Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Globalization and the decline of the welfare state in less developed countries
(USC Thesis Other)
Globalization and the decline of the welfare state in less developed countries
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
INFORMATION TO USERS This manuscript has been reproduced from the microfilm master. U M I films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer. The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send U M I a complete manuscript and there are missing pages, these w ill be noted. Also, if unauthorized copyright material had to be removed, a note w ill indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher qualify 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact U M I directly to order. Bell & Howell Information and Learning 300 North Zeeb Road, Ann Arbor, M l 48106-1346 USA 800-521-0600 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. R eproduced with permission of the copyright owner. Further reproduction prohibited without permission. NOTE TO USERS Page(s) not included in the original manuscript are unavailable from the author or university. The manuscript was microfilmed as received. 313 This reproduction is the best copy available. um t Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. R eproduced with permission of the copyright owner. Further reproduction prohibited without permission. GLOBALIZATION AND THE DECLINE OF THE WELFARE STATE IN LESS DEVELOPED COUNTRIES by Nita Rudra A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (Political Economy and Public Policy) August 2000 Copyright 2000 Nita Rudra Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI Number. 3041514 Copyright 2000 by Rudra, Nita A ll rights reserved. U M I* UM I Microform 3041514 Copyright 2002 by ProQuest Information and Learning Company. A ll rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. ProQuest Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, M l 48106-1346 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UNIVERSITY OF SOUTHERN CALIFORNIA THE GRADUATE SCHOOL UNIVERSITY PARK LOS ANGELES, CALIFORNIA 90007 This dissertation, written by .................N ita^ .R u d ra................................................ under the direction of h..*?.. Dissertation Committee, and approved by all its members, has been presented to and accepted by The Graduate School, in partial fulfillment of re quirements for the degree of DOCTOR OF PHILOSOPHY Dean of Graduate Studies _ December 18, 2000 D ate.................................... DISSERTATION COMMITTEE p L l O & M : Chairperson S D L d d s u ........................................ Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DEDICATION To my loving Ma and Bapi, Lina and Sujit Rudra Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ACKNOWLEDGMENTS I am enormously indebted to many people for helping me complete this dissertation. To each and every one of them, I would like to give a special note of thanks. First, I would like to thank Professor Hayward Alker, the chair of my dissertation committee. I am a great admirer of his mathematical genius, his astuteness in multiple fields and, very importantly, his kind heart. Not only did he render supreme guidance through the dissertation process, but he also warmed the human spirit with his extraordinary enthusiasm for intellectual challenges. His excitement for building knowledge and the insatiable curiosity of his mind were inspirational and had a tremendous impact on my own work. I am most grateful for his unique ability to strongly support me through the struggles of my dissertation and, at the same time, constantly challenge me to reach beyond my perceived limits. I am deeply indebted to John Odell who has been my mentor since the day I started the doctoral program at the University of Southern California. I have the utmost respect for John Odell. It is his distinction as both an educator and a researcher that has given me the willingness to withstand the pressures of both graduate school and a potential career in academia. His high standards of excellence, his sincere concern for his students, and his intellectual superiority has made a marked impression on both my work and career aspirations. He provided me with expert guidance and focused attention which went above and beyond his responsibilities as my committee member. Quite simply, this dissertation may not have been completed without his influence and leadership throughout the years. iii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. I am also extremely grateful to Gary Dymski who has made the whole process of my dissertation program a much more pleasant experience. It was his faith confidence in me that encouraged me to pursue my PhD. I will never forget his unfailing support and encouragement throughout my doctoral studies. He has been a wonderful friend and has provided excellent guidance in many aspects. My thanks also goes to Nora Hamilton and John Elliott. Their kind support and constructive criticism were invaluable. Nora Hamilton’s comments on the case studies during my arduous struggle to understand labor power in developing countries was especially helpful. I am also grateful to Saori Katada for her help with methodology, technical and academic advice. She has been very generous with her time and assistance with the dissertation and other related issues. Dissertations cannot be completed without strong external support. I am indebted to my friends without whom I could not have finished this dissertation. I have been extremely fortunate to have by my side, my good friend and colleague, Samira Salem. Samira has been an incredible asset to me in so many ways. She has read through endless revisions of my dissertation chapters, always given me her full support and encouragement, and never hesitated to engage in a dialogue with me about my ideas. I am very grateful for the emotional and intellectual strength that her friendship has provided me through the years. Cassandra Thomas with her keen sense of logic has been remarkably helpful. Her insightful comments were instrumental in framing some of the primary arguments in Chapter 2. Gitika iv Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Commuri, who suffered with me and laughed with me, always brought in a breath of fresh air. I cannot forget to recognize Lisa Snead who, though far away, always lent me her support, faith and understanding. My thanks similarly goes to Anna Ruth Worten-Fritz who always reminded me of the importance of having a good attitude. I also extend my appreciation to Ujjal Dutt for his constant reassurance and words of encouragement. Last but not least, my brothers, Krish Sundaram and Mithun Dutt, have been the strongest supporters of both me and my aspirations. I must give them the deepest thanks for having the wherewithal to withstand my dissertation stress. Finally, without hesitation, my love and my life, Ravi Sundaram, deserves the most credit for the completion of this dissertation. He never tired from encouraging me and pushing me forward. Alongside his own work, he spent endless hours perfecting the numerous graphs, charts and figures in this dissertation. I have learned so many things from him. But, most importantly, he has given me the confidence to pursue my dreams. Above all, my parents to whom this dissertation is dedicated deserve the ultimate recognition. It is my parent’s deep passion for knowledge that has always inspired me. And it will always be my parent’s wishes that I hope to fulfill. This journey would not have been possible without their unshakable love and faith in me. They have given me the world. I can never thank them enough for that. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE OF CONTENTS CHAPTER PAGE DEDICATION u ACKNOWLEDGMENTS iii LIST OF TABLES x LIST OF CHARTS xi LIST OF FIGURES xiii ABSTRACT xiv SECTION I: 1 INTRODUCTION 2 Theoretical Context and Alternative Hypotheses Thesis, Methodology and Research Design 9 21 2 THE LINK BETWEEN GLOBALIZATION AND WELFARE STATES IN LESS DEVELOPED COUNTRIES 25 The Economic Effects of International Markets on Labor The Political Effects of International Markets on Labor The LDC Model The Variables Independent Variable: Economic Globalization in LDCs [TRADE and KFLOW] Independent Variable: Factor Endowments and The Bargaining Power of Labor [UNSKILL And SURP[ (the structural variable) 27 32 39 42 47 49 V I Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER PAGE Control Variables 53 The Results 56 The Closed Economy Model of LDCs 56 The Open Economy Model of LDCs 59 The Comparison Between LDCs and OECDs 64 Implications 68 3 CASE STUDIES: GLOBALIZATION AND SOCIAL WELFARE IN INDIA 71 The Qualitative Methodology 77 The Dependent Variable: Government Social Welfare Commitment 81 The Indian Case 84 India’s Government Social Commitment 84 The Characteristics of India’s Social Welfare System 88 India’s Labor Market Policies 95 Independent Variable: Globalization in India 98 Trade Flows-Background 102 Independent Variable(s): Globalization and the Bargaining Power of Labor 116 Background 117 Lack of Uniform Labor Code 122 State Dominance Intripartite Bargaining 123 Shift in Demand for Informal Unskilled Labor 124 Growth in Demand for Formal and Informal Unskilled Labor 125 The Growth of the Informal Sector 134 Effects on Labor Power 141 Textile Sector 147 Conclusion 153 4 INDIA’S WELFARE STATE IN COMPARISON PERSPECTIVE: BRAZIL 155 Dependent Variable: Brazil’s Social Welfare Commitment Trends and Levels of Brazil’s Social Welfare Spending The Characteristics of Brazil’s Social Welfare System Brazil’s Labor Market Policies vii 156 157 161 164 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER PAGE The Independent Variables: Globalization and Bargaining Power ofLabor 167 Globalization in Brazil 169 Globalization and the Bargaining Power of Labor in Brazil 175 5 KOREA AND CASE STUDY SUMMARY 187 The Dependent Variable: Government Social Welfare Commitment in Korea 188 Trends and Levels of Korea’s Social Welfare Spending 188 The Character of Korea’s Social Welfare System 191 Korea’s Labor Market Policies 194 The Independent Variables: Globalization and the Bargaining Power of Labor 196 Globalization 197 Globalization and the Bargaining Power of Labor in Korea 203 Summary 211 SECTION 2: 6 REASSESSING THE RELATIONSHIP BETWEEN GLOBALIZATION AND WELFARE: WELFARE 216 Conceptual Concept 219 The Empirical Model 233 The Variables 238 Data Limitations and Notes on Methodology 247 Discussion of Results 249 Implications 264 SECTION 3: 7 THE GLOBALIZATION-LDC WELFARE STATE NEXUS AND ITS IMPLICATIONS 268 viii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER PAGE Contribution to Literature 281 Policy Implications and Future Research Agenda 283 BIBLIOGRAPHY 287 APPENDICES 309 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABSTRACT This dissertation assesses the impact of economic globalization on welfare states in less developed countries (LDCs). Using panel data for 53 LDCs from 1972- 1995, it is revealed that trade and capital flows adversely affect LDC government social spending. I argue that the collective action problems of labor, caused by a large population of unskilled and surplus laborers, offset their potential political gains from globalization. My findings show that if a nation has a comparative advantage in unskilled labor, then globalization will lead to a decline in welfare expenditures. In order to gather further support for my hypothesis, I explore the cases of India, Brazil and Korea in greater detail. These findings provide sufficient justification for scholars to reassess their recent claim that welfare states have been resilient to the pressures of globalization. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. SECTION I Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER I Introduction Are citizens of all nations protected from the risks and uncertainties of economic globalization? Indeed, a recent theme in scholarly literature has been to emphasize the resilience of the welfare state in developed countries, despite the pressures of globalization (Alesina & Perotti 1997; Garrett, 1998a,b; Immerfall, 1998; Lindert, 1996; Pierson, 1996; Rodrik, 1997, 1998). This is quite a departure from studies of the previous two decades that contemplated the inevitable doom of the contemporary welfare state (Brown, 1988; Morris, 1988; Sherer, 1987). Yet neither current nor previous systematic comparative research has determined the effects of economic globalization upon government welfare expenditures in less developed countries (LDCs). This study finds that the assumptions implicit in analyses of welfare states in developed countries, particularly in the era of globalization, are not relevant to LDCs. Rather, a new model must be constructed to incorporate certain structural factors specific to LDCs, exposing a link between international (export and capital) markets, labor power and the LDC welfare state. Such a model will explain why the pursuit of comparative advantage has a strong impact upon the political power of labor and welfare states in LDCs, while its political effect within developed countries might be negligible. This study ultimately extends beyond existing research by analyzing how 2 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. international markets, alongside domestic economic and political variables, influence the rate of government welfare expenditures in LDCs. The term “globalization” has been discussed broadly in the political economy literature as: (I) a process, i.e. linked with technological change and information technology (Strange, 1996, 1997) (2) both a process and structural change, i.e., changing state structures (Cemy, 1995), international ideological hegemony (Gill, 1995; Pasha, 1996), and a supraterritoriality emerging from cross-national economic . flows, ideas and people (Ohmae, 1993; Scholte, 1997; Tonelson, 1997). This study, however, focuses on the depth and scope of international markets in the 23-year period from 1972-1995.’ The extent of “economic globalization” of nations is operationalized by the level of international trade (exports and imports) and gross capital flows as a share of GDP. Including only a small sample of politically and economically developed countries in conventional analyses has created bias in studies of the sustainability of welfare spending (i.e., social security and welfare) in this era of economic globalization.2 The purpose of this research is to expand the database to include LDCs and provide the first 1 From a broad perspective, Eichengreen (1996) best explains that there are four phases of (economic) globalization: (1) Before WWI, when controls on international financial transactions were absent, and international capital flows were high. (2) The inter-war period and the widespread imposition of capital controls. (3) Post WWII, gradual relaxation of capital controls, while trade flows continued to rise. (3) The post Bretton Woods period, from 1972 to the present, when the shift to floating exchange rates became inevitable with increasing capital flows. In this study, this last phase is considered to be the current era of globalization, where both capital and trade flow s are rising. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. systematic study on. the effects of globalization upon LDC government welfare expenditures. It is systematic in that the theoretical model is empirically estimated on the basis of social security and welfare expenditures of 53 LDC governments, and it controls for the existence of unobservable country specific variables. Estimates are obtained using panel data and the fixed effects method in an effort to mitigate, to the extent possible, the disadvantages of making broad generalizations about LDCs. The impetus for this analysis is the fact that while trade and (gross) capital flows have been increasing in both rich and poor nations, social security and welfare expenditures have been rising in developed countries but not in LDCs. Charts l.l and 1.2 well illustrate this contrast between the two groups of nations.3 Richer countries expanded their average spending from an average of 12 percent of their GDP in 1972 to 16 percent in 1995. Meanwhile the LDCs spent an average of 3 percent in 1972-74 and 2.5 percent in 1994-95. 2 Rodrik (1997) briefly includes developing countries in chapter 7 of his analysis. However his estimates lack robustness in that they include only 23 developing countries, use S year averages, and do not take over time variations into account In his most recent publication on the topic, Rodrik (1998) focuses upon the effects of trade on total government spending in LDCs. He does not examine the impact of trade and capital flows on welfare spending, in particular. 3 Refer to Appendix A for a more detailed description of how trade and gross capital flows are calculated. 4 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chart 1.1: Globalization and Welfare in LDCs Trade, Capital Flows and Govt Welfare Expenditures in LDCs 1972*1995 (as % of GDP) N*53 L < > -Trade — Capital Flews ■ ■Govt Welfare Exp. Chart 1.2 Globalization and Welfare in Advanced Countries Trade, Capital Flows and Govt Welfare Expenditures In Advanced Nations 1972-1995 (as % of GDP) N* 14 2 3 I 20 C « a 1 0 if s I 0 — —Trade— Govt Welfare Expenditures — Capital Flows j Sources: Government Finance Statistics (IMF: various editions) Balance o f Payments Statistics (IMF: various editions) World Bank Development Indicators (World Bank: 1998) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The objective of this inquiry is to explain the declining trend of government welfare expenditures in LDCs since the early seventies. It is intriguing that government spending on social security and welfare as a share of GDP in developing countries has not been rising (see Chart 1.3). Chart 13 : Trend o f LDC Government Welfare Expenditures CL o 0 4 as 3 25 2 1.5 1 Q5 0 LDC GQMBmrat WHfare Bqaendtires1972-1995 (%ofGDP) N=53 •G cw tW fetfere Expendtires •Trendine i i i i t i i i i i i i i i i i i - 1 i i i i •Please see Appendix A for list of countries included and for further description of the data. Sources: International Fi nance Statistics &Govemment Finance Statistics (IMF: various editions) Unfortunately, only S3 countries could be included in this study because the data for the remaining non-OECD developing countries was either unattainable, or was not recorded as a time-series. This LDC data set is limited to the S3 developing countries for which comparable, standardized data for government social welfare expenditures 6 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. could be obtained. The classification of “LDC’ was set according to low, middle and high income non-OECD counties as categorized in 1997 World Development Indicators. Data used in this analysis cover 20 Latin American and Caribbean countries, 12 African countries, eight Middle Eastern countries, 10 Asian countries, and three European countries. Of these, the majority are low and middle income countries (i.e., 16 and 33, respectively), while the four remaining countries are high income non- OECD countries. Thus, this sample is biased in the sense that it excludes a comparable number of high income countries and several newly industrialized countries (NICs).4 For example, fast emerging international market participants such as China, Taiwan, and Hong Kong were amongst those that had to be excluded because of data unavailability. It is likely, however, that such an omission creates a more homogenous sample of developing countries since these missing countries tend to have exceptionally high levels of globalization.3 The question of why welfare benefits in LDCs have not been increasing over time is a puzzling one. Social security experts argue that implementation of one important kind of welfare expenditure, social insurance programs, is an evolutionary 4 Eastem European countries were intentionally excluded from this data set because their historical circumstances (particularly regarding state social spending) are distinct from the majority of non-OECD developing countries. These countries are not readily comparable to other LDCs because they are undergoing a unique historical experience with respect to then market transitions, have comparatively different functions for the state and welfare, and on average^ maintain much higher levels of spending on welfare relative to GDP. Therefore, inclusion of these countries would skew the data and bias the results. s Integration into the world economy for die NICs has been occurred at an exceptionally rapid pace compared to other LDCs (see Griffith-Jones and Stallings 199S and IM F World Economic Outlook 1997). 7 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. process. Initially, a few benefits are offered until, eventually, there is universal coverage (Midgely, 1984; Schmidt, 1995). Influential theorists of International Political Economy maintain that welfare states increase in tandem with greater exposure to international markets (Cameron, 1978; Garrett, I998a,b; Katzenstein, 1985). However, the opposite trend seems to be occurring in the LDC cases. This is in direct contrast to the developed world where welfare spending as a percentage of GDP has been steadily increasing. Analyzing both developed and developing countries reveals that factor endowments mark the crucial distinction between these nations in the current analytical context.6 Unlike developed countries, LDCs tend to be simultaneously labor abundant and capital scarce. Stiff international competition for capital and amongst low wage labor markets imposes strong constraints for most LDC governments. Consequently, LDC governments, contrary to their developed countries’ counterparts, have greater incentives to deny benefits to labor and limit their spending to ventures that directly enhance international competitiveness. This study presents three important conclusions about globalization and LDC welfare states. First, globalization has a negative overall effect on the capacity of LDC workers to demand social programs, depending upon the skill level of labor and their extant political strength. Their ability to pressure governments for greater welfare spending is impeded. Second, domestic historical-institutional constraints in LDCs 6 In this chapter, factor endowments refer specifically to unskilled labor, skilled labor and capital See Thompson (1995), Pastor (1998), and Wood (1994,1995,1997). 8 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. discourage greater and more efficient levels of welfare spending. Finally, LDC governments can still afford to increase spending on welfare programs, though labor’s collective action problems have prevented them from becoming a strong political group. Theoretical Context and Alternative Hypotheses Forming a theoretical basis to evaluate LDC welfare spending in the current era involves synthesizing several schools of thought, particularly since there is very little existing research on the topic.7 Theorists who do investigate the welfare state in LDCs are either prescriptive or historically reflective (Ahmad et al., 1991; Esping Andersen, 1996; Mesa Lago, 1994; Midgley, 1994; Schmidt, 1995; Usui, 1994; Wahl, 1994). Interestingly, such scholars tend to favor domestic level explanations for understanding welfare states. Although there are many others who have generally questioned the sustainability of welfare state spending in LDCs (Eichengreen, 1997; Herdia, 1997; Hurrell & Woods, 1997; Kothari, 1997; Pasha, 1996), none have adequately investigated the downward trajectory of government welfare expenditures or analyzed it a panel data empirical fashion. On the other hand, students of both International Political Economy (IPE) and Comparative Politics have long been discussing the resilience of the welfare state in developed countries. They have taken for granted two 7 Most theorists investigating the globalization-welfare state nexus in the industrialized world are too quick to dismiss the consideration of LDCs from their analyses. This negligence is based on the false assumption that, there is a lack of capacity and resources for the existence of LDC welfare states. (See Atkinson & Hills, 1996; Garrett, 1998b; Rodrik, 1996,1996 for examples). Social programs have, in fact, been implemented in LDCs as early as 1924, beginning with Chile, and since spread to as many as 70 developing countries. 9 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. crucial preconditions for institutional structures: states with sufficient revenues and a well functioning democracy based on a workable and enforceable constitution. While each of these works offer some valuable tools by which to analyze government welfare expenditures in LDCs, none can effectively capture the interaction between international markets and LDC welfare states. Generally, most theorists researching welfare state dynamics tend to focus on developed countries and favor either domestic or international conditions to explain the sources of changes in welfare state expenditures. Those who concentrate upon conditions within countries tend to emphasize a nation’s demographic, economic and political conditions to account for variation in welfare state expenditures. Alternatively, theorists who emphasize international causal variables suggest that government welfare spending is a function of a nation’s links to the global economy. Finally, most comparative scholars who investigate the sustainability of contemporary welfare states maintain their analytical focus upon the advanced industrialized countries. In other words, the political and economic circumstances in which they evaluate welfare spending are specific to the OECD countries and completely exclude 10 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LDCs. This dissertation investigates the combined effects of domestic and international factors on the welfare state in developing countries.8 Domestic level hypotheses: economic development, population growth and democracy Demographics, economic development and democracy are the most common domestic level explanations for variations in state spending on welfare. The social security literature argues that government spending on transfer programs, i.e., pensions and family benefits, increases in tandem with expanding demographic factors, particularly large working and/or elderly populations (Tang, 1996; The World Bank Policy Research Report 1994; Usui, 1994/ Political economists suggest that the extent of state social spending is based upon nations’ economic development, or GDP (Atkinson & Hills, 1991, Stern, 1991). This proposition is in accordance with Wagner’s Law 1883 states that the growth of the public economy is positively related to the economic affluence of the nation. Finally, the political science literature claims that the more democratic a nation is, the more responsive the government must be to the demands of the populace. Democracies generally give greater voice to the median voter and large organized groups (such as labor unions) which tend to favor larger welfare * Dependency theories are the closest approximation of past theoretical works that focused on the negative domestic effects of international markets (particularly trade) in LDCs. See Cardoso and Faletto (1979), Dos Santos (1969), Evans (1972), and Singer (1970) for examples. However, I argue that the manifestations of international market integration today are distinctive and counter many of the predictions posed by dependency theory. Less developed countries have witnessed positive growth rates— particularly Ac East Asian countries (NICs)— while terms of trade have varied over time. In short, traditional dependency theories have not been convincing with regards to the empirical validity of their assumptions and predictions. While some more recent dependency theorists do take growth possibilities of LDCs into account, they do not systematically examine the impact of trade and capital flows on government social spending. 11 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. state spending (Baneg'i & Ghanem, 1997; Eichengreen, 1996; Esping-Andersen 1996; Hicks & Swank, 1992). While these hypotheses might help to explain different levels of state social spending (as a percentage of GDP) between countries, they do not explain long run trends among developing nations. A brief review of the data suggests that the relationship between LDC welfare states and demographic, economic and political variables is more complex than these (domestic-level) theorists might assume. Contrary to the predictions of the domestic-level hypotheses, countries experiencing both population and economic growth (e.g., India, Kenya, Dominican Republic) have been reducing their welfare expenditures as a share of GDP since the mid 1980s. Interestingly, developing democracies (e.g., Botswana, Mauritius, Turkey) have also been experiencing similar welfare cutbacks. Thus, domestic conditions by themselves are unable to account for trends in LDC income transfers. International factors must be taken into consideration, alongside these national variables. International level hypotheses: Compensation versus efficiency effect J Theorists investigating the link between economic globalization and welfare state policies are divided. Some argue that efficiency linked welfare retrenchment will occur with globalization, while others expect compensation oriented welfare growth. 9 The categories of efficiency and compensation are based on Garrett (1998b). 12 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The “efficiency hypothesis” posits that economic globalization and increased market competition is the cause and consequence of lower state spending on redistribution programs. This hypothesis, recognized as the “conventional (neoliberal) wisdom,” predicts that there is an inverse relationship between the size of welfare states and the extent of globalization (Garrett, 1998a,b). Alternatively, the compensation hypothesis proposes that governments resort to increasing income transfers in order to compensate those who are negatively affected by globalization. Those who support the conventional “efficiency” argument can be further divided into two distinct categories.1 0 The first group, the structuralists, claim that welfare state retrenchment is a consequence of increased state pressure to be more efficient and, thereby, more competitive in global markets. States are virtually forced to distribute resources towards more competitive ventures, and away from the needs of citizens unable to adjust to market developments (Cemy, 1995; Eichengreen, 1997, Gill 1995; Hurrell & Woods, 1995; Kothari, 1997; Ruggie, 1994; Scholte, 1997; Strange 1996). The second efficiency group adheres to the fundamentals of neoclassical economics. They argue that more efficient distribution and production structures resulting from economic globalization are likely to increase overall economic growth and, thereby, reduce the need for welfare states. Redistributive government spending in the era of globalization is deemed distortionary and, hence, detrimental to market efficiency (Dunning 1997; IMF World Economic Outlook 1997; World Bank 1996). 1 0 This study uses the concepts of “efficiency effect,” “conventional wisdom” and “neoliberal logic”* interchangeably. 13 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Either way, globalization constrains state’s policy-implementing abilities by limiting the distribution of government resources to ventures that enhance the competitive position of their economies. These pareto improving government policies, in turn, encourage the process of globalization by attracting greater capital flows and improving export production (see Figure l.l). The expanding depth and breadth of globalization has engendered the conventional wisdom that all states, regardless of their partisan compositions and national differences, will come to embrace neoliberal policies in order to maintain national competitiveness in a globalizing world (Cemy, 1995; Evans, 1997; Gill, 1995; Grunberg, 1998; Huber & Stephens, 1998; Pontusson, 1995; and Ruggie, 1994). Accompanying this extraordinary growth of global export and capital markets has been the increasing voice of international investors and their neo-liberal ideology in national policy debates. Their capacity to withdraw and shift productive and financial capital with relative ease made more credible the threat of “exit.”1 1 Subsequently, the alleged ability of international investors’ to punish governments for “bad” economic and fiscal policies led to this prediction of convergence amongst national policies. The last two decades thus became witness to the reification of Lindblom’s “markets as prison” idea (Maxfield, 1998). As a result, all expenses affecting national revenues were expected to be redirected towards more efficient and “market friendly” programs. Government welfare expenditures become logical victims in this global paradigm shift. 1 1 Schwartz (1998) points out that the distinction between goods-producing capital and financial capital should be emphasized in the current era of globalization. He argues they each have distinct effects and different degrees of mobility. 14 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 1.1: The Efficiency Effect (Conventional Wisdom) -Sfetefeatfeff a Capital M obility Trade | Efficiency 4 Welfare I spending The “compensation” perspective challenges both efficiency views by suggesting that globalization neither seriously alters the state’s abilities to implement certain redistribution policies, and does it make welfare spending necessarily inefficient. These theorists present convincing empirical evidence on the compatibility of economic growth, political stability and welfare state spending (Cameron, 1978; Garrett, I998a,b; Garrett & Lange, 1991,1995; Rodrik, 1996, 1997b). Such works suggest a positive correlation between state social expenditures and economic globalization within developed countries. Nonetheless, these empirical investigations, by addressing only 14 economically and politically developed countries of the world, fail to capture how globalization is affecting the majority of the global population. Geoffrey Garrett (1998a,b) presents the most recent and convincing challenge to the conventional wisdom that welfare states are crumbling under pressures to maintain 15 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. competitive national economies in the era of globalization. By analyzing 14 nations of the Organization for Economic Cooperation and Development (OECD), he demonstrates that international market exposure actually induces greater government spending on redistribution programs that “compensate” for market generated inequalities. Predecessors to Garrett documented the positive relationship between the welfare states and international openness in developed countries for most of the 20* century (Cameron, 1978; Rodrik, 1997).1 2 Garrett extends the globalization welfare debate initiated by Katzenstein (198S) and Ruggie (1983).1 3 His work is particularly pathbreaking at a time in which the loss of national autonomy has become increasingly feared in a world of freely mobile capital. Significantly, such works reinforce the growth with equity thesis (one that is bypassed by conventional theorists). Nations with more equitable income distributions are politically stable and, thereby, attract more capital inflows. The net effect is greater economic growth (Garrett 1998a). Key to Garrett’s analysis is the ability of labor market institutions to effectively mediate between government policies and labor demand. Only if labor markets are highly centralized and well developed can labor and government coordinate economic performance with redistribution policies. From his empirical investigation, Garrett concludes that globalization has in fact strengthened organized labor and, as a result, 1 2 This historical relationship between globalization and the welfare states in developed countries makes it all the more intriguing why this same relationship does not seem to be materializing in LDCs. 1 1 The theoretical basis for this debate was laid by the seminal work of Karl Polanyi (1944) and forwarded by John Ruggie (1982) with his concept of “embedded liberalism.” They conjecture that the state must make a broader commitment to social welfare in order to temper the “pernicious effects” of international markets. 16 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. cross-national partisan differences in the developed world have been sustained (see Figure 12). Garrett’s recent work clearly defies conventional wisdom and provides the most compelling case for the resilience of partisan politics and the survival of the welfare state in the face of globalization pressures. Yet such analyses have two limitations: they are limited to states that have sufficient revenue flows to provide a high social wage, and states that have formalized democracies and, thus, the institutional structures necessary for dialogue between societal pressure groups and public policy makers. Once these preconditions are withdrawn, the fate of welfare states for the majority of nations in the world is once again in question. Indeed, the majority of developing nations are severely lacking on both of these fronts. Put simply, LDC governments do not have revenues to both compensate those disadvantaged by globalization and maintain economic competitiveness. While trying to participate effectively in global markets, most of these nations are also grappling with issues such as high debt, unemployment, privatization and other structural adjustments. Second, relative to developed nations, those most disadvantaged by globalization do not have the formal democratic means by which they can negotiate with governments.1 4 17 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The IPE literature thus tends to ignore the question of how nations relatively deficient in both political and economic resources are coping with the social consequences of globalization. Figure 1.2: The Compensation Effect (Garrett’s Hypothesis) Welfare spending •Yes t i Capital M obility 'Strong Labor .Institutions?, Trade Welfare spending No •Globalization Comparativists: Developed-country-centered hypotheses Comparativists have been making a related but separate argument with respect to the endurance of the contemporary welfare state. Pierson’s widely cited 1996 World Politics article provides the most compelling case for welfare state resilience.1 5 He distinguishes between the politics of welfare state expansion and retrenchment in 1 4 Statistical data in Ted Robert Gurr’s and Keith Jaggar’s Polity HI (1994) on democracy ratings makes valid this comparison between developed and developing countries. See also Hadenius (1992) for an index on democracy on 132 developing countries in year 1988. Note, in particular, the low ratings of LDCs in organizational freedoms and effective and correct elections. For more theoretical discussions on the limitations of democracies in developing countries, see Rueschemeyer et al. (1992) and Crisp (1994). 1 5 See also Clayton and Pontusson (1998) whosubstanriate this claim and move forward with a constructive reformulation of Pierson. 18 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. developed economies. Pierson (1994, 1996) argues that while well-organized labor groups were once the impetus for welfare state development, the politics of welfare state retrenchment must be assessed from an entirely different basis. The contemporary welfare state now enjoys widespread popular support, resulting in cross class cutting coalitions. Consumers of welfare benefits (i.e., the elderly, the disabled) have mobilized successfully to prevent the dismantling of the welfare state. Significantly, this research de-emphasizes the role that labor now plays in the maintenance of government spending on welfare states in developed countries. As a consequence, comparativists ultimately fall short of theorizing the situation of contemporary LDC welfare states where labor still remains the primary pressure group supporting it Although this literature approaches the welfare state from a different perspective, it too falls victim to the second general assumption. In order for Pierson’s theory to hold, average citizens must have the means and institutional support to mobilize and make demands upon government policy. Generally, non-labor groups in LDCs have yet to attain a comparable level of political development. The urban labor force, historically being the best organized in developing nations, still remains the most important group in LDCs with regards to their influence upon welfare states (Midgley, 1984; Schmidt, 1995).1 6 Therefore, the politics of welfare state expansion and retrenchment in LDCs are not as distinct as Pierson sees it in the advanced nations. 1 6 For more detailed discussions on power resource theory see Esping Andersen (1990), Korpi (1983), and O’Conner and Olsen (1998). See also Collier and Collier (1991) for a thorough discussion on the link between labor mobilization and the evolution of national politics. 19 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Compared to the more developed countries, LDCs not have the same institutional means for multi-class coalitions. Therefore, an LDC model of the welfare state must account for labor’s role in the existence of the welfare state.1 7 Such an analysis should be based on power resource theory that suggests that the success of welfare states depends upon a well organized labor movement Comparativists have thus far failed to consider how labor and LDC welfare states are being affected by the current era of globalization (Table 1.1). T ablel.l: Summary of Literature Literature Welfare State Variables Domestic Level Hypotheses • Social Security • Political Economy • Political Sdancs • Demographics • G D P • Democracy International Laval Hypotheses Efficiency Effect (C onvanttonaf AmuntnO • Conventional Economics * L a., Neoilberallsm • International Market Integration Compensation Effect * International Political Economy ► e.g., Garatt, Rodrik, Cameron » International Market Integration * Strong Labor Institutions Developed Countries Centered Hypotheses * Comparative Politics • e.g., Pierson * Strong Social & Political Institutions (P luralist M o d e l) 1 7 Unless specified otherwise, “labor'* will refer to both unskilled and skilled workers. 20 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Thesis, Methodology and Research Design This study applies both quantitative and qualitative methods to test the competing hypotheses and propose a new model about the effects of domestic and international variables on LDC governmental welfare expenditures. Factor endowments in LDCs give strong reasons to expect that the efficiency hypothesis will prevail in developing countries.1 8 Nations with similar factor endowments (i.e., an abundance of unskilled labor) are entering international markets at roughly the same time.1 9 Therefore, with growing global markets, developing countries are struggling to maintain the international competitiveness of their labor-intensive goods. LDC governments are being presented with greater incentives to repress labor costs. Also, low organizing capabilities of these unskilled laborers (Deyo, 1989; Henk, 1995; Ingersen, 1984; Shafer, 1994) enables governments to deny labor demands with greater ease than in more developed countries. The hypothesis is that i f a nation has a comparative advantage in (unskilled) labor, then globalization m il lead to a decline in government welfare expenditures. This study incorporates both quantitative and qualitative methodologies to test the competing hypotheses. Whether or not Garrett's (1998a,b) compensation thesis, 1 S Factor endowments are land, labor and capital. Research indicates that most LDC’s specialize in labor- intensive, basic manufacturing activities (Deyo 1989, UN Trade and Development l992/’93). 1 9 Wood (1994, 1995, 1997a, b) makes an important distinction between skilled and unskilled labor. He claims that since capital is mobile, the North does not necessarily enjoy a comparative advantage in capital-intensive production. Rather, trade is based on the availability of skills, not capital. Unskilled labor, in this analysis, refers to workers who have little or no education (referring to Wood’s BAS-EDs and NO-EDs, respectively). These workers ( with minimal education) are employable in basic manufacturing and are most affected by a large population of surplus laborers. 21 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. that globalization induces greater welfare state spending, holds true for developing countries is of particular interest The quantitative sections of this study test the relationship between economic globalization and LDC welfare states by way of regression analysis. The purpose of the qualitative analysis is not only to provide further evidence confirming (or rejecting) the proposed hypothesis, but also to trace “the decision process by which various initial conditions are translated into outcomes” (George 1979: 43). Detailed case studies will then be useful for attaining greater insights into the policy making process that connects international markets and welfare spending. This dissertation involves three sections and six chapters. The first section of this dissertation (Chapters 2, 3 and 4) uses econometric tools to test the proposed hypothesis. This section presents a LDC model in an effort to extract their economic and political particularities. Based on this model, regression results from LDCs will be compared with those from developed countries to isolate the primary independent variables that are causing the differing trends in government welfare spending amongst these two sets of countries. Using three case studies, India, Brazil, and Korea, Chapters 3 and 4 explore and analyze the results of the LDC model (presented in Chapter 2) with greater detail. The second section (Chapter 5) revisits the conclusions from chapter two and estimates the reverse impact, or feedback effects, of lower welfare expenditures on the level of globalization. This section subjects the conventional neo-liberal wisdom to a more rigorous test by, once again, applying econometric methods to assess the potential macroeconomic costs of greater social spending in LDCs. Chapter 5 thus 22 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. explores the possibility that LDCs can increase its range of social policy options in this era of globalization, without incurring severe consequences to their national competitiveness. Finally, the third section (Chapter 6) takes a full view of the LDC model, complete with the qualitative variables that emerge from the case studies and the feedback effects reported in Chapter 5, to understand and explain the relationship between globalization and LDC welfare spending in its totality. The normative goal of this investigation is to present viable policy prescriptions that encourage social and economic development in LDCs. Although international organizations have broadened their views on development, the impact that government welfare spending can have on both long run political and economic development is still being underestimated.2 0 Evidence supporting the efficiency hypothesis in the case of LDCs reveals how and why poor countries do not include government support of labor as part of “good government” policies. Answers to the proposed research question will provide policymakers and researchers with compelling evidence on which to base their assessments of welfare policies in the era of globalization. This analysis highlights the misguided nature of certain policies unfolding from globalization and, thereby, proposes a framework for policies that could strengthen labor movements and government social spending in LDCs. As evidenced by the advanced countries, such moves toward greater LDC social stability would, at the same time, support the expansion of international markets. Ultimately, this research seeks to isolate potential sources of domestic 2 0 As Garrett (1998a: 1S9) explains, most of these policy prescriptions limit government spending to “market friendly” measures such as the provision of basic public goods, securing property rights, bur enforcement of contracts, etc. 23 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. instability and will foster public policies that will simultaneously stabilize the international and domestic sphere. 24 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 2 THE LINK BETWEEN GLOBALIZATION AND WELFARE STATES IN LESS DEVELOPED COUNTRIES This chapter models the effects of globalization on LDCs. The objective is to incorporate, to the extent possible, their historical political and economic particularities, as well as to discriminate among the various domestic and international level hypotheses. It is found that when the model takes into account certain structural and political factors affecting labor in LDCs, the efficiency hypothesis prevails. In order to analyze the correlation between globalization and LDC government redistribution policies, the relationship between international markets, labor and LDC welfare spending must be hypothesized. The key causal link lies in the interaction effect between globalization and the bargaining power of labor. This study captures the political implications of LDC factor endowments and presents a new approach to approximating labor strength in LDCs (i.e., the ratio of unskilled to skilled labor). Simply put, the hypothesis is that the greater a nation's comparative advantage in 25 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. unskilled labor, then the increasing interaction between globalization and weak labor power creates a decline in government welfare expenditures. The concept of the “LDC welfare state” is adopted from Pfaller et al. (1991:2). They refer to “welfare statism” in the most general way as “the use of state power and responsibility towards the ends of protecting citizens against economic adversities and ensuring a certain standard of prosperity to all.” This analysis focuses specifically upon social security and welfare services which, according to Pfaller et al. (1991), is the most visible manifestation of welfare statism. This hypothesis emphasizes three points. First, globalization has led to intense competition amongst countries with similar factor endowments (i.e., unskilled labor). Second, confronted with growing international demand for their most abundant factor (unskilled labor), greater competition for international export and capital markets stimulates employment of unskilled labor and, thereby, increases UNSKILL. Finally, because the corresponding decrease in surplus labor is not sufficient and unskilled labor is harder to organize, labor as a whole has less relative influence on government fiscal decisions. Governments ultimately have greater incentives to shift fiscal resources away from programs labor favors (the welfare state) and redirect them towards satisfying foreign (and possibly domestic) investors at the expense of labor. Specific assumptions about the different motivations behind welfare spending are made in this study. First, it is assumed that both skilled and unskilled workers desire greater welfare benefits, while owners of capital prefer lower welfare spending. The latter’s predisposition is based on a preference for paying lower taxes and less 26 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. overall government interference. Finally, it is assumed that governments are primarily concerned with promoting the aggregate level of economic performance and, at the same time, arbitrate between the political demands of societal groups commanding power resources. The Economic Effects o f International Markets on Labor The proposed hypothesis is based upon the reasonable assumption that most LDCs are simultaneously labor abundant and capital scarce. Participation of LDCs in international markets eventually leads to greater productive activities involving unskilled labor. According to the Hecksher-Ohlin and the related Stolper-Samuelson theorem, labor abundant economies will specialize in and export labor-intensive goods in a global economy.1 It wouild therefore, be expected that LDCs would export either primary products or basic manufactured goods or both. However, exports of primary commodities have been declining as a share of world trade.2 1 In order to apply the Stolper-Samuelson Therorem here, it must be assumed that (1) the relatively abundant factor in LDCs is labor, and not natural resources (see, for example, Table 1) and (2) there is full employment This author agrees, along with Paus and Robinson (1997), with the reasonability of assumption one, but not assumption two. The surplus variable has been added to the LDC modeli order to address the inadequacy of assumption two. 2 According to an OECD (1997:11) report, primary commodity trade is declining because there is: (I) low income inelasticity of demand (mainly for food) and (2) declining intensity of raw materials used in economic activity. This also explains why some Iow-income countries have not experienced a growth in trade that is as intense as many of the other LDCs. Note also that by lowering the tariffs in many developed countries, the Uruguay Round Agreements of the General Agreements on Tariffs and Trade (GATT) served as an important impetus to the expansion of manufactured exports by LDCs (Safadi& Laird, 1996). 27 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. As a result, in several empirical studies, Wood (1994, 1995, 1997) confirms that nations with high levels of unskilled labor have been placing more emphasis on basic manufactured goods.3 This study focuses only upon the manufacturing sector to assess the economic (and political) effects of international markets on labor, since manufacturing is the growing hotbed of economic activity in LDCs, Table 2.1 illustrates the growth in labor intensive manufacturing exports from 1992-1996. Take particular note of the contrast between the low income countries and high income OECD countries. 3 The importance of this distinction between unskilled and skilled is emphasized by Wood (1994, 1995, I997a,b). He claims that since capital is mobile, the North does not necessarily enjoy a comparative advantage in capital-intensive production. Rather, trade is based on the availability of skills, not capital. Wood defines skilled workers as those with more than a basic general education, “which makes it a mixed bag of professional and technical workers, managers, and craftsmen, all of whom have advanced education or substantial training or work experience” (Wood 1994:6) Thus, skilled workers are generally employed in high-skill intensive manufacturing. Unskilled labor, in this analysis, refers to workers who have limited or no education (referring to both Wood’s BAS-EDs and NO-EDs, respectively). BAS- EDs, according to Wood, are employable in Iow-skOl manufacturing. I add that they are the group most affected by a large population of surplus laborers. 28 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 2.1: The growth of labor intensive manufacturing exports LOW INCOME COUNTRIES Growth ^•'S S Share of Exports (1996) Primary Products 12.7 60.1 Natural Intensive Manufacturing 12.6 13.4 Labor Intensive Manufacturing 18.9 14.6 Technology Intensive Manufacturing 132 3.6 Human Capital Intensive Manufacturing 14.7 2.8 LOW M IDDLE INCOME COUNTRIES Primary Products 9.4 56.6 Natural Intensive Manufacturing 7.6 4.6 Labor Intensive Manufacturing 14.8 20.8 Technology Intensive Manufacturing 212 11.3 Human Capital Intensive Manufacturing 11.6 6.0 HIGH MIDDLE INCOME COUNTRIES Primary Products 9.0 49.23 Natural Intensive Manufacturing 18.8 7.23 Labor Intensive Manufacturing 1 C 5 13 46 Technology Intensive Manufacturing 13.3 15.54 Human Capital Intensive Manufacturing 9.8 1ZO O HIGH INCOME OECD Countries Primary Products 8.07 23.86 Natural Intensive Manufacturing 10.21 4.79 Labor Intensive Manufacturing 3.57 9.07 Technology Intensive Manufacturing 12.14 35.47 Human Capital Intensive Manufacturing 9.S0 23.43 Source: U N IT C InfoBase: National Trade Performances by C ountry 29 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chart 2.1 demonstrates that LDC’s manufactured exports as a percentage of total merchandise exports have increased almost three-fold since the early 70’s. Consequently, the growing pressures of globalization upon the governments of these late integrators should not be underestimated. The entry of labor surplus economies, such as China and India, into the world market has dramatically increased the world supply of unskilled labor (Wood 1997). For example, the East Asian NICs, once world leaders in the exports of labor intensive commodities, are now being Chart 2.1: Trends in LDCs Manufactured Exports* M an u fac tu re d E x p o rts in LDCs (as % of M e rc h a n d ise E x p o rts) N = 53 50 45 40 | 35 3 30 | 25 | 20 « 15 10 5 0 7 2-75 76-80 81-'85 86-’ 90 91-'95 * J’: - u ■ - . **.;V 'i U . ' y i • • % _ * .■ *>t- V • ■ ' * See Appendix A for countries included Source: World Development Indicators (World Bank: 1998) 30 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. exports of labor intensive commodities, are now being challenged by the exports of countries such as Malaysia, Philippines, Sri Lanka, China, Vietnam, India, Pakistan and Honduras, to name a few (Gereffi 1995, UN Industry and Development 93/94).4 The stimulus of the world market has hence given rise to greater competition amongst the labor abundant economies of the developing world, both for attracting capital and expanding their low wage labor markets. As such, LDC governments are being presented with strong incentives to prioritize national competitiveness concerns over social spending.5 It is labor that is most adversely affected by this international market rivalry for capital inflows and export markets. This is especially true for laborers in countries that liberalized their markets in the late to mid eighties. These late integrators, which are the majority of LDCs, face more intense competition in low-skill manufacturing industries than did East Asia in the 1960s and 70s.6 Economists contend that simultaneous liberalization of large low income countries worsen their terms of trade in the contemporary world market (Bleaney, 1993; Dutt, 1992; Lutz & Singer, 1994). Based on this argument, several scholars have suggested the improbability that lower income countries might successfully follow the “East Asian miracle” (Amsden & Van Deer Hoeven 1996; Lee, 1995; Wood, 1997). Recall that the newly industrialized countries, anticipating international market * This is especially apparent in the textiles and apparel industries. 5 For a more detailed look at the relationship between welfare spending and national competitiveness, please see Chapter 5. 6 The terms labor intensive goods and low-skill manufactured goods are used interchangeably 31 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. opportunities, succeeded in shifting their comparative advantage towards goods of intermediate skill intensity.7 The question then arises: is the current environment of international competition in LDCs such that, by specializing in goods of low skill intensity and successfully keeping wages low, they see less economic advantage in raising the skill development of their work force? The Political Effects o f International Markets on Labor From an economic growth perspective, it seems that globalization should benefit labor in LDCs by increasing international demand and, eventually, the employment possibilities of unskilled labor. The determining variable for greater welfare benefits then becomes political or, more specifically, the bargaining power of labor. By applying the logic of the Stolper-Samuelson theorem to the political realm, Rogowski (1987, 1989) argues that greater international demand for labor should eventually result in greater political power. He takes for granted, however, that there is full employment of labor and the organizing potential of labor is strong. Since neither of these assumptions hold in most LDCs, globalization is much less likely to strengthen labor movements in labor surplus economies. 7 East Asia had combined increased education expansion with temporary protection of a sequence of industries that demanded greater skill intensity (Wood & Ridao-Cano, 1996). 32 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Data in Table 2.1 and Chart 2.2 suggest that international demand for labor intensive exports is indeed rising. Indeed, as the Hecksher-Ohlin model predicts, both labor-intensive exports and manufacturing employment in LDCs are increasing. Chart 2.2: LDC Employment Trends in M anufacturing 140 120 e o ^ 100 •a >> 80 O a E o Employment in Manufacturing (LDCs) N=53 60 £ & 40 20 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 Source: World Bank Development Data 1995 If low-skill manufacturing goods comprise a large percentage of a nation’s manufacturing production and exports, then it can be concluded that the majority of 33 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. employment growth in LDC manufacturing is being undertaken by unskilled labor.8 The implication is that the formal sector is expanding to include unskilled labor.9 This is in direct contrast to OECD nations which, as Wood (1994) argues, enjoy a comparative advantage in skilled labor. Based on this logic, Wood (1994) presents empirical evidence that international demand for skilled labor in industrialized nations is increasing. Chart 2.3 illustrates the striking contrast between developed and developing countries with regards to the growth of unskilled labor employment. 8 This analysis incorporates the assumption that labor intensive industries mostly employ unskilled laborers. Although it is recognized that there are skilled employees in labor intensive manufacturing industries (and unskilled laborers in skill intensive manufacturing), this categorization intends to capture the general composition of laborers. 