more than the nominal numbers. In general, our results with imperfect deflators were similar to the nominal results reported here. Summary In sum, then, we do find empirical evidence that over the last decade and a half, FDI has raised wages in China, but the impact is small in relation to the effects of domestic investment and to a lesser extent liberalization. The impact of foreign trade on wages is also significant, both statistically and economically. In fact, this latter variable might be picking up some of the impact of FDI on wages, since, as we discussed above, a great deal of FDI was directed toward the export sector. Even if one adds up the effects of these two variables, implicitly attributing all the export effect to FDI, the sum of the coefficients on FDI and trade is about 0.06, which is about 1/3 of the impact of investment on wages (0.16). Of course, this is an overestimate of the FDI effect because not all trade is due to FDI and foreign invested enterprises (though a significant amount is, perhaps as much as 60 percent). Employment One of the key reasons that local officials are so interested in attracting FDI in China is employment creation. With economic reforms, a growing
Bargaining Power and FDI in China 227
population, and the rising incidence of layoffs in the state sector, there is tremendous pressure on local authorities to stave off the social unrest that will result from an increasingly serious unemployment problem, particularly in urban areas. While official urban unemployment rates stand at about 5 percent, scholars estimate that the true unemployment rate could be two to five times of that (Jiang, 2002; Minqi, 2001).10 Whether FDI has actually contributed to employment creation in China is an open question, and so we take up this issue empirically here. The regression model is presented in equation (2), and, as in the section above, data are panel data for China’s 29 provinces between 1986 and 1999. Employment, measured as the total number of remunerated jobs (including the selfemployed) at year-end in a particular province, is a function of: per capita output, GDP/pop; investment, I (here again we use two measures of investment, gross and adjusted investment, which subtracts FDI from gross investment); foreign direct investment, FDI; total foreign trade (imports exports T ); and liberalization, L2, the ratio of state sector output to all industrial output. As in the wage regressions, provincial fixed effects are ’s, a time trend has been added to control for uniform shocks, and is a serially uncorrelated random error. The regression is run in logs to get elasticities, and first differences were used to address non-stationarity in the variables. The results are detailed in Table 9.5 below. Inempit i ln(GDP/pop)it lnIit lnFDIit lnTit lnLit2 + timetrend + it
(2)
Based on these regression results, whether one uses gross or adjusted investment, FDI has no independent effect on employment. And even though using the adjusted investment measure does improve FDI’s significance somewhat, the potential impact of FDI on employment is nonetheless very small. The force of gross and adjusted investment declines across regressions I–VA, and once again, we believe that the actual coefficient lies somewhere in between the two estimates of investment. Looking at the full model in VA, a 1 percent increase in adjusted investment will increase employment .016 percent, meaning that an increase of one standard deviation from the mean raises employment by 1.7 percent. A 1 percent increase in trade has a similar impact on employment as adjusted investment, about .017 percent, but an increase in trade of one standard deviation from its mean increases employment by less than investment: 1.4 percent. Still, this finding is a very significant one: trade has had a significant and positive impact on employment in China, much more so than FDI. The results on the liberalization variable are consistent with the notion that all else equal, nonstate enterprises have generated more employment than state enterprises since reforms began. According to the regression results in VA, a 1 percent (standard deviation) increase in state output as a
228 Table 9.5 Employment equations for 29 provinces, using data for 1986–99 (estimation with first differences and fixed effects; dependent variable: in employment) Variable InGDP/pop
I
II
IIA
III
0.033** (1.84)
0.003 (0.11) 0.025** (2.12)
0.012 (0.54)
0.000 (0.01) 0.023** (1.82)
InI InadjI
0.019** (1.69)
InFDI
IIIA 0.006 (0.27)
0.002 (0.91)
0.018*** (1.54) 0.003 (1.21)
InT
IV
IVA
0.000 (0.01) 0.018*** (1.36)
0.004 (0.16)
0.002 (0.94) 0.018** (2.16)
0.015 (1.18) 0.003 (1.17) 0.019** (2.36)
InL2 R2
0.165
0.166
0.162
0.173
0.171
0.194
0.193
N
403
373
373
361
361
332
332
T-statistics in parentheses. * Significant at the 99 percent level. ** Significant at the 95 percent level. *** Significant at the 90 percent level.
V
VA
0.020 0.020 (0.79) 0.78 0.017*** (1.29) 0.016*** (1.34) 0.002 0.003 (0.97) (1.16) 0.017** 0.017** (2.03) (2.16) 0.048* 0.050* (3.04) (3.15) 0.218 0.219 332
332
Bargaining Power and FDI in China 229
percent of all industrial output will decrease employment by 0.05 percent (1.8 percent). This result reflects two things: that state firms tend to be more capitalintensive than other forms of ownership that have come about as a result of market reforms, and that state firms have been shedding workers in recent years in response to the widespread impression that these firms are overstaffed (the state sector shed nearly one-quarter of its workforce between 1995 and 1999 (Zhao, 2001, p. 2)). An important caveat must be noted before this result can be taken as a basis for making predictions about liberalization and the future of employment. First, that liberalization has been associated with job creation says nothing about the relative quality of jobs. Although there is clear evidence that liberalization is associated with higher wages (as is FDI), there is also evidence that other measures of job quality such as security and benefits are better in the state sector, at least for less educated workers (Zhao, 2000, 2001). That said, there is no evidence that FDI has been a good source of employment creation, except, perhaps in so far as it has led to exports. The role of FDI in exports The importance of FDI in employment generation through exports, however, is worthy of further scrutiny. It is true that the share of FIE’s in exports has been growing rapidly: their exports increased from less than 2 percent of total Chinese exports in 1986 to 48 percent in 2000. However, it is important to note that their share of imports rose during the same period from 6 percent to 52 percent (UNCTAD, 2002, p. 155). As we noted above, since FIE’s are more capital intensive than local firms, their role in job creation is modest. For example, in 1996, these firms employed only 5.4 million workers, or about 0.8 percent of the total labor force (ibid). Thus, taking into account the role of exports in employment creation is not likely to modify our overall conclusion that FDI has not generated as much employment as other activities. When combined with the analysis of its impact on investment in the next section, though, the overall employment generating effect of FDI is placed in further doubt. Crowding in/crowding out In both the wage and employment regressions, FDI is included on its own as well as part of gross investment in order to separate the effects of FDI as a distinctive form of investment – one that may tend to pay higher wages or create more (or less) employment than other forms. But in order to fully appreciate its role, the relationship between FDI and domestic investment must be investigated, as it is not necessarily the case that FDI merely adds to labor demand as a part of gross investment. FDI might encourage or crowd in domestic investment, as when there are strong backward or forward linkages created by new foreign firms. Or, FDI could crowd out domestic investment, as when foreign firms compete with domestic firms and drive them out of business.
230 Elissa Braunstein and Gerald Epstein
Using panel data for the period 1970–96 in three developing regions, Asia, Africa, and Latin America, Manuel Agosin and Ricardo Mayer (2000) did an econometric study of whether foreign investment crowds in domestic investment. Their results indicate that in Asia, and to a lesser extent in Africa, there has been strong crowding in of domestic investment by FDI. In Latin America, FDI has had a strong crowding out effect. China is one of the countries in their study, and they find that FDI has had a “neutral” effect on domestic investment; that is, it has neither crowded in nor crowded out domestic investment. One problem with this study for China is the time period under consideration. China did not allow FDI until 1979, and even then policy restrictions were strict until the mid-1980s, so the China results deserve closer analysis. In terms of this type of work on China, Haishun Sun (1998) does a simple regression analysis of the determinants of domestic investment in ten coastal provinces between 1983 and 1995, when about 90 percent of FDI targeted China’s coastal region. Using income per capita (as a proxy for domestic savings), FDI, and other forms of foreign capital, he finds a strong significant positive correlation between FDI and domestic investment. Such a simple approach is unconvincing, however, as the problem of omitted variables (such as measures of policy) throws the causal link between FDI and domestic investment into serious question. Among qualitative studies of the relationship between FDI and domestic investment in China, the sentiment is less sanguine than Sun’s. Yasheng Huang (1998) argues that FDI probably crowds out domestic investment because FIEs tend to be highly leveraged and compete with local firms for domestic bank financing. This leveraging is a direct result of policy incentives that grant preferred status to FIEs – Chinese partners are more motivated to qualify for FIE status than to ensure adequate financial contributions from foreign partners. Thus the equity contribution by foreign investors often falls short of what is specified in the contract, and Chinese partners end up borrowing to cover the shortfall (Huang, 1998). And Nicholas Lardy, in a classic analysis of China’s economic reform process, claims that rather than financing increased levels of investment, China’s substantial capital inflows have been used for three other purposes: (1) to increase the foreign exchange holdings of China’s Central Bank, which in the three years between 1994 and 1996 increased by US$83.3 billion, and in 1997 by US$35 billion; (2) to provide funds for capital flight;11 and (3) to a much lesser extent, to finance investment abroad by Chinese firms (Lardy, 1998: 191–92). Lardy concludes that, despite China’s tremendous success in attracting FDI, it cannot be counted on to fulfill China’s substantial investment financing needs. We decided to test these conclusions by rerunning Agosin and Mayer’s (2000) econometric study on China, using our province-level panel data set
Bargaining Power and FDI in China 231
for the years 1986–99. The investment equation is as follows: Iit i 1FDIit 2FDIi,t 1 3FDIi,t 2 4Ii,t 1 5Ii,t 2 6Gi,t 1 7Gi,t 2 YD it
(3)
where: I gross investment/GDP; FDI FDI/GDP; G GDP growth; YD year dummies; ’s are provincial fixed effects; and is a serially uncorrelated random error. Since we are interested in the long-term effects of FDI on investment, the formula for the relevant coefficient (beta hat long-term): 3
ˆ LT
兺 ˆ
j1
1
j
5
兺
(4) ˆ j
j4
LT 1, there is no crowding in or out: an increase If we cannot reject that ˆ LT1, there in FDI raises total investment exactly by the same amount. If ˆ is evidence for crowding in. Likewise, if ˆ LT1, there is evidence for crowding out. Table 9.6 below lists the results. Table 9.6 Investment equations for 29 provinces, using data for 1986–99 (estimation with fixed effects; dependent variable: gross investment, I) Variable FDI FDIt 1 FDIt 2 It 1 It 2 Gt 1 Gt 2 Adjusted R-square Observations
ˆ LT
T-statistics in parentheses. * Significant at the 99 percent level. ** Significant at the 95 percent level.
0.523* (3.69) 0.212 (1.14) 0.234** (1.67) 0.878* (15.70) 0.294* (5.24) 0.109* (4.03) 0.003 (0.12) 0.771 343 0.185 (0.30)
232 Elissa Braunstein and Gerald Epstein
Using a bootstrap method12 to assess the statistical significance of ˆ LT , we found that we could not reject the hypothesis that it was statistically equivLT is (0.445, 0.696), alent to zero; the 95 percent confidence interval for of ˆ and therefore does not include 1. We therefore conclude that there is evidence that FDI had a negative impact on Chinese domestic investment. There is strong evidence that FDI crowded out domestic investment during the late 1980s and 1990s in China, a finding consistent with that of Huang (1998) and Lardy (1998). This is a very significant conclusion for assessing the overall promise of FDI to positively affect China’s economy. Although it has contributed to wage increases its tendency to crowd out domestic investment confirms that its role in employment creation has not been a positive one.
4
Bidding for investment
Bidding for investment is certainly not unique to China. For a number of years, economists and policy makers have studied and even decried the socalled “War Between the States,” a popular name given to the US inter-state competition for investment and jobs (Federal Reserve Bank of Minneapolis, 1994; Burke and Epstein, 2001). Moreover, an enormous amount has been written about tax competition in the European Union as well as internationally among nations. But the level to which this “War Between the Provinces” has risen in China is, perhaps, surprising, even to those well schooled in the many paths of capital mobility. This point was brought home to us rather starkly when one of us interviewed a Chinese official in a Northeastern coastal city, Dalian, about foreign investment.13 This official is the associate director of one of Dalian’s special economic zones, in this case, a high tech export processing zone. We asked him, “Who is your greatest competitor when it comes to trying to attract foreign investment?” expecting the answer to be Vietnam, or Malaysia or, perhaps, Beijing. But his answer startled us: “Our biggest competitor is the export processing zone down the street.”14 Not only does one province or one town compete with another; but in China, there are numerous zones – export processing zones, high tech zones, industrial zones – all of which compete for foreign investment. The result is cut throat competition. The nature of this competition takes a number of forms even while the central government tries to limit and regulate it, primarily in order to preserve its tax revenue and also to promote its industrial policies. The central government determines the tax rates that foreign invested enterprises must pay. The various special zones established with the central government’s permission at the local level offer preferential tax rates, especially for desired types of investments, but these rates are set by the central government.15 Despite this central government control, holding constant the amount of investment, government tax revenue will fall as there is more foreign investment
Bargaining Power and FDI in China 233
getting preferential rates. While much of this revenue goes to the central government, there is a complex revenue sharing mechanism which sends some of it back to the provinces (Wong, 2000). Presumably, provincial governments therefore receive a diminished amount of tax revenue to the extent that the amount going to the central government falls (Wong, 2000). But there are other, indirect and potentially costly ways in which this interzone, inter-city and inter-provincial competition occurs out of the reach of the central government’s control. According to an official whose job it is to attract foreign investment to a city in Southeastern China, “Under the general policies, the local and municipal governments can do some [things to attract companies]. For example, the local and municipal governments can reduce or return the taxes enjoyed by local governments to the enterprises which [the] city [is] eager to attract, including the income tax and value added tax [which] belong to local and municipal governments (25 percent of the total VAT).”16 This potential decline in VAT revenue can be significant. In addition, perhaps the most widely used incentive under the control of local and municipal governments is the provision of free or highly subsidized land, subsidies for electricity and other utilities, and sometimes, the building of roads and other infrastructure projects supporting the factory sites. According to this same official, “Subsidies for the land are the most used policies by municipal and local governments. Nearly for all projects, governments give the subsidies for the land. The selling price of land is lower than the development cost. And for the local government encouraged projects, the price is cheap, [sometimes close] to zero.” The official adds that “Local governments [have] less control over [the] price of water and energy. But they can reduce or exempt some fees charged by governments.”17 Secondary literature confirms this interview information. The so-called “five connections and one leveling” (wutong yipping) – connecting roads, telecommunications, water, electricity and ports and leveling of sites – were the main methods used to attract foreign investment to Shenzen, an SEZ near Hong Kong (Yeh, 2000, p. 52). These methods have been used in many other parts of the country as well. Once again, these can cost the government significant amounts of money. They can also lead to enormous waste. Literally hundreds of economic and technical development zones (ETDZs) have been set up around the country to attract foreign investment. Many are established in small towns and villages. Evidently, while some are successful, “most are left idle because the flow of foreign investment does not materialize, leading to much waste of valuable land resources” (Yeh, 2000, p. 56). The other ways in which competition occurs is by promising good political connections so that companies will be able to cut through the various types of red tape that the government throws at them. It is not uncommon for foreign investors to be wined and dined by top government officials or their children with the implicit promise that these connections, or guanxi, will help pave the way to an easier corporate life.18
234 Elissa Braunstein and Gerald Epstein
Why are these government officials so anxious to attract FDI? One reason must surely be the fact that they believe it has a positive impact on their communities. But another is that local officials who attract high levels of FDI have a much greater possibility of being promoted in the government and/or party.19 Furthermore, and not to be underestimated for its appeal, is that attracting foreign investment with an offer to give large subsidies and streamlined regulatory treatment create enormous opportunities for graft and corruption (see more on this below). One result of this competition may be an erosion of provincial government revenue, at least in the short run. As we have seen, this loss can result from lower tax rates officially offered to investors investing in an economic zone, as well as the loss of fees and other income from the land and other assets. To assess the impact of FDI on tax revenues, we returned to our panel and studied the evolution of provincial government revenue using regression methods. Table 9.7 estimates the impact of domestic (adjusted) investment, foreign direct investment, trade and liberalization on the provincial government revenue. All variables are in logs and are measured in first differences. The Table 9.7 Government revenue equations for 29 provinces, using data for 1986–99 (estimation with first differences and fixed effects; dependent variable: log of provincial tax revenue) Variable In GDP
I
II
III
V
VI
VII
0.069 (0.57)
In adjI
0.071 (0.47)
0.166 1.06
0.026 (0.16)
0.01 (0.06)
0.066 (0.35)
0.209* (2.73) 0.036* (2.75) 0.171* (3.42)
.209* (2.49) 0.035* (2.49) 0.175* (3.24) 0.091 (0.88)
0.117*** (1.64) 0.175** 0.167** 0.210* (2.34) (2.18) (2.76) 0.036* 0.036* (2.75) (2.76) 0.172* (3.44)
In adjIt 1 In FDIt 1 In Tt 1 In L2 In L2t 1 R2 N
VIII
0.013 (0.09)
0.114 (0.99)
In GDPt 1
IV
0.006 396
0.004 396
0.011 366
T-statistics in parentheses. * Significant at the 99 percent level. ** Significant at the 95 percent level. *** Significant at the 90 percent level.
0.020 338
0.045 327
0.082 326
0.028 (0.29) 0.082 326
0.082 326
Bargaining Power and FDI in China 235
regressions show that government revenue is negatively associated with both domestic and foreign direct investment. On the other hand, trade is positively associated with government revenue. These results are certainly consistent with the argument that bidding for investment – both foreign and domestic – appears to reduce provincial government revenue. This seems to be a significant cost of the decentralized nature of the investment bidding process. As with the other regression results we have presented, however, this conclusion is tempered to some extent by the positive impact of trade. To the extent that FDI contributes to trade, then, indirectly it might be contributing to tax revenue. But, these results suggest that in order to measure the positive impact of trade, one might need to subtract the bidding costs associated with attracting investment. It also suggests that as foreign investment becomes less and less export oriented, as we discuss below, these trade related gains are likely to be significantly eroded. Corruption We suggested in the introduction that one way in which the possible social gains from FDI can get dissipated is through graft and corruption. It is certainly the case that most Chinese people believe that corruption is a very large problem. Moreover, there have been some very large and well publicized corruption scandals in China in recent years, including one in Xiamen, Fujian province in which dozens of local officials were convicted of smuggling, and more than ten were given the death penalty. But, as far as we know, there have been no publicized arrests of officials for corruption in the bidding process for foreign investment. However, there is little doubt that such corruption is widespread. Daniel H. Rosen interviewed almost 100 expatriate managers of MNCs in China in the late 1990s. Rosen reports that many managers complained of bribery and bribery related payments to achieve approval of investments or win contracts (Rosen, 1999, pp. 218–26). He suggests, moreover, that having an effective and well connected patron in China, to help firms navigate around the maze of regulations, was perhaps the most important determinant of business success (ibid.). Shang Jin-Wei, (2000), in his provocatively titled paper, “Why Does China Attract So Little Foreign Investment,” claims that given China’s income and population, the amount of foreign investment it attracts falls well below the cross-country regression line which would predict the quantity that “ought” to flow there. He argues the difference is even greater if one considers that a significant amount of investment is not really foreign: it is Chinese investment, round-tripped through Hong-Kong (see the discussion above). He attributes the low level of foreign investment to corruption and other government induced barriers to foreign investment. Either way, it does seem that corruption may well be a costly institution which is reducing the benefits of foreign investment for the Chinese
236 Elissa Braunstein and Gerald Epstein
population. To the extent that such corruption might be an inevitable concomitant of industrial policy in China, these possible costs must be weighed against the benefits that a structure of industrial policy and allocation, performance requirements, and import substitution provides.20 And this becomes particularly important in light of the constraints which China’s entry into the WTO will place on its ability to conduct industrial policy.
5
China’s entry into the WTO
In November 1999, the Chinese and US governments agreed to accession terms by which China could join the WTO. This agreement set the tone for agreements that China made with other countries over the ensuing year and a half. In the fall of this year, after further intense negotiations, China finally was admitted into the WTO. The basic terms of the agreement are as follows (Lardy, 2002; U.S. Trade Representative). ●
●
● ●
●
●
●
●
●
Agricultural tariffs fall to an average of 17.5 percent by 2004, liberalization of imports of major agricultural commodities, import, and distribution rights granted to foreigners; Average tariff on industrial products falls to 9.4 percent by 2005, including elimination of all tariffs on high-technology products; auto tariffs fall from 80–100 percent to 25 percent by 2006; Eliminates import quotas and licensing requirements by 2005; Grants import and distribution rights to foreign corporations, allows them to set up wholly-owned distribution, sales (including retail), shipping, and service networks over a three-year phase-in period; Financial services – banking, insurance, and securities – increased access phased in over five years, culminating in full market access in all activities and regions, and national treatment for foreign banks; minority ownership in domestic securities firms and most insurance business; Telecom – ends ban on foreign investment, allows 50 percent ownership in value added (internet) and paging services within two years after accession; 49 percent ownership in mobile telecom, domestic and international services phased in over five to six years; Other services – increased market access for professional services, including accounting, consulting, engineering, medical, and information technology; Commits China to implement and enforce international standards on protection of intellectual property; provides for increased access and distribution rights for motion pictures, music, and software; United States agrees to extend China Permanent Normal Trade Relations (PNTR) as a WTO member, phase out quotas on imports of Chinese textiles and apparel by 2004 (all other sectors already are fully open as a part of annual extension of NTR).
Bargaining Power and FDI in China 237
Safeguards and protections for the United States ●
●
●
Allows the United States to continue to treat China as a non-market economy in anti-dumping cases for 15 years after accession; this methodology generally results in the application of larger anti-dumping margins against Chinese imports; Permits the United States to implement a product-specific safeguard to prevent large import surges from China; this safeguard, which allows the United States to impose restrictions on imports from China more easily than imports from other WTO members, would remain in effect for 12 years after China’s WTO accession; Allows the continuation of a safeguard mechanism to prevent textile import surges until the end of 2008.
As these last points make clear, the Chinese government made many concessions to join the WTO, and indeed, made more concessions than virtually all other countries that joined. (Lardy, 2002). Why were the Chinese leaders so anxious to join? Many have argued that the principle reason was because they believed that joining would accelerate the flow of foreign investment into China and that foreign investment has been the engine of growth. This certainly seems to be true for some of the policy makers. But, another view is more compelling. As an official of the Ministry of Foreign Trade and Cooperation (MOFTEC), perhaps the most active supporter of WTO admission in the government, pointed out in an interview, the most important reason to join the WTO is to keep the process of liberalization moving forward.21 China is entering a difficult period as it tries to move further down the capitalist road. As state-owned enterprises continue to lay off workers, social unrest and political pressure to reverse the liberalization process is certainly bound to increase. Having signed an international agreement to liberalize further, it will be difficult for these political forces to block further liberalization. Foreign investment is an important aspect of this process, but it is not the main goal of joining. WTO, FDI and employment generation As we indicated above, one of the motivations for joining the WTO was as a way to attract more foreign direct investment. Part of the motivation was to provide employment to help absorb employees laid-off as a result of the increased competition State Owned Enterprises (SOEs) will face from foreign competition. However, as we have shown, the role of FDI in employment generation has not been great. Some have argued, though, that the role can be increased with WTO entry because then China will have access to more export markets and exports can significantly expand. But as is shown in UNCTAD (2002) and other studies, the role of exports is severely constrained by the sheer size of the Chinese
238 Elissa Braunstein and Gerald Epstein
labor force. To absorb a large amount of Chinese labor, the Chinese world export share would have to climb to politically unsustainable levels in a number of industries. Moreover, the role of FDI in the balance of payments may also be a constraint on the success of this strategy. UNCTAD (2002) reports that profits earned by FIE’s in China exceed their export surpluses by a wide margin (UNCTAD, 2002). If these profits are exported back to the home country, then the inflows of new FDI, over the long term, may significantly harm the balance of payments. Finally, we have seen that FDI not only generates exports; it also generates a great deal of imports. In short, the labor-intensive exports are also import intensive, and as a result, not only do they not generate a large amount of employment, they also fail to generate a high level of value added (UNCTAD, 2002, pp. 155–6). The key, then, for FDI to play a substantial role in improving labor outcomes would involve the upgrading of FDI: to produce more domestic capital goods and to generate more inputs and value added at home. However, it is precisely this ability to restrict FDI to higher value added channels which will be constrained. WTO’s rule of national treatment means that China will have to eliminate performance requirements over FIE’s unless it can extend them to domestic firms as well.Moreover, there will be pressure on the Chinese government to extend tax breaks and other special treatment from FIE enterprises to local ones, further undermining the Chinese government’s ability to conduct industrial policy and generate the upgrading of FDI and domestic investment as well. Of course, such upgrading will occur to some extent through general market forces. But the Chinese government will not be able to accelerate and channel this process as it has done relatively successfully for the last several decades. Hence, just at a time when joining WTO is likely to generate large scale employment losses in SOE’s (see Minqi, 2002; UNCTAD, 2002) the promise of FDI as an employment generating substitute is unlikely to bear fruit.
5
Conclusion
Can 1.3 billion Chinese consumers tame the multinationals? The answers we provide here provides a mixed picture. We certainly do not find what one would expect if China had enormous bargaining power that it was able to exercise in the interests of Chinese labor and communities. We would find at least some of the following: that more FDI leads to higher domestic tax revenue that can be used for development purposes; that it generates higher wages and/or significantly more employment; that the government would be able to channel more FDI to areas that would generate technology and productivity spillovers. To be sure, we found some of these effects. But they were much smaller than would be expected if the bargaining power were able to deliver the benefits of FDI that FDI boosters claim.
Bargaining Power and FDI in China 239
Instead, our econometric results suggest that the direct impacts of foreign direct investment on employment and wages are relatively modest, certainly much less powerful than the impact of investment and exports. Moreover, we provide strong evidence that FDI crowds out domestic investment, and is associated with a decline in provincial tax revenue. In combination with the questionable results on economic growth and productivity that we discussed earlier in the chapter, this research must certainly temper the cheerleading for FDI that is so prevalent in China and elsewhere in the developing world. And they suggest that the potential bargaining power provided by 1.3 billion consumers, in the context of a highly decentralized political system that is rife with government corruption, appears far from being realized. In countries with a fraction of the bargaining power of China, which means most developing countries in the world, China’s “success” does not offer great hope for the promise of FDI. At the same time, our results suggest that liberalization and exports have had a positive impact on wages and employment. So, it does appear that China’s policy of managed opening, combined with industrial policy and an export orientation, has had success in generating economic growth. In terms of exports, foreign direct investment has undoubtedly played an important role. But as we have seen, this has been tempered by the growth of imports and low level of value added in these industries. Still, supported by periodic expansionary macroeconomic policies and controls of speculative capital flows, these policies resulted in significant successes in promoting economic growth in China. However, this low wage, labor-intensive export orientation, funded and organized by foreign investment, is also reaching its economic and political limits. Politically, there is opposition from countries in the developed world to continued huge trade surpluses with China. Economically, these low-wage platforms hold little promise for generating high wage, high value-added jobs. China’s entry into the WTO was an attempt by China’s leaders to sustain economic growth by locking in and expanding its policy of economic liberalization, thereby making it difficult for opposition forces within the country to block the next, more difficult stages as China travels down the capitalist road. As part of this plan, the leadership attempted to maintain high levels of FDI, this time with investment that would generate more technology transfer and higher value-added jobs. But, in return, China has had to promise to give full access to its market to foreign firms, to tie its hands in its ability to carry out industrial policy, and finally to pay huge costs in terms of intellectual property rights. This will severely limit China’s ability to generate the technological upgrading required to generate more value added and employment. Moreover, with the financial liberalization that is very likely to follow the entry of foreign financial firms, the dismantling of credit allocation and capital controls is almost
240 Elissa Braunstein and Gerald Epstein
certain to follow. In short, the major tools which the Chinese government has been able to use to manage its economy will be compromised once the WTO agreements are fully implemented. This appears to be highly problematic for the Chinese people. What about for the nature of conflict versus cooperation for workers in other parts of the globe? China’s opening and entry into the WTO, to the extent that it will enhance the value of China as a location for foreign investment, might severely reduce investment elsewhere in Asia, and thereby increase the already intense competition for investment there. However, given the low tariffs on Chinese goods, this result is far from certain. It still might be cheaper for firms to locate in Vietnam, for example, and export to Southwest China. This will depend on how investment friendly the Vietnamese are compared with the local party officials across the border. It will also depend on the intricacies of the Multi Fiber Agreement and where quotas are available. It also will depend on the relative labor costs: China has lower wages than most developing and developed countries, but its labor costs, taking into account productivity, is not necessarily lower. In manufacturing, for example, they are lower than many, but not all developing countries. For example, its costs are higher than Singapore, but lower than Bolivia, Chile, and Mexico (UNCTAD, 2002, p. 158). Hence, the impact of China’s joining the WTO will depend not only on the evolution of wages in China versus the rest of the world, but also the evolution of productivity and regulations. There are many uncertainties in this regard. But, still, there is no doubt that China will provide enormous competition for many developing countries. As for cooperation versus conflict with workers in the industrialized countries, this too is somewhat unclear. The export orientation of China’s development so far has certainly led to enormous potential conflicts with workers in the North. If the next phase of Chinese development involves a substantial futher export push to generate more employment, the conflict level will increase, possibly to the boiling point. But if the next phase of China’s development will entail a much greater level of investment for the Chinese market, it is possible that some of the conflict could be reduced. Certainly, northern FDI into China might be very large; but it will likely be primarily for the Chinese market. Of course, US workers are unlikely to benefit from this foreign investment, which should certainly line the pockets of the owners of Western corporations. But the level of trade competition in the US market could level off. The best outcome for both China and workers in the rest of the world, however, would be for the Chinese government to abandon its obsession with FDI and to re-orient its economic policy to economic development at home, focused first and foremost on domestic resources (Minqi, 2001). Expansionary macroeconomic policy, control of the financial sector, upgrading of domestic investment and production, environmental protection and
Bargaining Power and FDI in China 241
the enhancement of labor’s role in the economy. And unifying more democratic control over bargaining with MNC’s and reducing corruption will do much to harness the enormous potential bargaining power of the Chinese people vis à vis international capital.
Appendix A: data appendix Unless otherwise indicated, the data discussed below are a compendium of two key sources: a provincial-level data set generously given by Robert Feenstra that covered the years 1980–95, originally culled from the statistical yearbooks of individual provinces and regions in China, supplemented by data taken from the China Statistical Yearbook, published annually by the PRC’s National Bureau of Statistics. Because the latter is generally considered a more reliable source, where possible we have relied on these statistics. Current versus Constant Dollars. Unless otherwise discussed, all units are in nominal yuan. This choice was taken as a result of our inability to get reliable deflators for the full set of variables, especially trade. It should be noted, though, that when we tried the regressions using the consumer price index on wages and the GDP deflator on everything else, the results were consistent with the nominal numbers. Wages are figured as the average annual wage for staff and workers, which includes those who work for wages and salaries with the exception of: teachers in schools run locally and foreigners and persons coming from Hong Kong Macao, and Taiwan and working in state-owned economic firms. Wages are calculated based on the total remuneration paid to staff and workers, regardless of source or category. Employment is the total number of jobs held in a province during a particular year. It incorporates all individuals working for remuneration and earning business income, including: all staff and workers, re-employed retirees, self-employed workers, and employed rural workers. Labor Force is figured as the total population fifteen and over by province. This is distinct from the concept of “economically active population” in that the latter includes only those already in the paid labor force or looking for remunerative work. We chose the wider definition to take account of those too discouraged to look for work (a potentially sizeable number in light of China’s employment problems), as well as to incorporate the potential labor of individuals that move in and out of the workforce on a regular basis (in particular married women). Investment. This category, termed “total investment in fixed assets” in Chinese statistical records, refers to the total volume of activities in construction and the purchase of fixed assets in monetary terms, regardless of the source of funds (including state enterprises and FDI, but not government
242 Elissa Braunstein and Gerald Epstein
spending). It is classified into four parts: capital construction, innovation, real estate, and other. Foreign Direct Investment includes all investments inside China by foreign enterprises and economic organizations or individuals (including those from Hong Kong, Macao, and Taiwan). It includes reinvested earnings as well as funds that enterprises borrow from abroad. Our measure of FDI includes only actual (as opposed to contracted) investments. FDI is recorded in US dollars; we transformed these into yuan using the relevant exchange rate published in the IMF’s Yearbook of International Financial Statistics. Trade includes all goods imported and exported outside the boundary of China to and from the relevant province. Imports are calculated at CIF, while exports are calculated at FOB. This data was taken from the Almanac of Foreign Economic Relations and Trade(various years), and we transformed these figures from US$ into yuan using the relevant exchange rate published in the IMF’s Yearbook of International Financial Statistics. Government Revenue includes the following: tax revenue, special revenues (including special fees such as for education); other revenues, including revenue from repayment of loans, donations and grants; and planned subsidies for the losses of state-owned enterprises (a negative revenue). Industrial Output is the total volume of industrial products sold or available for sale in a given year. State industrial output is a subset of this category, and reflects all industrial output produced by state enterprises.
Notes 1. Hong Kong’s return to Chinese rule in 1997 has not altered its status as a distinct administrative and economic entity, so it is treated separately from the PRC in this chapter. 2. He uses number of enterprises instead of reported value of capital invested to “avoid problems associated with differences in valuations by date of investment” (Chen, 1996, p. 23). Categories are as follows. Labor-intensive sectors include: food processing; food manufacturing; textiles; clothing and other fibre products; leather and fur; timber processing; furniture; paper and paper products; printing; cultural, education and sports goods; rubber products; plastic product;, non-metal mineral products; metal products; and others. Capital-intensive sectors include: beverage manufacturing; tobacco; petroleum refining and coking; chemical materials; chemical fibres; ferrous metal smelting and pressing; non-ferrous metal smelting and pressing; and transport equipment. Technology-intensive sectors include: medical and pharmaceutical; general machinery; special machinery; electrical machinery and equipment; electronics and telecommunications equipment; and instruments and meters (Chen, 1997c, pp. 16–17). 3. We thank Robert Feenstra for supplying some of these data. 4. We also tried using the ratio of state sector workers to all employees, which performed in a similar manner to the state output measure, but the latter was more robust overall and made more sense in the employment regressions. Eventually we
Bargaining Power and FDI in China 243
5.
6.
7.
8.
9. 10.
11.
12.