9 Certainly, globalizadon also contributes to the expansion of the informal sector (see Grunberg, 1998). In the fhce of heightened global competition, large enterprises subcontract to unprotected workers to reduce labor costs and increase flexibility. However, this analysis focuses upon effects upon the formal sector since Midgely (1984), Mesa-Lago,(l994) and Esping-Andersen,(1996) argue that traditional welfare benefits exclude the non-salaried work force and often serve as privileges for labor in the formal sector only. 34 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chart 23 : The Growth o f Unskilled Labor Employment in Advanced Countries and LDCs* Employment In Low-Skill Manufacturing Industries (relative to high* skill intensive industries) N =67 (A o Q 4 r 1.2 3.5 - 1 3 2.5 - 0.8 2 w ^ - 0.6 1.5 - 0.4 1 0.5 - 0.2 0 - 0 •LDCs 'Advanced Countries (0 o a o O ■ o o o c < o > ■ o See Appendix A for countries included Source: UNIDO data base (provided by James Galbraith). Yet, contrary to Rogowski’s theoretical predictions, this microeconomic gain for unskilled labor in LDCs does not necessarily translate into greater political power with greater openness.1 0 Chart 2.3 verifies that the number of unskilled laborers in LDCs is increasing relative to skilled laborers. The UN Industry and Development Report 1 0 Indeed, Rogowski (1987, 1989) refers to capital, land, and labor as a whole in his analysis. However, the logic of his theory suggests that unskilled labor should enjoy political benefits because it is abundant and international demand for this group is increasing. 35 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1992/93, however, reports that this growth in employment has still failed to significantly reduce the vast amount of surplus labor that exists in LDCs. This, added to the fact that unskilled laborers in LDCs are generally very difficult to organize (relative to skilled labor), suggests that labor’s bargaining power is being weakened with globalization.1 1 Taken together, globalization intensifies the collective action problems of labor by expanding the base of unskilled laborers and, thereby, negatively affecting the potential for coalition building amongst the different labor groups. It consequently becomes much more difficult for labor to increase collective pressure on government for policies in their favor. Evidence against the compensation hypothesis in the case of LDCs is beginning to build. Thus, it is important to test whether or not LDCs are experiencing greater growth in employment of labor intensive manufacturing relative to skill intensive. If this is true, then it is possible that the overall level of labor’s bargaining power is declining as a result of globalization. Figure 2.1 diagrams the link between international markets and labor, assuming that labor surplus is high. 1 1 For detailed hypotheses on why unskilled labor groups in LDCs are difficult to organize, see Chapter 6 ofDeyo (1989), Gereffi (1995), Ingersen (1994), and Lok (1993). 36 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 2.1: Effects of international demand on the bargaining power of labor in high surplus labor economies No Nm « Employment of unskilled labor Nu « Employment of skilled labor Xy, « Export of labor Intensive manf. goods X,^ * Export of skin Intensive manu. goods Yes 'IS | Bargain Power f Bargain Power Exports of manufactured goods In sum, this hypothesis proposes that the challenge of increased international competition coupled with the influx of unskilled groups into the manufacturing sector is weakening labor groups and preventing coalition building in LDCs. Therefore, the LDC model tests the (efficiency) hypothesis that if a country’s major factor endowment is unskilled labor in the contemporary era of globalization, governments reduce social benefits to labor (see Figure 2.2). This implies that globalization changes the power base and organization of labor within the political system by affecting some types of labor differently than others. The net effect of this growing ratio of unskilled/skilled employment and the failure to absorb surplus labor is the lower bargaining power of labor. 37 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 2.2: Model of the Efficiency Hypothesis in LDCs Levels Variables Implications Global k Capital M obility f Trade increased eompeuon amongst labor abundant and capital scarce nations Increased global supply & demand of unskilled labor Domestic ^ ^ Unskilled/Skilled Employment -Hg) H igh Surplus Labor Decreased bargaining power of labor State Policy * ^ Welfare Spending Lower benefits to labor Hypothesis: The effects o f globalization on government welfare spending in LDCs. Leads to Increase I Decrease Summation: Both events are occurring Interacts with 38 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The LDC Model This model tests the impact of economic globalization, alongside relevant domestic variables, on governmental welfare expenditures in LDCs. Garrett’s (1998a, b) model of developed countries is used as a basis for building the LDC model. The primary explanatory variable, the interactive effect of globalization and the bargaining power of labor, is taken from Garrett’s (1998a) model of OECD countries. However, to more accurately assess the domestic effects of increased international market integration in the developing world, structural conditions affecting labor in LDCs are also taken into consideration. The proposed model specifically recognizes the importance of labor as the most abundant factor endowment in LDCs.1 2 This labor variable, it will be argued, is the most appropriate proxy for the bargaining power of labor in developing countries. Once this factor endowment variable is taken into account, it appears that globalization has a negative effect on government welfare expenditures in LDCs. Similar to the works of Rodrik (1997) and Garrett (1998a), this model will be estimated by pooling cross section and time series data. This estimation procedure has two benefits. First, the use of the fixed effects model makes it possible to control for unobservable country specific differences, eliminating much of the omitted variable 1 2 As mentioned earlier in the paper, LDCs are generally considered to be capital poor and labor rich, while OECD member countries tend to be labor poor and capital rich (Agenor & Montiel, 1996; World Development Report, 1995). Note that the factor endowment variable in the LDC model presented here captures labor abundance and does not explicitly control for capital scarcity. Instead, capital scarcity is an assumption built into the model. The factor endowment variable (of unskilled labor) is expected to be significant only if the country is capital scarce. 39 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. bias of cross section data.1 3 Second, important changes which have occurred over time for the same country can be assessed. Panel data ultimately combines the benefits of an increased number of observations with the ability to eliminate country specific fixed effects. The proposed hypothesis is applied to S3 LDCs (excluding countries with insufficient data) for the years from 1972 to 199S. Using the Hausman specification test, random effects are not significant in all cases. Therefore, each of the following equations is estimated using the fixed effects procedure. Year dummy variables are included as independent variables only when they are significant. The panel contains a host of different control variables, but many of them do not show significant correlation and therefore are not shown here. All of the variables that did have any impact on the observed empirical relationship between globalization and welfare state expenditures are reported in this chapter. The effects of globalization on LDC welfare spending will be assessed in a three-stage process. First, a “closed economy” model is presented to assess the validity of previous studies of LDC welfare states that exclude globalization variables as significant determinants of welfare spending. The second “open economy” model 1 3 The country specific fixed effects expected to be most relevant to this model are culture and history. Informal systems of old age support (i.e., extended families) are common methods of providing social security in many parts of Africa and Asia (World Bank Policy Research Report, 1994). Such variables are more significant in some countries, relative to others. Fixed effects allow us to control for the influence of such idiosyncratic differences between countries without having to model them explicitly. 40 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. examines the influence of both domestic and international level variables on government welfare expenditures. Finally, the open economy model is applied to a sample of developed countries to assess whether or not the factor endowment variables mark the distinction between political outcomes in LDCs and those in developed countries (as found by Garrett (1998a)). See Table 2.1 for a summarized description of all the variables. Model 1: the closed economy model welf/,= biwelfft./+ b2 unskill(t + bjsurp;, + bjlbjX,, + Ib*country, + \iit (1) Model 2: the open economy model welf( ,= b[Welfa ./+ hztradea./ + b3Icflowa./ + b4tradea./*unskillit+ (2) bjkflowft./* unskilla + b«surpft + bjZbjXtt + Zb*countryf + \ilt The b ’ s are parameter estimates in this equation, while the subscripts / and t represent the country and year of the observations respectively; bt is the lagged rate of welfare expenditures; p. is an error term. I X represents the vector of control variables. Note that logarithms are taken of all the variables. This type of relationship is useful in that it displays the property of constant elasticities between the variables. In addition, the international level variables are lagged in order to take the period of Adjustments’ into account 41 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The Variables Dependent variable: Social Security and Welfare [WELF] Since the primary intent of this study is to discriminate among different hypotheses (presented in Chapter 1) and analyze international market effects upon government redistribution policies, government expenditures on social security and welfare are used to estimate the welfare state in LDCs. This same measure of welfare is used in both developed and developing countries to facilitate comparative analysis on the domestic effects of globalization (see Chart 2.4). According to Schmidt (1995), the social expenditure/GDP ratio is an important indication of the extent of social policy management in a country. Indeed, social security and welfare expenditures command a fair percentage of GDP in many LDCs (e.g., Brazil, Argentina, Cyprus, Israel) although coverage is less extensive than in developed countries. This measure indicates, more importantly, the amount of national resources committed to welfare in relation to a country’s total resources (Olsen & O’Connor, 1998:22). 42 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 2.1; Concepts, Measurements and Definitions C om ets Measurements* Definition Social Security sod Welfare Government welfare expenditures as a percentage of GDP [WELF)- dependent variable 'Social security' consists o f income transfers, providing benefits in cash or in kind far old age, invalidity or death, survivors, sickness and maternity, work injury, unemployment, family allowance, and health care. ‘Welfare affairs and services' are defined a assistance delivered to clients or groups o f clients arfth specials needs, such as the young, the old, or the handicapped. Globalization The amount o f total trade (EX + IM/GDP) (-)[TRADE] and gross capital flows (-) [KHGW] as a percentage of GDP. Groaa capital flows Is the sum of all Inflows and outflows, using the finest classifications to avoid netting The bargaining power of labor The ratio o f the numbers employed m labor intensive manufacturing industries relative to numbers employed in skill intensive manufacturing industries (-) [UNSKILLI; The number o f surplus laboros in the economy (-) [SURPJ. 'Unskilled' refers to tboae who have no more than a primary or secondary education, and are likely to be employed in low skill manufacturing industries. 'Skilled* are those with more than a basic general education and are usually employed hi high skill manufacturing industries. 'Surplus labor' is the total labor fame divided by the amount o f arable land. Demographic variables The Age Dependency ratio (+) [DEPEN]; Urban population as a percentage o f total polulation (+) [URBAN]. 'Age Dependency istiof fa the number o f persons over 60 divided by number o f persons aged 20 to 59. ’ Urbanization' is the midyear papulation of areas defined as urban in each country. It is measured here as the percentage o f the total population. Economic development The Gross Domestic Product per capita (+) [GDP]; Revenue of state owned enterprises (+)[SOE]; Total debt service (-) [DEBT]. 'GDP'is the total gross domestic product of a country divided by total population.'SOE' economic activity is the value added of state enterprises, estimated as their tales revenue mfmn « tw rfW a f tlwtr IwVfW wittil, tnpnf Total'debf service is the sum o f principal repayments and interest paid in foreign currency, goods, or services on long-term debt and interest payments only on short-term debt Political development Indicator o f democracy (+) [DEMOC] Using scale 0-10; 10-strong democracy. This indicator Is derived flom the codings of the competitiveness o f political participation, the openness and competitiveness o f executive recruitment, and constraints on the chief executive. •The signs in the parentheses under measurements represent the expected direction of the reisdonship. For data sources and mote detailed definitions of some of the variables, see Appendix B. 43 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chart 2.4: Trends in Welfare Expenditures for LDCs and Advanced Nations* 3 O 0 1 e i G o v e r n m e n t W e l f a r e E x p e n d i t u r e s 1 9 7 2 - 1 9 9 5 ( % o f G D P ) •Advanced Countries •LDCs « u Sources: International Finance Statistics & Government Finance Statistics (IMF: Various Editions) Significantly, contrary to the works of noted scholars such as Cameron (1978), Rodrik (1997 and 1998), and Garrett (1998a), total government spending is not used to assess the domestic effects of globalization in this analysis.1 4 Research has shown that increased government spending does not necessarily translate into lower market efficiency (Alesina & Perotti, 1996; Garrett, 1998a; and Slemrod, 1995). Therefore, if globalization is putting pressure on LDC nations to adopt a “pure” neoliberal agenda, explaining variations in government spending per se may not be instructive in this 1 4 Rodrik (1997) takes a brief look at social insurance and welfare spending as the variable to be explained, as does Garrett (1998b). However, Agenor & Montiel (1996); Cameron (1978), Garrett (1998a) and Rodrik (1996,1997) focus upon government spending as their dependent variable. 44 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. analytical context1 5 For example, higher proportions of the government budget may be allocated towards funding more “competitive” ventures, while government redistribution programs, as witnessed by the transfer budget, might be decreasing.1 6 Furthermore, data shows that the relationship between total government spending and spending on social security and welfare in LDCs is not highly correlated (see Chart 2.5). These two variables move in tandem less than 30% of the time in the period between 1972 and 1995. It is thus no surprise that the estimated covariance (at a mere -.00004) is close to zero. This shows that variations in government welfare expenditures cannot be explained by focusing on total government spending. The implication is that the politics of government spending is distinct from the politics of welfare spending in LDCs. 1 5 This finding of Garrett’s (1998b) and Rodrik’s (1996) analysis precisely illustrated the point that globalization does not necessarily have a negative effect on government spending. This research does not challenge these findings, only takes a different approach. It is a more rigorous test of the constraints on government freedom to pursue redistributive social policies in the era of globalization. 1 6 For example, Grunberg (1998) argues that due to the pressures of globalization, public authorities must subsidize some exporting firms in LDCs so that they can be at par with similar firms competing internationally. In such a case, the “industrial subsidies” aspect of government spending might be increasing over and above (or at the cost of) the transfer budget 45 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chart 2.5: Government Spending and Welfare Spending in LDCs* T o t a l G o v . S p e n d i n g v s . W e l f a r e S p e n d i n g i n L D C s ( a s % G D P ) 4.0 - I 3.5 • & 3.0 • a 2.5 ■ i 2.0 - £ > 1.5 • a 1.0 ■ 0.5 - 0.0 # ■ s ' " J" 35 30 25 20 15 10 5 0 •Welfare Exp -Total Gov spending a e ‘o e 8. (0 § (9 1 * See Appendix A for countries included Sources: International Finance Statistics &Govemment Finance Statistics (IMF: various editions) In essence, a more robust test of the conventional wisdom would be to isolate globalization effects on the portion of the general revenue allocated towards social spending, rather than analyzing total government spending. As Garrett (1998b:2) best puts it, the transfer budget is most vulnerable to globalization constraints because “its purposes are purely redistributive and are hard to justify in solely economic terms.” Thus, investigating national social security and welfare expenditures is one sine method of assessing government priorities in the era of globalization. 46 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Independent Variable: Economic Globalization in LDCs [TRADE and KFLOW] Both capital mobility and trade are important aspects of economic globalization for the LDCs. Several empirical studies have declared LDCs as de facto open in capital and trade markets (Amirahmadi et al., 1992; Dean, 1995; Faruquee, 1992; Ghosh & Gray, 1995; Mamingi, 1997; Montiel, 1994). The conventional measure of openness is implemented in the LDC model: exports + imports as a percentage of GDP. However, contrary to Garrett (1998a) and Rodrik’s (1997) use of capital controls to measure capital mobility in the OECD, capital mobility in LDCs will be measured by gross capital flows (inflow plus outflows). An admittedly imperfect measure of capital mobility, this decision was made for two reasons: (1) it is an exogenously determined measure of capital mobility, whereas capital controls are not and (2) data are provided annually.1 7 The primary purpose for using both capital mobility and trade is an attempt to make a direct comparison to the path of globalization observed in the industrialized economies. The relative concurrence of trade and financial liberalization between 1985 and 1995 for LDCs is in sharp contrast to the experience of the economically advanced nations. The “late integrators” (countries of Latin America, South Asia and some parts of Southeast Asia), almost unilaterally liberalized their capital or trade regimes or both (Dean 1995; Greenaway et al., 1997; IMF World Economic Outlook 1997). In contrast, 1 7 See Montiel (1994, 1996) for a discussion on the four different ways of measuring capital mobility, their pros and cons. As Montiel points out, each of these measures suffer from shortcomings. 47 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Garrett (1998a) reveals that, historically, there has been an inverse relationship between exposure to trade and capital mobility in developed countries. He finds that developed countries that were heavy traders, such as Austria, Denmark, Norway and Sweden, tended to maintain capital controls until the late 1980s. Conversely, countries such as the US and Japan that had more open capital markets exhibited much lower trade to GDP ratios. These dissimilar trends in economic globalization between developed and developing countries may help to explain the different trends in welfare spending (see Chart 2.6). Chart 2.6: Globalization in LDCs and Advanced Countries* T r a d e & C a p ita l R o w s in A d v a n c e d C o u n tr ie s & L D C s ( a s % G O P ) c 4 e 6 0 7 0 6 0 9 0 4 0 3 0 20 1 0 0 4------ 1 -------> -------:------ 1 -------1 -------1 ------ 1 - A * I c " I a a 20 O -Trade Flaw in A dvanced Cntre X Trade Rows in LD Cs - C apital Ftaw in A dvanced Cnlrs Capital Flow s In LD C s * See Appendix A for countries included Source: Balance o f Payments Statistics QMS'. Various Editions) 48 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Independent Variable: Factor Endowments and the Bargaining Power of Labor [UNSKILL and SURP] (the structural variable) A measure of the factor endowment of unskilled labor [UNSKILL] is used to estimate the bargaining power of labor in LDCs. Assessing labor power in developing countries is quite a formidable task. This is in contrast to the OECD countries where the tools for assessing bargaining power of labor have reached a much higher level of sophistication.1 8 The most common method of assessing labor’s bargaining power (in LDCs) is by unionization levels (Devarajan et al., 1997; Galenson, 1994, 1962; Western, 1997). While observing trends in unionization can produce some insights into labor’s political prowess, there is a need for a better proxy. This unionization approach is not the best method for making comparisons of labor strength in developing nations for two reasons. First, state corporatist arrangements in many LDCs render quantitative assessments of labor power a rather complicated affair. Lacking the same level of independence granted to unions in the West, LDC unionization rates can provide misleading readings of labor strength. LDC unions tend to be highly politicized, co-opted by political parties, the state and/or the company (Galenson, 1962). States are often inclined to impose constraints on their demand making, leadership and internal governance (Collier cannot be used and Collier 1 8 In addition to establishing an index of left power, Garrett (1998a) incorporates a measure of “encompassment” to capture the degree of centralization of labor unions. 49 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1991). Secondly, from a statistical perspective, unionization levels of LDCs on a comparative basis because standardized cross-country measures simply do not exist.1 9 At this juncture, it is vital to differentiate the historical organizational tendencies of “labor” in developed countries from LDCs. Traditionally, labor movements in developed countries found their origins amongst the working class in both heavy and light industries during the Fordist era. The recruiting grounds for labor organization in LDCs, on the other hand, generally occurred in heavy industries, amongst white collar workers, and public service employees (Deyo, 1989, 1984; Manning, 1998; Shafer, 1994). Unskilled laborers in LDCs, unlike those in developed economies, faced graver collective action problems and were very difficult to organize (Deyo, 1989; Gereffi, 1995; Henk, 1995; Ingersen, 1984; Lok, 1993; and Shafer, 1994). This divergence in organizational tendencies emphasizes the importance of disaggregating labor in LDCs into two groups: skilled and unskilled.2 0 When it comes to making political demands in LDCs, labor’s power is not homogenous. The bargaining power of labor has many determinants. Thus, in order to draw some conclusions about the bargaining power of labor in LDCs, we turn to the logic of the power resource model.2 1 This theory (power resource theory or PRT) asserts that if labor as a collectivity has access to power resources, then it has the potential to alter 1 9 In fact, as of yet, no cross-country measure of labor’s bargaining power in LDCs has been constructed 2 0 See fh. 15 for the definition of skilled and unskilled. 2 1 The “Social Democratic Model,” the “Working Class Mobilization Model,” and the “Political Gass Struggle” theory are all known as part of Power Resources Theory (Olsen & O’Connor, 1998, 22). 50 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. distributional inequalities to a significant degree.2 2 However, it is the skilled labor groups in LDCs, usually low in supply, that generally hold power resources. They tend to be more organized, face greater levels of demand— particularly in this hi-tech era of globalization, and tend to have access to an increasingly valuable commodity: information and knowledge.2 3 Unskilled labor groups, on the other hand, are confronted with higher labor market disequilibrium. They are subject to greater competition for employment and low human capital acquisition. As a result, these groups are rife with collective action problems and, thereby, tend to have very low access to power resources. Stated simply, it seems that the more unskilled laborers there are relative to skilled laborers in LDCs, the less bargaining power labor as a whole would have. It is essential to understand labor power in the aggregate since the extent of welfare coverage has traditionally been based upon coalition building between labor groups. Esping-Andersen (1990:30) states: It is a historical fact that welfare-state construction has depended on political coalition-building. The structure of class coalitions is much more decisive than are the power resources of any single class...For social democracy, the challenge was to synthesize working-class and white-collar demands without sacrificing the commitment to solidarity. 2 2 Note that PRT has traditionally used unionization rates or electoral durability of parties on the left to measure labor power. However, contemporary PRT works are beginning to acknowledge that these indicators may well be inappropriate measures of labor strength in the globalization era (O’Connor & Olsen, 1998). Note also that PRT has been applied only to a small number of developed countries thus far. This study is an attempt to fill some of these gaps in PRT. 2 3 industrialization (ISI) period when domestic industries enjoyed protection and predictability (Deyo, 1989; Henk, 1995). 51 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Using a classification scheme developed by Adrian Wood (1998), the LDC model includes a proxy for the bargaining power of labor in LDCs as the ratio of unskilled to skilled labor (UNSKILL). Albeit only an approximation, such a measure facilitates a country by country comparison of labor power in LDCs. This model accounts for the fact that structural differences in labor markets of developing nations have political repercussions. The higher the ratio, the less likely that labor’s demands will translate into desired government policies. This focus on labor employment in manufacturing is appropriate for LDCs for several reasons. From an economic perspective, manufacturing is the most dynamic sector in these economies (UN Industry and Development Report 92/93 and 93/94). Employment in this sector tends to have a multiplier effect upon employment in other sectors, particularly services (UN Industry and Development Report 92193: Chapter 3). Secondly, research shows that it is in this sector where most labor organizing activities occur. Wood’s classification scheme is derived primarily from the manufacturing sector of LDCs and differentiates employment in skill intensive industries from labor intensive industries. This proxy for labor’s bargaining power in LDCs, however, is insufficient without an added estimate for surplus labor [SURP] in the economy (measured by total labor force/arable land). Ingersen (1984:231) argues that governments enforce labor legislation in industries only when “the balance of labor supply and labor demand is shifted in favor of the workers.” As long as a large surplus labor population exists, labor’s political power will be negatively affected. On the contrary, if the ratio of 52 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. unskilled to skilled labor is high, yet surplus labor is low, then labor will have more bargaining power ceteris paribus. Governments will then be more likely to respond to labors’ demands. Certainly, this research does not challenge Garrett’s (1998a) measure of the bargaining power of labor in OECDs. The proposed measure of bargaining power of labor in LDCs is not expected to be significant for OECD countries. As argued by Pierson (1996), there is a need to focus upon non-labor groups as well when analyzing welfare state retrenchment in developed countries. The existence of preconditions for institutional structures in OECD nations adds a different level of complexity to analyzing welfare states in these nations. Garrett’s assessment of labor strength (i.e., political power of left labor parties and the structural strength of national labor market institutions), therefore, is a more appropriate predictor of labor’s bargaining power in the developed economies. Control Variables The control variables are drawn from the alternative hypotheses (i.e., the domestic, international and developed country centered) presented in Chapter I. Notably, most of the control variables for the LDC model are similar to Garrett’s (1998a, b) OECD model, with a few notable additions. The demographic variables are incorporated into the LDC model when applicable. However, given that most of these nations are considerably underdeveloped, supplementary political and economic variables are included in order to adapt the model to the LDC cases. 53 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. I. Demographic variables [DEPEN, URBAN] The social security literature emphasizes the role of demographic factors in explaining variations in the welfare state (Esping Andersen 1990; Hallen, 1967; Lindert, 1996; Schmidt, 1995; Tang, 1996; World Bank Policy Research Report, 1994), The number of dependents on the working population [DEPEN] (which is defined as the ratio of dependents-the population under age 15 and above age 65-to the working-age population— those aged 15-64) and the level of urbanization [URBAN] are included in the LDC model. Since these variables contribute to a larger labor force and/or create greater demand for non-wage benefits, they are expected to be positively correlated with welfare spending. However, unemployment figures, also linked to higher welfare expenditures according to this literature, were not included in the LDC model. This is because of the unreliability of such figures in underdeveloped nations (Agenor & Montiel, 1996; McDiarmid, 1977). II. Economic Development [GDP, SOE, DEBT] The most commonly suggested determinant of welfare state spending is the existence of economic surplus (Atkinson & Hills, 1991; Chatterjee, 1999; Devarajan et al., 1997; Glyn, 1998; Lindert, 1996; Rodrik, 1996; Stem, 1991; Tang 1996, Taylor- Goodby, 1997; Usui, 1994). The rationale is based upon Wagner’s Law (1883) which states that growth of the public economy is positively related with the economic affluence of the nation. This line of reasoning suggests that the relationship between GDP and state welfare expenditures will be positive. 54 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In addition to GDP, other economic variables must necessarily be included in the LDC model. Schwartz (1998) makes the important point that debt servicing [DEBT] should be included in welfare models, since sufficient state revenues is an unvalidated assumption incorporated into analyses of welfare state spending. If he is correct, then debt servicing might take priority over welfare spending, especially in LDCs. Another alternative pressure on the LDC budget is privatization [SOE].2 4 Given that there is no consensus on whether or not privatization leads to lower or higher government revenues, its potential effect on welfare expenditures lends itself to an empirical question. HI. Political Variables [DEMOC] Theorists argue that the more democratic a nation is, the more responsive the government must be to the demands of the populace (Atkinson, & Hills, 1991; Baneiji & Ghanem 1997; Eichengreen, 1996; Hicks & Swank, 1992; Stem 1991). This proposition implicitly reflects the median voter theorem. In other words, electoral turnouts tend to favor the median voter who, in turn, generally favors greater welfare spending (Esping-Andersen, 1996; Hicks & Swank, 1992). Accordingly, there should be a positive correlation between democracy and government social expenditures. 2 4 It cannot be denied that privatization is a massive force that is confronting developing countries, particularly since the late ‘80’s. Privatization can potentially affect state budgets in multiple ways. However, for the purposes of clarity and the goals of this paper, the focus will be on the revenue effects (either positive or negative) of privatizing state owned enterprises (SOE). 55 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The Results A variety of regressions were run to test the strength of the proposed hypothesis and the robustness of the results. First, panel regression techniques show that the exclusion of international variables biased the effects of some of the domestic level variables. A second set of regressions demonstrated that LDC welfare spending is quite sensitive to variations in international variables. Surplus labor and the interaction effect between labor power and globalization variables are highly significant and negative in an open economy LDC model. When all the relevant domestic variables are controlled for, globalization clearly has an adverse effect on LDC welfare expenditures. In contrast, a final set of regressions finds no impact of the labor skill distribution on welfare spending in Garrett’s 14 OECD countries. The final set of regressions appreciably strengthens the conclusions of this study. While all of the preceding results on LDCs present findings that are contrary to Garrett’s (1998a, b) on developed countries, they do not fully expose the explanatory power of the primary causal variable. This last set of regressions verify that it is the political-structural difference between rich and poor nations that marks the variation in domestic outcomes of globalization for developed verses developing countries. The Closed Economy Model o f LDCs This purpose of this model is to test the empirical validity of previous analyses that focus on domestic level explanations and exclude the consideration of globalization 56 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. variables in their assessments of LDC welfare spending.2 5 The idea is to assess whether and how the significance of domestic level variables differs when international variables are taken into account. Table 2.2 presents the first set of regression results where b^and b3 (or UNSKILL and SURP) are < 0, as expected. Table 2.2: Regression results for LDCs in a closed economy model Independent variable Parameter Estimate Standard Error welf (lagged) 0.579*** 0.025 Unskill -0.216*** 0.067 Urban -0.405** 0.180 Surp -0.131* 0.077 gdp per capita -0.162*** 0.057 Democ 0.017** 0.007 Country effects Yes Year effects No Rz .916 N 1272 F value 4.73 Dependent variable: social security and welfare expenditures (LDCs closed economy) Fixed effects regression estimates. Figures in parenthesis are standard errors. ***p<0.01; **.01<p<0.05; *0.05<p<0.10. 2 3 This applies to both qualitative and qualitative works on the determinants of LDC welfare spending. 57 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Before discussing the above results, it is important to point out that if there actually are globalization effects on welfare, this model is not the correct one and is incapable of capturing the true relationship between welfare spending and domestic level determinants. Proceeding with the assumption that there are no international effects that are correlated with the included explanatory variable, the results in Table 2.2 show that the measure of LDCs factor endowments (UNSKILL and SURP) are very significant in explaining the level of welfare expenditures. Most of the control variables behave as expected. The most important variables are lagged welfare spending and GDP per capita. Interestingly, these results confirm the findings of Esping Andersen, Rodrik and Cameron who also reported the democracy variable and GDP per capita as positive and significant. Thus, the estimations support both the median voter theorem and Wagner’s law, respectively. Year dummies are not included in the closed economy model because they did not have a significant impact on LDC welfare expenditures. This suggests that shocks (with long run effects) to the economy, such as recessions and/or crisis, do not cause variations in overall LDC welfare state expenditures. The coefficients on age 58 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. dependency, debt and privatization variables are also insignificant, and therefore the results are not reported.2 6 The negative coefficient on urbanization, however, seems counterintuitive and stands contrary to the report by Tang (1996). In his cross-sectional analysis of 29 developing countries, Tang (1996) found only the urbanization variable to be significant and positively related to welfare spending. It is true that by contributing to an expanding workforce, urbanization should have a positive association with welfare. However, as some scholars like Paul Barioch (1973) and the World Development Report (1995) discuss, mass urbanization is a big problem in LDCs because these economies cannot maintain the necessary infrastructure to support it. Results from this model suggest that in the event of rapid urbanization, welfare spending may suffer at the expense of other government outlays. The reason for the discrepancy between these results and Tang’s is most likely that his analysis did not take country differences into account and thus suffers from heterogeneity bias. The Open Economy Model o f LDCs Results conditioned upon both international and domestic effects are reported in Table 2.3. Since scholars argue, based on the Hecksher-Ohlin model, that increasing 2 6 Data on privatization and age dependency are not gathered annually. Therefore, in order to obtain the best results, the dependency variable was re-estimated two different ways: by interpolation and by five year averages. Neither estimation was significant The coefficient of the privatization variable, however, must be interpreted with caution, since this data was only available on a much smaller subset of LDCs. Because of limited data, 5 year averages were estimated on only 25 LDCs. Thus, degree of freedom was compromised. 59 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. employment in labor intensive industries is associated with the level of globalization in LDCs, interaction terms between globalization and the factor endowment of unskilled labor are included. Thus, specification of an interactive variable is designed to capture the effects of the Hecksher-Ohlin theorem. The open economy model drops the assumption of a closed system and includes the international market variables. This is a fully specified model. If the LDC welfare expenditures are indeed a function of global market activity, then b < and bs (or the interactive effects of TRADE and UNSKILL and KFLOW and UNSKILL) will be negative. It is interesting that incorporating the effects of both trade and capital mobility diminished the significance of the GDP variable, but increased the significance of the surplus labor and democracy variables. The influence of urbanization, outweighed by the impact of international variables, dropped out of the model altogether. Thus, excluding the international variables in Model 1 biased the effects of some domestic level variables downward and some upward. While the value of the R2 should not be exaggerated in pooled time series and cross section data, it is helpful to note its level of increase from Model 1 to Model 2. These estimates cast doubt on previous analyses that have downplayed the importance of international level variables in their analyses of LDC welfare spending (Atkinson & Hills, 1991; Midgley, 1984; Schmidt, 1995; Tang, 1996; Wahl, 1994). The results in Table 2.3 show that over time and across LDCs, welfare states are very responsive to globalization variables, or specifically the interaction effects (of 60 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 2.3: Regression results for LDCs in an open economy model Independent Variable Parameter Estimate Standard Error welf (lagged) 0.343*** 0.031 trade*unskill (interactive variable) -0.160*** 0.026 kflow*unskill (interactive variable) -0.152** 0.040 Trade 0.019 0.012 Kflow 0.150*** 0.052 Surp -0.230** 0.099 gdp per capita 0.129* 0.068 Democ 0.026*** 0.010 Country effects Yes Year effects No R2 .987 N 1272 F value 8.88 Dependent variable: social security and welfare expenditures (LDCs open economy) Fixed effects regression estimates. Figures in parenthesis are standard errors. ***p<0.0l; **.0Kp<0.05; *0.05<p<0.10. 61 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. globalization Gagged) and the factor endowment variable).2 7 As hypothesized, b4 and bs are negative and significant at the 99% confidence level. That is, a relative increase in exposure to competitive global markets motivates a substitution away from expenditures towards labor. All the measures except capital flows have the anticipated signs. One explanation for the positive coefficient on capital flows could be, as recent proponents of globalization argue, that capital mobility brings in more revenue.2 8 Governments might be more willing to spend on welfare as their revenues increase. It is possible that the increase in economic resources encourages greater welfare spending. However, more relevant to the purposes of this analysis, social spending responds negatively to lagged increases in globalization and its interactive effects with unskilled labor. The estimated coefficients suggest that an increase in the share of trade and capital (as percentage of GDP) of 10 percentage points respectively, results in reduced government social spending by about 1.5 percentage points. Furthermore, the interaction effect between unskilled labor and globalization variables remains significant at the 99% confidence level. UNSKILL was also included as a separate regressor, but since the coefficient was statistically insignificant, it is not reported here. n The regressions were run on both lagged and non-Iagged international variables. Both results were significant However, it is theoretically more sensible to anticipate a lagged effect because of job shifting and retraining, and resources allocation (see Wood’s 1994). Also see Rodrik (1997) who uses lagged measures of openness. a Understanding this particular relationship between capital flows and welfare spending involves an investigation into the ‘revenue’ side of the public budget via taxation sources and rates (see Garrett 1998a). This analysis focuses only on the expenditure aspect of the public budget 62 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The F-tests reject the null hypothesis that country differences are not important. This suggests that country specific differences persist over time. Yet it is rather telling that even though these historical differences account for much of variations in welfare expenditures, the globalization and factor endowment variables still maintain high statistical significance. This fixed effects estimation procedure thereby increases confidence in these results. To summarize, this equation tests whether or not variations in welfare state expenditures depend on the interactive effect of globalization variables and the employment of unskilled labor. As hypothesized b$ and their product terms are negative (and significant) in nations exporting more labor-intensive goods. Notice that, once again, the year effects were insignificant This suggests that conditions in the world economy do not necessarily affect LDC welfare spending. One possible explanation for this is that in order for social spending to be affected by global shocks and/or recessions, nations must spend a much larger percentage of their GDP on welfare. The positive and significant coefficient of the democracy variable underscores the fact that political variables are a primary determinant of LDC welfare spending and, thereby, the model is not an economically deterministic one. Nevertheless, the elasticity of welfare spending suggests that increasing economic globalization, as it interacts with labor’s bargaining power, is still one of the fundamental determinants of LDC welfare states over and above some of the important domestic level variables. Simply put, the above results indicate that the effects of international markets on LDC welfare states are quite strong and should be heeded by scholars and policy makers alike. 63 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The Comparison Between LDCs and OECDs Using the same OECD countries as Garrett (I998a,b), Equation 2 is re-applied to OECD welfare states in order to assess whether UNSKILL has a similar affect on advanced countries. This final set of regressions determines whether it is the factor endowment variable that produces the differential (domestic) effects of globalization for developed and developing countries. There is scant evidence to suggest that the factor endowment of unskilled labor has any impact upon the politics of welfare spending in developed countries. Yet, interestingly, these results confirm Garrett (1998a) findings of the significant effects of globalization, as well as all of the same control variables upon welfare spending in OECDs. Table 2.4 shows how these variables perform in the open economy model of developed countries. The industrialized nations demonstrate the exact opposite results to LDCs. The primary independent variable, unskilled labor, is not a fundamental determinant of OECD welfare states. The effect of surplus labor on welfare states is similarly negligible. Therefore, in a basic model of OECD welfare states, the number of unskilled laborers do not matter to the welfare states of developed countries. Significantly, year effects did matter for welfare expenditures in OECDs while they had no affect in LDCs. This lends further credibility to the supposition that the more a nation spends on welfare as a percentage of GDP, the more susceptible their social programs will be to shocks in the world economy. Developed countries tend to spend, on average, about 15% of their GDP on welfare expenditures. 64 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. It is a little surprising that the level of democracy exhibits only some significance in advanced countries when international variables are considered. This is perhaps because of the way in which democracy is defined in this model. The Gurr and Jaggar’s (1994) measure of political development in Polity HI is derived from the competitiveness of political participation, the openness of executive recruitment and constraints on the chief executive. While such broad measures of democracy might be telling in the developing country cases, it seems more relevant to implement a more nuanced measure of democracy in the OECDs as did Esping-Andersen (1990), Lindert (1994) and Garrett (1998a). It is anticipated that the level of statistical significance will be higher when this variable is refined and is able to capture political partisanship and leftist power in these countries (see Esping Andersen, 1990; Garrett, 1998a). 65 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 2.4: Regression results for Advanced Countries in an open economy Independent Variable Parameter Estimate Standard Error welf (lagged) 0.594*** 0.040 trade*unskill (interactive variable) -0.037 0.030 kflow*unskill (interactive variable) -0.011 0.040 Trade 0.178*** 0.046 Kflow -0.023* 0.012 Old -0.637*** 0.127 gdp per capita 0.090** 0.044 Democ 0.488* 0.283 Unemp 0.085*** 0.017 Country effects Yes Year effects Yes Rz .987 N 1272 F value 8.88 Dependent variable: social security and welfare expenditures (Advanced Countries: open economy) Fixed effects regression estimates. Figures in parenthesis are standard errors. ***p<0.01; **.01<p<0.05; *0.05<p<0.10. 66 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Contrary to LDCs, b4 . bs and their product terms are not significant in nations exporting relatively more skill-intensive goods. These results were expected for several reasons. First, recall that the UNSKILL variable is derived from the manufacturing sector of each country. Manufacturing no longer serves as the hotbed of labor activity in developed economies (Galenson, 1994; Pontusson 1995). The “deindustrialization” phenomenon spreading throughout the industrialized world marked the shift from employment creation in manufacturing and agriculture to the service industries (UN Industry and Development Report 92/93). Secondly, it is not as necessary to consider separate supply and demand curves for skilled and unskilled labor in OECD economies, particularly since their surplus labor problems are not as acute. Finally, the significance of the OLD and UNEMP variables confirm Pierson’s (1996) argument that there is a need to focus on non-labor groups when analyzing contemporary welfare states in developed countries. In sum, the empirical evidence in Table 2.4 validates the theory that OECD welfare spending is unresponsive to the interactive effects of globalization and unskilled labor. Expectations about the inappropriateness of applying this measure of labor strength to assess OECD government welfare expenditures in this era of globalization are verified. Garrett’s more fully specified model, which demonstrates the positive effects of globalization in countries that do not enjoy a comparative advantage in labor, should be taken seriously. Thus, the statistical results presented in this paper endorse the hypothesis that the political-structural difference between developed and developing countries affects different domestic outcomes in the contemporary era of globalization. 67 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Implications The evidence presented in this chapter indicates that LDC welfare spending is quite responsive to globalization variables, as well as certain domestic level factors. These findings are in direct contrast to empirical works by scholars such as Garrett who have recently argued that international market integration actually induces greater government spending on programs that reduce market generated inequalities in advanced nations. This analysis demonstrates that globalization has a drastically different impact upon less developed countries. Stiff international competition for capital and low wage labor markets abets certain constraints on the governments of labor abundant economies that are absent in OECDs. Because globalization induces greater employment of unskilled labor which is more difficult to organize, LDC governments have greater incentives to replace labor benefits with spending on ventures that attract international capital. How is this analysis consonant with the conventional wisdom that predicts the end of welfare states with globalization? The empirical demonstration that labor’s bargaining power is weakened in LDCs lends some support to conventional assumptions. These theorists commonly presume that globalization significantly increases the bargaining power of capital, vis a vis labor. The LDC model illustrates that this dynamic is particularly feasible where governments do not have revenues to both compensate those disadvantaged by globalization and promote national competitiveness. 68 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Put differently, one valid interpretation of the empirical evidence might be that capital has indeed gained bargaining power at labor’s expense in LDCs. As Grunberg (1998) summarized, LDC governments are finding it difficult to generate revenue and satisfy the demands of investors. This implies greater “voice” or bargaining power for sectors holding foreign exchange, liquid asset-holders, export sectors, private foreign creditors, foreign financial intermediaries, and multinational financial institutions (Haggard & Maxfield, 1997). With labor’s bargaining power weakened by globalization, international capital mobility could well intensify government responsiveness to capital demands and redirect the distribution of scarce government resources away from labor. In turn, the net effect would be greater incorporation of LDCs into the global economy. Thus, the next step in this research agenda would be to investigate whether or not the level of globalization of LDCs has, in fact, increased as a response to lower allocations of fiscal resources to labor. It is anticipated that this project will significantly enhance the knowledge of the domestic consequences of globalization by incorporating developing countries. At best, this analysis mitigates the bias of previous studies in both IPE and comparative politics that exclude LDCs from their realm of investigation. An exploration of the globalization-welfare nexus in states that do not have either sufficient revenue flows to provide a high social wage or formalized democracies and, thus, the institutional structures necessary for dialogue between societal pressure groups and public policy makers, has been long overdue. 69 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The ultimate aim of this investigation, however, is not to settle the question of the political-structural determinants of the decline of LDC welfare state. By establishing a comprehensive account of the decline of LDC welfare states, this analysis is only a first attempt at understanding the pressures of globalization faced by these nations. Heterogeneity amongst LDCs is not to be underestimated and, thereby, the dangers of considering LDCs as a unit are fully recognized. Rather, the goal is to initiate a trend of research into the international sources of LDC government redistribution polices and, at the same time, to propose a proxy for the bargaining power of labor in LDCs usable for comparative purposes. The next step in this analysis is to apply the comparative method to provide further evidence to support or reject the hypothesis in this chapter. By selecting a representative case from the LDC sample and exploring why their social welfare expenditures have been declining, the various phases of the policy process can be understood. Following, comparing this standard case with two other countries which have had increasing welfare expenditures will pose an interesting check on the validity of the primary hypothesis. It is hoped that evidence from these case studies will deepen our understanding of how and why (or why not) globalization is linked to the welfare states in LDCs. 70 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 3 CASE STUDIES: GLOBALIZATION AND SOCIAL WELFARE IN INDIA Three case studies have been chosen in order to offer critical insight into the quantitative results in Chapter 2. The quantitative results showed strong statistical support for the conventional hypothesis in LDCs. The purpose of the case studies is two- fold: (I) to determine how well the LDC model and the estimated coefficients fit and account for relevant welfare developments in developing countries; and (2) identify relevant variables that might have been excluded from the LDC model. In particular, detailed case studies will shed greater light on both the primary and secondary findings that emerged from Chapter 2. The main focus of the case studies is to further investigate the interactive effects of globalization and the bargaining power of labor, which emerged as the most significant variables in the quantitative analysis (also with the largest estimated coefficients). At the same time, the significant control variables that were not central to the primary hypothesis are given closer attention in this section.1 These are, specifically, the country specific differences and the level of democracy. 1 Note that GDP is not included here, though it was found significant in the quantitative analysis. Instead, by selecting three countries that have been experiencing increased GDP over time, GDP is controlled for in the qualitative analysis. 71 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Indeed, one of the primary advantages of the fixed effects method used in the quantitative analysis is that it controls for differences between countries without having to model them explicitly. Yet beyond suggesting that the presence of country- specific differences persist over time, further interpretation of these dummy coefficients is rather limited. The best way to determine which country specific effects account for variations in national social policies and supplement these quantitative results is to conduct a relatively intense and systematic analysis of a small number of cases. Case studies will also provide greater insight into the significance of the democracy variable. Even though international markets for capital and low skilled labor exports puts pressure on labor and intensifies collective action problems in LDCs, Chapter 2’s econometric estimates revealed that regime type still matters. Authoritarian governments were less responsive to labor interests than democratic ones. Essentially, a more detailed look at select countries will be the most appropriate way to understand the various phases of the policy process and assess how they differ from one regime to another. This would not only provide further evidence supporting or rejecting the proposed hypotheses, but also help to trace “the decision process by which various initial conditions are translated into outcomes” (George 1979:43). It will be interesting to note which significant variables emerge from the process tracing method, and how they connect to the model in Chapter 2. 72 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. To explore more deeply the significance of the explanatory power of the proposed hypothesis and trace complex decision-making processes, three countries have been selected: India, Brazil, and Korea. India was selected to typify countries in the full LDC sample that have been experiencing declining welfare spending after policies encouraging globalization were adopted (which was the mid-eighties in the Indian case).2 Brazil and Korea were chosen because their increasing welfare expenditures mark them as deviant from the average trend in the S3 LDCs (please see Charts 3.1, 3.2 and 3.3). The factors leading the latter two countries to become exceptions to the norm (in LDCs) will be investigated. In addition, because both Brazil and Korea experienced relatively recent changeovers to democracy, the importance of this variable can be carefully evaluated. The bulk of the qualitative Chart 3.1 Social Security and Welfare in India Qovernat ant Spanning on Social Security and Welfare lit India a* «y i “ 5 oj 1 •• * . . i t p i ' i rtv a t i n i t u i*io ee-u rear Source: IMF, Government Finance Statistics: various editions 1 Policies encouraging globalization include government efforts towards both capital and trade liberalization. 73 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chart 3.2 Social Security and Welfare in Brazil a 12.000 ovemment Spending on Social Security and W elfare in Brazil i 10.000 ' i 1 1 ! ] A L 8.000 ' ” 6.000 ' 4 .0 0 0 2.000 ' 1 I i 1 ! Yam Source: IMF: Government Finance Statistics: various editions Chart 3.3: Government Welfare Spending in Korea Government Spending on Soelol Security end W elfare In Y oirs * Note that the IMF data on Korea recorded a six fold jum p in spending for the 1982 year. The speculates that this was a result o f a data calculation and/or recording error. Source: IMF, Government Finance Statistics: various editions Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. analysis will be devoted to India since it represents the full sample of LDCs in this study. India’s case will then be placed into comparative context by briefly investigating Brazil and Korea. Ultimately, the comparative method usefully complements the extant literature, as Kurt Weyland (1996) argues in the Latin American Research Review with reference to Brazil: [Systematic cross-country comparisons would greatly benefit “Brazilian studies,” which has remained a fairly insular field in which single-country analyses have predominated. Even Brazil’s special features become distinct only when contrasted with other nations...Although in-depth case studies are difficult to conduct in more than one country, they must be placed in comparative context. It is certainly not possible to be an expert on so many cases. The aim, however, is not to capture reality but rather to structure and simplify the analysis in a theoretically useful way. Some reductionism will be necessary in the effort towards theory formulation.3 It is hoped that the combination of both qualitative and quantitative methods will help us to interpret and understand complex social phenomenon. 3 For example, globalization in the statistical model referred to a process involving increasing trade flows and rapidly increasing in international financial flows, gaining momentum in the time period after Bretton Woods (1972 to 1995). In the case studies, the analysis of globalization will be confined to the distinct point in time when a structural change occurred in the data for trade and capital flows. 75 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The findings from the qualitative section answer the key question of why welfare expenditures in most LDCs have declined with globalization, while in a few select cases they have increased. AU three cases confirm that the interaction between globalization variables and the bargaining power of labor do have an effect on government welfare expenditures. The Indian case well illustrates the primary hypothesis by tracing the decline of welfare spending to expanding unskilled labor interacting with globalization. This case, however, does not confirm the proposition that democracy has had a positive effect on social spending. Brazil and Korea, on the other hand, provide evidence supporting both the globalization-unskilled labor and the democracy propositions. Welfare spending in both countries rose, as they underwent a changeover to democracy and as the exposure to globalization and the proportion of skilled labor increased.4 The interesting difference between the Korean and Brazilian case is that while both experienced upward trends in welfare spending, the level of government welfare spending as a percentage of GDP in Brazil was substantially higher than Korea’s (and India’s). Additional explanatory variables affecting LDC welfare expenditures that emerged from these case studies were the historical legacies of the state-labor relationship in each of the three countries, the rise of the informal sector, and welfare distribution issues associated with state corporatism. 4 Another way to put this is the proportion of unskilled labor fell and thus, the bargaining power of labor increased. 