13.
hope to improve the liberalization measure by exercising a principal components strategy that combines a number of different measures of liberalization policies, such as the proportion of output subject to the Plan, or the proportion of free prices. China is actually divided into 22 provinces, five autonomous regions (Guangxi, Tibet, Xinjiang, Inner Mongolia, and Ningxia), and three municipalities (Beijing, Tianjin, and Shanghai). Autonomous regions and municipalities have the same administrative rights as provinces; the entire group will be referred to as “provinces.” Tibet has been left out of the dataset. There are of course important exceptions. Using panel data that included both developing and developed countries, Paus and Robinson (1998) find that: FDI has a direct positive impact on real wages; that that impact is especially true in developing countries (but not in developed countries); and finally that this positive impact is true only for the period 1968–87, after which there is some evidence that the threat effect of relocating has had a negative effect on wage growth in industrialized countries. In a comparative study of Mexico, Venezuela, and the United States, Aitken et al. (1995) find that higher levels of FDI are associated with higher wages in all three countries, but in Mexico and Venezuela, this association was limited to foreign-owned firms. This lack of evidence of wage spillovers to domestic firms is consistent with the large wage differentials between foreign and domestically owned firms in these countries. Unless otherwise discussed, all units are in nominal yuan. This choice was taken because we could not get reliable deflators for the full set of variables, especially trade. It should be noted, though, that when we tried the regressions using the consumer price index on wages and the GDP deflator on everything else, the results were consistent with the nominal numbers. In general, our results with imperfect deflators were similar to the nominal results reported here. A Levin-Lin panel unit root test, used to determine nonstationarity in panel data, was applied. The test may be viewed as an Augmented Dickey-Fuller test for panel data (Levin and Lin, 1992). First differencing the variables addresses the fact that the variable means change over time, which could result in spurious correlations if left as is. Yaohui Zhao and Jianguo Xu, 2000. “Earnings Differentials Between State and Non-State Enterprises in Urban China,” Mimeo, cited in Zhao (2000, p. 8). Herein lies the reason that we did not incorporate measures of unemployment in any of our econometric work. Official rates are widely believed to be significant underestimates. Lardy notes that the World Bank believes that large unexplained errors and omissions in China’s Balance of Payments, almost US$18 billion in 1995 and almost US$15 billion in 1996, are a sign of large unrecorded capital outflows (Lardy, 1998, p. 191). Frank Gunter (1996) estimates capital flight from China during 1984–94, and his middle estimate is that over the sample period, total accumulated capital flight was somewhere between US$50 billion and US$175 billion. In 1994, his estimate was US$20–US$35 billion, indicating there has been an upward trend (Gunter, 1996, p. 93). Bootstrapping is a Monte Carlo method designed to produce estimates of the bias and variance of an estimator by taking repeated random subsamples of the data, accumulating estimates of the quotient ˆ LT , and using the standard deviations of those estimates to get the overall variance of the bootstrap estimate (Peters and Freedman, 1984; Kennedy, 1993). These interviews were carried out in the PRC by Gerald Epstein between August, 2000 and July, 2001.
244 Elissa Braunstein and Gerald Epstein 14. 15. 16. 17. 18.
Interview, August, 2000, Dalian, PRC. See Appendix B for more information about tax policy toward foreign companies. Interview with Foreign Investment Officer, Fujian Province, May, 2001. Ibid. Interview with FIE business manager, Xiamen, Fujian, PRC, April, 2001. In another interview, a foreign investment official reports: “[We] reduced the regulations and procedures to improve the efficiency of governments; this is an important policy. Xiamen’s municipal government set up the Xiamen foreign investment executive committee to examine and approve the foreign investment projects. Then the enterprises only need to apply to the committee and not more departments.” Interview, May, 2001. 19. Interview with Xiamen official, May, 2001. 20. Of course, as the experience of many other countries will show, an industrial policy structure is neither a necessary nor sufficient condition for government corruption. 21. Interview with MOFTEC official, Beijing, April, 2001.
References Aitken,Brian and Ann Harrison (1999) “Do domestic firms benefit from foreign investment? Evidence from Venezuela.” American Economic Review 89: 605–18. Aitken, Brian, Ann Harrison, and Robert E. Lipsey (1995) “Wages and Foreign Ownership: a Comparative Study of Mexico, Venezuela, and the United States.” NBER Working Paper 5102. (A later cite is 1996. Journal of International Economics (40): 345–371.) Agosin, Manuel R. and Ricardo Mayer (2000) “Foreign Investment in Developing Countries: Does it Crowd in Domestic Investment?” UNCTAD Discussion Paper No. 146. Barnet, Richard and John Cavanaugh (1994) Global Dreams: Imperial Corporations and the New World Order, New York: Simon and Schuster. Branstetter, Lee G. and Robert C. Feenstra (1999) “Trade and Foreign Direct Investment in China: A Political Economy Approach.” NBER Working Paper 7100. Braunstein, Elissa and Gerald Epstein (1999) “Creating International Credit Rules and the Multilateral Agreement on Investment: What are the Alternatives?” in Jonathan Michie and John Grieve Smiths (eds), Global Instability: The Political Economy of World Economic Governance, London and New York: Routledge Press. Burke, James and Gerald Epstein (2001) “Threat Effects and the Internationalization of Production.” http://www.umass.edu/peri Chen, Chunlai (1996) “Recent Developments in Foreign Direct Investment in China.” The University of Adelaide Chinese Economy Research Unit Working Paper No. 96/3. Chen, Chunlai (1997a) “Foreign Direct Investment and Trade: An Empirical Investigation of the Evidence from China.” The University of Adelaide Chinese Economics Research Centre Working Paper No. 97/11. Chen, Chunlai (1997b) “The Composition and Location Determinants of Foreign Direct Investment in China’s Manufacturing.” The University of Adelaide Chinese Economics Research Centre Working Paper No. 97/13. Chen, Chunlai (1997c) “Comparison of Investment Behaviour of Source Countries in China.” The University of Adelaide Chinese Economics Research Centre Working Paper No. 97/14. Chen, Chunlai (1997d) “The Evolution and Main Features of China’s Foreign Direct Investment Policies.” The University of Adelaide Chinese Economics Research Centre Working Paper No. 97/15.
Bargaining Power and FDI in China 245 Chen, Chung, Lawrence Chang and Yimin Zhang (1995) “The Role of Foreign Direct Investment in China’s Post-1978 Economic Development.” World Development 23 (4): 691–73. Cheng, Leonard K. and Yum K. Kwan (1999) “Foreign Capital Stock and Its Determinants,” in Yanrui Wu (ed.), Foreign Direct Investment and Economic Growth in China. Chelthenham, UK and Northampton, MA: Edward Elgar. Cheng, Leonard K. and Yum K. Kwan (2000) “The Location of Foreign Direct Investment in Chinese Regions. Further Analysis of Labor Quality,” in Takatoshi Ito and Anne O. Krueger (eds), The Role of Foreign Direct Investment in East Asian Economic Development, Chicago and London: University of Chicago Press. Chi Peter S. K. and Chang Kao (1995) “Foreign Investment in China: A New Data Set.” China Economic Review 6 (1): 149–55. Chiu, Lee-in Chen (1995) “The Pattern and Impact of Taiwan’s Investment in Mainland China,” in Sumner J. La Croix, Michael Plummer and Keun Lee (eds), Emerging Patterns of East Asian Investment in China From Korea, Taiwan and Hong Kong. Armonk, New York: M.E. Sharpe. Coughlin, Cletus C. and Eran Segev (2000) “Foreign Direct Investment in China: A Spatial Econometric Study.” World Economy 23 (1): 1–23. Crotty, James, Gerald Epstein and Trish Kelly (1998) “Multinational Corporations and the Neo-Liberal Regime,” in Dean Baker, Gerald Epstein and Robert Pollin (eds), Globalization and Progressive Economic Policy. Cambridge, UK: Cambridge University Press. Dees, Stèphane (1998) “Foreign Direct Investment in China: Determinants and Effects.” Economics of Planning 31: 175–94. Federal Reserve Bank of Minneapolis (1994) “Congress Should End the Economic War Among the States,” Annual Report. Feenstra, Robert C. and Gordon H. Hanson (1997) “Foreign direct investment in Mexico and relative wages: Evidence from Mexico’s maquiladoras.” Journal of International Economics 42: 371–93. Fung, K.C. (1996) “Accounting for Chinese Trade: Some National and Regional Considerations.” NBER Working Paper 5595. Greider, William (1997) One World, Ready or Not: The Manic Logic of Global Capitalism. New York: Simon & Schuster. Gunter, Frank R. (1996) “Capital Flight from the People’s Republic of China: 1984–1994.” China Economic Review 7 (1): 77–96. Haddad M. and A. Harrison (1993) “Are there positive spillovers from direct foreign investment?” Journal of Development Economics 42: 51–74. Hanson, Gordon H. (2001) “Should Countries Promote Foreign Direct Investment?” UNCTAD G-24 Discussion Paper Series No. 9. Hanson, Gordon H. and Ann Harrison (1995) “Trade, Technology and Wage Inequality.” NBER Working Paper No. 5110. Harrold, Peter and Rajiv Lall (1993) “China reform and development in 1992–93.” World Bank Discussion Paper No. 215. Washington, D.C.: The World Bank. p. 24. Head, K. and J. Ries (1994) “Inter-city competition for foreign investment: Static and dynamic effects of China’s incentive areas.” Mimeo, Faculty of Commerce, University of British Columbia. Henley, John, Colin Kirkpatrick and Georgina Wilde (1999) “Foreign Direct Investment in China: Recent Trends and Current Policy Issues.” The World Economy 22 (2): 223–43. Huang, Fanzhang (1995) “China’s Utilization of Foreign Capital and the Related Policies.” Journal of Asian Economics 6 (2): 217–32.
246 Elissa Braunstein and Gerald Epstein Huang, Yasheng (1998) FDI in China: An Asian Perspective, Singapore: The Institute of Southeast Asian Studies. Huang, Yasheng (1999) “The Institutional Foundation of Foreign-Invested Enterprises (FIEs) in China.” The Davidson Institute Working Paper Series, Working Paper No. 264. International Monetary Fund (Various years) Yearbook of International Financial Statistics, Washington, D.C.: IMF. Jiang, Xueqin (2002) “Letter from China.” The Nation 274 (8): 23–25. Kaiser, Stefan, David A. Kirby and Ying Fan (1996) “Foreign Direct Investment in China: An Examination of the Literature.” Asia Pacific Business Review 2 (3): 44–65. Kamath, Shyam (1990) “Foreign Direct Investment in a Centrally Planned Developing Economy: The Chinese Case.” Economic Development and Cultural Change. Kennedy, Peter (1993) A Guide to Econometrics, Third Edition. Cambridge, MA: MIT Press. Kim, Si Joong (1995) “Korean Direct Investment in China: Perspectives of Korean Investors,” in Sumner J. La Croix, Michael Plummer and Keun Lee (eds), Emerging Patterns of East Asian Investment in China From Korea, Taiwan and Hong Kong, Armonk, New York: M.E. Sharpe. Kueh, Y. Y. (1992) “Foreign Investment and Economic Change in China.” China Quarterly 131: 637–90. Kwok, Reginald Yin-Wang (1995) “Hong Kong Investment in South China,” in Sumner J. La Croix, Michael Plummer and Keun Lee (eds), Emerging Patterns of East Asian Investment in China From Korea, Taiwan and Hong Kong, Armonk, New York: M.E. Sharpe. La Croix, Sumner J. and Yibo Xu (1995) “Political Uncertainty and Taiwan’s Investment in Xiamen’s Special Economic Zone,” in Sumner J. La Croix, Michael Plummer and Keun Lee (eds), Emerging Patterns of East Asian Investment in China From Korea, Taiwan and Hong Kong, Armonk, New York: M.E. Sharpe. Lardy, Nicholas R. (1994) China in the World Economy, Washington, D.C.: Brookings Institution Press. Lardy, Nicholas R. (1998) China’s Unfinished Economic Revolution, Washington, D.C.: Brookings Institution Press. Lardy, Nicholas R. (2002) Integrating China into the Global Economy, Washington, D.C: Brookings Institution Press. Lee, Keun (1995) “Structural Changes in the Korean Economy and Korean Investment in China and ASEAN,” in Sumner J. La Croix, Michael Plummer and Keun Lee (eds), Emerging Patterns of East Asian Investment in China From Korea, Taiwan and Hong Kong, Armonk, New York: M.E. Sharpe. Levin, Andrew and Lin, Chien-Fu (1992) “Unit Root Tests in Panel Data: Asymptotic and Finite Sample Properities.” University of California at San Diego Discussion Paper No. 92–93. Lipsey, Robert E. (2000) “Affiliates of U.S. and Japanese Multinationals in East Asian Production and Trade,” in Takatoshi Ito and Anne O. Krueger (eds), The Role of Foreign Direct Investment in East Asian Economic Development, Chicago and London: The University of Chicago Press. Liu, X., H. Song, Y. Wei and P. Romilly (1996) “Country characteristics and foreign direct investment in China: A panel data analysis.” Weltwirschaftliches Archiv 133 (2): 313–29. Li, Minqi (2001) “China’s Urban Unemployment: Exposed ‘Disguised Unemployment’ or Insufficient Aggregate Demand?” PERI Working Paper Number 22. National Bureau of Statistics, People’s Republic of China (Various years) China Statistical Yearbook, Beijing: China Statistics Press.
Bargaining Power and FDI in China 247 Naughton, Barry (1995) Growing out of the plan. Chinese economic reform, 1978–1993, Cambridge and New York: Cambridge University Press. Naughton, Barry (1996) “China’s Emergence and Prospects as a Trading Nation.” Brookings Papers on Economic Activity 2: 273–344. Paus, Eva A. and Michael Robinson (1998) “Globalization and Labor: The Impact of Direct Foreign Investment on Real Wage Developments, 1968–1993.” Paper prepared for the XXI International Congress of the Latin American Studies Association, Chicago, Sept. 24–26. Peters, S. C. and D. A. Freedman (1984) “Some Notes on the Bootstrap in Regression Problems.” Journal of Business and Economic Statistics 2: 406–9. Rosen, Daniel H. (1999) Behind the Open Door; Foreign Enterprises in the Chinese Marketplace, Washington, D.C.: Institute for International Economics. Rozelle, S., Y. Ying and M. Barlow (Undated) “Targeting transaction costs: An evaluation of investment policies in China’s foreign trade zones.” Mimeo, Food Research Institute, Stanford University. Shan, Jordan, Gary Tiann and Fiona Sun (1999) “Causality Between FDI and Economic Growth,” in Yanrui Wu (ed.), Foreign Direct Investment and Economic Growth in China, Chelthenham, UK and Northampton, MA: Edward Elgar. Shirk, Susan L. (1994) How China Opened its Door: The Political Success of the PRC’s Foreign Trade and Investment Reforms, Washington D.C.: The Brookings Institution. Sun, Haishun (1998) “Macroeconomic Impact of Direct Foreign Investment in China: 1979–96.” The World Economy 21 (5): 675–94. Sun, Haishun (1999) “FDI, Trade and Transfer Pricing,” in Yanrui Wu (ed.), Foreign Direct Investment and Economic Growth in China, Chelthenham, UK and Northampton, MA: Edward Elgar. UNCTAD (2002) “China’s Accession to WTO: Managing Integration and Industrialization.” in Trade and Development Report, 2002, New York: UNCTAD. UNCTAD (Various years) World Investment Report, New York: UNCTAD. UNCTAD (2002) Trade and Development Report, 2002, New York: UNCTAD. Wei, Shang-Jin (1995) “The Open Door Policy and China’s Rapid Growth: Evidence from City-Level Data,” in Takatoshi Ito and Anne O. Krueger (eds), Growth Theories in Light of the East Asian Experience, Chicago: The University of Chicago Press. Wei, Shang-Jin (1996) “Foreign Direct Investment in China: Sources and Consequences,” in Takatoshi Ito and Anne O. Krueger (eds), Financial Deregulation and Integration in East Asia, Chicago: The University of Chicago Press. Wei, Shang-Jin (2000) “Why Does China Attract so Little Foreign Direct Investment?” in Takatoshi Ito and Anne O. Krueger (eds), The Role of Foreign Direct Investment in East Asian Economic Development, Chicago and London: University of Chicago Press. Wei, Sung Shou and Lishui Zhu (1995) “Growth of Investment in Xiamen,” in Sumner J. La Croix, Michael Plummer and Keun Lee (eds), Emerging Patterns of East Asian Investment in China From Korea, Taiwan and Hong Kong, Armonk, New York: M.E. Sharpe. Wong, Christine P. W. (2000) “Central-local Relations Revisited: The 1994 Tax Sharing Reform and Public Expenditure Management in China.” Mimeo World Bank Office, China. World Bank (1992) China Strategies for Reducing Poverty in the 1990s, Washington D.C.: World Bank. World Bank (1994) China Foreign Trade Reform, Washington D.C.: World Bank. World Bank (1997) China Engaged: Integration with the Global Economy, Washington, D.C.: The World Bank.
248 Elissa Braunstein and Gerald Epstein Yeh, Anthony Gar-on (2000) “Foreign Investment and Urban Development in China.” In Si-ming Li and Wing-shing Tang (eds), China’s Regions, Polity and Economy: A Study of Spatial Transformation in the Post-Reform Era, Hong Kong: The Chinese University Press, 2000, pp. 35–65. Yong, Wang (2000) “China’s Domestic WTO Debate.” China Business Review, January–February, 2000. Young, Alwyn (2000) “The Razor’s Edge: Distortions and Incremental Reform in the People’s Republic of China.” The Quarterly Journal of Economics 115 (4): 1091–135. Young, Stephen, and Ping Lan (1997) “Technology transfer to China through foreign direct investment.” Regional Studies 31 (7): 669–79. Zhao, Yaohui (2000) “Foreign Direct Investment and Relative Wages: The Case of China.” China Center for Economic Research Working Paper No. E2000013. Zhao, Yaohui (2001) “Earnings Differentials between State and Non-State Enterprises in Urban China.” China Center for Economic Research Working Paper No. E2001001. Zhu, Gangti and Ding Lu (1998) “Evidence of Spill-over Efficiency: Implication on Industrial Policies Towards Foreign Direct Investment in China.” The Singapore Economic Review 43 (1): 57–73.