76 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The Qualitative Methodology Each of the following three case studies follows a similar format. The first part of each country study begins with a detailed exploration of the dependent variable, or the government’s commitment to social welfare. Second, the historical evolution of policies and actual levels of the globalization variables, or trade and capital flows, are traced. Finally, the third part of each case study focuses on how the interaction between globalization and the bargaining power of labor has led to either a decrease or increase in welfare expenditures. The variable of interest in this study is the LDC state’s commitment to social welfare.5 In the statistical analysis, this dependent variable was represented by the trend of government welfare expenditure (as a percentage of GDP) over 23 years and across countries. While this is the best statistical indicator of the state’s commitment to welfare, the dependent variable can be defined more broadly in the qualitative investigation. In other words, aside from a closer look at changes in government social expenditures, both the types of social welfare programs and labor market policies will be assessed. This particular method will provide a broader indication of state social intentions. Thus, the dependent variable will be assessed as follows: (1) a the level and trend in welfare spending will be analyzed; (2) a brief review of each countries welfare programs will be conducted; and (3) an investigation of public policies toward labor (e.g., the right to strike, freedom of association, legal provisions for collective 1 As emphasized in the quantitative analysis, social commitment refers specifically to government policies/actions that promote redistribution towards lower income groups. 77 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. bargaining, lay-offs, etc). Spending on education and health are not included in the welfare definition applied in this analysis because they are considered less redistributive.6 Both government actions and policies will, thereby, be used to deduce government’s social welfare commitments. Government “actions” can be inferred by the level of social welfare expenditures, while the focus on ‘policies’ will be those directed towards social welfare, collection action and/or legislation governing industrial relations. This analysis posits that an examination of a combination of these factors is needed to evaluate national social policy goals. Based on the logic of power resource theories, the latter two variables are included to capture the state-labor relationship which, in turn, is likely to influence the size of the welfare state.7 The assumption is that a structural shift in welfare expenditures indicates that a A government’s commitment to social welfare goals has been significantly affected. It is expected that investigating the source of this change will give valuable insight into 6 Spending on health and education is an important element of social welfare and redistribution policies that will be addressed in future studies. These types of social spending would involve an endrely different dynamic that is beyond the scope of this dissertation. In fact, many argue that health and spending are productive capital investments (i.e., human capital) and, therefore, are not redistributive in the purest sense. 7 As noted in Chapter 2, one of the shortcomings of power resource theories (PRT) is that they do not apply their analysis to the welfare states in LDCs. Recall that PRT states that labor groups with command over power resources can influence the size of the welfare state. Applying the logic of this theory to LDCs requires that state-labor relations must be taken into account since labor tends to be the most powerful pressure group with respect to the welfare states, and because states tend to take dominant roles in the economy of late developing countries. Therefore, while welfare states benefit not only labor, but also other vulnerable groups, PRT asserts that labor power is critical to providing social welfare institutions. Notice that most of the benefits in LDCs are tied to employment (e.g. social security, unemployment, family allowances, unionization, etc). 78 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the validity of the hypothesis proposed in this thesis. In other words, it asks the question was globalization indeed an influential force in causing this change in LDC government welfare expenditures? The extent of the structural shift will be further surveyed by investigating concurrent changes in social policy, and policy reactions toward labor, or both. Significantly, the years surrounding the structural shift in social welfare expenditures will be the primary focus of the investigation. The primary causal variable-the interaction between globalization and the bargaining power of labor—and its effects on government social commitments will be traced over time. Whether or not changes in the primary causal variable occurred in tandem with structural shifts in government social commitments will be of fundamental interest. Slightly different from the statistical analysis, the effects of globalization will be assessed during the particular years (after 1972) when trade and capital flows begin to show substantial increase. It is important to recognize that the endogenous and exogenous nature of globalization is difficult to disentangle in the qualitative analysis. Because there is a strong correlation between levels of globalization and the liberalization decisions made by governments, globalization is in one sense endogenous.9 On the other hand, liberalization efforts are important components of the process of globalization. Many theorists argue that nations are “forced” to liberalize in this era of globalization in order * A structural change can be assessed by a major shift in the level of expenditures, rather than incremental change. 9 For example, Garrett (1999a) argues that government decisions to lift capital controls are correlated with globalization. 79 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. to attract capital flows and to remain competitive in export markets. From this perspective, globalization is clearly an exogenous phenomenon. Taking both sides into account, this analysis posits that liberalization strategies and levels of globalization are intertwined and likely to be mutually reinforcing. Changes in decisions to liberalize the domestic economy are strongly influenced by globalization, and vice-versa. The point here is that liberalization decisions are an important indicator when assessing the levels of globalization. The concept of bargaining power of labor is more difficult to qualitatively assess in developing countries. Encarnation (1989) correctly points out problematical nature of this concept since claims using it tend to be tautological.1 0 Thus, concurring with Encarnation (1989:20), bargaining power in this analysis refers to the ability of laborers to “improve the range of plausible outcomes available to each [negotiator], and to improve the probability of securing the outcome that each prefers.” In the statistical analysis, the number of unskilled laborers relative to skilled laborers measured labor strength. As this ratio increased, it was reasoned that labor as a whole would And it more difficult to organize and bargaining power would fall. The qualitative investigation more extensively assesses bargaining power by reviewing their ability to 1 0 Encarnation (1989) argues that “bargaining power” tends to be tautological because it is generally defined by the outcome. It is difficult to tell which party had more bargaining power if negotiations are “ won by those who win.” (also see Schelling I960). Encamation’s (1989) argues that this approach makes it difficult to differentiate between power and negotiated outcomes. 80 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. strike, numbers unionized, numbers employed in unskilled labor1 1 and policy outcomes affecting labor. The qualitative analysis is by no means conclusive. This is an exploratory investigation that seeks to further understand the significance of the variables applied in the econometric estimations. The ultimate objective of this section is to assess how the proposed variables perform within the reality of different country dynamics and to determine if there are alternative variables excluded from the primary hypothesis. The Dependent Variable: Government Social Welfare Commitment Comparative case studies reveal that the extent of government welfare commitment can be determined in three ways: levels and types of social expenditure, its changes or trend over time, and by government social and/or labor policies. Recall that the statistical analysis focused only on welfare spending trends over time. However, the case studies reveal that the distinction between trends and levels is an important factor to be considered in understanding the politics of welfare states. A country’s trend in social welfare spending might be increasing, yet the overall level of spending may still be low relative to the other countries. Observing both the trends and levels of government social welfare spending presents the first exploration of government social commitments. The primary question in this section is whether or not spending has been increasing or decreasing over time 1 1 This analysis applies the term low-skilled and labor-intensive goods/industries interchangeably. By the same token, skill-intensive, capital intensive and high-technology goods/industries are also used 81 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (1972-1995), though it remains significant if a country’s absolute level of spending is high or low. Ultimately, it becomes clear that the degree of the government’s social commitment (high or low) is quite complex. After understanding each country’s trends and levels of social welfare spending, a brief description of each countries social welfare programs and, next, their labor market policies are explored. This method of triangulation provides important benchmarks for comparison of the different welfare systems in the three countries. Exploration of the dependent variable in each of the countries reveals that India’s overall commitment to welfare is moderate and more similar to Brazil than to Korea. The level of welfare expenditures as a share of GDP in India is low and displays a declining trend since the early nineties. Yet, it turns out that the pursuit of social welfare in India is ultimately manifested in its strong policies favoring labor, not the state’s financial contribution to social welfare. Spending levels since the 70’s have been meager compared to both Brazil and Korea. In contrast to India, both Korea and Brazil have been experiencing upward trends in government welfare expenditures. Brazil stands as the most committed to welfare by maintaining the highest levels of spending relative to GDP. With 60 percent of its labor force covered by welfare, it also enjoys one of the largest percentage of the population receiving benefits in LDCs. However, an analysis of its corporatist system reveals that its labor market policies are weak and oftentimes repressive. Korea, represents the opposite extreme, and is considered to be a ‘welfare laggard’ by world standards. Relative to its level of interchangeably. 82 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. economic development, Korea’s social expenditures are far smaller than other upper- middle income countries (but higher than India’s and lower than Brazil’s). What is significant, however, is that neither Brazil nor Korea maintains the same level of regulatory protections for labor as India. See Figure 3.1 for a matrix of government social commitments in India, Brazil and Korea. Figure 3.1: Government Social Commitment Matrix for India, Brazil and Korea High social welfare benefits Low social welfare benefits 83 High labor market protections Low labor market protections Brazil India Korea Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The Indian Case The Indian case supports the primary hypothesis that welfare expenditures began to decline after its government policies shifted towards global market integration in the 1980s. Conversely, as long as India maintained its tradition of isolation from global economic forces, welfare expenditures were increasing. Most of this shift in the welfare spending trend can be attributed to labor’s bargaining power, which became progressively weaker as the numbers of unskilled labor increased with globalization. In addition, case evidence was found for three secondary factors that were not included in the quantitative analysis. The following variables also had an impact on the change in India’s social commitment: the historical state-labor relationship; the government’s ideological commitment to labor, and the expansion of the informal sector. India’ s Government Social Commitment Trends and Levels o f India's Social Welfare Spending India represents the average social welfare trend of the S3 LDC’s analyzed in this study. The majority of LDCs displayed a decline in government social spending in this era of globalization. Data on India indicates that its social spending as a percentage 84 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of GDP increased at first, but eventually fell over time. Even though real welfare spending in India has been increasing over time (see Table 3.1), it is the relative size of the social welfare sector that is of most interest here.1 2 India’s government welfare expenditures as a share of GDP rose steadily through the early 1980s, almost doubling the amount spent in the 1970s. After the late eighties, however, the data show a structural change in India’s welfare spending. Chart 3.1 illustrates a distinct decline in welfare spending as a percentage of GDP in the nineties. To be sure that trends in Charts 3.1, 3.2 and 3.3 correlate with government spending commitments and are not merely reflective of fluctuating growth rates, Chart 3.4 takes a second look at government social commitment. Here, welfare spending is recalculated as a percentage of total government spending. This chart reveals that a structural shift in India’s welfare commitment occurred earlier in the mid-eighties, closer to the period of policy reversals towards globalization. Drawing a conclusion from both charts (3.1 and 3.4), it is the period between 1985 and 1992 that is of most interest to this analysis. This time span covers the structural change in welfare spending, measured both as a percentage of GDP and as a percentage of total government spending. Interestingly, when calculated in the alternative way, both India and Korea’s welfare trends more or less mirror their 1 2 India’s octal welfare figures in terms of GDP and total government spending suggests that although real welfare spending may be increasing, welfare expenditures are not keeping pace with GDP or government budget priorities (see Charts 3.1 and 3.4). It is the latter two measure of “welfare commitment” that is of most interest to this study. 85 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. welfare to GDP ratios. Chart 3.4 shows that welfare spending trends in Brazil, however, now appear to be decreasing instead of increasing. Table 3.1: Real Government Welfare Spending in India Year Real Welfare Spending* 1974 7.4174 1978 10.829 1984 26.557 1988 32.951 1990 44.980 1994 49.128 'F igures are in 100,000s 'Consumer Price Index (CPI) taken from World Bank Development Indicators (1997) Source: IMF: Government Finance Statistics. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chart 3.4: Social Welfare Spending in India, Brazil and Korea* G o v e r n m e n t W e l f a r e S p e n d i n g ( a s % o f T o t a l G o v . S p e n d i n g ) 1 9 7 2 - 1 9 9 5 I 0.04 & & S? & # — Indls — Korea ■ ■ Brazil Years shown are averages based on data availability. Source: IMF, Government Finance Statistics: various editions An analysis based solely on social welfare spending alone would suggest that the Indian government is not very committed to national social welfare needs. The level of benefits itself is very small, averaging only about 0.6 percent of GDP. Yet moving beyond these more popular measure of welfare state commitment reveals otherwise. Although state financial contributions overall have been minor, social welfare has been an integral part of its national development strategies. The Indian government’s commitment to social welfare is revealed in its pro-labor legislation. In comparison to Brazil and Korea, India emerges as the most ideologically committed to welfare. 87 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The Characteristics of Indians Social Welfare System India’s emphasis on welfare is enshrined in its 1947 Constitution and was an integral part of each of the five-year economic development plans for the next four decades. The Directive Principles of State Policy contain many provisions with regards to social welfare, though they are not enforceable by any court of law. Fifteen Articles (Articles 36 to 51) in the constitution are devoted to social welfare. Article 38 best summarizes the State’s social philosophy: The state shall strive to promote welfare of the people by securing and protecting as effectively as it may a social order in which justice, social, economic and political, shall inform all the institutions of the national life (as quoted in Kohli & Sharma, 1996). Article 41 goes on to propose that the State shall direct its policy towards securing social protections. The state shall, within limits of its economic capacity and development, make effective provision for securing the right to work, to education, and to public assistance in cases of unemployment, old age, sickness, and disablement, and in other cases, of undeserved “wants” (as quoted in Momoria & Momoria 1983:5). The subsequent Five-Year plans were launched in an effort to realize the goals and aspirations enshrined in the Constitution. The inclusion of social welfare in the national development plans suggests that providing a minimum level of social security for the general poor was indeed a matter of national concern. Emphasis on the “less 88 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. privileged classes of society,” however, did not occur in the first two plans. Rather, these two Five-Year plans presented the ideological goal of social development alongside economic growth. For example, the first Five-Year plan (1951-56) called upon the State “to play an increasing role in providing (social) services for the welfare of people” (Bose, 1970). It also recognized that “the employer-employee relationship is essentially a partnership to promote the community’s economic ends, workers rights to association, organization and collective bargaining” (Mamoria & Mamoria 1983:341). To quote an instance, the Second Five-Year Plan states: The task before an underdeveloped country is not merely to get better results within the existing framework of economic and social institutions, but to mold and refashion these so that they contribute effectively to the realization of wider and deeper social values (Kohli & Sharma, 1996a). While the first and second year plans did not refer specifically to the “poor,” subsequent sections describe the programs and policies pertaining to welfare of women, children, under-privileged groups and the handicapped (Bose, 1970). It was not until the Third Plan (1961-1966) that the state began to recognize the need to eradicate “absolute poverty.” It argued for the need to target benefits for groups “altogether lacking in the means of livelihood and support” (Appasamy et al., 1996). The Fourth, Fifth and Sixth plans (spanning the years 1969 to 1985) began to implement direct anti-poverty programs such as those for drought-prone areas (DPAP), small and marginal farmers and agricultural labor (SMAL), and a series of rural 89 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. employment programs. Since then, all of the following economic development plans recognized the need to address the poorest members of the population. Despite the ideological commitment to welfare as witnessed by the Five-Year Plans, the State made little financial contributions to social welfare goals. The actual strategy undertaken by the State to meet these goals was by combining increased public ownership with economic growth. Most of these Five-Year Plans suggested that specific steps to “deal with their problems” were needed, but did not discuss the steps in any detail (Appasamy et. al., 1996). Ultimately, Bose (1970) points out that growth of the state social sector was a slow process and tended to fall short of the immediate objectives set forth by the Five-Year Plans. Appasamy et al. (1996:16) concludes: The State has failed to hold on to the initial vision of human development steadily and as a whole. There have been failures in terms of political commitment, policies, and implementation. Serious structural reforms to reduce poverty and inequality have been lacking....While welfare rights have been promoted, they are far from being assured as legal entitlements. As the Constitution and Five-year Plans set the philosophical foundations for India’s social welfare state, most of their social welfare legislation was established in the early years after independence. Occasionally, amendments were made to existing legislation in order to increase the rate of compensation or extend benefits to another industrial group.1 3 However, it was not until direct attempts to globalize the economy in the late eighties that another series of social policies were implemented. Hence, one 1 1 For example, The Fatal Accidents Act of 188S was the earliest welfare legislation, which was replaced in 1923 by the Workmen’s Compensation Act after having proven ineffective. The Act was then amended in 1995 to increase the rate of compensation (Indian Labour Yearbook 1995). 90 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. would expect that the social expenditures in the late 1980s would increase rather than decrease. Charts 3.1 and 3.4 reveal that this was certainly not the case in India, and that expenditures actually began to decrease fo r the first time in the mid to late eighties and early nineties. Why India’s social welfare commitment declined during the period of globalization, particularly after policies mandating greater state financial contributions to welfare were implemented, is a puzzle that needs to be addressed.1 4 By the early seventies, much of India’s welfare policies had been established. Most of these social policies were targeted towards labor, rather than groups that were unorganized and/or the more vulnerable members of society. Benefits granted shortly after independence in 1948 marked India’s distinct attempt to become a modem welfare state. Kuruvilla (1996) claims that India’s Factories Act of 1948 is among the most advanced in the world. It provided legislation for maternity leave and benefits, the provision of child care in all factories, and advanced legislation on health and safety. Much of India’s social welfare programs encompass standard labor benefits, and are primarily based on employer rather than state contributions. For example, the 1948 Act also made provisions for the long-term protection of the family of the worker. Sickness and maternity benefits, medical care and pensions constituted the majority of expenditures. These benefits were guaranteed under the 1948 Employee State Insurance (ESI) Act which provided institutionalized medical care, 1 4 Another interesting puzzle that might emerge is why the level of social welfare expenditures in India is low. However, recall that the quantitative analysis found GDP to be significant, suggesting that poorer countries spend less on welfare. While this does not completely address the levels question, it remains an interesting inquiry for future research. For the objective of this study, however, India’s trends is the primary focus. 91 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. sickness, maternity, disability and employment injury benefits. Also implemented before 1972 was the 1961 Central Maternity Benefit Act, the 1952 Provident Fund Act for retirement supplemented in 1971 with a family pension scheme, and in 1976 with insurance linked to deposits to the Fund. In addition, the 1972 Payment of Gratuity Act provided supplementary retirement benefits. The Indian government has implemented select programs that aim to assist non labor groups and the weakest members of society, but have not been successful. For instance, the Public Distribution System (PDS) was intended to protect the poor and vulnerable from higher prices of food staples (Ravallion & Suggarao, 1992). Yet since coverage remains untargeted and uneven, the fundamental purpose of the program has been ineffective. Broader benefits such as Unemployment and Family Allowances have not been incorporated into India’s social welfare system. Until recently, the government has funded temporary unemployment programs in only 11 of 25 states. It is significant that these types of social welfare programs generally demand greater financial commitments from the state. The most recent social security legislation was the Pension Fund implemented in 1995. This program faced much resistance because, originally, it was supposed to supplement existing benefits such as the provident fund and gratuity (Ratnam, 1996). Instead, the new Pension Scheme ended up replacing the existing family pension fund rather than adding an entirely new benefit for workers. Labor was much dissatisfied with this change, as revealed by the eruption of labor strikes protesting the new social 92 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. scheme (The Working Class Newspaper, January 1996). The fact that this new program did not lead to an increase in overall expenditures, measured as a percentage of GDP, suggests that labor was justified in protesting this substitution. Most of the social welfare schemes mentioned above involved limited financial contributions from the state.1 5 Table 3.2: Contributions to Social Security Receipts in India (in millions of national currency units) Year State contribution Employer Contribution Employee Contribution 1975 764 (3.2%) 16404* (70%) 1985 3012 58737 8562 (3.4%) (57%) (9%) 1989 2017 12174 10448 (4.5%) (28%) (23%) 1991 2529 16758 15085 (4%) (27%) (24%) • Numbers in parenthesis represent contributions as a percentage of total receipts. These figures do not add up to 100 percent because remaining contributions come from special social security taxes, participadon of other public authorities, income from capital and other receipts. However, contributions from these sources in LDCs tend to be very small. * Data for employee and employer contributions in 1975 is aggregated represents missing data • Years chosen based on data availability. Source; The Cost o f Social Security. International Labor Organization: Various years 1 5 Note that the charts disaggregating contributions to social security by the state, employer, and employee are the largest component of the numerical measure of the dependent variable. As indicated in Chapter 2, the dependent variable is social security and welfare. 93 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3.2 illustrates the proportion of state financing relative to those of employers and employees. State financial contributions to social benefits averages at only about 3.8 percent, while standard employers’ and employees’ contributions are at approximately 37 percent and 20 percent, respectively. It is interesting that state contributions (as a percentage of total receipts) begin since the early nineties, while employers witness a more dramatic decline in their proportion of benefit contributions dating back to the early eighties. In order to compensate for declining expenditures by the other participants, it is no surprise that those insured are being faced with greater financial responsibility for benefits. Some safety nets were implemented in the nineties to ease India’s transition to a more global economy. Faced with calls for restructuring and the closure of “sick” (or inefficient) industries, the National Renewal Fund (NRF) was implemented to enable enterprises to pay reasonable separation benefits for affected workers (Papola, 1994). The NRF also aimed to assist retrenched workers in their retraining and redeployment. One strategy used by the industrial sectors to encourage retrenchment and reduce their workforce was the Voluntary Retirement Schemes (VRS). Ultimately, these schemes were implemented with the intent to make “exit” easier. It has been observed, however, in many cases that workers benefited disproportionately from these compensation schemes. Rather than compensating those most adversely affected by the new economic policies of the late eighties, those accepting these benefits were mostly skilled workers or those close to retirement 94 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (Mamkootan, 1994). Younger and unskilled workers were receiving small entitlements, if any. India’ s Labor Market Policies The philosophy of social justice enshrined in India’s constitution has resulted in a higher degree of regulatory protection to workers in the organized sector, than in Brazil or Korea. State policies have been highly interventionist in the fields of labor and employment. This pro-labor bias is revealed by several legal enactments dealing with employment, conditions of work, wages, industrial relations, social security and labor welfare (Papola, 1994). Indian public policy has evolved to become very protective of labor and incorporated strict labor protections within enterprises. Furthermore, unlike many LDCs, trade unions in India have been legally recognized and are legitimate representatives of labor interests. The freedoms afforded labor in India creates a state-labor system that is far removed from corporatist type arrangements in many LDCs. It is important to note, however, that the scope of these labor laws has been limited to the organized sectors (Venkatachalam & Singh, 1982). Three important pieces of labor legislation have shaped the relations of business, labor and the state in India (Lansing & Kuruvilla, 1987). First and foremost, the Trade Unions Act of 1926 permitted the registration of unions, and granted trade union leaders immunity from criminal prosecution. This law was also significant because it provided legal rights to organize and it represented the first step towards union regulation. After independence in 1947, the Industrial Employment “Standing 95 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Orders” Act contained rules and regulations governing the terms and conditions of employment of workers in industry. This Act ensured a certain minimum standard of protection for labor. Finally, in 1947, the Industrial Disputes Act was intended to supplement the efforts of weak labor movements. It sought to guarantee workers a fair return for their services, as well as to promote industrial peace (Venkatachalam & Singh, 1982). Significantly, these three acts serve as the core of country’s labor and industrial legislation. The last two acts, although initially designed to regulate conflict and boost production, ultimately promoted an extremely inflexible and inefficient labor market. The remaining legislative acts governing the welfare and safety of workers were all passed in piecemeal fashion by the early sixties. They include Indian Railways Act 1890, the Dock Workers Act 1948, Merchant Shipping Act 1958, Motor Transport Workers’ Act 1961, and the Shops and Commercial Establishment Act (,Indian Labor Yearbook 1990). It is interesting to note that most of these protections apply to the service sectors only, rather than the manufacturing sectors where the majority of laborers in India are employed. It is the 1947 Industrial Disputes Act, as amended in 1976, that is in highest contention as India enters global markets. Originally, the primary intent of this Act was to defend workers from arbitrary lay-offs but, in the end, afforded extremely high protections to labor. For instance, compensation is required up to 50 percent of basic 96 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. wages plus dearness allowance1 6 for the days which the worker is laid off. The 1976 amendment states that no lay-off or retrenchment in any establishment employing 300 or more workmen can be enacted unless prior permission from the appropriate government officials has been obtained. It also stipulates that any change in the terms of work that is likely to adversely affect the worker, can only be implemented after 21 days prior notice (Papola, 1994). These consequences of these laws are bureaucratic delays, high labor costs and prevent easy “exit” policies, thereby reducing the flexibility of the Indian workforce. The problem is that these laws serve to disproportionately protect organized workers, which comprise only 10 to 15 percent of the workforce (Appasamy et al., 1996). Generally, the labor laws apply only to relatively large industrial establishments (Jagannadham, 1954). The largest category in the workforce, the self-employed and the casual workers, do not receive such protections and/or benefits. They comprise about 90 percent of the working economy (Harriss-White, 1999). Consequently, the cost of workers in the organized sectors is likely to be three to four times higher than unorganized sectors (Papola, 1994). Clearly, India has not yet adopted the necessary provisions that would allow for a more universal social welfare system. In sum, the extent of the Indian government’s social commitment is based on its generous regulatory provisions for labor. The problem is that these protections cover only a minute percentage of the labor force. It is only India’s labor market policies that 1 6 Dearness allowance allows adjustments for inflation 97 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. fulfill the doctrine of social justice embedded both in its Constitution and the ideology of the Five-Year plans. On this basis, India ranks as a moderate welfare state. Its level of welfare expenditures, as well as the current decline (starting from the early 1990s) in welfare trends illustrates that India’s social welfare system as a whole might be in crisis. How does India’s welfare state compare to Brazil and Korea? How are they similar and how are they different? Independent Variable: Globalization in India Like most LDCs, globalization in India began as a slow process and then took- off around the mid eighties. Trends in India’s trade and capital flows illustrate a steady increase after 1986. Historically, India has been extremely isolationist in its international economic policies. Its preference for the import substitution industrialization (ISI) development strategy and its socialist ideology resulted in very low levels of globalization prior to the 1980s. As restrictions on trade and capital flows were lowered, it became increasingly clear that India enjoyed a comparative advantage in manufacturing, particularly in low-skill exports. India’s trade levels as a share of GDP began to flourish after 1985. Capital flows in India, however, remained relatively small relative to GDP even after liberalization. It has been only recently (in the early to mid 1990s) that this trend has begun to improve. 98 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Background When did India begin to feel the impact of globalization? While it is impossible to pinpoint when international market forces first began to assert their influence on the Indian polity and economy, it is possible to ascertain when it became a significant force. The Indian case is particularly interesting given that after independence, its disassociation from the global economy was of conscious design. Embarking upon a socialist agenda and Soviet type strategy of inward oriented development, India struggled to maintain its economic isolationism for almost the first four decades after independence in 1947. While occasional attempts were made at opening up the economy during this time, it was not until the mid eighties that the Indian economy began to witness a distinct structural change in its export and capital markets. Ideologically, India was predisposed towards establishing its own path to development, one that eschewed international economic influences. Almost two centuries of British colonial rule culminated in a strong antipathy towards foreign influences. The principle of swadeshi, or self-reliance, became synonymous with Indian nationalism.1 7 Independence leader, Mahatma Gandhi, represented swadeshi himself by using only traditional methods to spin and weave his own clothes. The values of a globalizing market economy was thus not consistent with either Gandhi or the philosophy of India’s first prime minister, Jawharlal Nehru. 1 7 The first Swadeshi movement, from 1905-1908, called for a total boycott of foreign goods, particularly imports (Encamation, 1989). 99 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Nehru subsequently led the drive towards India’s economic independence from global market forces. During the Nehru era, India became relatively self-sufficient with respect to production of capital and intermediate goods. The logical strategy to adopt based on the prevailing ideology was heavy industry development based on import substitution industrialization (ISI). Between 1951 and 1966, or the first three Five-year Plan decades, the ISI strategy produced a strong domestic economy in which local producers were able to successfully manufacture most consumer durables (Rudolph & Rudolph, 1987). Thus, a fairly strong and diversified industrial base emerged by the end of the first decade after independence (Evans, 1995). As Rudolph and Rudolph (1987) point out, India’s attempts at isolation from the influences of globalization were particularly successful relative to other LDCs: The international environment is far less salient for explaining Indian economic or political development than is the case for other Third World countries, unless it is salient in a negative sense, as an environment to exclude or keep at bay. India’s economic and security dependency has been slight. Its participation in the international economy...has been highly selective. Its goal has been self-reliance, a goal that has led it to participate in world markets less than other large Third WofW countries, in order to maintain control of the sources of economic power (Rudolph & Rudolph, 1987: 3). Until the mid-eighties, India maintained this domestic economic strategy of autarky and remained, thereby, de-linked from the world economy. The first short attempt at liberalization in the 1960s was short-lived. It was only after the second attempt at liberalization in 1985 that India relinquished its control over the impact of the international environment on its domestic economic and political choices. 100 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The foreign economic policy dimensions of India’s self reliance tactics are revealed by its strong resistance to competition from abroad. The 1951 Industrial Development and Regulation Act (IDRA) provided the state with control over the direction of private investment through the extensive use of industrial licensing. In 1969, the Monopolies and Restrictive Trade Practices Act (MRTP) aimed to increase government control over access of (domestic and international) industrial conglomerates to the Indian market. Thus, by no mere exaggeration was the term “Permit Raj” coined to characterize the Indian government during the prolonged period of autarky. This was the complex system of industrial licensing and foreign exchange, price and distributional controls (Lai, 1999). This discriminatory licensing system enabled India’s top business houses to block competitors, particularly foreign firms and investors.1 8 Indian business houses thus fared comparatively well in battles with multinationals for access to scarce government licenses (Encarta, 1989). Heavy restrictions on trade and capital flows resulted from the licensing system. The policy origins of the import licensing system are traced to the Imports and Exports Control Act in 1947 and the Import Trade Control Order of 1955 (Ahluwalia, 1997). Almost all imports were brought under the purview of licensing with this early legislation and were subject to administrative scrutiny. By not allowing imports of goods domestically produced and prohibiting external competition, India’s domestic IS Encarta (1989:135) points out that between 1956 and 1966, the 20 largest business houses secured almost one-fifth of all government approved licenses. 101 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. markets enjoyed full protection. Only imports of capital and intermediate goods were allowed, while consumer goods were banned. Trade Flows -Background Not surprisingly, promoting exports was not an integral part of the country’s development strategy. The export pessimism that dominated much of the developing world in the post-World War II era was rooted in the belief that exports were not dependable engines of growth. This strategy could not remain effective for long, however. As Lai (1999) explains it, exporters that depended upon imports were damaged since the effective rates of exchange were much higher for importers. It was not until the third Five Year Plan (1961-1966) that planners attempted to redress this export bias. It was agreed that some level of export promotion was necessary to finance the essential imports by way of cash subsidies, different sorts of tax rebates, and entitlements to imports for exporters of select commodities (Balasubramanyam, 1984). Nonetheless, this did little to turn the focus of the country’s economy towards export markets. In the presence of heavily sheltered domestic markets, entrepreneurs had little incentive to participate in the highly competitive export markets. Regardless, India’s change in economic philosophy towards exports in the 1960s became an important factor in sustaining the steady increase in exports as a share of GDP. Clearly before the mid 1980s, India’s export sector lacked the dynamism and competitive strength displayed by many of the Asian countries (Balasubramanyam, 1984). It was not until the mid to late eighties that the rates of growth demonstrate a 102 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. steady increase (see Chart 3.5). This rising trend persists even as India’s exports are calculated in real terms. When measured in constant 1987 US dollars, India’s export data show a similar trend (not shown here). The apparent decline in Brazil and Korea’s exports as a share of GDP, however, can be attributed to rapid increase in GDP. Chart 3.5: India’s Export Trends in Comparative Perspective Source: World Bank (1997) World Development Indicators E x p o r t s in I n d i a , B r a z i l a n d K o r e a f r o m 1 9 7 2 - 1 9 9 5 ( a s % o f G D P ) o s c 45 40 35 30 25 20 15 10 5 0 j y . a * 16 14 121 10 m a c a 6 ■ ■o 4 £ 2 0 Korea Brazil ‘India | Their export growth, when measured in constant 1987 US dollars, show a steady increase. ISI policies of curbing the imports of foreign goods eventually resulted in a significant change in the composition of exports. Growth in capital intensive exports was the natural consequence of ISI, which strongly favored the establishment of high- skill intensive industries. It was such industries that were permitted to take advantage 103 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of the government’s export subsidies. India was able to move up the product cycle by the early 70’s with the addition of skilled exports—such as engineering goods and chemicals—alongside handicrafts, leather manufactures and cotton. Nonetheless, traditional labor-intensive exports remained the primary export sectors. As Balasubramaniam (1984: 200) points out, subsidy assisted exports (or the heavy industry goods) have grown at a much slower rate than unassisted exports. Srinivasan (1990) further argues that a diversified and dynamic export sector has yet to emerge in India, since just four of the manufactured products, leather, gems, garments, and textiles, accounted for 70 percent of the growth during the 1980s. Clearly, Table 3.3 illustrates the importance of labor intensive manufacturing as well as skill intensive exports in India. Primary industries, such as agriculture and beverage and tobacco, show a steady decline in exports. Whereas, chemicals, manufacturing, and textiles have been experiencing steady increase in exports. Thus, this table confirms that chemicals and manufacturing exports, which represent both low and high-skill intensive categories, comprise the industrial mix of India’s export base.7 9 India, then, has to some extent been successful at gearing its productive activity 1 9 Manufactured exports generally comprise commodities in STTC revision 1, sections 5 through 9 (chemicals and related products, basic manufactures, machinery and transport equipment, other manufactured articles and goods not elsewhere classified) excluding division 68 (non-ferrous metals). The skilled vs. unskilled categories is derived from Wood and Berge (1997) (see Chapter 2). 104 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. towards a few skill-intensive industries, as argued by Peter Evans (1995). Significantly, however, India’s primary comparative advantage continues to be in low-skill goods. Table 3.3: Indian Exports by Industrial Origin (Percentage of Total Value)* Industry 1975 1985 1990 1994 (I) Agriculture 23.9 22.3 14.3 10.9 (2) Mining 10.9 21.3 19.6 18.5 (3) Manufacturing a 65.2 56.4 66.1 70.6 Textiles 23.7 28.6 32.8 33.6 FoodBev Tobacco 17.8 5.9 5.0 5.7 Chemicals 3.6 9.2 12.1 12.1 a: The three primary manufacturing export categories in India are reported here. There are nine total. • Source: United Nations. Yearbook o f International Trade Statistics: Various Years 105 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Capital Flows-Background With regards to constraints on capital flows, India’s dirigiste impulse was just as strong. The aim of the Indian government was to place the financial sector directly under national control (Encamation, 1989). By 1969, the bulk of the Indian banking system had been nationalized. Mistry (199S) argues that public intervention and control over the oligopoly of institutional investors thwarted the development of an internationally competitive, market-based portfolio-fund management industry. Once again, the country’s economic strategy of prioritizing national social and economic objectives was served by this financial system, but at a considerable cost. The banking system was effective in domestic resource mobilization (i.e., raising savings of households), while it ultimately failed at efficient resource allocation.2 0 Severe restrictions existed on both inward and outward capital flows before the early 1990s. The capital market has been heavily controlled by the state and relatively isolated from international investors. In keeping with the protectionist thrust of India’s economic policies, government institutions or local markets were prohibited from actively participating in foreign capital markets (Encarnation, 1989). Capital flows consisted of aid flows, commercial borrowing, and nonresident Indian deposits (Chopra 3 0 Ahluwalia (1997: 277) points out that amongst the low income countries which savings data are recorded by institutions such as the World Bank, India placed at the upper end for such economies. 106 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. et. al., 1995). As a result, total private capital inflows in the 1960s and early 1970s comprised a negligible percentage of their GDP. Generally, because of weak capital markets, multinationals were the primary conduit for transferring foreign savings to most LDCs at that time (Encarnation, 1989). However, India was determined to remain de-linked from the international economy and gain both financial independence and managerial autonomy. After the 1960s, the Indian government imposed strict controls on foreign direct investment. These capital controls served to limit multinationals from access to newly protected domestic markets. By 1973, foreign financial collaboration reached its lowest point in more than a decade. The Foreign Exchange Regulation (FERA) of 1973 laid down strong guidelines for countries with the majority equity interest held by foreign nationals. Foreign equity was held to a maximum of 40 percent of total holdings of certain industries (Ahluwalia, 1997). FERA companies sought to control companies with foreign equity in excess of 40 percent. The agency in charge of enforcing FERA eventually compelled most companies to dilute foreign equity and to widen the shareholder base amongst Indians (Chandra, 1997). Due to these relatively success attempts at monitoring the financial sector, multinationals in India were relying on local sources for expansion and diversification by the mid 70’s. Thus, independence from foreign financing was the sine qua non after the amendment of FERA (1973) and until the next relaxation of government restrictions on foreign investment in 1980 (Encarnation, 1989). The chemicals and industrial machinery accounted for most foreign collaborations in India. FERA forced MNCs in 107 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. unprecedented numbers to rely on the Indian equity market for new financing since government exempted few firms from equity dilution. After FERA, most MNCs reduced their foreign shareholdings to 40 percent or L ess. They simply sold existing foreign equity to Indian nationals. Companies such as IBM and Coca-Cola that did not want to abide by India’s strict foreign regulations exited by 1977. Enter Globalization When did India begin to experience the effects of globalization and become a recognized player in international markets? The distinct upward swing in the globalization trends of the late eighties strongly suggests that there was a structural change that occurred in India’s export and capital markets at that time. According to Denoon (1998), three attempts were made at liberalization in the years since independence (1966-68, 1985-87, and 1991-4). However, data suggests that it was not until the second attempt beginning in 1985 that liberalization efforts were concomitant with the actual effects on export and capital markets. Significantly, many Indian economists argue that India would have shifted towards a more open economy even without the 1991 balance of payments crisis (Srinivasan, 1990). Their position is consistent with this analysis, which posits that there are elements of globalization that are exogenous. Not only was the trajectory of economic nationalism becoming increasingly unsustainable in a globalizing world, but also there was a rise in international demand for trade and capital flows at that time 108 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (Byres, 1999; Kumar, 1998; Patnaik, 1994; Sharma, 1996; and Srinivasan, 1990).2 1 Regardless, based upon the empirical data and the policy orientation as evidence, the influence of globalization on India has been highest since the late eighties. In the first liberalization episode, 1966-68, the extent of reform was relatively limited. Prime Minister Indira Gandhi’s effort to liberalize India’s economy entailed providing export incentives by devaluing the rupee and rationalizing decisions on importing. However, this devaluation and decontrol package announced in June of 1966 shortly began facing criticism from the political Left, civil servants, and businesses (Denoon, 1998). Mounting political opposition eventually led Gandhi to reverse course two years later in 1969. Ultimately, the first liberalization effort ended with the nationalization of banks and a return to the state centered controls-dominated policy. Many argue that India’s “real” turn to international markets began in the mid 80s, with India’s second attempt at liberalization (Brahmbhatt et al., 1996; Byres, 1997; Kumar, 1998; Lai, 1999; Mani, 1995; Srivastava, 1996). The 1985-7 liberalization episode was instigated by Rajiv Gandhi. Major structural changes incorporated in the domestic economy subsequently allowed for the effects of global markets. This turning point, evidenced by changes in India’s overall trade levels as well as capital flows, is of most concern to this analysis. Some of the major reforms undertaken by Gandhi 2 1 Other exogenous influences are the conclusion of the Uruguay Round in 1993, ongoing implementation of Internal Market in the European Union, implementation of the South Asian Preferential Trading Arrangement and India’s ascendance to Full Dialogue Parmer of ASEAN (see Sharma 1996). For discussions on globalization as an inexorable process (see Grieder (1998), Nader (1998) and Pasha (1996)). 109 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. included reduced licensing of industrial investments, easing of import regulations and stimulation of imports, replacing quotas with tariffs, exchange rate depreciation. Attempts at financial market deregulation also began by raising the interest rate on government securities and lowering tax rates to increase investment incentives and reduce evasion. Trade and capitals flows continued to rise after the mid-eighties, even though the new coalition in 1989, headed by V.P Singh and the National Front, attempted to reverse Rajiv Gandhi's earlier attempts at liberalization. The 1989 coalition sought to reinstate economic policies supporting greater government intervention and slowly arrested the implementation of Gandhi’s liberalization policies. These efforts were short-lived, however. The National Front government lost a parliamentary vote of confidence after only 11 months in office, followed by the temporary reign of a minority government led by Chandra Shekar. The last set of reforms in 1991 is labeled by Denoon (1998) as India’s third attempt at liberalization. By the early nineties, India’s centrist party Congress (I) under Narasimha Rao was back in power. The country, then faced with a high current account deficit, adopted even more far reaching reforms embracing globalization. These New Economic Policies (NEP) confirmed that the momentum towards India’s globalization was well underway. The objectives of the 1991 reforms were based on a clear need to integrate with the global economy based on the encouragement of greater trade and capital flows (Ahluwalia, 1995). Liberals hailed the New Industry Policy Statement (NIPS) as a 110 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. major departure from the dirigiste nature of earlier command policies (Mani, 1995). Its proclaimed goal was to improve India’s national competitiveness by reforming the industrial sector in general and the manufacturing sector in particular. Most importantly, India’s licensing system was abolished and the MRTP amended.2 2 Consequently, most of the quantitative restrictions on imports and exports were withdrawn. In 1991, export-production linkages were strengthened by policy measures such as the elimination of quantitative restrictions on intermediates and capital goods. It also dropped a large majority of tariff lines covering raw materials, freeing intermediates and capital goods from import licensing requirements and extending the facility of Special Import Licenses to export/trading houses (Sharma, 1996). The most successful part of the reforms however, concerned traded goods and services. India’s imports and exports rose from an average of 15 percent of GDP in the 1980s to 27 percent of GDP in 1994-5 (see Chart 3.6 ). Once again, Chart 3.6 illustrates that trade has been declining in Korea and Brazil as a share of GDP. As before, however, when checked against trade measured in constant 1987 dollars, real trade growth in all three countries have been increasing (not shown here). 2 2 The asset threshold that subjected to firms to MRTP was raised. I l l Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chart 3.6: India in Comparative Perspective: Exports and Imports T r a d e f l o w s in I n d i a , B r a z i l a n d K o r e a ( a s % o f G D P ) 1 9 7 2 - 1 9 9 5 90 80 70 60 | 50 I 40 30 20 10 0 — — Koras • - India - Brazil Source: World Bank (1997) World Development Indicators. With regards to capital liberalization, India has taken a much more gradualist approach. It was not until 1991 that India took its first true steps towards capital market deregulation. Overall, official attitudes towards foreign direct investment are more favorable than those toward portfolio flows. Yet, in comparison to other developing countries, India levels of capital inflows and outflows still remain relatively low. Foreign direct investment and portfolio investment flows have been quite modest in comparison with many other developing countries, although growth began to accelerate in the eighties. 112 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chart 3.7: Trends in India’s Foreign Direct Investment I n d i a 's N e t F D 1 1 9 7 2 - 1 9 9 5 ( a s % o f G D P ) 0.45 - I 0.4 \ 0.35 j 0.3 - I 0.25 1 0 2 - 0.15 - i 0.1 - 0.05 - 0 - -0.05 - - 0.1 - Source: IMF (various editions): Balance o f Payments Statistics Current policies towards capital flows suggest that Indian government officials privilege long term inflows of equity into the country, rather than short term capital flows. The new policies actively promote FDI. Restrictive provisions of various types which were applicable to companies with more than 40 percent foreign equity have been abolished and all companies incorporated in India are treated equally irrespective of their foreign equity holdings (Mani, 1995). Investments with little restriction are now allowed for a defined list of 34 industry groups subject to a limit of 51 percent foreign equity holding. India has also joined the Multilateral Investment Guarantee Agency (MIGA) and is currently negotiating bilateral investment treaties with a host of other countries. Thus, automatic approval is granted for foreign investment up to certain equity in selected sectors. One hundred percent equity is allowed in these and 113 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. all other sectors, but requires case by case approval by the newly established Foreign Investment Promotion Board (Nidugala, 1997). The approval of greater FDI inflows in late eighties resulted in a dramatic increase after 1990, from an extremely low level in the 1970s and 1980s. (see Chart 3.7). Overall, increases in restrictions in India’s short term capital flows, however, were modest. Severe regulatory limitations continue to exist, particularly for outflows of short-term capital for loans, purchase of bonds and shares. By courting foreign investors but still restricting domestic firms from investing abroad, the aim is to ensure that capital inflows mostly exceed outflows. Therefore, Indian companies seeking to invest abroad are relatively disadvantaged. They face a maze of regulations that provide strong resistance to the mobility of capital (The Economist 2000: 89). Neither are domestic residents permitted to buy stocks and bonds from international markets (Nidugala, 1997). Foreign institutional investors, however, are permitted to invest in the Indian capital markets. Yet there are still restrictions on acquisition of shares of companies by these investors. Selected companies have been allowed to raise equity in the international markets. Regardless of the existing restrictions on capital mobility, data clearly illustrate that since the late 80’s, India has witnessed a steady increase in gross capital flows, again from an extremely low level (see Chart 3.8). Trends in gross capital flows appear to fluctuate prior to that date. Chart 3.8 clearly shows that India has the lowest gross capital flows as a share of GDP relative to both Brazil and Korea. Although its increased share of capital flows amounts to only around one percent of inflows relative 114 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. to other Asian developing countries, there have been significant improvements since reform (Kumar, 1998). The composition of total investment has also changed. Private investment advanced from an average of 11 percent of GDP in the 1980s to over 14 percent in 1990-1994 (Denoon, 1998). Chart 3.8: Gross Capital Flows: India in Comparative Perspective Capital flows in India, Brazil and Korea (as % of Brazil Korea Source: IMF (various editions): Balance o f Payments Statistics It can be concluded then that globalization in India, indicated by both increasing trade and capital flows, gained momentum in the 1980s. More specifically, globalization began with growing capital and trade flows in the early and mid 1980s, respectively, and then experienced a sharp increase in FDI flows in the early 1990s. India was unique in its strong determinism to remain isolated from global markets. The levels of imports and exports have shown a structural shift in the mid-eighties. Significantly, even though India has obtained a diversified mix of skill-intensive and labor-intensive exports because of its history with ISI, it appears that India has been 115 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. exporting more labor intensive goods since globalization. Capital flows, on the other hand, have exhibited a much more gradual increase relative to GDP. Indeed, it has been FDI flows more than portfolio flows that have improved in India. The question now is how these different components of globalization interact with the bargaining power of labor to ultimately affect social spending. Independent Variable(s): Globalization and the Bargaining Power o f Labor How and why does the interaction between globalization and the bargaining power of labor negatively affect social welfare expenditures, as tested in econometric estimations of Chapter 2? More significantly, does this “interaction” explain the late 1980s decline in the Indian government’s social commitments—as discussed in the first section of this chapter? The objective of this section is to assess whether or not labor has indeed increased the “range of plausible outcomes available to [them]” and improved “the probability of securing the outcome that [they] prefer” in the era of globalization.2 3 The evidence suggests that globalization has played a prominent role in exacerbating problems for labor in India and their ability to extract benefits from the state. Consistent with the primary hypothesis, the number of India’s unskilled laborers has been increasing in tandem with globalization, and adversely affecting their bargaining power. Nonetheless, additional country specific factors emerge which 2 3 See Encamadon (1989:20) for more on assessing “bargaining power.” 116 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. impede the development of strong labor institutions in India, but were not directly included in the LDC model of Chapter 2. Most importantly, it is revealed that the size of India’s informal sector has been expanding with globalization and, thereby, adding heavily to its unskilled labor population. Other significant causal variables found in the Indian case are based on the historical legacies of the state-labor relationship (e.g. , multiplicity of unions). Labor’s low bargaining power cannot be fully understood without first taking such domestic level factors into account. Changes in labor power will be deduced by labor demonstrations via strikes, lockouts, lost work-days, etc., and increased (or decreased) demand for informal, unskilled labor. Such data must be analyzed in historical context, however, since a rise in labor strikes, demonstrations, etc. can be indicative of either increasing or decreasing bargaining power of labor. This section will then show how globalization and the low bargaining power of labor have been mutually reinforcing. Background Assessing the Indian government’s stance on labor power is not a simple task. Government support of labor was endemic to India’s socialist order after independence. However, the government-labor relationship has been riddled with contradictions. While government labor policies have succeeded in protecting a very small percentage of the population, the majority of unorganized workers (which is 90 percent of the workforce) have directly suffered as a result of such legislation. A labor market has emerged that is divided between the (over)protected organized sector and the neglected 117 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. unorganized (and informal) sector. In reality, it emerges that India’s labor laws have served to divide labor as a political group. Indeed, it appears that collective action problems have progressively intensified with the increasing duality has characterized India’s labor markets during globalization. Unlike many developing countries, labor groups in India have not been subsumed under a state corporatist structure. The rhetoric of successive Indian governments since independence has privileged labor in national politics. Labor’s political position can be attributed to two factors. First, the importance of prioritizing labor’s needs is based on the principle of social justice embodied in India’s socialist philosophy (Appasamy et al., 1996). Second, labor’s involvement in the independence struggle earned them the right of representation in legislative assemblies. Nehru incorporated this philosophy in the first Five-Year Plan after India’s independence from colonial rule. A review of this Plan reveals a fully comprehensive labor section dedicated to the well-being of the working class. By way of establishing labor’s rights, this section clearly serves the national objective of social justice. It states: The worker is the principal instrument in the fulfillment of the targets of the Plan and in the achievement of economic progress generally. His [src] cooperation will be an essential factor in creating an economic organization in the country which will best subserve [sic] the needs of social justice. Certain rights and obligations are associated with this distinct role. Adequate provision has to be made for the basic needs of the workers [...] (First Five-Year Plan as included in Myers, 1958). Second, labor’s incorporation into India’s competitive electoral system occurred even before 1947 as it became involved with the independence struggle. The 118 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. government of India Act of 1935 first granted labor the right to be represented in the legislative assemblies (Venkatachalam & Singh, 1982). Unions soon came to have close political affiliations with various segments of the nationalist movement. The major national political party at the time, the Indian National Congress (INC), grew sympathetic to the labor cause and included a clear labor policy in its election manifesto. These close ties which evolved between the union movement and the freedom struggle helped to institutionalize the labor movement. Labor’s fight ultimately became intertwined with the national fight for independence. After 1947, there emerged 10 large unions and each affiliated with different parties. Thus, instead of state corporatism befalling the fate of India’s labor, Candland (1995) labels it as a type of party incorporation. Do these labor laws in India render a more powerful labor movement at its base? The evidence overwhelmingly indicates that India’s extensive labor legislation serves to protect only a minor segment of the labor force. In addition to the increasing proportion of unskilled workers relative to skilled, labor’s weak bargaining power is also based on (1) the multiplicity of unions, (2) the lack of a uniform labor code, (3) the relatively disproportionate influence of the state in tripartite bargaining, and finally (4) the increase in the informal sector. While the first three aspects of labor’s difficulties have continued to exist since the early 1970s to the present, it is arguably the rise of the informal sector and size of the unskilled labor force that is being most affected by globalization. 119 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. M ultiplicity o f Unions Based on review of the labor union laws, India flexibility with regard to union formation is a double-edged sword. While laborers are permitted to form industry and enterprise-wide unions to articulate their demands, the increasing number of unions created disarray for labor in terms of facilitating collective action. The Trade Unions Act in 1926 paved the way for labor’s freedom of association to voice grievances and to demand redress of those grievances {Indian Labor Yearbook 1990). According to this Act, only seven employees are needed to form a trade union, must remain in existence for one year, and must have membership corresponding to at least IS percent of the total workforce in the unit. The problem with this Act was that it led to the decentralization of union power, allowing an unlimited number of unions in each factory. Many scholars agree that the existence of multiple unions to be problematic to India’s collective action problems (Lansing & Kuruvilla, 1987; Mathur, 1996; Papola, 1994; Ratnara, 1993, 1996; Sinha, 1994). The multiplicity of unions in India has exacerbated the problem of inter-union rivalry and thus, helped engender a conflict-ridden industrial relations climate. Another disadvantage to this 1926 Act was that it enabled politically motivated “outside” members to become union leaders. Labor leaders often have never been factory workers and are thus unfamiliar with the conditions faced by workers. (Sinha, 1994). Often times, their interests are compatible with their affiliated political parties rather than with the working class they are expected to present (Sinha, 1994). Industry 120 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. wide unions may then be more influential in national politics, yet they are dependent on political patronage and tend to privilege national issues over labor ones. Unions have thus become heavily politicized in India. Consequently, politicians have little interest in reducing the number of unions since they depend upon them for political support (Lansing & Kuruvilla, 1987). This situation appears to have worsened since India’s turn to globalization. Union numbers continue to increase. Table 3.4 shows India in comparative context with other countries that are considered to have strong labor movements (i.e., Brazil, Sweden and Finland). The relative difference in the number of Indian unions is striking. Notice that while the number of unions in India approximately doubles from 1980 to 1990, the percentage of employees unionized stays almost the same. This lends strong evidence to the conjecture that higher numbers of unions in India do not indicate higher labor power, and that India’s labor movement has progressively become more fragmented. 121 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3.4: Number of Unions: India in Comparative Context* India Sweden Finland Brazil Korea 1980 4435 (25.5%) • * 100 (••) • • 2618 (21%) 1985 7815 (-) • • 4897*** (27.7%) 2534 (16.8%) 1990 8828 (27.0%) 67** (84%) 117 (..) • • 7698 (18.4%) • Figure in parenthesis is m em bership as percentage o l total paid em ployees • m is sin g data • Years shows is based on availability of data. ♦♦Data listed for 1989 ♦♦♦Data listed for 1988 • Source: International Labor Organization Lack o f Uniform Labor Code The second complicating factor for labor is the lack of a uniform labor code. Indeed, there is much existing legislation addressing various needs of labor. Venkatachalam and Singh (1982) argue, however, that the large number of individual laws in force complicates the protection of labor because application of these laws have long been open to interpretation and dispute. Approximately 108 enactments exist on the statute books regarding labor. As a result, the multiple standards safeguarding the interests of the working class are creating loopholes in labor’s regulatory provisions (Dhyani, 1977). A law to redress this effect was proposed in 1978 entitled the Industrial Relations Bill. This bill would recognize a single bargaining agent in each organization and thus provide some cohesiveness to the labor struggle. 122 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Parliament, however, was reluctant to support this bill since it would ultimately restrict the number of unions which provide them with political support. As it stands, India’s dispute resolution system continues to be inefficient and does not serve to pacify industrial conflicts involving labor (Lansing & Kuruvilla, 1987). This problem thus continues to persist. No further efforts have been made to amend these laws since globalization became a larger force. State Dominance in Tripartite Bargaining Despite the autonomy granted to Indian unions and emphasis on tripartite bargaining, the state has been the main force in determining labor’s fate. Since independence, India has experimented with tripartite forums in an attempt to bring about a healthy industrial relations climate. These meetings between government, business, and labor representatives, however, have only served as “smoke and mirrors” to pacify current tensions and serve public opinion (Mathur, 1993; Ratnam, 1997; Sodhi 1993). The problem is that the state dominates both employer and labor organizations (Rudolph & Rudolph, 1989). Labor made recommendations, but has never really been able to affect national policies (Mathur, 1992). For example, both labor and employers demanded state contributions to welfare benefits under the Provident Fund passed in 1961 as well as labor welfare in 1946. Each time the state prevailed in mandating contributions primarily from employers and those funded (Mathur & Seth, 1969). This pattern did not change in the face of globalization. As labor unrest accelerated during the process of industrial restructuring, government responded by 123 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. calling more tripartite meetings. A Special Tripartite Committee was formed in December of 1991 during the process of reform. Both workers and employers became increasingly involved with discussion of structural adjustments and how to handle the dismantling of “sick” firms. Mathur (1993) argues, however, that the government tended to favor employer demands during the process of negotiations. They lobbied intensely for easier exit policies, lay off of employees and retrenchment of workers without government permission. Ultimately, he concludes that such consultation procedures were successful at deescalating tensions and endorsing government policies, while providing little in the way of reasoned discourse. Shift in Demand for Informal Unskilled Labor The original hypothesis predicts that a greater demand for unskilled labor in (labor abundant) LDCs during the era of globalization affects the ability of labor to successfully extract scarce government resources. This theory is based on the assumption that higher levels of employed unskilled labor, which is notoriously difficult to organize, makes it more difficult to for labor to form coalitions (i.e., with different labor groups, middle class, etc). As shown above, there are several domestic level forces that have weakened the Indian labor movement since independence. The evidence from the econometric model confirms, however, that it is indeed unskilled labor which has been most directly affected by the forces of globalization. Yet what has not been captured in the quantitative analysis is globalization’s impact on the growth of 124 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. informal sector. This effect appears to be associated with the greater demand for unskilled labor. In other words, more and more enterprises have turned to the informal sector in an effort to avoid labor legislation, and in their search for cheap, unorganized and low skill labor. India’s country specific factors, or its rigid labor markets and multiplicity of unions, have contributed to this dynamic and engendered a weaker political position for labor. This section will explore how globalization encouraged both employment of unskilled labor and the casualization of the labor force.2 4 First, the growth in exports, particularly goods based on low skill labor, will be traced in order to obtain a sense of international demand and changes in domestic production.2 3 The second section interpolates evidence that the informal sector has been increasing alongside globalization. Finally, data on strikes and lockouts, as well as the most recent labor policies are discussed to assess whether, in fact, there has been a visible change in India’s labor power. Growth in Demand for Formal and Informal Unskilled Labor Since a large part of India’s economy is based on manufacturing, many of the reforms are aimed at the industrial sector. Through most of its post-independence 2 4 “ Casual izatioiT refers to the growth of the informal sector. It is a term commonly used by Indian academicians and labor leaders interviewed for this research project 2 5 Observing an increase in exports suggests that either there has been change in domestic production (and/or labor productivity) or growth in international demand or both. It is virtually impossible to determine which force dominates. Moreover, the logic of greater economic integration suggests that both are occurring at the same time. 125 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. history, India focused on developing its heavy industry, thus moving its economy towards more capital intensive industries. However, the diversified industrial structure was generally not internationally competitive in terms of cost and quality of output (Srinivasan, 1990). It has been, rather, low-skilled goods which comprised the greatest proportion of India’s exports. Since its comparative advantage has been greater in labor intensive export-oriented manufactures, these exports have been increasing alongside globalization (Brahmbhatt et al., 1996; Gangopadhyay & Wadhwa, 1998 Sathe, 1997; Sharma, 1996). Until the early eighties, tea, jute, iron ore, cotton textiles, leather products, and sugar were India’s primary exports (CIA Factbook, various years). Exports of engineering goods came after almost two decades of investment in heavy and capital intensive industries. Yet Sathe (1997) still claims that it is the traditional industries that have gained the most from liberalization. The export intensity of most skill intensive items 26 still remains relatively low (see Chart 3.9). Thus, it is no surprise the an IMF study by Chopra et. al., (1995:56) concludes that a major shift towards skill intensive exports is not yet evident. On the otherhand, Charts 3.10 and 3.11 reveal that comparatively, the proportion of high skill to low skill exports is remarkably higher in Brazil and Korea. 2 6 The terms capital-intensive, skill-intensive, and technology-intensive are used interchangeably in this analysis. 126 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chart 3.9 : India’s Manufactured Exports India's Manufactured Exports (as % of total exports) 50 40 30 20 10 0 1974 1976 1980 1984 1987 1992 1 994 1995 B low Skill- basic m anufactures B n i g h Skill -m achines and transport equip Source: World Bank (1997): World Development Indicators 3 r e ! Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chart 3.10: Brazil’s Manufactured Exports Brazil’s M anufactured Exports (as % of Total Exports) 30 25 20 15 10 5 0 1976 1979 1983 1987 1993 1995 ■ Low Skill-basic m a n u fa c tu res ■ High Skill -m achines and tra n sp o rt equip Source: World Bank (1997): World Development Indicators C hart 3.11: K orea's M anufactured Exports Korea's Manufactured Exports (as % ofTotal Expo rts) 60 50 40 30 20 10 0 1 9 7 6 1 9 7 9 1 9 8 3 1 9 8 7 1 9 9 3 1 9 9 5 ■ Low Skill-basic manufactures ■ High Skill-machines and transport equip Source: World Bank (1997): World Development Indicators Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. India’s advantage in low-skill goods is not expected to change any time soon. Approximately three-quarters of India’s largest net foreign exchange earner (i.e., garments) are currently restricted to markets in the EU and US. It has been predicted that the elimination of the Multi-Fiber Agreement (MFA) will greatly benefit India’s export sector, since currently about 78 percent of its clothing exports are subject to NTBS by OECD countries (Brahmbhatt et al., 1996). Thus, it is expected that there will soon be a tremendous growth in labor intensive exports with the phasing out of the MFA The growth in manufacturing sector as a whole increased by 6.5 percent in 1993-4, 8.6 percent in 1994-5 and over 10 percent during April-December, 1995 (Sharma, 1996). The latest trends indicate that sharp improvement in manufacturing production has been based on export oriented sectors like leather, chemicals, metal products and machinery. How might foreign capital flows have influenced an increased demand for unskilled labor and lowered labor’s bargaining power? It is rather difficult to trace the direct impact of portfolio flows on labor markets. The question of whether or not actual portfolio flows respond negatively to countries with labor market protections is best answered by statistics and correlation analysis. Otherwise, there are two alternatives (for future research). First, interviews with foreign institutional and individual investors (e.g., nonresident Indians) would address this issue, but is extremely difficult to execute. The second alternative would be to research companies in India issuing the most global depository receipts (GDR)—are they predominantly low-skill or high-skill 129 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. companies? As India attempts to reform their labor markets, is there more investment in labor intensive companies? Regardless, the previous section of this chapter explained that portfolio flows in India have not yet matched the increase in FDI inflows. Contrary to the difficulties of tracing motivations behind foreign portfolio flows, assessing the influence of FDI upon labor markets is more feasible. India’s inflexible labor policies are likely to dampen investor interest It is expected that capital-intensive firms would attract more FDI flows than labor-intensive sectors. Indian hourly manufacturing wage costs of US $0.27 are already among the lowest in the world, in line only with those in China ($0.25) (Brahmbhatt, 1996). The main impediment to attracting FDI would be its rigid labor laws. As India works towards establishing a more flexible labor market, FDI inflows should shift to labor-intensive sectors.2 7 Kumar (1998) presents data on the sectoral distribution of FDI stock in India. His estimates indicate that between 1980 to 1990, FDI stock in low-skill industries such as textiles remained at 3.4 percent. However, FDI in more capital intensive industries like machinery, transport equipment and electronics showed sizeable increases. It is important to note that Kumar’s (1998) data ends in 1990. In 1991, the Indian government adopted a number of measures to increase investment in labor- intensive export processing zones. If foreign investors and multinational corporations respond to these incentives as expected, then there would be an increase in employment of unskilled labor. Until now, FDI inflows have gone to a more diverse set of 2 7 As Chopra et aL (1995) show, progress in this area has been slow. The Board of International and Financial Restructuring (BIFR) was implemented in 1987 in an effort to facilitate institutional reforms that provide for overhaul of labor legislation. However, its procedures are slow moving. 130 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. companies, including both labor intensive and capital intensive sectors. (Chopra et al., 1995; Kumar, 1998). Thus, the evidence suggests that India’s rigid labor markets continue to adversely affect FDI inflows to its low skill sectors.2 8 FDI inflows are thus not yet helping to absorb India’s high labor surplus population. It can be deduced that one of the reasons for expanding employment in India’s low-skill industries is export growth. Employment in both high and low skill industries has increased since 1985 (see Chart 3.12). These numbers are impressive, given the number of retrenchments and closures that occurred in the eighties. For example, Papola (1994) states that in the entire organized cotton textile industry, employment declined by 242,000 during the period of 1980-1987. The data in Chart 3.12, 3.13 and 3.14 show that employment in unskilled sectors has increased in India, but declined in Korea. Interestingly, Brazil and India seem to have similar proportions of unskilled to skilled labor. Yet Brazil’s social welfare expenditures have been increasing relative to GDP, while India’s has been declining. This study proposes that the expansion of the informal sector, alongside the growth of formal low-skill sector, has had a comparatively stronger affect on the bargaining power of India than in Brazil. 2 8 This does not mean to suggest that India should strive to maintain its comparative advantage in low- skill production goods. Rather, it is India’s large unemployment and underemployment problem that needs to be addressed. The advantage of labor intensive goods is that it has a positive effect on employment I f however, India is able to increase the skill level of its workforce, then it may successfully manage to decrease employment via the growth of skill-intensive industries. 131 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chart 3.12: India in Comparative Perspective: Employment in Manufacturing Employment in India's Manufacturing 4000000 3500000 3000000 2500000 2000000 1500000 1000000 500000 0 I v - j :;l .2 1 0 0 1] 3 Q 1 "3 I -3 1 1 1 :i 1 1 H I I 1 1 1 1 1 1 1 1 1 1 1 1 1 13 I 1 31 1 1 31 1 3 1 1 J 1 I ;j 1 H I II H 1 1 1 1 1 1 1 1 1 1 1 1 1 h ] l i n e :! 1 1 1 1 1 i " n j j j & d^ d S > * ■numbers employed In low- skill sectors ■ numbers employed in high- skill sectors Source: UNIDO data base (provided by James Galbraith). Chart 3.13: Brazil in Comparative Perspective: Employment Manufacturing 4000000 3500000 3000000 2500000 2000000 1500000 1000000 500000 0 Employment in India's Manufacturing S A ctn re J d a . i . ! -I j c:-j ;-j 1 1 1 1 H i 1 1 1 1 1 H 11 1 1 1 1 1 1 1 1 1 -S' N * 9 ■numbers employed in low-skill sectors ■numbers employed In high- skill sectors Source: UNIDO database (provided by James Galbraith) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chart 3.14: Korea in Comparative Perspective: Employment in Manufacturing Employment in Korea's Manufacturing Sectors 1600000 n 1400000 -| c 1200000 I 1000000 §• 800000 §■ 600000 iB 400000 200000 0 3 1 ll II H i r l : 3 f l i:! i i l ! i! 1 1 i ;| i f 1 i f ! 1 1 t;l kl q n 1 n ; 1 ; ! i | i i : i f I • i n r a • i > :: ^ :; t • < i o I o %1 ; i H P i i :1 11 \l i 11 i 1 M l i. i l l ; i J - h . i il; i i i .11 1 : 1 : 1 J 1 1 3 1 1 i 1 :3 1 1 1 J 1 £ J " J p *v numbers employed in low^ Bftll sectors numbers employed in high-skill sectors Source: UNIDO database (provided by James Galbraith) Indeed, the large numbers of formal sector unskilled laborers in India are crucial in terms of lowering the organizing potential of labor. Yet research suggests that a simultaneous increase in the informal sector has contributed to a larger gross unskilled labor population. There is a hidden story that unfolds behind the changing levels of formal unskilled labor employment. A relatively rapid increase in the share of casual workers is a phenomenon that has been plaguing India (Deshpande, 1998; Johri, 199S; Papola, 1994; Ratnam, 1993; and Singh, 1993). Interviews with Indian academicians and union leaders suggest that labor-intensive sectors are the most affected by the casualization of the labor force since globalization. How well then does Chart 3.12 133 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. capture increases in employment of unskilled labor, since most employment has been increasing in the informal, unorganized sector where data cannot be obtained? The Growth o f the Informal Sector As developing nations compete and the demand for unskilled labor increases, businesses seek even lower wages and more flexible labor legislation. India’s rigid labor markets are problematic for luring foreign investors as well as increasing the competitiveness of domestic export sectors. Indeed, it would seem that this aversion to labor market rigidities would be intensified with recent exposure to international competition. Now that the domestic economy is no longer insulated from the pressures of global market activity, there is a stronger push for cheap and flexible labor. Much of India’s industries are now turning towards the unorganized informal sector in order to maintain their competitiveness by evading the higher wages and benefits of the formal labor market.2 9 2 9 An alternative to contracting out to the informal sector in order to avoid higher wages is to switch to capital intensive production. However, it seems that in labor abundant economies, the incentive to utilize cheap labor in the informal sector would be greater than implementing more extensive capital intensive technologies. The empirical validity o f this conjecture would be best tested by future fieldwork. 134 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. As a result of the plethora of labor laws, India is well-known to have a rigid labor market. India is often accused of providing “excessive” protections to labor. According to Mamkoottam (1994): A key feature of employer-employee relations in India has been the overwhelming interventionist role played by the state through the enactment of a plethora of labor laws affecting almost every possible dimension of industrial relations. Many of these labor laws are not only archaic, but have also considerably reduced and restricted labor flexibility, and thereby slowed down the process of restructuring. Indian employers and potential foreign investors have been clamoring for radical changes in the existing labor legislation in India allowing for a free hand to practice a hire-and-fire policy (Mamkoottan, 1994:176-7). How might India compete successfully on international markets, given the inflexibility of their labor markets? Scholars argue that one way in which the industrial sector is coping with new exposure to competition is through increased casualization of the work force and thereby increased instability of unemployment (Deshpande, 1998; Harriss-White, 1999; Papola, 1994). Interviews with labor leaders also suggest that shift to employment in the informal sector has been one way that employers are confronting the growing pressure to be competitive. Expansion of labor intensive industries, in particular, are usually subcontracted to the unorganized sector. According to data from the International Labor Organization, this sector comprises almost S O percent of the Indian economy that are not protected (see Table 3.5). 135 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. It is impossible to get accurate data on the size of the informal sector. Papola (1994) claims that the tendency is to avoid application of labor laws and pressure from trade unions and to keep the work force small, employ workers on a more casual basis and subcontract work to smaller enterprises. A recent publication by the International Labour Office, or the Key Indicators o f the Labor Market (1999) is helpful in assessing the size of India’s informal sector relative to other countries. Unfortunately, there is no report of India’s trends over time. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 136 Table 3.5: Comparative Statistics on Size of the Informal Sector* Country Year Number of people employed in the urban informal sector (‘000) Urban informal sector employment as a percentage of urban employment India 1993 26813 44.2 Brazil 1990 16539 44.3 1994 560 47.3 Indonesia 1995 5566 20.6 Philippines 1995 539.3 17.0 Argentina 1990 1714 41.1 1995 • • 45.7 Turkey 1993 1354 15.0 Source: Key Indicators in the Labor Market (International Labour Office: 1999) * Years listed are based on data availability. Data on Korea was also unavailable. 137 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. One way to assess the expansion of informal sector over time is to observe the decline in total social security coverage. T.S. Papola best summarizes: Social protection o f labor, beyond a point, may prove counterproductive in so far as it discourages investment and growth and consequently employment. Further, when such protection is available to a few workers in one segment and not to others who, in feet, constitute the majority, as is the case in India, employment takes place in the unprotected segment and thus proportion of protected workers declines. In the process, the very objective of providing social protection to a large majority, if not at all, gets defeated (Papola ,1994: 119). The payment of social security represents an officially recognized work contract. Theoretically, increases in formal sector employment should reveal a rising trend in aggregate social security coverage. According to Telles (1992), the lack of social security generally means cheaper production, workers and employers are avoiding costs of paying health and unemployment benefits, paying less than minimum wage, and are hiring and firing on a casual basis (Telles, 1992). Charts 3.IS and 3.16 present India and Brazil’s social security data from the ILO. Data on Korea were not available. 138 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chart 3.15: India in Comparative Perspective: Social Security Coverage India's Total Social Security Receipts (as % of GDP) 4 1 3 \ 2 J 1 1 0 - I , ------- , ------- , ------- « -------, ------- , ------- , ------- , 1974 1975 1980 1982 1985 1988 1989 1991 Source: International Labor Office: The Cost o f Social Security. Various Tears Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chart 3.16: Brazil in Comparative Perspective: Social Security Coverage Brazil's Total Social Security Receipts (as % of GDP) 5.7 5.65 5.6 5.55 5.5 5.45 5.4 5.35 Source: International Labor Office: The C ost o f Social Security. Various Years 1981 1983 1986 1987 1989 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Applying Telles’ (1992) method of using social security coverage to assess the growth of the informal sector strongly suggests that there is a growing duality characterizing India’s labor markets, particularly after 1985. Effects on Labor Power Labor’s weak bargaining power seems to have become even weaker since the attempts at liberalization in the mid eighties. The Seventh Five-Year Plan (1985-1990) was the first plan, which focused less on social justice and more on improving efficiency the of labor. Rather than repeating previous calls for improving labor welfare, this Five-Year Plan emphasized higher productivity (see Singh, 1989). The New Industrial Policy adopted in 1991 aims to make India’s industry more competitive in international markets by way of improving the flexibility of the labor market. General data on labor activity suggests that labor strength has been decreasing in India. Chart 3.17 illustrates the declining trends in strikes, lockouts, workers involved and workdays lost. 141 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chart 3.17: Trends in India’s Labor Disputes W S _____ 8 7 8 ______t oo______ 8 89 ______1 9 8 8 wxfadays not wxked (thousands) nurrber of strikes and lockouts w orioers involved (thousands) Source: International Labor Office: Yearbook o f Labor Statistics. Various Years. However, taken by itself, this data provides little conclusive evidence that labor has been weakened since late eighties. This data can be interpreted in two ways. First, the declining number of strikes and lockouts, as well as less numbers of workers involved in disputes, may indicate that labor is relatively satisfied and their demands have been met. Otherwise, declining trends in the above categories may suggest that workers are facing collective action problems and articulate their demands. In order to resolve this ambiguity, a simultaneous assessment of India’s political atmosphere is necessary in order to ascertain what these trends indicate. Put simply, do the declining trends from the late eighties and onwards indicate that labor is less satisfied and “has less probability of securing the outcome that it prefers?” 142 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. It has already been shown that labor’s position had not been improved since globalization. The problems of multiple unions, the lack of uniform labor code, and state power in tripartite bargaining has continued to persist until the present. Yet the question is whether or not labor’s position has worsened in the last decade? It is during the period of reforms (1985- mid 1990s) that relations between labor, business and the state in India have been extremely antagonistic.3 0 The Indian government, for the first time in its post-independence history, has managed to impose policies that aim towards reforming the state-labor relationship. Data (i.e., declining social security coverage, less government social spending) strongly suggests that there has been a structural change in India’s labor markets. These changes are a consequence of labor’s growing inability to act collectively and affect political outcomes. The politics of labor and the economics of globalization have come head to head. With little question, labor has not been satisfied with the government’s attempts to reform the state-labor relationship. Labor has been actively opposing the new exit policies of the administration. Their primary argument is that labor is not the principal cause of industry sickness, but rather, inefficiencies are the result of low managerial productivity and unbridled influx of technologies and multinationals (Ratman, 1993). Some of labor’s demands in the early 1990s include continued budgetary support to public sector units referred to the Board of International and Financial Restructuring 3 0 This information is based on interviews with Vasant Gupte, Secretary of the Hind Mazdoor Sabha union. 143 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (BIFR) as “sick,” assurance that there would be no exit policy, retraining and voluntary retirement schemes to avoid retrenchment, and to halt privatization of public sectors (Ratman, 1993). In particular, labor has been opposed to the 1985 Sick Industrial Companies Act. This act was an effort to close down inefficient industries that were supported before India opened up to competition from international markets. By the end of 1988, 240,000 sick industrial units had been identified (Papola, 1994). The aim was to restructure Indian industry by targeting those industries that were not competitive due to the use of obsolete technology and the employment of excess manpower (Mamkootam, 1994). Labor feared the massive loss of jobs that would result. It appears that labor was successful in obtaining some minimal concessions. For example, the administration has been stalling efforts to close down the sick industries. Also, the Industrial Disputes Act, which require government approval of closures, lay offs, and retrenchment for all plants employing 100 plus workers, has not yet been amended. These two Acts have been highly contested by labor. Nevertheless, for the first time, these Acts have been placed as on the primary political agenda. The probability of labor securing the outcome that it prefers has been, thereby, lessened considerably. Significantly, labor has lost much of their previously held benefits as well as power. By aligning itself with the political left, labor has been part of the powerful coalition that has been able to successfully stall previous efforts at liberalization (see Denoon, 1998). India’s latest attempt at liberalization, however, has not yet been as 144 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. affected by labor’s discontent. In the end, labor been unable to arrest the process of liberalization and the growing exposure to global markets as it has in the past. Despite the fact that the BIFR efforts have so far not been extremely successful, the process of labor retrenchment has been occurring. Some estimate approximately an additional unemployment of eight to nine million people as a result of industrial restructuring and structural adjustments. Indeed, Mamkootam (1994) finds that many industries have successfully managed to “trim the fat.” It is possible that the Indian government is stalling on strengthening BIFR. and its handling of the sick private sector firms because it would cause unemployment in India to skyrocket and thereby, result in high levels of instability. This leads to the conjecture that it is not the bargaining power of labor that is slowing the pace of restructuring, but rather the government’s conscious political strategy. In addition, government measures to assist labor during the transition to liberalization have only managed to help a privileged few. The National Renewal Fund (NRF) was created in 1992 in order to protect workers affected by restructuring process. The Fund was also expected to retrain these affected workers. Yet in effect, only the most technically qualified people or those close to retirement were able to utilize the Voluntary Retirement Scheme (VRS) (Mani, 1995). Retraining would have been most beneficial to unskilled workers, yet the amount allocated to unskilled workers was entirely too small (Mamkootan 1994). Thus, the declining trends in Chart 3.17 (i.e., numbers of strikes, etc.) should be interpreted as decreasing bargaining power of labor. Labor is clearly not satisfied with 145 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the status quo. Documents from labor organizations such as The Monthly Journal o f the CITU reveal that labor is vehemently opposed to the process of globalization. They fear the “perilous moves to kill jobs and the surrender of self reliance.” In September of 1991, the following report was published: At the moment the biggest task before the working class is to make the people know what’s in store for them. Trade Union movement has to step up massive counter campaign exposing the real character of reforms... [T]he malice of our economy lies in the purposeful neglect to expand internal markets and not in its failure in getting structurally adjusted with so called “global economy.” At the present context, no economy of ours can survive competition with the transnational sharks. ...In view of the situation, the working class and the democratic section should make its position clear before the nation {The Working Class, September 1991, p. 20). This leads to the conclusion that the declining trends in Chart 3.17 indicate that labor’s ability to act collectively has decreased. In sum, labor’s power in India is in an interesting paradox. Strong labor laws that protect approximately 10 percent of the workforce, have at the same time, weakened the labor movement as a whole. Country specific factors such as the existence of multiple trade unions, lack of uniform labor code, and absence of “true” tripartite bargaining make it even more difficult for labor to organize post-globalization. At the same time, fundamental changes to labor legislation have not occurred alongside 146 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. globalization, encouraging more businesses to turn to the informal sector.3 1 Ultimately, labor’s main problem is the disproportionate protections received by the organized labor groups. As Waterbury (1993) argued, labor must be skilled, not easily replaceable and strategically located in order for unions to bargain from a position of strength. The majority of labor in India does not fit into this category. Textile Sector A brief look at India’s textile sector provides a more targeted investigation of the effects of the interaction between globalization and the bargaining power of labor. It is interesting how in this sector, the Indian government, on the one hand, has established one set of laws that favors labor, and, on the other hand, promoted another set of laws privileging small-scale production that thrives on informal labor. As a result, the proportion of those protected by labor laws and welfare benefits has diminished in this sector. Significantly, government attempts to increase competitiveness by privileging small-scale production has weakened the ability of labor to act collectively. The textile industry is one of the largest industries in India, particularly with respect to its share of exports and number of employees. Since the 1940s, however, India’s cotton mills have undergone a metamorphosis from a large integrated factories 3 1 When the relaxation of India’s labor laws will occur cannot be predicted at this time. It’s true that labor’s bargaining power has been weakened in India. However, the Indian government has obviously chosen a gradual path to labor market reform. The reasons for the slow pace of this reform is a subject for future research. 147 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. with hierarchical organization to decentralized small weaving factories. This shift from the large cotton mills to powerlooms characterizes the recent history of the textile industry in India. The smaller weaving factories (or powerlooms) employ mostly by unprotected, low-skilled laborers. This transformation is significant because of its effects on labor and its organizational capabilities. Evidence suggests that India’s globalizing economy is bound to witness a rise in powerlooms and, consequently, have a negative effect on labor power.3 2 The cotton mill industry dominated India in the 19th century. As one of the world’s largest by 1900, the 20th century began to witness a gradual dismantling of the mills (Roy, 1998). Three types of government policies were said to encourage the shift to the labor intensive power looms: restrictions on mills regarding capacity expansion in weaving in force by 1984-1985,3 3 protections for organized labor driving up labor costs in the cotton mills, and higher taxes on mill cloth (Roy, 1998). Eventually, Roy (1998) argues, employers of cotton mills would secretly collude with powerlooms in order to lower their labor costs. The value of powerlooms to India’s export economy is impressive, and continues to increase with India’s globalization in the early nineties. Currently, cloth and garment exports constitute about 30 percent of India’s exports and have been the 32 Evidence for this claim has been derived by interviews with union leaders and academicians from the Tata Institute for the Social Sciences in Bombay, India. Most of the data collected in this section is from the Maniben Kara Institute in Bombay. 148 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. greatest beneficiary of trade reforms since the early 1990s. For example, the Textile Order of 1993 eliminated licensing restrictions on powerlooms. Textiles production has surged since the mid 1980s (see Table 3.6). Powerlooms share of total exports of cotton cloth was 57.6 percent in 1995-6 (Adjankoar et a l., 1998). Significantly, the second half of the 1980s was a period of rapid growth in world exports of clothing/textiles (Ramaswamy & Gereffi, 1998). India’s competitors are China, Hong Kong, Taiwan, and South Korea, of which India and China offer the lowest labor costs. It is interesting that India has a comparative advantage in cotton cloth and its growth in garment exports has been achieved with essentially no reliance on imports. 3 3 This analysis suggests that the reason that the Indian government initiated policy protections for the powcrloom sector is to bolster competitiveness. Ultimately, it seems at least one objective of privileging the powerloom sector was to lower production costs of this labor-intensive good. Evidence to the contrary has not yet been found by this researcher. 149 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table: 3.6: India’s Textile Exports Year Exports of all commodities Export of textiles % of Share of textiles in total exports 1981-82 7998 1336 17.13 1985-86 10487 2086 21.47 1990-91 32555 8251 25.34 1994-95 82609 26599 32.20 Source: D. B. Ajgaonkar, MiC. Talukdar, Vasant Gupte. (1998) Profile o f Textile Industry in India. Maniben Kara Institute. Roy (1998: 898) states that “ it is fair to assume that nearly 100 percent of the non-cotton cloth exports, and nearly 100 percent of the cloths, which are converted into garments and exports from India, [also] come form powerlooms.” Now, hardly six or seven percent of production comes from composite mills (Ajgaonkar et al., 1998). The popularity of the powerlooms has increased because they have low labor costs and allows mill owners to escape from the system of labor regulation. Typically, powerlooms are small firms that have low capital needs. Weaving is the primary activity conducted within the powerlooms. The yam is bought and processed at the market. Significantly, workers in this sector are not covered by the Factory Act that provides labor protections. In 1991, Roy (1998) estimated that powerlooms employ four million workers and accounted for 20 percent of all wage labor in industry, and 33- 150 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 35 percent of wage labor in the informal sector industry. Cotton mills are thus notorious for their use of covert casual labor. The advantage of powerlooms is their moderate capital requirement, cheap and docile labor, low overheads, long working hours, no labor welfare obligations, flexibility in changeover of production patterns, and closure of powerlooms during slack periods (Ajgaonkar et al., 1998). Interestingly, the textile industry has historically had the most powerful labor movements in Indian industry. In the end, the ability to command government resources depends on the political power of unions. The problem is that unions protect a disproportionately small number of the population and this number is further dwindling with globalization. In the past, textile unions staged a number of successful strikes demanding higher wages and benefits. For example, the Coimbatore District Textile Workers’ Union led statewide strikes for medical allowance, casual leave, travel allowance and higher bonuses. According to Ajgaonkar et al. (1998) they were successful in securing a number of these demands. However, since 1983, labor has more been more cautious in articulating their grievances because of the looming threat of closing down “sick” industries (Ajgaonkar et al., 1998). The break up of the industry into smaller units resulted in the expansion of the informal sector. Ajgaonkar et al. (1998) claim that the second result of fragmentation is: ...the lack of labor organizations. The scattered nature of the industry, mobility of the employers as well as of labor, and constant fear of job termination are some of the negative factors that come in a big way in the efforts to organize the workforce. Thus the labor remains unorganized, giving free hand to the employers to exploit it through long hours 151 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. o f work, low wages, no continuity o f job, no social security, and no enforcement o f labor laws. These advantages, though mean, help the decentralized sector to enter into unfair competition with the organized sectors o f the industry who in turn get tempted to break up into small units and reap the advantages thereof. Thus, the vicious circle goes on and the formal organized sector gets thinner and thinner day by day (Ajgaonkar et al., 1998:54). According to Vasant Gupte, Director of the Maniben Kara Institute,3 4 the famous January 1981 strike of textile workers was a turning point in the textile sector’s labor power. Even though the strike for higher wages continued for one year, mill owners and the government did not bow to labor’s demands. Gupte claims that since then, labor has been on the defensive and has become more cautious in bargaining for benefits. He argued that increases in the practice of contracting out to the informal sector has made labor even more difficult to organize. Up until late eighties, more than 150 factories belonged to textile processing unions. In the last 10 years, according to Gupte, hardly 20 factories remain affiliated with their unions.3 3 Thus, India’s strict labor laws and government support of powerlooms have resulted in a no-win situation for labor. The relatively higher technology based composite mills are not as competitive as the lower technology powerlooms that employ informal labor. Larger factories that are surviving do so by contracting out most of their labor intensive work (i.e., weaving, spinning) to informal labor. As a result, formal employment is not being generated and the informal sector is increasing. Labor is, *The Maniben Kara Institute is the education and research institute of the Hind Mazdoor Sabha (One of the major Indian textile unions). 3 5 Note that the number of factories affiliated with unions has decreased, but as shown earlier, the number of unions has increased. This further illustrates the increased fragmentation of the Indian labor movement 152 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. therefore, finding it much more difficult to organize and pressure the government for policies in their favor. Conclusion Case evidence from this “average” LDC provides further confirmation of the conventional efficiency hypothesis presented in Chapter 2. India’s welfare spending trends were well captured by the LDC model. During the period of relatively higher social welfare commitment, India remained isolated from the international sphere by embracing autarchy and protectionist policies. Government social commitments, revealed by its pro-labor legislation and its trend in welfare spending, showed increasing trends until the mid 1980s when India began to globalize. As predicted, welfare expenditures in India began to fall as the ratio of unskilled labor increased during globalization. Bargaining power of labor was also weakened due to an expanding informal sector, the multiplicity of unions, state dominance in tripartite meetings, and the lack of a uniform labor code. These latter variables that emerged from the case studies represent relevant factors that were excluded from the LDC model. Overall, however, the LDC model and the estimated coefficients well accounted for the welfare developments in the average developing country. The next chapter investigates the cases of Brazil and Korea, which defied the average LDC’s welfare trends. Investigating these two countries brings greater understanding to the importance of democratization and the decline of the unskilled labor ratio. Similar to India, the state-labor relationship will also be paid close 153 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. attention. The objective of the following two case studies will be to assess why and how “deviant” countries followed a different trajectory with regard to changes in their welfare expenditures as a share of GDP. 154 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 4 INDIA’S WELFARE STATE IN COMPARATIVE PERSPECTIVE: BRAZIL Brazil deviates from the sample of 53 LDCs in two obvious ways. It exhibits a higher than average level of welfare spending and the trend of welfare expenditures relative to GDP increases over time. Nevertheless, Brazil’s welfare history illustrates the two main propositions of this study that: (I) globalization in Brazil increased alongside the gradual shrinking of the unskilled labor population; and (2) the changeover to democracy in the mid 1980s resulted in a distinct increase in welfare expenditures as a share of GDP. Other relevant variables that were not included in the LDC model but emerged from the Brazilian case study were the history of state corporatism and the nature of the state labor relationship that evolved. Interestingly, the country specific issues that emerge from the Brazilian case reveal that although welfare as a share of GDP has been increasing, their government’s social commitment is strongest for only a small minority of the population. As Brazil’s state-labor relationship evolved under the umbrella of a state corporatist system, the distribution of welfare was targeted towards a select labor aristocracy. Ultimately, the undermining of redistributive goals of such traditional social programs highlights the contradictory and complex nature of Brazil’s welfare system. 155 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Therefore, the Brazilian case provides added understanding that even though trends in LDC welfare expenditures might be increasing alongside globalization, this does not necessarily indicate that the compensation effect is occurring in LDCs. Evidence from Brazil’s welfare history precludes the conclusion that the Brazilian government is protecting the majority of its citizens from the risks and uncertainties of globalization. Instead, Brazil’s social welfare system suffers from an interesting duality. On the one hand, it maintains the appearance of a strong LDC social welfare system by maintaining high levels of benefits. Yet, on the other hand, Brazil’s social system privileges and protects a relatively small and elite labor group (i.e., skilled laborers, middle class strata). Thus, the Brazilian cases suggests that LDCs that show rising trends and levels of welfare spending as a share of GDP are not necessarily taking redistribution concerns into consideration and should be reviewed with caution. Dependent Variable: Brazil's Social Welfare Commitment Brazil’s welfare state is comparatively strong, with respect to its social welfare spending as a share of GDP. Brazil was one of the pioneers in the field of LDC social welfare development Unlike India, Brazil’s social sector maintains high state financial responsibility in welfare contributions and has recently tended towards universalization.1 The most important contrast to the Indian case emerges from Brazil’s success increasing the amount governments spend on welfare over time, as well 1 Universalization means that all laborers, and not just organized workers, are eligible. 156 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. as an improvement in the types of social welfare programs. Second, Brazil’s state corporatist system and its state-labor relations sets it apart from both India and Korea. Most independent labor movements in Brazil have been subsumed under state control, and have only recently been politically successful. By incorporating labor into the state system at an early stage in labor history (1930s), the characteristics of the labor institutions that emerged in Brazil were very different from both India or Korea. Trends and Levels of BraziTs Social Welfare Spending On average, Brazil’s level of expenditures as a share of GDP has been increasing steadily since the mid 1980s. There appears to be two structural shifts in state social spending. The first structural shift during the early 1980s constituted Brazil’s one-time drop in the amount spent on such programs. Many attribute this drop in expenditures to the debt crisis when international financial organizations pressured countries such as Brazil to adopt austerity measures (Huber, 1999). At this time, Brazil experienced a severe contraction in economic activity, shrinking of employment, falling real incomes and a decline in social services (Huber, 1996). The second structural shift occurred in the late eighties. Interestingly, it is during this time, in 1987, that the Brazilian re-democratized and the government adopted more generous labor and social policies. Thus, all three indicators of the state social commitment (i.e., trends, social and labor policies) suggest that an upward structural shift occurred in the late 1980s. However, Chart 4.1 suggests that Brazil’s government social welfare commitment has been ambivalent. Although welfare spending as a percentage of GDP 157 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. has been increasing, measuring social spending as a share of total government spending depicts a different reality. From this perspective, government commitment to social welfare fell after the debt crisis in the early 1980s. After this decline, welfare expenditures as a proportion of total government spending rose again in the late 1980s (similar to the share of GDP trend), but never reached the same level of spending as in the 1970s. Interestingly, overall government spending relative to GDP has also been increasing (not shown here). This suggests that as GDP rose (assuming that government revenues also rose), the Brazilian government began privileging other spending needs over those of labor (e.g., defense, subsidies to industry, etc). These two trends are not necessarily contradictory since, in all likelihood, real welfare spending continued to increase.2 One way to interpret these contrasting trends is that while the government’s social welfare commitment to labor may not have decreased, it has not been a spending priority since the debt crisis. Interestingly, Brazil’s social security expenditures as a percentage of GDP are at a level higher than the majority of LDCs. Spending fluctuated between 9 to 11 percent of the gross domestic product (GDP) within the last decade. 2 Trends in real welfare spending could not be shown here since the consumer price index (CPI) for Btazil over a long period of time (1972 to 1995) is difficult to obtain. 158 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chart 4.1: Government Welfare Spending in Brazil BraziTs Social Welfare Expenditures (1972-1995) s o lo,4 0 3 i 3 0 £i® 5 ” 1 0 T"------ -- ---------------------------- * r 120 ■ ion •ao a. o •ao O n -4.0 5 5 - zu -0.0 —— as%cf total GcvSpendng -*-as%ofGCP Source: IMF, Government Finance Statistics: various editions In contrast to India and most LDCs, the Brazilian state takes substantial financial responsibility for social welfare over time (see Table 4.1). 159 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 4.1: Contributions to Social Security Receipts in Brazil (in millions of national currency units) Year State contribution Employer Contribution Employee Contribution 1981 130427 1165989 69544 (9.2%) (83%) (5%) 1985 3.09 52.06 12.37 (4.3%) (72%) (17%) 1987 5.30 336 190 (0.7%) (52%) (29%) 1989 14267 36660 17499 (20%) (51%) (24%) • Numbers in parenthesis represent contributions as a percentage of total receipts. Note that remaining contributions come from special social security taxes, participation of other public authorities, income from capital and other receipts. However, contributions from these sources in LDCs tend to be very small, if any. • Years chosen based on data availability. Source: The Cost o f Social Security. International Labor Organization: Various years It is interesting to compare these patterns of financial contributions. From the onset, coverage in Brazil has been financed by compulsory contribution from all three sources: the employees, their employers, and the federal government. It was as early as the 1930’s during the Getulio Vargas regime, which mandated that all three sources should contribute equally (Malloy, 1979). In India, the states set contribution to welfare programs was more arbitrary. The Brazilian state contributed approximately 20 percent in 1989, in contrasted with India’s 4.5 percent state contributions in that same year (The Cost o f Social Security 1990). Based on the increasing amount of state contributions in 160 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. recent times, it is fair to say that both the state and the employer share the burden of social expenditures in Brazil.3 The 1989 figures confirm that a structural shift did occur in 1988 after democratization. State participation increased, while the proportion of employer contributions falls over time. Employee contributions, on the other hand, after increasing substantially early on, began to decline again in the late eighties.4 Thus, factors affecting the structural change in the late eighties when the state made a substantial shift will be of primary concern in the Brazilian case. The Characteristics o f BraziVs Social Welfare System A review of Brazil’s social welfare programs reveals that there are a greater variety of and more extensive benefits offered than India. Brazilian welfare programs date back to the early twenties. The first law or the Lei Eloy Chaves (LEC) was passed back in 1923. LEC laid the legal and conceptual base for the Brazilian social system (Malloy, 1979). Brazil offers a wide range of protections, from old age, disability,' and death, as well as sickness and maternity benefits, protection from work injury, unemployment and family allowances (Social Security Programs Throughout the World 1997). Most of these laws were implemented by the late seventies. 1 Although employer contributions in India were as high as 57 percent in 1985, recall the absolute level of welfare benefits is very small compared to Brazil. This percentage fell to 27 percent by 1991. 4 Recall that trends in India’s social security contributions suggest that, over time, the burden of social security payments are shifting towards the individual. 