10 Rethinking International Labor Standards Michael Piore
1
Introduction
Labor standards are moving to center stage in international trade policy, but the debate has been conducted largely within the framework of competitive economic theory. In that framework, working conditions are viewed as the outgrowth of an informed choice by workers and by firms. To make the labor standards of the advanced, industrial countries a condition for trade is at best the naïve imposition of the values of rich nations upon the poor in the developing world who can ill afford them. At worst, it is simply covert protectionism (Bhagwati, 2002). There is a certain abstract logic to this view, but it is not grounded in field experience. I have been looking at firms in Mexico over the last six years as they have tried to adjust, first to the opening of that economy to trade and then to NAFTA, and in Mexico at least these views seem completely out of touch with reality (Piore, Dussel-Peters and Ruiz-Duran, 1997; and Kuznetsov et al., 2001). Particularly telling are the firms in Ciudad Hidalgo, a small city in the mountains of the Mexican state of Michoacan. The major industry is woodworking, and I went there to visit furniture-making shops as part of a study we were conducting on the impact of the opening of the Mexican economy to trade upon traditional industries. The city had been hard hit by this process: there had once been over 3000 furniture shops, but their numbers had declined rapidly in the early 1990s, and by the time I was there in 1996 there were less than 1000 left. The shops were in many ways typical of production in traditional industries such as clothing, shoes, ceramics, and woodworking in Mexico and elsewhere in the developing world. They were attached to the living quarters of the proprietor, overflowing into what in other houses in the neighborhood was the living room or the dinning room or into outbuildings which would otherwise have served as a garage or to house farm animals. The work was laid out in a helter-skelter fashion, seemingly without order. And, of particular relevance to the debate about labor standards, there were children 249
250 Michael Piore
everywhere and of all ages. The older children helped with the work, fetching wood pieces on command for their parents, painting and varnishing, occasionally even cutting wood on electric saws, or at least learning how to do so. The younger children chased each other around the shops, jumping over open vats of glue and varnish; it was often hard to tell whether they were actually working or just playing games. In one shop, there were two babies sleeping on a pile of rags on the floor next to the table where their mothers worked sanding wood pieces. Health and safety was also a problem in these shops, certainly for the kids and often for the adults as well. The electric saws were unguarded. The glue and varnish emitted toxic fumes; the material was probably poisonous if ingested and with prolonged exposure carcinogenic – or at least I knew that this was the case in similar shops I had visited in Italy in the 1980s. But while the shops in Ciudad Hidalgo violated the rudimentary child labor norms and health and safety standards of advanced industrial economies right and left, they were not without a set of norms and standards of their own. Much of what was going on, in fact, was not child labor at all, but child care. The women who were working were also watching; they would stop work to nurse a baby when it awoke or mind the children if the games got out of hand. The line between play and work was fluid. As the children grew up, they were drawn progressively into the work process – the line between work and play shifted gradually in a way that one could imagine more or less protected the health and safety of the kids and respected the limits of their physical and mental capacities. For the older kids whom we saw being taught to use the electrical equipment, the work was still something of a game. Imposing international standards upon these shops would have been in this sense a kind of cultural imperialism. The shops of Ciudad Hidalgo were not, however, efficient; they were having real trouble competing in the international marketplace; and the blatant violation of international standards was not helping matters, indeed quite the contrary. In fact, we were following in the footsteps of a consultant who had been hired to help upgrade their operations, and part of our mission was to evaluate his impact. The owners explained to us what was pretty evident to anyone familiar with modern production techniques: that the layout needed to be altered to smooth the flow of work, and the aisles cleared. They showed us, as the consultant had showed them, that the open vats of glue and varnish collected sawdust, which spoiled the finish on the wood. The children were a distraction which continually threatened the quality and the consistency of the work. Most of the changes the consultant had recommended to improve efficiency would clearly have improved health and safety as well. One had to conclude that if the changes had been imposed by international labor standards, however imperialistic, they would have contributed to efficiency. Certainly, covering the barrels to protect them from the sawdust would have also protected the workers from the noxious
Rethinking International Labor Standards 251
fumes and reduced the chances that the kids would ingest the poisonous material. Clearing passageways through the shop and isolating the equipment in well-defined work stations would have made it easy to install guard rails. But an efficient layout would also have repositioned the mothers into places where they could no longer see their babies or watch the older children. No amount of reorganization could completely eliminate the distraction of small children running around the shop. Ultimately, an efficient shop by western standards would have sent the children home. None of these reforms were actually introduced in Ciudad Hidalgo. This was partly because the shops could not afford the continuing help of outside consultants that was required to upgrade their operations. But, more fundamentally, because the children were already home; they had nowhere else to go. Indeed, child labor is basically one symptom of a more fundamental problem of these shops: the way in which productive activity is interwoven with, and inseparable from, the household. Thus, business and household finances were intermingled; the shop which, by common consensus, had the best designs and the most refined craftsmanship had been driven to the edge of bankruptcy when the head of the household fell sick and the working capital of the business was diverted to pay for his surgery. Others families lost their homes or their transportation when the business went broke. Infrastructure – transport, water, sewage – was inadequate because most of the shops were located in residential neighborhoods and interspersed with private dwellings. The shops were, in fact, a perfect illustration of one of Max Weber’s basic points about industrial capitalism: that its emergence requires the introduction – indeed in early capitalism it required the invention – of a distinction between the business and the household realms (Weber, 1958). In late capitalism, at least in a country like the United States, it actually involves the creation of a third realm, a realm of education, which frees women from the responsibilities of childcare to enter the labor market. The conventional wisdom in development policy is that prohibitions on child labor are only effective when accompanied by the expansion of the school system (Weiner, 1991). Unable to restructure their production process, the furniture shops in Ciudad Hidalgo were responding to the pressures of the international marketplace unleashed by opening to trade by cutting costs, reducing employment levels, intensifying the pace of work, and cutting wages. The result was to compromise prevailing standards and weaken protective structures. As the pace of work increased, people were distracted from watching the children; even when they remained committed to coming to their aid in a crisis, they were less likely to anticipate problems. Children who normally never would have been allowed to touch dangerous equipment unsupervised began to do so. As the amount of adult labor in the shop was reduced, children were drawn into production prematurely or prematurely assigned tasks
252 Michael Piore
inappropriate for their age. Thus, if the production system prior to trade could be said to respect its own set of standards, however much they might have deviated from those of the industrial world, the impact of trade was to force them to violate not only international standards but also their own standards. They may have been better off materially than if they had failed to respond to the pressures of globalization at all – indeed, if that had happened they would have gone out of business. But they would have been much more competitive internationally, and better off materially as well, had they adopted a production system consistent with our labor standards. This is no accident: it follows directly from the nature of the global forces to which the opening to trade is subjecting the Mexican economy. Those forces are coming from the industrialized world. The problem the traditional industries are having in Mexico – and in the developing world more broadly – is that they cannot meet industrial standards of quality and reliability. They are forced therefore to compensate for their defects by reducing prices. They have to find some way of meeting those standards. Since the standards are coming from the industrialized world, it is clear that one way of meeting them is to adopt the production practices of the industrialized world as well. The labor standards of the industrialized world are consistent with those production practices because they were developed in combination with them. Indeed, where the production practices could not be modified to meet those labor standards, the standards were probably adjusted to the production practices. The “cultural” imperialism – if that is the right way to think about what is happening – is thus coming from international competition, and the developing countries opened themselves up to its dynamics when they opened their economies to trade. The real question is not whether the practices the industrial economies are pursuing are more effective than those which Ciudad Hidalgo is adopting, but rather whether there is actually a third way, a set of practices that would enable Ciudad Hidalgo to compete in the international marketplace without simply becoming a carbon copy of, to put it in the crudest terms, a small American city. One such alternative, more compatible with traditional household production and the labor standards which grew up around it in Ciudad Hidalgo, is to move upmarket into the production of customized items for specialty niche markets. This strategy had been developed extensively by the industrial districts of central Italy in the 1970s and almost a century earlier by the English districts described by Alfred Marshall (Pyke et al., 1990; Schmitz and Nadvi, 1999). Marshall maintained that in the English districts people absorbed the craft skills and sense of the materials from the very air they breathed, which is surely the experience of children growing up playing in the shops of Ciudad Hidalgo. This kind of production, of course, has exacting standards of quality as well, but they are very different from the quality standards of industrial production. The latter involve an emphasis on
Rethinking International Labor Standards 253
reliability and consistency: production to precise specifications, within a well defined and limited margin of error. The standards of niche markets are much closer to the aesthetic of craftsmanship: novelty and an absolute standard of perfection. These standards are consistent with the leisurely pace of work and the informal organization of the Ciudad Hidalgo shops. Production for particular market niches would also require a considerable adjustment on the part of the producers of Ciudad Hidalgo, but one which would in many ways be easier for the true craftsmen among them to understand and appreciate than the industrial standards, which focus on consistent production of medium quality goods and are often quite alien to the craft aesthetic. We actually encountered one furniture-maker in Michoacan (although not in Ciudad Hidalgo) who had quite successfully adopted a niche strategy of this kind. Also, the traditional industries of Jalisco have recently begun to systematically pursue an approach modeled on central Italy.
2
The new garment industry in Mexico
Most of the “traditional” industries of Mexico today, however, are really new. While traditional firms are sweating to survive in the face of international competition, and many are going out of business, a whole new productive structure composed of new firms and new plants is taking their place, especially in the garment industry. Mexico has become the largest foreign producer for the US clothing market. The new plants are tied closely to foreign firms, either owned directly or operating as subcontractors. The plants are laid out and organized in accordance with the most up-to-date engineering and management principles. Indeed, they are typically designed by engineers from the parent company, or in the case of subcontractors the principle client. These engineers also train the first generation labor force and managerial staff; and a small staff of engineers from the parent company or client is usually permanently out-stationed in these plants to make sure that standards are maintained over time. The plants thus look and feel very much like their counterparts abroad. They have the look and feel of modernity which contrasts sharply with traditional producers in Mexico and even with smaller garment shops in urban centers such as those in New York and Los Angeles. There are no small children in the shop; they are more spacious and have better ventilation; rudimentary health and safety is better too. But the plants are typically more labor intensive than the US plants upon which they are modeled; they use less capital equipment and work the equipment that they do have more intensively, so the pace of work is faster as well. Because of the pace and because there is less machinery to move bundles around, the work is a good deal more strenuous than it would be in a typical American shop. The workforce is composed largely of young women, most of whom start working in the industry before they have children, and children are not tolerated in the plant. The real danger is the impact of the
254 Michael Piore
strenuous and repetitive labor. It would be hard to say that the workers are rationally choosing to pay for their material well-being in this way, since the cost in terms of health is only apparent over time, and the factories have not been in place long enough for the effects to become apparent. Fairly typical of these new traditional industries are the garment plants in Tehuacan, which Natasha Iskander studied last summer (Iskander, 2000; Iskander, 2002). The workforce there is composed largely of migrants: The vast majority … are young campesinos, in their mid-teens or early twenties, from surrounding rural hamlets in the mountains. Their departure represents a death-knell for the dying rural economies that they abandon, making it much more likely that their move to Tehuacan will be permanent. The social setting that they confront in Tehuacan has challenges they were unlikely to face in their home communities. Drug use among maquilla workers is relatively widespread, with speed being the drug of choice as it makes it easier to withstand the intense pace of production lines. High levels of sexual activity among workers are commonplace, and community health workers are already expressing concern that maquillas are strong centers of HIV transmission. A more visible expression of the sexual activity among maquilla workers is the rate of single motherhood: according to the public health department (regidor de salud), approximately 80% of all women who work in maquillas are single mothers, a source of real concern for the municipality that may ultimately have to provide childcare as mothers have moved far away from the family structure that would otherwise have provided that kind of support. (Iskandar, 2000) Work standards are too narrow a focus to capture the full impact of the emergent garment industry which Iskander points out: … misses the impact that maquillas have on workers’ lives outside factory doors. In Tehuacan, these impacts were at least as damaging and more enduring than the hardships workers experienced on production lines. Tehuacan was once famous for its plentiful and pure mineral waters, said to have curative properties. Now, the water table drops by a meter and a half annually as factories that stone- and acid-wash the assembled jeans supplement their extravagant use of water with illegal wells that tap into rapidly diminishing underground reserves, threatening this arid and isolated region with permanent drought. . . . These industrial launderers dump the used water, largely untreated, into the local rudimentary drainage system and the opaque bright blue waters, heavy in lead and other chemical content, run through the town in shallow unlined canals. Children play in and around these waters, and the contaminants seep through the soil to nearby household wells that are the only source of
Rethinking International Labor Standards 255
supposedly potable water available to many families. Gastrointestinal problems are already a significant problem in many neighborhoods, and the longer term health impacts of this pollution are already beginning to manifest, with the municipality of Tehuacan displaying an alarming rate of birth defects. . . . Several branches of the drainage system flow directly into the waters that local farmers use to irrigate their fields, and in many areas the soil is stained blue. To address the threat of immediate contamination to crops posed by the industrial discharge, the municipality has ordered farmers to cultivate only “tall” crops such as maize and has banned altogether the harvest of “short” crops like tomatoes, cilantro and cabbage from many fields. Unfortunately, these measures do nothing to deal with the long-term impact that the contaminants will have on soil fertility, most likely turning them into inarable and dusty stretches of arid land. In addition to ecological damage, the maquillas in Tehuacan are responsible for significant social dislocation in the town itself and in the surrounding areas. With production that has doubled every year for the last five years, the maquillas have set off a landslide of new immigration that has literally overwhelmed the city. The municipality’s Department of Urban Planning estimates that about 14,000 new migrants move to Tehuacan every 6 months. The new arrivals settle in makeshift informal settlements at the town’s outskirts, quickly erecting houses of sheet metal and woodscraps. These new shantytowns lack even the most basic infrastructure, with their absence of potable water significantly complicating Tehuacan’s perpetual struggle against cholera outbreaks. (Iskandar, 2000) The way in which the new productive structure is wearing away at the human and physical infrastructure of Tehuacan is particularly troubling because the garment industry is notoriously fickle. It is easy for virgin industrial territories like Tehuacan to attract the industry because its location decisions are motivated almost entirely by low wages and abundant labor. The raw materials and finished products are light and easily transported in and out, and the capital equipment is minimal. But for these same reasons its roots in these new areas are shallow. And the growth which the industry creates when it first moves in has repeatedly proven to be short-lived. When wages rise and labor becomes scarce, the industry moves on to other parts. Examples of places that have been used and then abandoned include Puerto Rico, which grew miraculously in the 1950s and 1960s only to stagnate in the 1970s and 1980s when the industry moved out to other Caribbean islands, Asia, and El Paso, Texas, from whence many of the plants now in Mexico came. A strategy of garment-led industrial development makes sense if the prosperity is used to build a base which will attract a second generation of more sophisticated industry, but the uneducated, illegitimate children, overburdened urban infrastructure, and polluted natural resource
256 Michael Piore
base which the industry is producing in Tehuacan are not a legacy that is likely to have this effect.
3
China
Tehuacan, however, is not the only pattern of garment industry development. A very different model is found in China, one which offers a solution to much of what is problematic in the way the industry has developed in Mexico. In China, new garment plants are also staffed by migrants from rural villages, but the women live in dormitories built by the company adjacent to the plant, in special compounds surrounded by chain link fences. The women are sheltered from the corrupting influence of drugs and sex, the company bears the cost of new urban infrastructure required to support the expanding population and the development of shantytowns is forestalled. The physical damage to the territory occurs within a contained space. The earnings of the workers are transmitted directly to their families who remain behind in the villages from which the women came, and the women themselves return there to reenter village life when their tour of duty in the factories is complete. But the virtues of China’s approach to garment production is also its defect. The women are virtual prisoners in these factories; they do not have access to the outside world; nobody really knows what the pace of work is within, nor to what materials they might be exposed; and the workers are barred access to outsiders who might encourage them to question the conditions under which they work and to organize to seek redress. They are under enormous pressures from their families at home to earn as much as they possibly can. They have no effective voice in their work lives and no effective opportunity to exit either (Berger and Lester, 1997).