161 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The fact that Brazil has both family allowance and unemployment benefits reveals that it is a relatively advanced welfare state. Not only is it rare for a developing country to offer unemployment insurance, but it is even more unusual for government to take full responsibility for the whole costs of unemployment benefits. This is indeed exceptional, given that even governments of most developed countries do not accept the entire financial burden of unemployment. In sharp contrast to India, Brazil has achieved impressively large legal coverage of the labor population. Between 70 to 80 percent of the economically active population was covered by social security by the late 1970s (Huber, 1996). Brazil is also rare in that it provides benefits to portions of the rural sector through FUNRURAL. However, as Huber (1996) points out, the overall level of rural benefits is extremely low. The 1988 constitution implemented substantial changes with respect to social benefit programs. They were based on the principles of universal coverage, equivalent urban and rural benefits and guaranteed benefit levels (Barreto et al., 1994). Most of the previous social welfare laws were amended by 1991, providing for higher cash benefits in most of their existing programs (i.e., old age, disability and death, maternity benefits [in 1988], work injury and unemployment). For example, the new constitution guaranteed a minimum monthly income to the broader population, formerly available to the elderly and handicapped who had contributed at least once to the social insurance system (Barreto et al., 1994). Thus, in theory at least, social welfare became established as the right of all people in 1988. 162 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Despite these efforts, the informal sector is still excluded from social benefits in Brazil as it is in India. The problem of inequality in entitlements then holds for both Brazil and India. Yet, in Brazil, benefits allocated to the privileged sector are disproportionately higher relative to the rest of the covered population. For instance, earlier LEC only brought legal coverage to those employed in service sectors, or mostly white collar employees. Today, the privileged groups receiving the bulk of social benefits in Brazil are the civil servants, military, middle class strata and urban salaried employees (mostly white collar workers) ( Mesa-Lago 1978). India’s labor aristocracy, on the other hand, is not as well defined. Thus, although Brazil approximated a universalized welfare system by the late 80s, in reality it fell far short of this idea. Brazil’s relatively strong welfare state has evolved into a system of patronage and clientelistic favors. Huber (1996) states: [N]on-payment of contributions left many low-income earners without coverage, and the requirements for documenting a social security pension or social assistance claim were beyond the record-keeping abilities of most low income people. This situation perpetuated the tradition of patronage in Brazilian politics, as low-income claimants needed an intermediary to deal with the social s e c u r i t y and social assistance bureaucracies. Accordingly, control over social security agencies remained a coveted political prize for parties and individual politicians at all levels of government... (Huber, 1996: 171). Consequently, legal coverage was first brought to the most radical workers who were mostly skilled laborers (Malloy, 1979). Industrial workers who were in the least developed sectors of the economy were the last to receive protections. Mesa-Lago 163 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (1978) argues that, early on, the elite pressure groups (i.e., industrialists, upper-level professionals, military personnel, policemen, civil servants and white collar workers) grew to be the most significant force of social welfare. He argues that social security has been “manipulated to gain electoral support of a particular clientele, legitimate a spurious political regime, and to satisfy the needs and co-opt powerful pressure groups which threaten the status quo” (Mesa-Lago 1978: 8). Several scholars have supported this argument (, Midgley, 1984; Weyland, 1995,1996). Weyland (1995) convincingly argues that it is this clientelism that has prevented Brazil’s social welfare system from becoming more equitable. Reforms that would provide benefits as a universal right now pose a deep threat to powerful politicians and privileged beneficiaries alike. Thus, inequality has increased despite the changes implemented in the 1988 Constitution. This suggests that the higher benefits are still not reaching the segments of the population who are in most need, but rather further enriching the privileged labor class. Brazil's Labor Market Policies What explains the regressive nature of Brazil’s social welfare system? Overall welfare expenditures relative to GDP might be increasing, but a more equitable distribution of benefits and welfare as a share of total spending is not. The system of patronage that developed concurrently with Brazil’s welfare system cannot be fully understood without reference to its state-labor relationship. State corporatism in Brazil both subsidizes and controls the group whose interests are to be represented. It is a 164 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. nonpluralist system of group representation, deliberately designed to promote “collaboration” between labor and capital under the aegis of the state (Cohen, 1982; Collier & Collier, 1991). This system established a pattern in which social welfare benefits were granted in place of labor cooperation. More specifically, in exchange for the right to strike, form independent unions and collective bargaining, labor received state protection and benefits.5 It was Vargas in the 1930s who readily extended legal protections and labor benefits to a powerless working class in order to preempt the formation of a cohesive labor class (Cohen, 1982). Successive regimes muted the relationship between economic fluctuations and strike activity by imposing repressive labor tactics.6 Government held the power to interfere in labor relations and suppress their demands by stopping strikes, preventing union formation, and setting policies in favor of corporations (Sandovol 1993). 5 This is not to suggest that there were no independent labor movements that emerged in Brazil. The Workers’ party (Paratido dos Trabalhadores, PT), for example, emerged in the early 1980s based on support from increasingly autonomous and powerful movement of Brazilian unions. However, the PT and its relationship with the unions were often faced with crises and remained largely marginal to the political process through most of the 1980s. See Keck (1992) The Workers Party and Democratization in Brazil. 6 Sandovol (1993) discusses four phases of military rule. He characterizes the first phase after the 1964 coup (1964-1968) as highly repressive towards labor, as the number of strikes declined dramatically. The second phase (1969-1977) was characterized by heightened political repression and accelerated industrial growth. The third phase (1977-1980) was the Abertura Politico period marked by the beginning of redemocratization of the political system and resurgence of strike activity and popular discontent The last phase (1981-1989) was the end of the military government occurring alongside increased militancy and organization of the working class. 165 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Marquez (1995) lists the 1988 Constitutional reforms to the old Labor Code restrictions on union organization and collective bargaining: • Prohibition of government intervention or interference in union activities. • The right to strike was turned into a constitutional right of all workers in the country. • The formation of national unions, which also gained the right to argue the constitutionality of a law or act of government in the Supreme Court of Justice. • Workers gained the right to organize elected workers’ council that could negotiate with the employer, in enterprises with over 200 employees. Yet many scholars question if changes in the formal labor legislation in 1988 were merely cosmetic (Amadeo et al., 1995; Barreto, 1994; Hubert 1996; Marquez 1995; Weyland, 1995). Given the history of corporatism and patronage, these scholars suggest that many legal loopholes render the reforms as meaningless. In reality, they suggest that the state still maintains ultimate control over labor. Thus, until 1988, it was primarily the level of social benefits that best illustrate the extent of state’s commitment to welfare in Brazil. Benefits were high pre- democracy, but rose to even higher levels afterwards. Welfare data, a review of social welfare programs, and changes in labor regulations leave little question that a structural change in Brazil’s social commitments appears to have occurred alongside redemocratization in 1988. Indeed, policy changes towards labor and their social protections seem particularly impressive. Nonetheless, Brazils social history must be 166 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. taken into account in order to interpret these changes. Brazil’s state-labor relations have been more repressive than state-labor relations in India. Italian corporatism provided the Brazilian model, which adopted the goal of achieving labor cooperation by way of social welfare benefits distribution. The state became very interventionist in labor matters and, at the same time, established a strong system of patronage politics. Thus, Brazil’s late 1980s shift in social commitment must be reviewed with caution (see Weyland, 1996). The Independent Variables: Globalization and Bargaining Power of Labor Comparative case studies show that globalization has varying impacts on welfare policies in Brazil, India and Korea. Brazil does not prove to be a unique case with regards to understanding its rising trends in social benefits. Brazil’s success in increasing its skilled labor sectors relative to unskilled sectors with globalization has affected greater welfare expenditures. The relative growth of skilled labor has been higher than the average LDCs (i.e., India) is revealed by the proportional growth of skilled to unskilled exports, as well as an apparently smaller expansion in the informal sector.7 Brazil’s case also suggests that re-democratization in 1985 under President Jose Samey created new opportunities for labor. Thereby, the evidence suggests that 7 This is not to suggest that the informal sector has not been increasing in Brazil. Rather, the evidence suggests that it has not been increasing at the same rate as the informal sector in India. 167 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. labor was successful in negotiating greater freedoms and more social benefits with the recent turn to democracy. Finally, the historical realities of Brazil’s corporatist system has mediated between globalization and its effects on social welfare spending in Brazil. Their corporatist system has created a welfare system that caters to a privileged labor group, while the political rights of labor as a whole have been suppressed. With respect to their social welfare expenditures, three variables account for the divergent paths taken by Brazil and India. First, Brazil faced the pressures of international market competition many decades earlier. Encouraging and collaborating with foreign investors as early as the 1920s, Brazilian labor was never afforded the same amount of labor market protections as labor in India. At the same time, globalization in Brazil helped encourage the production and exports of high-skill goods in Brazil. Rising employment in these skilled sectors increased the potential for collective action and, subsequently, affected a growing labor movement. Second, Brazil’s re-democratization in the mid 1980s contrasts with India’s longstanding democracy. This transition, from a military government to a more democratic one, required the government to reform its earlier repressive tactics towards labor and continue benefits to privileged labor groups. Not only does Brazil continue to use benefits to maintain clientelistic relations with the labor aristocracy, but also democracy now requires that social spending be broadened to include larger labor groups. 168 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The final deviating factor in Brazil’s state-labor relationship was its history of state corporatism. This system created institutional rigidities with respect to the distribution of welfare benefits to privileged labor groups. These benefits became part and parcel of the patronage politics that evolved between the state and these particular labor groups. Such clientelistic relations served to segment the labor movement early on. Thus, while the level of labor benefits were high, the distribution of benefits was extremely uneven. Ultimately, Brazil emerges as a country where there the correlation between globalization, the bargaining power of labor and welfare spending is linked to its history of state corporatism. Brazil’s story is a complex one. While it appears that Brazil is making stronger headway towards welfare statism than India, in reality the gains have been unequally divided among social groups. Globalization in Brazil Brazil’s experience with globalization is quite different from that of India. It is true that both India and Brazil followed ISI strategies. Yet India achieved a level of isolation from the global economy that was far beyond levels attained by Brazil. The process of globalization began much earlier for Brazil than it did in India.8 Not only was Brazil more successful with their export markets, but Brazil also did not impose the 1 The concern here is not upon the exact date that Brazil established links with the international economy (whether or not it began in iate I9d > century with commodity exports). Rather, the focus of this study is the fact that such links were established before 1972. 169 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. same level of capital controls as India. Like India, Brazil initially restricted imports, but, unlike India, Brazil entertained foreign investors and enjoyed greater export growth. Since their experiences with globalization varied, it is highly possible that globalization impacted their domestic political economy differently. Brazil’s social policies faced the effects of global competition much before India and eventually fell I n d ia a n d B r a z i l : E x p o r t a n d I m p o r t s (% o f G D P ) 30 25 20 15 10 5 0 + ■ N ® N ® India's trade levels Brazil's trade levels Chart 4.2: Comparison of India’s and Brazil’s Exports and Imports Source: World Bank (1997) World Development Indicators. upon a path of dependent development (Evans 1995).9 Ultimately, Brazils levels of global integration since the early 1970s far surpassed India, particularly with respect to capital flows. 9 For instance, by allowing multinational corporations (MNCs) to participate in the Brazilian economy, ISI policies eventually resulted in increased reliance on foreign investment (Encamation 1989). 170 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Brazil intensified its links to global markets through a massive export drive which began in the early 1970s. Increasing manufactured exports became a deliberate policy based on export subsidies and other arrangements (Bradford, 1990). The result was a surge in trade-GDP share from 1967-1973 and 1981-84. In contrast, Lall (1999) argues that India’s complex system of export incentives created a host of distortions to the export sector. India would often end up exporting goods with high import content at a lower cost than “the foreign currency cost of imports embodied in them.”1 0 Chart 4.2 illustrates the higher levels of trade to GDP of Brazil, relative to India, during the earlier stages of ISI (or the 1970s and early 1980s). The decline of trade as a proportion of GDP in Brazil after the mid eighties can be attributed to growth in GDP, as well as the delayed impact of the import substitution programs of the 1970s.1 1 Baer (1989: 117) states that imports of consumer goods began to decrease after 1983 due to the greater domestic output of import-substitution industries. The fall of petroleum imports (by 9.7 percent) also contributed to this decline as domestic production of petroleum and the energy substitution programs increased (Baer 1989).1 2 1 0 Lall (1999) explains that export subsidies were usually tied to the import content of exports, instead of to those with high domestic value added. Thus, because there were different effective exchange rates for exports and imports, in addition to the practice of over-invoicing exports, goods were often exported for less than the cost of the imports embodied within them. 1 1 Recall that real export growth has been increasing from 1972 to 1995 (not shown here). 1 2 Longo (1997) argues that the Ming trend in exports in the early 90s, was due to the suspension of the foreign debt service. He argues that current recovery of the export sector depends upon the pay off of debt arrears, reviving exports, and accelerating exchange rate evaluations. 171 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Brazil and India do share like histories in one aspect of globalization. Similar to India, foreign trade in Brazil was not liberalized until the late eighties. Early in 1991, the Collor de Mello government announced among the most far reaching tariff reductions to be phased over the 1991-1994 period. Liberalization programs also entailed the elimination of administrative controls on imports, gradual import tariff reduction and changes in the exchange rate regime (Silber, 1997). Notice that trade liberalization did not have a structural effect on the Brazilian economy as it did on the Indian economy (see Chart 4.3). Steady increase in India’s trade levels after mid 1980s suggest that there was a structural shift in the economy. In Brazil, on the other hand, changes in trade levels after liberalization were incremental and thus do not indicate that a structural shift occurred. How has different trends in globalization affected their economy? It is reasonable to argue that both Brazil and India’s export profiles involve more technologically sophisticated goods. The difference is that since Brazil’s large industrial groups were less insulated from the global competition, they were forced to be more efficient and had considerably more success in producing manufactured products for export markets (Evans, 1995). Collaboration with technology intensive MNCs such as IBM became major contributors to Brazilian exports. Significantly, by 1985 exports of more skill intensive sectors, such as chemicals products, transportation, machinery increased from 43.5 percent to 67.1 percent Unskilled, low-technology exports such as textiles and apparels fell from 56.5 percent in early seventies to 32.6 percent (Silber, 172 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1997:121). Indian exports, on the other hand, displayed the opposite trend with increasing unskilled exports relative to skill-intensive exports. With respect to capital flows, Brazil has allowed greater capital flows in the form of FDI and lending from foreign banks. Unlike India, attracting foreign capital was labeled a priority for the state (Schlaghek, 1977). Extensive fiscal and financial incentives were implemented in order to attract FDI. Thus, foreign interests permeated the Brazilian economy via long term capital flows. By the early seventies and into the eighties, Encamation (1989) claims that MNCs in Brazil controlled over half the Brazilian market for chemicals, machinery and transport equipment. Whereas, during that same time period, India permitted MNCs only in the rubber and pharmaceutical industries. Similar to India, Brazil also has a history of capital controls on financial flows. Generally, Brazil’s link to the world capital market has been through borrowing capital Chart 4.3: Comparison of Capital Flows in Brazil and India Gross Capital Rows In India and Brazil (as % of GOP) ♦ 0.2 ' / 0.15 0.1 ' 0.05 1 1 ,,v r' V 1 — o - i k i i 1 I 1 1 1 1 I 1 I I 1 1 I i i t i t i i BrazTs qross caoital flow India's qrosscaoital flow 1 Source: World Bank (1997) World Development Indicators. 173 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and foreign investments, rather than through risk capital (Silber 1997). Private capital flows were virtually nonexistent in the 80’s but increased dramatically after 1991. In 1991, the Brazilian government adopted several measures to stimulate capital inflows. For example, much of the taxes on foreign funding were eliminated. As a result, medium and long term capital flow shares have declined while the importance of short term capital has been increasing. However, in 1995 Brazil once again attempted to discourage short-term capital flow by reimposing taxes and reserve requirements on foreign capital. Interestingly, relative to foreign trade, net capital inflow increases have been quite large in recent years. Monthly private capital flows in Brazil between 1988 and 1991 increased 25 times, although overall capital flows to Brazil have been quite modest relative to other Latin American countries (Agosin & French-Davis, 1996). In fact, since 1992, net foreign capital flows have been sufficient to increase foreign reserves and finance small current account deficits (Cardosa & Golfajn 1998). Overall, given their different histories with exposure to international capital and export markets, Brazil’s economy has been much more globalized than India (see Charts 4.2 and 4.3 ). Yet social welfare spending in Brazil has been increasing, contrary to the average trend in most LDCs. If the tenets behind the original hypothesis hold, then welfare spending in Brazil during globalization should rise if skilled labor has been increasing relative to unskilled. The evidence should indicate that Brazil has been more successful than India at altering its comparative advantage in labor-intensive goods for this to be confirmed. This analysis does, in fact, suggest that 174 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. this hypothesis is valid in Brazil. However, it emerges that the rise of skilled labor is not the only factor that has contributed to the size and trend of Brazil’s welfare state. In order to explain the increasing social welfare expenditures in Brazil, two additional variables must be taken into consideration alongside the primary hypothesis. Both the turn to re-democratization, as well as the use of welfare benefits for clientelistic favors, appear to affect the level and trend in Brazil’s social spending. Needless to say, the Brazilian welfare case emerges as one of the most complex and one that is riddled with contradictions. Globalization and the Bargaining Power o f Labor in Brazil Although the structural change in Brazil’s social commitment did not occur until 1988, data indicates that welfare spending relative to GDP has been increasing since the mid 1970s.1 3 Strong labor power interacting with globalization and re-democratization seems to best explain the gradual increase in social welfare spending since the early 1970s. The fact that the level of Brazil’s social expenditures is higher than the average LDC, however, can be attributed to the historical evolution of the state-labor relationship. This explains why benefits were high both before the constitution was rewritten in 1988 and during the process of globalization. The story behind increasing social benefits in Brazil is as follows. First, the bargaining power of labor as a whole increased incrementally because of rising 1 3 This excludes the temporary decline of welfare expenditures after the debt crisis in the early 1980s. 175 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. proportion, of skill-intensive employment Second, re-democratization in 1988 led to structural changes in the government’s commitment towards social and labor policies. Finally, social welfare payments represented an instrument of state corporatism used to maintain labor cooperation. This resulted in a clientelism that persisted even after democracy. Of course, these processes are interdependent and have influenced one another. Why has the breadth of Brazil’s social coverage has been increasing over time, unlike most LDCs? Huber (1999) claims that after five decades, Brazil’s social welfare system has matured and the benefits cover more than 60 percent of the economically active population. Patronage politics does not explain why the quality of social welfare in Brazil has been improving, however slightly. This analysis suggests that it is the bargaining power of labor which resulted from the expansion of skilled labor groups, in addition to re-democratization that best explains the structural change of the late 1980s. Several Latin American scholars suggest that labor located in the strategic high-skill sectors have gradually obtained disproportionate gains in social welfare ( Bronstein, 1997; Huber, 1996, 1999; Malloy, 1979; Mesa-Lago, 1978). There emerged a labor aristocracy consisting of a selective subgroup of skilled blue collar workers which had access to special privileges (i.e., social benefits) (Malloy, 1979; Mesa-Lago, 1978). It is in the most sophisticated manufacturing branches that the leverage of the working class increased the most. Drake (1996) claims that enhanced skill levels in these sectors tightened the labor market, led to higher wages and made collective 176 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. bargaining more plausible. In particular, the bargaining power of labor increased the most for skilled workers in the strategic FDI sector (Drake 1996). Sandovol (1993) presents convincing evidence of the discrimination of the Brazilian government towards labor in unskilled sectors verses the higher-skilled sectors.1 4 Using strike data, he argues that not only did the skilled sectors hold more strikes and protests, but also most of the massive interventions by the military government in unions were avoided in the skilled sectors. Sandovol (1993) concludes that: [S]ome modem sector workers, in particular the metalworkers, were better able to recover lost ground and to form new leadership, while most of the traditional sector unions were severely handicapped as a consequence of the imposition of conservative, pro-govemraent union bosses and closer surveillance of worker discontent (Sandovol, 1993: 41). Recall that the primary hypothesis claimed that if a country had a comparative advantage in unskilled labor in the era of globalization, then it was likely that welfare spending would decrease. Brazil, however, began to export high skilled goods from the early stages of globalization, as shown in Chart 4.4. Indeed, data indicates that increases in skill-intensive exports have managed to keep pace with low-skill exports in Brazil. Unlike India, there is almost an equal proportion of skilled and unskilled exports since the early seventies! Significantly, the trend of high-skill exports begins to peak in the mid 1980s, mirroring the structural change in Brazil’s social benefits also 1 4 Note that Sandovol’s (1993) categorization of unskilled and skilled sectors is traditional and modem, respectively. 177 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. in the mid 1980s. Hereby, this data suggest that Brazil has a comparative advantage in both skilled and unskilled manufactured goods. Chart 4.4: Breakdown of Brazil’s Manufactured Exports Brazil's Manufactured Exports (as % of total exports) 30 2 5 20 - 1 5 1 0 5 0 9 1976 1980 1983 1987 1993 1995 ■ Low Skill-basic m anufactures □ High Skill -m achines and transport equipment Source: United Nations (various editions) Yearbook o f International Trade Statistics Furthermore, Charts 4.4 and 4.5 show that the increase in the number of skilled workers is consistent with the rise in high-skill exports. Employment in both skilled and unskilled sectors has been increasing (see Chart 4.5).IS It would seem, however, that the relative demand for unskilled labor is less than India, given Brazil’s growing production of skilled exports. Moreover, Table 4.2 suggests that Brazil’s informal sector has been decreasing, indicating that the trend of unskilled labor in Chart 4.5 is somewhat reliable. India’s data on the trend of unskilled labor did not hold up to the same scrutiny. This is not to say the informal sector is not problematic in Brazil, 1 S Note that some sectoral data from the last four years are missing. The trend in employment is not declining to the extent that it appears in Chart 4.5 178 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. but theory and evidence suggests that their informal labor situation is less problematic than India’s. However, neither the history of corporatism nor growth of unskilled labor can explain the late 1980s structural shift in Brazil’s social spending. Although Chart 4.5 shows that the number of skilled employees has been rising, the increase seems incremental. Thus, the expansion of the skilled sector suggests why there has been a gradual change in social welfare expenditures over time. The late 1980s structural change in Brazil’s social commitment, however, is best explained by re democratization. 179 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chart 4.5: Employment of Unskilled and Skilled Labor in Brazil* 3000000 2500000 2000000 1500000 1000000 500000 0 ■ Numbers employed in unskilled labor D Numbers employed in skilled labor E m ploym ent in Brazil Source: UNIDO database. ♦Note: missing data for years 1986, 1987, 1988,1989, 1991. For years 1990, 1992, and 1993 some sectoral data for both unskilled and skilled is missing (four sectors in unskilled and three sectors in skilled) and thus these years have been underestimated. Table 4.2: Informal Sector Employment in Brazil Year Number of people employed in the urban informal sector (‘000) Urban informal sector employment as a percentage of urban employment 1990 • * 44.3 1994 • * 47.3 ♦Informal sector employment is defined as: all own-account workers (excluding administrative workers, professionals and technicians) and unpaid family workers, and employers and employees working in establishments of less than 5 or 10 persons engaged, depending on the available information. Paid domestic workers are excluded. 180 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Many argue that governments with a history of labor repression result in increased incentives for different labor groups to mobilize in the fight for democracy and greater labor rights (Drake, 1996; Sandovol, 1993). Despite heavy government constraints, Brazilian labor had pursued an important political role that defied state incorporation. Collier and Collier (1991:511) point out that the labor movement in Brazil had managed to “carve out an area of much greater political independence from state control than one might expect.” It was not the existence of a multiplicity of unions that was problematic, unlike the case of India. 1 6 Worker organizations sought to find alternative support organizations (i.e., neighborhood associations, church groups, and other grassroots organizations) to parallel the union structures dominated by the state (Sandovol, 1993). The Catholic Church, in particular, was a key source for the mobilization of support for working class political needs. Ultimately, strong control over unions and the denial of labor rights promoted a united cause for labor. The Brazilian labor movement “stitched together a multiclass front” in the early to mid 1980s, according to Drake (1996). They reached outside of the unions to other members of the working class (and the community) in order to ally themselves against military dictatorship. During the struggle for democracy, trade unions progressed in five major areas: base organization, breadth of coverage, interunion linkages, collective bargaining, and political mobilization (Drake 1996). 1 6 This is not to deny the extent of party polarization that evolved in Brazil after state incorporation. Please see Collier and Collier (1991). 181 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. As a result, unions have gained freedom of government control since democratic rule in 1988 (Amadeo et al., 1995). The new 1988 Constitution included numerous guarantees regarding employee rights and significantly expanded employee benefits such as the reduced size of work week, extended maternity leave, increased vacation pay, compensation for arbitrary dismissals, etc. (Palhano, 1995). In addition, strikes were declared to be a social right for all workers in 1989, considerably reducing previous restrictions imposed upon labor freedom. Thus, the declining trends in Chart 4.6 must be interpreted differently from similar trends that emerged in the Indian case. Recall that these trends in India were also declining after the late eighties. However, a review of India’s political situation revealed labor's bargaining power was actually declining. It was surmised that India’s declining were more of a reflection of labor’s growing collective action problems. In Brazil, it seems that an interpretation of declining trends is quite the opposite. Brazilian labor, on the whole, appears to be more satisfied relative to Indian labor. Take, for instance, the increase in real wages in Brazil (see Table 4.3). Furthermore, given the degree of government concession to labor demands after re-democratization, it can be concluded that the declining trends in labor disputes indicate that bargaining power of Brazilian labor has been increasing. 182 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chart 4.6: Trends in Labor Disputes in Brazil 140000 120000 100000 80000 60000 40000 20000 0 r 20000 " 15000 - 10000 3 a jg ■o w S a ' O e a ■5000 £ 1986 1987 1988 1989 1990 1991 1992 work days not worked (thousands) industrial disputes ~*~w orkere involved (thousands) e I I Source: International Labor Organization (Various editions) Yearbook o f Labor Statistics Table 4.3: Brazil: Real Manufacturing Wage Indices (base year=l990) Year Real Manufacturing Wage Indices 1980 88.6 1990 100 1993 94.6 1994 98 1995 102 Source: International Labour Organization (1999) Key Indicators o f the Labor Market. Geneva: International Labor Office. 183 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Both the political situation of the times and the trends in strike activities since 1989 (Chart 4.6) suggest that Brazilian labor is closer than India to resolving its collective action problem. Nevertheless, while new political spaces have opened for new forms of political and social organizations, traditional forms of elite domination have been maintained and reinforced. This “schizophrenic nature of Brazilian politics,” as Keck (1999) appropriately labels this dynamic, characterizes the contradictions in Brazil’s social welfare system. An examination of the history of state corporatism in Brazil reveals why the level of benefits are high, yet its labor protections are low. Indeed, corporatism established a system that exchanged labor benefits and protection for the right to strike and form autonomous unions. The Estado Novo (New State) dictatorship of Getulio Vargas established the Consolidation of Labor Laws (CLT) which served, in theory, to provide labor freedoms. For a developing country at that time, CLT represented very advanced labor legislation. It included hours of work, the setting of wages and salaries, discipline, hiring and firing, pensions, employment of women and minors, industrial health and safety, etc. These labor laws were consolidated in the forties, and were not altered until the 1988 constitution (Sandoval, 1993). However, French (1998) persuasively argues that there was a large chasm between law and reality in Brazil. He claims that “if the working world did in fact operate according to the CLT, Brazil would be the best place in the world to work” (French, 1998:186). In practice, the CLT system was extremely problematic, with 184 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. systematic and intrinsic difficulties. It is no surprise, then, that Amadeo et al (1995) conclude that Brazil’s labor market was flexible, despite rigid labor regulations. This makes sense, given that Brazil was facing international competition early on and needed labor market flexibility in order to be competitive. The ad-hoc distribution of benefits for clientelistic favors led to an inequitable system of social welfare benefits. Powerful social groups (i.e., the military, professionals and white collar workers) obtained these advantages at the expense of more needy and the less politically powerful (Barreto de Oliveira, 1994). Ultimately, Mesa-Lago (1978) claims social welfare has often been “manipulated in order to gain the electoral support of a particular clientele, to legitimate a spurious political regime, and to satisfy the needs and co-opt powerful pressure groups that threaten the status quo” (Mesa Lago, 1978:3). Weyland (1995, 1996) further finds that one of the main obstacles to welfare reform even after democracy has been the pervasive clientelism. The use of social welfare to fuel patronage politics has resulted in institutional constraints that makes it virtually impossible to implement more equitable social reform. Weyland (1995) argues that politicians are now the biggest obstacle to reform because universal benefits pose a deep threat to the electoral sustenance of many powerful politicians. In sum, state corporatist arrangements preceding democracy explain why social spending in Brazil has been relatively high. Brazil’s case suggests that newly emerging democracies that have a history of corporatism produce a social welfare system that is extensive, yet extremely inegalitarian. The increasing trends in welfare spending are 185 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. best accounted for by increasing bargaining power of labor, due to globalization and the size of employment in high-skill industries. In addition, structural changes in the government’s social commitment occurred after the fight for re-democratization. As a result, labor demands for greater freedoms and more benefits under democracy are being met. The Brazilian case revealed that rising welfare expenditures as a percentage of GDP in LDCs must be viewed with caution, however. The costly welfare benefits that have been extended by Brazils social welfare programs since 1919 have provided fertile ground for patronage politics to thrive. When compared to the case of India, it can be conjectured that since India’s benefits remained low (and began decreasing), a privileged labor group never managed to dominate India’s social system. Brazil’s attempts to reform these expensive privileges have since met with strong resistance (Barreto, 1994; Weyland, 1995, 1996). Thus, it must be concluded that increasing welfare expenditures does not necessarily indicate that the compensation effect is occurring in LDCs. 186 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 5 KOREA AND CASE STUDY SUMMARY Like Brazil, Korea is another deviant case with respect to its increasing trend of social welfare expenditures.1 Yet interestingly, it appears that the Korean case also confirms rather than disconfirms the proposed hypothesis. Korea, more than Brazil, has managed to alter its comparative advantage and increase the relative level of skilled labor in the population. Increasing bargaining power of labor has mitigated the adverse effects of globalization on welfare spending. The growth of labor power has, in turn, culminated in a stronger push for welfare statism during the turn from authoritarianism to democracy. However, the rising trends in Korea’s welfare spending should not obscure the fact that Korea has been a ‘welfare laggard’ compared to similar high income (both OECD and non-OECD) countries. Understanding Korea’s state-labor relationship sheds light upon the reasons behind their lower than average levels of welfare spending relative to GDP. By “deviant” I mean that the trend of social welfare spending deviates from the average trend displayed by the S3 developing countries in my sample. Recall that, in the average, social welfare spending was characterized by a downward trend in LDCs. 187 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The Dependent Variable: Government Social Welfare Commitment in Korea Overall, Korea’s commitment to welfare has been extremely weak. Although like Brazil, Korea’s trend in social welfare expenditures does not appear to bow in the face of globalization, its overall level of spending is very low. Moreover, its state labor relationship is best characterized as flexible, with priority afforded to capital over labor. Korea’s country specific differences were two: (1) the communal attitude of large companies towards their employees and (2) the existence of a large middle class and weak working class. Thereby, the Korean state, similar to the Indian state, played a stronger legislative, rather than financial role, in allocating social welfare. Korea’s laws, however, are not designed in favor of labor. Instead, Korea tends to encourage firms to take financial responsibility for social welfare. Trends and Levels o f Korea’ s Social Welfare Spending Korea’s pattern of welfare spending has been steadily increasing at low levels. The trend illustrates three phases. Spending seems to hold constant after increasing slightly between the 1970s and 1980s. Ultimately, the strongest structural shift appears to have occured in the late 80*s (see Chart S.l). Chart illustrates that there is little variation between welfare spending as share of GDP and as a share of total government expenditures. Just like in Brazil, the major structural shift coincides with democracy. 188 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chart S.l Korea’s Social Welfare Expenditures Korea's Social Welfare Expenditures > o 0 £ * o « 0 2 ? 12 10 8 6 4 2 0 m M M ■ n m m m m 2 1-5 Q . Q 1 ° ' S 0.5 0 j y < # ST ► J S 3 as % of Total Gov Expenditures — as % of GDP * Note that the IMF data on Korea recorded a six fold jump in spending for the 1982 year. The author speculates that this was a result of a data calculation and/or recording error. Source: IMF, Government Finance Statistics: various editions Korea is similar to the average LDC in that it maintains a very low level of benefits. However, it is different from India because they have a more flexible labor market. On average, Korea spends about one percent of GDP on welfare benefits. It was not until around 1987 that welfare spending rose above one percent. State 189 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. contributions relative to employers and employees have been miniscule. Prior to 1988, the Korean state lent minimal financial support to social welfare programs. This role shifted markedly after 1988 when the state made a visible commitment to greater participation in social welfare (see Table 5.1 ). Nonetheless, it is particularly striking that Korea’s level of social expenditures has been meager in proportion to its level of economic growth. Table 5.1: Contributions to Social Security Receipts in Korea (in millions of national currency units) Year State contribution Employer Contribution Employee Contribution 1990 835126 1057030 1660504 (21%) (26%) (41%) 1991 1551731 247751 2530954 (20%) (32%) (33%) 1993 930536 3613919 3289694 (9.7%) (3.7%) (3-4%) • Numbers in parenthesis represent contributions as a percentage of total receipts. Note that remaining contributions come horn special social security taxes, participation of other public authorities, income from capital and other receipts. However, contributions from these sources in LDCs tend to be very small, if any. • Years chosen based on data availability. * Gross figures unavailable. Source; The Cost o f Social Security. International Labor Organization: Various years 190 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The Character o f Korea*s Social Welfare System At first glance, the character of Korea’s welfare system seems similar to India’s. Overall, traditional social security benefits allocated to workers are rather limited. They involve mostly protection against work injury and medical insurance that is offered for most workers. The first law was implemented in 1953 for work injury and followed by provision of medical insurance in 1963 (Social Security Throughout the World 1998). Legislation on old age, disability and death did not arise until 1973, but amount of cash transfers were amended in the late 1980s and periodically through the 1990s. Finally the Industrial Accident Insurance introduced in 1962 currently covers only about half the employed (Social Security Throughout the World 1998). Like India, Korea does not provide funding for unemployment or family allowances. Although Korea provides limited social coverage overall, it does have more targeted benefits. The Public Assistance Program provides cash allowance and educational and medical support for families officially defined as poor. However, the problem is that a very small proportion of the population qualifies as “poor,” since the poverty line defined by the government is so low. Kwon (1997) claimed, for instance, that only 4.54 percent of the working population was covered in 1993 (Kwon 1997). Similar to the social insurance type programs, the state serves as the regulator of social programs and involves limited government financial resources. Coverage for social welfare did not expand until 1988, which coincides with Korea’s democratization. It was not until 1988 that a national social insurance system was implemented. Prior to this, coverage was allocated on an enterprise wide basis. In 191 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1987, Chun Doo-Hwan brought social welfare back to the forefront. Competitive presidential elections in 1987 forced governing party candidates to pay for those not covered by existing social security schemes (Kwon, 1998a). It was after 1988 that the government began to expand its role in social welfare financing. Specifically, the National Pension Program (NPP) and National Health Insurance (NHI) were implemented until 1992. Even then, only 26.7% of the working population was covered (Kwon, 1998b). Acceptance of these programs is particularly significant, given that attempts to launch programs such as the NPP in 1973 under the authoritarian Chung-Hee Park regime were unsuccessful.2 The government now participates financially in the NPP. It contributes as much as 50 percent to those who do not have employers such as farmers, self employed and workers in the urban informal sectors. In addition, the government also began to contribute subsidies to the Employment insurance scheme. They increased their level of contributions to job training programs and local job centers (Kwon, 1998b). The Y.S. Kim government in 1995 implemented only two programs, the old-age pension for peasants and fisherman and unemployment insurance (Song, 1999). Korea is very different from India and most LDCs in that the state has acted to achieve welfare goals by other means. Their state-led market development has resulted in a more equitable income distribution and a large middle class. As a result, the state focused more on providing educational and medical support, rather than income support 2 According to Kwon (1998b), the Park regime claimed to repeal NPP because of high inflation stemming from the oil crisis. 192 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (Kwon, 1997). Redistribution type welfare policies were effectively removed from the immediate agenda. Instead, the government imposed mandates upon employers to provide employment security and benefits. Workers were, thereby, encouraged to foster a greater dependence upon private sector benefits instead of the state. Thus, Korea’s social system has been the model for Neoliberals and Social Democrats alike (Gordon & Goodman, 1998). Neoliberals approved of the low cost of welfare system and its simultaneous policies of economic growth. According to them, this leads to less drain on state resources, while prioritizing economic efficiency. Social democrats, on the other hand, do not focus as much on the social system per se, but endorsed the overall positive effects of economic policies on income distribution. What has evolved in Korea, then, is a sense of mutual security between employers and employees. Prior to the late 80’s, it was the firm, community and family, which were expected to play the major role in welfare. Employers covered the bulk of the costs for old age, disability and death and work injury. As explained by White and Goodman (1998): [N]on-state agencies -community, firm, family—have been expected to play a major welfare role in both financing and providing welfare services in an ideolgoical context wherein self/mutual help is encouraged and dependence on the state discouraged, indeed stimatised (White & Goodman, 1998:14). Indeed, the income distribution argument, as well as the emphasis on corporate welfare responsibility, explains government motivations behind the low levels of welfare 193 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. expenditures. It does not, however, answer the question of why expenditures in Korea have been increasing steadily. Certainly, a survey of the trends reveals that Korea’s turn to democracy coincided with the later structural shift in social welfare spending. Thus, the causal relationship between democracy and changes in welfare spending cannot be easily dismissed. Yet data indicates that social spending in Korea has been increasing since the early seventies before democracy, suggesting that democracy alone cannot explain the trends of social welfare spending. Thus, the politics of the state- labor relationship must be taken into account. It appears that the status quo was ultimately not acceptable to Korea’s working class. In fact, Lindauer (1997) argues that labor protests were highest when labors economic circumstances were relatively good. Korea*s L abor M arket Policies Korea’s labor market policies were highly repressive. Deyo (1989), a strong advocate of this thesis, argues that labor repression was one of the most significant factors behind the East Asian miracle. Under the postwar American occupation, President Sygman Rhee adopted modem labor legislation in 1953. However, these efforts were the result attempts to create a Western style, politically moderate trade union movement. In practice, these rights did not enjoy legal recognition and were not implemented (Deyo 1989). The Yushin Constitution formed by Park Chung Hee laid the foundations for state-labor relations. This formal document banned future elections and mandated the end of political and labor activities. 194 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Following the Yushin Constitution, the Korean government relayed heavy- handed tactics with labor. Labor unions were severely repressed, as the government aimed to control unions and minimize any base for political opposition (Vogel & Lindauer, 1997). Thus, the bargaining power of labor was muted and thereby, in large part, released firms from having to deal with problems such as collective action and other labor-management issues (Lee, 1998). The nation’s labor market experience involved a very long work week in manufacturing, high industrial accidents and very limited labor rights. A new era for the Korean labor union movement emerged with political democracy. From 1987 to 1989, the Korean government refrained from intervention and a massive unionization drive occurred (Lee, 1988). President Kim Dae Jung appealed to the industrial working class in the presidential campaign. He promised the right of political participation at the same time (Song, 1998). The democratization of enterprise unions and the demise of company dominated union leadership occurred in many workplaces (Lee, 1998). Collective bargaining eventually emerged as the main channel for determining employment conditions in the unionized sector. In sum, the case of Korean social welfare is complicated. This country’s historical circumstances have engendered a state that has low welfare commitment in the traditional sense of welfare. Its policies have instead effected a more equitable income distribution and greater corporate responsibility that has required a low level of state spending. At the same time, Korea’s labor policies have been very repressive. Korea then scores low on all categories of social commitment applied in this analysis. 195 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. What stands out, however, is the fact that its social spending has been steadily increasing, particularly since its turn to democracy. This will be the focus of the analysis. The fact that Brazil and Korea increased the amount of expenditures consistently since the late 80s, while social spending in India began to decrease needs to be further investigated. Indeed, this section suggested that the levels of social spending are primarily a function of domestic politics and historical forces (i.e., state-labor relations, ideology, etc). The next section(s) asks whether or not international economic factors, alongside other relevant domestic variables, play an important role in determining the structural changes in social spending in India, Brazil and Korea. The Independent Variables: Globalization and the Bargaining Power of Labor It would be premature to approve Korea’s social welfare record solely on the basis of its increasing trends in social welfare spending. The fact that Korea is a welfare laggard and spends a comparatively small percentage of their GDP on social programs cannot be ignored. The growth of unskilled labor and the increasing momentum of labor organizations explain increasing welfare trends. Also, similar to Brazil, democracy is a key factor in explaining Korea’s welfare spending trends. These two cases suggest that new democracies spend more on social programs, since India, a longstanding democracy, has not yet managed to improve its social welfare spending. Nevertheless, although welfare spending has been increasing, Korea lacks a historical commitment to welfare and labor. In contrast to India, where social policies 196 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. have been strongly influenced by socialism and its ideological commitment to social development alongside growth, Korea pursued a “growth first” strategy with responsibility for welfare provision allocated to firms. Therefore, despite the fact that labor momentum in Korea has been increasing, they have not been able to make substantial changes to welfare spending because of historical constraints.3 Globalization Contrary to India, Korea pursued an economic strategy that was far removed from isolationism. The process of globalization began in the early twentieth century for Korea. President Park Chung Hee implemented specific ISI reform measures in the 1960s, of which the promotion of manufactured exports was one of the major changes. As a result, Korea began engaging in export-oriented industrialization. To bolster their export competitiveness, the government allotted special privileges to exporters, such as providing preferential access to import licenses and readily available credit. Significantly, the Korean experience with export production resulted in an efficient use of Korea’s resources, quite contrary to India’s experience (Kuznet, 1984). Thereby, Korea’s efforts at export promotion were more successful than those of India (see Lall, 1999). Many argue that Korea’s ability to exploit its comparative advantage in labor during the sixties was a major contributing factor to its economic success (Kim, 1997; Krueger, 1995; Kuznet, 1984; Perkins, 1997). 3 Social welfare spending never rose much above one percent o f GDP, even after democratization. 197 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. While neo-liberals tend to claim victory for Korea’s success on the basis of its outward-looking strategies, Korea was far from laissez-faire in its approach to trade (Perkins, 1997). One way Korea was similar to India was its application of ISI to limit imports of consumer goods. The difference between the Korean and Indian ISI strategies is that in Korea, protections afforded to import substituters were designed to be temporary (Perkins, 1997). Eventually, most manufacturers were expected to become exporters. It can be argued, then, that Korea was allowing its economy to be gradually exposed to global market forces. India, on the other hand, was insulating its economy from international competition and, thus, the process of globalization. Opening up to international markets came rather rapidly in India, with the onset of trade and financial liberalization in the mid eighties. It was clear that trade by itself was a “leading” engine of growth in Korea (Krueger, 1997). Exports and imports as a percentage of GDP grew from 45 percent in the early seventies to more that 70 percent of GDP by the mid-eighties (see Chart 5.2). This is a marked contrast to the Indian economy, where trade accounted for only 8 percent of GDP in the early seventies, and never reached beyond the 20 percent mark even after liberalization. 198 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chart 5.2: Comparison of Trade Flows in Korea and India Trade Flows in India and Korea (as % of GDP) 100 80 60 40 20 “I —i —. i i t > i—i i i r r*i' i i i ‘i - r 'i ’ Korea's trade levels India’ s trade levels Source: World Bank (1997) World Development Indicators. Despite this, why isn’t there a visible structural change in the trade data after Korea made the decision to liberalize its economy? Korea’s import liberalization efforts date back to 1967,4 but it was not until the early 1980s that a five-year import liberalization schedule was introduced to gradually reduce tariff rates and eliminate non-tariff barriers (Krueger, 1997: 356). This presents two possibilities. First, the level of integration into the global economy prior to liberalization was already quite high and thus, the structural change had already occurred during their switch to EOI in early ‘60’s. The second possible reason for the declining trade levels (relative to GDP) 4 In 1967, there was a changeover from a positive list system o f import controls to a negative list system. This reduced the number of restricted items from 26,000 to under 13,000 (Perkins, 1997). 199 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. is that there was a strong move towards offshore production, after Korean authorities abolished overseas investments restrictions in early 1989 (Chu, 1995). In order to avoid rising wages in the mid-eighties, a large number of Korean foreign investors fled to lower wage counties for establishing production and exports. Despite the fact that Korea has not recently witnessed a structural change in its export and import markets, it appears that the Korean economy has been exposed to the pressures of globalization for over three decades. In contrast, Korea has been much more gradual in permitting the free flow of capital across borders. Changes in the level of gross capital flows occurred soon after 1987 when Korea liberalized its capital markets. Like India, Korea has been much more restrictive towards capital flows and has promoted minimal dependence on foreign capital flows to finance development Both the Korean and Indian governments dominated the financial sector, regulated private investments in financial services, and nationalized commercial banks (Encamation, 1989). FDI, in particular, played a minor role in both their economies in the 1960s and 1970s. Between the period 1961-1986, FDI levels in Korea was only 7 percent of total capital flows (Stallings, 1990). How, then, did Korea obtain its foreign exchange earnings? More than half of the total foreign capital flows in Korea were in the form of public and private bank loans. From the mid 1960s to the mid 1970s, Korea received foreign capital mainly in the form of loans at concessionary rates of interest Notably, the Korean government directly managed the distribution of both private and commercial loans and, thereby, controlling their effects on the domestic economy. Kim (1997) argues that the Korean 200 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. state was active in managing the distribution process of foreign capital, even though commercial loans usually involve less state control. This is significant because it, once again, suggests that the Korean government was cautious in managing the gradual process of globalization and its effects on the domestic economy. Unlike India, it does not seem that the Korean government’s primary aim was to protect the economy from all possible foreign influences, but to be in command of their impact on the Korean economy. After 1980, the process of financial liberalization began. The Korean government sold its shareholding in the nation’s five largest commercial banks, as well as encouraged private expansion into other financial services (Encamation, 1989). Between 1982 and 1986, Korea aimed to liberalize the nation’s investment markets in order to raise the level of foreign investment. Foreign investors flocked to Korea when it removed its protections from domestic markets (Song, 1999). By the latter half of the eighties, private investors had almost unrestricted access to South Korean stock and securities market. Note, however, that foreigners were not allowed to own more than 10% of the shares in any Korean company (Perkins, 1997). Thus, although the pace of Korea’s incorporation in the global economy quickened by the first half of the 90’s, officials were still cautious in allowing the unbridled flow of foreign capital.s Nevertheless, Chart S.3 illustrates that in the late eighties, there was a distinct shift upwards in capital flows. The high level of flows in the early 1970s and early 5 Kim and Kim (1997) point out that the fear o f massive capital inflows and an anticipated won appreciation prompted authorities to tighten control on capital inflows once again in the late 1980’s. After 1989, the government once again began deregulating capital flows. 201 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1980s can be attributed to bank loans permitted by the country. In fact, Korea’s long term debt to banks increased from $200 million in 1970 to $9.7 billion in 1982 (Stallings 1990). Chart 5.3: Comparison of Capital Flows in India and Korea G ro ss Capital Flow s In India and Korea (as % of GDP) 14 T ^ ^ & J* ♦ India’ s gross capital flow —a — Korea's g ro ss capital flow Source: IMF (various editions) Balance o f Payments Statistics Korea has been paying back its debt since the late 80’s. Based on this chart, it does not at first appear that globalization is bringing in greater risks and uncertainties to Korea. The level of capital flows reached in the early nineties are not much higher than the I970’s. However, the composition of private capital flows have been increasing, thus raising the chances of instability. Given their recent policy changes, the upward trend after liberalization has been based on increases in both FDI and short-term capital inflows. For example, in 1977 commercial loans accounted for 93 percent of total gross capital flows and less than S percent was made up by FDI and portfolio flows (Balance o f Payments, various editions). By 1995, commercial loan had fallen to approximately one-half total flows, while FDI and 202 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. portfolio flows accounted for over 40 percent. Thus, FDI and portfolio flows are slowly replacing Korea’s earlier reliance on (commercial and public) bank loans. Although the level of gross capital flows in Korea have fluctuated over time, notice that the total level of capital flows has always been high relative to India. The question now emerges how social groups such as labor were able to successfully mediate between the slow, but constant, process of globalization and welfare? This question is particularly important from the early 1970s until the late eighties, when globalization begins to pick up momentum in capital markets. Globalization and the Bargaining Power of Labor in Korea This analysis suggests that as Korea began to change its comparative advantage to high skilled labor, labor was gradually able to pressure the government for greater labor rights and benefits. Striving to keep its products competitive in the international marketplace, the Korean state did not promote the same level of labor protectionism granted by the Indian state. Labor was initially key to their industrial success, beginning with primary export oriented industrialization in the 1960s. Labor began to organize and exert their demands for a welfare state only after Korea managed to shift its comparative advantage to more skill intensive industries in the 1970s. This analysis suggests that increasing labor discontent resulting from years of labor repression, and growing labor power based on skilled laborers, caused the labor movement to explode in the 1980s with the onset of democracy. 203 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The Park government by shifting from ISI to EOI in the mid 60’s, unleashed Korea’s comparative advantage in labor-intensive industries. The composition of exports shifted significantly in the early years of EOI. Manufactured exports overtook agricultural and mineral exports as they rose from 25 percent to 72 percent of total exports (Krueger, 1997). Some of the key manufacturing products which contributed to this increase were textiles, lumber and plywood, and steel and metal products. During this same period, the Park regime provided more room for state intervention and placed severe restrictions on union membership and action. Labor affairs were made an important target of government control by the Emergency Decree on National Security in 1971 and the Yushin system established in 1972 (Kim, 1997). These measures resulted in severe punishments for any labor activity, and banned o rg an izin g and collective action. Along with these repressive labor tactics, the state also kept wages low in order to promote international competitiveness of their labor intensive exports (Deyo, 1989). Deyo (1989) convincingly argues that much of Korea’s miracle was built on the backs of labor which worked longer hours, had no union rights, and overall little bargaining power. For example, Korean wages were less than one- third of Japanese wages in the same labor-intensive industries, while worker productivity was more than a third higher (Kuznet, 1984). By the 1970s, President Park inaugurated the heavy and chemical industry (HCI) drive. HCI signified Korea’s push towards the next stage of development by way of a shift in their comparative advantage. By the 1980s, export share of skill-intensive products such as machinery, transport equipment increased significantly, while labor- 204 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. intensive exports began to show declines. Indeed, by 1990, electronics had overtaken textiles and garments, whose share of total exports had fallen from 41 percent in 1970 to 22.6 percent in 1990 (Krueger, 1997:325) (See Chart 5.4). Significantly, by the early eighties, high skilled exports began to surpass low skill exports. Thus, as Jaimin Lee (1995) argued, Korea had successfully and K orea's M anufactured E xports (as % of total exports) 19 7 6 19 7 9 1983 198 7 1 9 9 3 1995 F low Skill- b a sic m a n u fa ctu re s D W g h Skill -m a c h in e s an d tran sp o rt equip Chart 5.4: Breakdown of Korea’s Manufactured Exports Source: UN (various editions) Yearbook o f International Trade Statistics strategically managed to shift its comparative advantage to high-skill exports. By the 1980s, two factors seriously affected the bargaining power of labor. First, Korea’s labor surplus began to disappear and the demand for labor began to outpace supply. Shortages of skilled workers were beginning to affect all industries. 205 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The HCI drive had increased the demand for skilled workers relative to unskilled and had, consequently, driven up wages for skilled labor (Krueger 1997). At the same time, the Korean labor market began to tighten after the mid-eighties (see Kim & Lee, 1997). Unlike most developing countries, Korean had absorbed its surplus labor and evolved from a situation of labor abundance to one of labor scarcity (see Chart S.S). Chart S.5: Numbers Employed in Skilled and Unskilled Industries in Korea 1600000 1400000 1200000 1000000 800000 600000 400000 200000 0 Inum bers employed in low- skill secto rs D num bers em ployed in high- skill sectors Source: UNIDO data base (provided by James Galbraith). Concurrent with the HCI drive in the 1970s and 1980s was the rapid growth of the labor movement alongside the skill-development of the labor force. Greater government investments in education during that time led to rapid increases in the skill levels of the Korean work force. HCI benefited from this governmental strategy since it required a more highly educated labor force (Kim, 1997). At the same time, the better 206 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1995199159 educated labor force provided fertile conditions for labor movements to organize and grow. It is interesting that the relative number of high skilled workers begin to increase in the mid eighties, about the same time as the push for welfare as a state priority. The labor movement essentially erupted in the 1980s. According to Kim (1997), labor was demanding a welfare state, which guaranteed the rights of labor and provided basic welfare services. Labor fought against the state suppression of wages, and for freedom of association, collective bargaining and collective action. The number of disputes began to increase drastically. Significantly, there was a documented shift in labor attitudes in 1980s, as suggested by Lindauer (1997:2). Labor unrest and dissatisfaction existed in the earlier decades, but starting mid-1987 the degree of militant behavior and the depth of worker anger seemed historically unprecedented. What is surprising about these developments is that no other nation has ever recorded as rapid and as sustained an increase in average wages and industrial employment opportunities as has the post-1965 economy. Korean workers shared in the nation’s growing prosperity, but dissatisfaction dominated their attitudes and fueled their militant actions. Chart 5.6 documents the explosion in labor protests in the mid 1980s. Workers involved in strikes and industrial disputes increased more than six fold. Afterwards, the number of industrial disputes, work day lost, and workers involved decreased dramatically. Similar to Brazil, however, the political situation suggests the change in strike pattern after 1989 was due to structural change in government social commitments and, subsequently, labor satisfaction. Recall that, to labors advantage, 207 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. real wages in Korea were rising (see Table 5.2 ). The labor movement in Korea had, by then, begun to mature. Notice, for example, the dramatic growth in union membership after 1987 (see Chart 5.7). Table 5.2 Real Wages in Korea (base year=l990) Year Real Manufacturing Wage Indices 1980 45.6 1990 100 1993 123 1995 140.8 Source: International Labour Organization (1999) Indicators o f the Labor Market. Geneva: International Labour Office. Choi and Kim (1997: 550) claim that after 1987, the structure of Korea’s social welfare system went through “three ground-breaking changes: enforcement of a nation wide health insurance system; the implementation of the national pension plan; and the enactment of the Minimum Wage Law.” They argue that implementation of these three systems was a major step forward for Korea’s social welfare system. In 1993, the regime’s welfare policy advanced even further. Legislation of an Employment Insurance Law, for example, served to guarantee workers an income security when unemployed. More significantly, however, it aimed to provide better employment 208 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. conditions for workers by training the unskilled, and by unemployment prevention and employment promotion programs (Choi & Kwon, 1997). The evidence, therefore, suggests that there was a change in the extent of government social commitments after democracy. Although far from establishing a universal welfare state, Korea began to modify its repressive stance towards labor rights and benefit distribution. The Fifth Regime (1982-1986) placed the welfare state as a major policy objective (Choi & Kwon, 1997). Thus, it seems that after 1985, Korea began to reorder its priorities towards social and economic development. 6000 5000 4000 3000 2000 1000 300 250 200 150 100 1978 1980 1985 1988 1990 1993 industrial disputes work days not worked (thousands) workers involved (thousands) Chart S.6: Trends in Labor Disputes in Korea Source: International Labor Organization (Various editions) Yearbook o f Labor Statistics 209 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chart 5.7: Unionization Trends in Korea 2500000 2000000 1500000 1000000 500000 0 & Union Membership in Korea < & > & & PUnion membership & Source: International Labor Organization In sum, the primary explanatory variables for the growth in social welfare expenditures are growing labor power and democratization. Both of these variables account for the structural change in the government’s social commitment in the late 1980s. Since labor has been historically repressed, the level of labor benefits in Korea has been notoriously low and changed only by plus or minus .19 percent from 1972 to 1985. The political climate of democracy prompted the state to begin prioritizing the welfare state and cater to labor’s growing demands. After democracy, the government refrained from repressing labor and interfering with labor-management issues. Annual wage and benefit packages were negotiated for labor shortly after, in the mid 1980s. 210 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Summary The cases of India, Korea and Brazil offer critical insights to the statistical results presented in Chapter 2. Both the rise of unskilled labor and democracy had important effects on welfare spending in the quantitative and qualitative analysis. The comparative case studies, however, provide a deeper understanding of how country specific differences and shifting of regime type from authoritarianism to democracy affect welfare states. This cross-regional investigation into the realities of welfare spending in three countries went beyond simple confirmation that globalization does indeed affect the politics of welfare spending. Further complexities underlying the various phases of the social policy process were revealed. Ultimately, it is the way in which political variables (i.e. state-labor relations, regime type) interact with a country’s economic variables (i.e., globalization, comparative advantage) that determines social outcomes. Three major conclusions can be drawn from the comparative case studies. First, globalization encourages the growth o f the informal sector in LDCs, which in turn, exacerbates the collective action problems o f labor. This particular conclusion is most valid in countries (e.g., India) that have overly protective labor markets, a high number of surplus laborers, and face a growing global demand for their labor-intensive products. The number of economically active unskilled laborers in such countries is grossly underestimated in quantitative analyses. It is possible, then, that if reliable statistics on informal sector activity were available, then the coefficients for labor’s bargaining power in the estimated model would be much larger. In addition, these case 211 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. studies revealed that the problem of rigid labor markets is a result of the historical (political) relationship between state and labor. Thereby, in order to establish stronger welfare states in this era of globalization, these types of countries have two related policy options: (I) increase government spending on education and training to raise the skill level of the workforce. Workers will then be employable in other sectors and some of the labor surplus will be absorbed (2) gradual reform of state-labor relations by relaxing the most inefficient labor protections.6 The second conclusion is that democracy is favorable fo r social welfare spending, but is not sufficient. The turn to democracy in both Korea and Brazil was a watershed for labor groups, resulting in more advanced labor rights and greater government commitment to social welfare. At the same time, it cannot be ignored that labor mobilization in both Brazil and Korea accompanied the democratization process. This relationship is clearly dynamic and, therefore, the link between globalization, labor mobilization and democracy warrants future investigation. Regardless, India with its longstanding democracy is the deviant case here. While it is true that the combination of both democracy and the socialist ideology resulted in strong labor protections, the democratic variable was ultimately unable to mediate between globalization pressures and declining social expenditures in India. The policy lesson to be learned is that democratic reform should be promoted alongside targeted (economic and political ) efforts to empower weaker labor groups. 6 The key word here is “gradual.” Drastic neo-Iiberal-type labor reforms, which India has in some cases proposed to do, will only create instability and protest from the protected labor groups. Gradual labor reforms should aim towards making labor policies more inclusive o f weaker labor groups. 212 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Finally, the third conclusion is that historical institutional constraints affect the equitable distribution o f social welfare spending. The historical realities of all three countries confirm the validity of this proposition, one that was completely concealed in the statistical analysis. India has developed a welfare system that privileges only the ten percent of the labor force that is organized. Brazil is even more complex, given the labor aristocracy and generous benefits that developed from the history of state corporatism. Currently, a strong clientelistic relationship with political officials continues to plague efforts to reform welfare spending. In Korea, the “growth first” strategy created an ideological barrier to the institutionalization of a welfare state, as well as the slow development of a cohesive labor movement. As a result, Korea’s level of spending on social welfare is exceptionally low relative to its level of economic development. Thus, increasing welfare expenditures does not always indicate that more vulnerable groups are receiving greater protections. Although, indeed, some “trickle-down” effects occur from greater state social commitments, privileged groups tend to enjoy disproportionate gains (e.g., Brazil and India). Following the statistical results, the qualitative analysis shows more forcefully that political factors are decisive in determining social outcomes. In the final analysis, no true outliers emerged. Globalization does affect welfare spending. However, the mediating agents are privileged social groups, labor groups benefiting from export diversification, and a more liberal political regime type. Interestingly, the timing of globalization has also had an important impact on government commitments to social welfare. It seems that countries that began 213 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. globalizing in the 1960s entertained more repressive labor policies. As a late globalizes India entertained strong labor protections and relatively higher benefits through most of the 1970s and 1980s. Brazil and Korea, on the other hand, confronted globalization pressures earlier in the 20th century. For Brazil, globalization gained momentum with the entrance of foreign investors while for Korea, globalization was propelled by its export oriented industrialization strategies. Subsequently, both of these nations implemented relatively more repressive labor policies until the onset of democracy. This leaves many avenues open for future research. Measurement of unskilled labor can be made more accurate by somehow accounting for informal sector populations. Second, sectoral data (i.e. number of strikes, government concessions, labor-management relations) can further verify how and why skilled sectors gain political advantages. Finally, extensive field work can provide more detailed policy tracing of a chosen effective and redistributive-type welfare policy. 214 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. SECTION 2 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 6 REASSESSING THE RELATIONSHIP BETWEEN GLOBALIZATION AND WELFARE: WELFARE SPENDING AND INTERNATIONAL COMPETITIVENESS IN LDCS Having established that welfare spending in LDCs is adversely affected by the pressures of globalization, this chapter questions whether or not such policy choices have positive macroeconomic consequences. How accurate is the logic of the conventional neoliberal agenda that celebrates the efficiency effects of lower welfare spending? Indeed, the evidence in this chapter challenges the conventional wisdom that government welfare spending in this contemporary era of globalization must be sacrificed for the sake of improving international competitiveness. Results from Chapter 2 indicate that the incentives to prioritize market oriented policies over social objectives (i.e., social security and welfare) are particularly high for governments of less developed countries (LDCs). Such economic policy choices are consistent with the chorus of scholars and policymakers who claim welfare state spending is inefficient and erodes a nation’s competitiveness in global markets. Yet a growing number of researchers assert that LDC fiscal policy choices do not in fact adversely affect their competitiveness. Rather, it is international market conditions and the indiscriminate 216 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. behavior of international investors that are stronger determinants of LDC’s national competitiveness and, subsequently, their level of economic globalization. Globalization is not (yet) an inexorable process and thus it is all the more important to investigate the dynamic relationship between economic globalization and welfare spending. If the conventional neoliberal agenda and the call for prudent fiscal policies is effective, then lower social spending should positively affect competitiveness and thereby effect stronger links with international markets.1 This investigation assesses the impact of welfare state spending on economic globalization across 44 LDCs from 1972-1995.2 The results of this section rejects the conventional view that reduced government welfare expenditures increase competitiveness with respect to trade and affects the competition for financial or productive capital flows. This chapter asks whether LDC social welfare policies affect trade and capital differently, if at all. This analysis relies on the assumption that a nation that is becoming more competitive will exhibit increasing levels of economic globalization ceteris paribus. Therefore, if a country is becoming more internationally competitive overall, then its economy is becoming more globalized or more penetrated by international trade or capital investment or both.3 This study posits that the 1 Although some advocates of the conventional thesis might not directly link lowering the share of the public sector to increasing international competitiveness, both improving fiscal discipline and export sales have been primary components of structural adjustment policies in developing countries. See, for example, the Institute for International Economics study by John Williamson 1990 .the World Development Report (1990 and 1995), IMF World Economic Outlook 1997, and Asian Development Outlook 1995. 2 These countries include large parts of Asia, Africa, Middle East, and Latin America, and cover the post- Bretton Woods era. Nine countries had to be dropped from the original sample because of insufficient data. See Appendix B for more details. 3 Recall that the emphasis of this study is on international competitiveness. It could be argued that trade- deficit Country A could become more domestically competitive without makings its economy more globalized (i.e., by lowering costs or improving efficiency of production). However, if this improvement 217 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. competitiveness of a nation is affected by higher (or lower) welfare spending if it is experiencing a corresponding decrease (or increase) in export levels and/or net capital flows. The model for exports is drawn from the debate between the World Bank and the International Labor Organization, while the model for the latter is constructed on the basis of the international (push) verses domestic (pull) source of capital flows. Careful empirical work on the political economy of both trade and capital flows in LDCs has long been needed. Scholars and policymakers alike stand to benefit from research that explores the feedback effects of lower state welfare spending in this era of economic globalization. This analysis accomplishes three things. First, it serves as a more rigorous test of the logic of the conventional wisdom. Research suggests that LDCs are indeed lowering their welfare expenditures in order to mollify international investors and increase national competitiveness. However, in order for the conventional argument to be complete, LDCs should be experiencing increased globalization as a response to this lower level of welfare spending. Second, this type of investigation into the globalization-welfare nexus provides greater insights into the complexity of these two variables and their interrelationship. Economic globalization, after all, is not an inevitable process that is immune to political pressures, particularly since capital has acquired institutional representation in transnational institutions. Finally, by disaggregating the different aspects of economic globalization, into increasing trade, and productive and financial capital mobility, this study explores the differential impact of LDC domestic policy on each of these variables. in competitiveness helps them not to just displace imports, but also to improve exports, then Country A 218 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Conceptual Context Conventional wisdom asserts but does not demonstrate how government welfare expenditures affect international competitiveness. Instead, there seems to be an implicit agreement in the globalization literature that spending on welfare is by definition inefficient and, consequently, dampens the competitiveness of national goods and services as well as incentives for capital investments. For example, Scholte (1997:448) evidently assumes welfare spending harms competitiveness: At a time when the financing of many social security systems were coming under strain..., the added pressure from global capital for reduced taxes and labor costs has driven many governments to cut back welfare programs. In the cause of bolstering global competitiveness, governments across the planet have since 1980 rolled back social democracy and dismantled state socialism. Such shrinkages have been the cornerstone of many “adjustment” packages in the South...Governments have generally implemented the greatest cuts in respect of sunk costs such as unemployment benefits, old-age pensions, and untied official development assistance. Effectively removing questions of “how” and “why” redistribution policies (i.e., welfare spending) affect globalization from the realm of investigation, Gill (1995:417) takes a similar approach: Driven to raise operating finance on the more globalized financial markets, governments are pressured into providing a business climate judged attractive by global standards in order to win and retain foreign direct investment... Traditional forms of state intervention in the economy to promote redistribution have declined, and the socialization of risk for the majority of the population has been eroding. is considered to have increased its international competitiveness. 219 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The problem with this scholarship is that it refers to the relationship between welfare-type fiscal policies and competitiveness in global markets at too high a level of abstraction. As a result, the reality that there are two distinct (albeit related) forces at play, one in export markets and the other in financial markets, and that welfare spending may have differential effects upon both is often overlooked. Underlying debates on each of the two aspects of LDCs economic globalization must be addressed before the conventional “race to the bottom” thesis can be fully validated.4 If the corresponding empirical tests fail to confirm conventional wisdom on both counts, then the idea that welfare spending negatively affects national competitiveness misguides both scholars and policymakers alike. Recall that the analysis in this study focuses specifically upon formalized public programs of income maintenance and welfare services which, according to Pfaller et al., (1991) are the most visible manifestations of welfare statism.5 Note that these schemes are generally financed by employer, employee, and state contributions and are considered to be a component of non-wage labor costs (or social wage). Interestingly, data shows that, on average, contributions from all three sources have fallen in LDCs since 1972 (see Chart 6.1).6 This study focuses specifically on government spending 4 Conventional theorists advocate the race to the bottom thesis. They argue that governments compete with each other by continuously lowering their taxes and labor costs in an effort to increase capital and trade flows. See, for example, Greider (1998) and Nader (1993), Scholte (1997), and Strange (1997). 3 These refer more specifically to income transfers such as pensions, unemployment benefits and family allowances. See Table I for further descriptions. 6 Because of limited data, this pattern could only be confirmed in a limited number of LDCs. Data on government welfare expenditures and economic globalization have been collected for a total of S3 LDCs. O f these, due to a tack of cross-national data on employer and employee contributions, the above pattern could only be verified in 45% of the cases. Thus, the conclusion that employer, employee and government spending in LDCs are declining is based on limited data. 220 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. on social welfare, since it is the government that is the ultimate guarantor of the size and extent of the LDC welfare state. Chart 6.1: Total Social Security Contributions in LDCs* Total Spending on Social Security 1972-1995 (as % of GDP) N= 23 r% y r S ? 5 rfS 1 ' , 8 ? r iiP ^ ^ •Total Spending “ ""“ “ Trendline Source: Government Finance Statistics (IMF: various editions) ♦For countries included, please see Appendix E. How, then, might fiscal policies in this vein affect trade and capital flows of developing countries? Does the empirical evidence confirm the tenets of conventional wisdom that claim welfare spending negatively affects national competitiveness? Put differently, has the rise in LDC exports and capital flows since 1972 occurred in part as a response to lower government spending on welfare? Careful empirical work on the effects of welfare spending on both trade and capital flows in LDCs is an important addition to the globalization-welfare debate. As of yet, there are no studies that have addressed these questions and convincingly demonstrated that the logic behind conventional theories is empirically sound. Responses to these inquiries must be drawn from two different lines of debate: one 221 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. emphasizes determinants of competitiveness in export markets, while the other discusses forces driving financial flows. In the first debate, economists from the World Bank and International Labor Organization deliberate the effects of welfare spending on international competition in export markets. They are particularly concerned with labor costs associated with greater welfare spending.7 The second debate revolves around the primacy of global “push” or domestic “pull” factors of capital flows. If LDC welfare expenditures are indeed affecting the competition for financial flows, then the implication is that domestic determinants (or the-“pull” factors) of capital inflows predominate over systemic ones (or the “push” factors). Welfare Spending and International Competition in Export Markets The debate between World Bank economists and economists from International Labor Organization (ILO) concerns the effects of welfare spending on export competitiveness.8 Advocates of the World Bank perspective argue that contemporary LDC welfare spending, particularly on such items as job security and mandated contributions to social funds, protect labor “excessively” and are thereby distortionary.9 Such interventions interfere with the efficient allocation of resources by driving up 7 There is an alternative argument in the globalization literature that welfare spending indirectly affects export competitiveness by causing an appreciation in the exchange rate. However, this effect is not generalizable since currency appreciation that occurs in response to increased government spending is contingent upon several variables, such as the source of government finance (borrowing instead of taxing or money creation) and on private sector reactions. This analysis is based on a general definition of performing competitiveness (see PfaQeret al., 1991). Therefore, the direct effect of welfare spending on performing competitiveness is the focus o f this paper. Refer to section on "International Competitiveness" (page 22) for a more precise definition of performing competitiveness. 8 The classification of World Bank vs. ILO are general ones, and mirror those used by Freeman 1993. Indeed, there are many outside these institutions who accept similar theoretical positions. At the same time; there are many within the respective institutions who do not accept the general views. 222 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. labor costs and encouraging rent-seeking activities. According to the World Development Report 1990: When governments intervene in the market for...labor, they often exacerbate the anti-labor bias of protection. Labor market policies— minimum wages, job security regulations, and social security—are usually intended to raise welfare or reduce exploitation. But they actually work to raise the cost of labor in the formal sector and reduce labor demand...and thus depress labor incomes where most of the poor are found (World Bank, 1990:63). The idea is that a state committed to few of such labor market interventions can instead devote resources to raising national output and promoting competitiveness.1 0 Freeing private enterprise of such “onerous burdens” is extremely important for improving labor’s performance.1 1 In contrast, the ILO argues that welfare spending has a positive effect on competitiveness. By providing the necessary income security, the corresponding improvements in social well-being ultimately result in higher labor productivity. It is, then, the World Bank perspective that provides fodder for the conventional belief that there is an inverse relationship between welfare spending and competitiveness in international export markets. The World Bank perspective focuses upon three supply-side arguments for why social welfare programs might adversely affect competitiveness. The first of the 9 Because World Bank views traditional pay-as-you-go social security and welfare systems as less efficient, their policy recommendations call for privatization of such schemes. For examples, see Estelle, 1992; Holzmann, 1997 and World Bank Policy Research Report 1994. 1 0 The idea is that better fiscal discipline will lead to lower inflation, stable exchange rates and more favorable export prices. Therefore, greater domestic economic stability is positively related to their international economic performance. See, for example; World Bank Report 1990 and, specifically, the discussion in Chapter 7. 1 1 Pfeller et aL (1991) notes that the challenge of competitiveness can be met in several ways: currency devaluation, reducing welfare statism, reduction of other costs—most of all wages, and redistribution of costs of welfare statism away from enterprises to households. The analysis in this study focuses only on the reduction of welfare statism and tests the effectiveness o f this political choice on international m arkets 223 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. arguments is that interventions tend to impose labor rigidities and drive up labor costs beyond equilibrium levels. The setting of labor standards causes misallocation of resources and prevents markets from reaching optimality conditions (Freeman, 1993). Riveros (1992) argues that the resulting level of the non-wage costs of labor significantly affects manufactured exports in LDCs. Unfettered labor markets, in contrast, set wages and levels of employment that are closer to being Pareto-efficient. Thus, reducing total labor costs would lead to increased economic openness, greater economic growth and, thereby, a higher standard of living for workers. The second line of reasoning of the World Bank is that such welfare programs indirectly influence export competitiveness by encouraging rent-seeking activities. Social welfare policies prompt capitalists, workers and even governments in LDCs to devote resources towards rent- seeking instead of raising national output. Many social security experts have argued that governments, in exchange for political influence, often distribute benefits to labor (Baneiji & Ghanem, 1997; Midgeley, 1984; Pedersen, 1997). At the same time, Esping Andersen (1996) implies that traditional elites have an interest in pressuring governments to maintain their traditional privileges (i.e., non-taxation of the rich in Latin America). Either way, whether welfare programs encourage rent- seeking in the form of income transfers from political decision makers to workers or capitalists, they require real resources that are withdrawn from productive activities.1 2 The net effect is a loss in (domestic) efficiency in the Paretian sense. Here, the link between welfare policies and export competitiveness is indirect in that there is a 1 2 See Pedersen (1997) for the theoretical link between rent-seeking and Bhagwati’s (1982) directly unproductive, profit-seeking activities (DUP). Pedersen also presents an interesting argument on the political economy o f distribution between capitalists, workers, and governments. He argues that political 224 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. “deadweight” loss of resources that could have been otherwise directed towards activities involving export production. The third argument refers to issues of labor productivity. Simply put, interventions in the labor market create a moral hazard problem by stifling incentives to work, save and invest. Unemployment benefits and compensation schemes, for instance, are said to moderate the disciplinary effect of unemployment on work intensity (Marshall, 1994). In another example, the World Bank Research Report (1994) argues that early retirement provisions reduce the supply of experienced workers and resulting in a labor force comprised mostly of relatively less experienced workers. Thus, workers subject to pure market forces perform more efficiently and raise national output. The idea is that in the contemporary era, the work force should become even more “disciplined and malleable” in order to keep up with global competition. Marshall (1994:55) describes the alleged trade-off between welfare spending and export competitiveness faced by nations: A new emphasis has been placed on the alleged need for greater flexibility and less regulation of dismissal and contracts of employment. Whether expressing the more sophisticated or popular form, these views assume that for export competitiveness to improve, labor costs must go down, the workforce must become more disciplined and malleable, and individual efforts must increase. In the context of such views, labor protection and trade union intervention in the labor market and at the workplace are perceived simply as obstructions to the achievement of those aims. Figure 6.1 illustrates the World Bank view. decision makers represent an interest group themselves and prefer income transfers that will serve their own interests. 225 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 6.1: The World Bank perspective | Welfare 1 Spending 1 r | Labor 1 Costs The World Bank perspective (representing conventional t wisdom) | Productivity 1 t | Exports ILO theories, on the other hand, argue that welfare programs are socially beneficial and thereby positively contribute to economic competitiveness and economic development. While this perspective is "more diffuse and less analytically grounded” than the World Bank’s, it provides theories strong enough to present a case in favor of interventions (Freeman, 1992). Theorists supporting the ILO view argue that it is all the more crucial to maintain labor standards in the era of globalization so that social stability can accompany greater market exposure and rapid economic growth. ILO justifications for welfare policies are both economic and non-economic. Support for social programs is first and foremost based on moral imperatives. The risks and uncertainties that accompany market development require that the state 226 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. provide a minimum level of economic security for its citizens. Citizens must be compensated for the negative externalities of the market, even if market integration may benefit society in the long run. Second, in contrast to the World Bank, the ILO argues that government mandated labor standards are actually Pareto improvements. Such benefits directly mitigate the principal-agent problem not considered in World Bank assessments of the costs and benefits of welfare programs. By heightening worker motivation and workplace cooperation, labor policies can increase workers’ attachments to the firm (Kenworthy, 1999; Marshall, 1994). Therefore, the ILO stresses that higher labor benefits can actually increase the productivity of labor. The consequent reduction in per unit labor costs results in a net improvement in competitiveness. Sengenberger (1993:327), in the International Labour Review, summarizes the net benefits of basic social welfare programs: Once assured of minimum protection firms and other members of the community have an incentive to search for other, more constructive responses to competitive pressures, such as the introduction of better products and processes, a more rational utilization of their physical and human resources and an improved infrastructure. [Minimum labour] standards can thus act as an inducement to endogenous development. Ultimately, according to the ILO, there is no trade-off between equity and efficiency (see Figure 6.2). By improving social well-being and increasing labor productivity, implementation of social welfare programs makes capitalism compatible with overriding social objectives (Pfaller et al., 1991). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 6.2: The ILO perspective | Exports | Productivity Welfare Spending conventional wisdom) Welfare Spending and the Competition fo r Capital Flows Conventional theorists such as Cemy (1995), Grieder (1998), Grunberg (1998) Nader et al. (1993), Rodnik (1997b), Scholte (1997), and Strange(l997)1 3 base their analyses on the assumption that fiscal policies, particularly concerning social welfare, are fundamental determinants of capital flows. These analysts adopt the domestic “puli’ interpretation of capital flows, or the conception that international capital flows are “pulled” in by an attractive domestic investment climate. Hereby, lower welfare expenditures, by signaling strong fiscal discipline and a more “friendly” tax environment, attract capital flows. This view, however, has been challenged by the 228 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. recent works of theorists such as and Doodley et al. (1996), Femandez-Arias (1994), Maxfield (1998), and Reisen (1996). They question this method of emphasizing domestic conditions over systemic ones in attracting capital inflows. Capital flows to LDCs, they argue, are more a function of the business cycle in developed countries rather than domestic LDC conditions. Consequently, it is unfavorable conditions in developed countries, such as low interest rates, that “push” capital flows towards LDCs. Much empirical work has been conducted on the merits of the push vs. pull hypothesis. Yet this is the first general cross-national study to model the particular effects of government welfare spending on capital flows. It presents an alternative way to analyze the conventional wisdom while putting the contending (global push and domestic pull) hypotheses to an empirical test. It is interesting to note that in the “pull” story, capital inflows are subject to the full control of policymakers. In this perspective, if the goal is greater access to capital markets, governments are responsible for implementing the appropriate fiscal policies. Otherwise, bond markets, for example, will tend to punish government policies such as welfare that can result in high inflation and large budget deficits. These pressures are now particularly high since the form of capital flows to LDCs has changed. Since the mid 1980's, most foreign capital flows to LDCs have consisted of foreign direct investment, and portfolio equity flows (Griffith-Jones & Stallings, 1995; Kuczynski, 1994; Reisen, 1996). This is in contrast to the predominance of commercial bank flows to LDCs in the 1970s. Thus, the current form of external capital flows matter because 1 3 Rodrik (1997 and 1999) is not a conventional theorist as referred to in this paper. However, he is included here because he argues that although greater openness induces greater government spending, capital mobility mitigates this effect o f trade. 229 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LDCs are increasingly subject to the judgement of international investors (Baer & Hargis, 1997). More emphasis is being placed on LDC fiscal policies as investors can easily signal their displeasure with government policies with “exit” or the threat of “exit.”1 4 Conventional theories suggest that the link between welfare spending and capital inflows also occur by way of the domestic tax structure. A more favorable tax environment is attractive to capital and thus serves as an important “pull” factor (see Figure 6.3). Rodrik (1997) argues that the relative ease of capital owners to move in and out of the domestic economy and avoid higher taxes causes the tax base to become dependent upon labor. When globalization moves beyond a certain point, governments can no longer finance social welfare (Rodrik, 1997: 55). He says: At the present[...], international economic integration is taking place against the background of receding governments and diminished social obligations. The welfare state has been under attack for two decades. Moreover, the increasing mobility of capital has rendered an important segment of the tax base footloose, leaving governments with the unappetizing option of increasing tax rates disproportionately on labor income (Rodrik, 1997:6). Governments may lower social insurance payroll taxes to reduce labor costs and provide a strong lure for direct investment, particularly in labor intensive industries. Grunberg (1998) points out that this growing tax competition between nations is accompanied by fiscal degradation. Less taxes eventually means less social spending since the loss of fiscal revenue makes it even more difficult to increase and/or sustain 1 4 While the threat o f exit is most applicable to portfolio investments, it also applies to foreign direct investments which can shift production to offshore factories with relatively low start up costs. 230 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. welfare expenditures. Relatediy, the international market driven preference for lower taxes leads to lower government incentives for maintaining welfare states. Figure 6.3 illustrates the domestic pull hypothesis: Figure 63: The Domestic Pull Hypothesis W elfare Spending | Capital Inflows to LDCs Attraction ot Domestic Investment Climate__ The Domestic Pull Hypothesis (representing conventional wisdom) G reater returns on Investments Capital Inflows to LDCs offering higher rates of return Prudent fiscal policies Concepts In contrast, the “push” theorists dismiss much of the conventional arguments by pointing out the theoretical and empirical weakness of their assumptions. They argue that capital inflows are to a large extent beyond the control of policymakers. Investors are presumed to be rational and have perfect information. Yet both Maxfield (1998) and Bertolini and Drazden (1997) argue that investors tend to become less discriminatory when world interest rates are low. As Maxfield (1998:1201) states, financial markets may, in fact, be irrational, and “psychology rather than economics drives capital flows.” This tendency is intensified by the problem of information asymmetry, which is particularly acute in the globalizing world (Reisen, 1996). Contagion effects, another 231 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. systemic phenomenon, also demonstrate the problem of herd behavior among international investors. Panic caused by financial conditions in one country triggers capital flight in several LDCs, usually based on incomplete information about domestic conditions (see Figure 6.4). These scholars conclude that it is wrong to assume that international investors are making investment decisions according to the creditworthiness of LDCs. Thus, with some country exceptions, Femandez-Arias (1996) and Doodley et al. (1996) find that international interest rates have been the dominant factor in explaining variations in annual private net capital flows to LDCs. Domestic policy, then, is not constrained by the inability to tax capital or implement social welfare policies. Rather, it is limited by high local interest rates and stable exchange rate expectations when global liquidity is tight (Maxfield 1998). Figure 6.4 represents the push hypothesis: Figure 6.4: The Global Push Hypothesis International Economic Conditions Greater risk taking Capital inflows to LDCs with higher rates or return t Capital Inflows I tA i n r . International Interest Rate 232 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The Empirical Model These models will be estimated by cross-section time series data. Panel data techniques and the fixed effects method have been chosen to assess the impact of welfare spending on national competitiveness for three reasons. First, panel data sets increase efficiency by using a large number of data points, increasing the degrees'of freedom and reducing the collinearity among explanatory variables. By following a given country over time as it changes status (e.g., from more welfare spending to less, or vice-versa), panels enable a proper recursive structure to study the before/after effects (Hsiao, 1986). Second, the fixed effects procedure allows one to control for the effects of missing or unobserved variables by utilizing information on the intertemporal dynamics as well as the individuality of entities. Such a procedure eliminates much of the omitted variable bias. For example, country effects capture land size, a variable that is highly correlated with trade.1 5 In another example, fixed effects are particularly important for capturing the tendency towards high inflation (especially for some Ladn American countries) in the model for capital flows. Ultimately, fixed effects controls for such country-specific differences without having to model them explicitly. The proposed model posits that LDCs do not satisfy the conditions of perfect competition. By lagging the necessary explanatory variables, this model takes into account the time lapse involved with political decision making, economic adjustments, and the allocation of resources. Moreover, one might argue that the causality is reversed and that private capital flows affect changes in gross domestic product (GDP), foreign > s In addition, for reasons o f linear dependency, land size could not be used as an independent regressor. Instead, controlling for country effects ‘swallows up’ the effects of land size and is therefore included in the model. 233 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. exchange reserves and the level of International Monetary Fund (IMF) credits.1 6 Yet it is not possible for current private inflows to affect these same variables in the previous period. Therefore, this estimation technique mitigates simultaneity problems by lagging some of the necessary variables such as welfare, GDP, IMF credits, and foreign exchange reserves. Both welfare spending and IMF credits were lagged up to 4 years. Four years was the limit chosen since some LDCs did not begin the process of liberalization until the early 90’s. Lagging any more than four years would interfere with analysis of the recursive effects of lower welfare spending. The Models The following equations model both the international and domestic determinants of exports and capital flows. See Table 6.1 for a summarized description of all the variables. Model 1 export= btexportj(t-i) + b2 welfi(t-i) + b3 netfdij(t-i) + bjSbjXu + Ibkcountry; + Zbiyeart + H it Model 2 netk= b inetkj(t-t) + b2 welf^.i) + b3 tradej(t-i) + b-dntdifflt + bjZbjXjt + Ibkcountryj + Zbiyeart +mt Model 3 fdi= b[fdij(t-i) + b2 welfi(t.i)+ b3 tradei(t-i) + bjEbjXjt + Zbtcountryi + Sbiyeart + pit 1 6 The democracy variable is not included here because it is not plausible that private inflows affect the level o f democracy in the same period. To check this, the democracy variable was tested both ways- by lagging and not lagging. The lagged democracy variable was insignificant each time. 234 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The b 's are parameter estimates in this equation, while the subscripts / and t represent the country and year of the observations respectively; bi is the lagged rate of openness, incorporated to alleviate problems of serial correlation across error terms; |i is an error term. UC represents the vector of control variables, or GDP, democracy, pressures from international finance institutions and foreign reserves. The international level variables are lagged in order to take the period of “adjustments” into account. Also, note that logarithms are taken of all the variables. This type of relationship is useful in that it displays the property of constant elasticities between the variables. Models 1, 2, and 3 test the conventional wisdom that LDC governments are meeting the challenge of increasing competitiveness (1972-1995) by reducing government welfare expenditures. Model 1 assesses the conventional wisdom as it applies to competitiveness in export markets. If the World Bank is correct, then b2 will be negative and the empirical evidence will support the conventional claims that lower welfare spending improves international competitiveness. However, if the ILO perspective is more accurate, then b2 will be positive and welfare spending actually supports an increase in international competitiveness. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 6.1; Concepts, Measurements and Definitions Concepts Measurements* Definition National Competitiveness The value of total exports [EXPORTS], net private capital flows [NETKJand net foreign direct investment [FDI] as a percentage o f GDP* dependent variables. Government welfare expenditures Social security and welfare as a percentage of GDP (+/-) [WELF] Exports is the level of exports divided by GDP and as a proportion o f total LDC exports (this total refers to the 44 LDCs in the full sample). Net capital flows consist of private debt and nondebt flows. Private debt flows include commercial bank lending, bonds and other private credits; nondebt private flows are foreign direct investment and portfolio equity investment divided by GDP). ‘Social security' consists of Income transfers, providing benefits in cash or in kind for old age, invalidity or death, survivors, sickness and maternity, work Injury, unemployment, family allowance, and health care. ‘Welfare affairs and services' are defined as assistance delivered to clients or groups of clients with specials needs, such as the young, the old, or the handicapped. Economic development Political Regime The Gross Domestic Product (+) [GDP]; 'GDP is the total gross domestic product of a country. Indicator o f democracy (+/-) [DEMOC] Using scale 0-10; 10 "strong democracy. This indicator is derived from the codings on the competitiveness o f political participation, the openness and competitiveness of executive recruitment, and constraints on the chief executive. Pressure Com Internariona! Financial Institutions Foreign reserves Interest rates Use o f Credits front the International Monetary Fund (+) [IMF] Foreign exchange reserves (+/-) [RESV] The difference between international Interest rates and domestic interest rates MflNTDIFF] Denotes repurchase obligations to the IMF for all uses of IMF resources. Includes enlarged access resources, trust firnd loans, and operations under structural adjustment facilities. Holdings of foreign gold as % of imports. reserves minus London Interbank Offer Rates represent the international interest rates. The interest rate for each developing country is represented by the ‘average* annual interest rate. •The signs in the parentheses under measurements represent the expected direction of the relationship. Multiple signs mean that there is an underlying debate regarding the expected direction of the relationship. For data sources, list of countries and years included, and more detailed definitions on some of the variables, see Appendix D. 236 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Models 2 and 3 assess the effects of specific LDCs’ fiscal policies on capital inflows. Private net capital flows are the favored proxy because they represent gross inflows minus amortization. They therefore convey some estimation of how much “capital” remains within the country.1 7 Net foreign direct investment ( FDI) flow, used to represent productive capital mobility, is the third dependent variable. While Model 2 analyzes the determinants of productive and financial capital flows, Model 3 isolates the effects of welfare spending on productive capital flows. If the logic of the conventional theorists is correct and prudent fiscal policies are domestic “pull” factors that affect capital inflows, then 6 2 will be negative.1 8 If, instead, the global push factors, such as international interest rates, are the major determinants of capital flows to LDCs, then b2 may be either positive or insignificant. Credibility of the global push hypothesis will be tested in two ways. The significance level of the dummy variables that take year specific effects into account will be the first indication. This series of year dummy variables capture the effects of common shocks experienced by all the developing countries in a given year. For instance, if the world interest rate was low and indeed driving capital flows towards LDCs, then without inclusion of the year dummies, the democracy and/or welfare variable would likely be biased upward-showing high significance levels and a positive coefficient. The inclusion of the year dummies, however, might generate positive 1 7 In contrast, gross capital flows is an estimate of the degree o f capital mobility (see Montiel, 1994), or the cross-border flows of capital. Therefore, capital mobility might be high, while the country’s ability to attract capital relative to others might be low. It is thus an insufficient indicator o f international competitiveness for capital l* GDP, foreign reserves, democracy are also considered pull factors, but are not the focus of this paper or o f the conventional theories on welfare. 237 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. coefficients for those years and render the democracy and/or welfare variable as insignificant. Thus, these coefficients will capture some of the variation in average annual global interest rates. If they are insignificant, then year effects do not exist and international conditions are less likely to be a determinant of LDCs capital flows. The second test of global push factors will be the interest rate differential between the London Interbank Offer Rates (LIBOR) and the domestic interest rate of developing countries.1 9 If this coefficient is negative, or capital inflows are greater when LDC domestic interest rates are higher than LIBOR, then the push hypothesis will be confirmed. The Variables International Competitiveness: Dependent Variables [NETK, FDI, and EXPORT] All three models operationalize national competitiveness by the level of exports and net (productive and financial) capital flows relative to GDP, based on Jeffry Hart’s Rival Capitalists (1992:7).2 0 To check the robustness of these findings, the results were replicated using an alternative measure of export competitiveness and competition for capital flows. Both exports and net capital flows as a share of the total LDC (using the entire 44-country sample) exports and share of total net capital flows, respectively, are applied as alternative indicators of competitiveness. These two 1 9 LIBOR commonly represents the international interest rate. 2 0 Hart (1997) argues that for the economy as a 'whole, rising real GDP, relatively stable inflation rates, stable or rismg export levels as a percent of GDP, trade surpluses, and rising labor productivity demonstrate increasing competitiveness. Note that Hart (1997) does not specifically suggest capital flows as a percentage o f GDP as a measure for national competitiveness. This measures is adopted from 238 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. measures may be related, but they are from identical. Nonetheless, no serious differences resulted from using this second indicator. Indeed, there is an underlying debate regarding the characterization of “competitiveness.” Economists are quick to remind us that firms are competitive, not nations. Perry and Robertson (1998) convincingly argue, however, that nations do compete for the confidence of capital market investors.2 1 By the same token, many scholars address the concept of competitiveness, as it is applied to trade performance, on an economy-wide basis. Alesina and Perotti (1997) monitor improvements in competitiveness by falls in the relative unit labor costs in manufacturing in one country relative to its competitors.2 2 The actual chance of selling products in an internationally contested market is then called its performing competitiveness.2 3 According to Pfaller et al. (1991:6), a country may be forced to reduce costly welfare expenditures in order to safeguard its performing competitiveness. Note that this concept of competitiveness operates under the assumption that governments are concerned with increasing aggregate economic performance. Significantly, although the proposed models analyze capital and trade flows independently, it cannot be denied that the two economic forces are interrelated. Net capital flows, for instance, are expected to have a negative effect on trade. Large capital inflows are often accompanied by inflationary pressures, real appreciation of the Fcmandez-Arias (1996) and Claessens and Naude (1993). See section on “Data Limitations” for further explanation. 2 1 Note that Perry and Robertson (1998) do not directly refer to states, but to political-economic systems. 3 2 Sanyal (1993) argues that a country's short-term competitiveness depends on the ability to produce at a lower cost Long-term competitiveness, on the other hand, depends on improvements in productivity. The analysis in this paper focuses on short-term competitiveness. 