4
New industry: electronics and software
The dormitory/factory communities in China are not, of course, a new invention. They are a revival of a pattern of labor recruitment and deployment that was characteristic of the New England textile industry in the early nineteenth century, before the great waves of foreign immigration, when American industry was dependent on drawing workers out of domestic agriculture. But there are actually patterns of work organization in developing economies that do not have longstanding historical roots. Virtually every major company in the global computer industry has located a facility in Guadalajara, Mexico, and all of them rely almost entirely on a system of contract labor that completely undermines conventional notions of industrial employment. Over 7000 people work in IBM’s El Salto computer facility for example, but only 612 are permanent IBM employees. All of the rest, from computer programmers at the top of the occupation hierarchy to mother
Rethinking International Labor Standards 257
board assemblers at the bottom, work either as independent contractors or as subcontractors who operate production lines inside IBM’s premises. Twelve thousand others work at adjacent manufacturing facilities (Kuznetsov et al., 2001). The pattern in Guadalajara is extreme, but the tendency toward employment arrangements of this kind is characteristic of the electronics industry throughout the world, in the United States, no less than in Asia and Latin America. It is very hard to know what to make of these arrangements. They abrogate the traditional notion of the employer and confuse legal responsibility. They have been used in other industries to evade conventional labor standards, and in the United States at least studies have shown that they lead to significant increases in industrial accidents (Kochan, Smith, Wells, and Rebitzer, 1994). But it has been argued with some cogency that the computer industry is organized around the concept of a project, which requires a continual shifting of the mix of skills in a manner comparable to the construction industry, and that the notion of permanent employment under these circumstances is obsolete. What can one conclude? What conclusions can one draw from these several examples for the question about international labor standards? The examples, it seems to me, make very dubious the claim of neoclassical economists that prevailing standards represent the informed choice of rational agents in the face of the constraints imposed by inherently limited resources in an underdeveloped economy. There is no particular reason to believe that the prevailing conditions are the best that these people can do even in the short run, and certainly no reason to believe that they contribute to long run economic development. The workers and the communities in which they live may just as well be the victims of the process of globalization as its beneficiaries. On the other hand, the specific problems presented in each of the four cases are very different. The first (Ciudad Hidalgo) presents the question of enforcing a set of indigenous standards under pressure from international competition. The second (Tehuacan) involves the enforcement of standards in a system of production imported from abroad, where the source country has a great deal of experience in dealing with the problems this strategy of production poses for working conditions. In all three cases, but particularly the contrast between Tehuacan and China, the question is the relationship between labor standards narrowly conceived and the broader choices which the country (or the locality) has to confront in the face of the challenge of economic development. The problems of evaluating labor standards in the electronics industry in Guadalajara are almost precisely the same as those of evaluating standards in the electronics industry in the United States, and in both cases regulations have to negotiate territory for which there is relatively little relevant historical experience. These problems are so different that it is difficult to imagine a single set of standards that would cover all of them or
258 Michael Piore
a single set of institutions that would address the problems they present, even on a national level, let alone a set of international institutions. The contrasting cases do, however, begin to suggest a set of principles which might guide the search for an institutional structure that would address the problems of labor standards. Four basic principles seem to arise directly: ●
●
●
●
5
First, one would want a system which respected the existing standards of production systems that were embedded in an established social order, such as that of the furniture industry in Ciudad Hidalgo before it came under the pressure of global competition. Second, one would like a system which was flexible enough so that established systems could adjust to changes in the competitive environment in a reasoned and informed way. In Ciudad Hidalgo, specifically, one would like the producers there to be able to choose a different productive system, even if one were willing to accept a debased version of the system already in place. Third, one would like to be able to draw upon the experience of industrialized countries in coping with the systems which had grown up there and were being exported to other parts of the world. It is one thing to say that the people of Tehuacan, or China, are deciding that they cannot afford to respect the health and safety standards of the United States and quite another for them not to know about the dangers to health and safety to which those standards are a response. Finally, one would like to encourage informed debate and discussion, particularly discussion and debate which linked labor standards to the broader issue of economic development. This last goal is in some ways implicit in the conventional economic argument that the developing countries whose labor standards we view as debased are making a calculated choice. But they seem to reinforce the value of informed choice itself.
A regulatory structure
While it is hard to imagine a single set of regulations and institutions that would handle all of these cases, it is possible to envisage pieces of a regulatory structure that would address many of them. Seemingly the easiest and most relevant are those governing traditional industries in the industrialized countries. Here the experience is extensive and the regulations were put in place at a much earlier moment of history when their income was low and the workers often lived on the margin of subsistence. The concern that rich countries are imposing their preferences on the developing world, which cannot afford the luxury of such labor standards, is minimal.
Rethinking International Labor Standards 259
These regulations could be imposed simply by extending the legal obligations of companies from the United States, for example, to their foreign direct investments. We do that already with legislation on bribery and on intellectual property; it is hard to distinguish these precedents from labor rights. The obligations of companies in the industrial world could be extended further to their subcontractors. US companies have been particularly resistant to this obligation, claiming that the monitoring of labor conditions in subcontracts imposes an intolerable burden upon them. But when one actually sees the relationship between the contractor and subcontractor in Mexico, that claim is patently absurd. They exercise the most stringent control over anything that affects the quality and reliability of the product. And, in fact, one of the conditions which they impose on their suppliers is that the supplier cut off the whole chain of secondary and tertiary contractors, which is typical of the garment industry in most of the world, precisely because it is so difficult to control quality and reliability along an extensive contracting chain (Gereffi et al., 2002). While these regulations have been slow to develop through law, they are in fact growing out of consumer movements, and nongovernmental regulatory processes have developed through negotiations with major brand name producers of clothing and shoes (O’Rourke, 2001; Fung et al., 2001).
6
Sweatshops
The longest-standing labor standards are those controlling the sweatshops in the garment industry. Sweatshops have their economic roots in a particular type of cost structure, a cost structure in which labor is the major cost of production and workers are paid by the piece. Where this is the case, the major cost which is independent of output is rent for the space in which the work takes place. For this reason, the employer tries to minimize rent by cramming as many workers and material into the production space as possible. The crowded conditions and the health hazards associated with them often have a detrimental effect on worker productivity, but the impact on the employers is minimized by piece rates, which effectively shift the cost of the sweatshop to the worker and eliminate any incentive which the employer might have to control working conditions at all. The ultimate logic of this system of production is industrial homework, where the worker even pays the rent and all costs are variable. The system also encourages child labor since the piece rate system compensates for a child’s low productivity. The system was finally brought under control by a complex set of regulations, ranging from health and safety standards which limited the impact of crowded conditions in the shop, to child labor laws, and a minimum hourly wage which controlled the piece rate system. In the United States, we actually outlawed industrial homework as well. The regulations were administered by a variety of different agencies but since they were all attacking the
260 Michael Piore
same basic system of production, the efforts of these agencies reinforced each other (Piore, 1990). The laws governing the sweatshop in the garment industry date from the early twentieth century and were the culmination of efforts which began at least twenty years before. These efforts were spearheaded by immigrant workers, at a time when their income was comparable to those in Mexico or China today. It is difficult to argue that the regulations constitute a luxury which only the high income workforce of modern industrial nations would rationally chose. But the extension of sweatshop regulations would do little to reach the problems suggested by the cases described above, and in some respects they would foreclose important alternatives. Their greatest impact would be on the production system that has developed in Ciudad Hidalgo under the impact of trade. That system is a sweatshop system, and an extension of the sweatshop regulations to Mexico would foreclose it. The problem is that the regulations might well foreclose the production system that had prevailed in Ciudad Hidalgo before the opening to trade and the kinds of adjustment patterns modeled on the shops of the industrial districts of central Italy. The regulatory structure might limit the abuses in the large garment factories of Tehuacan and China, but these production systems are not sweatshops. The fixed capital investment is large enough to foreclose a true sweatshop. Basic health and safety compares favorably – quite favorably in fact – to unionized shops in New York City. The major problem in these shops appears to be the pace of work and the length of the work day; and these characteristics of the production system are not really reached by the sweatshop regulations. Clearly a different regulatory system is required here, and to reach the issues involved in Ciudad Hidalgo, a much more subtle one, one which respects local variation and accommodates to it.
7
Respecting local differences
Collective bargaining in the United States It is possible to design a system of social regulation that respects local differences. Decentralized collective bargaining in the United States, although it has come under attack in recent decades for its bureaucratic rigidity, actually functioned in this way. The underlying theory upon which the system was built is that each shop was in fact a social system with its own code of behavior, and that the code represented a compromise which had evolved over time between the technical and the economic exigencies of production and the social standards of the work place. Pressures from one side or the other operated to push practice outside the bounds the code permitted. Collective bargaining was a means of amending and updating that code, through negotiations to accommodate these pressures. But it was also a framework for enforcing the code through the systematic adjudication of
Rethinking International Labor Standards 261
disputes. Adjudication, however, required experts skilled in uncovering and interpreting these local laws (or more precisely customs). A group of specialized labor arbitrators grew up who cultivated and practiced precisely these skills of local adjudication. At the apogee of the postwar system of collective bargaining in the United States, the US courts deferred completely to the judgment of these skilled arbitrators. As the Supreme Court held up in its 1960 decision: The collective bargaining agreement covers the whole employment relationship. It calls into being a new common law – the common law of a particular industry or a particular plant … [It] is an effort to erect a system of industrial self-government … [T]he grievance procedure is at the very heart of the system of industrial self-government. Arbitration is the means of solving of unforeseeable by molding a system of private law. [It is the end point in the process] by which meaning and content is given to the collective bargaining agreement. (United Steelworkers of America v. Warrior and Gulf Navigation Co, 1960)
Bureaucratic regulation in France The US system of labor law is among the most extreme in its decentralization and its deference to local practice. But the French labor code, while in theory at the very other end of the spectrum with a nominally uniform code and administered by a corps of bureaucrats in the Ministry of Labor, operates very much to the same effect. The code is so complex that the bureaucrats who enforce it, the inspecteurs du travail, cannot possibly apply it literally. They use it instead as an instrument for maintaining a social order, the terms of which vary from one region to another, and the industry, the shop, and each inspector attempts to balance the custom of the workplace against the pressures of the economic environment (Duprilot, 1975; Berger and Piore, 1980). Such a system, especially on an international scale, is not without its problems. First of all, it is difficult to imagine the American system at least without worker representation. It is, after all, the workers’ participation in collective bargaining which makes it possible to change the rules to adjust to variation in the business environment. But employers are at least as resistant to procedural protections for workers rights as they are to substantive standards. These rights have in fact been severely compromised by the decline in union membership in much of the industrial world over the last 20 years, most especially in the United States where legal protections have diminished as well. The French system substitutes a corps of highly trained government officials for union representation (Duprilot, 1975). But this kind of elite bureaucracy has proven very difficult for developing countries to create and sustain. The framework of rules which such a bureaucracy administers, moreover, gathers its legitimacy and much of its flexibility from the
262 Michael Piore
potential for amendment through the political process; and developing countries, most notably China, have been unwilling to guarantee the most basic rights of free speech and free press which would make such a process meaningful. These, however, are political obstacles which have nothing to do with the economists’ case against labor standards as a luxury of the rich or covert protectionism.
8
The inflexibility of flexible systems
The more telling problem, however, is economic, albeit not of the kind which has generally been raised as an objection to standards as a condition for trade: In recent years, both the French and American systems have proven to be much less flexible in practice than in earlier periods. This, it appears, is because of the radical changes in the business environment and productive technology which these countries, no less than Mexico, have sought to accommodate. The systems governing the workplace evidently are quite capable of compromise within the framework of a given production system by amending one rule or discarding another. But what has been at stake since the first oil crisis in the developed world is the ability to shift from one production system to another. Indeed, what existed in France and the United States is not perhaps all that different from what prevailed in Ciudad Hidalgo before the opening to trade. Evidently, it would seem to be a tendency for all social systems to generate work codes, sometimes embodied in national law, sometimes left to more local and less formally instituted processes, which govern work relations. The codes grow out of the interaction between market pressures and social norms and evolve over time. But the process of gradual evolution can be overwhelmed by the sudden surge of market pressure, such as those coming out of great depressions and world wars, or the sudden opening of previously closed economies like Mexico’s to trade, or the combined forces of trade, deregulation, and new technology with which American industries tried to cope in 1970s and 1980s. The economy cannot always adjust to these kinds of surges within an existing production system. When the changes in the environment are permanent, in fact, firms probably need to adopt or invent totally new systems of production to survive. For such radical adjustments, the moral code of the old system acts as a major inhibition. People resist new practices as immoral when, in fact, those practices lie totally outside the range which the moral system was designed to judge. What is required is not the application of the norms of the old system to the new, but the development of a moral code appropriate to the new work context. International experience could presumably facilitate such transition by providing a menu of alternative systems and moral codes which go with them and hence helping to distinguish between a shift in “moral compass” so to speak and a kind of moral anarchy.
Rethinking International Labor Standards 263
Ultimately, to accomplish a shift of this kind in a way that is perceived as moral requires that people recognize that morality exists on at least two levels. One of those levels is concrete, embodied in the specific code associated with a prevailing production system. But those concrete codes are embedded in a second, more fundamental, set of abstract moral principles; the concrete codes are basically an interpretation of the abstract principles in a particular context; and those principles get reinterpreted as one moves from one such context to another (Bloch, 1961; Piaget, 1965). It is not clear that such shifts in the level of moral judgment and across different concrete moral systems could be managed by a system of international jurisprudence. But the international community might nonetheless play a role. It might be able to articulate, and gain adherence to, a common set of abstract moral principles. The application of those principles to formulating concrete codes for particular production systems and the decision to move from one such code to another must grow out of very local discussions and debate. But the international experience could probably inform these local discussions as well. Viewed in these terms, one of the problems with the international regulations that have been discussed historically, particularly those promulgated by the ILO, is that they are linked to the particular production systems, essentially those associated with large-scale mass production. As such they are too specific. This is one of the reasons underlying the split between the formal economy, which in the past at least has been guided by the organizational principles of mass production and the informal economy, where a variety of other production systems prevail. This was less of a problem when we had a clear view of what industrial society was all about and where it was headed. The informal economy then seemed like the vestiges of a system which would eventually disappear. But in recent years, our vision of what a modern economy looks like and how it will evolve over time has become much more confused. Many of the production systems associated with the informal economy seem to be reappearing in the high tech start-ups.
9
The conventional framework
A framework for thinking about this problem is suggested by the French school of economic sociology known as the conventionalist school. Their work grows out of studies by Luc Boltanski of what he calls denunciation: attacks conceived within one system of moral judgement upon practices conceived in another (Boltanski et al., 1984). Boltanski then went on with Laurent Thévenot to develop a typology of moral systems and a research program which explores the relationship between these moral types and a parallel typology of production systems (Boltanski and Thévenot, 1987). This work begs the question of where the alternative moral systems come from in the first place and how they become linked to particular systems of production. The conjecture implicit in their work – and indeed in the
264 Michael Piore
argument I have been developing here – is that differences in moral standards governing the workplace reflect differences in production systems, not in national cultures. Their initial typology was abstracted from a reading of philosophical texts, guided by Boltanski’s early work on denunciation. But recently, Boltanski has been studying the emergence of a totally new moral system associated with the production paradigm growing up around information technology (Boltanski and Chiapello, 1999). This study suggests the rudiments of a theory of how new types of morality emerge from the production system itself. This research does not resolve the labor standards debate. It does, however, help us to identify a domain of common international standards for situations like those in the garment industry where there is considerable commonality across nations in the menu of production system from which they can chose. It also helps us to understand what the limits of commonality are. And it provides a framework for gathering international experience and for organizing it in a way in which people from different cultures can make informed, moral choices in situations which lie beyond the range of their own experience. It can be viewed not so much as an alternative to the conceptual framework of economics, which has dominated the standards debate, but as a complement, a framework which explicitly recognizes moral judgment and makes it a central focus. It might enable us, by comparison among the moral standards of many different production systems, to identity a set of abstract principles which underlie moral choice more broadly. If one were to develop the list of abstract principles which underlie the specific moral codes associated with particular production systems, the debate about labor standards suggests that one item on the list would be the principle of informed choice, implicit in the economists’ critique of international standards. Perhaps paradoxically, the research agenda of the conventionalist school may make it possible to approach the economists’ ideal in a way that the framework in which their critique of labor standards was conceived does not.
References Berger, Suzanne and Richard K. Lester (1997) (eds), Made by Hong Kong, Hong Kong: Oxford University Press. Berger, Suzanne and Michael Piore (1980) Dualism and Discontinuity in Industrial Society, New York: Cambridge University Press. Bhagwati, Jagdish (2002) “Free Trade and Labour,” Financial Times, August 29. Bloch, Marc (1961) Feudal Society, Volume 2, Chicago: University of Chicago Press, pp. 113–16. Boltanski, Luc and Eve Chiapello (1999) Le Nouvel Esprit du Capitalisme, Paris: Gallimard. Boltanski, Luc, Y. Darre and M-A Schiltz (1984) “La Dénonciation,” Actes de la recherche en sciences socials 51: 3–41. Boltanski, Luc and Laurent Thévenot (1987) Les Economies de la Grandeur, Paris: Presses Universitaires de France.