239 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. exchange rate, and deterioration in current accounts (Calvo et al., 1994). Simply put, high capital inflows might adversely affect trade ratios through their direct effect on the exchange rate. It is thus expected that domestic export sectors have incentives to pressure governments for increased capital controls. In contrast, it is plausible that high levels of trade have a positive effect on both productive and financial capital flows.2 4 A liberal trade and payments regime is often used as country creditworthiness indicators by international investors (Lensink & White, 1998). Additionally, the fact that inputs can be procured from the cheapest source and dividends can be repatriated boosts investor confidence (Asian Development Bank, 1995). Higher trade flows should then attract greater capital flows. Therefore, for both trade and capital flows, the lagged variables of the other are used as regressors. Capital flows are lagged in Model 1 to avoid simultaneity problems, while trade flows are lagged in Models 2 and 3 since risk ratings are made on the basis of historical data.2 3 Social Security and Welfare [WELF] LDCs’ government welfare spending is the primary variable of interest. Welfare spending, as it is measured in this analysis, provides an indication of both the prudent fiscal policy and the social wage. The social wage variable is of direct relevance to the ILO-World Bank debate because it focuses on benefits available to the working population and is therefore a component of total labor costs. Government welfare 2 3 Underlying competitiveness is the ability to provide qualitative excellence, while performing competitiveness refers to ability to sell on the world market. See Pfaller et aL (1991) for further details on the distinction between underlying competitiveness and performing competitiveness. 2 4 Here, “trade” refers to the level of imports and exports as percentage of GDP. 2 5 Interestingly, multicollinearity was not a problem here. Correlation between capital flows and trade was moderate (0.28). 240 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. expenditures in Model 2 also contributes to the push-pull debate of capital inflows. Reductions in welfare spending represents (potentially) lower inflation, less deficit spending and a more favorable tax environment for the pull theorists Therefore, if the pull hypothesis is valid, lower levels of spending represent prudent fiscal policy (consistent with the neoliberal agenda) and should therefore attract more capital flows. Here it is important to emphasize that it is government spending on social security and welfare that is of interest. Data illustrate that total contributions to social security have been falling since 1972 (refer to Chart 6.2).2 6 Notice that government, employer and employee contributions follow similar trendlines (see Charts 6.2, 6.3, and 6.4). Importantly, only government and employer contributions can affect labor costs.2 7 Governments hold the option of using general revenues to compensate for a fall in employer contributions to labor. Yet the data show strong indications that governments are following the same expenditure patterns as employers. It is not feasible to use data on employer contributions to assess the level of non-wage costs because such cross country data is unavailable and, when it does exist, there is a very low number of observations. Thus, for this analysis, government contributions to social security and welfare are used as a proxy for the level of social wages instead of employer contributions in LDCs (see, for example, Kenworthy, 1999).2 8 2 5 Note that government contributions apply to social security and welfare. Employer contributions exclude welfare. 2 7 It is interesting that employee contributions have also been M ing. Recall that it is the government that determines the size and extent of the welfare state. If governments opt to reduce the welfare state, then it is expected that contribution from all three sources will fall. 2 8 Kenworthy (1999) disaggregates government transfers and social wage. However, his definition of social wage parallels the concept of social security and welfare applied in this paper. 2*1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chart 6.2: Government contributions to Social Welfare in LDCs* G o v e r n m e n t C o n t r i b u t i o n s t o S o c i a l S e c u r i t y a n d Government Spend ng Trendine 0.015 0.010 ‘Please see Appendix E for countries included Source: Government Finance Statistics (IMF: Various editions) Chart 6.3: Total Employer Contributions to Social Security in LDCs* E m p l o y e r C o n t r i b u t i o n s t o S o c i a l S e c u r i t y 1 9 7 2 - 1 9 9 5 ( a s % o f G O P ) & & & * Employer Spending 1 “Trendline Please see Appendix E for countries included Source: Government Finance Statistics (IMF: Various editions) 2 4 2 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chart 6.4: Total Employee Contributions to Social Security in LDCs'" E m p l o y e e C o n t r i b u t i o n s t o S o c i a l S e c u r i t y 1 9 7 2 - 1 9 9 5 (UBS 0KB 0015 0.01 0005 0 I "'♦"E m ployee Spending “ Trendline ‘ Please see Appendix E for countries included Source: Government Finance Statistics (IMF: Various editions) Interest Rates [INTDIFF] The interest rate differential between developed and developing countries is, according to the push hypothesis, the fundamental determinant of capital inflows to LDCs. If this differential is high, then capital inflows to LDCs should be greater. Recall that conventional theorists do not place any emphasis on the push factors. Therefore, if the latter are correct, then INTDIFF, or b4 , will be insignificant. 243 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The push hypothesis, however, will be verified if b4 <0. LIBOR is commonly used to represent the international interest rate.2 9 Three different measures of the domestic interest rate are applied alternatively to ensure the robustness of the results: average interest rate, official interest rate and private interest rate.3 0 INTDIFF is estimated as follows: b4= LIBOR - domestic interest rate Thus, if the domestic interest rate is higher than LIBOR, then b4 will be negative as capital flows toward LDCs. Conversely, if LIBOR is higher than the domestic interest rate, then capital will flow towards the developed countries and away from LDCs (b4 , of course, will still be negative). Indeed, the INTDIFF variable is a very important indicator of the push hypothesis. Control Variables: Democracy [ D E M O C ] , GDP [ G D P ] , IMF credits [ IM F ] , Foreign Exchange Reserves [RESVj The vector LXjt represents the range of political and economic control variables. According to existing studies, the level of democracy, pressure from international institutions, the level of economic development, and foreign exchange reserves influence the levels of capital and trade flows. It is expected that all these control variables will be positively correlated with the dependent variables. Certainly, there are 2 9 This analysis uses the one year rates, which differ very little from one month, three-month and even six-month rates. 3 0 Since using the different measures did not show much difference in the estimations, the results in this paper reflect only the average interest rate. 244 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ongoing debates on the relationship between some of these control variables and the dependent variable (democracy and trade, for example). This study is useful in that, while its main concern is to analyze the effects of welfare spending on international competitiveness, the findings also contribute to some of the accompanying debates. However, in order to sustain the focus of this study, the parallel debates will be discussed only briefly. Most of the variables and their expected signs are self-explanatory. IMF credits [IMF] should be positively related to trade and capital flows, while GDP [GDP] is likely to be negative since larger economies tend to be self-sufficient and thereby, less dependent upon both types of flows. Foreign reserves [RESV], however, are included only in Model 2 as they are directly relevant to capital flows.3 1 The impact of reserves on capital flows is not as straightforward as one might expect. Certainly, the traditional wisdom is that RESV is positively related to capital flows. Countries with high foreign reserves are supposedly better able to adjust to the destabilizing effects of capital inflows because these stocks can be used as “shock absorbers.” Under such conditions, governments are less likely to restrict capital inflows and outflows. When foreign exchange is low, however, speculative attacks on its currency will lead governments to impose capital controls and create a dampening effect on inflows (Leblang, 1997).3 2 Haggard and Maxfield (1996), however, challenge this economic interpretation. They argue that balance of payments crises lend greater bargaining power to international-oriented sectors and thus lead to liberalization of capital 3 0 Since using the different measures did not show much difference in the estimations, the results in this paper reflect only the average interest rate. 245 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. accounts. Therefore, in this analysis, if capital inflows are positively associated with low foreign reserves (an indicator of crisis), then the Haggard and Maxfield hypothesis will be confirmed.3 3 The democracy variable [DEMOC] deserves a little more explanation. Building on the literature in political science, scholars debate whether democratic countries will be more competitive in the global economy than less democratic ones. Pro-democracy advocates (those who believe that democratic countries trade more) argue that the plurality of votes in a democracy is a check on the power of narrower interest groups. For example, Baneiji and Ghanem (1997) state that democratic countries trade more because they are less likely to pass inefficient labor legislation benefiting “insiders.”3 4 On the other hand, democracy skeptics (those who argue that democratic countries trade less) claim that such political freedoms have a negative effect on trade (Destler, 1995; Perry & Robertson, 1998; Verdier, 1998). They argue that because democracies encourage a greater role for interest groups (i.e., from the non-tradable sector) in the legislative process, democracies have deleterious consequences for trade. Thus, since there is no consensus on the issue, more empirical evidence is needed on the issue to help solve the debate. 3 3 Note that Model 2 assumes that low foreign reserves are one way of measuring crisis in LDCs. If foreign reserves are high enough to stabilize the effects of balance of payments shocks, then a crisis will not occur. 3 4 An example would be workers in privileged industries, or the urban labor elite. 246 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. There is a similar lack of consensus regarding the relationship between democracy and capital flows. One might expect that democracies suggest political stability and, thereby, encourage private capital. The improved transparency and availability of information linked to democracies should be a further incentive for investment. Yet Maxfield (1998) points out the possibility that soft authoritarian regimes might be preferable to democracies because of the greater ease in privileging capital’s policy demands over local constituents. As the editors of the Columbia Journal o f World Business (1994:6) put it, any “deviation from the ‘Washington doctrine’ [will] have a chilling effect [sic] capable of casting a pall over international investor enthusiasm.” It is thus easy to imagine why international investors might shy away from the political uncertainty still prevalent in the developing democracies of today, given the priorities placed on the neoliberal agenda. Here again, the findings of this study will contribute to the substance of this debate. Data Lim itations and Notes on Methodology The aim here is not to resolve the debate on what constitutes the competitiveness of a country or what is the best measure. Instead, it is to assess the impact of social welfare spending on a country’s ability to export more goods in international markets and to attract greater capital inflows. The ILO-World Bank debate, as well as the pull or push controversy, presents the appropriate frameworks in which to address this underlying agenda. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Some obvious difficulties exist in measuring national competitiveness. Indeed, as Krugman (1994) has stressed, the success of one country in the world market does not have to be at the expense of another country.3 5 Exports as a percentage of GDP, exports as a share of LDC exports, and the magnitude of the trade balance are popular proxies of a country’s general economic strength and competitiveness (Ezeala-Harrison, 1999; Hart, 1992; Haque et al., 1995). Yet although all three measures have their drawbacks, export intensity is a better proxy of competitiveness (over time) than trade balances. It is important to emphasize that the trade balance can be affected by factors other than competitiveness. For instance, South Korea is well known as a successful exporter. However, it ran a substantial trade deficit for many years because of its high imports of capital goods and technology needed for its heavy investments (Haque et al., 1995). Using exports as a measure of competitiveness, on the other hand, can also be problematic because temporary currency devaluations can cause a rise in exports. The upshot is that unlike trade balances, applying a fairly long time series analysis can mitigate this shortcoming of using exports/GDP or exports/total LDC exports. Even more limiting is the use of net capital flows as percentage of GDP to assess a country’s ability to attract foreign capital flows. First, wealthier countries (i.e., the United States) might show a low capital flow ratio because of the large size of their GDP. However, the data set used in this analysis is confined to developing countries and, thus, can avoid the problem of extremely high GDP’s skewing this ratio. Both Claessens and Naude (1993) and Femandez-Arias (1996) are good examples of works 248 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. that have successfully applied the net capital flows/GDP measure in developing countries to assess “competitiveness” as it is applied in this analysis.3 6 Alternatively, net capital flows are measured as a share of total LDC net capital flows in order to mitigate the problems using GDP ratios. An argument against the use of both these measures might be that countries with high savings rate might “need” less capital inflows. However, data indicates that high savings countries, such as South Korea and Singapore, have been experiencing increasing trends in capital flows. Finally, issues of data quality and comparability are likely to be problematic in cross-section, time-series data. The severity of this problem is mitigated by using mostly IMF and World Bank data. It is then reasonable to assume that data was systematically gathered for all 44 developing countries used in this data set. However, this approach does not completely eliminate problems of comparability (Harrison, 1994). Data definitions and coverage are subject to change even within the same country overtime. Discussion of Results Overview The following results are conditioned upon international and domestic determinants of trade and capital flows. Altogether, there is weak support for the conventional wisdom that greater welfare spending has an adverse effect on LDCs* 3 5 Krugman (1994) argues that national competitiveness is difficult to assess since higher exports can also mean higher imports, and also because a country may be competitive in one area and not in others. 3 6 This means that Femandez-Arias (1996) also uses the measure of net flows/GDP to address the push- pull debate. Claessens and Naude (1993), however, focus more on capital flight (or capital outflows) instead of net capital flows. Their intent is not to address the push-pull debate. However, the measurement concepts are relatively the same. 249 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. international competitiveness. The results from Model 1 do not favor either the World Bank or ILO perspective. According to the data, even a four-year cycle did not lead to are any efficiency effects from reduced welfare expenditures (or lower social wage). Similarly, welfare spending has no effect on a nation's competition for capital flows, either productive or financial. Instead, the results from Models 2 and 3 give very strong support to the push hypothesis, or the argument that international factors are the primary determinants of capital inflows to LDCs. In sum, welfare spending has differential effects on trade and capital flows, affecting the former more than the latter. M odel I: The World Bank vs. ILO The effects of welfare spending on the level of exports relative to GDP and total LDC exports are reported in Table 6.2. Not much difference occurred between the two competitiveness indicators, reinforcing the robustness of the results. The main variations in results were the variations in significance levels for the IMF. Welfare spending was lagged a maximum of two times in the second measure of competitiveness, since there were significant results that emerged. The results from both estimation provide greater confidence in the result that welfare spending does not international competitiveness. These findings support neither the conventional or the ILO wisdom. Several combinations of the explanatory variables were tested. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 6.2: Regression Results for Welfare Effects on Export Competitiveness Model 1 Dependent variable: export competitiveness E x p o rts (lagged) 0.607** * 0.721*** (0.031) (0.035) w elf* Gagged 4 years) -0.007 -0.011 (0.010) (0.013) n etfd i Gagged 1 year) 0.016* 0.018* (0.009) (0.011) gdp Gagged I year) 0.127*** -0.462*** (0.021) (0.036) D em oc 0.004 0.003 (0.003) (0.004) imf(lagged 3 years) 0.002* -0.002 (0.001) '0.001) 1 Country effects Yes” Yes” Year effects Yes” Yes” R2 .997 .998 N 1008 F value 10.06 Fixed effects regression estimates. ***p<0.01; ** 01<p<0.05; *0.05<p<0.10. Figures in parenthesis represent standard errors. a. The welfare variable in exports as a share of total LDC exports was only lagged two years before significant results emerged. *F test for fixed effects. p<0.01 251 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. No trace of the efficiency effects of lower welfare spending appears for three years. The fact that these results do not provide strong support for either the World Bank or the ILO’s perspective leaves much room for conjecture. How valid are government policies that reduce welfare spending if the effects on competitiveness have not yet been realized? Are there better and more immediate ways to increase efficiency and improve international competitiveness? If so, then why was reducing welfare spending the political choice of LDC governments? The fixed effects test revealed that country specific differences and year effects account for most of the growth in trade flows (significant at the 99% confidence level). This suggests that both historical and international conditions are the most important determinants of LDCs’ international competitiveness in export markets. It is likely that past economic policy choices, such as import-substitution industrialization or export oriented industrialization, are important historical elements affecting the current level of openness. The year dummies suggest that major fluctuations of the international business cycle have an impact on trade flows. For example, the negative coefficients on the years covering the late seventies and the early eighties (not reported here) indicate that global recession affected the level of trade in LDCs. Trade flows in LDCs began to pick up again in the late eighties when conditions in the world economy had significantly improved. The other significant determinants of trade were net foreign direct investment (FDI) and GDP. The GDP findings are the most ambiguous. It is commonly known that larger countries tend to memories self-sufficient an smaller ones more open. Thus, we would expect the negative coefficient for GDP in exports measured as a percentage 252 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of LDC exports. On the other hand, countries with large GDPs would have more potential exporters. On the other hand, countries with large GDPs would have more potential exporters, justifying a positive coefficient in column two of Table 1.2. One explanation for these contrasting signs could be a sample bias resulting from the exclusion of countries with large GDPs, such as China. Regardless, these seemingly contradictory results warrant future investigation into the effects of larger GDPs on trade in LDCs. The positive effect of foreign direct investment confirms the proposition that such ventures increase export levels in a nation (Amirahmadi and Wu, 1992). Yet again, the magnitude of this relationship should be kept in proper perspective, since the coefficient suggests that a 5% increase in FDI (which is quite large) effects only a .1 percent increase in trade. The significant effect of the IMF credits for only one of the dependent variables makes it difficult to draw strong conclusions about the role of international financial institutions. After a three year lag, the IMF variable effected export competitiveness (measured as percentage of GDP), suggesting that Fund programs do work. This directly challenges the arguments of theorists such as Bullard and Malhotra (198) who are extremely critical about the effectiveness international finance institutions and their “neo liberal crusades” in developing countries. Lending greater credence to their arguments, however, was the second set of results reported on export competitiveness. Even with lags up to four years, there was no effect of IMF credit on exports competitiveness measured as a share of total LDC exports. 253 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The insignificant results on the democracy variable were particularly surprising at first, given the ongoing debates linking democracy and trade. Yet it is important to realize that the tenets of these theories were formulated on the basis of well-developed democracies and the existence of interest group politics.3 7 Results in Table 6.2 suggests that the effects of low GDP and the high demand for capital flows in LDCs crowd out the effects of democracy. The implication is that poorer countries of the world economy are more susceptible to the pressures of globalization and are thus liberalizing their trade regimes, regardless of regime type. The most important lesson to be drawn from these results is that it is not relevant to link lower welfare spending and international competitiveness. This “gap” leaves much room for other alternatives to be explored in the name of improving efficiency and labor productivity. What stands out most is the seemingly negligible role for “politics” in the political economy of trade (as witnessed by the insignificance of the democracy variable and the low coefficient on welfare). Yet this would be a hasty conclusion to make for two reasons. First, it is highly possible that the strong significance of the country dummies encompasses some of the country specific political differences not necessarily captured by the democracy variable (such as the organizational strength of privileged domestic industries and/or labor unions). Second, recall that reducing welfare spending is a political choice made by LDC governments in 3 7 Banaji and Ghanem(1997) are an exception to this in that they find a positive correlation between democracies and trade in LDCs. However, there are problems with their estimations in that they do not 254 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. response to the pressures of globalization. Seen in this way, governments could just as well have opted to reduce other costs such as subsidies to inefficient industries, or chosen a different strategy altogether, such as currency devaluation (see Pfaller et al., 1991). Regardless, these findings indicate that it is incorrect to associate welfare spending with a precipitous decline in export competitiveness. Welfare spending is therefore not a primary factor affecting international competitiveness and need not be discouraged in the current era of globalization. M odel 2: Domestic Pull vs. Global Push The results in Tables 6.3, 6.4, and 6.5 provide strong evidence against the pull hypothesis and thus the conventional wisdom. This model was first estimated without the interest rate variable in order to disentangle the effects of the year dummies.3 8 Inclusion of the interest rate variable in Table 6.3 provides further and even stronger verification of the push hypothesis. Finally, Table 6.4 demonstrates that welfare spending is also not correlated with productive capital flows. These regression results taken together demonstrate that the effects of welfare spending on productive and financial capital flows are clearly inconsequential. The welfare variable failed to emerge as significant even after lagging it up to four years, and with several different combinations of explanatory variables. take differences over time into account, nor do they control for significant country specific differences (such, as land size). 255 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 6.3: Regression Results for Effects of Welfare Spending on Net Capital Flows Model 2 Dependent variable: net capital flows (excluding interest rate variable) netk (lagged) 0.244*** 0.038 w e lf (lagged 3 years) 0.040 0.047 tra d e (lagged) 0.407** 0.180 gdp (lagged) 0.312*** 0.103 Democ -0.041** 0.016 resv (lagged 2 years) -0.154** 0.062 Country effects Yes* Year effects Yes* R* .951 N 1008 F value 2.65 Fixed effects regression estim ates. ***p<0.01; **.01<p<0.05; *0.05<p<0. to. F test for fixed effects. *p<0.01 It was only after lagging welfare by three years and the effects of the level of foreign reserves by two years, that any significant results were yielded. Put simply, government spending on welfare does not yield a barrage of capital flight as charged by the conventional wisdom. In fact, it has no effect on capital flows. On the contrary, the significance of the year effects (at the 99% confidence level) suggests that international factors (or the push variable) influence the rate of capital flows. The year effects suggests that year-to-year changes in international interest rates are more important that 3 8 In other words, it is necessary to test if the year dummies “swallow up” the effects o f the interest rate differential between developed and developing countries. 256 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the domestic level variables. Annual changes in global interest rates are reliable proxies since they do not differ much from monthly or bi-yearly changes. This was cross checked with the annual LIBOR rates reported in the IMF’s International Finance Statistics which tended to vary only slightly (less than 1 percent) from the weekly and six month interest rates. Thus, although year effects do not directly capture the impact of short term changes in the interest rate, they do indicate how annual changes might affect the direction of capital flows (see Chart 6.5). Chart 6.5: Trends in LIBOR from 1972-1995 18 16 14 12 10 8 6 4 2 0 LIBOR (annual period averages in percentage per annum 1972-1995) : • :'X;- f/z m . •vr u . : • V - & & 'LIBOR Source: G lobal Development Finance (IMF: Cd-rom: 1999) Regression results indicate that capital inflows to LDCs clearly went up during most years that international interest rates were low. In the mid 1970’s, for example, 257 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the positive sign on the year effects (1974 and 1975) illustrates that capital inflows to LDCs increased in those years (not reported here). The year coefficients were negative in the mid to late eighties when international conditions were improving and LIBOR was relatively high. Finally, when interest rates in the developed worid went down again in the early nineties, year coefficients turned up positive. Yet despite the similar pattern between the direction of the year dummies and LIBOR, this relationship must be viewed with caution. After all, even though the impact of interest rate differentials might be a strong component of the year effects, these dummies capture other conditions in the world economy that might also be affecting capital flows (i.e., recession, crisis).3 9 Therefore, another set of regressions were run in order to capture the impact of the interest rate differential more directly. This time, however, the regressions were applied both ways, by including and excluding year effects, in order to avoid any multicollinearity problems. Year effects remained significant even when the interest rate variable was included, suggesting that there are other things going on the world economy (aside from interest rates) that affect capital inflows (i.e., crisis and global recessions). To check for robustness, a second measure of the competition for net capital flows was once applied (Table 6.4). 3 9 Also, there are some inconsistencies between the year effects and LIBOR. For example, in 1982 and 1983 when LIBOR was highest, year effects in LDCs were negative. 258 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 6.4: Regression Results for Effects of Welfare Spending on Net Capital Flows (Including world interest rate) Model 2 Dependent variable: net capital flows (year and interest rate variables) netk (lagged) 0.226*** (0.038) 0.226*** (0.038) welf (lagged 3 years) 0.022 (0.047) 0.022 (0.047) trade (lagged) 0.407** (0.177) 0.407** (0.177) gdp (lagged) 0.260** (0.103) 0.260** (0.103) Democ -0.042** 0.016** (0.016) (0.016) resv (lagged 2 years) -0.159** (0.038) -0.159** (0.061) Intdiff -0.549*** -0.549*** (0.137) (0.137) 1 Country effects Yes* Yes* Year effects Yes* Yes* .953 .524 N 1008 F value 2.79 2.18 Fixed effects regression estimates. ***p<0.0l; **.0l<p<0.05; *0.05<p<0.10. Figures in parenthesis are standard errors. *F test for fixed effects. p<0.01 The overall fit is very impressive in this table. Both indicators of competitiveness revealed almost identical results, lending even greater confidence to the econometric results. Most importantly, the interest rate variable is negative and significant at the 99 percent confidence level. Moreover, the magnitude of the impact of the push factor is quite large relative to the other variables. Interest rates in LDCs 259 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. that are higher than developed countries interest rate by say 2%, will increase capital flows to LDCs by 3.3%! These results are consistent with the works of Maxfield (1998) Reisen (1996), Femandez-Arias(1994) and Doodley et al. (1996). The significance level on all of the other variables, particularly welfare spending, remained unaffected. The remaining domestic level variables such as democracy, GDP, and foreign reserves, however, did have an effect on capital flows. Does this finding challenge the push hypothesis that domestic level variables matter less in the competition for capital? The most obvious answer is that while push factors might be the primary determinant of capital flows to LDCs in the aggregate, certain domestic level differences (e.g., regime type, economic development and foreign reserves) do matter for the distribution of those flows to individual countries. For example, these results challenge Maxfield’s conclusion in World Politics (1998) by suggesting that governments of soft authoritarian regimes are less likely to implement capital controls and attract greater capital flows.4 0 A few words on the negative coefficient of the foreign reserves variable is in order. The negative sign on foreign reserves provides strong empirical support for the work of Haggard and Maxfield (1996) on “The Political Economy of Financial Internationalization.” By taking political evidence into account, they convincingly challenge standard economic interpretations by arguing that a balance of payments crisis eventually leads to financial liberalization and thus greater capital inflows to 4 0 Perry and Robertson (1998) also provide empirical evidence that there is a trade-off between democracy and capital market efficiency. However, they focus their analysis on the developed economies. 260 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LDCs.4 1 It is during a crisis that political conditions are most fertile for the removal of capital controls, since the need to reassure creditors and investors is greater under such circumstances. Haggard and Maxfield contend the bargaining power of liberalization advocates both within the government and the private sector is consequently strengthened during crises. Thus, the inverse relationship between foreign reserves and capital inflows shown in Tables 6.4 and 6.S lend strong empirical confirmation to the Haggard and Maxfield hypothesis. Yet what happens when productive capital flows [FDI] are disaggregated from net capital flows? Does lower welfare spending then become an important pull factor for foreign direct investors? After all, as Schwartz (1998) points out, productive and financial capital have quite distinct effects and differing degrees of mobility. Logically, it would seem that both the social wage and domestic tax structure (a pull factor) would affect the investment decisions of productive capital. The driving force behind productive capital investment then represents a bridge between the ILO-World Bank and the push-pull debate. Yet Table 6.5 shows that welfare spending still does not adversely affect a country’s ability to attract capital flows. 4 1 Recall that this model assumes that balance of payments crisis and low foreign reserves must occur concurrently. 261 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 6.5: Regression Results for Effects of Welfare Spending on FDI Model 3 Dependent variable: net foreign direct investment H i S B fdi (lagged) 0.378*** 0.041 welf (lagged 2 years) -0.009 0.049 trade (lagged) 0.506*** 0.144 gdp (lagged) 0.146* 0.084 Democ 0.032** 0.015 imf (lagged 3 years) 0.014*** 0.004 Country effects Yes* Year effects Yes* R2 .737 N 1008 F value 3.10 Fixed effects regression estimates. ***p<0.0l; **.01<p<0.05; *0.05<p<0.10. F test for fixed effects. * p<0.01 These results are interesting in that they show the pattern of signs for most of the variables, except for democracy, were similar to those found for net capital i* inflows. More relevant for the purposes of this paper, however, is that there is no relationship between welfare spending and productive capital flows.4 3 The estimated coefficients suggest other pull factors besides reduced social spending influenced the level of productive capital flows to LDCs. To summarize, the magnitude of the effect of welfare spending is minimal on LDCs international competitiveness. The evidence in the previous four tables suggest 4 2 This model assumed that IMF credits were a more appropriate determinant of productive capital flows than foreign reserves. 4 3 As in the other regressions, welfare spending was tried in different combinations, lagging it up to four years. 262 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. two basic things about the political economy of trade and capital flows. First, the logic of the conventional wisdom, as it was represented by the World Bank perspective and the pull hypothesis, overall has very little empirical support. Cross-national historical differences and year-to-year variations in international economic conditions matter more for trade and net capital flows to LDCs (although productive capital did react more to the domestic level control variables). Second, these results call into question the dichotomous nature of the pull vs. push debate. It is true that the findings suggest international investors are not “rational” as is commonly assumed, particularly when international interest rates are low. However, while push factors are the primary determinant of capital flows to LDCs as a whole, country specific differences can “pull” in more capital flows relative to other LDCs. Either way, this paper has firmly established that government welfare spending is not a significant pull factor in the competition for either productive or financial capital flows. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 263 Implications The evidence presented in this chapter strongly indicates that globalization variables in LDCs are not responsive to changes in domestic welfare spending. There is a marked disjuncture between those who think welfare spending is good and/or necessary for globalization and those who think it discourages it. Reassessment of the globalization-welfare nexus casts much doubt on the conventional wisdom that greater government spending on social programs creates production inefficiencies and encourages capital flight. Simply put, causality between globalization and welfare is not unidirectional. A fundamental flaw in the logic of the conventional wisdom is revealed when cause and effect are investigated separately. Thus, the dynamic nature of the globalization-welfare relationship must be understood so that firm conclusions can be drawn. Results from the intertemporal models used in this paper emphasize the fact that LDC governments can make the political choice to increase social spending without affecting international competitiveness in the current era of globalization. This analysis accomplishes three things. First, it reveals that conventional theories on globalization and welfare do not withstand more rigorous tests. Although LDCs may be lowering their welfare spending in the era of globalization, they are not subsequently becoming more competitive in global markets. Second, this investigation uncovers some of the complexities behind the globalization-welfare nexus. It is ultimately not economic necessity but the political choice of governments that has affected the outcome of the globalization-welfare relationship in LDCs. This is perhaps because capital generally has more institutional representation in LDCs (through international financial institutions, for instance) than does labor and other marginalized 264 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. groups. Finally, this study exposes the different magnitudes of impact of domestic policy on trade, productive and financial capital flows. It is trade flows that are most affected by government spending on welfare, while capital flows are more highly correlated with international factors. One alternative explanation of these results could be that average LDC welfare spending as a share of GDP is too low to be a source of competitiveness problems. Advocates of the conventional wisdom are wrongly assuming that welfare spending are inevitable casualties in the era of globalization, at least in the case of LDCs. This is all the more reason why LDC governments should be aware of their flexibility in improving the levels of welfare spending. Increasing welfare expenditures may be incorporated with a host of other reforms leading to greater economic stability and, eventually, greater international competitiveness. So what are these “pressures” of globalization to which LDC governments are reacting? Why have LDC governments reduced spending on social programs in response to increased levels of globalization when LDC welfare states, in fact, do not severely affect international competitiveness? One interpretation of the empirical evidence in this paper could be that it is the reinvigorated political faith in the efficacy of markets combined with the underdevelopment of political institutions for labor that has created a de-emphasis on social spending. Put differently, the discourse of neo liberalism has gained momentum in LDCs. Rodrik (1997) is on target when he argued that “competitiveness” is too often used as an excuse for domestic reform. The analysis in this paper suggests that this is particularly true for LDCs. Rodrik (1997:80) states: 265 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Too often [...]the need to resolve fiscal or productivity problems is presented to the electorate as the consequence of global competitive pressures. This not only makes the required policies a harder sell—why should we adjust just for the sake of becoming better competitors against the Koreans or the Mexicans?—it also erodes the domestic support for international trade—if we have to do all these painful things because of trade, maybe trade isn’t such a wonderful thing anyhow! In sum, this study reveals that the relationship between the pressures of globalization and consequent social policy choices is by no means deterministic. These findings provide sufficient justification for scholars and policymakers alike to reassess their claim that welfare state spending is inefficient and erodes a nation’s competitiveness in global markets. A better balance between greater exposure to international markets and addressing domestic social needs in LDCs should be set. Scholarly research has shown that greater social spending protects citizens from the risks and uncertainties of globalization. Consequently, there needs to be a growing realization amongst LDC governments (and international finance institutions) that something more needs to be done with respect to their social programs. It is hoped that the findings in this chapter will help to convince LDC governments that they can afford to implement greater social spending in the current global era. In the end, international economic integration has not made national politics and policies in LDCs irrelevant. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. SECTION 3 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 7 THE GLOBALIZATION- LDC WELFARE STATE NEXUS AND ITS IMPLICATIONS Clearly, globalization has had a drastically different impact upon less developed countries than it has had on OECDs. Section one of this study demonstrated that rising international demand has stimulated the growth of unskilled labor relative to skilled labor in LDCs. This steadily increasing ratio indicates that the bargaining power of labor has been weakening over time, since unskilled laborers are not well-organized in most LDCs. Consequently, labor has not been effective in pressuring governments for greater welfare spending in the face of globalization. This was confirmed in the case of India, while from a different angle, Brazil and Korea proved that increases in skilled labor can elicit greater government response to labors demands. Section two took a closer look at the efficiency hypothesis and assesses if lower welfare spending in LDCs results in greater international competitiveness. Evidence supporting a significant and negative relationship between competitiveness and government welfare spending was relatively weak. The next logical question that emerges, then, is why falling levels of social welfare spending have not resulted in improvements in efficiency. This section consists of three parts and attempts to answer this question by conceptually collating the two- way causal flows of the LDC model. First, it summarizes the findings of sections 268 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. one and two and, second, it applies evidence from the three cases to tie sections one and two together. Finally, this section ends with a discussion of this study’s contribution to the literature, policy implications and future research agenda. Chapter 2 concludes that the following variables have a significant influence on welfare spending in LDCs: (i) the interactive effects of globalization and the bargaining power of labor, (ii) the level of democracy; and (iii) the extent of economic development. The first variable emerged as the primary mechanism intervening between social welfare expenditures and globalization. It has been found that globalization puts pressure on the organizational ability of workers in surplus labor economies and thus weakens their bargaining power. The following hypothesis has been proposed, tested and confirmed: the greater a nation's comparative advantage in (unskilled) labor, then the greater globalization’ s effect on the decline in government welfare expenditures. Comparative case studies in Chapters 3 and 4 supplemented the econometric findings in three ways. First, they confirmed that if a country enjoys a comparative advantage in unskilled labor, then welfare spending decreases with globalization. India presented a classic case of declining welfare expenditures, while welfare spending in Korea and Brazil increased in tandem with their promotion of high skill manufactured exports. Second, case studies emphasized the importance of the transition from authoritarianism to democracy. The cases of Brazil and Korea confirmed that democratic governments were more responsive to labor demands than authoritarian ones. However, when placed in comparative context with the Indian case, it emerged that while democracy might be a necessary condition for greater 269 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. welfare states, it is not a sufficient one. Finally, comparative case studies revealed an important aspect of welfare spending that was completely excluded from the quantitative analysis: the tendency towards inefficient outcomes of current LDC social policies. A review of the historical state-labor relationship in these countries has explained how and why some labor groups tend to gain more than others, despite the (traditional) redistributional goals of a welfare state. These detailed case studies shed greater light on two interesting findings, which emerged from the quantitative section of Chapter 2: the significance of country specific differences,1 and the level of democracy. First, the Indian, Brazilian and Korean cases suggested that the most important country-specific difference relevant to LDC government welfare expenditures is their history of state-labor relations. In India, past ideological commitments to social welfare resulted in state legislation that privileged labor. The consequences of globalization have been the expansion of the informal sector to circumvent rigid labor laws, as well as the gradual dismantling of the labor laws themselves. The rise in unskilled labor (both formal and informal) weakened labor’s organizing potential and resulted in reduced state welfare expenditures. Korea and Brazil, on the other hand, faced a history of repressive labor regimes. The difference was that, in Brazil, the corporatist system used social benefits in exchange for labor cooperation. Brazil, therefore, experienced an expanding welfare state. However, the distribution of benefits was highly unequal. 1 Recall that the fixed effects procedure, which controls for country and year effects, was the econometric method used in the analysis. 270 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In comparison, although Korea’s welfare spending has been increasing over time, the amount it spends on welfare has been significantly below the average of others in the same national income bracket. Historically, Korea has been considered to be a welfare laggard. Regardless, both Korea and Brazil have provided greater insight into the significance of the democracy variable. Even though skilled labor has been increasing and labor, in general, was gaining strength over time, the Brazil and Korea case studies revealed that regime type still matters. The democratic transition in both countries resulted in a distinct shift towards greater government social commitments. Therefore, it appears that authoritarian governments tend to be less responsive to labor interests than democratic ones. Based on the findings from both the quantitative and qualitative sections of Section 1, Figure 7.1 provides an overview of the policy process. While the growth of international demand for unskilled labor has been an important factor in determining the social outcome, this module shows how country specific factors and regime type are additionally influential. It attempts to illustrate how the various phases of the policy process are connected and how they differ from one regime to another. Not only does this module provide further evidence on the proposed hypotheses, but it also helps to trace the decision-making process by which various initial conditions are translated into social policy outcomes. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 7.1: The Social Welfare Policy Process in LDCs* Bogin Compute growth -N o - Qrowtng labor Is labor M usncsd by history of corporatism? I N O _ i ____ -Y os - tnexctwnge ter tabor coopemtfoftf dientettom in employment of unsifted tenor? T Yes I . A m low-sUl saports bails of _ eom pa raO vs advanjiga7 Yes 4 Oosa labor lacs othtr do cnsabc connaim (e g ., mutbpftaiy of unions, tabor imtkaf Mtadbtty, Ke.)? I Yes f Gro wth of In fo rm al and formal unakUad labor ♦ labor martat rtglOBas- tabor movsmanta waaJttnlng Growth of high s M I l asports and growth of s M U a d labor Growing labor m o w a r nar t ls ♦ parvaaw* dlantatlam 1 -► Hsadamooaeyoecurad? Y as- ■Yes No N O I n goasmmsnt aocl al Oscraaamg * This process model follows from methodology of Bennet and Alker (1977). 272 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Finally, it can be argued that Section 2 (Chapter 6) has presented the most significant finding of this study, with respect to policy implications for LDCs. Conventional theorists of globalization should be most intrigued by the empirical evidence that LDC welfare spending does not greatly affect its ability to compete in international markets for capital and/or exports. In essence, the globalization-welfare relationship has not necessarily been a vicious cycle as suggested by the neoliberal wisdom. The fact that the causal relationship between globalization and welfare has been unidirectional suggests that LDCs enjoy considerable governmental autonomy in making social welfare decisions. Thereby, in theory, increased LDC government welfare spending is sustainable in this era of globalization. Yet why was the efficiency hypothesis validated in Section 1, but not in Section 2? In other words, why is the relationship between globalization and welfare not a two-way process as the conventional efficiency literature predicts? This question is best addressed with evidence from the case histories of India, Brazil and Korea. In all cases, it is confirmed that welfare spending has not developed far enough to deflect international investors or hurt export-oriented production. The evidence strongly suggests that factors other than welfare spending per se affect national competitiveness. Three points address the reasons why government welfare spending is not affecting changes in international competitiveness. First, since welfare spending as a percentage of GDP in most LDCs (e.g., Korea and India) is still relatively small, 273 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. international competitiveness is not yet being affected.2 Average LDC labor costs are still very low compared to developed countries. Second, both the Brazilian and Indian case illustrate that welfare spending is only one aspect of the business climate considered by investors. A country’s international competitiveness is affected by other domestic factors. For example, Korea, has improved its competitiveness through other types of government spending such as education. Finally, all three cases confirm that high domestic interest rates of LDCs tend to attract capital investment and thus have a positive effect on competitiveness. Whether LDC social welfare expenditures have been increasing or decreasing as a percentage of GDP, most LDCs maintain very low labor costs. Both the governments of India and Korea spent below the LDC average of 2 percent of GDP on welfare. India’s average spending on welfare is 0.6 percent of GDP and Korea’s average is only 1.1 percent. Not only are such levels of welfare expenditures too low to deter foreign investors, but they also are too low to affect export prices. Brazil, on the other hand, spends 7 percent of its GDP on welfare, which is much above the LDC average. Yet Brazil’s social system, as will be later addressed, tends to be inegalitarian and regressive. India, in particular, maintains some of the lowest wages in the world. Brahmbhatt et al. (1996) state that India’s hourly labor costs for production workers in manufacturing at US $0.27 are in line with those of China at $0.25 per hour.3 This leaves China, where the size of the labor force exceeds that of India, Taiwan, Korea, 2 Recall that labor benefits (Le., unemployment, pensions, etc.) constitute a major portion of welfare in LDCs (see Chapter 6). 274 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hungary, and Thailand combined, as its main competitor (Brahmbhatt et al., 1996). Thus, it is unlikely that labor benefits at their existing levels have contributed to increasing labor costs in India. Ratman (1993) points out that the “number of employees covered by collective agreements on wages and working conditions works out barely to one percent of the total labor force in the country.”4 In addition, Chapter 3 demonstrated that one way India manages to keep its labor costs low is by reaching out to the informal sector. Squire and Suthiwatt-Narueput (1997) argue that labor and business are likely to collude on this arrangement in order to maintain employment and avoid the law. Interestingly, while India’s labor costs might be amongst the lowest in the world, labor productivity has not shown significant improvements (Brahmbhatt et al., 1996; Sodhi, 1993). India has thus not chosen the option of increasing productivity to bolster international competitiveness. The implication is that the Indian government has an incentive to maintain low labor costs as one type of international strategy. The Indian then attempts to prohibit labor benefits from rising to the point in which they would affect competitiveness. The second reason why low welfare expenditures have not affected international competitiveness is because many LDCs (e.g., Brazil and India) are in need of other reforms. Labor market rigidities, for example, continue to exist. Many scholars argue that amendments to India’s existing labor laws have not gone 3 These figures refer specifically to wages in the clothing industry at 1993 hourly wage costs. 4 Ratman’s (1993) reference to labor benefits does not apply to ah of the components of the dependent variable applied in this study. See Table 2.2 for the full definition of social security and welfare. 275 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. far enough to affect positive macroeconomic changes (Gangopadhyay & Wadhwa, 1998; John, 1995; Lai, 1999; Ratnam, 1997). Although bargaining power of labor has been considerably weakened in India, changes in the state-labor relationship in India have come only gradually. The Indian government is moving slowly on the full enforcement of its exit policies, much to the displeasure of domestic employers and international investors. Government planning and control of the economy continue to exist, despite ongoing reforms. Simply put, India’s economy is far from becoming a classical liberal economic system. As Denoon (1998:44) states: “India has had three periods when the government has experimented with reducing controls over trade, finance and industry, yet there is still a bewildering array of central and state regulations on foreign investment and trade in certain sectors.” While many legal changes have been made, such as the Monopolies and Restrictive Trade Practices Act (MRTP) and the Sick Industries Companies Act, and exit policies/labor laws, few have been enacted.3 Chopra et al. (1995:3) best summarize why the Indian domestic situation does not promote greater international competitiveness:6 5 For example, as Denoon (1998) points out, not one of the 150 companies recommended for closing has been forced to close down. 6 Scholars tend to find several explanations for low domestic and foreign business confidence in India’s post-liberalization economy, in addition to reforming its labor laws. Ratnam (1997) attributes extensive bureaucratic delays and poor infrastructure to hindering greater integration. Kumar (1998) suggests that inferior product quality and increasing technological obsolescence explains why promote even greater manufactured exports. Chandra (1997)(-Byres ediction) claims that competition can be intensifed if the quality and standardization of goods improve, the MRTP Commission is more responsive and enforceable when restrictive trade practices are discovered and (3) the system of licensing is thoroughly revamped in low-technology sectors. Denoon (1998) finds problems with the continuing government control. 276 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Although much has been accomplished, the reforms are by no means complete, and persisting obstacles constrain a stronger and broader investment response...[L]ittle has been done to address the serious distortions in the public enterprises, labor markets, and the agricultural sector (Chopra et al., 1995:3). Furthermore, greater government intervention in other types of social spending is more likely to affect competitiveness in LDCs. The Korean experience has shown that a more equitable distribution of income is an important factor in LDCs international competitiveness (p. 1994). Korea is, therefore, an interesting contrast to the average LDC case with respect to growth, competitiveness, and equity. Unlike Brazil and India, the principle of “shared growth” has accompanied its rapid economic development and helped improve its competitiveness. A large part of its success story is attributed to the Korean government’s educational investments (p. 1994). Universal primary education and widely available secondary education have contributed to rising endowments of human capital. Amsden (1994) concludes that the Korean government opted to increase productivity rather than reduce labor costs in an effort to gain international competitiveness.7 Therefore, although social welfare has been at remarkably low levels in Korea, its government has invested in other types of social spending (i.e., education) that have improved competitiveness. Brazil is a good example of an LDC that stands to benefit from such government attempts at improving income distribution. The regressive nature of 1 0 f course Amsden (1994) acknowledges that the East Asian miracle was also based on other macroeconomic factors (ue., high savings and investment ratios, export promotion, etc). However, she attributes government policies of subsidizing learning (rather than cutting real wages) to the rapid increases in competitiveness of the East Asian economies. 277 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Brazil’s social welfare system is indicative of the broader phenomenon of inequality that characterizes Brazil’s domestic economy. As evidenced by its Gini coefficients, income distribution in Brazil is extremely skewed (see Charts 7.1-7.4). Brazil’s Gini coefficients are almost twice as high as those of India and Korea. Inequality in India and Brazil, after the 1990s, went back up to a level higher than it had in the early seventies. By contrast, Korea shows a constant decline in inequality after 1980. Brazil’s struggle to reduce its income inequality has been an uphill battle (see Weyland 1996, 1997). The fact that its labor aristocracy has been stalwart in efforts to block the reform of the social welfare system is strong evidence supporting this claim. Not surprisingly, many claim that inequality has been a constraint on economic growth in Brazil. It is often argued that the inadequacies of Brazil’s educational system underlie its burgeoning inequality. Rezende (1998) claims that Brazil’s poor education system is due to insufficient capacity in poorer areas and high inefficiency in the state education sector. Brazil's schooling average of S.S years of schooling is well below the average in other Latin American countries, such as 7.5 years in Argentina, Uruguay and Paraguay (Rezende, 1998). Using Korea as a point of reference, Brazil’s international competitiveness would stand to gain if government resources were directed towards greater access to more and better education and away from its inefficient social welfare system. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Charts 7.1,7.2,7.3: Inequality in India, Brazil and Korea In d ia: T ren d s In Inequality a 33.00 S 32.00 2 31.00 I 30.00 U 29.00 1 28.00 ® 27.00 Brazil: Trends in Inequality 64.00 62.00 g 60.00 | 58.00 g 56.00 2 54.00 5 52.00 50.00 Korea: Trends in Inequality 40.00 | 38.00 36.00 34.00 S 32.00 30.00 1971 1978 1980 1982 1985 1988 Source: Deninger and Squire World Bank data set (1999) 3 i> x I Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Lastly, Chart 7.4 illustrates that the pull hypothesis presented in Chapter 6 has merit in the cases of India and Brazil (data on Korea were not available). Domestic interest rates in both these countries tended to be higher than the average international interest rate represented by LIBOR. Of the years listed below, Brazil’s average interest rates were consistently higher than LIBOR. India’s average interest rates fell below LIBOR only in 1983 and 1995. Not surprisingly, both Brazil and India experienced rising capital flows as a share of GDP from the eighties onwards (see Chapters 3 and 4). Therefore, Chart 7.4 boosts confidence in the Chapter 6 econometric results which found that competition for capital flows amongst LDCs is based on international conditions, not just LDC domestic factors. Chart 7.4: Domestic vs. International Interest rates in Brazil and UBOR vs average interest rates in Brazil and India 181 18 | 1 4 I r 1 1 0 I 8 I 6 4 I 2 < > J ■U bor ■Brazil Hhdla ! : i a v i :! ;; U ! 1 3 1 1 j ;i j i J i y j y 1972 1976 1979 1963 1986 1969 1992 1994 1995 India Source: World Development Indicators (1997) 280 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In sum, while the bargaining power of labor in LDCs has been falling and governments have been spending less on social welfare (relative to GDP), this policy decision has not helped to improve competitiveness. Up until now, welfare spending has not succeeding in affecting labor costs to a point that is either internationally advantageous or disadvantageous for LDCs. Rather, comparative evidence from India, Korea and Brazil suggested that other reforms are needed to positively affect competitiveness. Korea, in particular, was an important case which showed that LDC government social spending (e.g., education) can ultimately improve competitiveness. Thus, Section 3 yields the very important conclusion that it is neither necessary nor productive for labor to bear the brunt of the competitive pressures of globalization. Contribution to Literature This study has demonstrated that citizens of all nations are not protected from the risks and uncertainties of globalization. By establishing that welfare states in LDCs have not been as resilient to the pressures of globalization as OECDs, this research challenges the conclusions of existing studies on globalization. The empirical and comparative evidence has verified that global market forces are interacting with labor’s bargaining power and obviating the development of strong national social welfare policies in LDCs. Section one confirms that contrary to OECDs, the efficiency hypothesis prevails in LDCs. Yet Section two demonstrates an interesting paradox. While globalization has been negatively affecting social 281 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. policy decisions in LDCs, these social policy decisions are having no effect on the country’s level of competitiveness and, subsequently, their level of globalization. Section three illustrates why it is not surprising that the econometric results in Section 2 did not find a negative and significant coefficient for welfare spending in LDCs. This analysis has revealed that contrary to conventional wisdom, welfare spending is a government social policy choice that ultimately does not impact LDC’s ability to compete on international markets. LDC governments can, in fact, respond to international market integration by increasing government spending on programs that reduce market generated inequalities. This, in turn, can positively affect international competitiveness as demonstrated by the case of Korea. Significantly, then, constraints on government social spending have been political, not economic. This research plan has been designed to push the boundaries of existing conceptual and empirical work on this topic. The conceptual work was aimed to consolidate principal theoretical works on the political economy of trade and financial liberalization, comparative social policy, and state-labor politics. The empirical work was drawn from this literature and built a cross-national database on LDCs that incorporates a set of political, institutional and economic variables. Taken together, a new theoretical baseline from which a more complete understanding of globalization and its impact on domestic social policies has been set. Using research on the mature welfare states of the advanced industrial nations as a reference point for analyzing LDCs, the most important features relevant for the 282 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. impact of globalization on national social policy have been determined. On the empirical front, two databases have been essential for this analysis. The main database has incorporated 53 LDCs over a time span of 23 years (1972-1995), and 17 relevant political, institutional, and economic variables. The second database has paralleled the first, with the same time frame and used 14 variables for 14 OECD nations. Combined cross-country, time series data and the fixed-effects method have been applied to assess the validity of the different hypotheses. To summarize, this study has revealed three important facts about the globalization-welfare nexus in LDCs. First, globalization does indeed have a negative effect on LDC welfare states. The forces of globalization, due to the growth in demand for unskilled labor, further weaken labor’s bargaining power in LDCs. Second, comparative studies have shown, additionally, that domestic institutional constraints inhibit the distribution of higher and more efficient welfare spending. Finally, although it is difficult for labor to make social policy demands because of their greater immobilization, governments can still afford to increase spending on social programs. International economic integration has not made the expansion of the welfare state unfeasible in LDCs. Policy Implications and Future Research Agenda The broader objective of this research project is to improve LDC social policies and programs. The first step to this process was to understand how and why globalization affects national social policy, particularly in the case of LDCs. Government assisted welfare programs ensure a balance between domestic social 283 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and political stability and the uncertainties of globalization. This analysis convincingly revealed that stronger bargaining power for labor is needed to mediate between globalization and social policy. Consequently, the policy implications of this study are straightforward: (a) improve the skill level of the workforce; (b) encourage more democratic governments; and (c) improve state-labor relations. Chapter 6 has exposed a problem with these policy recommendations, however. The implications of the econometric results in this chapter were that LDC governments are not faced with sufficient incentives to increase social welfare spending. Government social spending has not harmed international competitiveness, but neither has it promoted it. It is rational to expect that government incentives to increase welfare spending would be encouraged if welfare spending had a positive effect on competitiveness.8 Based on the conclusions of this study, then, it cannot be certain that governments will be inclined to develop strategies to overcome some of the impediments to strong social policies in LDCs. Do LDC governments have sufficient incentives to invest in social welfare programs? Consequently, this study has yet to be completed. Considerable progress has been made on the first objective of this research project, or the understanding of how LDC’s institutional arrangements interact with international market forces to shape national welfare policies. An analytic framework that incorporates appropriate concepts and questions for evaluating LDC social policies in the contemporary era of * This is based on the assumption that governments must confront the issue of international competitiveness in the era of globalization. This assumption would hold even if governments pursue the twin objectives o f social and economic development 284 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. globalization has been established. Nevertheless, the overarching goal of developing strategies to overcome some of the impediments to strong social policies in LDCs has yet to be met. Further advancement must be made on both the conceptual and empirical tracks in order to fulfill this larger purpose. Goals fo r Future Research The second phase of this research project should begin by moving beyond the traditional concept of welfare (i.e., income transfers and welfare services) applied in the dissertation, and thus seek a fuller understanding of international effects on national social policy. Examining a broader concept of welfare in LDCs, one that is more germane to their socio-politico-economic situation, will ultimately provide a more useful premise for policy prescriptions. Specifically, it is expected that further research should suggest different ways to conceptualize welfare. The conception of welfare applied in the dissertation focused upon income transfers and welfare services. This was interesting in that it conformed to the definition of welfare responsibilities in advanced industrial states and thus provided a basis for direct comparison. Many scholars argue, however, that the Western notion of the welfare state is inappropriate for LDCs (Ahmad et al., 1991; Mesa Lago, 1994; Schmidt, 1995). This research would thus benefit greatly from applying a broader notion of welfare to LDCs. For example, spending on education and training might be a more reasonable definition of welfare in LDCs. Indeed, the implication of the hypothesis tested in the dissertation was that as developing countries face greater pressures to attract capital flows and increase low-wage labor exports, governments have greater incentives to 285 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. keep labor costs down and spend less on welfare. The following research project should draw upon these results and test whether or not in labor abundant economies, spending on other variables (such as skill development) might be more relevant to welfare in LDC. To push this line of logic, it would be valuable to test the affects of globalization on government spending on education, training and/or health. The hypothesis to be tested is that i f a country has a comparative advantage in unskilled labor, welfare expenditures w ill be low at first, but they can be expected to rise over time as the workforce accumulates skills with governmental assistance. Specifically, this will require an expansion of the database to include additional institutional variables, as well as more detailed comparisons of social policy across regions and across countries within regions. Only then, broad principles guiding designs for stronger institutions of social policy in LDCs can be developed. If the original results (or the negative relationship between globalization and welfare) no longer hold when the category for welfare expenditures is expanded, then the policy implications for LDC governments would become obvious: accumulate human capital. Recall that welfare spending helps create a balance between social stability and economic progress (see Garrett, 1998a). Empirical evidence confirming the proposed hypothesis would then suggest that governments have greater incentives to spend on education and training in the face of globalization. The overall research objective, to present viable policy prescriptions in LDCs that will simultaneously improve the processes of economic and social development, will then be better served. 286 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. BIBLIOGRAPHY Agenor, PR. (June 1996). “The Labor Market and Economic Adjustment.” IM F S ta ff Papers, 43(2):261-325. Agosin, M, & French-Davis, R. (1996). M anaging Capital Inflows in Latin America No. 8. NY: United Nations Development Program. Ahluwalia, I. (1997). “Contribution of Planning to Indian Industrialization.” In Terence Byres (Ed.) The State, Development Planning and Liberalization in India Delhi: Oxford University Press: 254-297. Ahmad, E, & Hussain, A. (1991). “Social Security in China.” In Social Security in Developing Countries. Ahmad, E. et at. (eds). Oxford: Oxford University Press. Ajgaonkar, D.B. and M. Talukdar and Vasant Gupte (1998) Profile o f Textile Industry in India Mubai: Mariben Kara Institute. Alesina, A, & Perotti, R. (August 1996). “Fiscal Adjustments in OECD Countries: Composition and Macroeconomic Effects.” NBER No.5730. Alesina, A, & Perotti, R. (December 1997). “The Welfare State and Competitiveness.” American Economic Review, 87(5): 921-939. Amadeo, Edward and Ricardo Paes e Barros and Jose Camargo and Rosane Mendonca (1995) "Brazil” in Reforming the Labor Market in a Liberalized Economy. Baltimore: John Hopkins University Press:35-78. Amirahmadi, H., & Wu W. (1992). Private Capital Flows and Developing Countries. Working Paper #36. Rutgers: Center for Urban Policy Research. Amsden, A, & Van Der Hoeven, R. (April 1996). “Manufacturing Output, Employment and real Wages in the 1980’s: Labour’s Loss until the Century’s End.” The Journal o f Development Studies, 32 (4): 506-530. Andrews, D. (1994) “Capital Mobility and State Autonomy: Toward A Structural Theory of International Monetary Relations. International Studies Quarterly, 38:193-218. Andrews, D, & Willett, W. (Summer 1997) “Financial Interdependence and the State: International Monetary Relations at Century’s End.” International Organization, 51(3): 479-511. 287 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Annex (1997). “Globalization in Historical Perspective.” IMF: World Economic Outlook, May, 112-116. Appasamy, Paul and S. Guhan and R. Hema, and M. Majumdar and A. Vaidyanathan (1996) Social Exclusion from a Welfare Rights Perspective in India. Geneva: International Labour Organization. Arenas de Mesa, A, & Bertranou, F. (1997). “Learning From Social Security Reforms: Two Different Cases, Chile and Argentina.” World Development, 25 (3): 392-348. Arrighi, G., & Silver, B. (1999). Chaos and Governance in the Modem World System. Minneapolis: University of Minnesota Press. Asian Development Bank (1996). Asian Development Outlook. Oxford: Oxford University Press. Atkinson, AB, & Hills, J. (1991). "Social Security in Developed Countries: Are There Lessons for Developed Countries?” In Social Security in Developing Countries. Ahmad et al. (Eds). Oxford: Clarendon Press, 81-111. Baer, W. (1989). The Brazilian Economy. New York: Praeger. Baer, W., & Hargis, K. (1997). “Forms of External Capital and Economic Development in Latin America: 1820-1997.” World Development, 25(11): 1805-1820. Balasubramanyam, VN. (1984). The Economy o f India. Boulder: Westview Press. Baneiji, A., & Ghanem, H. (1997). “Does the Political Regime Matter for Trade and Labor Market Policies?” World Bank Economic Review, 11(1): 171-193. Barioch, P. (1993). Economics and World History. Chicago: University of Chicago Press. Barreto de Oliveira, Fancisco and Kaizo Iwakami Beitrao, and Andre Cezar Medici (1994) “Financing Social Security in Brazil” in Barreto de Oliveira, Fancisco(Ed). Social Security Systems in Latin America Washington D.C.: Inter-American Development Bank: 59-100. Barreto de Oliveira^ancisco (1994) “Social Security in Argentina, Brazil, Chile, and Venezuela” in Barreto de Oliveira, Fancisco(Ed). Social Security Systems in Latin America Washington D.C.: Inter-American Development Bank. 288 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Barua, Nayan (1995) Social Security and Labour Welfare in India New Delhi: Ashish Publishing House. Bennet, James and Hayward Alker, Jr (1977) “When National Security Policies Bred Collective Insecurity.” In Edutsche, Karl et. al. (eds) Problems o f World Modeling. Cambridge: Ballinger. Bertolini, L, & Drazen, A (1997). “When Liberal Policies Reflect External Shocks, What Do We Learn?” Journal o f International Economics, 42:249-273. Bhattacharya, R. (1997). “Pace, Sequencing, and Credibility of Structural Reforms.” World Development, 25(7): 1045-1061. Bleaney, M. (1993). “Liberalization and the Terms of Trade of Developing Countries: A Cause of Concern?” World Economy, 16 (4)July: 453-466. Borchert, J. (Winter 1996). “Welfare State Retrenchment: Playing the National Card.” Critical Review, 10(1): 63-94. Bose, AB. (1970), Social Welfare Planning in India. New Delhi: Planning Commission. Bradford, C. (1990). “Policy Interventions and Markets.” In Manufacturing Miracles. Gary Gerefifl and Donald Wyman (Eds). New Jersey: Princeton University Press, 32-51. Brahmbhatt, Milan and T.G. Srinivasan and Kim Murrell (1996) India in the Global Economy. Policy Research Working Paper no, 1681. World Bank. Bronstein, Arturo (1997) “Labour Law Reform in Latin America: Between State Protection and Flexibility.” International Labour Review 136 (I) Spring: 5- 26. Brown, M. (1988). Remaking the Welfare State. M. Brown (Ed). Philadelphia: Temple University Press. Byres, Terence (1999) The State, Development Planning and Liberalization in India: Delhi: Oxford University Press. Calvo, G., Leiderman, L. & Reinhart, C. (1994) The Capital Inflows Problem. Occasional Papers (56). San Francisco: International Center for Economic Growth. Cameron, D. (1978). “The Expansion of the Public Sector. A Comparative Analysis.” American Political Science Review, 72(4): 1243-1261. 289 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Cardosa, E., & Goldfajn, I. (1998) “Capital Flows to Brazil: The Endogeneity of Capital Controls.” International Monetary Fund S ta ff Papers, 45(1): 161- 201. Cardoso, FH, & Faletto, E. (1979). Dependency and Development in Latin America. Berkeley : University of California Press. Cameiro, Francisco and Andrew Henley (1998) “Wage Determination in Brazil: The Growth of Union Bargaining Power and Informal Employment.” Journal o f Development Studies 34 (4) April: 117-138. Cates, J. (1983). Insuring Inequality: Administrative Leadership in Social Security, 1935-1954. Ann Arbor University of Michigan Press. Cemy, P. (Autumn 1995). “Globalization and the Changing Logic of Collective Action.” International Organization, 49(4): 595-625. Chandra, NK. (1997). “Planning and Foreign Investment in Indian Manufacturing.” In Terence Byres (Ed) The State, Development Planning and Liberalization in India Delhi: Oxford University Press, 369-419. Chaudhry, KA. (October 1994). ‘ ‘ Economic Liberalization and The Lineages of the Rentier State.” Comparative Politics, 1-25. Chibber, Vivek (1999) “ Building a Developmental State: The Korean Case Reconsidered.” P olitics & Society 27 (3) September. 309-246. Choi, K, & Kwon, S. (1997). “Social Welfare and Distribution Policies.” In The Korean Economy 1945-1995: Performance and Vision fo r the 21st Century. Dong-Se Cha, Kwang Suk Kim, Dwight Perkins (Eds). Korea: Korea Development Institute. Chopra, A, Collyns, C, Hemming, R., Parker, K., Chu, W. & Fratzscher, 0. (1995). India: Economic Reform and Growth. Washington D.C.: International Monetary Fund. Claessens, S, & Naude, D. (1993). Recent Estimates o f Capital Flight. World Bank Working Paper. Clark, WR, Rachert, UN, Lomas, SL, & Parket, K. (Winterl998). “International and Domestic Constraints On Political Business Cycles in OECD Economies.” International Organization, 52(1): 87-120. Clayton, R, & Jonas Pontusson, J. (October 1998). “Welfare State Retrenchment Revisited.” World Politics, 51: 67-98. 290 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Cohen, Youssef (1982) “The Benevolent Leviathon: Political Consciousness among Urban Workers under State Corporatism. The American Political Science Review. 76 (1) March: 46-59. Columbia Journal o f World Business (Summer 1994). Editor’s introduction. Erdman, P.B., and M. Brandmeyer (Eds). Crisp, B. (1994). “Limitations to Democracy in Developing Capitalist Societies: The Case of Venezuela.” World Development, 22(10): 1491-1509. Culpeper, R. (1998). “Private Markets and Social Equity in a Post Aid World.” Transforming Development. Das, Arvind (1994) Brazil and India: Comparing Notes on Liberalization. Economic and Political Weekly. 29 (8): 416-419. Dean, J. (1995). “From Protectionism to Free Trade Fever? Recent Reforms in Developing Countries.” Open Economies Review, 6: 369-385. Denoon, David (1998) “Cycles in Indian Economic Liberalization, 1966-1996.” Comparative Politics. October: 43-60. Destler, IM. (1995). American Trade Politics. Washington D.C. International Institute for Economics. Deyo, F. (1989). Beneath the Miracle. Berkeley: University of California Press. Deyo, F. (1984). “Export Manufacturing and Labor The Asian Case.” In Labor in the Capitalist World Economy. Bergquist, Charles (Ed). Vol. 7 in Political Economy of World Systems Annuals. New Delhi and London: Sage Publications. Dhyani, S. (1977). International Labor Organization and India: In Pursuit o f Social Justice. New Delhi: National. Dooley, M, Femandez-Arias, E & Kletzer, K. (January 1996). “Is the Debt Crisis History? Recent Private Capital Inflows to Developing Countries.” World Bank Economic Review, 10(1): 27-49. Dos Santos, T. (1969). “The Crisis ofDevelopment Theory and the Problem of Dependence in Latin America.” Siglo 1. Drake, Paul (1996) Labor Movements and Dictatorships: The Southern Cone in Comparative Perspective. Baltimore: John Hopkins University Press. 291 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Dreze, J, & Sen, A. (1991). “Public Action for Social Security. Foundations and Strategy.” In Ahmad, Ehtisham, Jean Dreze, Amartya Sen and John Hills eds. Social Security in Developing Countries. Oxfoxd: Clarendon Press, 3- 40. Dunning, J. (1997). “The Advent of Alliance Capitalism.” In The New Globalism and Developing Countries by John Dunning and Khalil Hamdani (Eds). (New York: UN University Press, 12-50. Dutt, A. (1992). “The NICs, Global Accumulation, and Uneven Development: Implications of a Simple Three Region Model.” World Development, 20(8): 1159-1171. Echeverri-Gent, John (1993) The State and the Poor: Public Policy and Political Development in India and the United States. Berkeley: University of California Press. Edwards, S. (September 1993). Openness, Trade Liberalization and Growth in Developing Countries. Journal o f Economic Literature, XXXI: 1358-1393. Eichengreen, B. (1996). Globalizing Capital. New Jersey: Princeton University Press. Eichengreen, B. (November 1997). “The Tyranny of Financial Markets.” Current History, 96(613): 377-382. Encarnation, Dennis (1989,) Dislodging M ultinationals: India's Strategy in Comparative Perspective. Ithaca: Cornell University Press. Erdman, PB, & Brandmeyer, M. (Summer 1994). Editor’s introduction. Columbia Journal o f World Business. Esping-Andersen, G. (1996). Welfare States in Transition. London: Sage Publications. Esping-Andersen, G. (1990). The Three World c f Welfare Capitalism. New Jersey: Princeton University Press. Estelle, J. (1997). “Income Security for Old Age." World Bank Working Paper. No. WPS 977.WorId Bank: Washington D.C. Evan, G. (1998). “Britain and Europe: Separate Worlds of Welfare?” Government and Opposition, 33(2): 183-198. Evans, P. (1979). Dependent Development: The Alliance o f Multinational, State and Local Capital m Brazil. Princeton: Princeton University Press. 292 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Evans, P. (October 1997). “The Eclipse of the State? Reflections on Stateness in an Era of Globalization.” World Politics, 50(1): 62-87. Ezeala-Harrison, F. (1999). Theory and Policy o f International Competitiveness. Connecticut: Praeger Publishers. Faruquee, H. (September 1992). “Dynamic Capital Mobility in Pacific Basin Developing Countries.” IMF S ta ff Papers, 39(2): 706-717. Fernandez-Arias, E. (1996). “The New Wave of Private Capital Inflows: Push or Pull?” In Journal o f Development Economics, 48: 389-418. Fierlbeck, K. (1994). “Economic Liberalization as a Prologue to Democracy: Case of Indonesia.” Canada Journal o f Development Studies, XV(2): 151-169. Freeman, R. (1993). “Labor Market Institutions and Policies: Help or Hindrance to Economic Development?” International Bank fo r Reconstruction and Development, 117-144. French, JD. (1998). “Drowning in Laws But Starving (for Justice?): Brazilian Labor Law and the Workers’ Quest to Realize the Imaginary.” Political Power and Social Theory, 12:181-218. Frieden, J. (1991). “Invested Interests: The Politics of National Economic Policies in a World of Global Finance. International Organization, 45(4): 425-451. Frieden, J, & Rogowski, R. (1996). “The Impact of the International Economy on National Policies: An Analytical Overview.” In Internationalization and Domestic Politics. Robert Keohane and Helen Milner (Eds). New York: Cambridge University Press. Gangopadhyay, Shubhashis and Wilima Wadhwa (1998) “Economic Reforms and Labour.” Economic and Political Weekly. May 30: L-40-L48. Garrett, G. (1998a). Partisan Politics in the Global Economy. Cambridge: Cambridge University Press. Garrett, G. (1998b) International Risk and Social Insurance: Reassessing the Globalization-Welfare State Nexus. Unpublished. Paper prepared for APSR, August-September 1996. Garrett, G. (Fall 1998c) “Global Markets and National Politics: Collision Course or Virtuous Cycle?” Paper prepared for International Organization 50th anniversary issue. 293 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Garrett, G, & Lange, P. (Autumn 1995). “Internationalization, Institutions, and Political Change.” International Organization, 49(4): 627-655. Garrett, G, & Lange, P. (Autumn 1991). “Political Responses to Interdependence: What’s “Left” For the Left?” International Organization, 45(4): 539-564. George, A. (1979). “Case Studies and Theory Development: The Method of Structured, Focused Comparison.” In Paul Gordon Lauren, ed., Diplomacy: New Approaches in History, Theory and P olity. New York: Free Press, 43- 68. Ghosh, A, & Gray, J. (1995). “The Current Account in Developing Countries: A Perspective from the Consumption- Smoothing Approach.” The World Bank Economic Review, 9(2): 305-333. Gill, S. (1995). “Globalization, Market Civilization, and Disciplinary Neoliberalism.” M illenium: Journal o f International Studies, 24(3): 399- 423. Goodman, R, & Peng, I. (1996). “The East Asian Welfare States: Peripatetic Learning, Adaptive Change and National Building.” In Welfare States in Transition. Gosta Esping-Andersen (Ed). London: Sage Publications. Graham, C. (1994). Safety Nets, Politics and the Poor. Washington D.C.: The Brookings Institution. Greenaway, D, Morgan, W & Wright, P. (1997). “Trade Liberalization and Growth in Developing Countries: Some New Evidence.” In World Development, 25(11): 1885-1892. Grieder, W. (1998). One World, Ready or Not. New York: Simon and Schuster. Griffith-Jones, S., & Stallings, B. (1995). “New Global Financial Trends: Implications for Developing Countries.” In Global Change, Regional Response. Eds. Barbara Stallings. New York: Cambridge University Press, 143-173. Grunberg, I. (1998). “Double Jeopardy: Globalization, Liberalization and the Fiscal Squeeze.” World Development, 4(26): 591-605. Gupte, Vasant. (1999) Interview. Maniben Kara Institute Director. Gurr, TR, & Jaggar, K. (1994). Polity HI: Regime Type and Political Authority. Ann Arbor: Inter-university Consortium for Political and Social Research [distributor]. 294 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hadenius, A. (1992). Democracy and Development. Cambridge University Press. Haggard, S. & Kaufman, R. (April 1997). The Political Economy of Democratic Transitions. Comparative Politics, 29(3): 263-283. Haggard, S , & Maxfield, S (1996). “The Political Economy of Finanacial Internationalization in the Developing World.” International Organization 50 (1): 35-68. Haggarty, L, & Shirley, M. (1997). “A New Data Base on State Owned Enterprises.” World Bank Economic Review, 11(3): 491-513. Hallen, G. (1967). Dynamics o f Social Security. Delhi: Shiksha Press. Haque, IU, Bell, M, Dahlman, C, Lall, S, & Pavitt, K. (1995). Trade, Technology, and International Competitiveness. World Bank: Washington D.C. Harrison, A. (1996). “Openness and Growth: ATime-Series, Cross-country Analysis for Developing Countries.” Journal o f Development Economics, 48: 419-447. Harriss-White, Barbara (1999) “ State, Market, Collective and Household Action in India’s Social Sector.” in Illfare in India: Essays on India’ s Social Sector in Honour o f S. Guhan. Barbara Harriss-White and S. Subramanian (eds). New Delhi: Sage Publications: 303-328. Hart, J. (1992). Rival Capitalists. Ithaca: Cornell University Press. Havrylyshyn, O. (1987). Exports o f Developing Countries. Washington D.C.: The World Bank. Heredia, B. (November 1997). “Prosper or Perish?” Current History, 96 (613). Hicks, A, & Swank, D. (September 1992). “Politics, Institutions, and Welfare Spending in Industrialized Democracies, 1960-1982.” American Political Science Review, 86(3): 658-674. Hobsbawm, E J. (1996). “The Future of the State.” Development and Change, 27: 267-278. Holando Barbosa, Fernando (1998) “Economic Development: the Brazilian Experience” in Development Strategies in East Asia and Latin America. Hosono, Akio and Neantro Saavedra-Rivano (Eds). New York: St. Martin’s Press: 69-87. 295 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Holzmann, R. (June 1977). “Pension Reform, Financial Market Development, and Economic Growth: Preliminary Evidence from Chile.” IM F S ta ff Papers, 44 (2): 149-177. Hsiao, C. (1986). Analysis o f Panel Data. New York: Cambridge University Press. Huber, Evelyne (1999) Globalization and Social Policy Developments in Latin America Paper presented for conference on Globalization and the Future of Welfare States: Interregional Comparisons. Brown University. October 22- 23. Huber, E. (1996). “Options for Social Policy in Latin America: Neoliberal vs. Social Democratic Models.” In Welfare States in Transition. Gosta Esping- Andersen (Ed). (London: Sage Publications), 141-191. Human Development Report, 1992 (New York: Oxford University Press, 1990). Hurrell, A, & Woods, N. (1995). “Globalization and Inequality. ’ ’Millenium: Journal o f International Studies, 24(3): 447-470. IMF (1986). A Manual on Government Finance Statistics. International Monetary Fund. IMF (May 1997). “Globalization and the Opportunities for Developing Countries.” World Economic Outlook, 72-92. Immerfall, S. (1998). “ The State of the European Welfare State: Three Cheers for the Defendant.” In Challenges fo r a Global Welfare System. Peter Herrmann (Ed.) New York: Nova Science Publishers. Ingerson, A. (1984). “The Textile Industry and Working Class Culture.” In Labor in the Capitalist World Economy. Bergquist, Charles (Ed). Vol. 7 in Political Economy of World Systems Annuals. New Delhi and London: Sage Publications. Jagannadham, V. (1954) Social Insurance in India Amsterdam: Uitgeverij Djambatan. John, C.K. (1995) “Work Ethos and Globalization of Indian Economy.” Indian Journal o f Industrial Relations 30 (3) January: 297-307. Julius, D. (1997). “Globalization and Stakeholder Conflicts: A Corporative Perspective.” International Affairs, 73(3): 453-468. 296 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Keck, Margaret (1992) The Worker’ s Party and Democratization in Brazil. New Haven: Yale University Press. Keller, W, & Pauly, L. (November 1997). “Globalization at Bay.” Current History, 96 (613): 370-376. Kenworthy, L. (March 1999). “Do Social Welfare Policies Reduce Poverty? A Cross-National Assesment.” Social Forces, 1119-39. Keohane, R, & Milner, M. (1996). Internationalization and Domestic Politics. New York: Cambridge University Press. Kim, EM (1997). B ig Business, Strong State. Albany: State University of New York Press. Kim, KS, & Kim, JK. (1997). “Korean Economic Development: An Overview.” In The Korean Economy 1945-1995: Performance and Vision fo r the 21* Century. Dong-Se Cha, Kwang Suk Kim, Dwight Perkins (Eds). Korea: Korea Development Institute. Kim, S, & Lee, JH. (1997). “Industrial Relations and Human Resource Development.” In The Korean Economy 1945-1995: Performance and Vision fo r the 21s t Century. Dong-Se Cha, Kwang Suk Kim, Dwight Perkins (Eds). Korea: Korea Development Institute. Kohli, AS. and S.R. Sharma (1996a) Encyclopaedia o f Social Welfare and Administration. Vol 1. New Delhi: Anmol Publication PVT LTD. Kohli, AS. and S.R. Sharma (1996b,) Encyclopaedia o f Social Welfare and Administration. Vol 2. New Delhi: Anmol Publication PVT LTD. Kohli, AS. and S.R. Sharma (1996c) Encyclopaedia o f Social Welfare and Administration. Vol 4. New Delhi: Anmol Publication PVT LTD. Kothari, R. (1997). “Globalization: A World Adrift.” Alternatives, 22:227-267. Krueger, A (1997). “Korean Industry and Trade over Fifty Years.” In The Korean Economy 1945-1995: Performance and Vision fo r the 21st Century. Dong-Se Cha, Kwang Suk Kim, Dwight Perkins (Eds). Korea: Korea Development Institute. Krugman,P. (1994). “Competitiveness: A Dangerous Obsession.” Foreign Affairs, 73(2): 28-44. Kuczynski, P-P. (Summer 1994). “Why Emerging Markets?” The Colombia Journal o f World Business, 8-13. 297 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Kumar, Nagesh (1998) “Liberalization and Changing Patterns of Foreign Direct Investments.” Economic and Political Weekly. May 30: 1321-1329. Kuruvilla, Sarosh (1996) “Linkages between Industrialization Strategies and Industrial Relations/Human Resource Policies: Singapore, Malaysia, The Philippines, and India.” Industrial and Labor Relations Review. July 49 (4): 635-658. Kuznets, Paul (1984) “Government and Economic Strategy in Contemporary South Korea. Pacific Affairs. October.: 44-67. Kwon, Huck-Ju (1999) The Welfare State in Korea. New York: St. Martin’s Press. Kwon, Huck-Ju (1998a) “ Democracy and the Politics of Social Welfare: A Comparative Analysis of Welfare Systems in East Asia,” in The East Asian Welfare Model. (Eds) Goodman, Roger and Grodon White and Huck-Ju Kwon. NY: Routledge. Kwon, Huck-Ju (1998b) “ The South Korean National Pension Prgoram: Fulfilling its Promise?,” in The East Asian Welfare Model. (Eds) Goodman, Roger and Grodon White and Huck-Ju Kwon. NY: Routledge. Kwon, Huck-Ju (1997) “Beyond European Welfare Regimes: Comparative Perspectives on East Asian Welfare Systems.” Journal o f Social Policy 26 (4): 467-484. Lai, Deepak (1999) Unfinished Business: India in the World Economy. Oxford: Oxford University Press. Lansing, Paul and S. Kuruvilla (1987) “Industrial Dispute Resolution in India in Theory and Practice.” Loyola o f Los Angeles International and Comparative Law journal 9 (2): 345-375. Leblang, D. (1997). “Domestic and Systemic Determinants of Capital Controls in the Developed and Developing World.” International Studies Quarterly, 41:435- 454. Lee, CH. (1998). “New Unionism and the Transformation of the Korean Industrial Relations System.” Economic and Industrial Democracy, 19(2): 347-373. Lee, E. (1996). “Globalization and Employment: Is Anxiety Justified?” International Labour Review, 135(5): 485-497. Lee, J. (1995). “Comparative Advantage in Manufacturing as a Determinant of Industrialization: The Korean Case.” World Development, 23(7): 1195-1214. 298 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Lensink, R, & White, H. (1998). “Does the Revival of International Private Capital Rows Mean the End of Aid?: An Analysis of Developing Countries’ Access to Private Capital” World Development, 26(7): 1221-1234. Lim, Hy-Sop (1997) “The Evolution of Social Classes and Changing Social Attitudes.” In (Eds) David Lindauer, Jong-Gie Kim, Joung-Woo Lee, Hy-Sop Lim, Jae-Young Son, and Ezra Vogel. The Strains o f Economic Growth: Labor Unrest and Social Dissatisfaction in Korea. New England: Harvard University Press: 13-34. Lindauer, David (1997a) “Labor Market Outcomes: An Overview.” In (Eds) David Lindauer, Jong-Gie Kim, Joung-Woo Lee, Hy-Sop Lim, Jae-Young Son, and Ezra Vogel. The Strains o f Economic Growth: Labor Unrest and Social Dissatisfaction in Korea New England: Harvard University Press: 35-54. Lindauer, David (1997b) “Introduction.” In (Eds) David Lindauer, Jong-Gie Kim, Joung-Woo Lee, Hy-Sop Lim, Jae-Young Son, and Ezra Vogel. The Strains o f Economic Growth: Labor Unrest and Social Dissatisfaction in Korea New England: Harvard University Press. Lutz, M, & Singer, HW. (1994). “The Link Between Increased Openness and the Terms Malloy, James (1979) The Politics o f Social Security in Brazil. Pittsburg: University of Pittsburg Press. Mamkoottam, Kuriakose (1994) “Globalization and Emerging Labor Management Relations.” In Labour and Unions in a Period o f Transition. C.S Benkata Ratnam, Gerd Botterweck and Pravin Sinha (Eds.) New Delhi: Friedrich Ebert Stiftung. Mani, Sunil (1995) “Economic Liberalization and the Industrial Sector.” Economic and Political Weekly. May 27: M38-M50. Marquez, Gustavo (1995) Reforming the Labor M arket in a Liberalized Economy(ed). Washington D.C.: Inter-American Development Bank. Marshall, A. (1994). “Economic Consequences of Labour Protection Regimes in Latin America.” International Labour Review, 133(1): 55-73. Martin, A. (1998). “What Does Globalization Have to Do With the Erosion of Welfare States: Sorting Out the Issues.” Harvard University. Program fo r the Study o f Germany and Europe, Working Paper Series #7.5. Mathur, Ajeet (1993) “The Experience of Consultation During Structural Adjustment in India (1990-92).” International Labour Review 132 (3): 331- 345. 299 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Mathur, Kamala and N.R. Sheth (1969J Tripartitism in Labour Policy: The Indian Experience. New Delhi: Shri Ram Centre for Industrial Relations. Maxfield, S. (1998). “Understanding the Political Implications of Financial Internationalization in Emerging Market Countries.” World Development, 26 (7): 1201-1219. McDiarmid, OJ. (1997). Unskilled Labor fo r Development. Baltimore: John Hopkins University Press. Mesa-Lago, C. (1994). Changing Social Security in Latin America. Boulder: Lynne Rienner Publishers. Mesa-Lago, C. (1991). “Latin America and the Caribbean.” In Social Security in Developing Countries. Ahmad, E. et al. (Eds). Oxford: Oxford University Press: 356-394. Mesa-Lago, Carmelo (1978) Social Security in Latin America: Pressure Groups, Stratification, and Inequality. Pittsburgh: University of Pittsburgh Press. Meyers, C. (1958). Labor Problems in the Industrialization o f India. Cambridge: Harvard University Press. Midford, P. (1993). “International Trade and Domestic Politics: Improving on Rogowski’s Model of Political Alignments.” International Organization. Autumn, 47(4): 535-564. Midgley, J. (1984). Social Security, Inequality and the Third World New York: John Wiley and Sons. Mistry, P. (1995). “Financial Sector Reform in India: Hesitant Pursuit of an Incomplete Agenda.” In India: the Future o f Economic Reform. Cassen, Robert and Vijay Joshi (Eds). Delhi: Oxford University Press. Mitchell, O, & Zeldes, S. (May 1996). “Social Security Privatization: A Structure for Analysis.” American Economic Review, 86(2): 363-367. Momoria, CB, & Momoria, S. (1983). Labour W elfare, Social Security and Industrial Peace in India. Allahabad: Kitab Mahal. Montiel, P. (1994). “Capital Mobility in Developing Countries: Some Measurement Issues and Empirical Estimates.” World Bank Economic Review, 8(3) September. 311-350. Morales, R. (1998). “Dependence or Interdependence: Issues and Policy Choices Facing Latin Americans and Latinos ” In Borderless Borders. Bonilla, Frank 300 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Edwin Melendez, Rebecca Morales, and Maria de los Angeles Torres (Eds). Philadelphia: Temple University Press, 1-13. Morgan, R. (1991). “Social Security in South Africa.” In Social Security in Developing Countries. Ahmad, E. et al. (Eds). Oxford: Oxford University Press, 395-465. Morris, R. (1988). Testing the Limits c f Social Welfare. R. Morris (Ed). London: Brandeis University Press. Mosley, P., Harrigan, J. & Toye, J. (1995). A id and Power, World Bank and Policy Based Lending. Vol. 1. New York: Routledge. Nader, R, Grieder, W, Atwood, Shiva, V. et al. (1993). The Case Against Free Trade. San Francisco: North Atlantic Books. Nidugala, Ganesh Kumar (1997) “Capital Liberalization and Currency Crisis: Experience of Mexico and India” Economic and Political Weekly 32 (29) July 19:1803-1811. Oatley, T, & Nabors, R. (Winter 1998). “Redistributive Cooperation: Market Failure, Wealth Transfers, and The Basle Accord.” International Organization, 52(1): 35-54. O’Connor, J, & Olsen, G. (1998,). Power Resources Theory and the Welfare State. Julia O’Connor and G. Olsen (Eds). Toronto: University of Toronto Press. OECD Working Papers. (1997). Integrating Developing Countries in a Globalising World Economy. OECD Working Papers, 5 (59). Osmani, SR. (1991). “Social Security in South Asia.” In in Social Security in Developing Countries. Ahmad, E. et al. (Eds). Oxford: Oxford University Press, 305-347. Palhano, D. (1995) “Labor-Law in Brazil and the United States-Statism and Classical Liberalism Compared.” Columbia Journal o f Transational Law 33 (3): 583-635. Palley, Howard (1992) “Social Policy and the Elderly in South Korea.” Asian Survey XXXH (9) September. 787-801. Papola, T.S. (1994) “Employment Growth and Social Protection of Labor in India.” Indian Journal o f Industrial Relations 30 (2) October 117-141. Papola, T.S. (1992) “Labor Institutions and Economic Development: The Case of Indian Industrialization” in T.S. Papola, and Gerry Rodgers (Eds) Labour 301 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Institutions and Economic Development in India. Geneva: International Institute for Labor Studies: 17-47. Park, Young-Bum (1997) “Privatization and Employment in the Republic of Korea.” In Lessons from P rivatization. (Eds) Van der Hoeven, Ralph and G. Sziraczki. Geneva: International Labor Office. .21- 45. Pasha, MK (1996). “Globalization and Poverty in South Asia.” Millenium: Journal o f International Studies, 25(3): 635-656. Pastor Jr., M. (1998). “Interdependence, Inequality, and Identity: Linking Latinos and Latin Americans.” In Borderless Borders. Bonilla, Frank Edwin Melendez, Rebecca Morales, and Maria de Ios Angeles Torres (Eds). Philadelphia: Temple University Press, 17-51. Patnaik, Prabhat (1994) “International Capital and National Economic Policy: A Critique of India’s Economic Reforms.” Economic and Political Weekly. March 19:683- 689. Paus, E, & Robinson, R. (1997). “The Implications of Increasing Economic Openness for Real Wages in Developing Countries, 1973-1990.” World Development, 25 (4): 537-547. Pedersen, K. (1997). “The Political Economy of Distribution in Developing Countries: A Rent-Seeking Approach.” Public Choice, 91: 351-373. Perkins, D. (1997). “Structural Transformation and the Role of the State.” In The Korean Economy 1945-1995: Performance and Vision fo r the 21st Century. Dong-Se Cha, Kwang Suk Kim, Dwight Perkins (Eds). Korea: Korea Development Institute. Perry, R, & Robertson, J. (1998). “Political Markets, Bond Markets, and the Effects of Uncertainty: A Cross-National Analysis.” International Studies Quarterly, 42:131-160. Pfaller, A, Gough, I, & Therbom, G. (1991). Can the Welfare State Compete? London: Macmillan Press. Pierson, P. (January 1996). “The New Politics of the Welfare State.” World Politics, 48:143-179. Pissarides, C. (1997). “Learning by Trading and the Returns to Human Capital in Developing Countries.” The World Bank Economic Review, 11(1): 17-31. 302 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Platteau, J-P. (1991). “Traditional Systems of Social Security and Hunger Insurance: Past Achievements and Modem Challenges.” In Social Security in Developing Countries. Ahmad, E. et al. (eds). (Oxford: Oxford University Press), 112-163. Prabhu, K. Seeta (1994) “The Budget and ‘Structural Adjustment with a Human Face.” Economic and Political Weekly. April 16-23:1011-1028. Ramaswamy, KV, & Gereffi, G. (January 17, 1998). “India’s Apparel Sector in the Global Economy.” Economic and Political Weekly, 33(3): 122-130. Rao, Govinda (1998) “Accommodating Public Expenditure Policies: the Case of Fast Growing Asian Economies.” World Development 26 (4): 673-694. Ratnam, V. (1997). “Adjustment and Privatization In India.” In Lessons from Privatization. Van der Hoeven, Ralph and G. Sziraczki (Eds). Geneva: International Labour Office. --------------- (1996) “ Tripartism and Structural Changes: The Case of India.” Indian Journal o f Industrial Relations 31 (3) January: 346-230. -------------- (1993) “Impact ofNew Economic Policy on the Role of Trade Unions.” Indian Journal o f Industrial Relations 29 (1) July: 55-77. Ravallion, Martin and K. Subbarao (1992) “Adjustment and Human Development in India.” Journal o f Indian School o f Political Economy. Jan-Mar: 55-79. Reisen, H. (1996). “Managing Volatile Capital Inflows: The Experience of the 1990s.” Asian Development Review, 14(1): 72-96. Remmer, K. (October 1997). “Theoretical Decay and Theoretical Development: The of Institutional Analysis.” World Politics, 50: 34-61. Reuschemeyer, D, Stephens, EH, & Stevens, JD. (1992). Capitalist Development and Democracy. Chicago: University of Chicago Press. Riveros, L. (1992). “Labor Costs and Manufactured Exports in Developing Countries: An Econometric Analysis.” World Development, 20(7): 991- 1008. Rodrik, D. (Summer 1997a). “Sense and Nonsense in the Globalization Debate.” Foreign Policy, 107: 19-37. Rodrik, D. (1997b) Has Globalization Gone Too Far? Washington D.C. institute for International Economics. 303 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Rodrik, D. (1996). Why Do More Open Economies Have Bigger Governments? NBER Working Paper No. 5537. Cambridge, MA: National Bureau of Economic Research. Rogowski, R (December 1987). “Political Cleavages and Changing Exposure to Trade.” American Political Science Review, 81(4): 1121-1137. Rosenau, J. (November 1997).”CompIexities and Contradictions of Globalization.” Current History, 96 (613): 360-363. Roy, T. (1998). “Development or Distortion? ‘Powerlooms’ in India, 1950-1997.” Economic and Political Weekly, 33(16). April 18:897-911. Rudolph, Lloyd and Susanne Rudolph (1987) ln Pursuit ofLakshmi: The Political Economy o f the Indian State. Chicago: University of Chicago Press. Ruggie, JG. (1994). “At Home Abroad, Abroad at Home: International Liberalization and Domestic Stability in the New World Economy. ” Millenium: Journal o f International Studies, 24(3): 507-526. Ruggie, JG. (1982). “International Regimes, Transactions and Change: Embedded Liberalism in Postwar Economic Order.” International Organization, 36(2): 379-415. Rush, Howard (1993) “Employment and Skills in Brazil: The Implications of New Technologies and Organizational Techniques.” International Labour Review 132 (1): 75-93. Safadi, R, & Laird, S. (1996). The Uruguay Round Arguments: Impact on Developing Countries.” World Development, 24(7): 1223-1242. Sandoval, Salvador (1993) Social Change and Labor Unrest in Brazil Since 1945. Oxford: Westview Press. Sanyal, K. (June 19, 1993). “Paradox of Competitiveness and Globalization of Underdevelopment.” Economic and Political Weekly: 1326-1330. Sathe, Dhanmanjiri (1997) “Import Intensity of India’s Exports: Some Fresh Evidence.” Economic and Political Weekly 32 (8): M31-M44. Schlagheck, J. (1977). The Political, Economic, and Labor Climate in Brazil. Philadelphia: Trustees of the University of Pennsylvania. Schmidt, S. (1995). “Social Security in Developing Countries: Basic Tenets and Fields of State Intervention.” International Social Work, 38: 7-26. 304 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Scholte, J. (1997). “Global Capitalism and the State.” International Affairs, 73(3): 427-453. Schwartz, H. (April 1998). “Social Democracy Going Down or Down Under.” Comparative Politics, 253-272. Sengenberger, W. (1993). “Local Development and International Economic Competition.” International Labour Review, 132(3): 313-329. Sharma, Krishna Murti (1976) Social Assistance in India Delhi: The Macmillan Company of India Limited. Sharma, 0 P (1996) “Performance, Policy Issues and Prospects of India’s Exports.” Economic and Political Weekly 31 (41-2) October 12-19: 2817-2822. Sherer, M. (1987). “Welfare States: An Overview of Problems and Prospects.” In M odem Welfare States: A Comparative View o f Trends and Prospects, Friedmann, R. et al. (Eds). New York: New York University Press: 290- 298. Silber, S. (1997). “The External Sector of the Brazilian Economy.” In The Brazilian Economy: Structure and Performance in Recent Decades. Maria Willumsen and Eduardo Giannetti da Fonseca (Eds.). Miami: North-South Center Press. Singer, H. (1970). “Dualism Revisited: A New Approach to the Problems of Dual Societies in Developing Countries.” Journal o f Development Studies, 7:60- 61. Singh, RCP. (1989). Labor Welfare Administration in India. New Delhi: Deep and Deep Publications. Singh, Seema (1993) “New Economic policy in India: Some Implications For Employment and Labour Market” Indian Journal o f Industrial Relations 28 (4) April: 311-325. Sinha, Pravin (1994) “ Unions in a Period of Transition.” In Labour and Unions in a Period o f Transition. C.S Benkata Ratnam, Gerd Botterweck and Pravin Sinha (Eds.) New Delhi: Friedrich Ebert Stiftung. Slemrod, J. (1995). “What Do Cross Country Studies Teach About Government Involvement, Prosperity, and Economic Growth.” Brookings Papers on Economic Activity, 2: 373-431. Social Security Administration. (1983). Social Security Throughout the World. Office of Research, Evaluation and Statistics, Washington D.C. 305 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Sodhi, J.S. (1993) “New Economic Policies and their Impact on Industrial Relations.” Indian Journal o f Industrial Relations 29 (1) July: 31-54. Song, Ho Keun (1999) “Politics of Crisis Management and Social Policy in South Korea.” Paper presented for conference on Globalization and the Future of the Welfare State: Interrregional Comparisons. Watson Center, Brown University. October 22-23. Srinivasan, TN. (1990). “External Sector in Development: China and India, 1950- 89.” American Economic Review, 80(2): 113-117. Stallings, B. (1990). “The Role of Foreign Capital.” la. M anufacturing Miracles. Gary Gerrefi and Donald Wyman (Eds). New Jersey: Princeton University Press. Stephens, J. (1996). “The Scandinavian Welfare States: Achievements, Crisis, and Prospects.” In Welfare States in Transition. Gosta Esping-Andersen (Ed) .(London: Sage Publications), 32-65. Stem, N, & Burgess, R. (1991). “Social Security in Developing Countries: What, Why, Who and How.” In Social Security in Developing Countries. Ahmad, E. et al. (Eds). Oxford: Oxford University Press, 41-80. Strange, S. (January 1997). “Erosion of the State.” Current History, 96(613): 365- 369. Strange, S. (1996). The Retreat o f the State. New York: Cambridge University Press. Tang, K-L. (1996). “The Determinants of Social Security in Developing Countries: A Comparative Analysis." International Social Work, 39: 377-393. Tcha, MoonJoong. (1998) “ Labor Disputes and Direct Foreign Investment: The Experience of Korea in Transition.” Economic Development and Cultural Change 46 (2) January: 305-327. Telles, Edward (1992) “Who Gets Formal Sector Jobs? Determinants of Formal- Informal Participation in Brazilian Metropolitan Areas.” Work and Occupations 19 (2). May: 108-127. Tewari, Meenu (1999) “ Successful Adjustment in Indian Industry, the Case of Ludhiana’s Woolen Knitwear Cluster.” World Development 27 (9) 1651- 1671. 306 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Thompson, H. (1995). “Free Trade and Income Redistribution in Some Developing Countries and NICs.” Open Economies Review, 6:265-280. Tonelson, A. (November 1997). “Globalization: The Great American Non-Debate.” Current History, 96(613): 353-359. Usui, C. (1994). “Welfare State Development in a World System Context: Event History Analysis of First Historical Social Insurance Legislation among 60 Countries, 1880-1960.” In The Comparative Political Economy o f the Welfare State. Janoski, Thomas and A. Hicks (Eds). Cambridge University Press. Van der Hoeven, Ralph and G. Sziraczki (1997) Lessons from Privatization. Geneva: International Labor Office. Venkatachalam, V., & Singh, R. (1982). The Political, Economic, and Labor Climate in India. Philadelphia: University of Pennsylvania. Verdier, D. (Winter 1998). “Domestic Responses to Capital Market Internationalization Under the Gold Standard.” International Organization, 52(1): 1-34. Vogel, Ezra and David Lindauer (1997) ‘ Toward a Social Compact for South Korean Labor.” In (Eds) David Lindauer, Jong-Gie Kim, Joung-Woo Lee, Hy-Sop Lim, Jae-Young Son, and Ezra Vogel. The Strains o f Economic Growth: Labor Unrest and Social Dissatisfaction in Korea New England: Harvard University Press: 93-121. Wahl, A-M. (1994). “Economic Development and Social Security in Mexico, 1845- 1985.” International Journal o f Contemporary Sociology, 1-2: 59-81. Waterbury, John (1993) Exposed to Innumerable Delusions: Public Enterprise and State Power in Egypt, India, Mexico, and Turkey. New York: Cambridge University Press. Weyland, Kurt (1996a) “Social Change and Labor Unrest in Brazil Since 1945.” Latin American Research Review. Summer 31 (3): 189-203. Weyland, K. (1996b). “Obstacles to Social Reform in Brazil. Comparative Politics, 1- 22. Weyland, Kurt (1995) “Social Movements and the State: The Politics of Health Reform in Brazil. World Development 23 (10): 1699-1712. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. White, Gordon and Roger Goodman (1998) “Welfare Orientalism and the Search for an East Asian Welfare Model,” in The East Asian Welfare M odel (Eds) Goodman, Roger and Grodon White and Huck-Ju Kwon. NY: Routledge. Williamson, J. (January 1990). The Progress o f Policy Reform in Latin America. Washington D.C.: Institute for International Economics. Wood, A. (Summer 1995). “How Trade Hurt Unskilled Workers.” Journal o f Economic Perspectives, 9(3): 57-80. Wood, A. (1994). North-South Trade, Employment, and Inequality. New York: Oxford University Press. Wood, A. (1997). “Openness and Wage Inequality in Developing Countries: The Latin American Challenge to East Asian Conventional Wisdom.” World Bank Economic Review, 11(1): 33-57. Wood, A, & Berge, K. (October 1997). “Exporting Manufactures: Human Resources, Natural Resources, and Trade Policy.” The Journal o f Development Studies, 34(1): 35-59. Wood, A, & Ridao-Cano, C. (Decemberl996). Skill, Trade and International Inequality, IDS Working Paper. World Bank Policy Research Project (1994). Averting the Old Age Crisis. Oxford: Oxford University Press. World Bank (1996). Global Economic Prospects and the Developing Countries. World Development Report (1995). Workers m an Integrating World. New York: Oxford University Press. Yeom, Jaeho (1998) “Economic Reform and Govemment-Business Relations in Korea: Towards an Institutional Approach” in Development Strategies in East Asia and Latin America Hosono, Akio and Neantro Saavedra-Rivano (Eds). New York: St Martin’s Press: 139-156. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDICES Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX A The following countries were included in all the graphs in Chapter 2. Less Developed Countries: Argentina, Bangladesh, Bolivia, Botswana, Brazil, Cameroon, Chile, Colombia, Costa Rica, Cyprus, Dominican Republic, Ecuador, Egypt Arab Rep., El Salvador, Fiji, Ghana, Greece, Guatemala, Guyana, Honduras, India, Indonesia, Iran, Israel, Jordan, Kenya, Korea Rep., Kuwait, Lesotho, Liberia, Malawi, Malaysia, Mali, Mauritius, Mexico, Morocco, Nepal, Nicaragua, Pakistan, Panama, Paraguay, Philippines, Singapore, Sri Lanka, Syria, Tanzania, Thailand, Trinidad and Tobago, Tunisia, Turkey, Uruguay, Venezuela, Zimbabwe Developed Countries (14 OECD nations analyzed by Garrett): Austria, Belgium, Canada, Denmark, Finland, France, Germany, Italy, Japan, Netherlands, Norway, Sweden, United Kingdom, and United States 310 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX B DATA SOURCES, CALCULATIONS AND ESTIMATIONS WELF: IMF, Government Finance Statistics and International Finance Statistics. KFLOW: IMF, Balance o f Payments Statistics. These include direct investment abroad, direct investment in country, portfolio investment and other investments (including trade credits, loans, currency and deposits). Early editions (early 70’s) include some counting of non-produced, non-financial assets in estimations of the financial account. TRADE: World Bank, World Development Indicators. UNSKILL: Classification scheme developed by Wood and Mayer (1998). Their export product classifications are based on the Standard International Trade Classification (SITC, Revision 2). The following list shows which International Standard Industrial Classification (ISIC) codes correspond to the SITC codes ( also supplied by Wood). The employment statistics for each manufacturing sector came from UNIDO Database o f Industrial Statistics. The final value of UNSKILL is based on the total numbers employed in low skill manufacturing production/ high skill manufacturing production. Manufactured Exports (NM) Low-sfall manufactures (1) Leather and rubber products (2) Wood and paper products (3) Textiles, clothing, footwear and travel goods (4) Non-metallic mineral products (5) Iron and steel and metal products (6) Furniture and plumbing equipment (7) Ships, bicycles and trains (8) Miscellaneous High-skill manufactures (9) Chemicals (10) Cut diamonds (11) Non-electrical machinery (12) Computers and office equipment (13) Communication equipment (14) Electrical machinery (15) Motor vehicles and aircraft (16) Scientific instruments, watches and cameras SITC2 categories 61-2 63-64 65,83-85 66 (less 667) 67,69 81-82 78 Cess 781-4), 79 (less 792) 89,9 (less 941,971) 5 (less 522.24,522.56,524) 667.29 71-74 75 76 77 781-784,792 87,88 311 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Note: the SITC 5-8 categories allocated to primary rather than manufactured exports are phosphorus pentoxide and phosphoric acids (522.24), alunriiuum hydroxide (522.56), radioactive material (524), pearls and precious stones, except cut diamonds (667 except 66129), and non-ferrous metals (68). Translated to M anufactured Production Low-sfdll manufacturing production ISICl categories (1) Leather and rubber products 323,355 (2) Wood products, except furniture 331,341 Paper and paper products (3) Textiles 321,323,322,324 Leather products Wearing apparel, except foc'wear Footwear, except rubber or plastic (4) Other non-metallic mineral products 369,362,361 Glass and products Pottery, china and earthenware (5) Iron and Steel 371,381 Fabricated metal products (6) Furniture, except metal 332 (7) - (8) Plastic products 356,390 Other manufactured products Higk-skill manufacturing production (9) Industrial chemicals 351,352,354,356 Other chemicals Misc. petroleum & coal products Plastic products (10) - (11)Fabricated metal products 381,382,383,384 Machinery, except electrical Machinery, electric (12) - (13)- SURP: World Bank, World Development Indicators. DEPEN: World Bank, World Development Indicators. GDP: World Bank, World Development Indicators. SOE: World Bank, World Development Indicators. DEBT: World Bank, World Development Indicators. DEMOC: Ted Robert Gurr’s and Keith Jaggar’s Polity III (1994) 312 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. NOTE TO USERS Page(s) not included in the original manuscript and are unavailable from the author or university. The manuscript was microfilmed as received. This reproduction is the best copy available. UMf Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX C The Dependent Variable: Government Social Commitment India Brazil Korea Variable -a A a -b B b c -c -c a = level of welfare b= changes in government welfare expenditures (increasing or decreasing) c= strong labor policies (a) = above LDC average -(a) = below LDC average Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX D DATA SOURCES WELF: IMF, Government Finance Statistics and International Finance Statistics. NETK: World Bank, World Development Indicators Cd-Rom. TRADE: World Bank, World Development Indicators Cd-Rom. FDI: World Bank, World Development Indicators Cd-Rom INTDIFF: IMF, Government Development Finance Cd-Rom GDP: World Bank, World Development Indicators Cd-Rom. DEMOC: Ted Robert Gurr’s and Keith Jaggar’s Polity III (1994) RES V: IMF: International Finance Statistics Cd-Rom IMF: World Bank, World Development Indicators Cd-Ro m Less Developed Countries: Argentina, Bangladesh, Bolivia, Brazil, Cameroon, Chile, Colombia, Costa Rica, Dominican Republic, Ecuador, Egypt Arab Rep., El Salvador, Fiji, Ghana, Guatemala, Guyana, Honduras, India, Indonesia, Jordan, Kenya, Lesotho, Liberia, Malawi, Malaysia, Mali, Mauritius, Mexico, Morocco, Nepal, Nicaragua, Pakistan, Panama, Paraguay, Philippines, Singapore, Sri Lanka, Tanzania, Thailand, Trinidad and Tobago, Tunisia, Turkey, Uruguay, Venezuela, Zimbabwe 315 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX E COUNTRIES INCLUDED Total Contributions to Social Security [Chart 1]: Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Cyprus, Dominican Republic, Egypt, Ghana, Honduras, Indonesia, Israel, Korea, Malaysia, Mexico, Nicaragua, Panama, Paraguay, Thailand, Trinidad and Tobago, Uruguay, Venezuela. Government Contributions to Social Security and Welfare [Chart 2]: Argentina, Bangladesh, Bolivia, Brazil, Cameroon, Chile, Colombia, Costa Rica, Dominican Republic, Ecuador, Egypt Arab Rep., El Salvador, Fiji, Ghana, Guatemala, Guyana, Honduras, India, Indonesia, Jordan, Kenya, Lesotho, Liberia, Malawi, Malaysia, Mali, Mauritius, Mexico, Morocco, Nepal, Nicaragua, Pakistan, Panama, Paraguay, Philippines, Singapore, Sri Lanka, Tanzania, Thailand, Trinidad and Tobago, Tunisia, Turkey, Uruguay, Venezuela, Zimbabwe. Employer Contributions to Social Security [Chart 3]: Argentina, Brazil, Colombia, Costa Rica, Cyprus, Egypt, Honduras, Israel, Korea Rep., Malaysia, Thailand, Trinidad and Tobago, Uruguay, Venezuela. Employee Contributions to Social Security [Chart 4]: Argentina, Brazil, Colombia, Costa Rica, Cyprus, Egypt, Honduras, Israel, Korea Rep., Thailand, Uruguay, and Venezuela. 316 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Analysis of the role of government in Taiwan's industrialization and economic development
PDF
By sea: The port nexus in the global commodity network (the case of the West Coast ports)
PDF
Economic transition, micro- and small -scale enterprise in Zimbabwe, and the significance of participatory development to institutional economics
PDF
Dependence on foreign labor, quality of education and unemployment in the GCC countries: In search of solutions
PDF
A comparative examination of the impact of business -government relations on labor market reform in Egypt and Mexico, 1975--1995
PDF
The Interventionist State Revisited: The Political Economy Of State-Business Relations In India
PDF
Before and after reform: The effects of structural adjustment on women's employment in Latin America
PDF
Reassessing the relationship between globalization and welfare: Welfare spending and international competitiveness in less-developed countries
PDF
Dialectical feminism and contradictory economic systems: Campesina daily life from Somoza to Aleman
PDF
The Role Of The State In The Development Of The Indonesian Textile And Garment Industries
PDF
The political economy of the state and development strategies: A case study of Turkey, 1930-1980
PDF
An empirical analysis of foreign aid and the provision of international public goods
PDF
Essays in environmental policy
PDF
Compatibility standards, market power, and antitrust policy: Proprietary interfaces and their alternatives
PDF
A multidimensional analysis of regional integration and cooperation: The case of the Economic Community of West African States
PDF
An analysis of the developmental state: The case of the Vietnamese textile and garment industries
PDF
A critique of Rawls's structure of justification
PDF
Key factors in appraising development project in Egypt
PDF
Carrots or sticks: The Taiwanese government's contrasting policy approaches to banking sector liberalization
PDF
The drivers of economic change: Technology, globalization, and demography
Asset Metadata
Creator
Rudra, Nita
(author)
Core Title
Globalization and the decline of the welfare state in less developed countries
School
Graduate School
Degree
Doctor of Philosophy
Degree Program
Political Economy and Public Policy
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
economics, labor,OAI-PMH Harvest,political science, general
Language
English
Contributor
Digitized by ProQuest
(provenance)
Advisor
Elliott, John E. (
committee chair
), [illegible] (
committee member
), Kalaba, Robert E. (
committee member
)
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c16-134073
Unique identifier
UC11334674
Identifier
3041514.pdf (filename),usctheses-c16-134073 (legacy record id)
Legacy Identifier
3041514-0.pdf
Dmrecord
134073
Document Type
Dissertation
Rights
Rudra, Nita
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA
Tags
economics, labor
political science, general