Rethinking International Labor Standards 265 Duprilot, Jean Pierre (1975) “Le controle adminstitratif des licenciements,” Droit Social, June (No. 6): 53–75. Fung, Archon, Dara O’Rourke, and Chuck Sabel (2001) “Realizing Labor Standards,” The Boston Review, New Democracy Forum: February/March. Gereffi, Gary, David Spener and Jennifer Bair (2002) (eds), Free Trade and Uneven Development: The North American Apparel Industry After NAFTA. Philadelphia: Temple University Press. Iskander, Natasha (2000) “Beyond factory walls: Labor Standards and the Case of Tehuacan, Mexico,” MIT Center for International studies Newsletter, Fall 2000. Iskander, Natasha (2002) “International Labor Standards in Local Struggles: Tehuacan, Mexico,” mimeo, Sloan School of Management, MIT. Kuznetsov, Yevgeni, Michael Piore, Clemente Ruiz Duran, and Charles Sabel (2001) “Think Globally, Act Locally: Decentralized Incentive Framework for Mexico’s Private Sector Development,” World Bank Informal Research Report #22643. O’Rourke, Dara (2001) “Sweatshops 101: Lessons in Monitoring Apparel Production Around the World,” Dollars and Sense, September/October. Piaget, Jean (1965) The Moral Judgment of the Child. Translated by Marjorie Gabain. New York: The Free Press. Piore, Michael J., Enrique Dussel Peters and Clemente Ruiz Duran (1997) (eds), Pensar Globalmente Pensar Globalmente y Actuar Regionalmente: Hacia un Nuevo Paradigma Industrial Para el Siglo XXI, Universidad Nacional Autónoma de México: Fundación Friedrich Ebert: Editoral Jus, Mexico. Piore, Michael J. (1990) “Labor Standards and Business Strategies,” in Stephen Herzenberg and Jorge Perez Lopez (eds), Labor Standards and Development in the Global Economy, Washington, D.C.: U.S. Department of Labor, Bureau of International Labor Affairs, pp. 35–50. Piore, Michael J. (2002). “Thirty Years Later: Internal Labor Markets, Flexibility and the New Economy,” Journal of Management and Governance, 6: 271–9. Pyke, Frank, Giacomo Becattini, and Werner Sengenberger (1990) (eds), Industrial Districts and Interfirm Cooperation in Italy, Geneva: International Institute for Labour Studies. Sengenberger, Werner, and Duncan C. Campbell (1994) Creating Economic Opportunities: The Role of Labour Standards in Industrial Restructuring, Geneva: International Institute for Labour Studies. Schmitz, Hubert and Khalid Nadvi (1999) “Clustering and Industrialization: Introduction,” World Development 27 (9): 1503–14. Thevenot, Laurent (1985) “Introduction,” in Laurent Thevenot et al., Conventions economiqus Cahier du center d’etudes de l’emploi, Presses Universitaires de France, pp. v–xviii. Thomas A. Kochan, Michal Smith, John C. Wells, and James B. Rebitzer (1994) “Human Resource Strategies and Contingent Workers: The Case of Safety and Health in the Petrochemical Industry,” Human Resource Management 33 (1) Spring: 55–78. United Steelworkers of America v. Warrior & Gulf Navigation Co., No. 443, Supreme Court of the United States, 363 U.S. 574; 80 S. Ct. 1347; 4 L. Ed. 2d 1409; 1960 US. Weber, Max (1958) The Protestant Ethic and the Spirit of Capitalism, New York: Scribner. Weiner, Myron (1991) The child and the state in India: Child labor and education policy in comparative perspective, Princeton, N.J.: Princeton University Press.
Index Aalfs, 72 acquisitions and mergers, 9, 190, 193, 199, 203 aerospace industry, 51, 53, 56n12 affiliates skill upgrading, 122, 124–8, 129–30, 131–3, 134, 136–40 South Africa, 188–90, 191, 192, 197–201, 202, 203 wage elasticity of labor, 9 see also multinationals Africa crowding in effect of FDI, 230 deindustrialization, 11, 28 FDI inflows, 187 FDI stock in China, 216 import substitution industrialization, 32 industrialization, 54 manufacturing, 24, 25, 26 stagnation, 29 trade integration, 37 agglomeration, 125, 129 Agosin, Manuel R., 230 Agreement on Textiles and Clothing, 61 Agreement on Trade Related Aspects of Intellectual Property Rights (TRIPS), 180 Agreement on Trade Related Investment Measures (TRIMS), 180, 186–7 agriculture, 22 Aitken, Brian, 128, 220, 243n6 Akyuz, Yilmaz, 11, 21–57 American Textile Trade Action Coalition (ATTAC), 67 anti-dumping, 237 anti-globalization movement, 1, 15 apartheid, 14, 182, 183, 186, 187, 194 apparel (garment) industry, 12, 58–87 China, 67, 83n6, 175n15, 256 Eastern European industrial downgrading, 3 gender wage gap, 158, 161
labor standards, 253–6, 259, 264 Mexico, 12, 59, 65, 66, 68–81, 82–3, 253–6 South Africa, 194–6 South Korea, 158, 161 Taiwan, 158, 161, 175n15 see also maquiladoras; sweatshops; textile industry Argentina commodity structure of exports, 46, 47, 48 competitiveness of manufacturers, 43, 45 exports and manufacturing value added, 30, 31 industrial upgrading, 11, 37, 54 international specialization, 53 labor costs, 41 labor productivity, 34, 36 manufacturing, 26, 27, 28, 37, 38 revealed comparative advantage, 49–50, 51 Artecona, Raquel, 149 ASEAN see Association of South-East Asian Nations Asia crowding in effect of FDI, 230 electronics industry, 257 garment industry, 255 labor productivity, 33 large firms, 9 manufacturing, 25, 26 see also East Asia Asian financial crisis, 165, 168, 175n16, 209 assembly production, 12, 59, 63, 64, 68, 83n2 assets intangible, 8 knowledge, 122, 123, 124, 140 Association of South-East Asian Nations (ASEAN), 216 ATTAC see American Textile Trade Action Coalition
266
Index 267 Australia exports, 24 gender wage ratio, 153 Austria, gender wage gap, 148 automobile sector industrial upgrading, 48, 51, 53 MERCOSUR, 56n12 South Africa, 193–4 Bacha, Edward, 102n10 Baily, Martin Neil, 128 Bair, Jennifer, 12, 58–87 Balakrishnan, Radhika, 13, 104–18 balance of payments, 23, 238, 243n11 banking, 196–7 bankruptcies, 67, 71 Barbie Dolls, 6 bargaining power capital, 223 China, 15, 211–12, 238, 239, 241 multinationals, 210–11, 212, 238, 241 privileged groups, 150 women, 151 batch production, 95, 97, 110, 111 BEA see Bureau of Economic Analysis Becker, Gary, 14, 146, 147, 149, 158, 169 BEE see “black economic empowerment” Behrman, Jere, 149, 174n2 Berg, Janine, 12, 88–103 Berik, Günseli, 14, 146–78 Berman, Eli, 141n9 Beyer, Harald, 90–1 Black, Anthony, 14–15, 179–206 “black economic empowerment” (BEE), 179, 196, 201–3, 205n26 Black, Sandra, 147, 148, 149, 162 Blau, Francine, 176n19 Blecker, R., 2 Bolivia competitiveness of manufacturers, 43 exports and manufacturing value added, 30, 31 labor costs, 41, 240 labor productivity, 34 manufacturing, 27 sectoral investment, 45 Boltanski, Luc, 263, 264 Borjas, George, 147 brain drain, 135, 142n14
Brainard, S. L., 9 Brainerd, Elizabeth, 147, 149, 162 branded manufacturers, 61, 62, 71, 78 Braunstein, Elissa, 15, 209–48 Braverman, H., 107 Brazil commodity structure of exports, 46, 47, 48 competitiveness of manufacturers, 43, 45 exports and manufacturing value added, 30, 31 industrial upgrading, 11, 37, 54 international specialization, 53 labor costs, 41 labor productivity, 34, 36 manufacturing, 26–7, 28, 29, 37, 38 multinationals, 94 positive signs, 2 productivity growth, 29 revealed comparative advantage, 49–50, 51 sectoral investment, 45 US multinational employees, 126 wage inequality, 102n10 Budd, John W., 141n8 Bureau of Economic Analysis (BEA), 125–6, 136–7 Burlington Industries, 67, 71–2 Bush, George, 67 buyer-driven commodity chains, 6, 60–1, 62, 111, 194, 195 Calvin Klein, 67, 77 Canada apparel industry, 66, 83n4 exports, 24 GDP per worker, 137 imported inputs, 5 capital accumulation, 21, 29, 45, 53, 54, 122, 127 bargaining power, 223 China, 230 flows, 134–5, 181, 186, 209 investment, 1–2, 11, 125, 127, 128, 130, 134 mobility, 1, 39, 116n13, 151–2, 161, 170 South Africa, 187
268 Index capital flight, 161, 230, 243n11 capitalism, 10, 60, 106, 107, 251 Caribbean Basin, 12, 65, 68–70, 82, 255 Caves, Richard E., 124 Chandler, A., 8 Chen, Chunlai, 217–18 Chen, Fen-ling, 170 Chen, Tain-Jy, 155–6, 175n15 Chen, Yi-Ping, 155–6, 175n15 child labor, 16, 249–52, 259 Chile capital accumulation, 54 commodity structure of exports, 46, 47, 48 competitiveness of manufacturers, 42, 43 cosmetics industry, 12, 88–103 exports and manufacturing value added, 30, 31, 32, 52 GDP per worker, 137 home workers, 114 industrial upgrading, 11, 37 international specialization, 53 labor costs, 41, 240 labor productivity, 34, 36 manufacturing, 26, 27, 28, 37, 38 positive signs, 2 productivity growth, 29 revealed comparative advantage, 49–50, 51 sectoral investment, 45 China, People’s Republic of apparel industry, 67, 83n6, 175n15, 256 Barbie Doll production, 6 bargaining power, 15, 211–12, 238, 239, 241 basic rights, 262 commodity structure of exports, 47 competitiveness of manufacturers, 42, 43 exports and manufacturing value added, 30, 31–2, 52, 53, 54 foreign direct investment, 15, 209–48 gender wage gap, 148, 174n4 industrialization, 26, 54 labor costs, 41 labor productivity, 34 labor standards, 256, 257, 258, 260 manufacturing, 25, 26, 27
Open Door Policy, 213–14, 220 relocation of Taiwanese industry to, 161, 175n15 revealed comparative advantage, 49–50 trade integration, 37 Chiu, Su-fen, 170 Chung, Wilbur, 128 Ciudad Hidalgo, 16, 249–53, 257, 258, 260 clothing see apparel industry Coase, R., 106 codes of corporate conduct, 1, 78, 81, 210 see also moral codes collective bargaining, 98, 99 United States, 260–1 women, 151, 171 Colombia commodity structure of exports, 46, 47, 48 competitiveness of manufacturers, 43 exports and manufacturing value added, 30, 31, 52 GDP per worker, 137 labor costs, 41 labor productivity, 34, 36 manufacturing, 27, 38 sectoral investment, 45 slow progress, 54 commodities, 46, 47 commodity chains buyer-driven, 6, 60–1, 62, 111, 194, 195 North American apparel industry, 58, 59–65, 72, 73, 76, 79–80, 81–3 producer-driven, 6, 60–1, 193 South African auto sector, 193 Compañía Industrial de Parras, 76 comparative advantage export promotion, 2 Latin America, 27 productivity growth, 40 revealed, 48–51 competition bidding for investment, 232–5 Chile, 12, 88, 93, 94–5, 100, 101 developing countries, 9, 252 discrimination, 146–8, 149
Index 269 competition – continued foreign direct investment, 128, 209, 210, 229, 230 gender wage gap, 14, 146–7, 149–52, 155, 161, 162, 168, 169 international competitiveness, 39–45 outsourcing, 8, 9 price-based, 113 protection from foreign, 39 South Africa, 194, 196 subcontracting, 13, 113, 114 competitive advantage apparel industry, 65 labor costs, 10 competitiveness, 39–45 gender wage gap, 151, 152, 162–5, 168, 169 Taiwanese Industrial and Commercial Census, 174 computer industry, 256–7 computerization, 10 Cone Mills, 76 conventionalist school, 263, 264 corporate conduct, 1, 78, 81, 210 see also moral codes corruption, 212, 234, 235–6, 239, 241 cosmetics industry, 12, 88–103 Costa Rica exports and manufacturing value added, 52 Intel facilities in, 133 productivity growth, 29 costs global production sharing, 6–7 labor, 10, 39–40, 41, 42, 43–4 China, 240 foreign direct investment, 123 subcontracting, 13, 104, 106, 107–8, 111–12, 113–14 sweatshops, 259 trade, 123 transaction, 8, 104, 106–7, 199 Côte d’Ivoire commodity structure of exports, 46, 47 competitiveness of manufacturers, 42, 43 exports and manufacturing value added, 30, 31 labor costs, 41
labor productivity, 34 manufacturing, 26, 27 craftsmanship, 252–3 creative destruction, 32 cultural imperialism, 250, 252 Cunningham, Wendy, 149 currency depreciation, 42, 179, 181, 192, 194 Curry, B., 163 de-skilling, 107 debt crisis, 28, 33, 55, 93 decentralization China, 15, 211–12, 213, 239 US labor law, 260, 261 Dees, Stèphane, 219 deindustrialization, 23, 32 Chilean cosmetics industry, 101 Latin America, 11, 26, 28 negative/positive distinction, 23, 37 productivity growth in manufacturing, 24, 26, 28 US apparel industry, 73 demand Chinese labor demand, 223, 225 gender wage gap, 158 industrialization, 22, 23 skilled labor, 13, 121, 122–33, 140 Deng Xiaoping, 214, 215 Denmark, GDP per worker, 137 denunciation, 263, 264 deregulation, 112, 113, 262 gender wage gap, 148 labor market, 91 see also liberalization; privatization developing countries basic rights, 262 competition, 9 exporting clusters, 82 foreign direct investment, 128, 134–5, 210, 215–16 industrial upgrading, 21–57 industrialization, 24 labor costs, 40 labor standards, 252 low-skilled labor, 39 manufacturing value added, 51 productivity growth, 32, 33 skill upgrading, 13–14, 122, 136–40 structural challenges to workers, 1, 16
270 Index developing countries – continued subcontracting, 112, 114 wage inequality, 12, 88, 89 development globalization, 59–60 uneven, 83 Dicken, Peter, 60 discrimination gender wage gap, 14, 146–9, 150–3, 158, 162–3, 168, 169–71 racial, 182 see also apartheid diversification Chilean cosmetics industry, 95, 97, 101 economic development, 23–4 Dollar, David, 134 drug use, 254 Dunning, J. H., 123 “Dutch disease”, 55n2 East Asia apparel industry, 58–9 commodity structure of exports, 46–8 economic “miracle”, 2 export promotion, 2, 182 export-investment nexus, 29 exports and manufacturing value added, 31 gender wage gap, 146, 150–71 industrialization, 2, 11, 24, 26 international specialization, 53 investment in South Africa, 189 manufacturing, 11, 24, 25–6 productivity growth, 32 trade integration, 37, 220 transfer of production to China, 217 women in employment, 115n5 see also Newly Industrializing Economies Eastern Europe, industrial downgrading, 3 economic development China, 240 labor standards, 258 structural change, 22–32 see also growth Ecuador
competitiveness of manufacturers, 42, 43 exports and manufacturing value added, 30, 31 labor costs, 41 labor productivity, 34 manufacturing, 27 education Chile, 89, 90 China, 226 late capitalism, 251 multinationals’ influence on, 122, 133 South Africa, 183–4 efficiency wage effects, 111 Eggertson, T., 107 Egypt commodity structure of exports, 47 competitiveness of manufacturers, 43 exports and manufacturing value added, 30, 31 labor costs, 41 labor productivity, 34 manufacturing, 26, 27 El Paso, 73–4, 77, 82, 255 electronics industry, 48, 161–2, 256–7 employment Chilean cosmetics industry, 97–8 China, 226–9, 238, 239, 241 foreign direct investment impact on, 226–9, 237–8, 239 industrialization, 22–3 manufacturing, 25, 26 maquiladoras, 68 service sector, 23 South Africa, 181–5, 188 US apparel industry, 66 US multinational affiliates, 132 women’s employment sectors, 158–62 see also labor; labor markets; unemployment; wages endogenous asymmetry hypothesis, 9 entry barriers, 9, 61, 113, 196 entry mode, 190, 193 environmental regulation, 114, 117n32 Epstein, Gerald, 15, 209–48 Europe FDI stock in China, 216 investment in South Africa, 186, 189, 190
Index 271 Europe – continued South African clothing exports, 195 exchange rate international competitiveness, 40, 42, 43–4 South Africa, 179, 181 export processing zones, 3, 64, 116n26, 232 export promotion, 2, 182, 187 export-oriented production, 10, 58, 59 China, 211, 213, 240 gender wage gap, 151, 170, 171 Mexican apparel industry, 12, 74–5, 77, 79 South Africa, 194, 203 exports capital accumulation, 29 China, 212, 217, 229, 237–8, 239 export-investment nexus, 29, 45, 48 exporting clusters, 82 gender ratios, 159–60, 161 gender wage gap, 14, 151, 155, 165, 168, 169, 170–1 growth, 11–12 industrial upgrading, 64 international competitiveness, 42, 43–4 international specialization, 48–53 manufacturing value added, 29–32 North American apparel industry, 68, 70, 74, 79, 82 reliance on, 23–4 revealed comparative advantage, 48–51 South Africa, 180, 185, 187, 191, 192, 193–6 upgrading, 45–8 see also trade externalization, 8 Farah, 77 FDI see foreign direct investment Feenstra, R., 13, 112, 241 feminism, 150, 151 FIEs see foreign investment enterprises financial services China WTO accession terms, 236 South Africa, 181, 184, 185, 187–90, 196–7, 198, 200 Financial Times 500 list, 9
Finland, GDP per worker, 137 flexibility arms-length relations, 8 global production sharing, 6 labor, 6, 91, 104, 108–9 labor standards regulation, 262 flexible specialization, 110 Flyer, Frederick, 129 Ford, 5 foreign direct investment (FDI) bidding for, 232–5 China, 15, 209–48 crowding in/crowding out of domestic investment, 229–32, 239 East Asia, 2 gender wage gap, 151, 156, 170 globalization, 209, 210–11 horizontal, 123 increase in, 16 labor standards regulation, 259 Latin America, 11, 36 skill upgrading, 13–14, 121–45 South Africa, 14–15, 179–206 vertical, 123 world, 4 see also greenfield investment; investment; multinationals foreign investment enterprises (FIEs), 214, 216–18, 226, 229, 230, 232, 238 see also multinationals; transnational corporations formal sector mass production, 263 South Africa, 182, 183–4 subcontracting, 108–9, 111 Fortune 500 Companies, 9 France bureaucratic regulation, 261–2 investment in South Africa, 189 research & development, 123 free trade apparel industry, 67 Chilean cosmetics industry, 93–5, 102 export promotion, 2 wage inequality, 88, 89–91 Free Trade Zones, 68 “full-package” production industrial upgrading, 59, 63–4
272 Index “full-package” production – continued Mexican apparel industry, 12, 69, 70–2, 76–7, 78–9, 80, 82 furniture industry, 16, 249–53, 258 Gap, 6, 61, 76, 77 garment industry see apparel industry GATS (General Agreement on Trade in Services), 180 GATT (General Agreement on Tariffs and Trade), 180 GCC see global commodity chains framework GDP see gross domestic product Gelb, Stephen, 14–15, 179–206 gender feminization of the workforce, 105, 115n4 subcontracting, 104 wage inequalities, 14, 105, 146–78 see also women General Agreement on Tariffs and Trade (GATT), 180 General Agreement on Trade in Services (GATS), 180 George, K. D., 163 Gereffi, Gary, 6, 12, 58–87, 111 Germany investment in South Africa, 189, 194 research & development, 123 structural similarity indices, 52, 56n13 Ghana commodity structure of exports, 47 competitiveness of manufacturers, 42, 43 labor costs, 41 labor productivity, 34 manufacturing, 26, 27 Ghosh, Jayati, 114 Gibbon, P., 195 Gindling, Thomas, 90 Gini coefficient, 102n4 Glen, J., 9 global commodity chains (GCC) framework, 58, 60–3, 64, 65, 81–3 see also commodity chains global production sharing, 3–9, 10 globalization commodity chains framework, 81–2 and development, 59, 83
financial, 116n13 foreign direct investment, 209, 210–11 gender wage gap, 14, 147, 170 global production sharing, 3–9, 10 industrialization, 2–3 of production, 2, 3–9, 14–15, 16, 179, 191–3 relativization of scale, 60 transformative influence, 58 victims of, 257 wage inequality, 88 governance, 60, 61 government decentralization, 211–12 Greece, GDP per worker, 137 greenfield investment Chilean cosmetics industry, 93 location choices, 129 South Africa, 188, 190, 192, 193, 199, 203 spillovers, 128 gross domestic product (GDP) manufacturing output, 25, 26, 27 manufacturing value added, 29–32, 37, 54 per worker, 137 South Africa, 181, 182–3, 204n5 world, 4 Grown, Caren, 105 growth China, 212, 218–20, 239 East Asia, 11 exports, 11–12 foreign direct investment, 141, 210, 212, 218–20, 239 impediments to, 54 incomes, 134 Latin America, 28–9 productivity, 2, 21, 22–9, 32–7, 39–40, 219–20 South Africa, 181, 182–3, 185 see also economic development Grupo Lajat, 76 guanxi, 233 Guilford Mills, 67, 72 Gunter, Frank, 243n11 Haddad, M., 128, 220 Hanson, Gordon H., 13, 132, 133, 219–20 Harrison, Ann, 128, 220
Index 273 Harrison, B., 3 Haskel, Jonathan E., 128 Hayes, Chuck, 72 health and safety, 250, 253, 258, 259, 260 Hecksher-Ohlin theory, 55n7, 89, 101 Hellerstein, Judith, 148 Henderson, George, 83n7 high-tech products, 50, 51, 53 HIV prevalence, 183, 254 home-based work, 105, 112, 113, 114, 151, 259 Hong Kong apparel industry, 59, 65, 83n6 Barbie Doll production, 6 exports and manufacturing value added, 52 foreign direct investment, 215, 216, 218, 235 gender wage ratio, 153 industrialization, 2, 11 return to Chinese rule, 242n1 transfer of production to China, 217 Huang, Yasheng, 230, 232 human capital foreign direct investment, 13, 122, 133, 134, 136, 199–201 industrial upgrading, 64–5 subcontracting, 110 Hymer, S., 7–8 hypermarkets, 93 IBM, 256–7 ICTs see information and communication technologies ILO see International Labor Office IMF see International Monetary Fund immigration barriers, 151, 199 import substitution industrialization, 2, 55 Chile, 12, 88, 89, 93, 94 China, 213, 214, 236 East Asia, 182 Latin America, 25–6, 27–8, 32 Mexico, 81 imports Chilean cosmetics industry, 93 China, 214, 229, 238, 239 gender ratios, 159–60, 161 gender wage gap, 14, 165, 168, 169, 170
machinery, 36, 90, 91 restrictions on Chinese imports to United States, 237 South Africa, 180, 193, 194, 195 US apparel imports, 69 see also trade India commodity structure of exports, 46, 47 competitiveness of manufacturers, 43 exports and manufacturing value added, 30, 31, 52 GDP per worker, 137 industrialization, 54 IT industry, 110 labor costs, 40, 41 labor productivity, 34, 36 manufacturing, 27 positive signs, 2 US multinational employees, 126 Indonesia Barbie Doll production, 6 competitiveness of manufacturers, 43 exports and manufacturing value added, 29, 30, 31 industrialization, 11 labor costs, 41 labor productivity, 34, 36 manufacturing, 11, 27, 37 industrial districts, 82, 252–3, 260 industrial downgrading, 3, 83n2 industrial pollution, 254–5 industrial upgrading, 3, 21–57 apparel industry, 59, 65–72, 76–9 Chile, 89 exports, 45–8 “full-package” production, 63–4 Latin America, 11 Mexico, 16 value added, 10 industrialization, 1–2, 11 and competitiveness, 39–45 developing countries, 24 economic development, 22–32 productivity growth relationship, 21, 22–9 trade integration, 37 see also import substitution industrialization industry concentration, 149, 150, 155–65, 168, 169–70, 175n12
274 Index inequality South Africa, 183 wages Chilean cosmetics industry, 12, 88–91, 98–102 gender wage gap, 14, 105, 146–78 trade liberalization, 1, 10, 45, 88 inflation, 180 informal sector alternative systems of production, 263 subcontracting, 108–9, 111, 112, 115n3, 116n15 information and communication technologies (ICTs), 127, 130–3, 142n12 input-output structure, 60 institutional context, 60, 82, 108–9, 210–11 see also regulation Intel, 130, 133 intellectual property, 180, 236, 239 International Labor Office (ILO), 58, 263 International Monetary Fund (IMF), 210 International Standard Industrial Category (ISIC), 136, 174 investment bidding for, 232–5 crowding in/crowding out of domestic investment, 229–32, 239 export-investment nexus, 29, 45, 48 Latin America, 45 manufacturing value added and exports, 29–32 multinationals, 101, 130, 134 productivity growth, 32 skilled-labor demand, 125, 127, 128 South Africa, 180, 181, 186 total investment in fixed assets, 241–2 see also foreign direct investment; greenfield investment; multinationals Ireland foreign direct investment, 130, 136 GDP per worker, 137 ISIC see International Standard Industrial Category Iskander, Natasha, 254 Italy, industrial districts, 252, 253, 260
Japan Barbie Doll production, 6 FDI stock in China, 216 foreign direct investment, 128 GDP per worker, 137 gender wage ratio, 153 imported inputs, 5 intra-firm trade, 7 investment in South Africa, 204n7 research & development, 123 structural similarity indices, 52 JC Penney, 61, 62, 72, 77, 78 job losses Chilean cosmetics industry, 95 Chinese state sector, 227, 229, 237, 238 maquiladoras, 79 US apparel industry, 66, 67–8, 73, 74, 75 women, 152 see also unemployment Joint Ventures ( JVs), 129, 141, 190, 192–3, 203, 213 K-Mart, 77 Kahn, Lawrence, 176n19 Kaltex, 72 Kaplinsky, Raphael, 61 Kentucky-Lajat, 76–7 Kenya competitiveness of manufacturers, 42, 43 exports and manufacturing value added, 30, 31 labor costs, 41 labor productivity, 34 manufacturing, 27 King, Elizabeth, 149, 174n2 knowledge assets, 122, 123, 124, 140 base, 33 spillovers, 13, 124, 128–9, 132–3, 141n5 Korea, Republic of (South Korea) apparel industry, 59 commodity structure of exports, 46, 47, 48 competitiveness of manufacturers, 42, 43 deindustrialization, 26
Index 275 Korea, Republic of (South Korea) – continued exports and manufacturing value added, 29, 30, 31, 51–3 FDI stock in China, 216 GDP per worker, 137 gender wage gap, 14, 146, 147, 149, 150–71 ICT industry, 131 industrial upgrading, 11, 37, 46–8 industrialization, 2, 11, 53 international specialization, 53 labor costs, 40, 41 labor productivity, 35, 36 manufacturing, 27, 37, 38 revealed comparative advantage, 48–51, 53 Korzeniewicz, Roberto, 104 Kozul-Wright, Richard, 11, 21–57 Kray, Aart, 134 Krugman, Paul, 10, 125 Kucera, David, 109 labor Chilean cosmetics industry, 96 Chinese labor demand, 223, 225 costs, 10, 39–40, 41, 42, 43–4 China, 240 foreign direct investment, 123 subcontracting, 13, 104, 106, 107–8, 111–12, 113–14 sweatshops, 259 demand for skilled, 13, 121, 122–33, 140 flexibility, 6, 91, 104, 108–9 industrialization, 23 international/global division of, 2, 104 Mexico, 74 mobility, 1, 125 productivity, 33–7 China, 219 growth, 2, 21 international competitiveness, 42–5 South Africa, 182 subcontracting, 114 women, 111 supply of skilled, 122, 133–6, 140 turnover, 80–1, 124, 141n6, 188–9 wage elasticity of, 9
see also employment; low-skilled workers; skilled workers; wages labor markets China, 226 deregulation, 91 formal-informal divide, 108–9 Mexico, 80–1 multinationals’ influence on, 133, 134 segmentation, 11, 109, 114, 116n17, 181, 226 South Africa, 179, 181–5, 197–203 see also employment labor standards, 1, 8, 249–65 China, 256, 257, 258, 260 East Asia, 2 Korea, 171 Mexico, 16, 78, 249–54, 256–7, 258, 260 regulation, 258–9, 260, 261–2 see also working conditions labor unions Chilean cosmetics industry, 98–100, 101 North American apparel industry, 67 right to unionize, 108, 116n16 United States, 261 land subsidies, 233 Langlois, R., 8 Lardy, Nicholas, 230, 232, 243n11 Latin America commodity structure of exports, 46–8 crowding out effect of FDI, 230 deindustrialization, 11, 26, 28 electronics industry, 257 exports and manufacturing value added, 29–31, 53 FDI stock in China, 215, 216 impediments to growth, 54–5 import substitution industrialization, 25–6, 27–8, 32 international competitiveness, 45 large firms, 9 manufacturing, 24, 25, 26–9, 37, 45 natural resources, 28, 36 productivity, 33–6, 45 sectoral investment, 45 trade integration, 37 lead firms, 8, 61, 62–3, 82 learning, 21, 33, 64
276 Index Lee jeans, 62, 72, 74, 75 legislation anti-discrimination, 147, 148, 170–1 Chinese FDI laws, 213 living wage, 210 South Africa, 180, 194, 195 subcontracting, 108, 109, 114 see also institutional context; regulation Levi Strauss and Company, 62, 65, 74, 75, 77, 78 liberalization “big-bang”, 32 Chile, 12, 89, 90, 91, 100, 101 China, 213–14, 219, 220–2, 225–6, 227–9, 236, 237, 239 foreign direct investment, 220–1 South Africa, 180–1, 185 trade disintegration of production process, 104 gender wage gap, 149 labor market effects of, 10 Mexico, 81 price-based competition, 113 relative wages, 11 South Africa, 185, 194–5 trade integration, 37 wage inequality, 1, 45, 88, 90, 91, 101 see also deregulation; privatization living standards, 2, 21, 210 living wage, 210 Liz Claiborne, 61, 62 “low-level equilibrium trap”, 3, 11 low-skilled workers, 3, 10–11 capital mobility, 39 Chilean cosmetics industry, 12–13, 89–90, 91, 95–7, 98, 100–1 China, 225–6 classification of, 102n9 South Africa, 184, 185, 195 subcontracted work, 105 wage inequalities, 45, 88, 89–90 Lu, Ding, 219 machinery imports, 36, 90, 91 investments, 95, 97–8, 101 see also mechanization; technology
McKendrick, David G., 132 macroeconomic stability, 135 “make or buy” decision, 13, 105, 106–7, 108 Malaysia Barbie Doll production, 6 commodity structure of exports, 46, 47, 48 competitiveness of manufacturers, 42, 43 exports and manufacturing value added, 29, 30, 31, 52, 53, 54 gender wage ratio, 153 ICT industry, 131 industrial upgrading, 11, 37, 48 industrialization, 11, 53–4 international specialization, 53 labor costs, 40, 41 labor productivity, 34, 36 manufacturing, 11, 27, 37, 38 productivity growth, 40 revealed comparative advantage, 49–50, 51 US multinational employees, 126 Mansfield, Edwin, 124, 129 manufacturing Chile, 12 China, 218, 240 competitiveness, 42–5 East Asia, 11 gender wage gap, 154–5, 156, 169 industrialization, 23, 24–8 “low-level equilibrium trap”, 3, 11 original brand name manufacturing, 63, 64, 65 original equipment manufacturing, 63, 64, 65 revealed comparative advantage, 48–51 South Africa, 182, 184, 185, 186, 189 US multinational affiliates, 126, 127 value added, 9, 11, 26–7, 29–32, 37–8, 42, 51–3, 54 women’s employment, 158 “workshop economies”, 24 see also branded manufacturers; production maquiladoras, 13, 36, 68, 74–7, 79, 254, 255 see also apparel industry
Index 277 market access, 236, 239 market exchange, 106, 107 market failure, 8, 129 market orientation, 191–2, 203 marketers, 61, 62, 63, 71, 77, 78 Marshall, Alfred, 124, 252 Martin, William, 104 Marx, Karl, 107 Marxism, 106, 107 Mayer, Jörg, 2, 11, 21–57 Mayer, Ricardo, 230 mechanization, 10, 16, 22 see also machinery Meng, Xin, 148, 174n4 MERCOSUR (Southern Common Market), 56n12 mergers and acquisitions, 9, 190, 193, 199, 203 Mexico apparel industry, 12, 59, 65, 66, 68–81, 82–3, 253–6 commodity structure of exports, 46, 47, 48 competitiveness of manufacturers, 42, 43 computer industry, 256–7 exports and manufacturing value added, 30, 31, 52, 53 foreign direct investment, 243n6 GDP per worker, 137 gender wage gap, 149 home workers, 114 ICT industry, 131 industrial upgrading, 11, 16, 37 international production networks, 54 international specialization, 53 labor costs, 40, 41, 240 labor productivity, 34, 36 labor standards, 16, 78, 249–54, 256–7, 258, 260 manufacturing, 11, 26–7, 28, 37, 38 multinationals, 94 outsourcing, 13 productivity growth, 40 revealed comparative advantage, 49–50, 51 subcontracting, 12, 80, 253, 259
US multinational employees, 126 Microsoft, 130 migrant workers, 73, 105, 183, 254, 255, 256 migration reduced barriers to, 1 reverse, 136 Milberg, William, 1–18, 109 Millikin, Roger, 67 mineral resources, 182, 185–6, 187 minimum wage, 108, 109, 114 Moore, Carlos, 67 moral codes, 11, 262–3, 264 Moran, T. H., 129–30 Morocco commodity structure of exports, 46, 47 competitiveness of manufacturers, 43 exports and manufacturing value added, 30, 31 foreign direct investment, 128 labor costs, 41 labor productivity, 35 manufacturing, 27, 220 Mousiolek, B., 3 Multi-Fiber Agreement, 61, 195, 240 multinationals bargaining power, 210–11, 212, 238, 241 Chile, 12, 88–9, 92, 93–4, 95–7, 98, 100, 101 China, 212, 238 productivity growth, 219–20 skill upgrading, 13–14, 121–45 small number from developing countries, 9 South Africa, 189, 203 subcontracting, 115 wage inequality, 88 see also affiliates; foreign direct investment; transnational corporations NAFTA see North American Free Trade Agreement national treatment, 236, 238 Naughton, Barry, 221 neo-Marxism, 106, 107
278 Index neoclassical economics, 106, 107, 116n10 gender wage gap, 147–50 labor standards, 257 neoliberalism Chile, 98 foreign direct investment, 210, 212 Mexico, 81 subcontracting, 112 networks apparel industry, 62–3, 65, 70–1, 74, 77, 79–80, 82 commodity chain framework, 60–1 “full-package” production, 63–4, 79 global production, 33, 54, 58, 59, 60, 179 global service, 196 institutional context, 60, 82 inter-firm, 60, 70, 74 Neumark, David, 148 New Zealand, gender wage ratio, 153 Newly Industrializing Economies (NIEs), 22, 53–4 apparel industry, 58–9 export-investment nexus, 29 FDI stock in China, 216, 217 international specialization, 53 manufacturing, 25–6 see also East Asia niche markets, 252–3 NIEs see Newly Industrializing Economies Nigeria competitiveness of manufacturers, 42, 43 exports and manufacturing value added, 30, 31 labor costs, 41 labor productivity, 35 manufacturing, 27 Nike, 5–6, 17n9, 61, 62 North America apparel industry, 12, 65–82 investment in South Africa, 189 see also Canada; Mexico; United States North American Free Trade Agreement (NAFTA), 12, 65–70, 72, 73–7, 79, 249 NuStart, 72
Oaxaca, Ronald, 174n3, 176n19 OBM see original brand name manufacturing OECD see Organization for Economic Co-operation and Development OEM see original equipment manufacturing oligopoly, 9, 10 OMJC see Original Mexican Jean Company Ong Aihwa, 104, 115n3 openness China, 219, 220, 221–2 East Asian “miracle”, 2 foreign direct investment, 210 gender wage gap, 149, 152, 162, 165, 169, 170 Organization for Economic Cooperation and Development (OECD), 109 original brand name manufacturing (OBM), 63, 64, 65 original equipment manufacturing (OEM), 63, 64, 65 Original Mexican Jean Company (OMJC), 72 outsourcing arms-length, 7–8, 10 Chilean cosmetics industry, 95–7, 101 competition, 8, 9 cost reduction, 113 institutional context, 109 international, 3–5, 6–7 Mexican apparel industry, 75 skill intensity, 112 South Africa, 189 see also subcontracting Pafer-Huichita, 77 Pakistan competitiveness of manufacturers, 43 exports and manufacturing value added, 30, 31 labor costs, 41 labor productivity, 35, 36 manufacturing, 27 Palpacuer, Florence, 82 Panama, GDP per worker, 137 Parisotto, Aurelio, 82 Paus, Eva A., 243n6
Index 279 People’s Republic of China see China, People’s Republic of Peru commodity structure of exports, 48 competitiveness of manufacturers, 42, 43, 45 exports and manufacturing value added, 30, 31 labor costs, 41 labor productivity, 35 manufacturing, 27 sectoral investment, 45 Philippines Barbie Doll production, 6 competitiveness of manufacturers, 43 exports and manufacturing value added, 30, 31, 52, 53 home workers, 114 ICT industry, 131 international production networks, 54 labor costs, 40, 41 labor productivity, 35, 36 manufacturing, 11, 27 productivity growth, 40 US multinational employees, 126 piece rate system, 259 Piore, Michael, 16, 110, 249–65 political issues China, 239 labor standards, 262 South Africa, 186 pollution, 254–5, 256 portfolio investment, 180, 181, 186, 209 poverty, 183 price-based competition, 113 primary sector industrialization, 22, 23 low productivity, 21 South Africa, 184, 188, 198, 200 terms-of-trade movements, 46 privatization Mexico, 81 South Africa, 180, 196 see also deregulation producer-driven commodity chains, 6, 60–1, 193 product promotion, 98 production assembly, 12, 59, 63, 64, 68, 83n2 batch, 95, 97, 110, 111
Chilean cosmetics industry, 94, 95, 97–8 commodity chains, 60–1 disintegration of, 104, 105, 107 export-oriented, 10, 58, 59 China, 211, 213, 240 gender wage gap, 151, 170, 171 Mexican apparel industry, 12, 74–5, 77, 79 South Africa, 194, 203 “full-package” industrial upgrading, 59, 63–4 Mexican apparel industry, 12, 69, 70–2, 76–7, 78–9, 80, 82 global production sharing, 3–9, 10 labor standards, 16, 250, 251–2, 258, 260 maquiladoras, 68 minimum efficient scale of, 110 moral codes, 262, 263, 264 radical changes, 262 South Africa, 14, 15, 179, 191–3 vertical integration, 8, 33 see also manufacturing productivity Chilean cosmetics industry, 97–8 China, 218–20 enhancement, 105, 112 export growth, 12, 45–6 gender wage gap, 147, 163, 175n8 growth, 21, 22–9, 32–7, 39–40, 219–20 international competitiveness, 42–5 labor, 33–7 China, 219 growth, 2, 21 international competitiveness, 42–5 South Africa, 182 subcontracting, 114 women, 111 spillovers, 124, 127–8, 187, 219, 220, 223 subcontracting, 109 total factor, 124, 128, 219, 220 profit maximization discrimination, 147–8 subcontracting, 106, 107–9 Puerto Rico, garment industry, 255 purchasing power parity, 42 Putterman, Luis, 106
280 Index quotas, 70, 236, 240 R&D (research & development), 6, 36, 123, 125, 127, 190, 201 race to the bottom, 210, 212 racial discrimination, 182 see also apartheid Ramey, Valerie, 147 Raynor, Bruce, 67 RCA see revealed comparative advantage Reagan, Ronald, 67 real estate, 218 Reebok, 61, 62 regionalism, 59–60 regionalization, 94, 95, 98, 101 regulation China, 211, 212 collective bargaining, 260–1 environmental, 114, 117n32 foreign direct investment, 210, 211 France, 261–2 labor standards, 258–9, 260 mass production systems, 263 South African banking sector, 196, 197 subcontracting, 104, 109, 112, 114, 115n2 see also deregulation; institutional context; legislation Reisen, Helmut, 135 research & development (R&D), 6, 36, 123, 125, 127, 190, 201 resources capital accumulation, 54 efficiency of use, 21, 39, 106 industrialization, 23 Latin America, 28, 36 South African mineral resources, 182, 185–6, 187 retailers buyer-driven commodity chains, 6, 61 Chilean cosmetics industry, 93 decentralized production networks, 111 “full-package” production, 63 North American apparel industry, 61, 62, 71, 77, 78 South African garment industry, 194, 195 retraining programs, 68
revealed comparative advantage (RCA), 48–51 Riker, D., 9 risk, 189, 203 Robbins, Donald, 90 Roberts, M., 9 Robinson, Michael, 243n6 Rodgers, William, 149 Rodriguez-Clare, Andres, 124, 146–78 Romeo, Anthony, 124, 129 Rosen, Daniel H., 235 rural areas, 80–1 Sabel, C., 110 Salinas, Carlos, 81 Sara Lee, 62, 65 Sayeed, Asad, 13, 104–18 SBTC see skill-biased technological change scale economies, 6, 21, 94, 110 Scandinavia exports, 24 natural resources, 28 Sears Roebuck, 61 Seguino, Stephanie, 105, 151, 152, 170 service sector foreign direct investment, 196 industrialization, 22–3 South Africa, 184, 186, 190 see also financial services sex discrimination, 14, 146–9, 150–3, 158, 162–3, 168, 169–71 SEZs see special economic zones Shan, Jordan, 219 Shaver, J. Myles, 129 SIC see Standard Industrial Classification Singapore exports and manufacturing value added, 52, 53 FDI stock in China, 216 gender wage ratio, 153 labor costs, 240 US multinational employees, 126 skill-biased technological change (SBTC), 90–1, 101, 124, 126–7, 138, 142n11 “skill-enhancing trade”, 10–11 skilled workers Chilean cosmetics industry, 89, 95–7, 98, 100, 101, 102
Index 281 skilled workers – continued China, 226 classification of, 102n9 demand for, 13, 121, 122–33, 140 innovations in the division of labor, 110 South Africa, 179, 184, 185, 197–9, 202, 203 subcontracting, 110, 112 supply of, 122, 133–6, 140 wage inequalities, 45, 88, 89–90 Slaughter, Matthew, 13–14, 121–45 social capital, 64 social protection, 210 SOEs see State Owned Enterprises Solow, Robert M., 128 Song, J., 124–5 Soto, Marcelo, 135 South Africa, foreign direct investment, 14–15, 179–206 South Korea see Korea, Republic of Southern Common Market (MERCOSUR), 56n12 Southern Cone countries, 24, 25, 26 Spain, GDP per worker, 137 special economic zones (SEZs), 213, 220, 232, 233 specialization, 2, 21, 39, 48–53 flexible, 110 vertical, 8 Spener, David, 74 spillovers knowledge, 13, 124, 128–9, 132–3, 141n5 labor market, 124–5 productivity, 124, 127–8, 187, 219, 220, 223 technological, 13, 101, 121–2 Sri Lanka, gender wage ratio, 153 Standard Industrial Classification (SIC), 142n12, 171, 172 standards see labor standards Standing, Guy, 112 State Owned Enterprises (SOEs), 237, 238 state sector, 221, 223, 225–6, 227–9, 237–8 Stolper-Samuelson theory, 10, 90 Strahan, Philip, 148 structural adjustment policies, 112, 113
Sturgeon, Timothy J., 63 subcontracting, 6, 7–8, 104–18 assembly model, 12, 59, 63, 64 branded manufacturers, 62 Chilean cosmetics industry, 95, 97 “full-package” model, 59, 63–4 labor standards regulation, 259 Mexico, 12, 80, 253, 259 “push/pull” explanation, 13, 106, 109–14 see also outsourcing subsidiaries, 94, 95–7 subsidies, 2, 194, 233, 234 Sun Apparel, 75, 77 Sun, Haishun, 230 supermarkets, 93, 95 supply chain, 70–1 supply of skilled labor, 122, 133–6, 140 sweatshops, 3, 10, 78, 81, 105, 114, 259–60 Sweden GDP per worker, 137 intra-firm trade, 7 Taiwan Barbie Doll production, 6 commodity structure of exports, 46, 47, 48 competitiveness of manufacturers, 42, 44 deindustrialization, 26 exports and manufacturing value added, 29, 30, 31, 52 FDI stock in China, 216 gender wage gap, 14, 146, 147, 149, 150–71 industrial upgrading, 11, 46–8 industrialization, 2, 11, 53 international specialization, 53 labor costs, 40, 41 labor productivity, 35, 36 manufacturing, 27, 37, 38 productivity growth, 40 revealed comparative advantage, 48–51, 53 transfer of production to China, 217 US multinational employees, 126 tariffs, 70, 83n7, 93, 113 China, 214, 236, 240 South Africa, 180, 193
282 Index tariffs – continued tax incentives for foreign direct investment, 213, 214, 232 revenue, 15, 232–3, 234–5, 238 Taylor, Lance, 102n10 technological progress, 2, 32–3 technology Chile, 28, 89, 91, 95, 101 China, 214 foreign direct investment, 121, 134, 190 gender wage gap, 152 high-tech products, 50, 51, 53 intensity, 46, 48, 218 machinery investments, 95, 97–8, 101 productivity growth, 32–3 skill-biased technological change, 90–1, 101, 124, 126–7, 138, 142n11 subcontracting, 110 technology transfer China, 239 skill upgrading in developing countries, 121, 122, 124, 125, 126–7, 128 South Africa, 187 vertical integration, 33 Tehuacan, 254–6, 257, 258, 260 telecommunications, 236 terms-of-trade, 23, 46 textile industry gender wage gap, 161 New England, 256 North America, 66–7, 71, 72, 73, 76, 78, 79 restrictions on Chinese imports to United States, 237 South Korea, 158, 161 Taiwan, 158, 161, 175n15 US international relations policy, 83n7 see also apparel industry TFP see total factor productivity Thailand competitiveness of manufacturers, 44 exports and manufacturing value added, 29, 30, 31 gender wage ratio, 153 home workers, 114
industrialization, 11, 54 labor costs, 40, 41 labor productivity, 35 manufacturing, 11, 27, 37 The Limited, 61 Thévenot, Laurent, 263 TNCs see transnational corporations Tommy Hilfiger, 76, 77, 78 Torreon, 72, 73, 74–81 total factor productivity (TFP), 124, 128, 219, 220 trade China, 217, 220–1, 226, 227, 240, 242 costs, 123 evolution of economic structure, 22 export-investment nexus, 29 gender wage gap, 14, 146–7, 149, 151, 155–6, 158–65, 168, 169–70 global production sharing, 3 government revenue, 234–5 integration, 37, 104, 220 intra-firm, 7–8 labor standards, 249, 252 liberalization disintegration of production process, 104 gender wage gap, 149 labor market effects of, 10 Mexico, 81 price-based competition, 113 relative wages, 11 South Africa, 185, 194–5 trade integration, 37 wage inequality, 1, 45, 88, 90, 91, 101 North American apparel industry, 12, 69, 74 skill-enhancing, 10–11 South Africa, 180, 185 terms of, 23, 46 textile industry, 67 world, 4 see also export-oriented production; exports; free trade; imports; tariffs Trade Related Aspects of Intellectual Property Rights (TRIPS) agreement, 180 trade unions see labor unions training, 68, 122, 133–4, 140, 199–201 transaction costs, 8, 104, 106–7, 199
Index 283 transnational corporations (TNCs) automobile sector, 48 China, 15 expansion of activities, 7 intra-firm trade, 7–8 producer-driven value chains, 6 technology transfer, 33 wage elasticity of labor, 9 see also multinationals TRIMS (Agreement on Trade Related Investment Measures), 180, 186–7 TRIPS (Agreement on Trade Related Aspects of Intellectual Property Rights), 180 Troske, Kenneth, 148 Turkey commodity structure of exports, 46, 47 competitiveness of manufacturers, 44 exports and manufacturing value added, 30, 31, 52 industrial upgrading, 11, 37 industrialization, 54 labor costs, 40, 41 labor productivity, 35, 36 manufacturing, 27, 37 positive signs, 2 Tybout, J., 9 UK see United Kingdom UNCTAD (United Nations Conference on Trade and Development), 29, 121, 187, 237 unemployment China, 227 industrialization, 23 low-skill workers, 13 push into subcontracting, 112 South Africa, 14, 15, 179, 181–5 see also job losses UNIDO (United Nations Industrial Development Organization), 36, 174 Union of Needletrades, Industrial and Textile Employees (UNITE), 67 United Kingdom (UK) GDP per worker, 137 imported inputs, 5 investment in South Africa, 185–6, 189
productivity spillovers, 128 research & development, 123 United Nations Conference on Trade and Development (UNCTAD), 29, 121, 187, 237 United Nations General Industrial Data Base, 136–7 United Nations Industrial Development Organization (UNIDO), 36, 174 United States (US) apparel industry, 65–8, 79, 82, 83n5 Barbie Doll production, 6 China WTO accession, 236, 237 clothing market, 253 collective bargaining, 260–1 electronics industry, 257 FDI stock in China, 216 foreign direct investment, 13, 243n6 gender wage gap, 149, 153 ICT products, 130–1, 132, 133, 142n12 imported inputs, 5 international outsourcing, 4, 5 intra-firm trade, 7 labor costs, 40 labor productivity, 33, 35, 36 labor standards regulation, 259 late capitalism, 251 multinationals, 13, 124, 125–6, 127, 130–3, 136–40 research & development, 123 South African clothing exports, 195 structural similarity indices, 52, 56n13 textile industry and international relations, 83n7 Treasury, 210 unskilled workers see low-skilled workers Uruguay competitiveness of manufacturers, 42, 44 exports and manufacturing value added, 30, 31 labor costs, 41 labor productivity, 35 manufacturing, 27 US see United States value added China, 238, 239
284 Index value added – continued foreign affiliates, 132 global value chains, 10 industrial upgrading, 21 lead firms, 62 low, 6, 9, 32, 54, 239 manufacturing, 9, 11, 26–7, 29–32, 37–8, 42, 51–3, 54 North American apparel industry, 70 South Africa, 185 value chains, 6, 9, 10 van der Meulen Rodgers, Yana, 14, 146–78 Venezuela competitiveness of manufacturers, 44 exports and manufacturing value added, 30, 31, 52 foreign direct investment, 128, 243n6 GDP per worker, 137 home workers, 114 labor costs, 41 labor productivity, 35 manufacturing, 27, 220 vertical integration, 8, 33 VF Corporation, 62, 72, 75, 77 Vietnam, foreign direct investment, 240 wages Chile, 13, 88, 89–91, 96, 97, 98–100, 101–2 China, 223–6, 238, 239, 240, 241 efficiency wage effects, 111 foreign direct investment impact on, 210, 222–6, 238, 239, 243n6 free trade, 88, 89–91 gap between skilled and unskilled workers, 39 gender wage gap, 14, 105, 146–78 inequality Chilean cosmetics industry, 12, 88–91, 98–102 gender wage gap, 14, 105, 146–78 trade liberalization, 1, 10, 45, 88 international competitiveness, 42–5 labor market segmentation, 114 living wage, 210 low, 105 Mexican apparel industry, 80 minimum wage, 108, 109, 114
multinationals, 125 productivity growth, 39–40, 42 South Africa, 185, 205n20 subcontracting, 108, 109, 110–11, 112 wage elasticity of labor, 9 see also labor, costs Wal-Mart, 6, 61, 111 Washington Consensus, 55 water pollution, 254–5 Weber, Max, 251 Wei, Shang-Jin, 219, 235 welfare optimization, 2 Winter-Ebmer, Rudolf, 148 women childcare, 250, 251, 254 Chinese garment industry, 256 feminization of the workforce, 105, 115n4 HIV prevalence, 183 home-based work, 105, 114 Indian IT industry, 110 labor flexibility, 109 maternity benefits, 116n14 Mexican garment industry, 253, 254 South African unemployment, 182 subcontracted work, 105, 111, 114 wage inequalities, 14, 146–78 wood-working industry, 16, 249–53 working conditions, 78, 105, 112, 249, 259 see also labor standards “workshop economies”, 24 World Bank, 135, 174n4, 187, 210 World Trade Organization (WTO) China accession to, 209, 212, 214, 215, 218, 236–8, 239–40 South Africa, 180 trade liberalization goal, 1 Wrangler, 62, 74, 75 WTO see World Trade Organization Yeh, Anthony Gar-on, 233 Zhao, Yaohui, 226 Zhu, Gangti, 219 Zimbabwe, trade with South Africa, 192 Zveglich, Joseph E. Jr, 14, 146–78